Economic Journal

Frey, Bruno, S.;Benz,, Matthias

doi: 10.1111/j.1468-0297.2005.01041.xpmid: N/A

Abstract Corporate governance is importantly based on agency theory and relies on extrinsic incentives to align the interests of managers, employees and shareholders. This article argues that in view of recent corporate scandals, private governance can learn from public governance: (1) Goal‐oriented intrinsic motivation of agents should be supported by fixed incomes and an extensive selection process of employees; (2) Extrinsic, but non‐monetary incentives (e.g. conferring orders and titles) can be used; (3) The power of actors should be restricted by a clear division of power, appropriate rules of succession and institutionalised competition for positions in firms. The way scholars think of corporate governance today has been importantly shaped by agency theory (Shleifer and Vishny, 1997; Daily et al., 2003). It takes the firm to be a web of voluntary contracts. The major task is to find the most efficient way to align the interests of the managers as the agents to the interests of the stockholders as the principals (Jensen and Meckling, 1976). The market for corporate control is taken to work well, so that this system is self‐regulating. There is no reason to assume ‘contractual failure’; the collective action problems faced by (dispersed) shareholders are overcome by various processes, the most important being unfriendly stock market takeovers. Agency Theory has sparked a huge literature, which has been very ably surveyed by e.g. Becht et al. (2002), Prendergast (1999) or Eisenhardt (1989). Agency theory not only dominates the academic discipline but has also been accompanied by applications of its major message in business practice. In particular, agency theory's emphasis on the need for managers’ interests to be aligned with those of stockholders has been accompanied by, if not been responsible for, a widespread effort to introduce performance incentive plans, in particular pay‐for‐performance. The idea has even spread to areas outside the market and the capitalist economy. The view that society should be run as if it were a firm has e.g. led to a strong movement called ‘New Public Management’, urging non‐profit firms and public administrations to adopt pay‐for‐performance programmes; see Pollitt and Bouckaert (2000); Weisbrod (1998). But there has recently been a sudden sobering up. The stock market has crashed and the corporate sector has been plagued by huge scandals relating to excessive manager compensation and fraudulent bookkeeping. Most importantly for agency theory, performance pay by linking salaries to stock options has led to an explosion of compensation due to the stock market boom and the trend has in many cases simply continued, even though economic conditions have changed. Management compensation has often increased still more, even though share prices have plummeted. This suggests that, in actual fact, the compensation of managers has little to do with performance. Rather, the reason for the steady increase in compensation is now widely seen in the fact that managers are able to exert considerable control over how much money they get (Bebchuk and Fried, 2003). Some managers even resorted to unlawfully misrepresenting their firms’ accounts in order to raise their private incomes. A particularly troubling aspect is that, in many instances, extended pay‐for‐performance plans have created the very incentives to commit fraud, by making it attractive to produce short‐term increases in share prices (Efendi et al., 2004; Erickson et al., 2003; Johnson et al., 2003). After the event, it can be said that agency theory has obviously neglected the possibility of managers distorting their own standards of performance: ‘[…] much of agency theory […] unrealistically assumes that earnings and stock prices cannot be manipulated. That is a major weakness of the theory […]’ (Becht et al., 2002, p. 47).1 These weaknesses and failures of actual corporate governance practice, as well as the incompleteness of Agency Theory in predicting them, suggest that it might be useful to approach the issue from a new perspective. This contribution argues that fresh insights for corporate governance can be gained from the way democratic government and public administration are organised. Corporate governance can learn from public governance, in the sense that institutions devised to control and regulate the behaviour of actors in the public sphere can give new insights into how corporate governance practice can be improved. This does not mean that public governance has produced ideal results – far from it. In public choice theory or modern political economy, many inefficiencies of democratic politics and public administration have been documented and analysed, like distortions due to rent seeking activities.2 These shortcomings, however, do not exclude some institutions of public governance being useful for corporate governance. While the reverse direction of learning from private to public governance has been extensively discussed in the past, leading, for example, to the introduction of New Public Management in at least some parts of public administration, useful insights may also be gained by applying public sector ideas to private governance. The analysis offered here is rooted in the tradition of constitutional political economy, the economic analysis of political institutions; for surveys, see e.g. Frey (1983); Mueller (1996); Cooter (2000). But is also greatly influenced by developments in psychological economics or behavioural economics; for surveys, see e.g. Rabin (1998); Frey and Stutzer (2002). The resulting alternative approach does not necessarily contradict classical agency theory, but rather aims at introducing new, forgotten or neglected aspects. As a consequence, however, the ideas for organisational design differ substantially from much of what is suggested by accepted theory. We propose that corporate governance can learn from public governance in three areas: Section 1 considers the possibilities of using goal‐oriented intrinsic motivation in organisations, Section 2 discusses extrinsic, but non‐monetary, incentives imposed from outside (e.g. titles and orders) and Section 3 looks at the restriction on power of actors. In each of these Sections, specific public governance institutions are discussed and the possibilities of introducing these institutions into corporate governance outlined. We conclude in Section 4 by discussing limitations of our arguments, and by summing up the main messages. 1. Supporting Goal‐Oriented Intrinsic Motivation Traditional Agency Theory builds primarily, or exclusively, on extrinsic motivation.3 In contrast, intrinsic motivation is attributed a substantial role in public governance. A substantial number of institutions have been designed in the public sector, serving to shape agents’ intrinsic motivation in order to produce the desired outcomes. The fact that such institutions exist is an important, but often neglected, aspect of politics and public administration. Four institutions serve this purpose, the first one relating to politicians, the second one relating to both politicians and public officials, and the two remaining ones relating to public administrators. 1.1. Popular Participation Rights Democracies are defined by giving citizens clearly determined participation rights. In representative democracies, the citizens can determine the parties and often the persons to represent them in parliament. In direct democracies, citizens can also determine substantive issues via initiatives and referenda. Citizens’ participation rights have important consequences for the behaviour of the politicians. Above all, participation rights ensure accountability but they also affect identification. The more extensive the rights are to participate politically, the stronger the extent of interaction is between the citizens and the (professional) politicians. The constant tendency for the leaders to establish a ‘classe politique’ is reduced. A similar positive effect on identification can be expected from co‐determination in firms, not least because the managers are induced to interact more intensively with their employees than they otherwise would, e.g., van den Berg (2004). Participation rights also affect feedback. The reaction of the voters to the actions taken by the politicians constitutes an essential part of the democratic political process. In referendum democracies, the politicians get direct feedback on how the voters evaluate specific policies. This feedback is of an informative, rather than a controlling nature, and is therefore likely to raise the intrinsic motivation of politicians to pursue policies in the interests of the citizens (Frey, 1997a). 1.2. Fixed Position and Income The career path and the income corresponding to the various positions in public administration are governed by formal rules. Advancement is regular and is largely determined by seniority. Promotion does not depend on any specific output performance. The job is guaranteed for fixed terms, and often for life; members of the public administration cannot be simply dismissed by their superiors. Public employees are therefore able to make suggestions for improvement, and to criticise the course of events, even if their superiors do not necessarily agree. Agency theory has identified such bureaucratic rules as an optimal response to dysfunctional behaviour, due to evaluation procedures occurring in multi‐tasking situations (Holmström and Milgrom, 1991, 1994). Agency theory responds by suggesting ‘subjective performance evaluation’ (Prendergast, 1999, p. 29–33; Gibbons, 1998, p. 120–3). But such subjective evaluation in firms shifts the discretion over employees’ pay to the superior. This dependence of employees on the goodwill of their superiors weakens or even totally suppresses any incentives they might have to monitor and criticise the behaviour of their superiors (Prendergast, 1993). Indeed, whistle‐blowing has proved to be rarely used under a regime in which the superiors are able to determine the wages of the persons working for them. But such monitoring of the superiors by their inferiors plays an important function, because the inferiors are normally well informed about the tricks played and wrongdoings committed by the management. This problem is to some extent mitigated in the public administration. While there is a well‐defined bureaucratic hierarchy, the superiors have to follow well‐defined rules and, in principle, have no discretion concerning the pay of their inferiors. This advantage is indeed seen by e.g. Prendergast (1999, p. 37), when he states with respect to bureaucracy: ‘[…] rules are used to allocate resources rather than allowing individuals’ discretion over resource allocation’. An administrative career according to seniority and with fixed compensation allows its members to concentrate on work content. Fixed incomes have the important advantage of serving as ‘redistribution constraints’ (Hansmann, 1996). They free employees from fighting over earnings and so contribute to the organisations’ common good. In contrast, a system characterised by pay‐for‐performance strongly induces employees to devote time and effort to influencing their variable income. Employees rationally engage in rent seeking activities in order to manipulate the performance standards and therewith their income. While they can seek higher income by increased effort, it is often easier and less demanding to influence the measuring rod, even to the extent of distorting and falsifying the figures. This has turned out to be of particular importance with respect to manager compensation. Several empirical studies have shown that there is a strong relationship between the extent of variable, stock‐based executive compensation and the incidence of corporate fraud (Osterloh and Frey, 2005a). Johnson et al. (2003), for example, show that managers involved in accounting frauds between 1992–2001 had a 69% higher pay‐for‐performance sensitivity than managers not involved in frauds, a result of their much higher stock and stock option compensation, approximately US$ 4.4 million more at the median (for similar findings see Efendi et al., 2004; Erickson et al., 2003). Agency theory has, to a large extent, failed to see this rational reaction of managers subjected to pay‐for‐performance, an observation that is now largely acknowledged (Becht et al., 2002, p. 47; Jensen et al., 2004, p. 98). The administrative approach to the allocation of career advancement and pay avoids fundamental problems connected with pay‐for‐performance (for on overview, see e.g. Osterloh and Frey (2000)), some of them well known in agency theory: performance is rarely easily defined, but is often subject to interpretation and influence;4 in many cases, only some aspects of performance are measurable, leading to the multi‐tasking problem; employees must constantly be monitored in order to be able to pay them, which can lead to a perception of being controlled and tends to crowd‐out any existing intrinsic motivation; and the work content as such is of less interest to employees, but is regarded as instrumental to pay. Fixed incomes as used in public governance can avoid these fundamental problems of performance pay. In addition, the public sector approach to compensation seeks to guide agents’ behaviour by enabling and shaping intrinsic motivation to work in the organisation's interest. 1.3. Extensive Selection Process of Agents To become a member of the classical public bureaucracy, as it used to exist in Germany and has been described by Weber (1978), is a formalised and arduous process. It takes many years. Before a person can become a full member (‘ein Beamter’), he or she has to pass through many, clearly regulated stages, similar to being accepted into a religious order. The prospective members must have passed exams specifically designed as initiations into this select profession. In many cases, these exams have little or nothing to do with the task to be later performed as a member of the public administration. The famous Chinese bureaucracy required people to be able to prepare poems (Tullock, 1964); in Germany a public official had to have, and to some extent still has to have, an education in law, while France puts more emphasis on formal (mathematical) education in the ‘Grandes Ecoles’. The long and arduous process promotes a specific self‐selection for anyone wanting to become a member of public bureaucracy. Because of the ‘deferred compensation’, persons with very highly developed intrinsic motivation to work in the public sector are attracted, while short‐term materialists have no incentive to engage themselves. The difficult and long drawn out process strongly socialises the persons according to the specific culture of the public sector. The goals of the organisation are at least partly internalised. The members develop a distinct and often marked sense of loyalty, and intrinsic motivation is crowded in (Frey, 1997b). The long formal selection process, moreover, provides the successful applicants with confidence in their own abilities. This sense of competence is not so much based on outcome related performance (it already exists before the applicants take up their position) but on having got through the process. This feeling can sometimes also result in arrogance, a trait often attributed, for instance, to top‐level French public officials coming from the ENA (‘Ecole Nacionale d'Administration’) or ‘Enarques’. The democratic process also leads to a particular selection of traits in politicians. Voters have a strong tendency to evaluate contenders for political office in terms of their presumed characters (Brennan and Hamlin, 2000; Cooter, 2003). Decentralised democracies allow voters to select the characters of the politicians they want better, because there are a large number of elections in which the contenders have to present themselves to the citizens. The same holds for elections of specific persons, rather than closed lists prepared by the parties. It has been argued that firms should pay more attention to selecting the appropriate characters for the tasks at hand, rather than mainly or exclusively relying on external incentive systems (Cooter and Eisenberg, 2000–2001). 1.4. Autonomy Within Rules Bureaucratic rules provide directions but within them public employees enjoy a clearly defined measure of autonomy. They have the opportunity to evaluate and make decisions based on what they see to be correct and appropriate for the long‐term goals of the organisation. As long as public employees adhere to the rules, they are ideally protected from intervention on the part of their superiors. Such autonomy contributes to intrinsic motivation, as an innate need in individuals (Deci and Ryan, 1985; Ryan and Deci, 2000). Employees in capitalist firms do not enjoy such autonomy based on well‐defined rules. As long as superiors do not violate the law, they can, to a large extent, instruct an employee to do anything they like. The only option open to a dissenting employee is to leave the firm, an action that often involves substantial costs. The recent bookkeeping scandals have made clear that, in practice, CEOs are sometimes even able to instruct their employees to commit illegal actions. Employees in private firms thus have to accept a considerable ‘zone of indifference’. 1.5. Evaluation Public governance uses a completely different system to align agents’ behaviour with principals’ goals. It relies on fixed compensation, self‐selection, socialisation and rules to bring about an internalised intrinsic behaviour of agents consistent with the goals of the public organisation. Traditional agency theory has predominantly focused on how this alignment can be reached, by setting the right extrinsic (most importantly monetary) incentives. Bureaucratic rules may appear inefficient ex post, but are not necessarily ex ante. This point has been noted in accepted agency theory, but is only related to an effort to avoid inefficient rent seeking activities (Prendergast, 1999, p. 38; Milgrom and Roberts, 1988; Tirole, 1992). The approach on which public governance is based places more emphasis on guiding agents’ behaviour by intrinsic motivation, which has the considerable advantage that the problems arising from having to measure performance standards appropriately and fully (among which multi‐tasking has received most attention in agency theory) are avoided. There are several obvious drawbacks connected with the public sector approach of relying on intrinsic motivation. Most importantly, fixed compensations and careers fail to provide strong extrinsic incentives for agents to exert effort, which often leads to the impression that the public sector is lacking in effectiveness and innovativeness. Moreover, the intrinsic motivation of politicians and public administrators does not necessarily serve the interests of the citizens, e.g. if behaviour is predominantly directed towards rent‐seeking. It is important, however, to see that the public sector approach points out a useful role that intrinsic motivation can, in principle, play in organisations, an aspect largely neglected in agency theory. In addition, the intrinsic motivation of politicians and public administrators will further the interests of citizens or rather harm them, depending largely on the political system and political institutions under which they act. 2. Extrinsic Incentives Through Outside Recognition Public governance also relies on externally mediated incentives to align agents’ behaviour with the goals of the organisation. While agency theory focuses mainly on monetary incentives, because they are the most fungible and therefore seem to be the most efficient, public governance uses various kinds of awards as extrinsic motivations; see, more generally, Frey (2005). Two types of specific institutions are used in public governance as extrinsic motivators: 2.1. Titles People like titles to indicate their place in a hierarchy clearly (Frank, 1985, pp. 99–102; Gould, 2002). They have added value when they can be transferred to life outside of the organisation they work in, especially in the context of their social life with family and friends, but also with complete strangers. In former times, this condition was met by conferring such titles as e.g. ‘Geheimrat’ in Germany and Austria. Titles are transferable if persons outside the organisation have a sense of the distinction conferred. They do not necessarily have to know exactly what they mean, but they have to be impressed. Therefore awards must be conferred in a restricted way. In recent years, most leading private corporations have given up conferring transferable titles. Instead, they use ‘functional’ titles, which are of little or no use outside of the organisation. Corporations have also shifted to conferring titles in a rather unrestricted way. If, for example, a considerable percentage of managers are ‘vice‐presidents’, the title is no longer worth as much (it must then be amended by adding ‘senior’, ‘executive’ etc.). A similar inflation can be observed with the title ‘Chief Executive Officer’, which is nowadays used by many firms for several people within management heading a division. 2.2. Orders Public governance confers orders, medals and other decorations on people extensively as an incentive device. They are often awarded at the end of bureaucratic careers as a tribute to life‐long devotion to one's tasks, and not for specific performance.5 Orders are not contractible when entering a public career but they are given on the basis of clear rules. They can be transferred to one's social life, as they are virtually worn outside, on the dinner jacket and even on normal suits (e.g. the French ‘Légion d'Honneur’). Corporations use a similar incentive when they appoint ‘The employee of the month’. But these awards are designed to relate directly to specific performance. They are therefore likely to be understood as solely instrumental. Moreover, they are difficult to transfer outside of the organisations (and are often considered to be rather ridiculous). They have the added disadvantage that they tend to be handed out in a rather inflationary way, so that those not receiving them can be demotivated. 2.3. Evaluation The extrinsic incentives by awards conferred by public governance represent an overall evaluation, and are of an ex post non‐contractual nature. Care is taken not to make them appear directly instrumental. The incentive structure applied by public governance skilfully combines outside recognition and status with crowding‐in goal oriented intrinsic motivation. It might be argued that titles and orders have lost importance. This may be true but revealed preference suggests that they are still highly appreciated today. Thus, many leading economists were, and still are, delighted to have titles bestowed on them, Lord Keynes and Sir Tony Atkinson being just two such examples. There are major differences between conferring awards and giving money as compensation for performance. First, conferring awards is intended to honour long‐term, even life‐long performance, while, almost of necessity, performance pay relates to short‐term achievements. Second, awards are mainly given in a process‐oriented way, serving as a reliable feedback for performance, which is likely to raise intrinsic motivation (Deci and Ryan, 1985; Frey, 1997b). In contrast, performance pay relies on outcomes that are subject to many other systematic and random effects outside the influence of the recipient. The feedback is thus less reliable, and may not contribute to crowding‐in intrinsic motivation. Third, the value of awards is less easy and less straightforward to compare than monetary income. Awards are therefore less prone to being devalued by processes of social comparison and hedonic adaptation, which have been found to decrease the utility from monetary income considerably (Frank, 1985, 2000; Easterlin, 2001; Stutzer, 2004). The discussion suggests that awards have quite different motivational implications than monetary compensation, in particular pay‐for‐performance. In many cases, the incentives produced by conferring awards, especially the increase in intrinsic motivation induced, is of great value to an organisation and should therefore be considered a part of management. 3. Restriction of Power Public governance aims at producing intrinsic motivation in agents but at the same time it has devised many institutions that serve to discipline and control the behaviour of public officials. ‘Disciplining agents’ is a core task of democratic government. Corporate governance shares the same goal with respect to ‘disciplining management’ (Becht et al., 2002, pp. 21–2). In the case of public governance, not only the government politicians, but also the members of parliament and administration must be constrained from abusing their power. The major reason for the accumulation of uncontrolled discretion in both areas of governance is the strongly asymmetric state of information of the persons occupying leading positions. This accumulation of power threatens the interests of citizens, as well as shareholders, and leads to authoritarian, or even dictatorial, forms of governance. For centuries, democracies have developed various efficient institutions to restrict the accumulation of power. Three institutions are of particular importance and provide new insights as to how corporate governance can be improved. 3.1. Division of Power Democratic states distribute the right to act among the three classical decision‐making bodies: the executive, legislative and judicial branches. The constitutions actively promote the principle of checks and balances. This does not prevent one branch from dominating for a period of time, but it does ensure that the other branches can reassert themselves in due time. This principle is clearly visible in, among others, the American constitution. A close analogy has often been drawn between private corporations and the public sector: the CEO corresponds to the head of government; the shareholder meeting to parliament. A more appropriate analogy would, however, be to see the shareholder meeting as a town council meeting, in which the citizens themselves convene and no representatives are needed. The company board may be seen to correspond to the members of the cabinet. In corporate governance, the principle of division of power is applied much less strictly than in public governance. In many countries, for example in the US, France and Switzerland, it is common practice that the CEO of the firm is, at the same time, the chairman of the board and therewith of the shareholder meeting. This blurs the division between the top agents (CEOs) and the principals (shareholders). In the same vein, until recently, not much attention has been paid to a clear division of control over core aspects of the firm, like the independence of compensation and audit committees. Division of power is an area where corporate governance can learn from public governance and, indeed, it already has to some extent. In public governance, there is an independent institution controlling the executives, the ‘court of accounts’. In many countries, their competencies are quite restricted. These courts of accounts derive their independence from being part of the judicial branch. In Switzerland, by contrast, an interesting kind of court of accounts exists, which derives its independence from being directly elected by the citizens. Empirical evidence shows that such directly elected courts of accounts have a considerable impact on the quality of government (Schelker and Eichenberger, 2003, 2004). It seems that they successfully restrain local governments from abusing their power and induce them to act more in the citizens’ interests. The corporate sector has often not clearly separated the executive and external auditing functions, at least until recently. In many cases, CEOs determined the auditing firms that were supposed to control them. At the same time, the auditing firms were, and still are, paid for advising jobs for the CEO and general management (The Economist, 2004). As a result of the huge scandals produced by this system, there are now government‐imposed rules in many countries, more clearly separating the executive from the auditing branch, like the Sarbanes‐Oxley Act in the US (Securities and Exchange Commission, 2003). This is obviously an area where corporate governance has incorporated insights from public governance, but only after having incurred huge costs. The public governance perspective suggests, however, that learning from the public sector could go a step further. The independence of the auditing process could be further improved by relying on the democratic mechanism of direct elections for (a)the members of the audit committee and (b)the auditing firm by the shareholders. This reflects the basic democratic idea that the independence of a committee cannot be judged by abstract formal criteria alone,6 but ultimately has to be based on the fact that it has been freely chosen by the people who have an interest in it being independent – in this case the shareholders. The reasoning can also be applied to compensation committees and the choice of the auditing firm; we discuss the idea of competitive elections in more detail in Section 3.3. Evidence from the public sector shows that the direct election of independent bodies leads them to take the citizens’ interests into account better than when they are simply appointed (and are more likely to be ‘captured’ by those they are supposed to control); for the case of public regulators, see Besley and Coate (2003), and for courts of accounts, see Schelker and Eichenberger (2003, 2004). The most important area where corporate governance violates the principle of division of power is CEO duality, i.e. when the CEO of the firm is at the same time the chairman of the board. From a public governance perspective, this seriously blurs the distinction between the management and the board who is supposed to control it. In contrast to this view, however, the existing empirical evidence shows that CEO duality does not, in general, lead to disastrous consequences. While researchers have found a weakly positive relationship between CEO duality and the incidence of corporate fraud (Erickson et al., 2003; Uzun et al., 2004; Beasley, 1996; Dechow et al., 1996), a large number of empirical studies document that firm performance is essentially unaffected by the combination of the chairman and CEO positions (Dalton et al., 1998). According to this evidence, one might be led to conclude that the public governance approach overstates the importance of division of power in firms. However, the existing empirical literature simply assumes independence of the CEO and chairman positions if they are held by different persons (Dalton et al., 1998, pp. 271–2). This empirical strategy is likely to seriously underestimate the importance of division of power for firms, because a large majority of the presumably ‘independent’ chairmen are actually former CEOs of the same firm, as well as former or current executives (The Corporate Library, 2004). Such arrangements can hardly be considered a true division of power, given, among other factors, the important role that outgoing CEOs play in the determination of their successors (Shen and Cannella, 2002) and the fact that current executives are supposed to monitor their own bosses. 3.2 Succession In Top Positions Democratic constitutions constrain their agents not only by division of power but also by extensive rules of law that regulate the succession and the rotation in leading positions. Three rules are of particular importance: (a) Restricted terms of office. The members of parliament and directly elected presidents are (normally) elected for four years. At the end of this period, their term in office ends automatically; no further decision is needed. (b) Re‐election restrictions. Many constitutions know the provision that a president may only be re‐elected for one additional term. This is a very strong constraint; it can safely be assumed that many, if not most, presidents would have been re‐elected for more terms. (c) Rotation of positions. Some parties (the German Green party is an example) instituted an automatic change in positions between those inside and those outside parliament and government. The Japanese MITI, a regulatory authority, automatically rotates its leading members in order to make corruption more difficult. The basic idea behind these rules is that they are able to restrict the power of public agents effectively. Moreover, they also open positions to newcomers and, therefore, to fresh ideas. Of the three rules, the one relating to restricted terms in office is the most commonly used in public governance; it is a part of essentially all existing democratic constitutions. But also the requirement of re‐election restrictions is common, most notably in the form of the two‐term limit for the US President. Corporate governance also knows either self‐imposed or government‐imposed rules but they are much less far reaching than those used in public governance, mainly because the market is supposed to control firms. In principle, the terms of office of agents in private corporations are limited but in practice, this is just a formal provision of no real consequence.7 Formal term limits can therefore play a useful role in corporate governance. Their main advantage is that they entail an automatic end of office, where no further discussion and decision is needed, and that they bring about a binding re‐election constraint. Term limits can be envisaged for board members but also for the top executive function of the CEO, e.g. in the form of a two or four‐year term in office. Naturally, such term limits, in particualar for CEOs, would lead to certain advantages as well as disadvantages. One the one hand, four‐year term limits would probably lead CEOs to manipulate company fundamentals in such a way that the firm can be presented in a favourable light at re‐election time. In the political realm, politicians have been found to produce ‘political business cycles’ in a similar way; for an overview, see e.g. Frey and Benz (2003). On the other hand, the current pay‐for‐performance systems arguably give CEOs incentives to act in an even more short‐sighted way, as the recent corporate scandals have made clear (see Section 1). Seen from this perspective, well‐defined terms in office of four years have several advantages. First, they reinstall an incentive to develop a long‐term view on business, as CEOs are basically granted a four‐year period to achieve their goals. If CEOs perform well, they can be confident of being re‐elected for a second or subsequent term, based on a long‐term assessment of their performance. Second, the increased job security can lead top managers to invest more in firm‐specific human capital, which cannot be sold to other firms and thus benefits shareholders (Harris, 1990). And third, term‐limits create strong incentives for the persons electing a CEO to be very careful when choosing a top manager. Thus, term limits do not need to be an alien element in the corporate world. 3.3. Competitive Elections Probably the most important area where corporate governance can learn from public governance is from the latter's strong emphasis on institutionalised competition. Democratic governance can be understood to be the competition by parties for votes (Schumpeter, 1942; Downs, 1957). This competition is closely regulated, but it is fundamentally an open competition. There are three main features: – Voting rights. Only citizens may participate, and each citizen has one vote. Elections are individually oriented, as the voters can determine which persons will sit in parliament. In some cases (especially at the local and provincial level), the voters are able to also choose the persons in the executive branch. – Competitive process. Elections must be open and the citizens must have a choice between several different options, i.e. parties and persons. – Voting rules. Various mechanisms for aggregating votes are used, the best known being ‘first past the post’, leading to strong majorities, but tending to exclude minorities (the system used in the US and the UK), and the proportional system (used in most European countries). The latter sometimes guarantees seats for minorities, or excludes parties with less than a certain percentage of votes (e.g. 5% in Germany). The voting and representation processes used in the public sector and in stock companies share several similarities. In both spheres, there is a collective action problem related to dispersed ‘ownership’. Corporations use elections by shareholders to determine the members of the board, and the board then elects the top management and the external auditing firm. But there is a very big difference in the election process, as we know it, distinguishing the corporate sector from the public sphere: in most corporations, there is generally no choice between various alternatives. As a matter of course, the shareholders are offered one person to be elected for one position on the board, and only one external auditing firm can be chosen, and the CEO cannot be chosen at all. We suggest that corporate governance can learn from public governance with respect to the following three aspects: Voting rights. In principle, each share has one vote. However, this principle is often violated by privileged shares or by non‐voting shares. Such devices are often used to prevent unfriendly takeovers (Seligman, 1986; Grossman and Hart, 1988). Their abolition would strengthen corporate control and secure truly ‘democratic’ shareholder representation. Voting rights may, in principle. also be given to non‐shareholder groups, giving rise to a considerable theoretical literature on multi‐constituency boards (Becht et al., 2002, pp. 48–57). The employees are one group that may be represented. A regime of co‐determination can be seen as a formal recognition of ‘corporate citizenship’ or, more broadly, of ‘organisational citizenship’. The literature has recognised that such employee participation is likely to be efficient in firms with considerable firm specific human capital (Furubotn, 1988; Roberts and Van den Steen, 2000; Osterloh and Frey, 2005b). Moreover, co‐determination seems, in general, not to damage firm performance (Addison et al., 2004). To the extent that employees’ firm specific human capital becomes more important in the ‘knowledge economy’, corporations can be expected to develop an increasing interest in the public sector experience with broad representation practices, and also an increasing interest in the experience many European countries have made with co‐determination. Competitive process. Democracy is not well developed within corporations. The essential element of competition, namely that the voters can choose between relevant alternatives, hardly exists. For instance, for a truly democratic process, the persons with voting rights in the firm must have the option to choose between various persons willing to serve as directors. Similarly, they must be able to choose between several competing firms offering external auditing. In both cases, the competitors must be willing to clearly state their interests and programme, and must be able to convince the corporate voters that they are capable of fulfilling the tasks required. It is difficult to see why such a competitive process exists in the political sector but is often assumed to be impossible within corporations. Paradoxically, the way in which capitalist corporations today select their most important representatives brings to mind former communist regimes: there is one option to choose from and it gets chosen by a huge majority. Corporate governance can learn from public governance by rediscovering the importance and the power of institutionalised competition. Competitive elections seem, at the very least, obvious for positions on the board. Board members are the representatives of shareholders and it is hardly conceivable why shareholders should not have the possibility to exercise their right of free choice; see also Bebchuk (2003, 2004). It is a simple but powerful aspect of public governance that good representation can only be secured if voters have the opportunity to choose their representatives freely. This insight, however, seems to have become completely forgotten in corporate governance. Competitive elections may not only apply to board members but can be further extended to core areas of the firm. For example, a strengthening of corporate governance can be expected if shareholders are given the right to determine the board members that specifically sit on the auditing and the compensation committees, and to choose between different auditing firms. Such elections would greatly improve the independence of the respective actors. It vests them with a unique, institutionally based legitimacy to take an independent stance; at the same time, it secures their accountability to shareholders in important corporate matters. To see the potential of competitive elections for corporations, one may even go a step further. Instead of the board members being faced with a choice of top managers chosen by the former CEO, and possibly by a small group of directors aided by headhunters, the selection of a new CEO could take the form of an open competition. Even more extreme, the whole management group could be open to competition from individuals or firms prepared to fill certain positions like the CEO. The electoral competition then serves to select the most efficient group (it may be the former managers), i.e. the group the corporate voters believe to be the most capable relative to the compensation demanded. In difficult situations, the price asked may be high but, unlike in many cases today, such compensation would truly reflect expected performance rather than rent sharing. Naturally, such a far‐reaching proposal raises diverse issues, like the problem of a reduced confidentiality in the application process, but this should not distract from the potential value the idea has for corporations. Voting rules. Public governance tends to be rather conservative. It is difficult to extend the area of democratic participation or to introduce new voting rules. The major reason is that the established politicians, parties and interest groups fear losing from such changes. The corporate sector, being proud of being more dynamic than the public sector, should find it easier to consider new voting mechanisms for shareholder votes or for decisions taken by the board. Examples are voting by veto (Mueller, 1978) or storable votes (Casella, 2002), but there are many others. Firms can choose the respective innovative voting rules where they are most appropriate, while sticking to simple majority, qualified majority or unanimity elsewhere. 3.4. Evaluation The many different institutional devices developed in public governance to restrict the power of public agents can serve as a pool of ideas to improve corporate governance. The latter has relied too much on the notion that competitive market forces are quite capable of effectively restricting executives. In view of the recent corporate scandals, it seems reasonable to consider novel approaches to control the behaviour of managers, like an improved division of power, well‐defined rules of succession and institutionalised competition in core areas of the corporation. Recent changes in corporate governance rules have indeed co‐opted some of the mechanisms of public governance, like the division of power between management and external auditing firms introduced by the Sarbanes‐Oxley Act. 4. Conclusions In this article, it has been argued that fresh insights for corporate governance can be gained from the way democratic government and public administration are organised. Corporate governance can learn from public governance in such areas as goal‐oriented intrinsic motivation, extrinsic motivations imposed from outside (like titles and orders), and institutional restrictions of power. Counter‐arguments against the very idea of transferring elements of public governance to corporate governance, and even more so against specific institutions, are possible and, in any case, necessary. We have discussed some of these objections in the text. To conclude, we wish to address three more general counter‐arguments that are often brought against new ideas, in order to discuss the limitations as well the potential of the proposals in a broader context. 4.1. ‘The Ideas are Fundamentally Wrong’ This view claims that corporate governance has nothing to learn from public governance; the two sectors are fundamentally different and thus it does not make sense to transfer institutions. It can be argued that the public sector emphasis on intrinsic motivation is outmoded. Even worse, public governance mechanisms may substantially damage corporate governance, because they are alien to the corporate world, thereby lowering productivity and raising the cost of doing business. There are indeed situations where corporate governance can safely disregard public governance ideas. This is, in particular, the case in a perfectly competitive market in which managers have very little discretionary room, and where the goods, managerial and financial markets align their incentives perfectly with those of the shareholders. In a competitive equilibrium, private firms select efficient governance mechanisms, and any intervention by political and regulatory agencies is supposed to be harmful. The disciplinary forces of markets, however, are not likely to be enough to effectively discipline corporate agents and institute efficient governance mechanisms. Corporate governance systems all over the world rely not only on self‐regulation, but also, most importantly, on corporate law and regulatory measures (such as those imposed by the SEC). Indeed, ideally competitive markets are certainly the exception rather than the rule. Practically all markets allow for substantial discretionary behaviour on the part of executives, which sometimes goes as far as illegal actions, as the recent corporate scandals have made clear. This leads to substantial areas within the firm in which power is wielded and in which politics enters the game. Also, a large number of organisations acting on markets are not purely for profit but take an intermediate position somewhere between shareholder wealth maximisation and pure public ownership. There is thus a broad area where insights and institutional ideas derived from public governance are of interest for corporate governance. We have concentrated mainly on developing our arguments with respect to the classical private corporation; the arguments, however, could be applied to an even larger extent to not‐for‐profit firms and firms with a varying degree of governmental influence, which may substantially profit from institutions derived from public governance. 4.2. ‘The Ideas May (Partly) Be Correct but Impossible To Introduce’ This objection raises the concern that rough and ready devices to overcome the problems faced by corporate governance today cannot easily be derived from public governance. Mechanisms would have to be carefully adapted to the needs of firms, but this would be very difficult, if not impossible. There is indeed substantial merit in this objection. Firms act in a different environment and are dominated by a different social decision‐making mechanism from governments; this important distinction has always been a part of Modern Political Economy (Dahl and Lindblom, 1953). We nevertheless think it is not fundamentally impossible to use ideas from the public sphere for more effective corporate governance. Recent regulatory changes embodied in the Sarbanes‐Oxley Act, for example, create a division between executives and external auditors that very much embraces the democratic idea of division of power. Also, it is difficult to see why the interests of principals and agents should be almost exclusively aligned by extrinsic incentives, especially as intrinsic motivation can be formed by institutions and socialisation in a goal‐oriented way. 4.3. ‘The Ideas are Not New’ This charge is certainly true. The ideas presented here are based on fundamental insights from the public sector that hardly any scholar of political economics would find particularly novel. Their application to the corporate world, however, may be relatively new, as may the novel approaches to the governance of firms thereby gained. The notion that elements of public governance can be transferred to corporate governance stands in odd contrast to what is taken to be ‘modern’ today. This holds, in particular, for New Public Management, according to which the public sector should adopt ideas from the corporate sector, or for pay‐for‐performance, where market ideas are transferred within firms. Our arguments suggest that learning can go in a fruitful reverse direction: private governance can learn from public governance. Footnotes 1 " Other authors, among them major contributors to agency theory, tend to defend the existing corporate governance system, but most of them admit major weaknesses. An example is Holmström and Kaplan (2003, p. 2) stating that: ‘[…] while parts of the U.S. corporate governance system failed under the exceptional strain of the 1990’s, the overall system, which includes oversight by the public and the government, reacted quickly to address the problems’. It should, in particular, be noted that the idea of a self‐regulating corporate system based on competitive markets is not seen to be sufficient. 2 " For introductions to public choice theory and modern political economy, see e.g. Persson and Tabellini (2002) and Mueller (1997, 2003). The shortcomings of public bureaucracy are discussed in e.g. Niskanen (1971) or Wintrobe (1997), and rent seeking in Tullock et al. (1988) and Tollison and Congleton (1995). 3 " The extensive surveys by Becht et al. (2002) and Prendergast (1999), which adequately represent the state of research in Agency Theory, virtually disregard intrinsic motivation. In contrast, intrinsic motivation has received considerable attention in the business economics literature, see e.g. Frey and Osterloh (2001) or Davis et al. (1997). 4 " Most of the existing empirical studies analysing the effect of pay on performance, e.g. Lazear (2000), Paarsch and Shearer (1999) or Fehr and Goette (2002), relate to very simple jobs, like windscreen fitting, planting trees or bicycle courier services. Exceptions are Lavy (2002), in his analysis of performance wages for schoolteachers, and certainly also the large literature on performance pay for managers, e.g. Murphy (1999). 5 " Exceptions are the few orders conferred on the battlefield but they are small in number. 6 " This is the current approach, e.g. embodied in the Sarbanes‐Oxley Act, where the term ‘independent director’ is legally defined (Securities and Exchange Commission, 2003). 7 " According to recent empirical evidence, board seats were almost never contested in the US in 1996–2002 (Bebchuk, 2003). Automatic re‐election seems to be the rule, despite the fact that board members formally have to stand for re‐election every year at the shareholder meeting. Bebchuk (2003) gives several reasons for why board seats are almost never contested. Most importantly, the nomination of director candidates in proxy contests is very costly and subject to a public good problem. Moreover, many boards in the US are ‘staggered’, i.e. only a fraction of board members stand for re‐election every year. References Addison , J. , Schnabel , C. and Wagner , J. ( 2004 ). ‘The course of research into the economic consequences of German work councils’ , British Journal of Industrial Relations , vol. 42 ( 2 ), pp. 255 – 81 . Google Scholar Crossref Search ADS WorldCat Beasley , M. ( 1996 ). ‘An empirical analysis of the relation between the board of director composition and financial statement fraud’ , The Accounting Review , vol. 71 ( 4 ), pp. 443 – 65 . OpenURL Placeholder Text WorldCat Bebchuk , L. ( 2003 ). ‘The case for shareholder access to the ballot’ , The Business Lawyer , vol. 59 , pp. 43 – 66 . OpenURL Placeholder Text WorldCat Bebchuk , L. ( 2004 ). ‘The case for empowering shareholders’ . Working Paper, Harvard Law School . Bebchuk , L. and Fried , J. ( 2003 ). ‘Executive compensation as an agency problem’ , Journal of Economic Perspectives , vol. 17 ( 3 ), pp. 71 – 92 . Google Scholar Crossref Search ADS WorldCat Becht , M. , Bolton , P. and Röell , A. ( 2002 ). ‘Corporate governance and corporate control’. ECGI Working Paper Series in Finance 02/2002, also in ( G. Constantinides, M. Harris, and R. Stulz, eds.), Handbook of the Economics of Finance , pp. 1 – 109 .Amsterdam: North Holland . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Besley , T. and Coate , S. ( 2003 ). ‘Elected versus appointed regulators: theory and evidence’ , Journal of the European Economic Association , vol. 1 ( 5 ), pp. 1176 – 205 . Google Scholar Crossref Search ADS WorldCat Brennan , G. and Hamlin , A. ( 2000 ). Democratic Devices and Desires , Cambridge: Cambridge University Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Casella , A. ( 2002 ). ‘Storable votes’ , NBER Working Paper No. 9189. Cooter , R. ( 2000 ). The Strategic Constitution , Princeton: Princeton University Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Cooter , R. ( 2003 ). ‘Who gets on top in democracy? Elections as filters’ , mimeo, School of Law , University of California at Berkeley. Cooter , R. and Eisenberg , M. ( 20002001 ). ‘Fairness, character, and efficiency in firms’ , University of Pennsylvania Law Review , vol. 149 , pp. 1717 – 33 . Google Scholar Crossref Search ADS WorldCat Dahl , R. and Lindblom , C. ( 1953 ). Politics, Economics and Welfare , New York: Harper . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Daily , C. , Dalton , D. and Cannella , A. ( 2003 ). ‘Corporate governance: decades of dialogue and data’ , Academy of Management Review , vol. 28 ( 3 ), pp. 371 – 82 . Google Scholar Crossref Search ADS WorldCat Dalton , D. , Daily , C., Ellstrand , A. and Johnson , J. ( 1998 ). ‘Meta‐analytic reviews of board composition, leadership structure, and financial performance’ , Strategic Management Journal , vol. 19 , pp. 269 – 90 . Google Scholar Crossref Search ADS WorldCat Davis , J. , Schoorman , D. and Donaldson , L. ( 1997 ). ‘Toward a stewardship theory of management’ , Academy of Management Review , vol. 22 ( 1 ), pp. 20 – 47 . Google Scholar Crossref Search ADS WorldCat Dechow , P. , Sloan , R. and Sweene , A. ( 1996 ). ‘Causes and consequences of earnings manipulation: an analysis of firms subject to enforcement actions by the SEC’ , Contemporary Accounting Research , vol. 13 ( 1 ), pp. 1 – 36 . Google Scholar Crossref Search ADS WorldCat Deci , E. and Ryan , R. ( 1985 ). Intrinsic Motivation and Self‐Determination in Human Behavior , New York: Plenum Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Downs , A. ( 1957 ). An Economic Theory of Democracy , New York: Harper and Row . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Easterlin , R. ( 2001 ). ‘Income and happiness: towards an unified theory’ , Economic Journal , vol. 111 ( 473 ), pp. 465 – 84 . Google Scholar Crossref Search ADS WorldCat Economist, The ( 2004 ). ‘Half measures: the auditing industry still needs more reform’ , vol. 373 ( 8402 ), November 20, p. 16 . Efendi , J. , Srivastava , A. and Swanson , E. ( 2004 ). ‘Why do corporate managers misstate financial statements? The role of option compensation, corporate governance, and other factors’ . Working Paper, Texas A&M University. Eisenhardt , K. ( 1989 ). ‘Agency theory: an assessment and review’ , Academy of Management Review , vol. 14 ( 1 ), pp. 57 – 74 . Google Scholar Crossref Search ADS WorldCat Erickson , M. , Hanlon , M. and Maydew , E. ( 2003 ). ‘Is there a link between executive compensation and accounting fraud?’ . Working Paper, Graduate School of Business , University of Chicago. Fehr , E. and Goette , L. ( 2002 ). ‘Do workers work more if wages are high? Evidence from a randomized field experiment’ . Working Paper No. 125, Institute for Empirical Research in Economics , University of Zurich. Frank , R. ( 1985 ). Choosing the Right Pond , Oxford: Oxford University Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Frank , R. ( 2000 ). Luxury Fever: Money and Happiness in an Era of Excess , Princeton: Princeton University Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Frey , B.S. ( 1983 ). Democratic Economic Policy , Oxford: Blackwell . Frey , B. S. ( 1997a ). ‘A constitution for knaves crowds out civic virtues’ , Economic Journal , vol. 107 ( 443 ), pp. 1043 – 53 . Google Scholar Crossref Search ADS WorldCat Frey , B. S. ( 1997b ). Not Just for The Money. An Economic Theory of Personal Motivation , Cheltenham, UK and Brookfield, US: Edward Elgar . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Frey , B. S. ( 2005 ). ‘Knight Fever: Towards an economics of awards’ , Working Paper, Institute for Empirical Research in Economics , University of Zurich. Frey , B. S. and Benz , M. ( 2003 ). ‘Business cycles: political business cycle approach’, in ( H. Vane and B. Snowdown, eds.), An Encyclopaedia of Macroeconomics , pp. 89 – 92 , Cheltenham: Edward Elgar . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Frey , B. S. and Osterloh , M. ( 2001 ). Successful Management by Motivation. Balancing Intrinsic and Extrinsic Incentives , Berlin, New York: Springer . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Frey , B. S. and Stutzer , A. ( 2002 ). Happiness and Economics , Princeton: Princeton University Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Furubotn , E. ( 1988 ). ‘Co‐determination and the modern theory of the firm: a property‐rights analysis’ , Journal of Business , vol. 61 ( 2 ), pp. 165 – 81 . Google Scholar Crossref Search ADS WorldCat Gibbons , R. ( 1998 ). ‘Incentives in organizations’ , Journal of Economic Perspectives , vol. 12 ( 4 ), pp. 115 – 32 . Google Scholar Crossref Search ADS WorldCat Gould , R. ( 2002 ). ‘The origins of status hierarchies: a formal theory and empirical test’ , American Journal of Sociology , vol. 107 ( 5 ), pp. 1143 – 78 . Google Scholar Crossref Search ADS WorldCat Grossman , S. , and Hart , O. ( 1988 ). ‘One share/one vote and the market for corporate control’ , Journal of Financial Economics , vol. 20 , pp. 175 – 202 . Google Scholar Crossref Search ADS WorldCat Hansmann , H. ( 1996 ). The Ownership of Enterprise , Cambridge, MA: Belknap Press of Harvard University Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Harris , E. ( 1990 ). ‘Anti‐takeover measures, golden parachutes, and target firm shareholder welfare’ , RAND Journal of Economics , vol. 21 ( 4 ), pp. 614 – 25 . Google Scholar Crossref Search ADS WorldCat Holmström , B. and Kaplan , S. ( 2003 ). ‘The state of U.S. corporate governance: what's right and what's wrong?’ , NBER Working Paper No. 9613. Holmström , B. and Milgrom , P. ( 1991 ). ‘Multitask principal agent analyses: incentive contracts, asset ownership and job design’ , Journal of Law, Economics and Organization , vol. 7 , pp. 24 – 52 . Google Scholar Crossref Search ADS WorldCat Holmström , B. and Milgrom , P. ( 1994 ). ‘The firm as an incentive system’ , American Economic Review , vol. 84 , pp. 972 – 91 . OpenURL Placeholder Text WorldCat Jensen , M. and Meckling , W. ( 1976 ). ‘Theory of the firm: managerial behavior, agency costs and ownership structure’ , Journal of Financial Economics , vol. 3 , pp. 305 – 60 . Google Scholar Crossref Search ADS WorldCat Jensen , M. , Murphy , K. and Wruck , E. ( 2004 ). ‘Remuneration: where we've been, how we got to here, what are the problems, and how to fix them’ , ECGI Finance Working Paper No. 44/2004. Johnson , S. , Ryan , H. and Tian , Y. ( 2003 ). ‘Executive compensation and corporate fraud’ , Working Paper, Louisiana State University. Lavy , V. ( 2002 ). ‘Paying for performance: the effect of teachers financial incentives on students scholastic outcomes’ , mimeo , Hebrew University Jerusalem. Lazear , E. ( 2000 ). ‘Performance pay and productivity’ , American Economic Review , vol. 90 ( 5 ), pp. 1346 – 61 . Google Scholar Crossref Search ADS WorldCat Milgrom , P. and Roberts , J. ( 1988 ). ‘An economic approach to influence in organizations’ , American Journal of Sociology , vol. 94 , pp. 154 – 79 . Google Scholar Crossref Search ADS WorldCat Mueller , D. ( 1978 ). ‘Voting by veto’ , Journal of Public Economics , vol. 10 , pp. 57 – 75 . Google Scholar Crossref Search ADS WorldCat Mueller , D. ( 1996 ). Constitutional Democracy , Oxford: Oxford University Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Mueller , D. (ed.) ( 1997 ). Perspectives on Public Choice , Cambridge: Cambridge University Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Mueller , D. ( 2003 ). Public Choice III , Cambridge: Cambridge University Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Murphy , K. ( 1999 ). ‘Executive compensation’, in ( O. Ashenfelter, and D., Card, eds.), Handbook of Labor Economics , pp. 2485 – 563 , Amsterdam: Elsevier . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Niskanen , W. ( 1971 ). Bureaucracy and Representative Government , Chicago and New York: Aldine Atherton . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Osterloh , M. and Frey , B. S. ( 2000 ). ‘Motivation, knowledge transfer, and organizational form’ , Organization Science , vol. 11 , pp. 538 – 50 . Google Scholar Crossref Search ADS WorldCat Osterloh , M. and Frey , B. S. ( 2005a ). ‘Yes, managers should be paid like bureaucrats’ , Journal of Management Inquiry , vol. 14 ( 1 ), pp. 96 – 111 . Google Scholar Crossref Search ADS WorldCat Osterloh , M. and Frey , B. S. ( 2005b ). ‘Shareholders should welcome employees as directors’ . Working Paper, Institute for Organization and Theories of the Firm , University of Zurich. Paarsch , H. and Shearer , B. ( 1999 ). ‘The response of worker effort to piece rates: evidence from the British Columbia tree‐planting industry’ , Journal of Human Resources , vol. 34 ( 4 ), pp. 643 – 67 . Google Scholar Crossref Search ADS WorldCat Persson , T. and Tabellini , G. ( 2002 ). Political Economics: Explaining Economic Policy , Cambridge, MA: MIT Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Pollitt , C. and Bouckaert , G. ( 2000 ). Public Management Reform: A Comparative Analysis , Oxford: Oxford University Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Prendergast , C. ( 1993 ). ‘A theory of ‘‘yes men'’’ , American Economic Review , vol. 83 ( 4 ), pp. 757 – 70 . OpenURL Placeholder Text WorldCat Prendergast , C. ( 1999 ). ‘The provision of incentives in firms’ , Journal of Economic Literature , vol. 37 ( 1 ), pp. 7 – 63 . Google Scholar Crossref Search ADS WorldCat Rabin , M. ( 1998 ). ‘Psychology and economics’ , Journal of Economic Literature , vol. 36 ( 1 ), pp. 11 – 46 . OpenURL Placeholder Text WorldCat Roberts , J. and Van den Steen , E. ( 2000 ). ‘Shareholder interests, human capital investments and corporate governance’ , Stanford University, Graduate School of Business Working Paper No. 1631. Ryan , R. and Deci , E. ( 2000 ). ‘The ‘‘what’’ and ‘‘why’’ of goal pursuits: human needs and the self‐determination of behavior’ , Psychological Inquiry , vol. 11 ( 4 ), pp. 227 – 68 . Google Scholar Crossref Search ADS WorldCat Schelker , M. and Eichenberger , R. ( 2003 ). ‘Starke Rechnungsprüfungskommissionen: Wichtiger als direkte Demokratie und Föderalismus? Ein erster Blick auf die Daten’ , Swiss Journal of Economics and Statistics , vol. 139 ( 3 ), pp. 351 – 73 . OpenURL Placeholder Text WorldCat Schelker , M. and Eichenberger , R. ( 2004 ). ‘Independent and competing institutions: an effective way to control government’ . Working Paper, University of Fribourg. Schumpeter , J. ( 1942 ). Capitalism, Socialism and Democracy, 1st edn. , New York: Harper . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Securities and Exchange Commission ( 2003 ). Final Rule on Standards Relating to Listed Company Audit Committees , http://www.sec.gov/rules/final/338220.htm . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Seligman , J. ( 1986 ). ‘Equal protection in shareholder voting rights: the one‐share‐one‐vote controversy’ , George Washington Law Review , vol. 54 , pp. 687 – 724 . OpenURL Placeholder Text WorldCat Shen , W. and Cannella , A. ( 2002 ). ‘Revisiting the performance consequences of CEO succession: the impacts of successor type, postsuccession senior executive turnover, and departing CEO tenure’ , Academy of Management Journal , vol. 45 ( 4 ), pp. 717 – 33 . OpenURL Placeholder Text WorldCat Shleifer , A. and Vishny , R. ( 1997 ). ‘A survey of corporate governance’ , Journal of Finance , vol. 52 ( 2 ), pp. 737 – 83 . Google Scholar Crossref Search ADS WorldCat Stutzer , A. ( 2004 ). ‘The role of income aspirations in individual happiness’ , Journal of Economic Behavior and Organization , vol. 54 ( 1 ), pp. 89 – 109 . Google Scholar Crossref Search ADS WorldCat The Corporate Library ( 2004 ). Split CEO/Chairman Roles – March, 2004 . http://www.thecorporatelibrary.com/Governance‐Research/spotlight‐topics/spotlight/boardsanddirectors/SplitChairs2004.html (as of December 2004). Tirole , J. ( 1992 ). ‘Collusion and the theory of organizations’, in ( J. Laffont, ed.), Advances in Economic Theory: Sixth World Congress , Cambridge University Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Tollison , R. and Congleton , R. (eds) ( 1995 ). The Economic Analysis of Rent Seeking , Aldershot: Edward Elgar . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Tullock , G. ( 1964 ). Politics of Bureaucracy , Washington D.C.: Public Affairs Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Tullock , G. , Tollison , R. and Rowley , C. (eds) ( 1988 ). The Political Economy of Rent Seeking , Boston: Kluwer . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Uzun , H. , Szewczyk , S. and Varma , R. ( 2004 ). ‘Board composition and corporate fraud’ , Financial Analysts Journal , vol. 60 ( 3 ), pp. 33 – 44 . Google Scholar Crossref Search ADS WorldCat van den Berg , A. ( 2004 ). ‘The contribution of work representation to solving the governance structure problem’ , Journal of Management and Governance , vol. 8 ( 2 ), pp. 129 – 48 . Google Scholar Crossref Search ADS WorldCat Weber , M. ( 1978 ). Economy and Society: an Outline of Interpretative Sociology , Berkeley: University of California Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Weisbrod , B. ( 1998 ). To Profit or Not to Profit: The Commercial Transformation of the Nonprofit Sector , New York: Cambridge University Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Wintrobe , R. ( 1997 ). ‘Modern bureaucratic theory’, in ( D. Mueller, ed.), Perspectives in Public Choice: A Handbook , pp. 429 – 54 , Cambridge: Cambridge University Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Author notes " We thank Margit Osterloh and Reiner Eichenberger for extensive discussions on corporate governance issues and we are grateful to two anonymous referees, Christoph Engel, René L. Frey, Jesse Fried, Henry Hansmann, Bernd Helmig, Stephan Meier, Dennis Mueller, Daniel Rubinfeld, Alois Stutzer and conference participants in Zurich and Berkeley for helpful remarks. The second author thanks the Boalt School of Law, University of California at Berkeley, for its hospitality during the preparation of this article and acknowledges financial support by the Swiss National Science Foundation. © Royal Economic Society 2005.

doi: 10.1111/1468-0297.00051-i1pmid: N/A

Article PDF first page preview Close This content is only available as a PDF. © Royal Economic Society 2005.

doi: 10.1111/j.1468-0297.2005.01044.xpmid: N/A

The inclusion of a book or publication in this list does not preclude publication in a future issue of the Journal. Abolafia (Mitchel Y.) (Ed). Markets. Aldershot and Lyme, NH: Edward Elgar, 2005. Pp. xvii+526. £135.00 hardback. ISBN 1 84376 051 7. Aichholzer (Georg) and Burkert (Herbert) (Eds). Public Sector Information in the Digital Age: Between Markets, Public Management and Citizens‘ Rights. Aldershot and Lyme, NH: Edward Elgar, 2004. Pp. xvi+342. £69.95 hardback. ISBN 1 84376 383 4. Alesina (Alberto). Institutional Reforms: The Case of Colombia. Massachusetts Institute of Technology: The MIT Press, 2005. Pp. x+373. £48.95 hardback, £22.95 paperback. ISBN 0 262 01214 6, 0262 51182 7. Ali (Abbas J.). Islamic Perspectives on Management and Organization. Aldershot and Lyme, NH: Edward Elgar, 2005. Pp. x+260. £59.95 hardback. ISBN 1 84376 766 X. Alm (James), Martinez‐Vazquez (Jorge) and Indrawati (SriMulyani) (Eds). Reforming Intergovernmental Fiscal Relations and the Rebuilding of Indonesia: The ‘Big Bang’ Program and its Economic Consequences. Aldershot and Lyme, NH: Edward Elgar, 2004. Pp. viii+360. £69.95 hardback. ISBN 1 84376 451 2. Anderson (Kym) and Josling (Tim) (Eds). The WTO and Agriculture: Volume 1 and 2. Aldershot and Lyme, NH: Edward Elgar, 2005. Pp. xxiv+1120. £275.00 hardback. ISBN 1 84376 279 X. Ando (Ken‐Ichi). Japanese Multinationals in Europe: A Comparison of the Automobile and Pharmaceutical Industries. Aldershot and Lyme, NH: Edward Elgar, 2005. Pp. x+201. £59.95 hardback. ISBN 1 84376 655 8. Armitage (Seth). The Cost of Capital: Intermediate Theory. Cambridge and New York: Cambridge University Press, 2005. Pp. xv+353. £50.00 hardback, £24.99 paperback. ISBN 0 521 80195 8, 0 521 00044 0. Arnone (Marco) and Ugolini (Piero). Primary Dealers in Government Securities. Washington, DC: International Monetary Fund, 2005. Pp. v+92. US $25.00 paperback. ISBN 1 58906 379 1. Arora (Ashish) and Gambardella (Alfonso) (Eds). From Underdogs to Tigers: The Rise and Growth of the Software Industry in Brazil, Chine, India, Ireland, and Israel. Oxford: Oxford University Press, 2005. Pp. xii+313. £50.00 hardback. ISBN 0 19 927560 2. Asefa (Sisay) (Ed). The Economics of Sustainable Development. Kalamazoo, Michigan: W.E. Upjohn Institute, 2005. Pp. vii+191. US $40.00 hardback, US $15.00 paperback. ISBN 0 88099 321 9, 0 88099 320 0. Atkinson (Robert D.). The Past and Future of America's Economy: Long Waves of Innovation that Power Cycles of Growth. Aldershot and Lyme, NH: Edward Elgar, 2004. Pp. vii+357. £75.00 hardback. ISBN 1 84376 955 7. Austen‐Smith (David) and Banks (Jeffrey S.). Positive Political Theory II: Strategy and Structure. Ann Arbor: University of Michigan Press, 2004. Pp. xvii454. £15.50 paperback, US $29.95 paperback. ISBN 0 472 06894 6. Ayres (Ian). Optional Law: The Structure of Legal Entitlements. Chicago and London: University of Chicago Press, 2005. Pp. x+284. £34.50 hardback, US $49.00 hardback. ISBN 0 226 03346 5. Bailey (R. E.). The Economics of Financial Markets. Cambridge and New York: Cambridge University Press, 2005. Pp. xix+528. £60.00 hardback, US $100.00 hardback, £29.99 paperback, US $50.00 paperback. ISBN 0 521 84827 X, 0 521 61280 2. Baltagi (Badi H.). Econometric Analysis of Panel Data. New York and Chichester: John Wiley & Sons Ltd., 2005. Pp. xii+302. ISBN 0 470 01456 3. Bardsen (Gunnar), Eitrheim (Oyvind) and Jansen (Eilev S.) (Eds). The Econometrics of Macroeconomic Modelling: Advanced Texts in Econometrics. Oxford: Oxford University Press, 2005. Pp. xiii+338. £60.00 hardback. ISBN 0 19 924694 1. Bateman (Ian J.), Lovett (Andrew A.) and Brainard (Julii S), et al. Applied Environmental Economics: A GIS Approach to Cost‐Benefit Analysis. Cambridge and New York: Cambridge University Press, 2005. Pp. xxi+335. £45.00 hardback, US $70.00 hardback, £24.99 paperback, US $43.00 paperback. ISBN 0 521 67158 2. Bazerman (Max H.) (Ed). Negotiation, Decision Making and Conflict Management: Volume 1, 2 and 3. Aldershot and Lyme, NH: Edward Elgar, 2005. Pp. xviii+1992. £450.00 hardback. ISBN 1 84376 377 X. Bebchuk (Lucian) and Fried (Jesse). Pay Without Performance: The Unfulfilled Promise of Executive Compensation. Cambridge, Mass. and London: Harvard University Press, 2004. Pp. xii+278. US $24.95 hardback. ISBN 0 674 01665 3. Bernardi (Luigi), Chandler (Mark W. S.) and Gandullia (Luca) (Eds). Tax Systems and Tax Reforms in New EU Members. London and New York: Routledge, 2005. Pp. xxvi+258. £65.00 hardback. ISBN 0 415 34988 5. Bernholz (Peter) and Vaubel (Roland) (Eds). Political Competition, Innovation and Growth in the History of Asian Civilizations. Aldershot and Lyme, NH: Edward Elgar, 2004. Pp. xii+225. £59.95 hardback. ISBN 1 84376 919 0. Bery (Suman), Bosworth (Bary) and Panagariya (Arvind) (Eds). India Policy Forum 2004: Volume 1. :, 2004. Pp. xxvi+332. £24.00 paperback. ISBN 1 8157 0881 5. Bierens (Herman J.). Introduction to the Mathematical and Statistical Foundations of Econometrics. Cambridge and New York: Cambridge University Press, 2005. Pp. xvii+323. £50.00 hardback, US $90.00 hardback, £22.99 paperback, US $39.99 paperback. ISBN 0 521 83431 7, 0 521 54224 3. Bill (Jenkins) and Page (Edward C.) (Eds). The Foundations of Bureaucracy in Economic and Social Thought: Volume 1 and 2. Aldershot and Lyme, NH: Edward Elgar, 2004. Pp. xxxix+1434. £395.00 hardback. ISBN 1 84064 015 4. Birkinshaw (Julian) (Ed). Strategic Management: Volume 1 and 2. Aldershot and Lyme, NH: Edward Elgar, 2004. Pp. xix+1038. £290.00 hardback. ISBN 1 84376 278 1. Blair (Roger D.) and Lafontaine (Francine). The Economics of Franchising. Cambridge and New York: Cambridge University Press, 2005. Pp. xii+338. £25.00 hardback, US $40.00 hardback. ISBN 0 521 77252 4. Bonoli (Giuliano) and Shinkawa (Toshimitsu) (Eds). Ageing and Pension Reform Around the World: Evidence from Eleven Countries. Aldershot and Lyme, NH: Edward Elgar, 2005. Pp. xi+279. £69.95 hardback. ISBN 1 84376 771 6. Bosworth (Derek L.). Determinants of Enterprise Performance. Manchester: Manchester University Press, 2005. Pp. xv+447. £60.00 hardback. ISBN 0 7190 6774 X. Brooks (Robin) and Razin (Assaf). Social Security Reform: Financial and Political Issues in International Perspective. Cambridge and New York: Cambridge University Press, 2005. Pp. viii+375. £55.00 hardback, US $75.00 hardback. ISBN 0 521 84495 9. Brown (Allan), Hossain (Moazzem) and Nguyen (Duc‐Tho) (Eds). Telecommunications Reform in the Asia‐Pacific Region. Aldershot and Lyme, NH: Edward Elgar, 2004. Pp. xxi+264. £69.95 hardback. ISBN 1 84064 691 8. Burda (Michael) and Wyplosz (Charles). Macroeconomics: A European Text, Fourth Edition. Oxford: Oxford University Press, 2005. Pp. xix+576. £36.99 paperback. ISBN 0 19 926496 1. Burke (Ronald J.) and Mattis (Mary C.) (Eds). Supporting Women's Career Advancement: Challenges and Opportunities. Aldershot and Lyme, NH: Edward Elgar, 2005. Pp. ix+359. £75.00 hardback. ISBN 1 84376 633 7. Caprio (Gerard), Fiechter (Jonathan L.) and Litan (Robert E.) (Eds). The Future of State‐Owned Financial Institutions. Washington, DC: Brookings Institution Press, 2005. Pp. x+382. £28.00 paperback. ISBN 08157 1335 5. Carlberg (Michael). International Economic Policy Coordination. Berlin, Heidelberg and New York: Springer, 2005. Pp. xvi+312. £65.00 hardback. ISBN 3 540 24445 X. Carrier (James G.) (Ed). A Handbook of Economic Anthropology. Aldershot and Lyme, NH: Edward Elgar, 2005. Pp. xvi+584. £135.00 hardback. ISBN 1 84376 175 0. Chappell JR. (Henry W.), Mcgregor (RobRoy) and Vermilyea (Todd A.). Committee Decisions on Monetary Policy: Evidence from Historical Records of the Federal Open Market Committee. Massachusetts Institute of Technology: The MIT Press, 2005. Pp. xv+313. £25.95 hardback. ISBN 0 262 03330 5. Chernoy (Lev). Globalization: Past or Future. New York: Pleiades Publishing, Inc. 2004. Pp. 381. ISBN 1 931938 28 8. Chernoy (Lev). Economy Market the State. New York: Pleiades Publishing, Inc., 2002. Pp. 225. ISBN 1 931938 08 3. Cheshire (Paul C.) and Duranton (Gilles). Recent Developments in Urban and Regional Economics. Aldershot and Lyme, NH: Edward Elgar, 2004. Pp. xxvi+607. £150.00 hardback. ISBN 1 84064 953 4. Chiswick (Barry R.). The Economics of Immigration: Selected Papers of Barry R. Chiswick. Aldershot and Lyme, NH: Edward Elgar, 2005. Pp. xxi+400. £80.00 hardback. ISBN 1 84376 458 X. Collins (Susan M.) and Graham (Carol) (Eds). Brookings Trade Forum 2004: Globalization, Poverty, And Inequality. Washington, DC: Brookings Institution Press, 2004. Pp. xxxi+311. £19.00 paperback. ISBN 0 8157 1286 3. Cooke (Philip) and Piccaluga (Andrea) (Eds). Regional Economies as Knowledge Laboratories. Aldershot and Lyme, NH: Edward Elgar, 2004. Pp. xxx+241. £69.95 hardback. ISBN 1 84376 821 6. Cowen (Tyler). Markets and Cultural Voices: Liberty vs. Power in the Lives of Mexican Amate Painters. Ann Arbor: University of Michigan Press, 2005. Pp. x+189. £40.00 hardback, US $70.00 hardback, £15.50 paperback, US $26.95 paperback. ISBN 0 472 09889 6, 0 472 06889 X. Dellink (Rob B). Modelling the Costs of Environmental Policy: A Dynamic Applied General Equilibrium Assessment. Aldershot and Lyme, NH: Edward Elgar, 2005. Pp. vii+305. £69.95 hardback. ISBN 1 84542 109 4. Di Tommaso (Marco R.) and Schweitzer (Stuart O.) (Eds). Health Policy and High‐Tech Industrial Development: Learning from Innovation in the Health Industry. Aldershot and Lyme, NH: Edward Elgar, 2005. Pp. xiii+285. £69.95 hardback. ISBN 1 84376 757 0. Dinello (Natalia) and Squire (Lyn) (Eds). Globalization and Equity: Perspectives From The Developing World. Aldershot and Lyme, NH: Edward Elgar, 2005. Pp. xxii+253. £59.95 hardback. ISBN 0 84376 884 4. Dopfer (Kurt) (Ed). The Evolutionary Foundations of Economics. Cambridge and New York: Cambridge University Press, 2005. Pp. xiii+577. £60.00 hardback, US $110.00 hardback. ISBN 0 521 62199 2. Dunning (John H.) and Narula (Rajneesh). Multinationals and Industrial Competitiveness: A New Agenda. Aldershot and Lyme, NH: Edward Elgar, 2004. Pp. x+287. £65.00 hardback. ISBN 1 84376 686 8. Dustmann (Christian) and Glitz (Albrecht) (Eds). Immigration, Jobs and Wages: Theory, Evidence and Opinion. London: Centre for Economic Policy Research, 2005. Pp. xiv+66. £25.00 paperback. ISBN 1 8988128 87 1. Ebrahim (Alnoor). NGOs and Organizational Change: Discourse, Reporting, and Learning. Cambridge and New York: Cambridge University Press, 2005. Pp. x+181. £40.00 hardback, US $55.00 hardback, £19.99 paperback, US $34.99 paperback. ISBN 0 521 82486 9, 0 521 67157 4. Eggertsson (Thrainn). Imperfect Institutions: Possibilities and Limits of Reform. Ann Arbor: University of Michigan Press, 2005. £37.00 hardback, US $65.00 hardback, US $37.00 paperback. ISBN 0 472 11456 5, 0 472 03039 6. Elzen (Boelie), Geels (Frank W.) and Green (Ken) (Eds). System Innovation and the Transition to Sustainability: Theory, Evidence and Policy. Aldershot and Lyme, NH: Edward Elgar, 2004. Pp. xx+315. £69.95 hardback. ISBN 1 84376 683 3. Evans (Alan W.). Economics and Land Use Planning. Oxford, Cambridge, MA and Victoria, Australia: Blackwell Publishing, 2004. Pp. xii+209. £37.50 paperback. ISBN 1 4051 1861 X. Evans (Alan W.). Economics, Real Estate and the Supply of Land. V Oxford, Cambridge, MA and Victoria, Australia: Blackwell Publishing, 2004. Pp. xiv+258. £39.50 paperback. ISBN 1 4051 1862 8. Fayolle (Alain), Kyro (Paula) and Ulijn (Jan) (Eds). Entrepreneurship Research in Europe: Outcomes and Perspectives. Aldershot and Lyme, NH: Edward Elgar, 2005. Pp. xi+366. £75.00 hardback. ISBN 1 84376 599 3. Forster (Nick). Maximum Performance: A Practical Guide to Leading and Managing People at Work. Aldershot and Lyme, NH: Edward Elgar, 2005. Pp. xix+594. £35.00 paperback. ISBN 1 84542 378 X. Franses (PhilipHans) and Richard (Paap). Periodic Time Series Models: Advanced Texts in Econometrics. Oxford: Oxford University Press, 2004. Pp. xiv+147. £53.00 hardback, £26.50 paperback. ISBN 0 19 924202 X, 0 19 924203 8. Gallant (M. Michelle). Money Laundering and the Proceeds of Crime: Economic Crime and Civil Remedies. Aldershot and Lyme, NH: Edward Elgar, 2005. Pp. viii+153. £45.00 hardback. ISBN 1 84376 951 4. Geels (Frank W.). Technological Transitions and System Innovations: A Co‐Evolutionary and Socio‐Technical Analysis. Aldershot and Lyme, NH: Edward Elgar, 2005. Pp. x+318. £79.95 hardback. ISBN 1 84542 009 8. Grauwe (PaulDe). Exchange Rate Economics: Where Do We Stand?. Massachusetts Institute of Technology: The MIT Press, 2005. Pp. xvi+347. £29.95 hardback. ISBN 0 262 04222 3. Green (Ken), Miozzo (Marcela) and Dewick (Paul) (Eds). Technology, Knowledge and the Firm: Implications for Strategy and Industrial Change. Aldershot and Lyme, NH: Edward Elgar, 2005. Pp. x+308. £69.95 hardback. ISBN 1 84376 877 1. Grimsey (Darrin) and Lewis (Mervyn K.). The Economics of Public Private Partnerships. Aldershot and Lyme, NH: Edward Elgar, 2005. Pp. xl+584. £150.00 hardback. ISBN 1 84376 249 8. Grobher (Gernot) and Powell (Walter W.) (Eds). Networks: Volume 1 and 2. Aldershot and Lyme, NH: Edward Elgar, 2004. Pp. xxxi+1286. £305.00 hardback. ISBN 1 84376 035 5. Gruber (Harald). The Economics of Mobile Telecommunications. Cambridge and New York: Cambridge University Press, 2005. Pp. xviii+323. £50.00 hardback, US $80.00 hardback. ISBN 0 521 84327 Guerrieri (Paolo), Iapadre (P. Lelio) and Koopmann (Georg) (Eds). Cultural Diversity and International Economic Integration: The Global Governnance of the Audio‐Visual Sector. Aldershot and Lyme, NH: Edward Elgar, 2005. Pp. x+296. £65.00 hardback. ISBN 1 84376 807 0. Guesnerie (Roger). Assessing Rational Expectations 2: Eductive Stability in Economics. Massachusetts Institute of Technology: The MIT Press, 2005. Pp. xxviii+455. £32.95 hardback. ISBN 0 262 07258 0. Hall (Robert E.) and Papell (David H.) (Eds). Macroeconomics: Economic Growth, Fluctuations, and Policy. Sixth Edition. : Norton Books, 2005. Pp. xxxi+521. £37.99 paperback. ISBN 0 393 92751 2. Hanson (David). CE Marking, Product Standards and World Trade. Aldershot and Lyme, NH: Edward Elgar, 2005. Pp. xiii+236. £59.95 hardback. ISBN 1084376 773 2. Hayami (Yujiro) and Godo (Yoshihisa). Development Economics: From the Poverty to the Wealth of Nations, Third Edition. Oxford: Oxford University Press, 2005. Pp. xviii+430. £22.95 paperback. ISBN 0 19 927271 9. Henderson (J. Vernon). New Economic Geography. Aldershot and Lyme, NH: Edward Elgar, 2005. Pp. xv+617. £150.00 hardback. ISBN 1 84376 038 X. Hensher (David A.), Rose (John M.) and Greene (William H.). Applied Choice Analysis: A Primer. Cambridge and New York: Cambridge University Press, 2005. Pp. xxiv+717. £70.00 hardback, £35.00 paperback. ISBN 0 521 84426 6, 0 521 60577 6. Hippel (EricVon). Democratizing Innovation. Massachusetts Institute of Technology: The MIT Press, 2005. Pp. x+204. £19.95 hardback. ISBN 0 262 00274 4. Hoti (Suhejla) and Mcaleer (Michael). Contributions to Economic Analysis: Modelling the Riskiness in Country Risk Ratings. Amsterdam: Elsevier, 2005. Pp. xix+492. £79.00 hardback, US $127.00 hardback. ISBN 0 444 51837 1. Hudson (Michael). Global Fracture: The New International Economic Order. London: Pluto Press, 2005. Pp. xxxi+296. £19.95 paperback. ISBN 0 7453 2394 4. Irwin (Douglas A.). Free Trade Under Fire. Princeton, NJ: Princeton University Press, 2005. Pp. xi+290. £12.95 paperback. ISBN 0 691 12247 4. Jewson (Stephen), Brix (Anders) and Ziehmann (Christine). Weather Derivative Valuation: The Meteorological, Statistical, Financial and Mathematical Foundations. Cambridge and New York: Cambridge University Press, 2005. Pp. xvii+373. £45.00 hardback, US $75.00 hardback. ISBN 0 521 84371 5. Jha (Raghbendra) (Ed). Economic Growth, Economic Performance and Welfare in South Asia. Basingstoke and London: Palgrave, 2005. Pp. xv+407. £65.00 hardback. ISBN 1 4039 4361 3. Julian (Craig C.). International Joint Venture Performance in South East Asia. Aldershot and Lyme, NH: Edward Elgar, 2005. Pp. viii+364. £69.95 hardback. ISBN 1 84376 094 0. Just (Richard E.), Hueth (Darrell L.) and Schmitz (Andrew). The Welfare Economics of Public Policy: A Practical Approach to Project and Policy Evaluation. Aldershot and Lyme, NH: Edward Elgar, 2004. Pp. xvii+688. £125.00 hardback. ISBN 1 84376 688 4. Kanbur (Ravi) and Venables (Anthony J.) (Eds). Spatial Inequality and Development. Oxford: Oxford University Press, 2005. Pp. xx+412. ISBN 0 19 927 863 6. Karlsson (Charlie), Flensburg (Per) and Horte (Sven‐Ake) (Eds). Knowledge Spillovers and Knowledge Management. Aldershot and Lyme, NH: Edward Elgar, 2004. Pp. ix+510. £85.00 hardback. ISBN 1 84376 785 6. Keating (Michael). Regions and Regionalism in Europe. Aldershot and Lyme, NH: Edward Elgar, 2004. Pp. xv+692. £175.00 hardback. ISBN 1 84376 127 0. Keen (Steve). Debunking Economics: The Naked Emperor of the Social Sciences. London: Pluto Press, 2001. Pp. 335. US $29.85 Kehoe (Timothy J.), Srinivasan (T. N.) and Whalley (John) (Eds). Frontiers in Applied General Equilibrium Modeling. Cambridge and New York: Cambridge University Press, 2005. Pp. xi+436. £55.00 hardback, US $90.00 hardback. ISBN 0 521 82525 3. Kenney (Douglas S.) (Ed). In Search of Sustainable Water Management: International Lessons for the American West and Beyond. Aldershot and Lyme, NH: Edward Elgar, 2005. Pp. xviii+185. £49.95 hardback. ISBN 1 84376 944 1. King (Christopher T.) and Mueser (Peter R.). Welfare and Work: Experiences in Six Cities. Kalamazoo, Michigan: W.E. Upjohn Institute, 2005. Pp. xii+195. US $40.00 hardback, US $17.00 paperback. ISBN 0 88099 319 7, 0 88099 318 9. Klosko (George). Political Obligations. Oxford: Oxford University Press, 2005. Pp. x+266. ISBN 0 19 925620 9. Knight (John) and Song (Lina) (Eds). Towards a Labour Market in China. Oxford: Oxford University Press, 2005. Pp. xvi+258. £55.00 hardback. ISBN 0 19 924527 4. Koundouri (Phoebe). Econometrics Informing Natural Resources Management: Selected Empirical Analyses. Aldershot and Lyme, NH: Edward Elgar, 2004. Pp. xiii+391. £79.95 hardback. ISBN 1 84376 922 0. Krarup (Signe) and Russell (Clifford S.) (Eds). Environment, Information and Consumer Behaviour. Aldershot and Lyme, NH: Edward Elgar, 2005. Pp. xiii+299. £65.00 hardback. ISBN 1 84542 011 X. Lavoie (Marc) and Seccareccia (Mario) (Eds). Central Banking in the Modern World: Alternative Perspectives. Aldershot and Lyme, NH: Edward Elgar, 2004. Pp. xiii+296. £65.00 hardback. ISBN 1 84376 641 8. Letto‐Gillies (Grazia). Transnational Corporations and International Production: Concepts, Theories and Effects. Aldershot and Lyme, NH: Edward Elgar, 2005. Pp. xiii+252. £59.95 hardback. ISBN 1 84376 856 9. Lippit (Victor D.). Capitalism. London and New York: Routledge, 2005. Pp. xvi+188. £65.00 hardback. ISBN 0 415 27394 3. Little (I. M. D.). Ethics, Economics and Politics: Principles of Public Policy. Oxford: Oxford University Press, 2003. Pp. xvi+162. £12.50 paperback. ISBN 0 19 926872 X. Lorsch (Jay W.), Berlowitz (Leslie) and Zelleke (Andy) (Eds). Restoring Trust in American Business. Massachusetts Institute of Technology: The MIT Press, 2005. Pp. 184. £25.95 hardback, £10.95 paperback. ISBN 0 262 24048 3, 0 262 74027 3. Maas (Harro). William Stanley Jevons and the Making of Modern Economics. Cambridge and New York: Cambridge University Press, 2005. Pp. xxii+330. £45.00 hardback, US $75.00 hardback. ISBN 0 521 82712 4. Mansell (Robin) and Collins (Brian S.) (Eds). Trust and Crime in Information Societies. Aldershot and Lyme, NH: Edward Elgar, 2005. Pp. xvi+459. £79.95 hardback. ISBN 1 84542 177 9. Marcelle (Gillian M.). Technological Learning: A Strategic Imperative for Firms in the Developing World. Aldershot and Lyme, NH: Edward Elgar, 2004. Pp. xiii+212. £55.00 hardback. ISBN 1 84376 692 2. Matthews (Kent) and Thompson (John). The Economics of Banking. New York and Chichester: John Wiley & Sons Ltd., 2005. Pp. x+ 246. ISBN 0 470 09008 1. Mcconnell (John J.) and Denis (David J.) (Eds). Corporate Restructuring: Volume 1 and 2. Aldershot and Lyme, NH: Edward Elgar, 2005. Pp. xxii+1226. £295.00 hardback. ISBN 1 84376 727 9. Mckinnon (Ronald I.). Exchange Rates Under the East Asian Dollar Standard: Living with Conflicted Virtue. Massachusetts Institute of Technology: The MIT Press, 2005. Pp. x+279. £22.95 hardback. ISBN 0 262 13451 9. Meier (Gerald M.) and Rauch (James E.). Leading Issues in Economic Development. New York: Oxford University Press, 2005. Pp. xviii+650. £33.50 paperback.ISBN 0 19 517960 9. Menard (Claude) (Ed). The International Library of The New Institutional Economics: Volume 1, 2, 3, 4, 5, 6, and 7. Aldershot and Lyme, NH: Edward Elgar, 2004. £950.00 hardback. ISBN 1 84064 382 X. Minsky (Hymen P.) and Papadimitriou (Dimitri B.) (Eds). Induced Investment and Business Cycles. Aldershot and Lyme, NH: Edward Elgar, 2004. Pp. xvi+272. £75.00 hardback. ISBN 0 84376 216 1. Mirowski (Philip). The Effortless Economy of Science?. Durham, NC: Duke University Press, 2004. Pp. 463. £70.00 hardback, £18.95 paperback. ISBN 0 8223 3310 4, 0 8223 3322 8. Moizer (Peter) (Ed). Governance and Auditing. Aldershot and Lyme, NH: Edward Elgar, 2005. Pp. xxv+513. £150.00 hardback. ISBN 1 84376 830 5. Mooij (Jos) (Ed). The Politics of Economic Reforms in India. New Delhi: Thousand Oaks/London: Sage Publications, 2005. Pp. 362. ISBN 0 7619 3343 3. Mooslechner (Peter), Schuberth (Helene) and Schurz (Martin) (Eds). Economic Policy under Uncertainty: The Role of Truth and Accountability in Policy Advice. Aldershot and Lyme, NH: Edward Elgar, 2004. Pp. xi+322. £69.95 hardback. ISBN 1 84376 485 7. Morth (Ulrika) (Ed). Soft Law in Governance and Regulation: An Interdisciplinary Analysis. Aldershot and Lyme, NH: Edward Elgar, 2004. Pp. ix+224. £55.00 hardback. ISBN 1 84376 571 3. Moul (Charles C.). A Concise Handbook of Movie Industry Economics. Cambridge and New York: Cambridge University Press, 2005. Pp. viii+214. £25.00 hardback, US $35.00 hardback. ISBN 0 521 84384 7. Mulder (Peter). The Economics of Technology Diffusion and Energy Efficiency. Aldershot and Lyme, NH: Edward Elgar, 2005. Pp. vi+264. £65.00 hardback. ISBN 1 84376 823 2. Nooy (WouterDe), Mrvar (Andrej) and Batagelj (Viladimir). Exploratory Social Network Analysis with Pajek. Cambridge and New York: Cambridge University Press, 2005. Pp. xxvii+334. £50.00 hardback, US $90.00 hardback, £22.99 paperback, US $39.99 paperback. ISBN 0 521 84173 9, 0 521 60262 9. Perdikis (Nicholas) and Read (Robert) (Eds). The WTO and the Regulation of International Trade: Recent Trade Disputes between the European Union and the United States. Aldershot and Lyme, NH: Edward Elgar, 2005. Pp. xvii+295. £65.00 hardback. ISBN 1 84376 200 5. Posen (Adam S.) (Ed). The Euro at Five: Ready for a Global Role?. Washington, DC: Institute for International Economics, 2005. Pp. xi+204. US $23.95 paperback. ISBN 0 88132 374 8. Proops (John) and Safonov (Paul) (Eds). Modelling in Ecological Economics. Aldershot and Lyme, NH: Edward Elgar, 2004. Pp. ix+213. £55.00 hardback. ISBN 1 84376 222 6. Puig‐Junoy (Jaume) (Ed). The Public Financing of Pharmaceuticals: An Economic Approach. Aldershot and Lyme, NH: Edward Elgar, 2005. Pp. xiii+255. £65.00 hardback. ISBN 1 84542 088 8. Quaddus (M.A.) and Siddique (M.A.B) (Eds). Handbook of Sustainable Development Planning: Studies in Modelling and Decision Support. Aldershot and Lyme, NH: Edward Elgar, 2004. Pp. xii+347. £115.00 hardback. ISBN 1 84064 879 1. Razin (Assaf), Sadka (Efraim) and Nam (ChangWoon). The Decline of the Welfare State: Demography and Globalization. Massachusetts Institute of Technology: The MIT Press, 2005. Pp. viii+133. £18.95 hardback. ISBN 0 262 18244 0. Reisch (Lucia A.) and Ropke (Inge) (Eds). The Ecological Economics of Consumption. Aldershot and Lyme, NH: Edward Elgar, 2004. Pp. xi+259. £65.00 hardback. ISBN 1 84376 512 8. Ritter (Jay R.) (Ed). Recent Developments in Corporate Finance: Volume 1 and 2. Aldershot and Lyme, NH: Edward Elgar, 2005. Pp. xx+839. £220.00 hardback. ISBN 1 84376 797 X. Roncaglia (Alessandro). The Wealth of Ideas: A History of Economic Thought. Cambridge and New York: Cambridge University Press, 2005. Pp. xiv+582. £60.00 hardback, US $110.00 hardback. ISBN 0 521 84337 5. Roy (Sumit). Globalisation, ICT and Developing Nations: Challenges in the Information Age. New Delhi: Thousand Oaks/London: Sage Publications, 2005. Pp. 245. ISBN 0 7619 3345 X, 0 7619 3346 8. Sacerdoti (Emilio), El‐Masry (Gamal) and Khandelwal (Padamja), et al. Mauritius: Challenges of Sustained Growth. Washington, DC: International Monetary Fund, 2005. Pp. vii+102. US $25.00 paperback. ISBN 1 58906 416 X. Salemi (Michael K.) and Hansen (W. Lee). Discussing Economics: A Classroom Guide to Preparing Discussion Questions and Leading Discussion. Aldershot and Lyme, NH: Edward Elgar, 2005. Pp. xv+318. £59.95 hardback. ISBN 1 84376 449 0. Samuels (Warren), Biddle (Jeff E.) and Emmett (Ross B.) (Eds). A Research Annual: Research in the History of Economic Thought and Methodology Volume 23A. Amsterdam: Elsevier, 2005. Pp. xi+259. £66.00 hardback, US $105.00 hardback. ISBN 0 7623 1164 9. Samuels (Warren J.) (Ed). Documents From F. Taylor Ostrander: Research in the History of Economic Thought and Methodology Volume 23B. Amsterdam: Elsevier, 2005. Pp. ix+400. £72.00 hardback, US $116.00 hardback. ISBN 0 7623 1165 7. Samuels (Warren J.) (Ed). Further University of Wisconsin Materials: Further Documents of F. Taylor Ostrander: Research in the History of Economic Thought and Methodology Volume 23C. Amsterdam: Elsevier, 2005. Pp. ix+367. £72.00 hardback, US $116.00 hardback. ISBN 0 7623 1166 5. Sartorius (Christian) and Zundel (Stefan) (Eds). Time Strategies, Innovation and Environmental Policy. Aldershot and Lyme, NH: Edward Elgar, 2005. Pp. x+358. £75.00 hardback. ISBN 1 84542 090 X. Schramm‐Nielsen (Jette), Lawrence (Peter) and Sivesind (KarlHenrik). Management in Scandinavia: Culture, Context and Change. Aldershot and Lyme, NH: Edward Elgar, 2004. Pp. vii+204. £55.00 hardback. ISBN 1 84376 431 8. Segura (Julio) and Braun (CarlosRodriguez) (Eds). An Eponymous Dictionary of Economics: A Guide to Laws and Theorems Named after Economists. Aldershot and Lyme, NH: Edward Elgar, 2004. Pp. xxviii+280. £90.00 hardback. ISBN 1 84376 029 0. SeroaDaMotta (Ronaldo), Thomas (Alban) and Hazin (LilianSaade), et al. Economic Instruments for Water Management: The Cases of France, Mexico and Brazil. Aldershot and Lyme, NH: Edward Elgar, 2004. Pp. x+151. £49.95 hardback. ISBN 1 84376 964 6. Setterfield (Mark) (Ed). Interactions in Analytical Political Economy: Theory, Policy and Applications. Armonk, New York and London: M.E. Sharpe, 2005. Pp. ix+281. US $89.95 hardback. ISBN 0 7656 1482 0. Shephard (Neil) (Ed). Stochastic Volatitlity: Selected Readings. Oxford: Oxford University Press, 2005. Pp. viii+525. £60.00 hardback. ISBN 0 19 925719 1. Shionoya (Yuichi). The Soul of the German Historical School: Methodological Essays on Schmoller, Weber and Schumpeter. Berlin, Heidelberg and New York: Springer, 2005. Pp. xv+207. £80.50 hardback, US $129.00 hardback. ISBN 0 387 23083 1. Sorge (Arndt) (Ed). The Global and the Local: Understanding the Dialectics of Business Systems. Oxford: Oxford University Press, 2005. Pp. viii+306. £55.00 hardback. ISBN 0 19 927890 3. Stabile (Donald R.). Forerunners of Modern Financial Economics: A Random Walk in the History of Economic Thought, 1900‐1950. Aldershot and Lyme, NH: Edward Elgar, 2005. Pp. viii+173. £49.95 hardback. ISBN 1 84542 101 9. Steinberg (Richard). The Economics of Nonprofit Enterprises. Aldershot and Lyme, NH: Edward Elgar, 2004. Pp. xxx+549. £135.00 hardback. ISBN 1 84376 036 3. Stern (Nicholas), Dethier (Jean‐Jacques) and Rogers (F. Halsey). Growth and Empowerment: Making Development Happen. Massachusetts Institute of Technology: The MIT Press, 2005. Pp. xxxiv+463. £29.95 hardback. ISBN 0 262 19517 8. Stern (Scott). Biological Resource Centers: Knowledge Hubs for the Life Sciences. Washington, DC: Brookings Institution Press, 2004. Pp. vii+128. £18.00 paperback, US $24.95 paperback. ISBN 0 8157 8149 0. Streeck (Wolfgang) and Thelen (Kathleen) (Eds). Beyond Continuity: Institutional Change in Advanced Political Economies. Oxford: Oxford University Press, 2005. Pp. xviii+290. £19.99 paperback. ISBN 0 19 928046 0. Sun (Chia‐Hung). The Growth Process in East Asian Manufacturing Industries: A Re‐examination. Aldershot and Lyme, NH: Edward Elgar, 2004. Pp. xiii+222. £55.00 hardback. ISBN 1 84376 775 9. Szymanski (Stefan) and Zimbalist (Andrew). National Pastime: How Americans Play Baseball and the Rest of the World Plays Soccer. Washington, DC: Brookings Institution Press, 2005. Pp. xii+263. £15.99 hardback. ISBN 0 8157 8258 6. TheWorldBank. World Development Report 2004: Making Services Work for Poor People. Washington, DC: World Bank, 2003. Pp. iv+26. ISBN 0 8213 5637 2. Toniolo (Gianni) and Clement (Piet). Central Bank Cooperation at the Bank for International Settlements, 1930–1973. Cambridge and New York: Cambridge University Press, 2005. Pp. xxii+729. £75.00 hardback, US $125.00 hardback. ISBN 0 521 84551 3. Tsoukalis (Loukas) (Ed). What Kind of Europe. Oxford: Oxford University Press, 2005. Pp. xii+248. £14.99 paperback. ISBN 0 19 927948 9. Tullock (Gordon). Public Goods, Redistribution and Rent Seeking. Aldershot and Lyme, NH: Edward Elgar, 2005. Pp. vii+153. £45.00 hardback. ISBN 1 84376 637 X. Unido. International Yearbook of Industrial Statistics 2005. Vienna: Unido, 2005. Pp. vi+566. £150.00 hardback. ISBN 1 84542 201 5. UnitedNations (Ed). International Investment Instruments: A Compendium: Volume xi–2003, xii–2003, xiii–2005, xiv–2005. New York: United Nations, 2005. Pp. xxxxvi+1254. ISBN 92 1 112616 9. Velupillai (K. Vela) (Ed). Computability, Complexity and Constructivity in Economic Analysis. Oxford, Cambridge, MA and Victoria, Australia: Blackwell Publishing, 2005. Pp. viii+324. £24.99 paperback. ISBN 1 4051 3078 4. Vroman (Wayne) and Brusentsev (Vera). Unemployment Compensation Throughout the World: A Comparative Analysis. Kalamazoo, Michigan: W.E. Upjohn Institute, 2005. Pp. ix+273. US $41.00 hardback, US $20.00 paperback. ISBN 0 88099 323 5, 0 88099 322 7. Watson (Robert) (Ed). Governance and Ownership. Aldershot and Lyme, NH: Edward Elgar, 2005. Pp. xxii+553. £140.00 hardback. ISBN 1 84376 831 3. Wei (YingqiAnnie) and Balasubramanyam (V. N.) (Eds). Foreign Direct Investment: Six Country Case Studies. Aldershot and Lyme, NH: Edward Elgar, 2004. Pp. xvii+218. £49.95 hardback. ISBN 1 84376 467 9. Weigand (Jurgen) and Neumann (Manfred) (Eds). The International Handbook of Competition; Aldershot and Lyme, NH: Edward Elgar, 2004. Pp. viii+352. £95.00 hardback. ISBN 1 84376 054 1. Weisskoff (Rishard). The Economics of Everglades Restoration: Missing Pieces in the Future of South Florida. Aldershot and Lyme, NH: Edward Elgar, 2005. Pp. xii+345. £65.00 hardback. ISBN 1 84376 224 2. Wijen (Frank), Zoeteman (Kees) and Pieters (Jan) (Eds). A Handbook of Globalisation and Environmental Policy: National Government Interventions in a Global Arena. Aldershot and Lyme, NH: Edward Elgar, 2005. Pp. xxi+734. £150.00 hardback. ISBN 1 84376 913 1. Wilkinson (Nick). Managerial Economics: A Problem‐Solving Approach. Cambridge and New York: Cambridge University Press, 2005. Pp. xxi+533. £75.00 hardback, US $120.00 hardback, £33.00 paperback, US $60.00 paperback. ISBN 0 521 81993 8, 0 521 52625 6. Winter (Harold). Trade‐Offs: An Introduction to Economic Reasoning and Social Issues. Chicago: University of Chicago Press, 2005. Pp. xiv+126. £10.00 hardback, US $14.00 hardback. ISBN 0 226 90225 0. Witztum (Amos). Economics: An Analytical Introduction. Oxford: Oxford University Press, 2005. Pp. xiv+554. £33.99 paperback. ISBN 0 19 927 163 1. Wray (L. Randall) and Forstater (Mathew) (Eds). Contemporary Post Keynesian Analysis. Aldershot and Lyme, NH: Edward Elgar, 2004. Pp. x+356. £79.95 hardback. ISBN 1 84376 460 1. © Royal Economic Society 2005.

Dustmann,, Christian;Fabbri,, Francesca;Preston,, Ian

doi: 10.1111/j.1468-0297.2005.01038.xpmid: N/A

Abstract Using data from the British Labour Force Survey this article provides an empirical investigation of the way immigration affects labour market outcomes of native born workers in Britain, set beside a theoretical discussion of the underlying economic mechanisms. We discuss problems arising in empirical estimation, and how to address them. We show that the overall skill distribution of immigrants is remarkably similar to that of the native born workforce. We find no strong evidence that immigration has overall effects on aggregate employment, participation, unemployment and wages but some differences according to education. The possible negative effects of immigration on wages and employment outcomes of native workers is one of the core concerns in the public debate on immigration. Economic theory is well suited to help understand the possible consequences of immigration for receiving economies, and the theoretical aspects of the possible effects of immigration for the receiving economies’ labour markets are well understood. That is not to say that predictions of theory are clear‐cut, however. It is compatible with economic models that changes in the size or composition of the labour force resulting from immigration could harm the labour market prospects of some native workers; however, it is likewise compatible with theory that immigration even when changing the skill composition of the workforce has no effects on wages and employment of native workers, at least in the long run. Economic models predict that labour market effects of immigration depend most importantly on the structure of the receiving economy, as well as the skill mix of the immigrants, relative to the resident population. Without empirical test, predictions of theoretical models remain at best well‐reasoned speculation, and are not suited to guide policy. To quantify the effects of immigration on wages and employment of resident workers is therefore a main concern of economic analysis. A considerable number of papers address this issue, most of them for the US, with some studies for other European countries.1 Most papers find effects of immigration on wages and employment prospects of native workers which are either modest or absent. However, the general conclusion some draw from this evidence, that immigration has, at most, modest adverse effects on employment and wages, is not undisputed, and there is an ongoing debate about measurement and identification (Borjas, 2003). While there are many empirical studies for the US, and some work for other European countries, no analysis exists for Britain. Given the difference in recent migration history, settlement, and type of immigrants to Britain, it would be wrong to infer from other studies the possible effects of immigration on the British labour market. One purpose of this article is to fill the gap in evidence for Britain. We commence by pointing out the circumstances under which we should expect immigration to have an effect on labour market outcomes of native workers, and the circumstances under which such effects may not be expected. We then describe our empirical strategy. Our empirical model is directly derived from the theoretical work and allows therefore a straightforward interpretation of parameters within the framework set out by the theory. The dominant methodology in the literature, which we follow in this article also, is to seek to infer labour market effects from spatial correlations between local immigrant inflows and local changes in the labour market outcomes of natives. At the stage of empirical implementation, this methodology raises a number of important issues. Most of these relate to a clear isolation of the effect of immigration on native labour market outcomes from other associated phenomena, particularly in a context where immigrant inflows are themselves the outcome of economic decisions. We shall discuss the appropriate empirical strategies to solve these problems, and implement them as far as our data allows us to do so. For our analysis, we use data from the British Labour Force Survey (LFS). We commence in the next Section with a brief account of the relevant economic theory that underlies the subsequent empirical work, and a discussion of the problems which occur on the empirical level. Next we describe the skill distribution of immigrants to Britain. We then explain the data sources we use and report results of our empirical analysis. Finally, we draw conclusions and suggest avenues for future work. 1. Theory The theoretical analysis of the labour market effects of immigration sees effects as arising from the changes it introduces in supply of skills and consequent change in labour market equilibrium. Typically a distinction is drawn between skilled and unskilled labour. Immigration inflows affect the skill composition of the labour force if the skill composition of immigrants does not match the skill composition of natives. This change in skill composition leads to disequilibrium between supply of and cost‐minimising demand for different labour types at existing wages and output levels. Restoration of equilibrium should be expected therefore to involve short‐run changes in wages and employment levels of different skill types and may or may not require long‐run changes as we allow the economy's output mix also to adjust.2 The literature includes different approaches to theoretical modelling of these processes, implying different conclusions about the nature of long‐run effects. The main differences in assumptions made involve (i) differences in the number of goods produced and therefore in the flexibility of the economy to adapt through changes in mix of outputs, and (ii) differences in openness of the goods sector to trade and therefore in the extent to which output prices are set locally or on world markets. Models assuming limited flexibility of output mix or closedness to international trade tend to predict that immigration will have long‐run wage and employment effects. Such features are typical of the underlying framework used as a motivation for empirical work in this literature; see, for example, the models of Borjas (1999) or Card (2001).3 On the other hand, models assuming a sufficiently high degree of flexibility in the mix of output produced in the traded goods sector predict an absence of long‐run effects on labour market outcomes, at least to small scale immigration. For illustration, consider first the effects of migration into an economy which produces, with a constant returns to scale technology, one output good only, sold at a price set on world markets, and using three factors of production: capital, skilled labour, and unskilled labour.4 Assume also that capital supply is perfectly elastic (which would be the case if the rate of return to capital is set on world markets) and labour supply of both skill groups is completely inelastic. Finally, assume that the skill composition of immigrants differs from that of native workers, and consider for illustration the case where all immigrants are low skilled. In this case, immigration will lead to a decrease in wages of low skilled native workers as the economy moves down the marginal product of labour curve for unskilled workers. If the change is more than marginal, then the immigrants are paid less than their average product and the owners of other factors enjoy a surplus from immigration. Since the return to capital is fixed this surplus accrues to skilled workers whose wages rise while those of native unskilled workers fall. There is therefore an aggregate gain but also redistribution with one labour type losing while the other gains. More generally, in such an economy, and if immigrants differ in their skill composition from natives, per capita income of the native population will increase as a consequence of migration, but the gains of migration are unequally distributed. Supposing now that labour supply is actually somewhat elastic, immigration may also cause (voluntary) unemployment among those natives whose wages fall and who choose therefore to withdraw from supplying labour. Finally, notice that any wage effects are a consequence of immigration changing the skill structure of the workforce. No effects are to be expected if immigrants resemble resident workers in their skill composition. Below we will investigate this empirically for Britain. More generally, the lack of flexibility in an economy with a homogeneous traded goods sector means that there are insufficient degrees of freedom to accommodate changes in the skill mix through changes in the output mix. Wage changes are therefore not zero even in the long run. Now contrast these conclusions with those appropriate to an economy with a heterogeneous traded goods sector in which output prices are fixed on world markets (and which, therefore, has relatively high flexibility in the output mix of traded goods). Assume again that labour supply is inelastic and that migration is unskilled. Holding outputs fixed, immigration would, as before, drive down wages of unskilled workers (and increase wages of skilled labour). This however drives up profits in a sector which uses unskilled labour more intensively and should therefore lead to a relative expansion of production in that sector, pushing up demand for unskilled labour and hence unskilled wages. Assuming the eventual equilibrium continues to involve positive production in all traded goods sectors, wages should return to the initial pre‐immigration equilibrium.5 Rather than impacting on wages, long‐run effects of immigration are felt in the output mix with production of the good using unskilled labour relatively intensively expanding according to predictions of the Rybczinski (1955) theorem. Leamer and Levinsohn (1995) refer to this as the hypothesis of factor price insensitivity.6 This possible adjustment mechanism is sometimes mentioned in studies on the labour market impact of immigration; see, for example, Chiswick (1993), Borjas (1999), Card (2001), Friedberg and Hunt (1995) or Pischke and Velling (1997). Several recent contributions lay more stress on the need for models with multiple goods and openness to trade; see, for example, Kuhn and Wooton (1991), Scheve and Slaughter (2001), Hanson and Slaughter (1999; 2001), Gaston and Nelson (2000; 2001). If labour supply is elastic, there may be both employment and wage effects in the short run, before output mix can fully adjust. Again, as in the one output case, no effects of migration on wages and employment are to be expected if the composition of migrant labour resembles that of the resident pre‐migration population. This exposition shows that a variety of possible outcomes are compatible with economic theory. Immigration may depress wages and employment of natives. However, it is by no means inconsistent with economic theory to think that long run responses to immigration may involve no effect, or that immigration increases wages of workers complementary to immigrant labour. As for the long‐run effects, what matters is the openness of the economy to trade and the flexibility of the economy to adjust in respects other than wages and in particular through the mix of output produced.7 2. Empirical Implementation The dominant approach to estimation of such a model in the literature is that referred to by Borjas (1999) as the ‘spatial correlations’ approach. Effects of immigration are identified from the spatial correlation between immigrant labour inflows and changes in native or overall labour market outcomes (or between immigrant population shares and levels of these outcomes). Spatial units are intended to correspond to geographical labour markets. The key problem in empirical analysis is to compare the economic outcomes of certain groups of the resident population in particular cells after immigration with the counterfactual outcomes that would be observed had migration not taken place. While the first measure is observable, the second is not, and needs to be constructed. Construction of this counterfactual involves always assumptions which are debatable. The thought experiment in developing an empirical strategy based on local labour market variation in immigrant populations is that immigrants are randomly allocated across local labour markets, and that variation in economic outcomes is related to variation in immigrant densities. Problems arise with this strategy because levels of immigrant shares and levels of labour market outcomes may be spatially correlated because of common fixed influences, leading to a positive or negative statistical correlation between immigrant concentration and economic outcomes, even in the absence of any genuine effects of immigration. Elimination of common fixed influences could be achieved by using changes in economic outcomes and relating them to changes in immigrant concentrations. However, the direction of causality between immigrant inflows and labour market outcomes is not necessarily clear‐cut. Immigrants may be attracted to those areas that are enjoying current economic success. This selective settlement would lead to an upwardly biased estimate of the effects of immigrants’ concentration on labour market outcomes of natives. A possible solution to this problem is to use measures of historic settlement patterns as instruments for immigrant inflows. The underlying justification is that immigrants will be attracted to settle where there are existing networks and the presence of individuals with the same cultural and linguistic background as themselves, inducing immigrants to settle in areas with already high immigrant concentrations. Pre‐existing immigrant concentrations are unlikely to be correlated with current economic shocks if measured with a sufficient time lag, since existing concentrations are determined not by current economic conditions, but by historic settlement patterns of previous immigrants.8 The assumption that lagged values of immigrant stocks are correlated with employment changes only through their relation with immigrant inflows is an identifying assumption that is not testable. It could be problematic if local economic shocks were persistent and instruments were insufficiently lagged. The strength of correlation between lagged concentrations and current inflows is observable in data and can therefore be assessed. A further problem with studies that rely on relatively small sample sizes to compute immigrant concentrations and economic conditions on local level is measurement error. This is likely to be the case in an analysis that is based on a survey of relatively small sample size. The consequences of any measurement error is aggravated when using differenced or within groups estimation. This problem is addressed by instrumental variable estimation as long as the dependence on the regressor is linear, as it is in our case. The identification strategy we point out above should address measurement error as well. Local labour markets are not closed economies and native workers are free to move out. If immigration does drive down local wages for certain skill groups then one would expect there to be pressure for native workers of that skill type to move elsewhere. This will tend to disperse the impact of immigration through the national economy and undermine the ability to identify the impact from looking at effects within localities, leading to downward biased estimates of the effect of immigration on, e.g., employment of native workers. This point has been stressed in numerous contributions. The US literature contains conflicting opinions on the seriousness of the problem. Borjas (2003) regards it as more serious than Card (2001, this issue). For Britain, there is some evidence that mobility is, in general, low. Gregg et al. (2004) show that mobility amongst low skill/education people is limited, and often constrained by the housing market. Hatton and Tani (this Feature) use data from the International Passenger Survey and the National Health Service Registration Data to quantify the relationship between net inflows from abroad and the flows of residents within Britain. Their findings suggest a negative correlation between immigration to one region from abroad and in‐migration from other British regions, which is significant for the South‐East. On the level of estimation, the problem is one of an omitted term in the estimated equation. The most attractive resolution to this problem is available if native outflows are observable and therefore amenable to incorporation directly into the estimation – a strategy we follow below. However such outflows are likely to be correlated with shocks to local economic conditions for the same reasons as immigrant flows, discussed above, creating a further simultaneity issue. These outflows therefore also need instrumenting and it is theoretically less clear what would serve as a suitable instrument. In practice we rely on lags. 3. Immigrant and Native Skill Compositions In our discussion of the underlying theoretical model we emphasise that there are no effects of immigration to be expected on labour market outcomes of residents if immigration does not affect the skill composition of the resident labour force, and if capital supply is perfectly elastic. In the US, migration over recent decades has been predominantly unskilled (Borjas, 1999; Card and Lewis, 2005). The situation for Britain is different, however, with immigrants being more similar in their education and skill distribution to the resident population. We illustrate this by computing the percentage of native born workers, immigrants, and recent immigrants as of 2000 in three different education categories. Numbers are based on the LFS, which we describe in more detail below. We define recent immigrants as individuals who entered Britain over the last decade (between 1991 and 2000). Low education refers to no formal qualification; intermediate education to O‐levels (or equivalent); and advanced education to A‐levels or college/university degrees. We provide mean percentages of individuals in each of the three education groups in Table 1. While the percentage of native born workers in the highest education category is higher than the percentage of both immigrants and recent immigrants, the latter two groups are higher in the intermediate education category. For the unqualified, the percentages of immigrants and natives are fairly similar, while the percentage of recent immigrants is slightly lower. Table 1 Educational and Occupational Distribution, Immigrants and Natives Education . Advanced education . Intermediate education . Low education . Natives 0.509 0.318 0.172 Immigrants 0.423 0.393 0.183 Recent immigrants 0.304 0.551 0.145 Occupation Skilled Semiskilled Unskilled Natives 0.246 0.397 0.356 Immigrants 0.313 0.361 0.326 Recent immigrants 0.312 0.363 0.324 Education . Advanced education . Intermediate education . Low education . Natives 0.509 0.318 0.172 Immigrants 0.423 0.393 0.183 Recent immigrants 0.304 0.551 0.145 Occupation Skilled Semiskilled Unskilled Natives 0.246 0.397 0.356 Immigrants 0.313 0.361 0.326 Recent immigrants 0.312 0.363 0.324 Source: British Labour Force Survey 2000. Open in new tab Table 1 Educational and Occupational Distribution, Immigrants and Natives Education . Advanced education . Intermediate education . Low education . Natives 0.509 0.318 0.172 Immigrants 0.423 0.393 0.183 Recent immigrants 0.304 0.551 0.145 Occupation Skilled Semiskilled Unskilled Natives 0.246 0.397 0.356 Immigrants 0.313 0.361 0.326 Recent immigrants 0.312 0.363 0.324 Education . Advanced education . Intermediate education . Low education . Natives 0.509 0.318 0.172 Immigrants 0.423 0.393 0.183 Recent immigrants 0.304 0.551 0.145 Occupation Skilled Semiskilled Unskilled Natives 0.246 0.397 0.356 Immigrants 0.313 0.361 0.326 Recent immigrants 0.312 0.363 0.324 Source: British Labour Force Survey 2000. Open in new tab These figures suggest that immigrants to Britain are fairly similar in their educational background to native born workers, at least on the national level. An alternative measure for the distribution of immigrants across labour market skill groups is their observed occupational distribution.9 Again using data from the 2000 LFS, we have ranked 17 occupational groups by their mean earnings. We have then split the sample into three groups, which we refer to as ‘skilled’, ‘semi‐skilled’ and ‘unskilled’. The category ‘skilled’ includes the professions with the highest hourly wages: employers and managers, professional workers, employees with the armed forces. The category ‘semiskilled’ includes intermediate non‐manual workers, junior non‐manual workers, and foreman and supervisors. Finally, the category ‘unskilled’ includes farmers and farm workers, manual workers and personal service workers. The numbers in the second panel of the Table show a remarkable similarity in the skill distribution across the three groups of natives, immigrants and recent immigrants. The similarity between groups is stronger than for the educational classification. These results suggest that immigrants to Britain have a similar skill distribution to the native workforce. Based on these figures, we may conclude that at the national level there is no evidence that past or more recent immigration led to an increase of the ratio of unskilled to semiskilled or skilled workers. However, this does not imply that the skill distribution of immigrants across local labour markets is likewise similar to that of the native population; any conclusion that we should not expect labour market effects cannot be drawn on the basis of this evidence. 4. Data and Descriptives The data set we use for our analysis is the British Labour Force Survey (LFS). The LFS is a household survey, conducted by the Office for National Statistics (ONS) on behalf of the Department for Education and Employment (DfEE). It provides a wide range of data on labour market statistics and related topics such as training, qualifications, income and disability. The LFS has been carried out in Britain since 1973. Between 1973 and 1983 it has been on a biennial basis, changing into an annul survey from 1983 onwards. The sample size is about 60,000 households in each survey, or around 0.5% of the population. From 1992 onwards, the survey changed to a rotating quarterly panel, with the same individuals being interviewed for five consecutive quarters. Each quarter about 59,000 households are interviewed with about 138,000 respondents. The quarterly LFS contains information on gross weekly wages and number of hours worked for the fifth quarter wave (1992–6) or the first and the fifth quarter (1997 onwards). The British LFS contains spatial information only at regional level, except for a brief interval between 1997 and 1999 when data was made available at county level. We therefore aggregate data at regional level, according to the region definition available in the LFS.10 For consistency across the years, we use 17 regions, namely, Tyne and Wear, Rest of Northern Region, South Yorkshire, West Yorkshire, Rest of Yorkshire and Humberside, East Midlands, East Anglia, Greater London, Rest of the South East, South West, West Midlands, Rest of West Midlands, Greater Manchester, Merseyside, Rest of North West, Wales and Scotland. In our empirical analysis, we first focus on employment, defined as the proportion of the working age population employed. Below we will also present results for participation (the proportion of the working age population employed or looking for work), and unemployment (the proportion of those active in the labour market who are not employed). For these analyses we use data from the LFS from 1983 onwards.11 Wage information became available only in 1992 and we use data from 1992 until 2000 for the wage analysis. We use the log of (gross) hourly wages for the working population and for the skill subgroups.12 In Table 2 we present some summary statistics for the data we use for employment, unemployment and participation analysis. The summary statistics are based on our aggregate data at regional level. The time interval is from 1983 to 2000. The analysis for wages considers a shorter period, and means are displayed in the second panel. In the third panel we report the means of the regressors used in our analysis, based on the sample used for employment analysis. Table 2 Descriptive Statistics, LFS 1983–2000 Variable . Mean . Standard deviation . Employment Total 0.705 0.050 Advanced education 0.808 0.033 Intermediate education 0.712 0.043 Unqualified 0.558 0.075 Unemployment Total 0.095 0.033 Advanced education 0.064 0.022 Intermediate education 0.100 0.036 Unqualified 0.147 0.044 Participation Total 0.779 0.032 Advanced education 0.863 0.020 Intermediate education 0.791 0.026 Unqualified 0.652 0.068 Wages (1992–2000) Total 2.076 0.128 Advanced education 2.237 0.122 Intermediate education 1.817 0.144 Unqualified 1.699 0.134 Immigrant‐native ratio 0.072 0.079 ln advanced/unqualified 0.311 0.572 ln intermediate/unqualified 0.052 0.452 Mean native age/100 0.377 0.010 Mean immigrant age/100 0.386 0.018 No. Obs. 306 Variable . Mean . Standard deviation . Employment Total 0.705 0.050 Advanced education 0.808 0.033 Intermediate education 0.712 0.043 Unqualified 0.558 0.075 Unemployment Total 0.095 0.033 Advanced education 0.064 0.022 Intermediate education 0.100 0.036 Unqualified 0.147 0.044 Participation Total 0.779 0.032 Advanced education 0.863 0.020 Intermediate education 0.791 0.026 Unqualified 0.652 0.068 Wages (1992–2000) Total 2.076 0.128 Advanced education 2.237 0.122 Intermediate education 1.817 0.144 Unqualified 1.699 0.134 Immigrant‐native ratio 0.072 0.079 ln advanced/unqualified 0.311 0.572 ln intermediate/unqualified 0.052 0.452 Mean native age/100 0.377 0.010 Mean immigrant age/100 0.386 0.018 No. Obs. 306 Open in new tab Table 2 Descriptive Statistics, LFS 1983–2000 Variable . Mean . Standard deviation . Employment Total 0.705 0.050 Advanced education 0.808 0.033 Intermediate education 0.712 0.043 Unqualified 0.558 0.075 Unemployment Total 0.095 0.033 Advanced education 0.064 0.022 Intermediate education 0.100 0.036 Unqualified 0.147 0.044 Participation Total 0.779 0.032 Advanced education 0.863 0.020 Intermediate education 0.791 0.026 Unqualified 0.652 0.068 Wages (1992–2000) Total 2.076 0.128 Advanced education 2.237 0.122 Intermediate education 1.817 0.144 Unqualified 1.699 0.134 Immigrant‐native ratio 0.072 0.079 ln advanced/unqualified 0.311 0.572 ln intermediate/unqualified 0.052 0.452 Mean native age/100 0.377 0.010 Mean immigrant age/100 0.386 0.018 No. Obs. 306 Variable . Mean . Standard deviation . Employment Total 0.705 0.050 Advanced education 0.808 0.033 Intermediate education 0.712 0.043 Unqualified 0.558 0.075 Unemployment Total 0.095 0.033 Advanced education 0.064 0.022 Intermediate education 0.100 0.036 Unqualified 0.147 0.044 Participation Total 0.779 0.032 Advanced education 0.863 0.020 Intermediate education 0.791 0.026 Unqualified 0.652 0.068 Wages (1992–2000) Total 2.076 0.128 Advanced education 2.237 0.122 Intermediate education 1.817 0.144 Unqualified 1.699 0.134 Immigrant‐native ratio 0.072 0.079 ln advanced/unqualified 0.311 0.572 ln intermediate/unqualified 0.052 0.452 Mean native age/100 0.377 0.010 Mean immigrant age/100 0.386 0.018 No. Obs. 306 Open in new tab Employment is higher for the better educated, as well as for males, with an average employment rate of about 81%. Unemployment and participation varies in a similar manner across education groups, with those who are better educated having a stronger labour market attachment as well as lower rates of unemployment. Wages are, as expected, considerably higher for those with an advanced education. In the last panel we display means of variables we use as regressors in our analysis. As was mentioned above, survey data may be characterised by small sample sizes when analysing specific groups in the population (like immigrants, in particular when breaking them down by education group, gender, or other demographic characteristics). This is due to the fact that immigrants represent a small fraction of the population (about 7.2% across regions and years, as shown in Table 2, and 9% in LFS 2000), and that their geographical distribution in Britain appears to be very uneven (about 60% of immigrants of working age are concentrated in the Greater London and South East regions, against 29% of nonimmigrants). 5. Estimation Strategy The model we derive in the Appendix suggests a relationship between labour market outcomes and the share of immigrants in the labour market. The estimation specification we adopt follows directly from (5), (6), (7) and (8) in the Appendix. Wages and labour supply measures are related to immigrant population share π and to relative sizes of native skill groups, with additional controls for age composition of the population: (1) where Oit denotes the economic outcome under consideration (we consider employment, participation, and wages), πit denotes the ratio of immigrant to native population, nit denotes a vector of native skill group populations and ait denotes a vector of average ages, all in the ith region in the tth period. Here and are year and region effects and is a disturbance term. Homogeneity is imposed on the native skill group effects by omitting one skill category and expressing the others as ratios with the size of the omitted skill group. We report results using the OLS estimator, a difference estimator, and the IV estimator in differences.13 With OLS, the effect of immigration on economic outcomes is identified from the period‐by‐period cross sectional correlation between relative immigrant stocks and employment and wage levels. This offers a basic and straightforward point of comparison. Estimating the relationship in differences removes the influence of the fixed effects . Identification of the effect is now from changes over time in the pattern of cross sectional variation. Although more robust than simple OLS, it still has problems with measurement error and simultaneity. Combining estimation in differences with use of instrumental variables addresses both the issues of measurement error and simultaneity. These final estimates are calculated by GMM imposing the moment restrictions that are uncorrelated with the chosen instruments, which in each case are three and four‐period lags of the endogenous regressors πit and nit. Weighting of restrictions and calculation of standard errors recognises the anticipated first order serial correlation in the differenced residuals. In all estimated specifications we include a full set of year effects so that aggregate time series variation is completely absorbed. We also include controls for average age of immigrants and natives. These are taken as given in subsequent discussion. Size of native skill groups are also entered as controls in order to allow for the effect of native outflows.14 Tests are reported for first and second order serial correlation of residuals and for the overidentifying restrictions implied by the choice of instruments. For all IV estimates reported below there is clear evidence of first order serial correlation, as should be expected given differencing of the residuals but absence of second order serial correlation cannot be rejected at usual significance levels. The overidentifying restrictions are rejected in none of the specifications reported. 6. Results Table 3 presents a series of different estimates of effects on total native employment. OLS regression shows a slight positive relationship between employment and the immigrant–native population ratio. Removing persistent correlated effects by differencing switches the sign of the relationship, indicating that immigrants tend to be in areas with favourable employment conditions. Immigration is now associated with a decrease in employment. For the final and most robust of these estimates (columns 3), the hypothesis of no effect can not be rejected. An increase in immigration amounting to one per cent of the native population would lead, according to this result, to a decrease of 0.07 percentage points in the native employment rate but this estimated effect is far from significantly different from zero at conventional levels. Table 3 Effect of Immigration on Employment LFS 1983–2000 . OLS . Differences . IV . Variable Coeff StdE Coeff StdE Coeff StdE Immigrant–native ratio 0.088 0.040 −0.154 0.083 −0.070 0.096 ln advanced/unqualified 0.090 0.012 0.048 0.014 0.034 0.068 ln intermediate/unqualified 0.081 0.014 0.006 0.013 0.057 0.043 Mean native age/100 1.933 0.480 0.170 0.255 0.186 0.391 Mean immigrant age/100 0.198 0.101 −0.007 0.056 0.003 0.062 M1 13.805 p = 0.000 −4.059 p = 0.000 −3.256 p = 0.001 M2 12.890 p = 0.000 −1.383 p = 0.167 1.283 p = 0.200 W1  = 411.023 p = 0.000  = 24.998 p = 0.000  = 10.014 p = 0.075 W2  = 254.827 p = 0.000  = 426.004 p = 0.000  = 337.295 p = 0.000 S  = 0.451 p = 0.930 Sample size 306 289 238 . OLS . Differences . IV . Variable Coeff StdE Coeff StdE Coeff StdE Immigrant–native ratio 0.088 0.040 −0.154 0.083 −0.070 0.096 ln advanced/unqualified 0.090 0.012 0.048 0.014 0.034 0.068 ln intermediate/unqualified 0.081 0.014 0.006 0.013 0.057 0.043 Mean native age/100 1.933 0.480 0.170 0.255 0.186 0.391 Mean immigrant age/100 0.198 0.101 −0.007 0.056 0.003 0.062 M1 13.805 p = 0.000 −4.059 p = 0.000 −3.256 p = 0.001 M2 12.890 p = 0.000 −1.383 p = 0.167 1.283 p = 0.200 W1  = 411.023 p = 0.000  = 24.998 p = 0.000  = 10.014 p = 0.075 W2  = 254.827 p = 0.000  = 426.004 p = 0.000  = 337.295 p = 0.000 S  = 0.451 p = 0.930 Sample size 306 289 238 Notes: 
All regressions include full set of time dummies. M1 is a test for first‐order serial correlation, asymptotically distributed as a standard normal Instruments: 3 and 4‐period lags of πit and nit M2 is a test for second‐order serial correlation, asymptotically distributed as a standard normal W1 is a Wald test for joint significance of the reported regressors W2 is a Wald test for joint significance of the unreported time dummies S is a χ2 test of the overidentifying restrictions implied by choice of instruments underlying IV estimates Open in new tab Table 3 Effect of Immigration on Employment LFS 1983–2000 . OLS . Differences . IV . Variable Coeff StdE Coeff StdE Coeff StdE Immigrant–native ratio 0.088 0.040 −0.154 0.083 −0.070 0.096 ln advanced/unqualified 0.090 0.012 0.048 0.014 0.034 0.068 ln intermediate/unqualified 0.081 0.014 0.006 0.013 0.057 0.043 Mean native age/100 1.933 0.480 0.170 0.255 0.186 0.391 Mean immigrant age/100 0.198 0.101 −0.007 0.056 0.003 0.062 M1 13.805 p = 0.000 −4.059 p = 0.000 −3.256 p = 0.001 M2 12.890 p = 0.000 −1.383 p = 0.167 1.283 p = 0.200 W1  = 411.023 p = 0.000  = 24.998 p = 0.000  = 10.014 p = 0.075 W2  = 254.827 p = 0.000  = 426.004 p = 0.000  = 337.295 p = 0.000 S  = 0.451 p = 0.930 Sample size 306 289 238 . OLS . Differences . IV . Variable Coeff StdE Coeff StdE Coeff StdE Immigrant–native ratio 0.088 0.040 −0.154 0.083 −0.070 0.096 ln advanced/unqualified 0.090 0.012 0.048 0.014 0.034 0.068 ln intermediate/unqualified 0.081 0.014 0.006 0.013 0.057 0.043 Mean native age/100 1.933 0.480 0.170 0.255 0.186 0.391 Mean immigrant age/100 0.198 0.101 −0.007 0.056 0.003 0.062 M1 13.805 p = 0.000 −4.059 p = 0.000 −3.256 p = 0.001 M2 12.890 p = 0.000 −1.383 p = 0.167 1.283 p = 0.200 W1  = 411.023 p = 0.000  = 24.998 p = 0.000  = 10.014 p = 0.075 W2  = 254.827 p = 0.000  = 426.004 p = 0.000  = 337.295 p = 0.000 S  = 0.451 p = 0.930 Sample size 306 289 238 Notes: 
All regressions include full set of time dummies. M1 is a test for first‐order serial correlation, asymptotically distributed as a standard normal Instruments: 3 and 4‐period lags of πit and nit M2 is a test for second‐order serial correlation, asymptotically distributed as a standard normal W1 is a Wald test for joint significance of the reported regressors W2 is a Wald test for joint significance of the unreported time dummies S is a χ2 test of the overidentifying restrictions implied by choice of instruments underlying IV estimates Open in new tab A Sargan test of the overidentifying restrictions is comfortably passed (as it is in all specifications which we estimate for the article) and the evidence of Table 4 suggests that instruments do predict the endogenous regressors well, particularly the changes in immigrant–native ratio.15 Table 4 Significance of Excluded Instruments in First Stage Regression LFS 1983–2000 Variable . Wald test . Immigrant–native ratio  = 254.946 p = 0.000 ln advanced/unqualified  = 11.863 p = 0.065 ln intermediate/unqualified  = 32.374 p = 0.000 Variable . Wald test . Immigrant–native ratio  = 254.946 p = 0.000 ln advanced/unqualified  = 11.863 p = 0.065 ln intermediate/unqualified  = 32.374 p = 0.000 Open in new tab Table 4 Significance of Excluded Instruments in First Stage Regression LFS 1983–2000 Variable . Wald test . Immigrant–native ratio  = 254.946 p = 0.000 ln advanced/unqualified  = 11.863 p = 0.065 ln intermediate/unqualified  = 32.374 p = 0.000 Variable . Wald test . Immigrant–native ratio  = 254.946 p = 0.000 ln advanced/unqualified  = 11.863 p = 0.065 ln intermediate/unqualified  = 32.374 p = 0.000 Open in new tab In Table 5 we report similar series of regressions for other economic outcomes.16 The results in the Table suggest that OLS results suggest an overly optimistic effect of immigration on the various economic outcomes – similar to results on employment. Again, persistence in economic conditions and immigrant concentrations explains these results. Eliminating this factor by estimating differences changes the sign of the relationship for both unemployment and participation, suggesting a positive association between immigration and unemployment and a negative association between immigration and participation. Coefficients are however not significantly different from zero. The relationship between wages and immigration remains positive. Table 5 Effect of Immigration on Unemployment, Participation and Wages LFS 1983–2000 Variable . OLS . Differences . IV . Coeff . StdE . Coeff . StdE . Coeff . StdE . Unemployment −0.050 0.026 0.106 0.067 0.066 0.103 Participation 0.057 0.028 −0.082 0.071 ‐0.035 0.088 Wages 0.802 0.107 0.198 0.677 0.909 0.583 Variable . OLS . Differences . IV . Coeff . StdE . Coeff . StdE . Coeff . StdE . Unemployment −0.050 0.026 0.106 0.067 0.066 0.103 Participation 0.057 0.028 −0.082 0.071 ‐0.035 0.088 Wages 0.802 0.107 0.198 0.677 0.909 0.583 Notes: Reported coefficients are for immigrant–native ratio. All regressions include full set of time dummies and controls for native skill group sizes and mean native and immigrant ages. Instruments: 3 and 4 period lags of πit and nit For both IV results, the Sargan test of overidentifying restrictions fails to reject at 5% significance level in all specifications. Open in new tab Table 5 Effect of Immigration on Unemployment, Participation and Wages LFS 1983–2000 Variable . OLS . Differences . IV . Coeff . StdE . Coeff . StdE . Coeff . StdE . Unemployment −0.050 0.026 0.106 0.067 0.066 0.103 Participation 0.057 0.028 −0.082 0.071 ‐0.035 0.088 Wages 0.802 0.107 0.198 0.677 0.909 0.583 Variable . OLS . Differences . IV . Coeff . StdE . Coeff . StdE . Coeff . StdE . Unemployment −0.050 0.026 0.106 0.067 0.066 0.103 Participation 0.057 0.028 −0.082 0.071 ‐0.035 0.088 Wages 0.802 0.107 0.198 0.677 0.909 0.583 Notes: Reported coefficients are for immigrant–native ratio. All regressions include full set of time dummies and controls for native skill group sizes and mean native and immigrant ages. Instruments: 3 and 4 period lags of πit and nit For both IV results, the Sargan test of overidentifying restrictions fails to reject at 5% significance level in all specifications. Open in new tab The last column presents IV results. Point estimates decrease slightly and suggest that there is no strong evidence of impact on native unemployment or participation rates. Estimated wage effects are positive and substantial but the preferred IV estimates are again statistically insignificant. The results for wages should be treated with particular caution given the smaller range of years available for estimation. 6.1. Distinguishing Between Different Education and Demographic Groups Table 6 reports separate results for workers in different education groups. Educational classification follows the definitions in Section 3. We only report estimates obtained from the IV estimator. For the employment, unemployment and participation regressions the dependent variable is defined as the numbers employed, unemployed and participating in the group concerned divided by total relevant native population. This has the interpretive advantage that the estimated coefficients (roughly)17 add up to the total effects (because of the common denominator) and therefore provide a breakdown of the total effect across education groups. Table 6 Effect of Immigration by Education Group LFS 1983–2000 . IV, Differences . Advanced . Intermediate . Unqualified . Variable Coeff StdE Coeff StdE Coeff StdE Employment 0.111 0.068 −0.179 0.052 −0.028 0.058 Unemployment 0.001 0.044 0.098 0.043 −0.034 0.075 Participation 0.108 0.061 −0.108 0.050 −0.063 0.073 Wages 0.930 0.990 0.153 1.044 3.798 3.397 . IV, Differences . Advanced . Intermediate . Unqualified . Variable Coeff StdE Coeff StdE Coeff StdE Employment 0.111 0.068 −0.179 0.052 −0.028 0.058 Unemployment 0.001 0.044 0.098 0.043 −0.034 0.075 Participation 0.108 0.061 −0.108 0.050 −0.063 0.073 Wages 0.930 0.990 0.153 1.044 3.798 3.397 Notes: As for Table 5. Open in new tab Table 6 Effect of Immigration by Education Group LFS 1983–2000 . IV, Differences . Advanced . Intermediate . Unqualified . Variable Coeff StdE Coeff StdE Coeff StdE Employment 0.111 0.068 −0.179 0.052 −0.028 0.058 Unemployment 0.001 0.044 0.098 0.043 −0.034 0.075 Participation 0.108 0.061 −0.108 0.050 −0.063 0.073 Wages 0.930 0.990 0.153 1.044 3.798 3.397 . IV, Differences . Advanced . Intermediate . Unqualified . Variable Coeff StdE Coeff StdE Coeff StdE Employment 0.111 0.068 −0.179 0.052 −0.028 0.058 Unemployment 0.001 0.044 0.098 0.043 −0.034 0.075 Participation 0.108 0.061 −0.108 0.050 −0.063 0.073 Wages 0.930 0.990 0.153 1.044 3.798 3.397 Notes: As for Table 5. Open in new tab Estimated employment, unemployment and participation effects are individually statistically significant only for the intermediate education group – those with O‐levels but no higher – for whom the effects consistently suggest a depressive effect on labour market activity and probability of working. Nonetheless this seems to be offset by increasing employment of the more educationally qualified – with the net effect being not significantly different from zero, as our aggregate results above have shown. For the unqualified effects are very weakly determined. Albeit that the effects here are typically not very precisely estimated, the evidence does fit with the fact that immigration appears to have expanded the intermediate education group in particular, as discussed in Section 3. Similar regressions for wages show consistently positive but weakly determined effects, which again are least beneficial for the group with intermediate education. The small sample sizes on immigrants when distinguishing between populations with different characteristics suggest to interpret these results with caution. In none of these specifications have the dynamics of the relationship been explored. We have been unable to find statistically reliable and well determined estimates of dynamic specifications and have therefore refrained from commenting on differences between short‐run and long‐run effects. We note however that considerations of economic theory suggest that long run adjustments to immigration are likely to lower the magnitude of effects and that the estimates here are likely to overestimate long‐run responses. 7. Discussion and Conclusion This article provides a first analysis of the way immigration may affect labour market outcomes of native workers in Britain. We commence by reviewing and discussing the theoretical background. These considerations suggest that the effects immigration may have on the labour market outcomes of resident workers are by no means clear‐cut: they depend most importantly on the way immigration affects the skill mix of the resident population, as well as the way the economy may adjust to changes in the skill mix. These considerations emphasise the risk in drawing conclusions for Britain from analysis of other countries’ labour markets as both the composition of immigrant inflows as well as the adjustment mechanisms differ across countries. Moreover, theoretical considerations like those discussed in this article assume that the labour market is in equilibrium before and after immigration. However, migrations are often a consequence of disequilibrium situations – for instance the large migrations to Europe in the period between 1955 and 1973 were a response to an excess demand for labour; see Dustmann (1996) for details. This is likely to change the results of any empirical investigation – again, to an extent crucially depending on the type and magnitude of initial disequilibrium. The importance of careful consideration of possible differences in migration types is illustrated in our article. Unlike the US or some continental European countries, immigration to Britain is not concentrated at the lower end of the skill distribution but immigrants (recent immigrants as well as the existing immigrant population) resemble quite closely the skill composition of the resident native workforce. This is interesting and has in our view not yet received sufficient attention in the debate about possible effects of immigration. Empirical analysis of the effects of immigration on outcomes of native born workers faces a number of challenges, as we do not directly observe outcomes for native born workers that would have occurred in the absence of immigration. The approach we follow in this article is to use variation in immigration to different spatial areas and to instrument this by variation in historical settlement patterns. Our analysis focuses on a range of labour market outcomes: employment, unemployment, participation and wages. The main result is that we find little evidence of overall adverse effects of immigration on native outcomes. If there is evidence of negative effects on employment in any group, then it is for those with intermediate education levels, but this is offset in the aggregate by positive effects on employment among the better qualified. Estimated wage effects, based on a shorter run of data, are if anything positive but statistically poorly determined. We have drawn attention to many weaknesses in the available data and conceptual problems in the empirical analysis all of which should urge caution before drawing strong conclusions. We consider our investigation as a first step in analysing this important issue for Britain. We have repeatedly hinted at the relatively poor quality of data available for research of this type in Britain. The possibility of accessing a finer regional breakdown in the LFS, might for instance be one step towards an improved analysis. Appendix: Immigration and The Labour Market A1. Labour Market Equilibrium We outline here a simple model of the effect of immigration on the labour market. Let N denote total native population and M total immigrant population. Suppose there are two labour types, skilled and unskilled, earning wages wS and wU. Numbers of workers of the two types are where Ni is total native workforce of the type and Mi is total immigrant workforce of the type. Hence, assuming ratio of immigrant to native population, π = M/N, is small, where βi = (MiN/NiM) is relative skill share of immigrants. Supply of labour is then xili(wi,p), i ∈ I, where xi is number of workers of the ith type and li(wi,p) is a labour supply function. Capital is assumed elastically supplied at a return to capital, r, which is fixed on world markets. We consider two cases differing in the number of goods produced by the economy. Either the economy produces one good in quantity y0 or two goods in quantities y0 and y1. We denote the set of goods by J which therefore equals {0} or {0,1}. These goods are assumed traded and the economy small so that their prices p0 and p1 are therefore set on world markets.18 Assuming constant returns to scale and excluding the possibility of joint production, we write the unit cost function for the jth output as cj(wS,wU,r), j ∈ J. Letting denote the derivative ∂cj/∂wi, demand for the ith type of labour is therefore by Shephard's lemma. Wages and outputs are determined by two equilibrium conditions. Firstly, labour market equilibrium requires equality of demand and supply of labour ie (2) and secondly, firms earn zero profits and therefore (3) A2. One Output Good Considering first the case with only one output, we have where denotes a labour demand elasticity, denotes a factor share and ηi = ∂ ln li/∂ ln wi denotes a labour supply elasticity. Hence, by substitution, (4) (5) Negativity of the denominator in (4) follows from concavity of the cost function19 if we assume also that ηS, ηU < 0. Unskilled immigration therefore depresses unskilled wages and raises skilled wages. Effects on overall mean native wages depend on the proportions of natives in the two groups. Note also that it is change in relative size of native skill groups that matters to wages (given the assumptions of perfectly elastic capital supply and constant returns to scale). Effects on employment then follow from (6) (7) and clearly depend on the magnitude of labour supply elasticities. If ηU and ηS are zero then there are no equilibrium employment effects even if wages are affected. In particular there need be no equilibrium effect on proportion of the native population employed unless labour supply responds to wage changes at the extensive margin. A3. Two Output Goods Take now the case with two types of output. Considering only (3), we have from which it follows immediately that dlnwU/dπ = dlnwU/dπ = 0. This result, essentially an implication of the factor price equalisation theorem (Samuelson 1948), is what Leamer and Levinsohn (1995) call factor price insensitivity. Wages are determined solely by prices through the zero profit condition. Effects on employment are also zero in long run equilibrium. Rather than impacting on wages, long‐run effects of immigration are felt in the output mix. These responses can also be deduced and follow from (3) given unchanged factor prices: where denotes a sectoral share in a factor market. Therefore and unskilled immigration leads to a relative expansion of the sector using unskilled labour relatively intensively, in line with the Rybczinski (1955) theorem. For fixed levels of output, labour market equilibrium would imply wage changes. However these would lead to positive profits being earned in sectors using intensively labour types which become cheaper. Output in such sectors would be expected to expand driving wages back up and long‐run equilibrium will not be restored until wages are driven back to their initial levels. The nature of the solution in general depends upon a comparison between the numbers of goods produced and of labour types. This observation can be generalised beyond the case of only two labour types and can also be extended to allow for non‐traded goods.20 What is at issue is the ability of the economy to respond to immigration through flexibility in its output mix. A smaller number of traded goods mean that there are insufficient degrees of freedom to accommodate changes in the skill mix through changes in the output mix and wage changes are therefore non‐zero even in the long run. However with sufficient number of traded goods there is no need for immigration to induce factor price changes. Footnotes 1 " Studies for the US include Altonji and Card (1991), Borjas (1987; 2003), Butcher and Card (1991), Card (1990; 2001) and this Feature, and LaLonde and Topel (1991). Studies for Europe include Pischke and Velling (1997) for Germany, Hunt (1992) for France, Carrington and de Lima (1996) for Portugal and Winter‐Ebmer and Zweimüller (1996; 1999) for Austria, Friedberg (2001) and Cohen‐Goldner and Paserman (2004) for Israel. See Dustmann and Glitz (2005) for an extensive survey of the literature. Other surveys include Borjas (1994; 1999) and Friedberg and Hunt (1995). 2 " A less common approach (Lalonde and Topel, 1991) treats immigrant and native labour as different labour types. In such a model the effect of immigration depends on substitutability between immigrant labour and native labour of different skill levels. The form of equations arising for estimation are nonetheless similar to those under the more common approach. 3 " In this, these models share the features of standard models used in the broader literature on wage determination. See, for example, the papers of Katz and Murphy (1992), Murphy and Welch (1992) and Card and Lemieux (2001). 4 " Technical details are given in the Appendix. 5 " In the extreme case, for sufficiently large scale immigration of unskilled labour, the economy may specialise in producing only the good that uses the immigrating factor more intensively. Factor price insensitivity will therefore prevail only as long as the factor endowments remain in the original ‘cone of diversification’; see Bhagwati et al. (1998, chapter 28). 6 " This result is related to the well known factor price equalisation result of trade theory – see, for example, Woodland (1982), Samuelson (1948)– although it is a weaker result. 7 " Card (this Feature), drawing on Lewis (2003), reports that there is little evidence for the US that changes in industry structure are taking place. Lewis (2003) suggests that employers possibly adapt to the relative supply of different skill groups in their local market by introducing innovations that take advantage of more readily available factors, even in the absence of relative wage changes. 8 " Work following this approach (Card, 2001) has been influenced by the findings of Bartel (1989) who argued that immigrants in the US tend to settle in areas where immigrant settlement is already strong. 9 " In this paper, we choose to use education classification in our regressions. Information on occupation is only available consistently for individuals who are employed. 10 " The LFS provides information on the usual region of residence. 11 " Information on education is available consistently only from 1983 onwards. 12 " Hourly wages are derived dividing gross weekly wages by the number of hours worked. 13 " All estimates are calculated in GAUSS using DPD98 (Arellano and Bond, 1991, 1998). 14 " We impose the assumption that equiproportionate changes in all skill groups will have no effect. Coefficient estimates for these terms are generally not statistically significant. 15 " Wald tests for the irrelevance of excluded instruments reject strongly at any conventional significance level for immigrant native ratio and the intermediate/unqualified ratio and at the 10% level for the advanced/unqualified ratio. 16 " In this and subsequent Tables we suppress full reporting of coefficients on other regressors and associated test statistics. These results are available on request from the authors. 17 " Adding up is not exact because not all individuals in the LFS can be classified educationally but the discrepancies are small. 18 " In the context of regional labour markets we need only think of p being set in inter‐regional trade. 19 " Note that which is a positive multiple of a quadratic form in the second derivatives of the cost function and therefore negative. 20 " The relevant algebra can be drawn from trade theory models – see Ethier (1984), Woodland (1982). References Altonji , J.G. and Card , D. ( 1991 ). ‘The effects of immigration on the labor market outcomes of less‐skilled natives’, in ( J.M. Abowd and R. B. Freeman, eds.), Immigration, Trade and Labor , pp. 201 –34, Chicago: University of Chicago Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Arellano , M. and Bond , S.R. ( 1991 ). ‘Some tests of specification for panel data: Monte Carlo evidence and an application to employment equations’ , Review of Economic Studies , vol. 58 , pp. 277 – 97 . Google Scholar Crossref Search ADS WorldCat Arellano , M. and Bond , S.R. ( 1998 ). ‘Dynamic panel data estimation using DPD98 for Gauss: a guide for users’ , mimeo , London. Bhagwati , J.N. , Panagariya A. and Srinivasan , T. N. ( 1998 ). Lectures on International Trade . 2nd edition, Cambridge MA: MIT Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Bartel , A.P. ( 1989 ). ‘Where do the new U.S. immigrants live?’ , Journal of Labor Economics , vol. 7 , pp. 371 – 91 . Google Scholar Crossref Search ADS PubMed WorldCat Borjas , G.J. ( 1987 ). ‘Immigrants, minorities, and labor market competition’ , Industrial and Labor Relations Review , vol. 40 ( 3 ), (April), pp. 382 – 92 . Google Scholar Crossref Search ADS WorldCat Borjas , G.J. ( 1994 ). ‘The economics of immigration’ , Journal of Economic Literature , vol. 32 ( 4 ), pp. 1667 – 717 . OpenURL Placeholder Text WorldCat Borjas , G.J ( 1999 ). ‘The economic analysis of immigration’ . Chapter 28, Handbook of Labor Economics , vol. 3 , pp. 1697 – 760 , place: publisher. Google Scholar Crossref Search ADS WorldCat Borjas , G.J. ( 2003 ). ‘The labor demand curve is downward sloping: reexamining the impact of immigration on the labor market’ , Quarterly Journal of Economics , vol. 118 , pp. 1335 – 74 . Google Scholar Crossref Search ADS WorldCat Butcher , K. and Card , D. ( 1991 ). ‘Immigration and wages: evidence from the 1980s’ , American Economic Review , vol. 81 ( 2 ), pp. 292 – 6 . OpenURL Placeholder Text WorldCat Card , D. ( 1990 ). ‘The impact of the Mariel Boatlift on the Miami labor market’ , Industrial and Labor Relations Review , vol. 43 , pp. 245 – 57 . Google Scholar Crossref Search ADS WorldCat Card , D. ( 2001 ). ‘Immigrant inflows, native outflows, and the local labor market impacts of higher immigration’ , Journal of Labor Economics , vol. 19 ( 1 ), pp. 22 – 63 . Google Scholar Crossref Search ADS WorldCat Card , D. and Lemieux , T. ( 2001 ). ‘Can falling supply explain the rising return to college for younger men? A cohort‐based analysis’ , Quarterly Journal of Economics , vol. 116 ( 2 ), pp. 705 – 46 . Google Scholar Crossref Search ADS WorldCat Card , D. and Lewis , E.G. ( 2005 ). ‘The diffusion of Mexican immigrants during the 1990s: explanations and impacts’ , mimeo , University of California at Berkeley. Carrington , W. J. and de Lima , P.J.F. ( 1996 ). ‘The impact of 1970s repatriates from Africa on the Portuguese labor market’ , Industrial and Labor Relations Review , vol. 49 ( 2 ), pp. 330 – 47 . Google Scholar Crossref Search ADS WorldCat Chiswick , B.R. ( 1978 ). ‘The effect of Americanization on the earnings of foreign‐born men’ , Journal of Political Economy , vol. 86 , pp. 897 – 921 . Google Scholar Crossref Search ADS WorldCat Chiswick , B.R. ( 1993 ). ‘Review of immigration and and the work force: economic consequences for the United States and source areas’ , Journal of Economic Literature , vol. 31 , pp. 910 – 1 . OpenURL Placeholder Text WorldCat Cohen‐Goldner , S. and Paserman , D. ( 2004 ). ‘The dynamic impact of immigration on natives’ labor market outcomes: evidence from Israel’ , IZA working paper 1315. Dustmann , C. ( 1996 ). ‘Return migration: the European experience’ , Economic Policy , vol. 22 , pp. 215 – 50 . OpenURL Placeholder Text WorldCat Dustmann , C. and Glitz , A. ( 2005 ). ‘Immigration, jobs and wages: theory, evidence and opinion’ , CEPR‐CReAM publication , London. Ethier , W. J. ( 1984 ) ‘Higher dimensional issues in trade theory’, in ( R. W. Jones and P. B. Kenen, eds.), Handbook of International Economics , vol. 1, pp. 131 – 84 , Amsterdam: North Holland Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Friedberg , R.M. ( 2001 ). ‘The impact of mass migration on the Israeli labor market’ , Quarterly Journal of Economics , vol. 116 ( 4 ), pp. 1373 – 408 . Google Scholar Crossref Search ADS WorldCat Friedberg , R.M. and Hunt , J. ( 1995 ) ‘The impact of immigration on host country wages, employment and growth’ , Journal of Economic Perspectives , vol. 9 , pp. 23 – 44 . Google Scholar Crossref Search ADS WorldCat Gaston , N. and Nelson , D. ( 2000 ). ‘Immigration and labour‐market outcomes in the United States: a political‐economy puzzle’ , Oxford Review of Economic Policy , vol. 16 ( 3 ), pp. 104 – 14 . Google Scholar Crossref Search ADS WorldCat Gaston , N. and Nelson , D. ( 2001 ). ‘The employment and wage effects of immigration: trade and labour economics perspectives’ , University of Nottingham, Leverhulme Centre for Research on Globalisation and Economic Policy Research Paper 2001/28 . Gregg , P. , Machin , S. and Manning , A. ( 2004 ), ‘Mobility and joblessness’, in ( R. Blundell, D. Card, R. Freeman, eds.), Seeking a Premier League Economy , pp. 371 – 410 , Chicago: University of Chicago Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Hanson , G.H. and Slaughter , M.J. ( 1999 ). ‘The Rybczinski Theorem, factor‐price equalization, and immigration: evidence from U.S. states’ , National Bureau of Economic Research Working Paper No. 7074. Hanson , G.H. and Slaughter , M.J. ( 2001 ). ‘Labor market adjustment in open economies: evidence from U.S. states’ , Journal of International Economics , forthcoming. OpenURL Placeholder Text WorldCat Hunt , J. ( 1992 ). ‘The impact of the 1962 repatriates from Algeria on the French labor market’ , Industrial and Labor Relations Review , vol. 45 ( 3 ), pp. 556 – 72 . Google Scholar Crossref Search ADS WorldCat Katz , L.F. and Murphy , K.M. ( 1992 ). ‘Changes in relative wages, 1963‐1987: supply and demand factors’ , Quarterly Journal of Economics , vol. 107 , pp. 35 – 78 . Google Scholar Crossref Search ADS WorldCat Kuhn , P. and Wooton , I. ( 1991 ). ‘Immigration, international trade and the wages of native workers’, in ( J.M. Abowd and R. B. Freeman, eds.), Immigration, Trade and Labor , Chicago: University of Chicago Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Lalonde , R.J. and Topel , R.H. ( 1991 ). ‘Labor market adjustments to increased immigration’, in ( J.M. Abowd and R. B. Freeman, eds.), Immigration, Trade and Labor , Chicago: University of Chicago Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Leamer , E. E. and Levinsohn , J. ( 1995 ). ‘International trade theory: the evidence’, in ( G.M. Grossman and K. Rogoff, eds.), Handbook of International Economics , vol. 3, pp. 1339 – 94 , Amsterdam: North Holland Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Lewis , E.G. ( 2003 ). ‘Local open economics within the U.S. how do industries respond to immigration?’ , Federal Reserve Bank of Philadelphia Working Paper. Murphy , K.M. and Welch , F. ( 1992 ). ‘The structure of wages’ , Quarterly Journal of Economics , vol. 107 ( 1 ), pp. 285 – 326 . Google Scholar Crossref Search ADS WorldCat Pischke , J. and Velling , J. ( 1997 ). ‘Employment effects of immigration to Germany: an analysis based on local labor markets’ , Review of Economics and Statistics , vol. 79 ( 4 ), pp. 594 – 604 . Google Scholar Crossref Search ADS WorldCat Rybczinski , T. M. ( 1955 ). ‘Factor endowments and relative commodity prices’ , Economica , vol. 22 , pp. 336 – 41 . Google Scholar Crossref Search ADS WorldCat Samuelson , P. A. ( 1948 ) ‘International trade and the equalization of factor prices’ , Economic Journal , vol. 48 , pp. 163 – 84 . Google Scholar Crossref Search ADS WorldCat Scheve , K.F. and Slaughter , M.J. ( 2001 ). ‘Labor market competition and individual preferences over migration policy’ , Review of Economics and Statistics , vol. 83 ( 1 ), pp. 133 – 45 . Google Scholar Crossref Search ADS WorldCat Winter‐Ebmer , R. and Zweimüller , J. ( 1996 ) ‘Immigration and the earnings of young native workers’ , Oxford Economics Papers , vol. 48 , pp. 473 – 91 . Google Scholar Crossref Search ADS WorldCat Winter‐Ebmer , R. and Zweimüller , J. ( 1999 ) ‘Do immigrants displace young native workers: the Austrian experience’ , Journal of Population Economics , vol. 12 ( 2 ), pp. 327 – 40 . Google Scholar Crossref Search ADS WorldCat Woodland , A. ( 1982 ). International Trade and Resource Allocation , Amsterdam: North‐Holland . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Author notes " We thank the British Home Office for financial support. Dustmann and Preston gratefully acknowledge the support of the Economic and Social Research Council (grant RES‐000‐23‐0332). Fabbri also acknowledges financial support from the Munich Graduate School of Economics/Deutsche Forschungsgemeinschaft. We are grateful to Yasmin Dolatabadi for excellent research assistance. Important comments and advice were received from David Card, Wendy Carlin, Hidehiko Ichimura, Joanne Lindley, Costas Meghir, Jonathan Wadsworth, Carole Willis, and Frank Windmeijer. We thank the UK Data Archive at Essex for providing the data. All errors remain our own. © Royal Economic Society 2005.

Hatton, Timothy, J.;Tani,, Massimiliano

doi: 10.1111/j.1468-0297.2005.01039.xpmid: N/A

Abstract Most studies that look across local labour markets have found the effects of immigration to be benign. One possibility is that immigrants to a specific area simply push non‐immigrants onwards elsewhere, thereby diffusing the labour market effects. Examining net internal migration between 11 regions of Britain over two decades, we find consistently negative displacement effects but their magnitude and significance varies. The effects are somewhat stronger for the southern regions where immigration from abroad is concentrated. The results suggest that internal migration is one of the mechanisms through which regional labour markets adjust to immigration shocks. Mounting concern about the potential impact of immigration on the earnings and employment opportunities of non‐immigrant workers has provided the background for a growing literature that has sought to measure these effects. One approach has been to examine the effects of immigration by looking across local or regional labour markets, within a given country, that have different rates of immigrant inflow from abroad to see if those with higher rates of immigration also have lower rates of wage increase or slower growth of employment for non‐immigrant workers. This approach, sometimes called the ‘spatial correlations’ approach, has generated a lively literature on the effects of immigration in the US and some other countries such as Germany. The typical findings are that the wage and employment effects of immigration are negligible and this has led to further debate about what adjustment mechanisms account for these apparently small labour market effects. In Britain this has not been such an important issue compared with these other countries until quite recently. But, as Figure 1 shows, immigration to the UK has been on the increase. In 1976–80 there was a net inflow of 37,200 foreign citizens per annum; by 1996–2000 this had increased to a net inflow of 110,200 per annum. Immigration increased particularly rapidly during the 1990s, and it is responsible for about half of the population growth during that decade. Recently, reforms to UK immigration policy have sought to make immigration easier for certain groups and there is active consideration of future policy initiatives. This puts the issue of the labour market effects of immigration firmly on the agenda. Dustmann et al. (this Feature) find that immigration has positive but largely insignificant effects on unemployment in Britain and small positive effects on wage rates. Thus, in line with much of the US literature, immigration seems to have relatively benign effects on the labour market. Frijters et al. (this Feature) examine the mechanics of immigrant job search in Britain, finding that immigrants use similar methods but are somewhat less successful than natives in gaining employment. Fig. 1. Open in new tabDownload slide Net Immigration of Foreign Citizens to the UK, 1971–2000 Fig. 1. Open in new tabDownload slide Net Immigration of Foreign Citizens to the UK, 1971–2000 Our purpose in this article is to carry forward this research by investigating one possible mechanism through which the British labour market adjusts to immigration. One argument put forward in the US context is that local labour markets adjust to an immigrant inflow through outflows to other localities of the native‐born population and previous immigrant cohorts. In the absence of such mobility, any negative effects of immigration should be observed in differences across localities. But with perfect mobility across localities, the labour market impacts of immigration will not be observed locally because the effects will be spread across the entire country. In this article we investigate this issue by estimating the effects of net immigration on net internal migration between 11 regions of Britain over the 1980s and 1990s. We find that these effects are potentially important, especially in southern England where the inflow of immigrants is largest relative to the regional population. 1. Immigration Effects and Non‐Immigrant Labour Supply Many studies have looked for the effects of immigration by comparing outcomes across local labour markets that have different rates of immigration. The earliest and most influential studies for the US found that the wage and employment effects of immigration were small (Altonji and Card, 1991; Lalonde and Topel, 1991).1 These findings have been largely replicated in further research on the US and they are upheld for Britain by Dustmann et al. (this Feature). However, local area studies of the impact of immigration in Europe, and Germany in particular, have found some negative wage and unemployment effects (Winklemann and Zimmermann, 1993; De New and Zimmermann, 1994; Angrist and Kugler, 2003), possibly a reflection of differences in labour market institutions in Europe as compared with the US.2 Studies of large immigrant influxes such as the Algerian repatriates to France in the 1960s and those returning to Portugal in the 1970s also find negative wage effects, although they are surprisingly modest (Hunt 1992; Carrington and di Lima, 1996). The fact that immigrant‐induced labour supply shocks are often found to be benign seems inconsistent with the widely held view that labour demand curves slope down and this has led to a questioning of the methodology used and a search for other channels of adjustment that might help to explain these apparently anomalous findings. 3 One argument is that if the supply of other factors to a locality is elastic then it will only affect the wage rates of particular skill groups insofar as it alters the skill mix. However, recent studies for the US suggest that even immigrant‐induced changes to the skill composition have little effect on relative wages (Card, this Feature). A possible explanation for this is that, since local economies are very open, the output mix adjusts and any skill‐specific wage effects are muted by the changing specialisation in inter‐regional trade. This argument is examined in the contributions to this Feature by Card and by Dustmann et al. An alternative adjustment mechanism is through the labour market itself: immigration to one locality causes inter‐regional migration that spreads the wage effects of immigration throughout the economy. In his study of large scale Cuban immigration to Miami at the time of the Mariel Boatlift in 1980, Card (1990) found that this mass influx had very little effect on the wages and employment of the native‐born – implying that adjustment took place through inter‐regional migration. Although Filer (1992) found that immigrants crowded out inter‐regional migrants almost one for one, other estimates disaggregating by skill‐level find little evidence of such effects (Card and DiNardo, 2000; Card, 2001). More recently Borjas (2003) has examined the effects of immigration using nation‐wide education and experience groups rather than local areas as the units of observation. Since there is less scope for mobility across these categories at the national level there should be less downward bias in the estimate. He finds larger negative effects on wage rates than do previous studies: an increase in immigration equivalent to 10% of the labour force leads to a decline in the wage for native born workers of 3%. Thus a key issue is whether (and if so, how) the effects of immigration on wages and/or unemployment may be masked by inter‐regional labour mobility. To illustrate this more formally it is worth setting out a simple perfectly competitive model of labour supply and demand in a local labour market that includes inter‐regional migration. In regional labour market i, where immigrant and non‐immigrant labour are perfect substitutes, the change in total employment at time t can be represented as: (1) where mi,t = ΔMi,t/(M + N)i,t−1 is the growth in employment of immigrants as a proportion of the initial labour force (immigrant plus non‐immigrant), and ni,t = ΔNi,t/(M + N)i,t−1 is the growth of non‐immigrant employment, also as a proportion of the initial labour force. is the shift in labour demand, wi,t is the change in the (log) wage, the parameter α is the elasticity of labour demand with respect to the wage, and vi,t is a random disturbance. The growth of immigrant labour supply is given by: (2) where is the (log) national average wage, zi,t is a region‐specific immigration shock, and ui,t is a random disturbance term. Thus immigration to a region from abroad depends on the region's relative wage and other factors, absorbed in zi,t, that could include conditions in source countries or previous immigrant inflows. Non‐immigrant labour supply is also represented by a migration equation, although here migration is inter‐regional rather than international: (3) Here, non‐immigrant migration is determined by the region's relative wage, but there is also a direct negative effect of immigration. While inflows from abroad could crowd out non‐immigrants from the region by reducing the local wage, there are other channels that might include the housing market, congestion effects, or even self‐selected ethnic segregation effects. Using (1) and (3) we can solve for the regional wage level (written here in terms of wage change) as a function of immigration and other variables: (4) In the special case where internal migration is unresponsive to the relative wage or to the direct displacement effect of immigration, γ1 = γ2 = 0, and (4) reduces to: (5) Studies of the labour market impact of immigration that assume away inter‐regional mobility typically estimate some version of (5). A number of things are worth noting about (4) and/or (5). The first is that, in either (4) or (5), mi,t will be correlated with the error term, vi,t − ei,t, since, from (2), it is a function of the regional wage, wi,t. One approach would be to instrument mi,t using some component of zi,t. Most studies have used as instruments past immigrant flows to the region or the stock of immigrants at the beginning of the period to reflect the well‐documented tendency for new immigrants to join existing immigrant communities. The second methodological point is that most studies that use (5) omit the region‐specific demand shock, . Instead they assume that it can be accounted for by a nationwide shock, often represented in panel estimates by year dummies, plus a random regional component that is absorbed into the error term. If for some reason is correlated with mi,t, omitting the former may bias the coefficient estimate on the latter (the same applies to (4)). The third point is that in (5) the coefficient on mi,t is simply the inverse of the labour demand elasticity (hence the need to find an instrument from the immigrant supply function). The fact that the estimated coefficients are often close to zero seems inconsistent with the evidence from direct estimates of the labour demand elasticity, which often find it to be less than one; hence its reciprocal should be larger than one rather than close to zero. However it is possible that, at the regional level, labour demand could be very elastic, either because goods markets are highly integrated or because other factors, such as capital, are mobile.4 The fourth point, more directly relevant to what follows, is that if (4) is the correct specification, then the coefficient on mi,t will be a downward biased estimate of the inverse labour demand elasticity; the larger are γ1 and γ2, the greater is the downward bias. It is worth noting also that specification (4) should include the ratio of the national to the (lagged) regional wage. The fifth point is that the wage effect of immigration depends on both direct and indirect effects. If the direct effect, γ2, is close to one, then the wage effect in (4) evaporates. If this direct crowding out effect is large enough then there will be no wage effects, even with relatively inelastic labour demand. It is useful to solve (1) and (3) for the change in non‐immigrant labour supply to give: (6) If γ1 and/or γ2 is greater than zero then the coefficient on mi,t in (6) will be negative (otherwise zero). It is also worth noting that the effects of immigration on non‐immigrant mobility are not independent of the labour demand elasticity; if α is ‘large’ then the effect of mi,t in (6) could be close to zero even though γ1 is substantially larger than zero. Finally there are two, more general, points to make. The first is that, for simplicity, in the framework set out above, the market clears and there is no unemployment. However a broadly analogous system could be set out where wages do not adjust and where immigration and internal migration are determined by inter‐regional differences in unemployment rates. But if immigrants compete less effectively in the search for jobs as suggested by Frijters et al. (this Feature) then the effect on the mobility of non‐immigrants would be muted. Secondly, finding negligible effects of immigration on internal migration, not controlling for employment shocks, could be consistent with the results of studies that find interregional migration is large in response to shifts in labour demand (Katz and Krueger, 1992).5 2. Immigration and Internal Migration In Britain Growing interest in the effects of immigration has coincided with a sustained increase in immigration. The overall net inflow of foreign citizens to the UK is graphed in Figure 1. These data come from the International Passenger Survey (IPS), which is based on a sample of travellers arriving and departing at UK airports and the Channel tunnel. International migrants are defined as those arriving in the UK (or leaving) for at least a year after having resided for at least a year abroad (in the UK). Net immigration of foreign citizens increased from 38,200 per annum in the 1970s rising to a 44,600 in the 1980s and then more sharply to 85,300 per annum in the 1990s.6 These net immigration figures are the balance of much larger flows. In the 1990s the gross inflow averaged 175,600 per annum while the gross outflow averaged 90,300 per annum. It is worth noting that the definition of immigration used here excludes the emigration and immigration of British citizens, which produced an annual net outflow of 21,600 per annum in the 1990s.7 The increase in the immigration of foreign citizens largely reflects a rise in immigration from the European Union where no visa restrictions apply. But there have also been increases in net immigration from most other parts of the world. Among these the most important areas are the New Commonwealth, particularly the Indian subcontinent, Old Commonwealth countries (including South Africa), and the US. Studies of these immigration flows show that they are determined by economic conditions at home and abroad, operating through the ‘filter’ of immigration policies. The pattern of flows to and from different parts of the world is also conditioned by the source country composition of the existing stock of foreign‐born residents in the UK (Hatton, 2005). Our focus is on differences in the net inflow of foreign citizens to different regions. Table 1 shows gross and net inflows from abroad into the eleven Standard Statistical Regions of Britain per thousand of the regional populations. While gross inflows are sometimes sizeable, especially if cumulated over a number of years, the annual average net inflow is typically fairly small. During the 1980s it was less than 1 per thousand of the population in all regions with the exception of Greater London, where it averaged nearly 3 per thousand and the rest of the Southeast at 1.7 per thousand. By comparison, the figures for the 1990s reflect a general rise in immigration to all regions, with East Anglia, the West Midland and Yorkshire and Humberside now experiencing inflows of foreign citizens of more than one per thousand. But again London stands out, with the annual inflow in the 1990s exceeding 4 per thousand. Table 1 Gross and Net International Migration by Region, 1981–2000 (Per thousand of the regional population) . 1981–90 . 1981–90 . 1981–90 . 1991–2000 . 1991–2000 . 1991–2000 . Gross Inflow . Gross Outflow . Net Inflow . Gross Inflow . Gross Outflow . Net Inflow . London 7.11 −4.14 2.96 9.93 −5.88 4.05 Rest of S. East 5.78 −4.04 1.73 6.17 −3.30 2.89 East Anglia 1.73 −1.21 0.52 2.49 −1.37 1.12 South West 1.47 −1.07 0.41 1.76 −1.03 0.74 West Midlands 1.21 −0.72 0.50 2.06 −0.80 1.25 East Midlands 1.00 −0.51 0.49 1.47 −0.87 0.60 Yorks & Humber 1.08 −0.52 0.56 1.87 −0.81 1.06 North West 1.28 −0.65 0.63 1.62 −0.86 0.76 North 0.68 −0.37 0.31 1.18 −0.56 0.56 Wales 0.94 −0.70 0.23 1.32 −0.61 0.71 Scotland 1.31 −1.24 0.07 1.93 −1.24 0.70 . 1981–90 . 1981–90 . 1981–90 . 1991–2000 . 1991–2000 . 1991–2000 . Gross Inflow . Gross Outflow . Net Inflow . Gross Inflow . Gross Outflow . Net Inflow . London 7.11 −4.14 2.96 9.93 −5.88 4.05 Rest of S. East 5.78 −4.04 1.73 6.17 −3.30 2.89 East Anglia 1.73 −1.21 0.52 2.49 −1.37 1.12 South West 1.47 −1.07 0.41 1.76 −1.03 0.74 West Midlands 1.21 −0.72 0.50 2.06 −0.80 1.25 East Midlands 1.00 −0.51 0.49 1.47 −0.87 0.60 Yorks & Humber 1.08 −0.52 0.56 1.87 −0.81 1.06 North West 1.28 −0.65 0.63 1.62 −0.86 0.76 North 0.68 −0.37 0.31 1.18 −0.56 0.56 Wales 0.94 −0.70 0.23 1.32 −0.61 0.71 Scotland 1.31 −1.24 0.07 1.93 −1.24 0.70 Source: Data from the International Passenger Survey supplied by the Office for National Statistics. Open in new tab Table 1 Gross and Net International Migration by Region, 1981–2000 (Per thousand of the regional population) . 1981–90 . 1981–90 . 1981–90 . 1991–2000 . 1991–2000 . 1991–2000 . Gross Inflow . Gross Outflow . Net Inflow . Gross Inflow . Gross Outflow . Net Inflow . London 7.11 −4.14 2.96 9.93 −5.88 4.05 Rest of S. East 5.78 −4.04 1.73 6.17 −3.30 2.89 East Anglia 1.73 −1.21 0.52 2.49 −1.37 1.12 South West 1.47 −1.07 0.41 1.76 −1.03 0.74 West Midlands 1.21 −0.72 0.50 2.06 −0.80 1.25 East Midlands 1.00 −0.51 0.49 1.47 −0.87 0.60 Yorks & Humber 1.08 −0.52 0.56 1.87 −0.81 1.06 North West 1.28 −0.65 0.63 1.62 −0.86 0.76 North 0.68 −0.37 0.31 1.18 −0.56 0.56 Wales 0.94 −0.70 0.23 1.32 −0.61 0.71 Scotland 1.31 −1.24 0.07 1.93 −1.24 0.70 . 1981–90 . 1981–90 . 1981–90 . 1991–2000 . 1991–2000 . 1991–2000 . Gross Inflow . Gross Outflow . Net Inflow . Gross Inflow . Gross Outflow . Net Inflow . London 7.11 −4.14 2.96 9.93 −5.88 4.05 Rest of S. East 5.78 −4.04 1.73 6.17 −3.30 2.89 East Anglia 1.73 −1.21 0.52 2.49 −1.37 1.12 South West 1.47 −1.07 0.41 1.76 −1.03 0.74 West Midlands 1.21 −0.72 0.50 2.06 −0.80 1.25 East Midlands 1.00 −0.51 0.49 1.47 −0.87 0.60 Yorks & Humber 1.08 −0.52 0.56 1.87 −0.81 1.06 North West 1.28 −0.65 0.63 1.62 −0.86 0.76 North 0.68 −0.37 0.31 1.18 −0.56 0.56 Wales 0.94 −0.70 0.23 1.32 −0.61 0.71 Scotland 1.31 −1.24 0.07 1.93 −1.24 0.70 Source: Data from the International Passenger Survey supplied by the Office for National Statistics. Open in new tab Table 2 shows the regional location of UK‐born and foreign‐born as shares of regional populations from the 1991 census. It illustrates that the bias towards London is reflected in the population stock by place of birth. While nearly 40% of the foreign‐born were located in Greater London, only 10% of the UK‐born lived there in 1991. And while the Rest of the Southeast, East Anglia and the East Midlands had foreign‐born shares that are similar to their shares of the UK‐born, the foreign‐born are under‐represented in all other regions. These patterns differ somewhat by place of overseas birth. In particular the over‐representation of the foreign‐born in London is greatest among those from New Commonwealth countries and least among those from the EU. In the rest of the Southeast those born in the Old Commonwealth and EU are over‐represented while those from the New Commonwealth are under‐represented. Those born in the New Commonwealth are over‐represented in the East Midlands while those from the Old Commonwealth are over‐represented in Scotland. With a few exceptions, those from the EU and Other Foreign countries have regional concentrations similar to the foreign‐born as a whole. Table 2 Regional Shares of GB Population by Birthplace, 1991 (Percentage of each birthplace category) Region . UK Born . Non‐UK Born . Old CW . New CW . EU . Other Foreign . London 10.2 38.4 30.1 43.3 27.7 37.4 Rest of S. East 19.2 18.4 24.6 14.7 25.5 19.5 East Anglia 3.7 3.1 3.6 1.7 4.2 4.3 South West 8.6 5.2 9.0 3.6 8.5 5.4 West Midlands 9.4 9.0 4.5 11.8 5.4 7.5 East Midlands 7.3 5.5 4.1 6.3 5.6 4.5 Yorks & Humber 9.1 5.5 4.3 6.3 5.8 4.6 North West 11.7 7.3 6.0 6.9 6.2 8.4 North 5.8 1.7 2.5 1.4 2.6 1.6 Wales 5.4 2.0 2.6 1.4 3.3 2.3 Scotland 9.5 3.9 8.7 2.6 5.2 4.5 Region . UK Born . Non‐UK Born . Old CW . New CW . EU . Other Foreign . London 10.2 38.4 30.1 43.3 27.7 37.4 Rest of S. East 19.2 18.4 24.6 14.7 25.5 19.5 East Anglia 3.7 3.1 3.6 1.7 4.2 4.3 South West 8.6 5.2 9.0 3.6 8.5 5.4 West Midlands 9.4 9.0 4.5 11.8 5.4 7.5 East Midlands 7.3 5.5 4.1 6.3 5.6 4.5 Yorks & Humber 9.1 5.5 4.3 6.3 5.8 4.6 North West 11.7 7.3 6.0 6.9 6.2 8.4 North 5.8 1.7 2.5 1.4 2.6 1.6 Wales 5.4 2.0 2.6 1.4 3.3 2.3 Scotland 9.5 3.9 8.7 2.6 5.2 4.5 Source: 1991 Census Report for Great Britain (Part 1), London: HMSO (1993). Open in new tab Table 2 Regional Shares of GB Population by Birthplace, 1991 (Percentage of each birthplace category) Region . UK Born . Non‐UK Born . Old CW . New CW . EU . Other Foreign . London 10.2 38.4 30.1 43.3 27.7 37.4 Rest of S. East 19.2 18.4 24.6 14.7 25.5 19.5 East Anglia 3.7 3.1 3.6 1.7 4.2 4.3 South West 8.6 5.2 9.0 3.6 8.5 5.4 West Midlands 9.4 9.0 4.5 11.8 5.4 7.5 East Midlands 7.3 5.5 4.1 6.3 5.6 4.5 Yorks & Humber 9.1 5.5 4.3 6.3 5.8 4.6 North West 11.7 7.3 6.0 6.9 6.2 8.4 North 5.8 1.7 2.5 1.4 2.6 1.6 Wales 5.4 2.0 2.6 1.4 3.3 2.3 Scotland 9.5 3.9 8.7 2.6 5.2 4.5 Region . UK Born . Non‐UK Born . Old CW . New CW . EU . Other Foreign . London 10.2 38.4 30.1 43.3 27.7 37.4 Rest of S. East 19.2 18.4 24.6 14.7 25.5 19.5 East Anglia 3.7 3.1 3.6 1.7 4.2 4.3 South West 8.6 5.2 9.0 3.6 8.5 5.4 West Midlands 9.4 9.0 4.5 11.8 5.4 7.5 East Midlands 7.3 5.5 4.1 6.3 5.6 4.5 Yorks & Humber 9.1 5.5 4.3 6.3 5.8 4.6 North West 11.7 7.3 6.0 6.9 6.2 8.4 North 5.8 1.7 2.5 1.4 2.6 1.6 Wales 5.4 2.0 2.6 1.4 3.3 2.3 Scotland 9.5 3.9 8.7 2.6 5.2 4.5 Source: 1991 Census Report for Great Britain (Part 1), London: HMSO (1993). Open in new tab Table 3 gives a measure of the relative skills of British and foreign employees by region, averaged over the years 1992–7, derived from the Labour Force Survey. The proportion of foreign employees with relatively high education is rather similar to that of the native employees overall, although there are some differences across regions. There is some tendency for the education levels of foreign employees to be lower than that of natives in the south and higher than that of natives in the north and Wales. However, as Table 2 showed these regions have only very small proportions of foreigners present. Overall, it appears that compositional differences between natives and foreigners are not large enough to have very decisive effects at the regional level. While it would be interesting to compare the skill mix of the immigrant inflow with that of existing residents, unfortunately it is not possible to observe relative skills in the immigration flow data by region. Table 3 Percentage of Employees with High Education, 1992–7 Region . British . Foreign . Region . British . Foreign . London & S. East 28.9 24.6 Yorks & Humberside 21.7 22.2 East Anglia 21.2 17.4 North West 23.8 22.4 South West 24.2 22.4 North 21.7 32.8 West Midlands 21.4 17.3 Wales 23.0 29.8 East Midlands 21.4 22.4 Scotland 26.4 26.9 Region . British . Foreign . Region . British . Foreign . London & S. East 28.9 24.6 Yorks & Humberside 21.7 22.2 East Anglia 21.2 17.4 North West 23.8 22.4 South West 24.2 22.4 North 21.7 32.8 West Midlands 21.4 17.3 Wales 23.0 29.8 East Midlands 21.4 22.4 Scotland 26.4 26.9 Notes: Percentage of employees who remained in education to at least 17 years of age. Average for the years 1992–7 by region obtained from EUROSTAT and derived from the Labour Force Survey. Here, foreign is defined by citizenship rather than by birth. Open in new tab Table 3 Percentage of Employees with High Education, 1992–7 Region . British . Foreign . Region . British . Foreign . London & S. East 28.9 24.6 Yorks & Humberside 21.7 22.2 East Anglia 21.2 17.4 North West 23.8 22.4 South West 24.2 22.4 North 21.7 32.8 West Midlands 21.4 17.3 Wales 23.0 29.8 East Midlands 21.4 22.4 Scotland 26.4 26.9 Region . British . Foreign . Region . British . Foreign . London & S. East 28.9 24.6 Yorks & Humberside 21.7 22.2 East Anglia 21.2 17.4 North West 23.8 22.4 South West 24.2 22.4 North 21.7 32.8 West Midlands 21.4 17.3 Wales 23.0 29.8 East Midlands 21.4 22.4 Scotland 26.4 26.9 Notes: Percentage of employees who remained in education to at least 17 years of age. Average for the years 1992–7 by region obtained from EUROSTAT and derived from the Labour Force Survey. Here, foreign is defined by citizenship rather than by birth. Open in new tab How does the net flow of migrants from abroad compare with the flow of migrants within Britain? Table 4 shows the net and gross flows to and from different regions. These are based on National Health Service registration data for those whose registration changes from one region to another. These include all individuals regardless of nationality, place of birth or labour force status, and they represent movements within the UK including Northern Ireland. On this basis, about 2% of the population changes region each year. As with the international migration flows the net balances are small relative to the gross flows. Perhaps the most striking feature is the persistent net outflow from London. Other regions in the south and east exhibit persistent net inflows while the northern regions and Scotland have small net outflows. It is tempting to suggest that net immigration to London from abroad has been associated with a displacement of population from London to other regions. However, there is little evidence, at least in the regional cross‐section, of a strong inverse correlation between immigration and regional in‐migration across the other regions.8 Table 4 Gross and Net Inter‐Regional Migration, 1981–2000 (Percentage of regional population) . 1981–1990 . 1981–1990 . 1981–1990 . 1991–2000 . 1991–2000 . 1991–2000 . Gross Inflow . Gross Outflow . Net Inflow . Gross Inflow . Gross Outflow . Net Inflow . London 2.33 3.02 −0.69 2.26 2.99 −0.73 Rest of S. East 2.56 2.30 0.27 2.49 2.25 0.23 East Anglia 3.08 2.35 0.70 2.91 2.43 0.48 South West 2.80 2.09 0.71 2.73 2.19 0.54 West Midlands 1.49 1.68 −0.18 1.66 1.81 −0.16 East Midlands 2.26 2.02 0.23 2.44 2.18 0.26 Yorks & Humber 1.59 1.66 −0.07 1.79 1.85 −0.06 North West 1.29 1.56 −0.27 1.48 1.66 −0.18 North 1.43 1.60 −0.17 1.55 1.65 −0.10 Wales 1.86 1.62 0.24 1.87 1.75 0.12 Scotland 0.88 1.03 −0.15 0.96 0.96 0.00 . 1981–1990 . 1981–1990 . 1981–1990 . 1991–2000 . 1991–2000 . 1991–2000 . Gross Inflow . Gross Outflow . Net Inflow . Gross Inflow . Gross Outflow . Net Inflow . London 2.33 3.02 −0.69 2.26 2.99 −0.73 Rest of S. East 2.56 2.30 0.27 2.49 2.25 0.23 East Anglia 3.08 2.35 0.70 2.91 2.43 0.48 South West 2.80 2.09 0.71 2.73 2.19 0.54 West Midlands 1.49 1.68 −0.18 1.66 1.81 −0.16 East Midlands 2.26 2.02 0.23 2.44 2.18 0.26 Yorks & Humber 1.59 1.66 −0.07 1.79 1.85 −0.06 North West 1.29 1.56 −0.27 1.48 1.66 −0.18 North 1.43 1.60 −0.17 1.55 1.65 −0.10 Wales 1.86 1.62 0.24 1.87 1.75 0.12 Scotland 0.88 1.03 −0.15 0.96 0.96 0.00 Source: National Health Service registration data supplied by the Office for National Statistics. Open in new tab Table 4 Gross and Net Inter‐Regional Migration, 1981–2000 (Percentage of regional population) . 1981–1990 . 1981–1990 . 1981–1990 . 1991–2000 . 1991–2000 . 1991–2000 . Gross Inflow . Gross Outflow . Net Inflow . Gross Inflow . Gross Outflow . Net Inflow . London 2.33 3.02 −0.69 2.26 2.99 −0.73 Rest of S. East 2.56 2.30 0.27 2.49 2.25 0.23 East Anglia 3.08 2.35 0.70 2.91 2.43 0.48 South West 2.80 2.09 0.71 2.73 2.19 0.54 West Midlands 1.49 1.68 −0.18 1.66 1.81 −0.16 East Midlands 2.26 2.02 0.23 2.44 2.18 0.26 Yorks & Humber 1.59 1.66 −0.07 1.79 1.85 −0.06 North West 1.29 1.56 −0.27 1.48 1.66 −0.18 North 1.43 1.60 −0.17 1.55 1.65 −0.10 Wales 1.86 1.62 0.24 1.87 1.75 0.12 Scotland 0.88 1.03 −0.15 0.96 0.96 0.00 . 1981–1990 . 1981–1990 . 1981–1990 . 1991–2000 . 1991–2000 . 1991–2000 . Gross Inflow . Gross Outflow . Net Inflow . Gross Inflow . Gross Outflow . Net Inflow . London 2.33 3.02 −0.69 2.26 2.99 −0.73 Rest of S. East 2.56 2.30 0.27 2.49 2.25 0.23 East Anglia 3.08 2.35 0.70 2.91 2.43 0.48 South West 2.80 2.09 0.71 2.73 2.19 0.54 West Midlands 1.49 1.68 −0.18 1.66 1.81 −0.16 East Midlands 2.26 2.02 0.23 2.44 2.18 0.26 Yorks & Humber 1.59 1.66 −0.07 1.79 1.85 −0.06 North West 1.29 1.56 −0.27 1.48 1.66 −0.18 North 1.43 1.60 −0.17 1.55 1.65 −0.10 Wales 1.86 1.62 0.24 1.87 1.75 0.12 Scotland 0.88 1.03 −0.15 0.96 0.96 0.00 Source: National Health Service registration data supplied by the Office for National Statistics. Open in new tab Previous analyses of internal migration find that population or labour force flows take place in response to variables representing regional labour market conditions. The key variables used in these analyses are unemployment and vacancy rates, wages and house prices (McCormick, 1997). One study using the NHS registration data up to the mid‐1980s found effects on net in‐migration in the expected direction arising from unemployment and vacancy rates as well relative house prices, but not from relative wage rates (Jackman and Savouri, 1992). Studies of regional mobility that use individual level data do tend to find conventional wage effects but sometimes find weaker or perverse effects for some of the other key variables (Pissarides and Wadsworth, 1989; Hughes and McCormick, 1994). Overall, these studies indicate that interregional migration does serve as an equilibrating response to persistent differences in labour demand, but only to a limited extent. Evaluating these equilibrating effects Pissarides and McMaster (1990) found that inter‐regional migration would take ten years to eliminate half of an initial unemployment differential. More recently the focus has shifted to the effects on mobility of regional house price differentials. These consistently work in the opposite direction to the wage incentives: higher relative house prices discourage net in‐migration. Muellbauer and Cameron (1998) find that while the level of house prices has a strong negative effect on in‐migration, the expectation of capital gains through house price increases has a positive effect. Less attention has been paid to shifts on the supply side of the labour market. Only one study has examined the effects of immigration on inter‐regional migration. Focusing on the Southeast region, Muellbauer and Murphy (1988) found a strong negative relationship between the change in immigration and the level of inter‐regional in‐migration. To the extent that immigration raises house prices, one would expect that any direct displacement effect in inter‐regional migration would be greater when house prices are omitted than when they are included. 3. Explaining Inter‐Regional Migration Flows In order to explore the correlates of inter‐regional mobility we examine migration flows between each of the 11 UK regions. Our estimating framework can be expressed as follows: (7) where nij,t is the net inter‐regional migration rate from region j to region i at time t. The term (xi,t − xj,t) represents a vector of variables determining migration from j to i, expressed as the difference between region i and region j at time t, with coefficient vector λ. The term (mi,t−1 − mj,t−1) is the inflow rate of foreign immigrants to region i minus the inflow rate to j. Our regions are the Standard Statistical Regions excluding Northern Ireland but separately distinguishing Greater London and the rest of the Southeast as in Tables 1 and 2. The dependent variable is the net migration between one region and another from National Health Service registration flow data, divided half the combined populations of the two regions at time t − 1. The immigration series are the IPS series for net immigration of foreign citizens from overseas, each divided by half the combined regional populations at t − 1. It is important to note that these are population flows, not labour force flows, although previous studies suggest that labour market variables are important in explaining them. While it is not possible to investigate the effects by skill group or by labour force status, these data do have the advantage of being directly observed flows rather than differences between imperfectly measured population or labour force stocks. The labour market variables include the claimant count unemployment rate and the labour exchange/job centre vacancy rate. These are taken from the NOMIS database and are divided by the labour force base from the same source. We also include the vacancy inflow rate as suggested in studies of labour market matching (Coles and Smith, 1998). For the regional wage rate we use average weekly earnings net of overtime from the New Earnings Survey. Finally, following the literature on migration and house prices, we include the level and the change in the regional house price index (see Appendix). The maximum period for which all these series could be assembled is the period from 1981 to 2000, which (allowing for one lag), gives a panel of 1,045 observations (19 years × 55 region pairs). We use panel regression with 55 region pair dummies and we include a full set of year dummies. This takes out the effects of economy‐wide conditions on internal mobility that some studies have identified (Jackman and Savouri, 1992). In initial estimation we instrumented the net immigration rates, using their lagged values.9 But the results we present here simply include the one period lagged value of net immigration to each region, rather than using it as the instrument for the contemporaneous variable. There are also potential concerns about the endogeneity of the other explanatory variables and we experimented with lags of these too.10 Since it made very little difference to the main results, we enter them as contemporaneous variables, with the exception of the wage.11 The estimate in column (1) of Table 5 excludes all other variables except the region pair and year dummies (not reported). The coefficient on the difference between the immigration rates to regions i and j is negative as expected but it is not significant. Most of the explanatory power comes from the bilateral pair dummies and the year dummies. The labour market and housing variables are added in column (2), with the effect of increasing the size and significance of the immigration variable. This suggests that it is important to control for these additional variables and that in their absence the coefficient on immigration is biased downwards. The difference between regional vacancy inflow rates is an important determinant of regional in‐migration and it has the expected positive sign. By contrast, the vacancy stock was generally insignificant and so it was eliminated. Both the unemployment rates and lagged earnings give the expected sign although the coefficient on the former is small and insignificant. Table 5 Estimates of Bilateral Net Inter‐Regional In‐Migration Rates, 1982–2000 (Dependent variable: inter‐regional net migration to i from j) . 11 Regions . 6 Southern Regions . (1) . (2) . (3) . (4) . Net immigration rate (i minus j at t − 1) −0.024 −0.064 −0.128 −0.162 (0.9) (2.1) (1.5) (2.1) Log vacancy inflow/labour force (i minus j at t) 0.024 0.044 (4.9) (2.9) Log unemployment rate (i minus j at t) −0.009 0.001 (1.2) (0.0) Log average earnings (i minus j at t − 1) 0.081 0.308 (2.0) (2.2) Log house price (i minus j at t) −0.064 −0.212 (4.1) (3.1) Δ Log house price (i minus j at t) 0.063 0.079 (3.4) (0.9) Adj. R2 0.92 0.92 0.93 0.93 No of observations 1,045 1,045 285 285 . 11 Regions . 6 Southern Regions . (1) . (2) . (3) . (4) . Net immigration rate (i minus j at t − 1) −0.024 −0.064 −0.128 −0.162 (0.9) (2.1) (1.5) (2.1) Log vacancy inflow/labour force (i minus j at t) 0.024 0.044 (4.9) (2.9) Log unemployment rate (i minus j at t) −0.009 0.001 (1.2) (0.0) Log average earnings (i minus j at t − 1) 0.081 0.308 (2.0) (2.2) Log house price (i minus j at t) −0.064 −0.212 (4.1) (3.1) Δ Log house price (i minus j at t) 0.063 0.079 (3.4) (0.9) Adj. R2 0.92 0.92 0.93 0.93 No of observations 1,045 1,045 285 285 Note: ‘t’ statistics in parentheses calculated from robust standard errors. Columns (1) and (2) have 55 region pair dummies and columns (3) and (4) have 15 region pair dummies. 19 year dummies are also included throughout but not reported. Open in new tab Table 5 Estimates of Bilateral Net Inter‐Regional In‐Migration Rates, 1982–2000 (Dependent variable: inter‐regional net migration to i from j) . 11 Regions . 6 Southern Regions . (1) . (2) . (3) . (4) . Net immigration rate (i minus j at t − 1) −0.024 −0.064 −0.128 −0.162 (0.9) (2.1) (1.5) (2.1) Log vacancy inflow/labour force (i minus j at t) 0.024 0.044 (4.9) (2.9) Log unemployment rate (i minus j at t) −0.009 0.001 (1.2) (0.0) Log average earnings (i minus j at t − 1) 0.081 0.308 (2.0) (2.2) Log house price (i minus j at t) −0.064 −0.212 (4.1) (3.1) Δ Log house price (i minus j at t) 0.063 0.079 (3.4) (0.9) Adj. R2 0.92 0.92 0.93 0.93 No of observations 1,045 1,045 285 285 . 11 Regions . 6 Southern Regions . (1) . (2) . (3) . (4) . Net immigration rate (i minus j at t − 1) −0.024 −0.064 −0.128 −0.162 (0.9) (2.1) (1.5) (2.1) Log vacancy inflow/labour force (i minus j at t) 0.024 0.044 (4.9) (2.9) Log unemployment rate (i minus j at t) −0.009 0.001 (1.2) (0.0) Log average earnings (i minus j at t − 1) 0.081 0.308 (2.0) (2.2) Log house price (i minus j at t) −0.064 −0.212 (4.1) (3.1) Δ Log house price (i minus j at t) 0.063 0.079 (3.4) (0.9) Adj. R2 0.92 0.92 0.93 0.93 No of observations 1,045 1,045 285 285 Note: ‘t’ statistics in parentheses calculated from robust standard errors. Columns (1) and (2) have 55 region pair dummies and columns (3) and (4) have 15 region pair dummies. 19 year dummies are also included throughout but not reported. Open in new tab As other studies have suggested, house prices are important variables driving inter‐regional migration. The levels effect is strongly negative, suggesting that housing costs are a key component of the cost of living (consistent with the positive coefficient on earnings). However the effect of the change in the house price index is positive. This perhaps reflects the effect on inter‐regional migration of prospective capital gains in the housing market. If immigration causes crowding out through the housing market, then its effects should be smaller when house prices are included. The increased size of the coefficient on net immigration may reflect the fact that the effect of house prices on inter‐regional migration cuts in both directions – raising living costs on the one hand, and representing prospective capital gains on the other. As we have seen the gross flows of international migration tend to be larger relative to population in the southern regions, particularly London – a fact that is also reflected in population stocks. We might expect that where immigration flows are larger their effects would be more clearly discernible. In order to explore this issue, regressions restricted to six southern regions are presented in columns (3) and (4). The regions included are: London, the Rest of the Southeast, East Anglia, the Southwest, the West Midlands and the East Midlands. The coefficient in column (3) is larger than that of column (1), consistent with the notion that immigration effects are larger in the Southeast, but it is only significant when the other variables are included, as column (4) shows. It is also notable that, with the exception of the unemployment rate, the other coefficients also increase in size for the southern regions alone. Although the main focus here is on net migration, we also explore the effects of immigration on gross inter‐regional flows. Since there are two gross flows for every net flow, there are twice as many observations. Each is treated as an inflow so that the signs of the coefficients should be the same as for the net flow estimates, but their magnitudes should be smaller for gross flows.12 Columns (1) and (2) in Table 6, for the full 2,090 observations, produce the expected sign for the immigration variable but both coefficients are insignificant. All the variables in column (2) give smaller and less significant coefficients than does the same regression for net inter‐regional migration in Table 5. For the six southern regions the coefficients in columns (3) and (4) are also smaller and less significant than those for the corresponding net flows in Table 4. While the coefficients are expected to be smaller for gross as compared to net inter‐regional flows, there also seems to be greater heterogeneity. The most that can be said is that the results for gross inter‐regional migration are not inconsistent with those for the net flows, although they are much weaker. Table 6 Estimates of Bilateral Gross Inter‐Regional In‐Migration Rates, 1982–2000 (Dependent variable: inter‐regional gross migration to i from j) . 11 Regions . 6 Southern Regions . (1) . (2) . (2) . (3) . Net immigration rate (i minus j at t − 1) −0.004 −0.003 −0.015 −0.051 (0.02) (1.2) (0.4) (1.2) Log vacancy inflow/labour force (i minus j at t) 0.016 0.029 (4.5) (3.9) Log unemployment rate (i minus j at t) −0.002 0.005 (0.5) (0.3) Log average earnings (i minus j at t − 1) 0.019 0.135 (0.9) (2.1) Log house price (i minus j at t) −0.038 −0.089 (7.5) (5.1) Δ Log house price (i minus j at t) 0.045 0.054 (5.9) (1.7) Adj. R2 0.98 0.98 0.98 0.98 No of observations 2,090 2,090 570 570 . 11 Regions . 6 Southern Regions . (1) . (2) . (2) . (3) . Net immigration rate (i minus j at t − 1) −0.004 −0.003 −0.015 −0.051 (0.02) (1.2) (0.4) (1.2) Log vacancy inflow/labour force (i minus j at t) 0.016 0.029 (4.5) (3.9) Log unemployment rate (i minus j at t) −0.002 0.005 (0.5) (0.3) Log average earnings (i minus j at t − 1) 0.019 0.135 (0.9) (2.1) Log house price (i minus j at t) −0.038 −0.089 (7.5) (5.1) Δ Log house price (i minus j at t) 0.045 0.054 (5.9) (1.7) Adj. R2 0.98 0.98 0.98 0.98 No of observations 2,090 2,090 570 570 Note: ‘t’ statistics in parentheses calculated from robust standard errors. Columns (1) and (2) include 110 region pair dummies and columns (3) and (4) include 30 region pair dummies. 19 year dummies are included throughout but not reported. Open in new tab Table 6 Estimates of Bilateral Gross Inter‐Regional In‐Migration Rates, 1982–2000 (Dependent variable: inter‐regional gross migration to i from j) . 11 Regions . 6 Southern Regions . (1) . (2) . (2) . (3) . Net immigration rate (i minus j at t − 1) −0.004 −0.003 −0.015 −0.051 (0.02) (1.2) (0.4) (1.2) Log vacancy inflow/labour force (i minus j at t) 0.016 0.029 (4.5) (3.9) Log unemployment rate (i minus j at t) −0.002 0.005 (0.5) (0.3) Log average earnings (i minus j at t − 1) 0.019 0.135 (0.9) (2.1) Log house price (i minus j at t) −0.038 −0.089 (7.5) (5.1) Δ Log house price (i minus j at t) 0.045 0.054 (5.9) (1.7) Adj. R2 0.98 0.98 0.98 0.98 No of observations 2,090 2,090 570 570 . 11 Regions . 6 Southern Regions . (1) . (2) . (2) . (3) . Net immigration rate (i minus j at t − 1) −0.004 −0.003 −0.015 −0.051 (0.02) (1.2) (0.4) (1.2) Log vacancy inflow/labour force (i minus j at t) 0.016 0.029 (4.5) (3.9) Log unemployment rate (i minus j at t) −0.002 0.005 (0.5) (0.3) Log average earnings (i minus j at t − 1) 0.019 0.135 (0.9) (2.1) Log house price (i minus j at t) −0.038 −0.089 (7.5) (5.1) Δ Log house price (i minus j at t) 0.045 0.054 (5.9) (1.7) Adj. R2 0.98 0.98 0.98 0.98 No of observations 2,090 2,090 570 570 Note: ‘t’ statistics in parentheses calculated from robust standard errors. Columns (1) and (2) include 110 region pair dummies and columns (3) and (4) include 30 region pair dummies. 19 year dummies are included throughout but not reported. Open in new tab Analysing bilateral inter‐regional flows does not give a direct estimate of the overall displacement effect of immigration to a region. Since immigrant‐induced displacement effects spill over into ten other regions the overall effects should be about ten times the size of the coefficient on the bilateral effect. However, the bilateral models estimated in Tables 5 and 6 do not account for third region effects. An alternative is to estimate the total net migration to one region from all others, giving a panel of 209 observations (19 years × 11 regions). This is closer to the model of equation (6) above where the coefficient on net immigration measures the overall displacement effect. Including the year dummies is equivalent to estimating on deviations from the cross‐sectional mean and so there is no need to define the variables as differences from the national average, since these effects will be taken out by the dummy variable for each year. Estimates of the overall effects for each region are presented in Table 7. Column (1), which excludes all other variables, gives a negative coefficient on foreign immigration, but it is not significant at conventional levels. For what it is worth, the coefficient suggests that an increase of 100 in net immigration to a region from abroad generates a net out‐migration to other regions of about 35. Thus, the displacement effect could be substantial. When other variables are added in column (2) the coefficient is of the same order of magnitude but still not significant. The coefficient on the vacancy inflows is strongly positive and the coefficients implies that a 10% increase in the vacancy inflow rate increases net in‐migration from other regions by 0.41 per thousand of the population. Unemployment and earnings both give the expected signs but neither is significant. As in the estimates of bilateral net flows, the housing variables prove to be significant and with opposite signs on the level and the change. Table 7 Estimates of the Net Regional In‐Migration Rate, 1982–2000 (Dependent variable net in‐migration) . 11 Regions . 6 Southern Regions . (1) . (2) . (3) . (4) . Net immigration rate (t − 1) −0.350 −0.300 −0.504 −0.444 (1.6) (1.2) (2.8) (2.2) Log vacancy inflow/labour force (t) 0.412 0.482 (3.8) (2.7) Log unemployment rate (t) −0.043 0.129 (0.3) (0.5) Log average earnings (t − 1) 0.012 2.232 (1.4) (1.8) Log house price (t) −0.670 −0.653 (2.6) (1.8) Δ Log house price (t) 1.249 0.785 (4.0) (1.4) Adj. R2 0.84 0.87 0.89 0.90 No of observations 209 209 114 114 . 11 Regions . 6 Southern Regions . (1) . (2) . (3) . (4) . Net immigration rate (t − 1) −0.350 −0.300 −0.504 −0.444 (1.6) (1.2) (2.8) (2.2) Log vacancy inflow/labour force (t) 0.412 0.482 (3.8) (2.7) Log unemployment rate (t) −0.043 0.129 (0.3) (0.5) Log average earnings (t − 1) 0.012 2.232 (1.4) (1.8) Log house price (t) −0.670 −0.653 (2.6) (1.8) Δ Log house price (t) 1.249 0.785 (4.0) (1.4) Adj. R2 0.84 0.87 0.89 0.90 No of observations 209 209 114 114 Note: ‘t’ statistics in parentheses calculated from robust standard errors. 11 region pair dummies are included in columns (1) and (2) and 6 region pair dummies in columns (3) and (4). 19 year dummies are included throughout but not reported. Open in new tab Table 7 Estimates of the Net Regional In‐Migration Rate, 1982–2000 (Dependent variable net in‐migration) . 11 Regions . 6 Southern Regions . (1) . (2) . (3) . (4) . Net immigration rate (t − 1) −0.350 −0.300 −0.504 −0.444 (1.6) (1.2) (2.8) (2.2) Log vacancy inflow/labour force (t) 0.412 0.482 (3.8) (2.7) Log unemployment rate (t) −0.043 0.129 (0.3) (0.5) Log average earnings (t − 1) 0.012 2.232 (1.4) (1.8) Log house price (t) −0.670 −0.653 (2.6) (1.8) Δ Log house price (t) 1.249 0.785 (4.0) (1.4) Adj. R2 0.84 0.87 0.89 0.90 No of observations 209 209 114 114 . 11 Regions . 6 Southern Regions . (1) . (2) . (3) . (4) . Net immigration rate (t − 1) −0.350 −0.300 −0.504 −0.444 (1.6) (1.2) (2.8) (2.2) Log vacancy inflow/labour force (t) 0.412 0.482 (3.8) (2.7) Log unemployment rate (t) −0.043 0.129 (0.3) (0.5) Log average earnings (t − 1) 0.012 2.232 (1.4) (1.8) Log house price (t) −0.670 −0.653 (2.6) (1.8) Δ Log house price (t) 1.249 0.785 (4.0) (1.4) Adj. R2 0.84 0.87 0.89 0.90 No of observations 209 209 114 114 Note: ‘t’ statistics in parentheses calculated from robust standard errors. 11 region pair dummies are included in columns (1) and (2) and 6 region pair dummies in columns (3) and (4). 19 year dummies are included throughout but not reported. Open in new tab The estimates in columns (3) and (4) are for the six southern regions only. When no other variables are included the displacement effect is 0.5 and is significant at the 5% level. Adding the labour market variables and house prices reduces the coefficient slightly but it remains significant. The estimate of the displacement effect for the southern regions alone suggests that for every 100 additional immigrants to these regions, 44 others are displaced to locations elsewhere in the UK. This is consistent with the findings for the bilateral flows in Table 5 and it suggests that, where immigration is sufficiently large, the effects on inter‐regional migration can be more clearly discerned. 4. Conclusion We have investigated for Britain the hypothesis that net immigration is a determinant of inter‐regional migration flows. If there are large displacement effects then this will contribute to understanding why the local labour market effects of immigration seem to be relatively benign. The evidence indicates consistently negative correlations between immigration to one region from abroad and in‐migration from other regions. But they are only significant for the southern regions where immigration of foreign citizens is most concentrated. Nevertheless they suggest that inter‐regional migration may be an important mechanism through which the British labour market adjusts to immigration. One implication of these findings is that even though the effects of immigration on unemployment and earnings in one particular region may be small, this may be a poor guide to their effects on the whole economy. The results reported here are consistent with the modest wage and employment effects of immigration at the local level reported by Card for the US and by Dustmann et al. for Britain in this Feature. However, these authors argue that inter‐regional migration is not an important adjustment mechanism and this suggests that that further research would be useful in the UK context. Although our analysis has the advantage of using directly observed flow data, it does have two major limitations. One is that we cannot examine migration at a truly local level such as travel to work areas. The other is that we are unable to disaggregate by skill, education and labour force status. Future investigation along these lines would be worthwhile but it awaits improvements in the available data. Appendix: Summary Statistics for Explanatory Variables, By Region Region . Unemployment Rate (%) . Average Earnings (£/week) . Vacancy Rate (%) . Vacancy Inflow Rate (%) . Hse Price Index 1990=100 . Change in Hse Price (%) . London 7.10 325.4 0.54 0.063 86.4 0.0854 (2.04) (129.2) (0.21) (0.015) (37.4) (0.0956) Rest of South East 5.24 261.4 0.70 0.063 84.2 0.0767 (1.99) (96.9) (0.23) (0.061) (32.4) (0.0961) East Anglia 5.69 232.3 0.66 0.064 83.6 0.0709 (1.78) (81.8) (0.21) (0.009) (29.9) (0.0984) South West 6.30 233.2 0.73 0.071 84.4 0.0708 (2.10) (82.8) (0.23) (0.011) (31.0) (0.0932) West Midlands 8.51 231.1 0.64 0.059 83.0 0.0699 (2.81) (83.0) (0.31) (0.016) (32.4) (0.0835) East Midlands 6.95 225.1 0.62 0.061 82.0 0.0688 (2.02) (79.2) (0.23) (0.009) (29.8) (0.079) Yorkshire and Humberside 8.45 226.7 0.57 0.061 83.2 0.0675 (2.20) (78.1) (0.22) (0.011) (31.3) (0.0781) North West 9.16 235.6 0.72 0.083 82.2 0.0645 (2.79) (83.7) (0.27) (0.015) (30.4) (0.072) North 10.59 225.3 0.70 0.071 87.9 0.0638 (2.72) (76.2) (0.27) (0.013) (32.0) (0.0599) Wales 9.00 224.5 0.92 0.086 83.2 0.0649 (2.72) (76.9) (0.30) (0.019) (29.6) (0.0735) Scotland 8.97 232.7 0.86 0.083 96.6 0.0616 (2.49) (80.1) (0.24) (0.012) (33.2) (0.0379) Region . Unemployment Rate (%) . Average Earnings (£/week) . Vacancy Rate (%) . Vacancy Inflow Rate (%) . Hse Price Index 1990=100 . Change in Hse Price (%) . London 7.10 325.4 0.54 0.063 86.4 0.0854 (2.04) (129.2) (0.21) (0.015) (37.4) (0.0956) Rest of South East 5.24 261.4 0.70 0.063 84.2 0.0767 (1.99) (96.9) (0.23) (0.061) (32.4) (0.0961) East Anglia 5.69 232.3 0.66 0.064 83.6 0.0709 (1.78) (81.8) (0.21) (0.009) (29.9) (0.0984) South West 6.30 233.2 0.73 0.071 84.4 0.0708 (2.10) (82.8) (0.23) (0.011) (31.0) (0.0932) West Midlands 8.51 231.1 0.64 0.059 83.0 0.0699 (2.81) (83.0) (0.31) (0.016) (32.4) (0.0835) East Midlands 6.95 225.1 0.62 0.061 82.0 0.0688 (2.02) (79.2) (0.23) (0.009) (29.8) (0.079) Yorkshire and Humberside 8.45 226.7 0.57 0.061 83.2 0.0675 (2.20) (78.1) (0.22) (0.011) (31.3) (0.0781) North West 9.16 235.6 0.72 0.083 82.2 0.0645 (2.79) (83.7) (0.27) (0.015) (30.4) (0.072) North 10.59 225.3 0.70 0.071 87.9 0.0638 (2.72) (76.2) (0.27) (0.013) (32.0) (0.0599) Wales 9.00 224.5 0.92 0.086 83.2 0.0649 (2.72) (76.9) (0.30) (0.019) (29.6) (0.0735) Scotland 8.97 232.7 0.86 0.083 96.6 0.0616 (2.49) (80.1) (0.24) (0.012) (33.2) (0.0379) Note: Average annual values, 1981–2000, standard deviations from mean values in parentheses. The unemployment rate is the claimant count divided by the workforce jobs definition of the labour force, which includes the self‐employed. The vacancy rate is the average of quarterly vacancy stocks at job centres divided by the labour force and the vacancy inflow rate is the average monthly inflow over the year divided by the labour force. These data were obtained from the NOMIS database available at: http://www.nomisweb.co.uk. Data for average annual earnings of full‐time manual employees from the New Earnings Survey were provided by the Office for National Statistics. Regional house price indices were obtained from the Office of the Deputy Prime Minister at http://www.housing.odpm.gov.uk/statistics/live/houseprices. Region . Unemployment Rate (%) . Average Earnings (£/week) . Vacancy Rate (%) . Vacancy Inflow Rate (%) . Hse Price Index 1990=100 . Change in Hse Price (%) . London 7.10 325.4 0.54 0.063 86.4 0.0854 (2.04) (129.2) (0.21) (0.015) (37.4) (0.0956) Rest of South East 5.24 261.4 0.70 0.063 84.2 0.0767 (1.99) (96.9) (0.23) (0.061) (32.4) (0.0961) East Anglia 5.69 232.3 0.66 0.064 83.6 0.0709 (1.78) (81.8) (0.21) (0.009) (29.9) (0.0984) South West 6.30 233.2 0.73 0.071 84.4 0.0708 (2.10) (82.8) (0.23) (0.011) (31.0) (0.0932) West Midlands 8.51 231.1 0.64 0.059 83.0 0.0699 (2.81) (83.0) (0.31) (0.016) (32.4) (0.0835) East Midlands 6.95 225.1 0.62 0.061 82.0 0.0688 (2.02) (79.2) (0.23) (0.009) (29.8) (0.079) Yorkshire and Humberside 8.45 226.7 0.57 0.061 83.2 0.0675 (2.20) (78.1) (0.22) (0.011) (31.3) (0.0781) North West 9.16 235.6 0.72 0.083 82.2 0.0645 (2.79) (83.7) (0.27) (0.015) (30.4) (0.072) North 10.59 225.3 0.70 0.071 87.9 0.0638 (2.72) (76.2) (0.27) (0.013) (32.0) (0.0599) Wales 9.00 224.5 0.92 0.086 83.2 0.0649 (2.72) (76.9) (0.30) (0.019) (29.6) (0.0735) Scotland 8.97 232.7 0.86 0.083 96.6 0.0616 (2.49) (80.1) (0.24) (0.012) (33.2) (0.0379) Region . Unemployment Rate (%) . Average Earnings (£/week) . Vacancy Rate (%) . Vacancy Inflow Rate (%) . Hse Price Index 1990=100 . Change in Hse Price (%) . London 7.10 325.4 0.54 0.063 86.4 0.0854 (2.04) (129.2) (0.21) (0.015) (37.4) (0.0956) Rest of South East 5.24 261.4 0.70 0.063 84.2 0.0767 (1.99) (96.9) (0.23) (0.061) (32.4) (0.0961) East Anglia 5.69 232.3 0.66 0.064 83.6 0.0709 (1.78) (81.8) (0.21) (0.009) (29.9) (0.0984) South West 6.30 233.2 0.73 0.071 84.4 0.0708 (2.10) (82.8) (0.23) (0.011) (31.0) (0.0932) West Midlands 8.51 231.1 0.64 0.059 83.0 0.0699 (2.81) (83.0) (0.31) (0.016) (32.4) (0.0835) East Midlands 6.95 225.1 0.62 0.061 82.0 0.0688 (2.02) (79.2) (0.23) (0.009) (29.8) (0.079) Yorkshire and Humberside 8.45 226.7 0.57 0.061 83.2 0.0675 (2.20) (78.1) (0.22) (0.011) (31.3) (0.0781) North West 9.16 235.6 0.72 0.083 82.2 0.0645 (2.79) (83.7) (0.27) (0.015) (30.4) (0.072) North 10.59 225.3 0.70 0.071 87.9 0.0638 (2.72) (76.2) (0.27) (0.013) (32.0) (0.0599) Wales 9.00 224.5 0.92 0.086 83.2 0.0649 (2.72) (76.9) (0.30) (0.019) (29.6) (0.0735) Scotland 8.97 232.7 0.86 0.083 96.6 0.0616 (2.49) (80.1) (0.24) (0.012) (33.2) (0.0379) Note: Average annual values, 1981–2000, standard deviations from mean values in parentheses. The unemployment rate is the claimant count divided by the workforce jobs definition of the labour force, which includes the self‐employed. The vacancy rate is the average of quarterly vacancy stocks at job centres divided by the labour force and the vacancy inflow rate is the average monthly inflow over the year divided by the labour force. These data were obtained from the NOMIS database available at: http://www.nomisweb.co.uk. Data for average annual earnings of full‐time manual employees from the New Earnings Survey were provided by the Office for National Statistics. Regional house price indices were obtained from the Office of the Deputy Prime Minister at http://www.housing.odpm.gov.uk/statistics/live/houseprices. Footnotes 1 " The earlier literature, particularly that for the US, has been summarised by Friedberg and Hunt (1995) and Borjas (1997) among others. 2 " However Pischke and Velling (1997) find little evidence of any effects of immigration on unemployment across local labour markets in Germany. 3 " Borjas (1997) puts it thus: ‘One could easily argue that this literature has failed to increase our understanding of how labour markets respond to immigration. If we take the empirical evidence … at face value, the implications are disturbing: either we need different economic models to understand how supply shocks affect labour markets in different periods … or the regression coefficients are simply not measuring what we think they should be measuring’ (p. 1740). 4 " Another possibility is that immigration shifts labour demand upwards. In that case the coefficient on mi,t should be more negative when is added to the equation, since the latter should capture the immigrant‐induced demand shift. Chapman and Cobb‐Clark (1999) examine the conditions under which the demand effects of immigration will outweigh the labour supply effects on the job prospects of residents. 5 " For Europe it has been found that adjustment occurs largely through variations in participation (Decressin and Fatás, 1995) although recent evidence suggests that migration may be more important than previously thought (Tani, 2003). These studies, however, focus on the effects of demand shocks rather than on the supply shocks from immigration that are at issue here. To date there has been little work in this vein that integrates shocks from both sides of the market. 6 " The IPS data is generally thought to underestimate some classes of immigrants such as asylum seekers and visitor switchers (those who arrive initially for a short period but stay longer). Adjusting for these and for the movement of Irish nationals who are not enumerated in the survey, the Office for National Statistics estimates net immigration of foreign citizens at 135,400 per annum 1991–2000. However, adjusted estimates are not available back to the 1980s and so we rely here on the unadjusted figures. 7 " The long‐term upward trend in net immigration across all citizenships since the early 1970s is partly due to the fall in net emigration among UK citizens (Hatton, 2004). 8 " Conurbations outside London also exhibit net gains from overseas and net losses to the rest of the UK but these are masked in the wider regions (Champion, 1999). 9 " When instrumental variables are used the negative coefficients on immigration were always larger than in the OLS regressions in Tables 5–7, although the standard errors were also larger. 10 " Thus, for example, Saiz (2003) finds that housing rentals increased in Miami in the wake of the Mariel Boatlift immigration. However, we find that lagging or eliminating house prices makes very little difference to our estimate of the coefficient on immigration. 11 " Some studies find evidence that spatial inter‐dependence between labour markets increases with contiguity and declines with distance, presumably due to migration; see Burgess and Profit (2001) for evidence on matching functions for travel to work areas in Britain. In initial experiments, we scaled the explanatory variables by the share of each bilateral flow in total interregional flows over the whole period, so that the coefficients effectively become larger as the degree of association between two regions increases. However, this did not improve the explanatory power of the model. As an alternative, we provide separate estimates for the southern regions alone. 12 " Thus the gross flow from j to i is modelled as gij,t =β(xi,t − xj,t) + γ(mi,t−1 − mj,t−1) + μij,t, and the gross flow from i to j as gji,t = β(xj,t − xi,t) + γ(mj,t−1 − mi,t−1) + μji,t. These theoretical parameters are related to those in equation (6) above such that: λ = 2β and ϕ = 2γ. References Altonji , J. G. and Card , D. ( 1991 ). ‘The effect of immigration on the labor market outcomes of less‐skilled natives’, in ( J. M. Abowd and R. B. Freeman eds.) Immigration, Trade and the Labor Market , Chicago: University of Chicago Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Angrist , J. D. and Kugler , A. D. ( 2003 ). ‘Productive or counterproductive? Labour market institutions and the effect of immigration on EU natives’ , Economic Journal vol. 113 (June), pp. F302 – 37 . Google Scholar Crossref Search ADS WorldCat Borjas , G. J. ( 1997 ). ‘The economic analysis of immigration’, in ( O. Ashenfelter and D. Card, eds.) Handbook of Labor Economics , vol. 3A, New York: North Holland . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Borjas G. J. ( 2003 ). ‘The labor demand curve is downward sloping: re‐examining the impact of immigration on the labor market’ , Quarterly Journal of Economics , vol. 118 (November), pp. 1135 – 74 . Google Scholar Crossref Search ADS WorldCat Burgess , S. and Profit , S. ( 2001 ). ‘Externalities in the matching of workers and firms in Britain’ , Labour Economics , vol. 8 (June), pp. 313 – 33 . Google Scholar Crossref Search ADS WorldCat Card , D. E. ( 1990 ). ‘The impact of the Mariel Boatlift on the Miami labor market’ , Industrial and Labor Relations Review , vol. 43 (January), pp. 245 – 57 . Google Scholar Crossref Search ADS WorldCat Card , D. E. ( 2001 ). ‘Immigrant inflows, native outflows and the local labor market impacts of higher immigration’ , Journal of Labor Economics , vol. 19 (January), pp. 22 – 64 . Google Scholar Crossref Search ADS WorldCat Card , D. E. (this Feature). ‘Is the new immigration really so bad?’ . Card , D. E. and DiNardo , J. E. ( 2000 ). ‘Do immigrant inflows lead to native outflows?’ , American Economic Review , vol. 90 (May), pp. 360 – 73 . Google Scholar Crossref Search ADS WorldCat Carrington , W. J. and di Lima , P. J. F. ( 1996 ). ‘The impact of 1970s repatriates from Africa on the Portuguese labor market’ , Industrial and Labor Relations Review , vol. 49 (January), pp. 330 – 47 . Google Scholar Crossref Search ADS WorldCat Champion , T. ( 1999 ). ‘Migration and British cities in the 1990s’ , National Institute Economic Review , vol. 170 (October), pp. 60 – 77 . Google Scholar Crossref Search ADS WorldCat Chapman , B. and Cobb‐Clark , D. ( 1999 ). ‘A comparative static model of the relationship between immigration and the short‐run job prospects of unemployed residents’ , Economic Record , vol. 75 (December), pp. 358 – 68 . Google Scholar Crossref Search ADS WorldCat Coles , M. G. and Smith , E. F. ( 1998 ). ‘Marketplaces and matching’ , International Economic Review , vol. 39 (February), pp. 239 – 54 . Google Scholar Crossref Search ADS WorldCat Decressin , J. and Fatás , A. ( 1995 ). ‘Regional labour market dynamics in Europe’ , European Economic Review , vol. 39 (December), pp. 1627 – 55 . Google Scholar Crossref Search ADS WorldCat De New , J. and Zimmermann , K. F. ( 1994 ). ‘Native wage impacts of foreign labor: a random effects panel analysis’ , Journal of Population Economics , vol. 7 (June), pp. 177 – 92 . Google Scholar Crossref Search ADS PubMed WorldCat Dustmann , C. , Fabbri , F. and Preston , I. (this Feature). ‘The local labour market effects of immigration in the UK’ . Filer , R. K. ( 1992 ). ‘The impact of immigrant arrivals on migratory patterns of native workers’, in ( G. J. Borjas and R. B. Freeman, eds.) Immigration and the Work Force: Economic Consequences for the United States and Source Areas , Chicago: University of Chicago Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Friedberg , R. M. and Hunt , J. ( 1995 ). ‘The impact of immigrants on host country wages, employment and growth’ , Journal of Economic Perspectives , vol. 9 (Winter), pp. 23 – 44 . Google Scholar Crossref Search ADS WorldCat Frijters , P. , Shields , M. A. and Wheatley Price , S. (this Feature). ‘Job search methods and their success: a comparison of native and immigrants in the UK’ . Hatton , T. J. ( 2004 ). ‘Emigration from the UK, 1870–1913 and 1950–1998’ , European Review of Economic History , vol. 8 (August), pp. 149 – 71 . Google Scholar Crossref Search ADS WorldCat Hatton , T. J. ( 2005 ). ‘Explaining trends in UK immigration’ , Journal of Population Economics , (forthcoming). OpenURL Placeholder Text WorldCat Hughes , G. and McCormick , B. ( 1994 ). ‘Did migration in the 1980s narrow the North‐South divide?’ , Economica , vol. 61 (November), pp. 509 – 27 . Google Scholar Crossref Search ADS WorldCat Hunt , J. ( 1992 ). ‘The impact of the 1962 repatriates from Algeria on the French Labor Market’ , Industrial and Labor Relations Review , vol. 45 (April), pp. 556 – 72 . Google Scholar Crossref Search ADS WorldCat Jackman , R. and Savouri , S. ( 1992 ). ‘Regional migration in Britain: an analysis of gross flows using NHS Central Register data’ , Economic Journal , vol. 102 (November), pp. 1433 – 50 . Google Scholar Crossref Search ADS WorldCat Katz and Krueger ( 1992 ). ‘The effect of the minimum wage on the fast food industry’ , NBER Working Papers 3997. LaLonde , R. J. and Topel , R. H. ( 1991 ). ‘Labor market adjustments to increased immigration’, in ( J. M. Abowd and R. B. Freeman, eds.) Immigration, Trade and the Labor Market , Chicago: University of Chicago Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC McCormick , B. ( 1997 ). ‘Regional unemployment and labour mobility in the UK’ , European Economic Review , vol. 41 (April), pp. 581 – 9 . Google Scholar Crossref Search ADS WorldCat Muellbauer , J. and Cameron , G. ( 1998 ). ‘The housing market and regional commuting and migration choices’ , Scottish Journal of Political Economy , vol. 45 (September), pp. 420 – 46 . Google Scholar Crossref Search ADS WorldCat Muellbauer , J. and Murphy , A. ( 1988 ). ‘UK house prices and migration: economic and investment implications’ , Shearson Lehman Hutton Discussion Paper. Pischke , J‐S. and Velling , J. ( 1997 ). ‘Employment effects of immigration to Germany: an analysis based on local labor markets’ , Review of Economics and Statistics , vol. 79 (November), pp. 594 – 604 . Google Scholar Crossref Search ADS WorldCat Pissarides , C. A. and McMaster , I. ( 1990 ). ‘Regional migration, wages and unemployment: empirical evidence and implications for policy’ , Oxford Economic Papers , vol. 42 (October), pp. 812 – 23 . Google Scholar Crossref Search ADS WorldCat Pissarides , C. A. and Wadsworth , J. ( 1989 ). ‘Unemployment and the inter‐regional mobility of labour’ , Economic Journal , vol. 99 (September), pp. 739 – 55 . Google Scholar Crossref Search ADS WorldCat Saiz , A. ( 2003 ). ‘Room in the kitchen for the melting pot: immigration and rental prices’ , Review of Economics and Statistics , vol. 85 (August), pp. 502 – 21 . Google Scholar Crossref Search ADS WorldCat Tani , M. ( 2003 ). ‘Have Europeans become more mobile? A note on regional evolutions in the EU: 1988–1997’ , Economic Letters , vol. 80 (July), pp. 23 – 30 . Google Scholar Crossref Search ADS WorldCat Winklemann R. and Zimmermann , K. F. ( 1993 ). ‘Ageing, migration and labour mobility’, in ( P. A. Johnson and K. F. Zimmermann, eds.) Labour Markets in an Ageing Europe , pp. ? , Cambridge: Cambridge University Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Author notes " We thank members of the Office for National Statistics for providing some of the data used here, in particular Mike Harris, Alison Snell, Ian Parker, Robert Palmer, Kanak Ghosh, Rhian Tyler, Alastair Davies and Chris Hunt. Hatton also thanks the British Academy for support through a Research Readership. We are grateful to Alison Booth, Deborah Cobb‐Clark and to participants at the Royal Economic Society Conference at Warwick, 2003, for many useful comments. We have also benefited from valuable comments from the editor and three anonymous referees. © Royal Economic Society 2005.

doi: 10.1111/j.1468-0297.2005.01043_2.xpmid: N/A

Article PDF first page preview Close This content is only available as a PDF. © Royal Economic Society 2005.

Card,, David

doi: 10.1111/j.1468-0297.2005.01037.xpmid: N/A

Abstract This article reviews the recent evidence on US immigration, focusing on two key questions: (1) Does immigration reduce the labour market opportunities of less skilled natives? (2) Have immigrants who arrived after the 1965 Immigration Reform Act been successfully assimilated? Overall, evidence that immigrants have harmed the opportunities of less educated natives is scant. On the question of assimilation, the success of the US‐born children of immigrants is a key yardstick. By this metric, post‐1965 immigrants are doing reasonably well. Even children of the least educated immigrant origin groups have closed most of the education gap with the children of natives. Over the past two decades economists’ perceptions of US immigrants have shifted. In the 1970s, immigrants were viewed in a mainly positive light. Chiswick (1978) found that immigrant men earned as much as natives, despite having less education, and concluded that investments in on‐the‐job training made up for the gap in formal schooling. Grossman (1982) examined the impact of immigration on native wages and concluded that the effects were small. Subsequent research – most notably by Borjas (1985, 1995, 1999, 2003)– has chipped away at both conclusions and gradually led to a more negative picture of US immigration. The shift in perceptions has closely tracked changes in the national origin of US immigrants, often attributed to the 1965 Immigration Reform Act, and a widening gap between the language and culture of natives and immigrants (Borjas, 1999; Lazear, 1999). Concerns over immigration have also been heightened by the decline in low skilled wages in the US and the belief that some of this may be due to immigrant competition (Borjas et al., 1997). This article presents an overview and update of the US immigration literature, focusing on two central questions: Do immigrants harm the labour market opportunities of less skilled natives? How do today's immigrants perform in the US labour market, and are they successfully being ‘assimilated'? These questions are at the heart of the debate about immigration in many other countries – including most European nations – and insights from the recent US literature may prove useful in answering the questions elsewhere. My conclusion is that the ‘revisionist’ view of recent US immigration is overly pessimistic. The evidence that immigrants harm native opportunities is slight, while the fear that post‐1965 immigrants will never assimilate is belied by the rather surprising educational success of their children. 1. The Characteristics of Immigrants Most of the immigrants in Chiswick's (1978) landmark study had entered the US under the provisions of the Immigrant and Nationality Acts of 1924, which established national origin quotas with a strong bias in favour of Northern Europeans.1 In the 1970 data analysed by Chiswick, 63% of immigrants were born in Europe or Canada (Card et al., 2000, Table 6.3). The vast majority of working age immigrants in the US today arrived after the 1965 Immigration Act, which relaxed the quota system and established preferences for people with family members already in the country. The new law, coupled with declining supplies of potential immigrants from traditional source countries such as Britain, Germany and Italy, and increasing potential supplies from Mexico, Central America and Asia, have led to a shift in the ethnic composition of immigrants.2 In 2000, only 13.6% of adult immigrants in the US were born in Europe, while 32% were born in Mexico, 16% in Central America or the Carribean, and 26.6% in Asia. As emphasised by Borjas (1985, 1995), the skill characteristics of immigrants in the US are strongly related to their country of origin. For example, country of origin dummies explain 30% of the variation in average education levels among immigrants in the 2000 Census.3 Reflecting the high fraction of immigrants from countries like Mexico, the Philippines, Vietnam and El Salvador (all of which supply relatively low education immigrants) immigrants as a whole have lower average schooling than natives. Table 1 compares the education distributions of natives and two subgroups of immigrants – those who had been in the US at least 5 years at the time of the 2000 Census, and those who had arrived more recently. The recent arrival group is not very different from the earlier arrivals, reflecting the relative stability of immigrant inflow composition since the late 1970s. Nevertheless, both groups have a much higher fraction of people with very low schooling than natives. The excess concentration of immigrants in the ‘less than high school’ category is balanced by relative shortfalls in the number who completed high school but have no further formal schooling (24% of immigrants versus 39% of natives) and the number with 1–3 years of college (about 15% of immigrants versus 24% of natives). At the upper end of the education distribution immigrants and natives are very similar, though immigrants are slightly more likely than natives to hold an advanced degree. Table 1 Educational Attainment of Natives and Immigrants in the 2000 Census Highest Education . All . Natives . Immigrants: . All . In US 5+ yrs. . In US <5 yrs. . Dropouts 17.8 14.7 38.2 37.6 40.3 Of which:  1–8 Years completed schooling 5.3 2.8 21.5 21.3 22.2  9–11 Years completed schooling 12.6 11.9 16.7 16.3 18.1 High school diploma 37.2 39.2 24.0 24.3 23.1 Some college (including Associates degree) 22.6 23.7 15.5 16.3 12.5 Bachelors degree 14.8 15.0 13.2 12.9 14.0 Advanced degree 7.7 7.5 9.2 8.8 10.1 Of which:  Masters degree 5.2 5.2 5.4 5.1 6.3  Professional degree 1.7 1.6 2.3 2.2 2.4  Doctorate 0.8 0.7 1.5 1.5 1.4 Highest Education . All . Natives . Immigrants: . All . In US 5+ yrs. . In US <5 yrs. . Dropouts 17.8 14.7 38.2 37.6 40.3 Of which:  1–8 Years completed schooling 5.3 2.8 21.5 21.3 22.2  9–11 Years completed schooling 12.6 11.9 16.7 16.3 18.1 High school diploma 37.2 39.2 24.0 24.3 23.1 Some college (including Associates degree) 22.6 23.7 15.5 16.3 12.5 Bachelors degree 14.8 15.0 13.2 12.9 14.0 Advanced degree 7.7 7.5 9.2 8.8 10.1 Of which:  Masters degree 5.2 5.2 5.4 5.1 6.3  Professional degree 1.7 1.6 2.3 2.2 2.4  Doctorate 0.8 0.7 1.5 1.5 1.4 Note. Based on tabulations of individuals age 18–64 in the 2000 Census. High school diploma group includes people with less than 1 year of college (8% of the overall sample). Open in new tab Table 1 Educational Attainment of Natives and Immigrants in the 2000 Census Highest Education . All . Natives . Immigrants: . All . In US 5+ yrs. . In US <5 yrs. . Dropouts 17.8 14.7 38.2 37.6 40.3 Of which:  1–8 Years completed schooling 5.3 2.8 21.5 21.3 22.2  9–11 Years completed schooling 12.6 11.9 16.7 16.3 18.1 High school diploma 37.2 39.2 24.0 24.3 23.1 Some college (including Associates degree) 22.6 23.7 15.5 16.3 12.5 Bachelors degree 14.8 15.0 13.2 12.9 14.0 Advanced degree 7.7 7.5 9.2 8.8 10.1 Of which:  Masters degree 5.2 5.2 5.4 5.1 6.3  Professional degree 1.7 1.6 2.3 2.2 2.4  Doctorate 0.8 0.7 1.5 1.5 1.4 Highest Education . All . Natives . Immigrants: . All . In US 5+ yrs. . In US <5 yrs. . Dropouts 17.8 14.7 38.2 37.6 40.3 Of which:  1–8 Years completed schooling 5.3 2.8 21.5 21.3 22.2  9–11 Years completed schooling 12.6 11.9 16.7 16.3 18.1 High school diploma 37.2 39.2 24.0 24.3 23.1 Some college (including Associates degree) 22.6 23.7 15.5 16.3 12.5 Bachelors degree 14.8 15.0 13.2 12.9 14.0 Advanced degree 7.7 7.5 9.2 8.8 10.1 Of which:  Masters degree 5.2 5.2 5.4 5.1 6.3  Professional degree 1.7 1.6 2.3 2.2 2.4  Doctorate 0.8 0.7 1.5 1.5 1.4 Note. Based on tabulations of individuals age 18–64 in the 2000 Census. High school diploma group includes people with less than 1 year of college (8% of the overall sample). Open in new tab The patterns in Table 1 point to two important conclusions. First, labour market competition from immigrants is most intense for natives with the lowest levels of education. While immigrants comprised only 13% of the working age population in 2000, they made up 28% of the population with less than a high school diploma, and over half of all those with less than 8 years of schooling. For this reason, most studies of immigrant competition have focused on the impacts on very low skilled natives.4 Second, the positive effect of immigrants on the relative supply of the people with the lowest levels of education is offset by negative effects on the relative supply of people in the middle of the education distribution, with no effect on the relative supply of those with a bachelor's degree or higher. Arguably, then, immigrant inflows have exerted upward pressure on the wage gap between high school graduates and dropouts, and downward pressure on the college–high school wage gap. 2. Immigrant Competition and The Labour Market Outcomes of Low Skilled Natives 2.1. Conceptual Issues There are two main approaches in the literature to estimating the impact of immigration on native workers. The first – pioneered by Grossman (1982)– relates differences in the relative structure of wages in different local labour markets to differences in the relative supply of immigrants.5 The advantage of this approach is that there are many local labour markets in the US with different fractions of immigrants and samples from the Decennial Censuses can be used to estimate relatively rich models of the local wage structure. The disadvantage is that cities are not isolated economies: people, goods and services all flow between cities, and depending on how sensitive these flows are to differences in local wages or prices, comparisons across cities may reveal a lot or little about the underlying parameters that theoretically determine the effects of immigration on native opportunities. The second approach is a time series method, relating changes over time in immigrant densities to economy‐wide measures of relative labour market outcomes. The advantage of this approach is that it can potentially reveal the impact of immigration even when the local markets approach ‘fails’ because of intercity factor mobility or trade (Borjas et al., 1996).6 The disadvantage is the absence of a clear counterfactual. Inferences from the macro time series approach rely on assumptions about the trends in factors like the degree of skill bias in recent technological change. Early studies using the local labour markets approach (Grossman, 1982; Borjas, 1987; Altonji and Card, 1991; Lalonde and Topel, 1991) treated ‘immigrants’ as one type of labour and distinguished between various subgroups of natives in the same city. While simple and intuitively appealing, there are two obvious problems with this framework. One is immigrant heterogeneity: in some cities immigrants are actually more highly skilled than natives, whereas in others the reverse is true.7 Thus, it is important to classify the immigrant populations in different cities according to their skill levels. A second problem is that conventional economic models imply that immigrants should only affect relative wages to the extent that they distort the relative supplies of different skill groups. If inflows of unskilled immigrants cause unskilled natives to move out, for example, there may be little discernable effect of immigration on the local wage structure, even though relative demand curves at the local level are downward sloping. A potentially better way to model the impact of immigration is to assign immigrants and natives to skill groups and to assume that within skill groups, immigrants and natives are perfect substitutes (Jaeger, 1996; Card, 2001).8 Following this approach, the first step in evaluating the impact of immigration is to assess the effect of immigrants on the relative supplies of different skill groups in different cities. The second step is to then relate the relative wages for different skill groups to the relative supplies in the local labour market. The maintained assumption – that immigrants and natives are perfect substitutes within skill groups – can be tested by examining the stability of immigrant–native wage differences across different labour markets.9 2.2. Effects of Immigration On Relative Supply of Low Skilled Labour Some indication of the impact of immigration on the relative size of the dropout labour force in different cities is presented in Table 2. The Table shows the fractions of immigrants in all cities and in 15 selected cities in 1980 and 2000, along with the fractions of immigrants and natives with less than 12 years of completed schooling, and the overall fractions of the working age population in each city with less than 12 years of schooling. The data for all cities in the first row of the Table reveals three interesting facts.10 First, the fraction of immigrants in US cities has roughly doubled since 1980, from 9.5% to 18%. Second, in both 1980 and 2000, slightly more than one‐third of immigrants had less than a high school education. Third, the fraction of natives with less than a high school education has fallen sharply, more than offsetting the inflow of less educated immigrants.11 Thus, despite the upward pressure on the relative supply of dropouts caused by immigration, the overall fraction of dropouts in urban areas fell from 24.3% in 1980 to 17.7% in 2000. Table 2 Immigrant Densities and the Relative Fractions of Less Educated Workers, Selected Cities 1980 and 2000 . % Dropouts 1980 . % Dropouts 2000 . % Immigrants in city . Among immigrants . Among natives . % Dropouts in city . % Immigrants in city . Among immigrants . Among natives . %Dropouts in city . All Cities 9.5 38.9 23.0 24.3 18.0 37.8 13.0 17.7 New York 23.2 39.6 26.4 29.5 41.8 32.0 17.5 23.6 Los Angeles 25.3 49.2 19.5 27.0 47.8 47.2 14.4 30.1 Chicago 11.8 44.0 23.7 26.1 21.2 37.7 11.8 17.3 Philadelphia 4.9 31.1 25.2 25.5 8.3 21.9 13.3 14.0 Detroit 6.3 34.3 25.8 26.4 8.6 26.2 14.4 15.5 Houston 9.4 46.1 25.1 27.1 26.0 51.6 15.5 24.9 Dallas 5.1 43.7 24.3 25.3 19.7 54.2 13.6 21.6 Washington DC 9.6 18.3 16.8 16.9 20.6 25.8 9.9 13.2 Boston 10.3 35.6 15.6 17.6 17.8 24.0 7.9 10.7 San Francisco 17.0 28.4 14.3 16.7 36.4 26.6 6.9 14.0 Miami 41.1 38.5 23.3 29.6 61.2 33.3 18.6 27.6 Atlanta 3.1 14.8 24.9 24.6 12.1 34.0 13.6 16.1 Pittsburgh 2.6 28.1 21.5 21.7 2.6 12.5 10.4 10.5 Cleveland 5.8 34.5 24.0 24.6 5.6 19.7 14.2 14.5 . % Dropouts 1980 . % Dropouts 2000 . % Immigrants in city . Among immigrants . Among natives . % Dropouts in city . % Immigrants in city . Among immigrants . Among natives . %Dropouts in city . All Cities 9.5 38.9 23.0 24.3 18.0 37.8 13.0 17.7 New York 23.2 39.6 26.4 29.5 41.8 32.0 17.5 23.6 Los Angeles 25.3 49.2 19.5 27.0 47.8 47.2 14.4 30.1 Chicago 11.8 44.0 23.7 26.1 21.2 37.7 11.8 17.3 Philadelphia 4.9 31.1 25.2 25.5 8.3 21.9 13.3 14.0 Detroit 6.3 34.3 25.8 26.4 8.6 26.2 14.4 15.5 Houston 9.4 46.1 25.1 27.1 26.0 51.6 15.5 24.9 Dallas 5.1 43.7 24.3 25.3 19.7 54.2 13.6 21.6 Washington DC 9.6 18.3 16.8 16.9 20.6 25.8 9.9 13.2 Boston 10.3 35.6 15.6 17.6 17.8 24.0 7.9 10.7 San Francisco 17.0 28.4 14.3 16.7 36.4 26.6 6.9 14.0 Miami 41.1 38.5 23.3 29.6 61.2 33.3 18.6 27.6 Atlanta 3.1 14.8 24.9 24.6 12.1 34.0 13.6 16.1 Pittsburgh 2.6 28.1 21.5 21.7 2.6 12.5 10.4 10.5 Cleveland 5.8 34.5 24.0 24.6 5.6 19.7 14.2 14.5 Note. Based on tabulations of 1980 and 2000 Census public use files. ‘All cities’ includes 272 Standard Metropolitan Areas in 1980 and 325 Metropolitan Statistical Areas in 2000. Boundaries of some cities change between 1980 and 2000. Samples include individuals age 18–64 only. Open in new tab Table 2 Immigrant Densities and the Relative Fractions of Less Educated Workers, Selected Cities 1980 and 2000 . % Dropouts 1980 . % Dropouts 2000 . % Immigrants in city . Among immigrants . Among natives . % Dropouts in city . % Immigrants in city . Among immigrants . Among natives . %Dropouts in city . All Cities 9.5 38.9 23.0 24.3 18.0 37.8 13.0 17.7 New York 23.2 39.6 26.4 29.5 41.8 32.0 17.5 23.6 Los Angeles 25.3 49.2 19.5 27.0 47.8 47.2 14.4 30.1 Chicago 11.8 44.0 23.7 26.1 21.2 37.7 11.8 17.3 Philadelphia 4.9 31.1 25.2 25.5 8.3 21.9 13.3 14.0 Detroit 6.3 34.3 25.8 26.4 8.6 26.2 14.4 15.5 Houston 9.4 46.1 25.1 27.1 26.0 51.6 15.5 24.9 Dallas 5.1 43.7 24.3 25.3 19.7 54.2 13.6 21.6 Washington DC 9.6 18.3 16.8 16.9 20.6 25.8 9.9 13.2 Boston 10.3 35.6 15.6 17.6 17.8 24.0 7.9 10.7 San Francisco 17.0 28.4 14.3 16.7 36.4 26.6 6.9 14.0 Miami 41.1 38.5 23.3 29.6 61.2 33.3 18.6 27.6 Atlanta 3.1 14.8 24.9 24.6 12.1 34.0 13.6 16.1 Pittsburgh 2.6 28.1 21.5 21.7 2.6 12.5 10.4 10.5 Cleveland 5.8 34.5 24.0 24.6 5.6 19.7 14.2 14.5 . % Dropouts 1980 . % Dropouts 2000 . % Immigrants in city . Among immigrants . Among natives . % Dropouts in city . % Immigrants in city . Among immigrants . Among natives . %Dropouts in city . All Cities 9.5 38.9 23.0 24.3 18.0 37.8 13.0 17.7 New York 23.2 39.6 26.4 29.5 41.8 32.0 17.5 23.6 Los Angeles 25.3 49.2 19.5 27.0 47.8 47.2 14.4 30.1 Chicago 11.8 44.0 23.7 26.1 21.2 37.7 11.8 17.3 Philadelphia 4.9 31.1 25.2 25.5 8.3 21.9 13.3 14.0 Detroit 6.3 34.3 25.8 26.4 8.6 26.2 14.4 15.5 Houston 9.4 46.1 25.1 27.1 26.0 51.6 15.5 24.9 Dallas 5.1 43.7 24.3 25.3 19.7 54.2 13.6 21.6 Washington DC 9.6 18.3 16.8 16.9 20.6 25.8 9.9 13.2 Boston 10.3 35.6 15.6 17.6 17.8 24.0 7.9 10.7 San Francisco 17.0 28.4 14.3 16.7 36.4 26.6 6.9 14.0 Miami 41.1 38.5 23.3 29.6 61.2 33.3 18.6 27.6 Atlanta 3.1 14.8 24.9 24.6 12.1 34.0 13.6 16.1 Pittsburgh 2.6 28.1 21.5 21.7 2.6 12.5 10.4 10.5 Cleveland 5.8 34.5 24.0 24.6 5.6 19.7 14.2 14.5 Note. Based on tabulations of 1980 and 2000 Census public use files. ‘All cities’ includes 272 Standard Metropolitan Areas in 1980 and 325 Metropolitan Statistical Areas in 2000. Boundaries of some cities change between 1980 and 2000. Samples include individuals age 18–64 only. Open in new tab There is a lot of cross‐city variation in these patterns, however. The dramatic increase in the fraction of low education immigrants in Los Angeles, for example, led to a rise in the fraction of high school dropouts in the local population. In Pittsburgh and Cleveland, on the other hand, immigrant densities are low and have fallen over the past two decades, so the the trend in the overall fraction of dropouts closely parallels the trend among natives. Most high immigration cities, including New York, Houston, San Francisco and Miami, experienced relatively small declines in the fraction of dropouts between 1980 and 2000, whereas most low immigration cities, including Philadelphia, Detroit and Atlanta, experienced bigger reductions. The question of whether inflows of unskilled immigration have systematically affected the relative supply of dropout labour in different cities is addressed in Figure 1. As motivation for this Figure, note that the share of dropouts in the local working age population in city c, sd(c), is the sum of the share of native dropouts sdN(c) and the share of immigrant dropouts, sdI(c). An interesting descriptive regression relates the overall dropout share in a city to the share of immigrant dropouts: where e(c) is a residual. If inflows of less educated immigrants are offset by outflows of native dropouts (or if less educated immigrants tend to move to cities where there is a bigger positive trend in the educational attainment of the native population), immigration will have little impact on the overall dropout share and the coefficient β will be close to 0. If mobility flows of native dropouts (and trends in native educational attainment) are uncorrelated with the inflow rate of low skilled immigrants, the coefficient β will be close to 1. Fig. 1. Open in new tabDownload slide Fraction of Immigrant Dropouts and Overall Fraction of Dropouts Fig. 1. Open in new tabDownload slide Fraction of Immigrant Dropouts and Overall Fraction of Dropouts As suggested by the data in Table 2, the scatter of points in Figure 1 is more consistent with a value of β = 1 than β = 0. For reference, the graph shows a restricted regression line in which the slope is set to 1. This benchmark provides a reasonable fit, although it tends to under‐predict the fraction of dropouts in cities with few immigrant dropouts. In fact, a univariate regression across 325 cities yields an estimate of β equal to 0.79 (with a standard error of 0.03). When controls are added for city size and the fraction of blacks in the city population, the estimate of β rises to 1.01 (with a standard error of 0.03).12 Findings similar to those in Figure 1 are reported in Card (2001), using data for 175 cities from the 1990 Census, and defining low skilled workers as those who are predicted to work in low wage occupations based on their age, education, gender, race, ethnicity, and country of origin. In that paper I looked specifically at mobility responses of natives to recent immigrant inflows, and concluded that each new immigrant in the lowest skill group adds about 1 to the net supply of low skilled workers in a city. Focusing on longer term mobility, Card and DiNardo (2000) use a three skill group taxonomy to examine the effect of immigrant inflows on native migration rates between 1980 and 1990. Again, the conclusion is that native mobility has virtually no offsetting effect on the relative supply shocks created by immigration. Indeed, once controls are introduced for city‐specific trends in native population growth, the data suggest that native mobility responses may slightly reinforce the relative supply effects of immigration (Card and DiNardo, 2000, Table 2). A concern with the interpretation of the data in Figure 1 is that unskilled immigrants may be drawn to cities where the relative demand for dropout labour is increasing. One way to address this concern partially is to use historical immigration patterns as instruments for current inflows (Altonji and Card, 1991; Card, 2001). For example, Card and DiNardo (2000, Table 2) report models in which low skilled immigration inflows between 1980 and 1990 are instrumented by the fraction of Mexican immigrants in the city in 1970. These models give no indication that the OLS estimates are biased by local demand shocks. Although the results in Figure 1 suggest that immigration has a powerful effect on local labour markets, some researchers have argued the opposite. Borjas et al. (1997), for example, claim that native mobility effectively undoes any local impact of immigrant inflows.13 Importantly, however, Borjas et al. (1997) focus only on total population, not on the relative size of different skill groups. Looking at the California and Texas cities in Table 2 it is very hard to argue that immigration has not had some impact on the fraction of less educated people in the local labour market. To the best of my knowledge, in fact, all studies that have looked at the relative supply impacts of immigration find very large effects on local labour markets. 2.3. Impacts On Less Skilled Natives Once it is established that low skilled immigration increases the relative supply of unskilled workers in local labour markets, the next step is to measure the effects on the relative labour market outcomes of less skilled workers. A simple theoretical framework for this analysis consists of a local production function and a set of per capita labour supply functions for members of each skill group.14 Such a model implies that the relative wages and relative employment rates of workers in any two skill groups depend on the relative fractions of the groups in the local population. For example, comparing high school graduates to dropouts, a relative supply/demand model implies that (1a) (1b) where wd and wH are the mean wages of dropouts and high school graduates in a city, Ed and EH are the mean employment‐population rates of the two groups, and sd and sH are the shares of dropouts and high school graduates in the local population. The coefficients b1 and b2 depend on the elasticity of substitution between skill groups (σ) and on the elasticity of the per‐capita labour supply functions (ɛ): If local labour supplies are perfectly inelastic (ɛ = 0) then (1a) reduces to the familiar model used in studies of education‐based wage gaps (Katz and Murphy, 1992). Apart from the potential problem posed by unobserved relative demand and supply shocks, the key problem for interpreting estimates based on (1a) and (1b) is that the slope of the local relative demand curve may be uninformative about the degree of substitutability between skill groups. In particular, a Hecksher‐Olin style model of local economies suggests that relative wages may be uncorrelated with relative labour supplies, even though at the national level relative wages are negatively related to relative supplies. I return to this point below. Some simple reduced form evidence on the impact of unskilled immigration on relative wages and relative employment of low skilled natives is presented in Figures 2 and 3. Figure 2 shows the gap in mean log wages between native male high school graduates and native male dropouts in each of the 175 largest US cities in 2000, plotted against the fraction of immigrant dropouts in the local labour market.15Figure 3 shows the log of the ratio of the employment‐population rates of high school graduates and dropouts, also plotted against the fraction of immigrant dropouts.16 The graph for relative wages shows little connection between native wages and the fraction of immigrant dropouts, while the graph for relative employment suggests a slightly positive correlation. Estimated regression models fitted over all 325 cities confirm this impression: the estimated regression slope for wages is statistically insignificant (slope = −0.06; standard error = 0.06) while the estimated slope for relative employment is significantly positive (slope = 0.07, standard error = 0.02), suggesting a small negative impact of more unskilled immigrants on native dropout employment. Fig. 3. Open in new tabDownload slide Relative Employment of High School Graduates and Dropouts vs. Fraction Low Education Dropouts Fig. 3. Open in new tabDownload slide Relative Employment of High School Graduates and Dropouts vs. Fraction Low Education Dropouts Fig. 2. Open in new tabDownload slide Relative Wage of High School Graduates and Dropouts vs. Fraction Low Education Immigrants Fig. 2. Open in new tabDownload slide Relative Wage of High School Graduates and Dropouts vs. Fraction Low Education Immigrants Formal estimation results for the local demand/supply system are presented in Table 3, which shows both ordinary least squares (OLS) estimates and instrumental variables (IV) estimates using the fraction of immigrant dropouts as an instrument for the relative supply of high school versus dropout labour. The IV results are quite similar to the OLS results, and nearly as precise, reflecting the strong first stage.17 The estimates suggest there is no relationship between the relative supply of high school dropouts and their relative wages, but point to a small negative impact of relative supply on relative employment. These findings are quite similar to the results in Card (2001) using data for 1990 and occupation‐based skill groupings.18 As in most of the previous work looking at local labour market impacts of immigration, there is a surprisingly weak relationship between immigration and less skilled native wages (Friedberg and Hunt, 1995; Borjas, 1994). This pattern has persisted despite steady inflows of relatively unskilled immigrants that have created ever greater differences across cities in the relative supply of dropouts over the past two decades. Table 3 Effects of Relative Supply on the Relative Wages and Employment of Native Male Dropouts . Relative outcomes of native male dropouts: . Fraction employed last year . Mean log hourly wage . OLS . IV . OLS . IV . Log relative supply of dropouts vs. high school graduates −0.013 −0.012 0.006 0.010 (0.003) (0.003) (0.009) (0.010) R‐squared 0.056 0.035 0.001 0.003 . Relative outcomes of native male dropouts: . Fraction employed last year . Mean log hourly wage . OLS . IV . OLS . IV . Log relative supply of dropouts vs. high school graduates −0.013 −0.012 0.006 0.010 (0.003) (0.003) (0.009) (0.010) R‐squared 0.056 0.035 0.001 0.003 Note. Standard errors in parentheses. All models fit to sample of 325 Metropolitan Statistical Areas using weighted least squares. City data are derived from the 2000 Census public use files and pertain to individuals aged 18–64. Outcomes are adjusted differences in employment‐population or mean log wages between high school dropouts and high school graduates – see text. Instrument is fraction of low education immigrants in city. Open in new tab Table 3 Effects of Relative Supply on the Relative Wages and Employment of Native Male Dropouts . Relative outcomes of native male dropouts: . Fraction employed last year . Mean log hourly wage . OLS . IV . OLS . IV . Log relative supply of dropouts vs. high school graduates −0.013 −0.012 0.006 0.010 (0.003) (0.003) (0.009) (0.010) R‐squared 0.056 0.035 0.001 0.003 . Relative outcomes of native male dropouts: . Fraction employed last year . Mean log hourly wage . OLS . IV . OLS . IV . Log relative supply of dropouts vs. high school graduates −0.013 −0.012 0.006 0.010 (0.003) (0.003) (0.009) (0.010) R‐squared 0.056 0.035 0.001 0.003 Note. Standard errors in parentheses. All models fit to sample of 325 Metropolitan Statistical Areas using weighted least squares. City data are derived from the 2000 Census public use files and pertain to individuals aged 18–64. Outcomes are adjusted differences in employment‐population or mean log wages between high school dropouts and high school graduates – see text. Instrument is fraction of low education immigrants in city. Open in new tab 2.4. Explaining the Absence of Local Labour Market Impacts A variety of explanations have been offered for the finding that wages of less skilled natives are insensitive to the relative supply pressure created by unskilled immigrants. The first is unobserved relative demand shocks, which enter the relative wage and employment equations and are potentially correlated with the relative share of low skilled workers.19 The leading solution to this problem is to instrument relative supply (or the relative number of low skilled immigrants) with information on historical immigration patterns. Immigrants from a given source country tend to go to the same places they went many years ago, and relative skill levels of the immigrants from a country are highly correlated over time, so instruments based on historical immigration patterns have reasonable predictive power. My reading of the evidence is that instrumenting sometimes moves the coefficients in the ‘right direction’, but does not change the conclusion that immigrant impacts are small; see for example, Lewis (2003) who analyses changes in relative wages in major cities between 1980 and 1990 using OLS and IV methods. A second explanation is that, as predicted by a Hecksher‐Ohlin (HO) model, variation in the relative supply of unskilled labour across local labour markets is absorbed by changing industry structure. As pointed out by Lewis (2003), the magnitude of any HO‐style adjustments can be calculated by looking at data on industry shares across cities. To illustrate this point, start with an identity that expresses the overall fraction of dropouts employed in a given city, sd(c), as a weighted sum of the industry shares in the city, times the dropout intensity in each industry: (2) where N(c) is total employment in city c, is the number of dropouts employed in industry i in city c, Ni(c) is total employment in industry i in city c, λi(c) ≡  Ni(c)/N(c) is the employment share of industry i in city c, and is the share of dropout workers in industry i in city c. It follows that the gap between sd(c) and the national average fraction of dropouts, sd, can be written as the sum of a ‘between industry component’B representing shifts in the relative fractions of different industries in the city, a ‘within industry component’W, representing shifts in the relative fraction of dropout workers in each industry, and an interaction component I: (3) where The HO theorem states that under certain conditions all of the variation in the share of dropout labour across cities can be absorbed by expansion or contraction of high‐dropout‐intensity industries (i.e., via the B(c) term), with no city‐level variation in relative wages or the dropout intensity of any particular industry.20 In Card and Lewis (2005), we used data on employment classified by 3 digit industry from the 2000 Census to compute the terms in (2) for each of 150 larger cities. We then performed a series of regressions: (4a) (4b) (4c) Since (3) is an identity, bw + bB + bI = 1. A strict version of HO implies bB = 1. Figure 4 plots the between‐industry component B(c) against the excess fraction of dropouts in each of the 150 larger MSAs. For reference, the Figure also shows a horizontal line: if changing industry structure accounted for none of the absorption of immigrants the points would lie along this line. Although the points suggest an upward‐sloping relationship, the slope is relatively modest, suggesting that changing industry structure accounts for only a small share of the absorption of dropouts. Indeed, the OLS estimate, reported in the first column of Table 4, is 0.22, and is significantly different from 1. By contrast, Figure 5 plots the within‐industry component W(C) against the excess fraction of dropouts in each city. This component is more highly correlated with the dropout share: as shown in column 2 of Table 4, the estimate of bw is 0.76. Though not shown in a Figure, the interaction terms are relatively small and essentially uncorrelated with differences across cities in the share of dropout workers. The estimate of bI in column 3 of Table 4 is 0.02 (with a very small R‐squared, 0.03). Fig. 5. Open in new tabDownload slide Contribution of Within‐Industry Component to Absorption of Dropouts Fig. 5. Open in new tabDownload slide Contribution of Within‐Industry Component to Absorption of Dropouts Fig. 4. Open in new tabDownload slide Contribution of Between‐Industry Component to Absorption of Dropouts Fig. 4. Open in new tabDownload slide Contribution of Between‐Industry Component to Absorption of Dropouts Table 4 Regression Models Measuring Cross‐City Absorption of Excess Dropout Workers . Effect across all industries: . Industry‐specific expansion: . Between industry . Within industry . Interaction . Agric. . Apparel & textiles . Low skill services . (1) . (2) . (3) . (4) . (5) . (6) . Excess fraction of dropout employment in city 0.22 0.76 0.02 0.09 0.05 0.03 (0.02) (0.03) (0.01) (0.02) (0.01) (0.01) R‐squared 0.37 0.84 0.03 0.17 0.24 0.33 . Effect across all industries: . Industry‐specific expansion: . Between industry . Within industry . Interaction . Agric. . Apparel & textiles . Low skill services . (1) . (2) . (3) . (4) . (5) . (6) . Excess fraction of dropout employment in city 0.22 0.76 0.02 0.09 0.05 0.03 (0.02) (0.03) (0.01) (0.02) (0.01) (0.01) R‐squared 0.37 0.84 0.03 0.17 0.24 0.33 Note. All models estimated on sample of 150 larger MSAs, using 264 industry cells per city in columns 1–3. Regressions are weighted by city size. See text for definitions of industries used in columns 4–6. Open in new tab Table 4 Regression Models Measuring Cross‐City Absorption of Excess Dropout Workers . Effect across all industries: . Industry‐specific expansion: . Between industry . Within industry . Interaction . Agric. . Apparel & textiles . Low skill services . (1) . (2) . (3) . (4) . (5) . (6) . Excess fraction of dropout employment in city 0.22 0.76 0.02 0.09 0.05 0.03 (0.02) (0.03) (0.01) (0.02) (0.01) (0.01) R‐squared 0.37 0.84 0.03 0.17 0.24 0.33 . Effect across all industries: . Industry‐specific expansion: . Between industry . Within industry . Interaction . Agric. . Apparel & textiles . Low skill services . (1) . (2) . (3) . (4) . (5) . (6) . Excess fraction of dropout employment in city 0.22 0.76 0.02 0.09 0.05 0.03 (0.02) (0.03) (0.01) (0.02) (0.01) (0.01) R‐squared 0.37 0.84 0.03 0.17 0.24 0.33 Note. All models estimated on sample of 150 larger MSAs, using 264 industry cells per city in columns 1–3. Regressions are weighted by city size. See text for definitions of industries used in columns 4–6. Open in new tab The MSAs with relatively high dropout shares are labelled in Figures 4 and 5. Interestingly, most of these MSAs are comprised of counties in California and Texas with substantial agricultural employment. Since agriculture relies on the availability of land resources, it is debatable whether the high employment shares of agriculture in these MSAs represents a reaction to abundant supplies of less‐educated labour. Rather, it seems more likely that the relative supplies of less‐educated labour in these MSAs are driven by the availability of farm jobs. The framework of (3) can be used to examine the contribution of the changing scale of specific industries to the absorption of local supplies of dropout labour. For example, the contribution of industry i is , which is excess employment share of the industry in city c relative to its national average share, multiplied by the average dropout intensity of the industry. Columns 4–6 of Table 4 show estimates of models similar to (4a), focusing on the absorption contributions of agriculture, textiles apparel and footwear industries, and a set of low‐skilled service industries.21 These 3 industry clusters together account for most of the total between industry effect: agriculture alone accounts for nearly one‐half. Overall, though there is some evidence that textiles and apparel manufacturing tends to cluster in cities with high dropout shares, these results suggest that most of the absorption of unskilled labour across cities occurs within very narrow industries. Apart from a few small sectors it is difficult to find much evidence of HO‐style industry adjustment across cities. Similar conclusions were reached by Lewis (2003), who examined changes in the relative absorption of 4 education groups over the 1980–90 period. Lewis used Census data to estimate first‐differenced versions of (4b) for each skill group.22 He also compared OLS estimates to IV estimates that used immigrant inflows based on historical immigration patterns as instruments for the changes in the relative shares of each skill group. As in the 2000 cross‐section, the industry composition effects over the 1980–90 period are only weakly related to local skill‐group‐specific population growth. Lewis’ estimates of bB for manufacturing industries (which are arguably best able to respond to local factor availability) are very close to 0, while his estimates for all industries range from 0 to 0.08. He also reports parallel specifications in which the dependent variable is the within‐industry relative employment term. These are much more strongly correlated with relative population growth, accounting for 90% of the adjustment to skill‐group specific relative supply shocks. The evidence suggests that HO‐style changes in industry structure play relatively little role in explaining how cities like Los Angeles were able to absorb massive inflows of relatively uneducated immigrant workers over the past two decades. Instead, most of the less‐educated labour was absorbed by city‐specific within‐industry increases in dropout intensity, which took place despite any corresponding changes in the relative wages of dropout workers. One possible explanation for this pattern is that local relative demand shocks for dropout workers are ‘caused’ by the presence of low skilled immigrants. For example, Acemoglu's (1998) model of endogenous technological change suggests that firms will innovate in a direction to take advantage of more readily available factors, even in the absence of relative wage changes. Beaudry and Green's (2003) model of technological adoption has a similar flavour. Lewis (2004) presents some of the first direct evidence for this mechanism, using data on the number of advanced technologies adopted by manufacturing plants in the late 1980s and early 1990s. He finds that controlling for very detailed (4 digit) industry effects, the adoption of advanced technologies by individual plants is significantly reduced by the presence of a greater relative supply of unskilled labour in the local labour market. These results are potentially consistent with the evidence on within‐industry absorption in Table 4. More work is clearly needed to understand better how firms choose which technologies to use and whether the choice is influenced by the relative availability of different skill groups. 2.5. Aggregate Evidence On Relative Wages of Dropouts My reading of the evidence is that the two main mechanisms that economists have proposed to explain the adjustment of local labour markets to immigration‐based supply shocks – selective mobility and HO‐style realignment of local industry structure – are relatively unimportant, and that the bulk of the absorption occurs within industries. In view of the weak correlation between local wages and local immigrant supplies, some researchers – notably Borjas et al. (1996, 1997) and Borjas (2003)– have argued that aggregate time series analyses are required to measure the full impacts of immigration on native wages. A complete analysis of aggregate trends is beyond the scope of this article. However, in light of the data in Table 1 showing the relative education distribution of immigrants, it is useful to briefly examine trends in the relative wages of high school dropouts. Figure 6 plots two measures of the wage gap between high school dropouts and high school graduates: the mean log wage differential between the groups, and the average return per year of schooling among those with 12 or fewer years of schooling, multiplied by 4. These wage gaps refer to the hourly earnings of men age 18–64 in the 1980–2002 March Current Populations Survey (CPS), and are estimated from models that include controls for a cubic in potential experience and dummies for black race and Hispanic ethnicity. For reference Figure 6 also plots the college–high school wage premium, estimated from samples of men with 12 or 16 years of schooling. Since 1979 the wage premium for high school graduates relative to dropouts has fluctuated in the range of 25% to 30%, with a modest rise in the early 1980s and more or less steady declines since then. The return per year of schooling for those with 0–12 years of school has fluctuated between 7% and 8%, and also increased slightly in the early 1980s. In contrast, the college–high school wage premium has varied a lot more, rising by about 12 log points in the early 1980s and nearly 22 log points over the past two decades. Fig. 6. Open in new tabDownload slide College/High School and High‐School/DropoutWage Gaps Fig. 6. Open in new tabDownload slide College/High School and High‐School/DropoutWage Gaps Although immigration presumably exerts downward pressure on the relative wages of dropouts, the wage gap between dropouts and high school graduates has been nearly constant since 1980, and has fallen by more than 50% relative to the gap between high school graduates and holders of bachelor's degrees.23 The absence of an aggregate trend in the relative wages of high school dropouts is consistent with the remarkable stability of the relative wage of dropouts across different local labour markets. Of course, even taking account of unskilled immigrant inflows the relative supply of dropouts has declined over the past two decades, so depending on what is assumed about the rate of growth of relative demand for dropouts versus high school graduates, one can argue that immigration lowered the wages of the least educated natives relative to the counterfactual trend.24 Without knowing the trend in relative demand for dropouts, however, the aggregate data are uninformative, so estimates of the effect on native wages amount to simply multiplying the relative share of dropouts attributable to immigration by some estimate of the elasticity of substitution (Johnson, 1980; Borjas, 2003). 3. Assimilation of Immigrants While immigrant men in the 1970 Census earned about as much as natives, a wage gap opened up over the 1970s and has persisted. Currently, immigrant men's hourly wages are about 20% lower than natives’, while immigrant women's wages are about 10% lower.25 Given the gap in education between immigrants and natives, and the importance of education in the US wage structure, this is not too surprising. Moreover, the quality of education in many of the major immigrant sending countries is arguably below the quality in the US (Bratsberg and Terrell, 2002), and many immigrants have limited English skills, implying that immigrant human capital is even lower than observed education would suggest. Following Chiswick (1978) there is an extensive literature on the question of whether the immigrant‐native earnings gap narrows with time in the US. Such ‘earnings assimilation’ could be due to formal or informal training, acquisition of language skills, or a variety of other processes.26Borjas (1985, 1995) noted that a synthetic cohort analysis like Chiswick's will overstate earnings growth if more recent immigrant arrival cohorts have lower unmeasured skill characteristics than earlier arrivals, as seems to have been true in 1980 and 1990. Moreover, many immigrants return to their home country within a few years, and others move back and forth, further complicating inferences from cross‐sectional data. Limited evidence from true longitudinal data (Lubotsky, 2000) suggests that immigrant earnings rise with time in the US, though the gains (about 10–15% in the first 20 years in the US) are not enough to offset the 35–40% immigrant–native earnings gap at arrival. Although the precise magnitude of immigrant earnings assimilation will probably be debated for many years (see Duleep and Regets (2002) for a recent analysis), few of the 40% of immigrants who arrive in the US as adults without a high school credential will ever earn as much as average natives. Likewise, the 22% of immigrants with a college degree or more will earn more than average natives.27 In my opinion, a more interesting question is how well the US‐born children of immigrants are doing. Focusing on the status of immigrants’ children is important for a number of reasons. Second generation immigrants are a growing fraction of the population, accounting for 10% of teenagers nationwide.28 Nearly all of them will spend their entire lives in the US, and will pay taxes and receive income support payments. Thus, the success of immigrant children is an important component of the long‐run costs and benefits of immigration. For these and other reasons the relative success of the second generation provides a key gauge of the extent to which their parents assimilated into the US. Table 5 presents some simple descriptive regression models showing the relative status of immigrants and second generation immigrants in the 1995–2002 CPS. (I define second generation immigrants as people born in the US with at least one foreign‐born parent.) The upper panel of the Table shows results for men, while the lower panel shows results for women. The first two columns shows models for years of schooling and the probability of working in the previous year, fitted over the entire population of 21–64 year olds, while columns 3–6 show models for log hourly wages, fitted to workers only. Table 5 Education and Earnings Gaps Between Immigrants, Second Generation and Others . Fitted to all individuals . Fitted to workers only . Years of education . % Working . Models for log hourly wage (coefficients × 100) . (1) . (2) . (3) . (4) . (5) . (6) . Estimates for men:  Immigrant −1.24 −0.6 −18.3 −23.4 −11.1 −8.0 (0.02) (0.2) (0.4) (0.4) (0.4) (0.4)  Second generation 0.45 0.8 8.0 3.6 1.5 2.3 (0.02) (0.2) (0.5) (0.5) (0.5) (0.5)  Controls for age Yes Yes Yes Yes Yes Yes  Controls for Region/urban No No No Yes Yes Yes  Control for education No No No No Yes Yes  Control for race/ethnicity No No No No No Yes Estimates for women:  Immigrant −1.37 −13.6 −11.8 −18.6 −7.1 −5.4 (0.01) (0.2) (0.4) (0.4) (0.4) (0.4)  Second generation 0.31 0.5 8.3 3.0 1.2 1.9 (0.02) (0.3) (0.5) (0.5) (0.5) (0.5)  Controls for age Yes Yes Yes Yes Yes Yes  Controls for region/urban No No No Yes Yes Yes  Control for education No No No No Yes Yes  Control for race/ethnicity No No No No No Yes . Fitted to all individuals . Fitted to workers only . Years of education . % Working . Models for log hourly wage (coefficients × 100) . (1) . (2) . (3) . (4) . (5) . (6) . Estimates for men:  Immigrant −1.24 −0.6 −18.3 −23.4 −11.1 −8.0 (0.02) (0.2) (0.4) (0.4) (0.4) (0.4)  Second generation 0.45 0.8 8.0 3.6 1.5 2.3 (0.02) (0.2) (0.5) (0.5) (0.5) (0.5)  Controls for age Yes Yes Yes Yes Yes Yes  Controls for Region/urban No No No Yes Yes Yes  Control for education No No No No Yes Yes  Control for race/ethnicity No No No No No Yes Estimates for women:  Immigrant −1.37 −13.6 −11.8 −18.6 −7.1 −5.4 (0.01) (0.2) (0.4) (0.4) (0.4) (0.4)  Second generation 0.31 0.5 8.3 3.0 1.2 1.9 (0.02) (0.3) (0.5) (0.5) (0.5) (0.5)  Controls for age Yes Yes Yes Yes Yes Yes  Controls for region/urban No No No Yes Yes Yes  Control for education No No No No Yes Yes  Control for race/ethnicity No No No No No Yes Notes. Standard errors in parentheses. Models estimated on pooled sample of 1995–2002 March Current Population Surveys. Samples include individuals age 21–64 only. Hourly wage is estimated from data on wage and salary earnings last year, weeks worked last year and usual hours per week last year. Wages are censored below at $2/hour (in 2002 dollars) are set to $2 and above at $200 per hour in 2002 dollars. Controls for Region/urban are 8 region dummies and dummy for living in Metropolitan Area. Control for education is linear term in years of education. Controls for race/ethnicity are dummies for black race and Hispanic ethnicity. Open in new tab Table 5 Education and Earnings Gaps Between Immigrants, Second Generation and Others . Fitted to all individuals . Fitted to workers only . Years of education . % Working . Models for log hourly wage (coefficients × 100) . (1) . (2) . (3) . (4) . (5) . (6) . Estimates for men:  Immigrant −1.24 −0.6 −18.3 −23.4 −11.1 −8.0 (0.02) (0.2) (0.4) (0.4) (0.4) (0.4)  Second generation 0.45 0.8 8.0 3.6 1.5 2.3 (0.02) (0.2) (0.5) (0.5) (0.5) (0.5)  Controls for age Yes Yes Yes Yes Yes Yes  Controls for Region/urban No No No Yes Yes Yes  Control for education No No No No Yes Yes  Control for race/ethnicity No No No No No Yes Estimates for women:  Immigrant −1.37 −13.6 −11.8 −18.6 −7.1 −5.4 (0.01) (0.2) (0.4) (0.4) (0.4) (0.4)  Second generation 0.31 0.5 8.3 3.0 1.2 1.9 (0.02) (0.3) (0.5) (0.5) (0.5) (0.5)  Controls for age Yes Yes Yes Yes Yes Yes  Controls for region/urban No No No Yes Yes Yes  Control for education No No No No Yes Yes  Control for race/ethnicity No No No No No Yes . Fitted to all individuals . Fitted to workers only . Years of education . % Working . Models for log hourly wage (coefficients × 100) . (1) . (2) . (3) . (4) . (5) . (6) . Estimates for men:  Immigrant −1.24 −0.6 −18.3 −23.4 −11.1 −8.0 (0.02) (0.2) (0.4) (0.4) (0.4) (0.4)  Second generation 0.45 0.8 8.0 3.6 1.5 2.3 (0.02) (0.2) (0.5) (0.5) (0.5) (0.5)  Controls for age Yes Yes Yes Yes Yes Yes  Controls for Region/urban No No No Yes Yes Yes  Control for education No No No No Yes Yes  Control for race/ethnicity No No No No No Yes Estimates for women:  Immigrant −1.37 −13.6 −11.8 −18.6 −7.1 −5.4 (0.01) (0.2) (0.4) (0.4) (0.4) (0.4)  Second generation 0.31 0.5 8.3 3.0 1.2 1.9 (0.02) (0.3) (0.5) (0.5) (0.5) (0.5)  Controls for age Yes Yes Yes Yes Yes Yes  Controls for region/urban No No No Yes Yes Yes  Control for education No No No No Yes Yes  Control for race/ethnicity No No No No No Yes Notes. Standard errors in parentheses. Models estimated on pooled sample of 1995–2002 March Current Population Surveys. Samples include individuals age 21–64 only. Hourly wage is estimated from data on wage and salary earnings last year, weeks worked last year and usual hours per week last year. Wages are censored below at $2/hour (in 2002 dollars) are set to $2 and above at $200 per hour in 2002 dollars. Controls for Region/urban are 8 region dummies and dummy for living in Metropolitan Area. Control for education is linear term in years of education. Controls for race/ethnicity are dummies for black race and Hispanic ethnicity. Open in new tab Looking first at the education models, immigrants have about 1.2–1.4 fewer years of education then natives, standardising for their age. (The raw gaps are a little smaller.) On the other hand, second generation immigrants have 0.3–0.4 years more education than people whose parents were born in the US (the ‘third and higher’ generation). Among men, immigrants are only slightly less likely to work than members of third and higher generation, while second generation men are a little more likely to work. Among women there is a larger immigrant gap in the probability of working (a 13.6% lower annual employment rate than natives) but again second generation women are a little more likely to work than third and higher generation natives. The models in columns 3 and 4 present wage models that control for age and geographic location but not for education. Among men, immigrants have 18–23% lower wages than third and higher generation natives, while second generation immigrants have 4–8% higher wages. The wage gaps for second generation women are about the same as for second generation men, but for immigrant women the wage gap is smaller than for immigrant men, perhaps reflecting the relative selectivity of labour force participation among immigrant women. When controls are added for education, the wage gap for immigrant men falls to about 11% while the gap for immigrant women falls to 7%. The wage gaps for second generation men and women also fall, to under 2% in each case. Thus, the higher wages of second generation immigrants are largely explained by their geographic location and their higher education. The models in the final column of Table 5 add two additional controls for black race and Hispanic ethnicity. Reflecting the fact that many immigrants are Hispanic, and that third and higher generation Hispanics earn a little less than non‐Hispanics, these added controls reduce the immigrant wage gaps slightly, to 8% for men and 5% for women, and slightly increase the wage advantage of second generation immigrants. The results in Table 5 suggest a couple of conclusions. First, immigrant workers in the US labour market today (over 90% of whom arrived after 1965) earn less than natives but the magnitude of the wage gaps are not enormous. After controlling for education, which explains about an 11% gap in immigrant earnings for both men and women, the gaps are under 10%– comparable to the wage gaps for blacks or native Hispanics. Second, the children of immigrants do well, on average, with most of their wage advantage relative to natives attributable to higher education. Despite the lower education of their parents, children born to immigrant parents seem to catch up and even surpass the levels of children born to US natives. An interesting perspective on this catch‐up phenomenon is provided by examining differences across parental source countries (Borjas, 1993; Card et al., 2000). Looking in the recent CPS data, one can identify second generation men and women whose parents were from different countries, and compare the earnings or educational attainment of each second generation group against the corresponding outcomes for their parents. This idea is illustrated in Figures 7a and 7b, which show mean education levels for second generation younger adults (age 21–40) in the 1995–2002 March CPS by country of origin of their father, plotted against mean levels of education for fathers of children age 0–15 from the same source country in the 1980 Census. For reference, I have also shown the point representing members of the third and higher generation, and the estimated regression line across the 39 country of origin groups shown in the Figures.29 Fig. 7. Open in new tabDownload slide (a) Father–Son Intergenerational Correlation in Education. (b) Father–Daughter Intergenerational Correlation in Education Fig. 7. Open in new tabDownload slide (a) Father–Son Intergenerational Correlation in Education. (b) Father–Daughter Intergenerational Correlation in Education The Figures suggest that there is a strong intergenerational correlation in education that is similar for sons and daughters. Indeed, the coefficient estimates and R‐squared statistics are nearly identical for sons and daughters (slope = 0.30 for men, standard error = 0.03, R‐squared = 0.77; slope = 0.29 for women, standard error = 0.03, R‐squared = 0.77). Interestingly, the coefficient of 0.3 for the effect of fathers education on either sons or daughters is almost identical to the estimates obtained in a micro level regression using samples of men and women from the General Social Survey.30 Thus, the intergenerational transmission of education is about the same for families of immigrants as for other families in the US. In particular, there is no evidence that second generation immigrants’ education outcomes regress toward the mean more slowly than other children. Even more interestingly, in both Figures 7a and 7b the fitted line for the second generation group over‐predicts the outcomes for natives: by 0.71 years for men and by 0.77 years for women. This means, for example, that second generation sons whose fathers had as little as 10.4 years of schooling (2.3 years below the average for native fathers) ended up ahead of their third generation peers. Even sons of Mexican immigrants, whose fathers had 5.5 years of schooling less than native‐born fathers in 1980 (7.3 years versus 12.8 years for native‐born fathers) ended up with 12.2 years of schooling, closing 80% of the education gap faced by their fathers.31 Finally, it is interesting to compare the results in Figures 7a and 7b with similar results from an earlier generation of immigrant children. Card et al. (2000) conduct a parallel analysis using 1970 Census data for second generation immigrants, and 1940 Census data for their parents. The estimated intergenerational coefficients in education are 0.41 (standard error 0.10) for men and 0.47 (standard error 0.08) for women. These point estimates are a little higher than the ones for more recent cohorts, though relatively imprecise. If anything, however, they suggest that the rate of assimilation (which is 1 minus the intergenerational correlation) is slightly faster for more recent cohorts than for older ones. These results paint a relatively optimistic picture of the success of post‐1965 immigrants. Conditional on their parents’ human capital, the US‐born children of these immigrants have done remarkably well. Indeed, of the 39 largest country‐of‐origin groups, sons from 33 groups and daughters from 32 groups have higher average educational attainment than the children of natives. 4. Conclusions Immigration is a major policy concern in many countries around the world. Two important questions that economic research can answer concern the impact of immigrants on the labour market opportunities of natives, and the relative success of immigrants in integrating into the domestic economy. Economists have struggled with both questions for the past couple of decades, with varying degrees of success, and the lessons from the US literature provide potentially valuable lessons to researchers in other contexts. On the question of immigrant competition the US has a structural advantage, since there are many large US cities, with widely varying levels of immigration, and samples from the Decenniel Censuses can be used to develop detailed models of local labour market outcomes. New evidence from the 2000 Census reconfirms the main lesson of earlier studies: although immigration has a strong effect on relative supplies of different skill groups, local labour market outcomes of low skilled natives are not much affected by these relative supply shocks. Recent evidence on the response of local industry structure to immigration‐induced supply shocks shows that the absorption of unskilled immigrants takes place within industries in high‐immigrant cities, rather than between industries, as implied by simple trade models. It remains a fascinating question how firms in a given industry can adapt their production technology so closely to local supplies of different types of labour without substantial changes in relative wages. As the evidence has accumulated over the past two decades that local labour market outcomes are only weakly correlated with immigrant densities, some analysts have argued that the cross‐city research design is inherently compromised by intercity mobility of people, goods, and services. Underlying this argument is the belief that labour market competition posed by immigration has to affect native opportunities, so if we do not find an impact, the research design must be flawed. The leading alternative to a local labour market approach is a time series analysis of aggregate relative wages. Surprisingly, such an analysis shows that the wages of native dropouts (people with less than a high school diploma) relative to native high school graduates have remained nearly constant since 1980, despite pressures from immigrant inflows that have increased the relative supply of dropout labour, and despite the rise in the wage gap between other education groups in the US economy. While the counterfactual is unknown, it is hard to argue that the aggregate time series evidence points to a negative impact of immigration unless one starts from that position a priori. On the question of immigrant assimilation, a major constraint in the US literature has been the absence of true longitudinal data. Nevertheless, I believe that a narrow focus on immigrant earnings is misplaced. Few of the 40% of immigrants who come to the US without completed high school education will ever catch up with the average earnings of natives. Most of their US‐born children, however, will catch up with the children of natives. Evidence on the intergenerational progress of immigrants’ children is now becoming available and points to above‐average levels of educational attainment, even for children whose fathers had much lower schooling than native‐born fathers. The relatively strong educational progress of second generation immigrants, together with the limited evidence of adverse effects on less skilled natives, suggest that the new immigration may not be so bad after all. Footnotes 1 " The law was influenced by research of Brigham (1923), who classified immigrants into four racial categories: ‘Nordic’, ‘Alpine’, ‘Mediterranean’ and ‘Asian’, and argued that members of the Alpine and Mediterranean races had lower intelligence than Nordics. The influence of Brigham's work is illustrated by a headline announcing the new law in the Los Angeles Times (13 April 1924): ‘Nordic Victory is seen in Drastic Reduction’. 2 " The 2000 US Census has information on exact arrival year, and I used this information to examine changes in the fraction of immigrants from different countries before and after 1965. The fraction of Mexican immigrants, for example, is 20.1% for 1963–5 arrivals and 20.1% for 1966–8 arrivals. A sharp impact of the law is not discernable in these data, though there is a trend between 1950 and 1975. 3 " Education levels of immigrants are correlated with education levels in the home country but there are many interesting exceptions. For example, immigrants from India have the highest average education (average of 15.6 years of completed schooling). Immigrants from Russia are a very close second. 4 " If immigrants had the same education and other skill characteristics as natives, and if capital is elastically supplied to the relevant labour market, then standard economic models would predict no impact on native wages – see Altonji and Card (1991). 5 " This approach is closely related to work on internal migration and local wage structures, including Sjaastad (1962), Topel (1986) and Dahl (2002). 6 " Of course the same arguments about intercity trade and factor mobility also apply across countries. Models of international trade often imply that relative wages in a country are independent of the relative supplies of different skill groups, at least in some range. See Kuhn and Wooten (1991). 7 " Card (2001, p. 23) notes that immigrant men earned more than native men in one third of the largest US cities in 1990. 8 " An alternative approach is to assume that workers with different characteristics sell ‘bundles’ of skills, where the number of latent skills is small. This approach has been suggested to study the structure of wages by age and education (Welch, 1969) but becomes complex once allowance is made for non‐linear pricing of the bundles (Heckman and Scheinkman, 1987). 9 " Under the perfect substitutes assumption, for example, the wage gap between immigrants with less than 12 years of schooling and native high school dropouts should be constant (controlling for age, time in the US etc.). Even if the perfect substitutes assumption is true, the wage gap could vary across cities if immigrants in different cities possess different unobserved skills. 10 " The set of all cities includes 272 Standard Metropolitican Statistical Areas in 1980 and 325 Metropolitan Statistical Areas in 2000. 11 " This is largely a cohort effect, reflecting the steady rise across cohorts in the fraction of high school graduates until cohorts born in the 1950s. See Card and Lemieux (2000, 2001). 12 " The size controls are the log of the adult population and its square. The regressions are estimated by weighted least squares using the size of the population as weights. 13 " Likewise Frey (1995, 1996) reports a strong correlation between immigrant inflows and native outflows. Wright et al. (1997) re‐examine Frey's specifications and show that his results disappear once controls for city size are added to the model. 14 " See e.g., Card (2001). Specifically, consider a production function for local output y = [Σj(ejNj)(σ − 1)/σ]σ/(σ − 1) and supply functions  log (Nj/Pj) = ɛ log wj + ϕj, where Nj is the number of people employed in skill group j, wj is the wage of group j, ej is a relative demand shock, Pj is the population of skill group j, and ϕj is a local supply shock. These equations imply a relative labour demand curve and a relative labour supply curve 15 " The wage gaps were estimated as follows. First, separate models were estimated for mean log wages of native male dropouts and high school graduates, including unrestricted city dummies (for 325 MSAs) a quartic function of age, dummies for black race and Hispanic ethnicity, and interactions of the black and Hispanic dummies with age and ages squared. The wage gaps are estimated as the differences in the city dummies from these two models (re‐normalised to have the same mean as the raw data). 16 " The local employment population rates of the two groups were estimated as the city dummies in separate linear probability models for the event of working in the previous year, in models with the same control variables as the first stage wage models. 17 " The coefficient of the fraction of immigrant dropouts in a model for the log relative supply of high school versus dropout labour is −6.10, with a standard error of 0.20 (F‐statistic = 902). The fraction of immigrant dropouts explains 74% of the variation in the relative supply variable across the 325 MSAs in the 2000 Census. 18 " In my 2001 study I presented estimates for 6 occupation groups in 175 cities. Interestingly, the estimated relative supply effects were typically smaller when the sample was restricted to low skilled occupations. 19 " For example, using the model outlined in footnote 14, the residual in the relative wage equation is 20 " These conditions would include infinitely elastic supplies of capital, perfectly integrated product markets, and the existence of at least one industry that produces a tradeable good or service that has a dropout intensity that exceeds the maximum dropout share in any city. 21 " We include textiles, apparel, knitting mills, footwear, and leather industries as apparel, and the following as ‘low skilled services’: building services, landscaping services, carwashes, landscaping, dry cleaning and laundry services, private household services, and other personal services. 22 " One difference is that Lewis regresses the between‐industry effects on the population share of the skill group in the local labour market, rather than the employment share. An advantage of a first differenced approach is that it eliminates the confounding caused by permanent factors like differences in the amount of agricultural land in an MSA. 23 " According to the data in Table 1, the presence of immigrants increased the relative supply of dropouts in 2000 by about 21%, reduced the relative supply of high school graduates by about 5%, and had no net effect on the relative supply of people with a college degree or more. Assuming that the elasticity of substitution between education groups is −1.4 (Borjas, 2003; Katz and Murphy, 1992) and ignoring labour supply effects, the presence of immigrants in the US labour market should have raised the wage premium for high school graduates relative to dropouts by about 26/1.4 = 18 log points, in the absence of other factors. 24 " In the 1980 Census, 26.3% of the population age 18–64 were dropouts, 39.2% had exactly 12 years of schooling, 19% had some college and 15.6% had a college degree or more. Comparing these numbers to those in Table 1 there was a 35% decline in the log relative supply of dropout versus high school labour between 1980 and 2000. 25 " These numbers come from an analysis of March CPS data from 1995 to 2002. 26 " Cortes (2004) shows that recent immigrant arrivals have relatively high rates of participation in schooling. She finds that 1975–80 immigrant arrivals show a gain in English proficiency between 1980 and 1990. Manning (2003, ch. 6) notes that some fraction of life cycle earnings growth is due to accumulated ‘search capital’. Immigrants may start off with less efficient search and gradually catch up to natives. 27 " Using March CPS data for 1995–2002, I estimate that immigrants with at least a college degree earn about 30% more than average natives. 28 " Since 1994, the CPS has asked individuals where their parents were born. Using March 1995–2002 CPS files, I estimate that about 11% of people age 16–19 were born in the US with at least one immigrant parent. 29 " I selected countries of origin with at least 50 observations for second generation sons and daughters. The largest group is Mexico (4,998 second generation children). Italy, Canada, Cuba, Germany and the Philippines also have at least 500 second generation children. The smallest origin groups are Panama (54 observations), Austria (53 observations) and Israel (51 observations). 30 " I used the 1972–96 GSS. The sample has 6,667 men and 7,745 women between the ages of 21 and 45 with observed father's education. In a regression controlling for age and age‐squared, the effect of father's education is 0.32 for men (standard error 0.01) and 0.30 for women (standard error 0.01). 31 " The mean level of education of third and higher generation sons is 13.29 (14.4 for daughters), while the mean level of education of second generation Mexican sons is 12.19 (12.41 for daughters). References Acemoglu , Daron ( 1998 ). ‘Why do new technologies complement skills? Directed technical change and wage inequality’ , Quarterly Journal of Economics , vol. 114 (November), pp. 1055 – 89 . Google Scholar Crossref Search ADS WorldCat Altonji , Joseph G., and Card , David ( 1991 ). ‘The effects of immigration on the labor market outcomes of less skilled natives’, in ( John M. Abowd and Richard B. Freeman, eds.), Immigration, Trade, and the Labor Market , pp. 201 – 31 , Chicago: The University of Chicago Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Beaudry , Paul and Green , David ( 2003 ). ‘Wages and employment in the United States and Germany: what explains the differences?’ , American Economic Review , vol. 93 (June): pp. 573 – 603 . Google Scholar Crossref Search ADS WorldCat Borjas , George ( 1985 ). ‘Assimilation, changes in cohort quality, and the earnings of immigrants’ , Journal of Labor Economics , vol. 3 , pp. 463 – 89 . Google Scholar Crossref Search ADS WorldCat Borjas , George ( 1987 ). ‘Immigrants, minorities, and labor market competition’ , Industrial and Labor Relations Review , vol. 40 , pp. 382 – 92 . Google Scholar Crossref Search ADS WorldCat Borjas , George ( 1993 ). ‘The intergenerational mobility of immigrants’ , Journal of Labor Economics , vol. 11 , pp. 113 – 35 . Google Scholar Crossref Search ADS WorldCat Borjas , George ( 1994 ). ‘The economics of immigration’ , Journal of Economic Literature , vol. 32 , pp. 1667 – 717 . OpenURL Placeholder Text WorldCat Borjas , George ( 1995 ). ‘Assimilation and changes in cohort quality revisited: what happened to immigrant earnings in the 1980s?’ , Journal of Labor Economics , vol. 13 , pp. 201 – 45 . Google Scholar Crossref Search ADS PubMed WorldCat Borjas , George ( 1999 ). Heaven's Door , Princeton, NJ: Princeton University Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Borjas , George ( 2003 ). ‘The labor demand curve is downward sloping: reexamining the impacts of immigration on the labor market’ , Quarterly Journal of Economics , vol. 118 (November), pp. 1335 – 74 . Google Scholar Crossref Search ADS WorldCat Borjas , George , Freeman , Richard B and Katz , Lawrence ( 1996 ). ‘Searching for the effect of immigration on the labor market’ , American Economic Review , vol. 86 (May), pp. 246 – 51 . OpenURL Placeholder Text WorldCat Borjas , George , Freeman , Richard B and Katz , Lawrence ( 1997 ). ‘How much do immigration and trade affect labor market outcomes?’ , Brookings Papers on Economic Activity , vol. 1 , pp. 1 – 67 . Google Scholar Crossref Search ADS WorldCat Bratsberg , Bernt. , and Terrell , Dek ( 2002 ). ‘School quality and returns to education of US immigrants’ , Economic Inquiry , vol. 40 (April), pp. 177 – 98 . Google Scholar Crossref Search ADS WorldCat Brigham , Carl C . ( 1923 ). A Study of American Intelligence , Princeton NJ: Princeton University Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Card , David ( 2001 ). ‘Immigrant inflows, native outflows and the local labor market impacts of higher immigration’ , Journal of Labor Economics , vol. 19 (January), pp. 22 – 64 . Google Scholar Crossref Search ADS WorldCat Card , David and DiNardo , John E. ( 2000 ). ‘Do immigrant inflows lead to native outflows?’ , American Economic Review , vol. 90 (May), pp. 360 – 67 . Google Scholar Crossref Search ADS WorldCat Card , David and DiNardo , John E ( 2002 ). ‘Skill biased technical change and rising wage inequality: some problems and puzzles’ , Journal of Labor Economics , vol. 20 (October), pp. 733 – 83 . Google Scholar Crossref Search ADS WorldCat Card , David , DiNardo, John E., and Estes , Eugena ( 2000 ). ‘The more things change: immigrants and the children of immigrants in the 1940s, the 1970s, and the 1990s’, in ( George Borjas, ed.), Issues in the Economics of Immigration , Chicago: University of Chicago Press for NBER . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Card , David and Lemieux , Thomas ( 2000 ). ‘Dropout and enrollment trends in the post‐war period: what went wrong in the 1970s?’, in ( Jonathan Gruber, ed.), Risky Behavior Among Youth: An Economic Analysis , Chicago: University of Chicago Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Card , David and Lemieux , Thomas ( 2001 ). ‘Can falling supply explain the rising return to college for younger men? A cohort‐based analysis’ , Quarterly Journal of Economics , vol. 116 (May), pp. 705 – 46 . Google Scholar Crossref Search ADS WorldCat Card , David and Lewis Ethan, G. ( 2005 ). ‘The diffusion of Mexican immigrants in the 1990s: patterns and impacts’ , UC Berkeley Center for Labor Economics, unpublished working paper, February. Chiswick , Barry R . ( 1978 ). ‘The effects of Americanization on the earnings of foreign‐born men’ , Journal of Political Economy , vol. 86 , pp. 897 – 921 . Google Scholar Crossref Search ADS WorldCat Cortes , Kalena E ( 2004 ). ‘Are refugees different from economic immigrants? Some empirical evidence on the heterogeneity of immigrant groups in the United States’ , Review of Economics and Statistics , forthcoming. OpenURL Placeholder Text WorldCat Dahl , Gordon B . ( 2002 ). ‘Mobility and the return to education: testing a Roy model with multiple markets’ , Econometrica , vol. 70 (November), pp. 2367 – 420 . Google Scholar Crossref Search ADS WorldCat Duleep , Harriet and Regets , Mark C ( 2002 ). ‘The elusive concept of immigrant ‘‘quality’’: evidence from 1970–1990’ , IZA Working Paper No. 631, November. Friedberg , Rachel M and Hunt , Jennifer ( 1995 ). ‘The impact of immigration on host country wages, employment, and growth’ , Journal of Economic Perspectives , vol. 9 (Spring), pp. 23 – 44 . Google Scholar Crossref Search ADS WorldCat Frey , William H . ( 1995 ). ‘Immigration and internal migration ‘‘flight’’ from US metropolitan areas: toward a new demographic balkanisation’ , Urban Studies , vol. 32 , pp. 733 – 57 . Google Scholar Crossref Search ADS WorldCat Frey , William H . ( 1996 ). ‘Immigration, domestic migration, and demographic balkanization in America: new evidence for the 1990s’ , Population and Development Review , vol. 22 (December), pp. 741 – 63 . Google Scholar Crossref Search ADS WorldCat Grossman , Jean, B . ( 1982 ). ‘The substitutability of natives and immigrants in production’ , Review of Economic and Statistics , vol. 64 , pp. 596 – 603 . Google Scholar Crossref Search ADS WorldCat Heckman , James J . and Scheinkman , Jose ( 1987 ). ‘The importance of bundling in a Gorman‐Lancaster model of earnings’ , Review of Economic Studies , vol. 54 , pp. 243 – 55 . Google Scholar Crossref Search ADS WorldCat Jaeger , David ( 1996 ). ‘Skill differences and the effect of immigrants on the wages of natives’ , US Bureau of Labor Statistics Working Paper No. 273, March. Johnson , George E . ( 1980 ). ‘The labor market effects of immigration’ , Industrial and Labor Relations Review , vol. 33 (April), pp. 331 – 41 . Google Scholar Crossref Search ADS WorldCat Katz , Lawrence and Murphy , Kevin M ( 1992 ). ‘Changes in relative wages, 1963–1987: supply and demand factors’ , Quarterly Journal of Economics , vol. 107 (February), pp. 35 – 78 . Google Scholar Crossref Search ADS WorldCat Kuhn , Peter and Wooten , Ian ( 1991 ). ‘Immigration, trade, and the wages of natives’, in ( John M. Abowd and Richard B. Freeman, eds.), Immigration, Trade, and the Labor Market , pp. 285 – 304 , Chicago: The University of Chicago Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC LaLonde Robert, J. and Topel Robert, H. ( 1991 ). ‘Labor market adjustments to increased immigration’, in ( John M. Abowd and Richard B. Freeman, eds.), Immigration, Trade, and The Labor Market , pp. 167 – 200 , Chicago: University of Chicago Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Lazear , Edward P . ( 1999 ). ‘Culture and language’ , Journal of Political Economy , vol. 107 ( 6 ), pp. S95 – 126 . Google Scholar Crossref Search ADS WorldCat Lewis , Ethan G . ( 2003 ). ‘Local open economies within the US: how do industries respond to immigration?’ , Federal Reserve Bank of Philadelphia Working Paper No. 04‐1, (December). Lewis , Ethan G . ( 2004 ). ‘The effect of local skill mix on new technology adoption in US manufacturing’ , mimeo, Federal Reserve Bank of Philadelphia (December). Lubotsky , Darren ( 2000 ). ‘Chutes or ladders? A longitudinal analysis of immigrant earnings’ , Princeton University Industrial Relations Section Working Paper No. 445, August. Manning , Alan ( 2003 ). Monopsony in Motion , Princeton NJ: Princeton University Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Sjaastad , Larry A . ( 1962 ). ‘The costs and returns of human migration’ , Journal of Political Economy , vol. 70 ( 2 ) (October), pp. S80 – 93 . Google Scholar Crossref Search ADS WorldCat Topel , Robert H . ( 1986 ). ‘Local labor markets’ , Journal of Political Economy , vol. 94 ( 2 ) (June), pp. S111 – 43 . Google Scholar Crossref Search ADS WorldCat Welch , Finis ( 1969 ). ‘Linear synthesis of skill distribution’ , Journal of Human Resources , vol. 4 (Summer), pp. 311 – 27 . Google Scholar Crossref Search ADS WorldCat Wright , Richard , Ellis , Mark and Reibel , M ( 1997 ). ‘The linkage between immigration and internal migration in large metropolitan areas in the United States’ , Economic Geography , vol. 73 , pp. 234 – 53 . Google Scholar Crossref Search ADS PubMed WorldCat Author notes " I am grateful to Christian Dustmann, Thomas Lemieux, Ethan Lewis, Stephen Machin and two anonymous referees for helpful suggestions, and to Florence Neymotin for outstanding research assistance. Partial funding for this work was provided by the NICHD. © Royal Economic Society 2005.

Misina,, Miroslav

doi: 10.1111/j.1468-0297.2005.01043_4.xpmid: N/A

The motivation behind the present work can be summarised by quoting the author: …while to some extent irrelevant, economics is not ‘mostly harmless’: (p. 311) The source of potential harmfulness of economics, according to the author, lies in the fact that policy recommendations of mainstream economists on questions such as privatisation, monopoly regulation, labour unions etc., are based on flawed models. The author's objective is to demonstrate that models typically used by mainstream (neo‐classical) economists are built on unsound foundations. Concepts as basic as supply and demand suffer from a number of internal inconsistencies and logical contradictions, which casts doubts on the professed scientific foundation of the recommendations and raises the possibility that the advice offered is ‘as likely to do general harm as it is to lead to the general good’ (p. 4). The author, however, does not put all the blame on the economists – the guilt lies to a large extent in the educational system by means of which the mainstream economists are ‘educated into ignorance’. They are not exposed to different views and not encouraged to critically examine the theories presented to them. One purpose of this work is to bring together various criticisms of the mainstream economics that the author thinks should form a part of advanced education in economics. The presentation is divided into 3 parts. In Part 1 the author deals with basic concepts of supply and demand and wishes to demonstrate that these basic concepts suffer from a number of inconsistencies, which makes them useless. In Part 2 the author deals with issues such as the problem of aggregation of capital (Chapter 6), problems with static analysis and treatment of time in economics (Chapter 8), and critique of Keynesian economics, as summarised by the IS/LM model. In Part 3 alternatives to mainstream economics are discussed. Chapter 12 offers a summary of the argument, especially as it pertains to fundamental problems discussed in Part 1. The claim that mainstream economics suffers from fundamental inconsistencies is a serious one. Successful demonstration of these claims would imply that the theory is flawed on logical grounds and that it cannot be used as a basis for coherent analysis of the policy issues mentioned above. In light of the importance of the issues, in this review we focus on chapters in which the inconsistency arguments are presented. The key question is, then, to what extent does the author succeed in identifying the alleged contradictions and inconsistencies? Unfortunately, to paraphrase the author, ‘the mathematical economists did their job properly’ (p. 48). What they showed is that most of the propositions criticised in the book do not hold generally but only under certain conditions. Arguing that these conditions are unrealistic or absurd (p. 27), does not amount to demonstrating that the theory is internally inconsistent. Indeed, behind the superficially objective analysis of logical problems in mainstream economics lies a strong commitment to methodological views that differ from the stated methodology of mainstream economics. If one summarises methodological commitments of mainstream economists as methodological individualism plus instrumentalism, the author's preferred methodology seems to be some version of holism plus scientific realism. Most of the arguments presented in the book are best understood as criticisms of some of the basic tenets of the neo‐classical research programme.1 The arguments presented can be broadly divided in two categories:2 • in the first category are the arguments in which the author correctly identifies some of the underlying assumptions necessary for a proposition in question to hold and then argues that the theory is invalid or should be abandoned since the assumptions are not realistic. in the second category are the arguments in which the author fails to identify some of the key conditions necessary for a proposition to hold and then rediscovers these. In these cases, contradictions are established by claiming that the rediscovered conditions establish that the theory does not hold. We will now provide illustrations of these arguments.3 Inconsistencies and Contradictions? Market Demand Curve(Chapter 2) The objective of this chapter is to invalidate methodological individualism by appealing to some of the issues that arise in aggregating individual demands into market demand. It is well known that only under certain conditions the market demand will possess the properties of individual demand. The main condition is that the utility functions have Gorman form, which author correctly identifies.4 Interpreted in behavioural terms, this functional form implies that agents are allowed to differ only to a limited extent. The author believes that the necessary assumptions are unrealistic (p. 27), and that this demonstrates that methodological individualism is flawed, since society is something more than a sum of its parts. Supply Curve (Chapter 3) The author identifies diminishing marginal product as the key condition needed in order to derive an upward‐sloping supply curve, and then proceeds to argue that constant returns are the norm, and that ‘since theory must be realistic, production model based on the unrealistic assumption of diminishing marginal product should not be used, (p. 65) Monopoly Vs. Competition (Chapter 4) While in the discussion of market demand the author appeals to more advanced material in order to establish aggregation conditions, there is no such appeal in this chapter. The author fails to mention that the competitive (price‐taking) argument is a limiting argument5 Rather, he sets up an example of non‐atomistic economy, and then discovers that changes in quantity by any one firm will, if not compensated by corresponding opposite changes in quantity by other firms, result in changes in equilibrium price. Since in competitive (price‐taking) markets it is assumed that one firm's output decision does not influence output decision of other firms, the compensating change in quantity will not occur, and thus the equilibrium price will be affected. The author could have argued, in line with his approach in Chapter 2, that one of the conditions necessary to reach competitive outcome is that the economy be atomistic, and that since this is clearly an unrealistic assumption the theory should be abandoned. On the contrary, the author decided that it is necessary at this point to provide a math primer for economists –‘infinitesimals ain't zero’ (p. 97). Critique of Walras’ Law (Chapter. 9) In the spirit of discussion of competition, the author fails to identify the conditions under which Walras Law is assumed to hold, the essential condition here being non‐satiation.6 Rather than stating that Walras Law is a property that is assumed to hold under certain conditions the author starts the discussion by claiming that economists regard Walras’ Law irrefutable (p. 192) and then proceeds to show that under some circumstances Walras’ Law does not hold. The circumstances discovered here are precisely those where non‐satiation property does not hold. Realism Vs. Instrumentalism We have seen above that in most of the cases where the conditions under which particular propositions or properties are supposed to hold are correctly identified, the main objection is the lack of realism. The author is aware that this criticism is easily dismissed by appeal to some version of Friedman's methodological views. Consequently, in Ch. 7 he proposes to dispose of this methodology, and to show Friedman's methodological views are simply ‘bad philosophy’ (p. 149). The author proposes to deal with the issue in a straightforward manner. Based on the classification of assumptions into negligibility, domain, and heuristic assumptions, the author argues that the types of assumptions usually made in economics are the domain assumptions specifying the conditions under which a particular theory will apply. If these conditions (for example, diminishing marginal product) are not present, the theory is inapplicable.7 This would not convince proponents of instrumentalism, since this is reduced to the argument that the realism of assumptions (in some form) does matter, whereas they argue that it is the predictive power that matters. And we are back where we started. Interestingly, the author claims that ‘economists genuinely believe that their theories describe reality’ (p. 149). One way to interpret the whole argument then would be to say that economists are, in spite of their professed adherence to instrumentalism, in fact proponents of scientific realism, who simply fail to see how unrealistic their assumptions are. Conclusion Methodological debates in economics are certainly not new, and we believe that any contributions in this area would be welcome. What we find problematic with the present work is that a methodological debate is presented as ‘right versus wrong’ (p. 14) argument, which claims to deal with the logic of economic theory. This presentation in effect obscures the fundamental issues and makes the book largely unsuitable for at least one segment of its intended audience, viz. current students of economics. Much better understanding of issues in this debate can be acquired by consulting the literature dealing explicitly with methodological issues; Blaug (1980) is a good place to start. The situation with other segments of the intended audience is, we are afraid, not much better. Readers with knowledge of advanced economics but without exposure to methodological issues raised in this book will most likely dismiss it after having discovered problems with the argument. The rest will probably use it to support their preconceptions about the state of economics and/or its critics. This would be unfortunate since the author's concerns seem genuine. Footnotes 1 " For a discussion of scientific research programmes in economics, see Latsis (1976). Sawyer (1989) provides a discussion of macroeconomics in terms of scientific research programmes. 2 " One could possibly include another category, ‘Miscellaneous’. In this category we would include aruments such as ‘profits and time’ in Ch. 3, which we invite the reader to examine. 3 " In our illustrations we will use MasColell et al. (1995) as the reference point for advanced topics in mainstream economics. This book will be referred to as MWG. 4 " This turns out to be the key condition only if we limit ourselves to the consideration of aggregate demand as a function of the mean level of wealth. Weaker conditions can be imposed if aggregate demand is allowed to depend on some other aggregates. None of these issues are mentioned by the author. See MWG, Proposition 4.B.1 and the discussion that follows it. Incidentally, the author mislabels the aggregation conditions as Sonnenschein‐Mantel‐Debreu conditions. This is rather curious, given that the author provides Shafer and Sonnenschein (1982) as a reference. For a discussion of Sonnenschein‐Mantel‐Debreu theorem, see MWG, Sec. 17.E. 5 " This is easily seen in a typical Cournot oligopoly game‐in the limit the outcome converges to the ‘competitive’ case. See MWG, Example 12.C.I. 6 " See MWG, Proposition 17.B. 1. 7 " The author fails to mention that there are other possible classifications of assumptions. For a good discussion of these issues, see Blaug (1980) References Blaug , Mark ( 1980 ). The Methodology of Economics, or How Economists Explain , Cambridge: Cambridge University Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Latsis , Spiro (ed.) ( 1976 ). Method and Appraisal in Economics , Cambridge: Cambridge University Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC MasColell , Andrew , Winston , Michael, and Green Jerry ( 1995 ). Microeconomic Theory , New York: Oxford University Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Sawyer , John ( 1989 ). Macroeconomics Theory: Keynesian and Neo‐Walrasian Models , London: Harvester Wheatsheaf . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Shafer , W. , and Hugo Sonnenschein ( 1982 ). ‘Market demand and excess demand functions’, Chapter 14 in ( Kenneth Arrow and Michael. D. Intrigillator, eds.) Handbook of Mathematical Economics, Vol. 2 , Amsterdam: North‐Holland . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC © Royal Economic Society 2005.

Frijters,, Paul;Shields, Michael, A.;Price, Stephen, Wheatley

doi: 10.1111/j.1468-0297.2005.01040.xpmid: N/A

Abstract A major issue in the immigration debate concerns whether immigrants take jobs away from natives. In this article, we present new evidence on the job search methods used by immigrants, and their relative success in gaining employment, in the UK. We find that: (1) immigrant job search is less successful than that of natives; (2) immigrants are as likely to gain employment through informal methods as via verifiable routes; (3) the probability of success increases with years since migration. The finding that immigrants do not effectively compete for jobs may help explain why immigration has little impact on native employment. Many countries, such as Australia, Canada, Germany, the UK and the US, have continually struggled with the same question: should more immigration be allowed? Central to this debate is the public perception that immigration has a negative impact on the labour market opportunities of natives (Borjas, 2003). However, the principle finding in the US‐dominated literature is that there is little evidence that immigrants actually harm the job opportunities of natives; see Card (2005) and Friedberg and Hunt (1995), for a review of the literature. Although it is not universal; see, for example, Borjas (2003). This issue clearly remains controversial, with the existing European evidence also being a long way from a consensus; see, for example, De New and Zimmermann (1994); Pischke and Velling (1997) and Zorlu and Hartog (2005). So far very little research on this issue has been undertaken in the UK context. In a recent contribution, Dustmann et al. (2005) found that native wages and unemployment rates are largely unaffected by the presence of immigrants in the UK. One potential explanation for the lack of a significant impact is that natives may adjust to changed employment opportunities resulting from immigration by moving away from the immigrant‐receiving areas (Borjas, 2003). If this is the case the labour market effect of immigration would be diffused across all regions of the country. Hatton and Tani (2005) explore this hypothesis and find evidence that it may hold for some Southern regions of the UK. A further avenue to explore in the UK context is whether or not immigrants actually pose a competitive labour market threat to natives. Such a threat would be evident if immigrants use job search methods that lead them to have job finding rates equal (or even higher) than natives. Conversely, if immigrants are less successful in gaining employment, then the effect of immigration on the employment opportunities of natives may be considerably dampened, and this might help explain the findings of previous studies.1 In this article we provide new evidence of the relative success of natives and immigrants in the job search process using a large panel of ILO unemployed men drawn from the UK's Quarterly Labour Force Surveys between 1997 and 2001.2 In particular, we explore separately for the main immigrant groups, whether or not they employ different job search methods from natives, and importantly whether this choice then impacts on their duration of unemployment. An important innovation in this article is that we are able to test the relative effectiveness of verifiable (e.g. using a job centre) or non‐verifiable methods (e.g. through friends and family) of job search for the natives and immigrants in our sample directly and also establish whether immigrant job search is more successful as the number of years since immigration increases (Chiswick, 1982).3 In terms of the implications for immigration policy, if we find that immigrant groups are equally (or even more) successful in their job search, and therefore effectively obtaining jobs that would otherwise be filled by UK born workers, this might justify the tightening of immigration controls. However, there is a clear policy trade‐off – if it is the immigrant groups who are at a significant disadvantage in accessing employment, government policy might wish to address whether their job search could be made more effective. One clear motivation would be to reduce the cost to the public budget from a high rate of immigrant unemployment. A limitation of our analysis is that the link between differential job search success and job competition only exists under certain views of the labour market. One such view is where natives and immigrants apply to randomly created jobs where employers rank job applicants (Blanchard and Diamond, 1994). In the situation that employers, for whatever reason, prefer natives, immigrants will only get a job if no natives happened to have applied for that job. In this case the job finding probability would be strictly lower for an immigrant than for a native. Hence, in a random job‐creation story with ranking of applicants, the relative job finding success of different groups is informative about whether one group competes for a pool of jobs with the other groups, or whether one groups fills the jobs not applied for by another group. In other views of the labour market such a link between job search success and job competition cannot be so easily established. Our focus in this article is to describe and examine the differences in job search methods and their success between immigrant groups and natives. Therefore, it is important to recognise that drawing specific conclusions about job competition between the different groups implicitly depends on assumptions about the nature of the labour market. The remainder of the article is organised as follows. In Section 1 we introduce our data source and define our key variables. The descriptive analysis is presented and discussed in Section 2. In Section 3 we outline and discuss our decomposition analysis of the determinants of job search success across the different immigrant groups. Section 4 draws some policy conclusions. 1. Data and Key Definitions Our sample is derived from the Quarterly Labour Force Survey (QLFS) of the UK. The Labour Force Survey has been undertaken since 1973. Its primary purpose is to collect internationally comparable employment and unemployment data at a regional and national level for the UK. At the beginning of 1992 a quarterly element was introduced. The total number of households successfully questioned each quarter is approximately 64,000, amounting to some 167,000 persons. Each household is questioned for five successive surveys, so that if the household is first surveyed in the Spring (wave 1 – interviews conducted between March and May) of one year interviews will be attempted with that household for each successive quarter (waves 2, 3 & 4) up to (and including) the Spring of the following year (wave 5).4 The panel element of the QLFS has been relatively under‐utilised in empirical work but is the largest source of panel data on the labour market activity of immigrants in the UK.5 A clear limitation, in this context, is that individuals who move location in order to secure a job are not distinguishable from other sources of attrition. We constructed a series of 16 overlapping panel datasets, the first of which comprises those individuals who are first successfully interviewed in the Spring QLFS of 1997, following them through to the Spring QLFS of 1998. The next panel was first sampled during the Summer QLFS of 1997 (interviews conducted between June and August), and completed its duration in the panel in the Summer QLFS of 1998. Our sixteenth and final panel comprises individuals whose first interview took place during the Winter QLFS of 2000 and whose final interview was undertaken in the Winter QLFS of 2001 (between December 2001 and February 2002). The specific sample we utilise comprises males, aged 16–65, who are resident in the UK and not engaged in full‐time education. We select those who report currently experiencing a spell of ILO unemployment6 at least once, during their time in one of the QLFS panels described above, and who report country of birth information. The resultant sample of 60,890 observations is based on 16,435 individuals, who are present for an average of 3.7 quarters (or waves). We classify our groups of interest primarily according to country of birth and self‐reported ethnicity. We distinguish between males who report their country of birth as the UK, according whether their ethnicity is White (termed White UK born) or other than White (termed ethnic minority UK born). Immigrants, those born outside the UK, comprise four groups: White immigrants report their ethnicity as white; Black immigrants report an ethnicity other than white and were born in the Caribbean and Africa; South Asian immigrants were born in Bangladesh, India and Pakistan, and do not report a White ethnic origin, or were born in East Africa and report an Indian, Pakistani or Bangladeshi ethnicity; Other immigrants constitute all other males born outside the UK, mainly in China and Southeast Asia, who report an ethnicity other than White. The main methods of job search are defined as follows. The variable Job Centre includes the three questionnaire categories of visiting a Job Centre/Job Market or Training and Employment Agency Office, visiting a Careers Office and visiting a Job Club. Adverts/Newspapers indicates the job search activities of advertising for jobs in newspapers, journals or on the internet, answering advertisements in those sources and studying situations vacant columns in the same media. The next two variables (Direct to Employer and Social Networks) represent just one questionnaire category each, namely that of applying directly to an employer and that of asking friends, relatives, colleagues or trade unions about job opportunities. The final variable (Agency/Other) covers the six remaining questionnaire categories of job search which are: having your name on the books of a private employment agency (which accounts for the largest proportion of responses for this variable), waiting for the results of an application for a job, looking for premises or equipment for a job, seeking any kind of permit to be able to do a job, trying to get a loan or other financial backing for a job or business and doing anything else to find work. The duration of job search is given by the number of months since the individual last left their last job or full‐time education, as appropriate. 2. Sample Characteristics and Descriptive Analysis 2.1. Sample Characteristics In Table 1 we present the salient features of our sample. White UK born men comprise the vast majority of the sample with 14,126 individuals (86.0%) and are observed for an average of 3.74 quarters. Ethnic minority UK born males contribute just 627 individuals (3.8%) and are present for 3.26 quarters on average. Ideally we would compare immigrants with UK born members of the same ethnic group but our sample is too small for reliable disaggregated analyses. The remaining 10.2% of our sample is comprised of immigrants whose average length in the panel varies between 3.35 and 3.65 waves. The fact that these groups have higher rates of attrition in our data suggests some caution should be exercised in interpreting out results. If it results from immigrants being more likely or more willing to move location in order to obtain a job than UK born Whites we will be under‐estimating their job‐finding probabilities. Table 1 Sample Means . White UK born . Ethnic minority UK born . White immigrants . Black immigrants . South Asian immigrants . Other immigrants . Age 36.1 25.4 38.0 39.5 39.7 39.4 Single 45.4 76.0 37.5 30.0 13.1 35.4 Married/Cohabiting 44.9 20.1 52.1 49.9 81.8 50.7 Widowed/Divorced/ Separated 9.7 3.9 10.3 20.1 5.1 13.0 Number of children 0.58 0.64 0.59 0.72 1.56 0.63 Limiting long‐term illness 20.0 12.9 21.1 18.0 22.3 20.1 Degree or equivalent 9.3 14.1 12.7 19.8 8.7 11.6 Higher vocational qualification 5.2 4.2 3.8 4.4 3.1 3.0 ‘A’ level or equivalent 24.5 19.7 19.8 14.5 10.9 16.1 ‘O’ level or equivalent 20.9 26.0 12.4 10.1 9.3 10.7 Other qualification 16.2 16.2 28.5 34.6 34.3 34.0 No qualifications 23.9 19.8 22.8 16.6 33.7 24.6 Foreign highest qualification 0.0 0.0 44.3 62.0 45.9 56.3 Years since immigration 0.0 0.0 21.1 16.7 21.5 20.0 Number of observations 52,829 2,042 2,332 1,043 1,818 826 Number of individuals 14,126 627 672 311 498 245 Percentage of sample 86.0 3.8 4.1 1.9 3.0 1.5 Average length in panel 3.74 3.26 3.47 3.35 3.65 3.37 . White UK born . Ethnic minority UK born . White immigrants . Black immigrants . South Asian immigrants . Other immigrants . Age 36.1 25.4 38.0 39.5 39.7 39.4 Single 45.4 76.0 37.5 30.0 13.1 35.4 Married/Cohabiting 44.9 20.1 52.1 49.9 81.8 50.7 Widowed/Divorced/ Separated 9.7 3.9 10.3 20.1 5.1 13.0 Number of children 0.58 0.64 0.59 0.72 1.56 0.63 Limiting long‐term illness 20.0 12.9 21.1 18.0 22.3 20.1 Degree or equivalent 9.3 14.1 12.7 19.8 8.7 11.6 Higher vocational qualification 5.2 4.2 3.8 4.4 3.1 3.0 ‘A’ level or equivalent 24.5 19.7 19.8 14.5 10.9 16.1 ‘O’ level or equivalent 20.9 26.0 12.4 10.1 9.3 10.7 Other qualification 16.2 16.2 28.5 34.6 34.3 34.0 No qualifications 23.9 19.8 22.8 16.6 33.7 24.6 Foreign highest qualification 0.0 0.0 44.3 62.0 45.9 56.3 Years since immigration 0.0 0.0 21.1 16.7 21.5 20.0 Number of observations 52,829 2,042 2,332 1,043 1,818 826 Number of individuals 14,126 627 672 311 498 245 Percentage of sample 86.0 3.8 4.1 1.9 3.0 1.5 Average length in panel 3.74 3.26 3.47 3.35 3.65 3.37 Open in new tab Table 1 Sample Means . White UK born . Ethnic minority UK born . White immigrants . Black immigrants . South Asian immigrants . Other immigrants . Age 36.1 25.4 38.0 39.5 39.7 39.4 Single 45.4 76.0 37.5 30.0 13.1 35.4 Married/Cohabiting 44.9 20.1 52.1 49.9 81.8 50.7 Widowed/Divorced/ Separated 9.7 3.9 10.3 20.1 5.1 13.0 Number of children 0.58 0.64 0.59 0.72 1.56 0.63 Limiting long‐term illness 20.0 12.9 21.1 18.0 22.3 20.1 Degree or equivalent 9.3 14.1 12.7 19.8 8.7 11.6 Higher vocational qualification 5.2 4.2 3.8 4.4 3.1 3.0 ‘A’ level or equivalent 24.5 19.7 19.8 14.5 10.9 16.1 ‘O’ level or equivalent 20.9 26.0 12.4 10.1 9.3 10.7 Other qualification 16.2 16.2 28.5 34.6 34.3 34.0 No qualifications 23.9 19.8 22.8 16.6 33.7 24.6 Foreign highest qualification 0.0 0.0 44.3 62.0 45.9 56.3 Years since immigration 0.0 0.0 21.1 16.7 21.5 20.0 Number of observations 52,829 2,042 2,332 1,043 1,818 826 Number of individuals 14,126 627 672 311 498 245 Percentage of sample 86.0 3.8 4.1 1.9 3.0 1.5 Average length in panel 3.74 3.26 3.47 3.35 3.65 3.37 . White UK born . Ethnic minority UK born . White immigrants . Black immigrants . South Asian immigrants . Other immigrants . Age 36.1 25.4 38.0 39.5 39.7 39.4 Single 45.4 76.0 37.5 30.0 13.1 35.4 Married/Cohabiting 44.9 20.1 52.1 49.9 81.8 50.7 Widowed/Divorced/ Separated 9.7 3.9 10.3 20.1 5.1 13.0 Number of children 0.58 0.64 0.59 0.72 1.56 0.63 Limiting long‐term illness 20.0 12.9 21.1 18.0 22.3 20.1 Degree or equivalent 9.3 14.1 12.7 19.8 8.7 11.6 Higher vocational qualification 5.2 4.2 3.8 4.4 3.1 3.0 ‘A’ level or equivalent 24.5 19.7 19.8 14.5 10.9 16.1 ‘O’ level or equivalent 20.9 26.0 12.4 10.1 9.3 10.7 Other qualification 16.2 16.2 28.5 34.6 34.3 34.0 No qualifications 23.9 19.8 22.8 16.6 33.7 24.6 Foreign highest qualification 0.0 0.0 44.3 62.0 45.9 56.3 Years since immigration 0.0 0.0 21.1 16.7 21.5 20.0 Number of observations 52,829 2,042 2,332 1,043 1,818 826 Number of individuals 14,126 627 672 311 498 245 Percentage of sample 86.0 3.8 4.1 1.9 3.0 1.5 Average length in panel 3.74 3.26 3.47 3.35 3.65 3.37 Open in new tab The White UK born sample has an average age of 36.1 years whilst all the immigrant groups are slightly older, averaging around 39 years old. In striking contrast, the average age of the ethnic minority UK born is only 25.4, broadly reflecting the relatively recent immigration of their parents. Naturally this group is also most likely to be single whereas immigrants, notably South Asian men, are more likely to be in domestic partnerships than the White UK born. In addition, all immigrant groups, except the Whites, have more children on average but there are no substantial differences in the prevalence of limiting long‐term illnesses amongst the older groups. There are clear differences in the educational attainment of the men in our sample. Ethnic minority UK born men, White and Black immigrants are more likely to possess a degree and less likely to have no educational qualifications than the White UK born. In contrast South Asian immigrants have a much weaker educational attainment profile with over a third reporting no qualifications. Furthermore, approximately half of all immigrants in our sample attained their highest educational qualification before they immigrated to the UK. Their average number of years since immigration is 16.7 for Black immigrants and around 20 for all other immigrant groups. The proportion of the sample that report currently being ILO unemployed in any spell is on average 54.1%. This proportion declines gradually with duration in the panel, from 60.2% in wave 1 to 50.1% in wave 5, reflecting both attrition and exits from unemployment. Within the data we observe 25.2% of UK born Whites, who were unemployed in one quarter, reporting being in work in the subsequent quarter. This observed exit rate into employment is substantially lower for ethnic minority UK born men (20.7%) and all immigrant groups (22.7% for White immigrants, 17.6% for Black immigrants, 16.7% for South Asian immigrants and 23.6% for Other immigrants). 2.2. Methods of Job Search In Table 2 we report a descriptive analysis of the job search activity of our sample of currently ILO unemployed men. All respondents are looking for a job as an employee or for self‐employment (or for both) in the four complete weeks prior to interview, though some may be waiting to begin employment which has already been secured. The percentage of the unemployed reporting to use one of the five broad job search activities as their main method of looking for work is provided in Table 2. Additionally, in the right‐hand column, we report the mean number of broad methods used by each ethnic/immigrant group. The greatest number of significant differences in these means, relative to the White UK born, is evident for White and South Asian immigrants. Table 2 Main Method of Job Search and Total Number used by Current ILO Unemployed Percentage . Job Centre . Adverts/ Newspapers . Direct to employer . Social networks . Agency/ Other . Total number . All 35.9 35.1 8.2 9.8 9.2 3.53 (0.3) (0.3) (0.2) (0.2) (0.2) (0.01) White UK born 36.2 35.9 7.9 9.5 8.8 3.54 (0.3) (0.3) (0.2) (0.2) (0.2) (0.01) Ethnic minority UK born 36.9 31.3* 8.9 9.1 12.6* 3.64* (1.4) (1.3) (0.8) (0.8) (1.0) (0.04) White immigrants 28.6* 33.5x 11.0* 11.8* 12.9* 3.27* (1.3) (1.3) (0.9) (0.9) (0.9) (0.04) Black immigrants 35.9 32.9 7.1 11.3 12.0* 3.32* (1.9) (1.9) (1.0) (1.3) (1.3) (0.05) South Asian immigrants 38.3 24.4* 12.2* 16.1* 8.5 3.46* (1.5) (1.3) (1.0) (1.1) (0.9) (0.04) Other immigrants 26.9* 33.2 14.7* 12.7 10.8 3.32* (2.1) (2.2) (1.6) (1.5) (1.4) (0.07) Percentage . Job Centre . Adverts/ Newspapers . Direct to employer . Social networks . Agency/ Other . Total number . All 35.9 35.1 8.2 9.8 9.2 3.53 (0.3) (0.3) (0.2) (0.2) (0.2) (0.01) White UK born 36.2 35.9 7.9 9.5 8.8 3.54 (0.3) (0.3) (0.2) (0.2) (0.2) (0.01) Ethnic minority UK born 36.9 31.3* 8.9 9.1 12.6* 3.64* (1.4) (1.3) (0.8) (0.8) (1.0) (0.04) White immigrants 28.6* 33.5x 11.0* 11.8* 12.9* 3.27* (1.3) (1.3) (0.9) (0.9) (0.9) (0.04) Black immigrants 35.9 32.9 7.1 11.3 12.0* 3.32* (1.9) (1.9) (1.0) (1.3) (1.3) (0.05) South Asian immigrants 38.3 24.4* 12.2* 16.1* 8.5 3.46* (1.5) (1.3) (1.0) (1.1) (0.9) (0.04) Other immigrants 26.9* 33.2 14.7* 12.7 10.8 3.32* (2.1) (2.2) (1.6) (1.5) (1.4) (0.07) Notes: Standard error of mean in parentheses (adjusted for repeated individual observations). *Indicates the mean is significantly different from that of the White UK born at the 5% level. Open in new tab Table 2 Main Method of Job Search and Total Number used by Current ILO Unemployed Percentage . Job Centre . Adverts/ Newspapers . Direct to employer . Social networks . Agency/ Other . Total number . All 35.9 35.1 8.2 9.8 9.2 3.53 (0.3) (0.3) (0.2) (0.2) (0.2) (0.01) White UK born 36.2 35.9 7.9 9.5 8.8 3.54 (0.3) (0.3) (0.2) (0.2) (0.2) (0.01) Ethnic minority UK born 36.9 31.3* 8.9 9.1 12.6* 3.64* (1.4) (1.3) (0.8) (0.8) (1.0) (0.04) White immigrants 28.6* 33.5x 11.0* 11.8* 12.9* 3.27* (1.3) (1.3) (0.9) (0.9) (0.9) (0.04) Black immigrants 35.9 32.9 7.1 11.3 12.0* 3.32* (1.9) (1.9) (1.0) (1.3) (1.3) (0.05) South Asian immigrants 38.3 24.4* 12.2* 16.1* 8.5 3.46* (1.5) (1.3) (1.0) (1.1) (0.9) (0.04) Other immigrants 26.9* 33.2 14.7* 12.7 10.8 3.32* (2.1) (2.2) (1.6) (1.5) (1.4) (0.07) Percentage . Job Centre . Adverts/ Newspapers . Direct to employer . Social networks . Agency/ Other . Total number . All 35.9 35.1 8.2 9.8 9.2 3.53 (0.3) (0.3) (0.2) (0.2) (0.2) (0.01) White UK born 36.2 35.9 7.9 9.5 8.8 3.54 (0.3) (0.3) (0.2) (0.2) (0.2) (0.01) Ethnic minority UK born 36.9 31.3* 8.9 9.1 12.6* 3.64* (1.4) (1.3) (0.8) (0.8) (1.0) (0.04) White immigrants 28.6* 33.5x 11.0* 11.8* 12.9* 3.27* (1.3) (1.3) (0.9) (0.9) (0.9) (0.04) Black immigrants 35.9 32.9 7.1 11.3 12.0* 3.32* (1.9) (1.9) (1.0) (1.3) (1.3) (0.05) South Asian immigrants 38.3 24.4* 12.2* 16.1* 8.5 3.46* (1.5) (1.3) (1.0) (1.1) (0.9) (0.04) Other immigrants 26.9* 33.2 14.7* 12.7 10.8 3.32* (2.1) (2.2) (1.6) (1.5) (1.4) (0.07) Notes: Standard error of mean in parentheses (adjusted for repeated individual observations). *Indicates the mean is significantly different from that of the White UK born at the 5% level. Open in new tab Amongst the UK born Job Centre and Adverts/Newspapers are by far the most common main job search methods, being used by about 35% of respondents each. Each of the other three broad job search activities are utilised by approximately 10% of the ILO unemployed UK born samples as their main method. For White immigrants and Other (ethnic minority) immigrants Adverts/Newspapers are the commonest form of main job search method, whereas for Black and South Asian immigrants Job Centres are most likely to be used. Only 24.4% of ILO unemployed South Asian immigrants make use of Adverts/Newspapers as their main job search method which may partially be due to their lower levels of English language fluency. In contrast, this group of immigrants are the most likely to rely mainly upon their Social Networks for job search. All immigrant groups use this strategy to a greater extent than the UK born. All immigrant groups also use direct approaches to employers as the main method of job search more commonly, than do the UK born, with the exception of Black immigrants who use this method the least. All groups, other than South Asian immigrants, rely on Agency and Other job search methods to a greater extent than White UK born men. Interestingly, there is some indication that immigrant job search activity is more limited in scope than that of the UK born as the total number of job search methods reported is significantly lower for all immigrant groups. 2.3. Successful Job Search Methods In Table 3 we report the percentage of previously ILO unemployed individuals, subsequently observed in work, who report the actual job search method that resulted in them getting a job.7 Interestingly, successful employment outcomes do not vary enormously by the method used. For White UK born men Social Networks (27.3%) account for the largest proportion of successful job search methods, followed by Adverts/Newspapers (23.5%), Job Centres (19.7%), Agency/Other (18.6%) with Direct approaches to Employers (10.9%) being the least fruitful activity. Interestingly, Social Networks are also the most profitable avenue of job search for all immigrant groups, with around 25% of White and Black immigrants, and more than 36% of South Asian and Other immigrants, who obtained a job, using this method. For ethnic minority UK born men, Job Centres accounted for 26.5% of job outcomes, with direct approaches to employers also being more successful for this group than for White UK born. For all immigrant groups Agency and Other categories of job search proved the second most successful method, whilst Direct to Employer was also a more profitable employment route for ethnic minority immigrants than for White UK born men. Again we find that White and South Asian immigrants are significantly different from the White UK born in several job search categories. Of those who found employment, the vast majority were in permanent employment. This proportion was highest for White UK born men (73.9%), South Asian (73.0%), Other (70.9%) and White (70.5%) immigrant men, but lowest for ethnic minority UK born (68.9%) and Black immigrant men (68.2%). Table 3 Actual Successful Job Search Methods Used Percentage . Job Centre . Adverts/ Newspapers . Direct to employer . Social networks . Agency/ Other . All 19.8 22.9 11.0 17.4 18.9 (0.5) (0.5) (0.4) (0.5) (0.5) White UK born 19.7 23.5 10.9 27.3 18.6 (0.5) (0.5) (0.4) (0.6) (0.5) Ethnic minority UK born 26.5* 20.2 12.6 22.7 18.1 (2.9) (1.6) (2.1) (18.1) (2.5) White immigrants 21.5 19.9 7.2* 26.7 24.7* (2.6) (2.5) (1.6) (2.8) (2.7) Black immigrants 18.0 19.1 16.9 24.7 21.4 (4.1) (4.2) (4.0) (4.1) (4.4) South Asian immigrants 13.4* 16.2* 12.7 36.6* 21.1 (2.9) (3.1) (2.8) (4.1) (2.7) Other immigrants 14.5 11.8* 13.2 36.8 23.7 (4.1) (3.7) (3.9) (5.6) (4.9) Percentage . Job Centre . Adverts/ Newspapers . Direct to employer . Social networks . Agency/ Other . All 19.8 22.9 11.0 17.4 18.9 (0.5) (0.5) (0.4) (0.5) (0.5) White UK born 19.7 23.5 10.9 27.3 18.6 (0.5) (0.5) (0.4) (0.6) (0.5) Ethnic minority UK born 26.5* 20.2 12.6 22.7 18.1 (2.9) (1.6) (2.1) (18.1) (2.5) White immigrants 21.5 19.9 7.2* 26.7 24.7* (2.6) (2.5) (1.6) (2.8) (2.7) Black immigrants 18.0 19.1 16.9 24.7 21.4 (4.1) (4.2) (4.0) (4.1) (4.4) South Asian immigrants 13.4* 16.2* 12.7 36.6* 21.1 (2.9) (3.1) (2.8) (4.1) (2.7) Other immigrants 14.5 11.8* 13.2 36.8 23.7 (4.1) (3.7) (3.9) (5.6) (4.9) Notes: Calculated for individuals who reported gaining employment in the 3 months prior to interview. Standard error of mean in parentheses (adjusted for repeated individual observations). * indicates the mean is significantly different from that of the White UK born at the 5% level. Open in new tab Table 3 Actual Successful Job Search Methods Used Percentage . Job Centre . Adverts/ Newspapers . Direct to employer . Social networks . Agency/ Other . All 19.8 22.9 11.0 17.4 18.9 (0.5) (0.5) (0.4) (0.5) (0.5) White UK born 19.7 23.5 10.9 27.3 18.6 (0.5) (0.5) (0.4) (0.6) (0.5) Ethnic minority UK born 26.5* 20.2 12.6 22.7 18.1 (2.9) (1.6) (2.1) (18.1) (2.5) White immigrants 21.5 19.9 7.2* 26.7 24.7* (2.6) (2.5) (1.6) (2.8) (2.7) Black immigrants 18.0 19.1 16.9 24.7 21.4 (4.1) (4.2) (4.0) (4.1) (4.4) South Asian immigrants 13.4* 16.2* 12.7 36.6* 21.1 (2.9) (3.1) (2.8) (4.1) (2.7) Other immigrants 14.5 11.8* 13.2 36.8 23.7 (4.1) (3.7) (3.9) (5.6) (4.9) Percentage . Job Centre . Adverts/ Newspapers . Direct to employer . Social networks . Agency/ Other . All 19.8 22.9 11.0 17.4 18.9 (0.5) (0.5) (0.4) (0.5) (0.5) White UK born 19.7 23.5 10.9 27.3 18.6 (0.5) (0.5) (0.4) (0.6) (0.5) Ethnic minority UK born 26.5* 20.2 12.6 22.7 18.1 (2.9) (1.6) (2.1) (18.1) (2.5) White immigrants 21.5 19.9 7.2* 26.7 24.7* (2.6) (2.5) (1.6) (2.8) (2.7) Black immigrants 18.0 19.1 16.9 24.7 21.4 (4.1) (4.2) (4.0) (4.1) (4.4) South Asian immigrants 13.4* 16.2* 12.7 36.6* 21.1 (2.9) (3.1) (2.8) (4.1) (2.7) Other immigrants 14.5 11.8* 13.2 36.8 23.7 (4.1) (3.7) (3.9) (5.6) (4.9) Notes: Calculated for individuals who reported gaining employment in the 3 months prior to interview. Standard error of mean in parentheses (adjusted for repeated individual observations). * indicates the mean is significantly different from that of the White UK born at the 5% level. Open in new tab 2.4. Job Search Methods By Duration of Unemployment So far our description of immigrant job search behaviour has overlooked how the main method of job search is affected by the duration of unemployment. In Figures 1 to 6 the distribution of the main search category used, at various durations of unemployment, is plotted for each group of interest. It is clear that the proportion using each main job search method does change as the duration of unemployment increases but not equally for all groups. White UK born men are more likely to rely on Job Centres and Adverts/Newspapers and depend less on Direct approaches to Employers, Social Networks and Agency/Other avenues, as their duration of unemployment increases. In comparison, White immigrants initially rely mainly on Job Centres and Adverts/Newspapers, then increasingly use Social Networks and Agency/Other methods, between 3 and 10 months of unemployment duration, and finally revert more to the former methods. Black immigrants follow a broadly similar pattern, but with a much greater initial reliance on Adverts/Newspapers and private job agencies or other methods. South Asian immigrants depend to a greater extent on Adverts/Newspapers and Job Centres, and less on all other methods, as their unemployment duration increases. All long‐term unemployed men rely mainly on Job Centres and Adverts/Newspapers, with the proportions of these and other categories broadly similar across all groups. Fig. 1. Open in new tabDownload slide Actual Main Job Search Method – White UK Born Men Fig. 1. Open in new tabDownload slide Actual Main Job Search Method – White UK Born Men Fig. 2. Open in new tabDownload slide Actual Main Job Search Method – Ethnic Minority UK Born Men Fig. 2. Open in new tabDownload slide Actual Main Job Search Method – Ethnic Minority UK Born Men Fig. 3. Open in new tabDownload slide Actual Main Job Search Method – White Immigrant Men Fig. 3. Open in new tabDownload slide Actual Main Job Search Method – White Immigrant Men Fig. 4. Open in new tabDownload slide Actual Main Job Search Method – Black Immigrant Men Fig. 4. Open in new tabDownload slide Actual Main Job Search Method – Black Immigrant Men Fig. 5. Open in new tabDownload slide Actual Main Job Search Method – South Asian Immigrant Men Fig. 5. Open in new tabDownload slide Actual Main Job Search Method – South Asian Immigrant Men Fig. 6. Open in new tabDownload slide Actual Main Job Search Method – Other Immigrant Men Fig. 6. Open in new tabDownload slide Actual Main Job Search Method – Other Immigrant Men 3. Results From Econometric and Decomposition Analysis 3.1. Unemployment Duration and Choice of Job Search Method Clearly the actual patterns reported above are confounded by differences in the average characteristics across groups and by the selection process (into work) over time, as some individuals get jobs whilst others remain ILO unemployed. Hence we have undertaken a more structured investigation of the choice of main job search method and the influence it has on unemployment duration; see Frijters et al. (2003) for further details. Our estimates from a standard Proportional Hazard duration model (see Appendix A.1 for details) are provided in Appendix B. They show a differential speed of exit between UK born Whites and some immigrant groups, conditional on search method. In particular, South Asian immigrants are significantly less likely to exit to a job whilst White immigrants are significantly more likely to do so. No significant differences are found for the other groups. A clear policy‐relevant finding from the duration analysis results is that job search methods that cannot be objectively verified (Direct to Employer, Social Networks and Agency/Other) are generally preferable to directly verifiable methods of jobs search (Job Centres, Adverts/Newspapers). Specifically we find no evidence of a significant differential unemployment exit rate by main job search method for either Black or South Asian immigrants. Hence an insistence on verifiable search effort is likely to be counter‐productive. Such a finding concurs with the results of a Dutch policy‐experiment into rewarding observable search effort, evaluated by van den Berg and van der Klaauw (2002). However, some caution should be given to our finding as it might be that it is lower ‘quality’ individuals who predominantly use Job Centres as their main job search method. These results are limited in their usefulness because they do not directly compare the efficiency of the different job search methods for immigrants relative to the White UK born nor allow for the endogeniety of the job search method choice to the duration of unemployment. Therefore we next we estimated a standard multinomial probit choice of job search method model (see Appendix A.2 for details), which allows for differences across ethnic/immigrant groups. Our findings from estimating this model appear sensible but given that these results comprise 212 parameter estimates, whose effects involve complex interactions, we do present the results here. Rather we combine these results with those from the duration analysis in a decomposition framework. 3.2. Outline of Decomposition Methodology We define 6 individuals, each with average characteristics for the different UK born/immigrant groups in our data. We then explore the determinants of their job‐finding probabilities, at durations of 6 months and 2 years, with the aid of four simulations (see Appendix A.3 for details). These use the results from the duration and job search method choice models discussed above: Simulation 1: We calculate the probability of a transition to a job, for these typical individuals, directly from their respective estimated hazard rates using the duration model alone. Simulation 2: We simulate the probability of a transition to a job, for each representative of the UK born and immigrant groups, as if they possessed average White UK born characteristics. Importantly, the choice of job search method, and unemployment duration parameters used are group specific. Simulation 3: As in Simulation 2, we compute the probability of a transition to a job, for the representative individuals, with their characteristics changed to equal the mean White UK born values. However, here we additionally allow these characteristics to influence the choice of job search method, whilst keeping the unemployment duration parameters group specific. Simulation 3 shows the importance of individual characteristics on job‐finding probabilities through their effect on the choice of job search method. Together with Simulation 2 these findings reveal the full importance of individual characteristics. Simulation 4: We simulate what the job‐finding probabilities would be, for each representative individual at durations of 6 months and 2 years, if their observed characteristics and their influence on job search method choice and success would be precisely that of the White UK born. Residual: As a final calculation we examine the influence of group‐specific unobserved factors on the job finding probabilities of the representative individuals. The residual we report is equal to the job finding probabilities of the White UK born (from Simulation 1) minus the results from Simulation 4. It provides an indication of the relative search effectiveness, or job‐finding success, between White UK born and other ethnic/immigrant groups in the UK, after controlling for the full impact of observables on both the choice of job search methods and unemployment duration. 3.3. Decomposition Results We present the results of the decomposition analysis in Table 4 and discuss the main findings here. Firstly, it is clear from all the simulation results that a typical White male immigrants’ probability of finding a job is most similar to that of an average White UK born men, and that the gap is narrower at 24 months unemployment duration, than at 6 months. Secondly, Other (ethnic minority) immigrants have a similar job‐finding probability to that of ethnic minority UK born men, at both unemployment durations (from Simulation 1). Thirdly, the ethnic minority UK born unemployed have more favourable observable characteristics than the White UK born, on average (compare Simulation 2 with that of Simulation 1) and their lower probability of successful job search cannot be explained by their choice of job search method (compare Simulations 3 and 4 with that of Simulation 2). Hence the unemployment situation of these men would be even worse, were it not for their relative youth and higher qualification levels. This lack of success in exiting unemployment may indicate discriminatory behaviour on the part of employers. Fourthly, Black immigrants and, especially, South Asian immigrant men are much less likely to find a job than all other groups, regardless of the time horizon. The differences in job‐finding probabilities between these two immigrant groups maybe due to lower average English language ability amongst South Asian immigrants; see Dustmann and Fabbri (2003); Leslie and Lindley (2001); Shields and Wheatley Price (2001, 2002), for evidence. Table 4 Job‐Finding Probabilities for Typical Individuals Duration . Simulation 1 . Simulation 2 . Simulation 3 . Simulation 4 . Residual . After 6 months of unemployment White UK born 0.501 – – – – Ethnic minority UK born 0.436 0.370 0.368 0.378 0.123 White immigrants 0.479 0.494 0.482 0.477 0.024 Black immigrants 0.399 0.409 0.375 0.406 0.096 South Asian immigrants 0.374 0.412 0.385 0.357 0.144 Other immigrants 0.431 0.455 0.428 0.430 0.071 After 24 months of unemployment White UK born 0.916 – – – – Ethnic minority UK born 0.877 0.817 0.816 0.819 0.097 White immigrants 0.906 0.915 0.907 0.905 0.011 Black immigrants 0.855 0.864 0.835 0.842 0.074 South Asian immigrants 0.792 0.832 0.818 0.790 0.126 Other immigrants 0.878 0.895 0.869 0.862 0.053 Duration . Simulation 1 . Simulation 2 . Simulation 3 . Simulation 4 . Residual . After 6 months of unemployment White UK born 0.501 – – – – Ethnic minority UK born 0.436 0.370 0.368 0.378 0.123 White immigrants 0.479 0.494 0.482 0.477 0.024 Black immigrants 0.399 0.409 0.375 0.406 0.096 South Asian immigrants 0.374 0.412 0.385 0.357 0.144 Other immigrants 0.431 0.455 0.428 0.430 0.071 After 24 months of unemployment White UK born 0.916 – – – – Ethnic minority UK born 0.877 0.817 0.816 0.819 0.097 White immigrants 0.906 0.915 0.907 0.905 0.011 Black immigrants 0.855 0.864 0.835 0.842 0.074 South Asian immigrants 0.792 0.832 0.818 0.790 0.126 Other immigrants 0.878 0.895 0.869 0.862 0.053 Open in new tab Table 4 Job‐Finding Probabilities for Typical Individuals Duration . Simulation 1 . Simulation 2 . Simulation 3 . Simulation 4 . Residual . After 6 months of unemployment White UK born 0.501 – – – – Ethnic minority UK born 0.436 0.370 0.368 0.378 0.123 White immigrants 0.479 0.494 0.482 0.477 0.024 Black immigrants 0.399 0.409 0.375 0.406 0.096 South Asian immigrants 0.374 0.412 0.385 0.357 0.144 Other immigrants 0.431 0.455 0.428 0.430 0.071 After 24 months of unemployment White UK born 0.916 – – – – Ethnic minority UK born 0.877 0.817 0.816 0.819 0.097 White immigrants 0.906 0.915 0.907 0.905 0.011 Black immigrants 0.855 0.864 0.835 0.842 0.074 South Asian immigrants 0.792 0.832 0.818 0.790 0.126 Other immigrants 0.878 0.895 0.869 0.862 0.053 Duration . Simulation 1 . Simulation 2 . Simulation 3 . Simulation 4 . Residual . After 6 months of unemployment White UK born 0.501 – – – – Ethnic minority UK born 0.436 0.370 0.368 0.378 0.123 White immigrants 0.479 0.494 0.482 0.477 0.024 Black immigrants 0.399 0.409 0.375 0.406 0.096 South Asian immigrants 0.374 0.412 0.385 0.357 0.144 Other immigrants 0.431 0.455 0.428 0.430 0.071 After 24 months of unemployment White UK born 0.916 – – – – Ethnic minority UK born 0.877 0.817 0.816 0.819 0.097 White immigrants 0.906 0.915 0.907 0.905 0.011 Black immigrants 0.855 0.864 0.835 0.842 0.074 South Asian immigrants 0.792 0.832 0.818 0.790 0.126 Other immigrants 0.878 0.895 0.869 0.862 0.053 Open in new tab For a typical individual from each immigrant group, when compared to an average White UK born unemployed male, it is evident that the vast majority of the difference in the probability of transition from ILO unemployment to a job cannot be explained by: (a) Differences in average observable characteristics between each immigrant group and White UK born men (compare the results of Simulation 2 with that of Simulation 1); (b) Differences in the combined effect of these average characteristic differences and their influence on the choice of main job search method and its success across groups (compare the results of Simulation 4 with that of Simulation 1). As the calculated residuals make clear it is unobserved differences, between White UK born men and the respective immigrant groups, which largely account for the gap in their relative job‐finding probabilities. In other words, immigrant job search appears to be less effective than that of equivalent UK born job seekers. Surprisingly the large differences we noted earlier in education and family circumstances, across these groups, explain only a small minority of their difference in job search success. These findings are consistent with Chiswick's (1982) hypotheses of immigrant job search. An alternative potential explanation is that immigrants are searching in different parts of the labour market to White UK born men. For instance, they might be searching mainly for jobs amongst members of their own ethnic/immigrant group or employment at different skill levels to the UK born. These arguments would imply that the size of the job market that immigrants are searching over is much smaller than that explored by the UK born, resulting in a lower probability of success, regardless of the choice of search method. A further possibility is that employers prefer White UK born workers and hence immigrants inevitably appear to have lower job‐finding probabilities. Unfortunately, our data do not allow us to attempt to distinguish between the various hypotheses. 3.4. The Influence of Years Since Immigration On Immigrant Job Search Behaviour In order to explore whether the length of time that an immigrant has been in the UK influences their average unemployment duration we estimated an additional duration model that included all the human capital variables, and well as immigrant status, and an interaction variable between immigrant status and years since migration.8 For the interaction variable, we found that South Asian immigrants (coefficient = 0.023, t‐statistic = 2.6) and White immigrants (coefficient = 0.011, t‐statistic = 2.1) have significantly higher job‐finding hazards the greater the number of years since immigration. Specifically, for every additional year spent in the UK, South Asians are about 2% more likely to find a job in any 3‐month period, and White immigrants are around 1% more likely. Interestingly, no such significant effect is found for either Black immigrants or Other immigrants. Doing a back‐of‐the‐envelope calculation, whereby we presume that these linear effects can be extrapolated, reveals that it takes a White immigrant 48.5 years after immigration to achieve the same job‐finding probability as an otherwise identical native. It takes a South Asian immigrant 42.6 years to ‘catch up’ with natives. These very long time periods reveal that the first generation immigrants never truly catch up with the White UK born, which is an indirect indication that they are in different labour markets than natives. Whilst these findings are interesting, and may support the contention that lack of language fluency hinders South Asian immigrant job search success, they are also difficult to interpret since, as is well‐known in the immigration literature, parameter estimates for the interaction variable may be confounded by cohort and selection effects; see Borjas (1994) for a review. 4. Conclusion In this article we have explored new evidence concerning whether immigration is a threat to the employment prospects of UK born workers. We have presented a descriptive portrait of the type of job search activities undertaken by different ethnic and immigrant groups in the UK. We have also utilised a large panel data sample of UK born and immigrant ILO unemployed men to investigate the relative effectiveness of these different groups in their job search behaviour. It is quite clear from our analyses that male immigrants in the UK have more trouble finding jobs than White UK born males. Furthermore our decomposition analysis reveals that the choice of search methods explains virtually none of the difference in job‐finding probabilities between UK born Whites and the different immigrant groups. Our results also shed some light on whether verifiable or non‐verifiable job search methods are more successful in securing immigrant employment. We find no support for a policy of constraining immigrant job search to be mainly via verifiable methods, such as using government Job Centres. Non‐verifiable or informal methods, such as the use of Social Networks or Direct approaches to Employers, are at least as effective for these groups. Immigrants therefore appear to do worse from all forms of job search, not just verifiable methods or those requiring social networks. Our findings are consistent with other recent research that has found no evidence of a large negative impact of immigration on native employment in the UK; see Dustmann et al. (2005); Hatton and Tani (2005). Indeed we have shown that male immigrants in the UK, in complete contrast to their public perception as a competitive threat in the labour market, actually experience substantially greater difficulties in accessing employment than do UK born Whites. This may be due to their human capital being less suited to the UK labour market, their search for more restricted types of employment or employer preferences restricting their likelihood of success. Further research is needed to identify how the UK government might best be able to help immigrants into work. Appendix A: Duration Model and Decomposition Analysis A.1. Duration Model The hazard rate θ of individual i with UK born/immigrant group j at time t is modelled as: (1) which means we take the standard Proportional Hazard specification. Here, λt is the baseline hazard (which is allowed to be non‐parametric, i.e. it is taken to be piece‐wise constant); xit is a vector of individual characteristics that are not group‐specific, such as education, family circumstances, and year dummies; zit is a vector of variables that differ in content between the UK born and immigrant groups, such as indicators for group status and the job search method used. The hazard of observing someone making a transition to work between Tk−1 and Tk is then: (2) The advantage of this hazard model is that it naturally takes account of right‐censoring. We allow for several time‐varying characteristics, such as the number of children, marriage, and year effects.9 A.2. Choice of Job Search Method Model We model the choice of main search method as an optimal choice problem, whereby the expected pay‐off to individual i of UK born/immigrant group j at time t of choosing search method s equals (3) where wit now includes group‐specific variables which are not however all the same as zit. Here, the pay‐off Πijst implicitly includes the probability of finding a job with search method s and the costs of using search method s. We take eist to have a normal distribution and is taken to be orthogonal to any eilt with l ≠ s. The probability of observing an individual at time t choosing method s is thus the probability that method s has the highest pay‐off: (4) where Sit is the observed main search method at time t by individual i, and Φ[] denotes the standard normal cumulative density function. Estimating this structural model requires 5‐dimensional integration, which we tackle by simulated likelihood in Gauss. For this model we include in xit the same time‐invariant and time‐variant individual characteristics (excluding the search method variables) as in the duration model. We therefore estimate 16 parameters per search method s. In wit we include a baseline function, defined on the same 6 intervals as in the duration analysis, which is specific to each of the 6 separate UK born/immigrant groups. Here we estimate 36 variables per search method s. We normalise δ1 = ηj1 = 0 and normalise σ1 to equal 1. This leaves 52 × 4 parameters and 4 variances to be estimated. A.3. Decomposition Analysis We explore the determinants of job‐finding probabilities for typical individuals. We define 6 different typical individuals, one for each UK born/immigrant group in our data. The White UK born typical individual for instance will have a hazard rate equal to where and denotes the average individual and search characteristics of the White UK born. We will conduct four simulations with these 6 typical individuals. Simulation 1: We calculate the probability of a transition to a job within 6 months or within 2 years for each of these typical types, directly from their respective hazard rates. Note that this statistic cannot really be attained without using a duration model because the data do not actually have the specific information on employment transitions at precisely 6 months or 2 years. Simulation 2: We compute the probability of a transition to a job within 6 months or within 2 years for each UK born/immigrant group whilst setting the for every group equal to . The hazard rate for Black Immigrants is then for instance equal to Simulation 3: We compute the probability of a transition to a job within 6 months or within 2 years for each UK born/immigrant group whilst setting the for every group equal to the predicted . This means inserting for each individual in (3), and then predicting for each individual and each time period the average search behaviour given White UK born characteristics. These predictions require micro‐simulations because they involve all possible values of the 5‐dimensional error‐structure in (3). The hazard rate for Black Immigrants then for instance becomes equal to . These outcomes tell us about the importance of on job‐finding probabilities via the choice of search method. Simulations 2 and 3 together reveal the importance of on job‐finding probabilities. Simulation 4: We compute for each UK born/immigrant group what their job‐finding probabilities over the two specified periods would be if their search behaviour would be precisely that of the White UK born. This for instance means for the Black immigrants that their hazard rates would be . This reveals the importance of group status on job finding probabilities via search choice. Residual: The residual of these simulations, which is equal to the job finding probabilities of the White UK born minus the outcomes of simulation 4, tell us what the importance is of the parameters γj, i.e. how much the differences in search effectiveness amongst UK born/immigrant groups are important for the probability of finding a job. Together with Simulation 4, the residual is informative about the importance of unobserved factors. Appendix B: Duration Model Results 
 Main Duration Analysis of Length of Unemployment Spell 
 General variables . Search method variables . β . |t| . . β . |t| . Unemployed 0–3 months −1.642 19.7 White UK born Unemployed 4–6 months −1.145 14.7 Adverts/Newspapers 0.210 4.9 Unemployed 7–9 months −1.201 14.8 Direct to employer 0.833 15.7 Unemployed 10–12 months −1.394 16.3 Social networks 1.418 34.0 Unemployed 13–24 months −1.553 19.5 Agency/Other 1.147 25.0 Unemployed >24 months −2.307 28.5 Ethnic Minority UK born Number of children −0.058 3.9 Adverts/Newspapers −0.374 1.6 Age −0.018 12.8 Direct to employer −0.283 1.1 Married/Cohabiting 0.219 5.6 Social networks −0.265 1.2 Widowed/Divorced/Separated −0.227 3.6 Agency/Other −0.567 2.4 1998 0.015 0.3 White Immigrants 1999 0.037 0.7 Adverts/Newspapers −0.398 1.8 2000 0.092 1.6 Direct to employer −1.147 3.6 2001 0.158 2.6 Social networks −0.432 2.1 2002 −0.114 0.7 Agency/Other −0.394 1.9 Degree or equivalent 0.786 14.9 Black Immigrants Higher vocational qualification 0.635 9.3 Adverts/Newspapers −0.369 0.9 ‘A’ level or equivalent 0.520 11.6 Direct to employer 0.561 1.4 ‘O’ level or equivalent 0.480 10.8 Social networks −0.064 0.2 Other qualification 0.456 9.8 Agency/Other −0.493 1.3 Foreign qualification −0.050 0.5 South Asian Immigrants Limiting long‐term illness −0.313 7.5 Adverts/Newspapers 0.458 1.3 Direct to employer 0.278 0.7 Ethnic minority UK born −0.119 0.8 Social networks 0.316 1.0 White immigrant 0.285 1.8 Agency/Other 0.574 1.7 Black immigrant −0.152 0.5 Other Immigrants South Asian immigrant −0.750 2.8 Adverts/Newspapers −1.256 2.6 Other immigrant 0.340 1.1 Direct to employer −0.748 1.7 Social networks −0.196 0.5 Agency/Other −0.430 1.1 Mean log likelihood −0.47332 Quarterly observations 23,772 General variables . Search method variables . β . |t| . . β . |t| . Unemployed 0–3 months −1.642 19.7 White UK born Unemployed 4–6 months −1.145 14.7 Adverts/Newspapers 0.210 4.9 Unemployed 7–9 months −1.201 14.8 Direct to employer 0.833 15.7 Unemployed 10–12 months −1.394 16.3 Social networks 1.418 34.0 Unemployed 13–24 months −1.553 19.5 Agency/Other 1.147 25.0 Unemployed >24 months −2.307 28.5 Ethnic Minority UK born Number of children −0.058 3.9 Adverts/Newspapers −0.374 1.6 Age −0.018 12.8 Direct to employer −0.283 1.1 Married/Cohabiting 0.219 5.6 Social networks −0.265 1.2 Widowed/Divorced/Separated −0.227 3.6 Agency/Other −0.567 2.4 1998 0.015 0.3 White Immigrants 1999 0.037 0.7 Adverts/Newspapers −0.398 1.8 2000 0.092 1.6 Direct to employer −1.147 3.6 2001 0.158 2.6 Social networks −0.432 2.1 2002 −0.114 0.7 Agency/Other −0.394 1.9 Degree or equivalent 0.786 14.9 Black Immigrants Higher vocational qualification 0.635 9.3 Adverts/Newspapers −0.369 0.9 ‘A’ level or equivalent 0.520 11.6 Direct to employer 0.561 1.4 ‘O’ level or equivalent 0.480 10.8 Social networks −0.064 0.2 Other qualification 0.456 9.8 Agency/Other −0.493 1.3 Foreign qualification −0.050 0.5 South Asian Immigrants Limiting long‐term illness −0.313 7.5 Adverts/Newspapers 0.458 1.3 Direct to employer 0.278 0.7 Ethnic minority UK born −0.119 0.8 Social networks 0.316 1.0 White immigrant 0.285 1.8 Agency/Other 0.574 1.7 Black immigrant −0.152 0.5 Other Immigrants South Asian immigrant −0.750 2.8 Adverts/Newspapers −1.256 2.6 Other immigrant 0.340 1.1 Direct to employer −0.748 1.7 Social networks −0.196 0.5 Agency/Other −0.430 1.1 Mean log likelihood −0.47332 Quarterly observations 23,772 Notes: Omitted reference group is single white UK born with no qualifications, no pre‐school age children, no limiting long‐term illness, using Job Centre as main search method in 1997. General variables . Search method variables . β . |t| . . β . |t| . Unemployed 0–3 months −1.642 19.7 White UK born Unemployed 4–6 months −1.145 14.7 Adverts/Newspapers 0.210 4.9 Unemployed 7–9 months −1.201 14.8 Direct to employer 0.833 15.7 Unemployed 10–12 months −1.394 16.3 Social networks 1.418 34.0 Unemployed 13–24 months −1.553 19.5 Agency/Other 1.147 25.0 Unemployed >24 months −2.307 28.5 Ethnic Minority UK born Number of children −0.058 3.9 Adverts/Newspapers −0.374 1.6 Age −0.018 12.8 Direct to employer −0.283 1.1 Married/Cohabiting 0.219 5.6 Social networks −0.265 1.2 Widowed/Divorced/Separated −0.227 3.6 Agency/Other −0.567 2.4 1998 0.015 0.3 White Immigrants 1999 0.037 0.7 Adverts/Newspapers −0.398 1.8 2000 0.092 1.6 Direct to employer −1.147 3.6 2001 0.158 2.6 Social networks −0.432 2.1 2002 −0.114 0.7 Agency/Other −0.394 1.9 Degree or equivalent 0.786 14.9 Black Immigrants Higher vocational qualification 0.635 9.3 Adverts/Newspapers −0.369 0.9 ‘A’ level or equivalent 0.520 11.6 Direct to employer 0.561 1.4 ‘O’ level or equivalent 0.480 10.8 Social networks −0.064 0.2 Other qualification 0.456 9.8 Agency/Other −0.493 1.3 Foreign qualification −0.050 0.5 South Asian Immigrants Limiting long‐term illness −0.313 7.5 Adverts/Newspapers 0.458 1.3 Direct to employer 0.278 0.7 Ethnic minority UK born −0.119 0.8 Social networks 0.316 1.0 White immigrant 0.285 1.8 Agency/Other 0.574 1.7 Black immigrant −0.152 0.5 Other Immigrants South Asian immigrant −0.750 2.8 Adverts/Newspapers −1.256 2.6 Other immigrant 0.340 1.1 Direct to employer −0.748 1.7 Social networks −0.196 0.5 Agency/Other −0.430 1.1 Mean log likelihood −0.47332 Quarterly observations 23,772 General variables . Search method variables . β . |t| . . β . |t| . Unemployed 0–3 months −1.642 19.7 White UK born Unemployed 4–6 months −1.145 14.7 Adverts/Newspapers 0.210 4.9 Unemployed 7–9 months −1.201 14.8 Direct to employer 0.833 15.7 Unemployed 10–12 months −1.394 16.3 Social networks 1.418 34.0 Unemployed 13–24 months −1.553 19.5 Agency/Other 1.147 25.0 Unemployed >24 months −2.307 28.5 Ethnic Minority UK born Number of children −0.058 3.9 Adverts/Newspapers −0.374 1.6 Age −0.018 12.8 Direct to employer −0.283 1.1 Married/Cohabiting 0.219 5.6 Social networks −0.265 1.2 Widowed/Divorced/Separated −0.227 3.6 Agency/Other −0.567 2.4 1998 0.015 0.3 White Immigrants 1999 0.037 0.7 Adverts/Newspapers −0.398 1.8 2000 0.092 1.6 Direct to employer −1.147 3.6 2001 0.158 2.6 Social networks −0.432 2.1 2002 −0.114 0.7 Agency/Other −0.394 1.9 Degree or equivalent 0.786 14.9 Black Immigrants Higher vocational qualification 0.635 9.3 Adverts/Newspapers −0.369 0.9 ‘A’ level or equivalent 0.520 11.6 Direct to employer 0.561 1.4 ‘O’ level or equivalent 0.480 10.8 Social networks −0.064 0.2 Other qualification 0.456 9.8 Agency/Other −0.493 1.3 Foreign qualification −0.050 0.5 South Asian Immigrants Limiting long‐term illness −0.313 7.5 Adverts/Newspapers 0.458 1.3 Direct to employer 0.278 0.7 Ethnic minority UK born −0.119 0.8 Social networks 0.316 1.0 White immigrant 0.285 1.8 Agency/Other 0.574 1.7 Black immigrant −0.152 0.5 Other Immigrants South Asian immigrant −0.750 2.8 Adverts/Newspapers −1.256 2.6 Other immigrant 0.340 1.1 Direct to employer −0.748 1.7 Social networks −0.196 0.5 Agency/Other −0.430 1.1 Mean log likelihood −0.47332 Quarterly observations 23,772 Notes: Omitted reference group is single white UK born with no qualifications, no pre‐school age children, no limiting long‐term illness, using Job Centre as main search method in 1997. Footnotes 1 " Another explanation for a dampened effect of immigration, which is not explored in this article, is that a sizeable proportion of immigrants in the UK may not actually compete for the same jobs (or in the same labour markets) as natives. It is widely observed that non‐white immigrants groups in the UK are highly over‐represented in low paid jobs in certain manufacturing and service industries; see, for example, Shields and Wheatley Price (2002). However, it would then be the case that new immigrants might have the greatest impact on the employment and wage opportunities of existing immigrants. 2 " Previous research into the job search activities of immigrants, and their relative effectiveness, has been very limited. Most of the literature that has looked at job search behaviour in the general population has paid little attention to issues of immigrant status; see, for example, Holzer (1988) for the US; Osberg (1993) for Canada; Gregg and Wadsworth (1996) for Britain and more recently, Boheim and Taylor (2001) for Britain; Addison and Portugal (2002) for Portugal; and Weber and Mahringer (2002) for Austria. However, Dex (1982) focuses on Black/White differences in job search behaviour in Britain and how it is influenced by employer racial discrimination. Beggs and Chapman (1990) formulate some hypotheses of immigrant job search behaviour, which are largely compatible with those of Chiswick (1982) but provide evidence only on unemployment outcomes for immigrants in Australia. Korenman and Turner (1996) show that personal contacts are less likely to result in ethnic minority youths finding employment in the US. Daneshvary et al. (1992) find evidence that immigrants in the US use job search information to the same extent as the US born just 12 years after immigration. 3 " For example, it may be the case that immigrants employ fewer job search methods than the native born due to their limited knowledge of the local labour market institutions (Chiswick, 1982) or smaller social networks (Beggs and Chapman, 1990). In addition, immigrants’ chosen job search methods may not be as effective if they lack appropriate language skills or if their human capital acquired prior to immigration does not transfer perfectly across national labour markets (Chiswick, 1982). We investigate whether these factors dissipate over time using data on years since migration as an empirical proxy. 4 " One important limitation of the QLFS is that approximately 30% of interviews are conducted with a proxy, usually the partner of the actual respondent. However, our variables of interest are likely to be well known to such proxies. 5 " The British Household Panel Survey, used in Boheim and Taylor (2001), yields a sample of 655 males, who experienced an unemployment spell. Immigrants would constitute little more than 10% of this sample. 6 " According to the internationally recognised International Labour Office (ILO) standard, a person is unemployed if they are of working age, without a paid job, are available to start work in the next two weeks and have either looked for work in the previous four weeks, or are waiting to begin a job which has already been secured; see Sly (1994) technical note. 7 " This variable is only available for a small proportion of our sample, namely, those who started a job in the three months prior to interview. Hence we cannot use it in the econometric analyses. The percentage of those who obtained a job reporting the successful search method to be the same as the main reported method of job search used in the previous quarter is 43.9%. This compares with the 63.2% who report using the same main method of job search from one quarter to the next. 8 " The estimates from this additional duration model are available from the authors on request. 9 " Our data do not allow us to include unobserved heterogeneity: including unobserved heterogeneity would force us to look only at those we observe entering unemployment (the flow sample). The number of new entrants we have for each UK born/immigrant group is simply too small, and even then we would still be hampered by the fact that many durations less than 3 months would not be observed. Assuming no unobserved heterogeneity, whilst using very rich observed heterogeneity, has the advantage that the problem of lacking many small spells does not affect the estimates of observed characteristics. It also reduces the importance of functional form for our estimates, because the identification of unobserved heterogeneity in single‐spell data is known to be heavily dependent on functional form (Baker and Melino, 2000). References Addison , J.T. and Portugal , P. ( 2002 ). ‘Job search methods and outcomes’ , Oxford Economic Papers , vol. 54 , pp. 505 – 33 . Google Scholar Crossref Search ADS WorldCat Baker , M. and Melino , A. ( 2000 ). ‘Duration dependence and nonparametric heterogeneity: a Monte Carlo study’ , Journal of Econometrics , vol. 96 , pp. 357 – 93 . Google Scholar Crossref Search ADS WorldCat Beggs , J.J. and Chapman , B.J. ( 1990 ). ‘Search efficiency, skill transferability and immigrant relative unemployment rates in Australia’ , Applied Economics , vol. 22 , pp. 249 – 60 . Google Scholar Crossref Search ADS WorldCat van den Berg , G.J. and van der Klaauw , B. ( 2002 ). ‘Counselling and monitoring of unemployed workers: theory and evidence from a controlled social experiment’ , mimeo , Free University Amsterdam. Blanchard , O.J. , and Diamond , P.A. ( 1994 ). ‘Ranking, unemployment duration, and wages’ , Review of Economic Studies , vol. 61 , pp. 417 – 34 . Google Scholar Crossref Search ADS WorldCat Boheim , R. and Taylor , M. ( 2001 ). ‘Job search methods, intensity and success in Britain in the 1990s’ , Working Paper no. 2001, Essex: Institute of Social and Economic Research . Borjas , G.J. ( 1994 ). ‘The economics of immigration’ , Journal of Economic Literature , vol. 32 , pp. 1667 – 717 . OpenURL Placeholder Text WorldCat Borjas , G.J. ( 2003 ). Author to complete. Card , D. ( 2005 ). ‘Is the new immigration really so bad?’ , Economic Journal , this issue. OpenURL Placeholder Text WorldCat Chiswick , B.R. ( 1982 ). ‘The employment of immigrants in the United States’, in ( W. Fellner, ed.), Contemporary Economic Problems 1982 , pp. 1 – 37 , Washington DC: American Enterprise Institute . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Daneshvary , N. , Herzog Jr, H.W., Hofler , R.A. and Schlottmann , A.M. ( 1992 ). ‘Job search and immigrant assimilation: an earnings frontier approach’ , Review of Economics and Statistics , vol. 74 , pp. 482 – 92 . Google Scholar Crossref Search ADS WorldCat De New , J.P. and Zimmermann , K.F. ( 1994 ). ‘Native wages impacts of foreign labour: a random effects panel analysis’ , Journal of Population Economics , vol. 7 , pp. 177 – 92 . Google Scholar Crossref Search ADS PubMed WorldCat Dex , S. ( 1982 ). ‘Discrimination and job search theories for black British young men’ , Applied Economics , vol. 14 , pp. 533 – 44 . Google Scholar Crossref Search ADS WorldCat Dustmann , C. and Fabbri , F. ( 2003 ). ‘Language proficiency and labour market performance of immigrants in the United Kingdom’ , Economic Journal , vol. 113 , pp. 695 – 717 . Google Scholar Crossref Search ADS WorldCat Dustmann , C. , Fabbri , F. and Preston , I. ( 2005 ). ‘The impact of immigration on the British labour market’ , Economic Journal , this issue. OpenURL Placeholder Text WorldCat Friedberg , R. and Hunt , J. ( 1995 ). ‘The impact of immigration on host country wages, employment and growth’ , Journal of Economic Perspectives , vol. 9 , pp. 23 – 44 . Google Scholar Crossref Search ADS WorldCat Frijters , P. , Shields , M.A. and Wheatley Price , S. ( 2003 ). ‘Immigrant job search in the UK: evidence from panel data’ , Discussion Paper no. 902, Bonn: Institute for the Study of Labor . Gregg , P. and Wadsworth , J. ( 1996 ). ‘How effective are state employment agencies? Jobcentre use and matching in Britain’ , Oxford Bulletin of Economics and Statistics , vol. 58 , pp. 443 – 67 . Google Scholar Crossref Search ADS WorldCat Hatton , T.J. and Tani , M. ( 2005 ). ‘Immigration and inter‐regional mobility in the UK, 1982–2000’ , Economic Journal , this issue. OpenURL Placeholder Text WorldCat Holzer , H. ( 1988 ). ‘Search method used by unemployed youth’ , Journal of Labor Economics , vol. 6 , pp. 1 – 20 . Google Scholar Crossref Search ADS WorldCat Korenman , S. and Turner , S.C. ( 1996 ). ‘Employment contracts and minority‐white wage differences’ , Industrial Relations , vol. 35 , pp. 106 – 22 . OpenURL Placeholder Text WorldCat Leslie , D. and Lindley , J. ( 2001 ). ‘The impact of language ability on the employment and earnings of Britain's ethnic communities’ , Economica , vol. 68 , pp. 587 – 606 . Google Scholar Crossref Search ADS WorldCat Osberg , L. ( 1993 ). ‘Fishing in different pools: job search strategies and job‐finding success in Canada in the early 1980s’ , Journal of Labor Economics , vol. 11 , pp. 348 – 86 . Google Scholar Crossref Search ADS WorldCat Pischke , J.S. and Velling , J. ( 1997 ). ‘Employment effects of immigration to Germany: an analysis based on local labour markets’ , Review of Economics and Statistics , vol. 79 , pp. 594 – 604 . Google Scholar Crossref Search ADS WorldCat Shields , M.A. and Wheatley Price , S. ( 2001 ). ‘Language fluency and immigrant employment prospects: evidence from Britain's ethnic minorities’ , Applied Economics Letters , vol. 8 , pp. 741 – 5 . Google Scholar Crossref Search ADS WorldCat Shields , M.A. and Wheatley Price , S. ( 2002 ). ‘The English language fluency and occupational success of ethnic minority immigrant men living in English metropolitan areas’ , Journal of Population Economics , vol. 15 , pp. 137 – 60 . Google Scholar Crossref Search ADS WorldCat Sly , F. ( 1994 ). ‘Ethnic groups and the labour market’ , Employment Gazette , vol. 102 (May), pp. 147 – 59 . OpenURL Placeholder Text WorldCat Weber , A. and Mahringer , H. ( 2002 ). ‘Choice and success of job search methods’ , Working Paper no. 125 (Economics series). Vienna: Institute for Advanced Studies . Zorlu , A. and Hartog , J. ( 2005 ). ‘The effect of immigration on wages in three European countries’ , Journal of Population Economics , vol. 18 , forthcoming. OpenURL Placeholder Text WorldCat Author notes " The data are used with the permission of the Office for National Statistics, United Kingdom and the Data Archive at the University of Essex. The authors are grateful for helpful comments received from the editor, Steve Machin, two anonymous referees, Jonathan Haskel (discussant) and other participants at the Royal Economics Society Annual Conference at the University of Warwick in April 2003. The usual disclaimer applies. © Royal Economic Society 2005.

Bowles,, Samuel;Park,, Yongjin

doi: 10.1111/j.1468-0297.2005.01042.xpmid: N/A

Abstract We investigate the manner in which a desire to emulate the rich influences individuals’ allocation of time between labour and leisure, greater inequality inducing longer work hours as a result. Data on work hours in ten countries over the period 1963–98 show that greater inequality is indeed associated longer work hours. These ‘Veblen effects’ are large and the estimates are robust using country fixed effects and other specifications. Because consumption inequality is a public bad, a social welfare optimum cannot be implemented by a flat tax on consumption but may be accomplished by more complicated (progressive) consumption taxes. At the close of the nineteenth century, Thorsten Veblen proposed what he termed pecuniary emulation as the foundation of a theory of consumption. Spending, he maintained, is driven by relative status considerations, that is by the desire to be a particular type of person as much as by the desire to enjoy the consumer goods per se. The Joneses, with whom one had to keep up, were not the neighbours but the rich; their level of living became the never‐attainable objective in a consumption arms race among the less well‐to‐do. In The Theory of the Leisure Class, he wrote: The motive is emulation – the stimulus of an invidious comparison ... especially in any community in which class distinctions are quite vague, all canons and reputability and decency and all standards of consumption are traced back by insensible gradations to the usages and thoughts of the highest social and pecuniary class, the wealthy leisure class. (1934, p. 81). While valued by some economists as capturing common‐sense aspects of consumption as a form of status seeking, Veblen's view of social preferences was soon eclipsed by the simpler and more tractable neoclassical theory of the consumer. Relegated to the underworld of economics, Veblen's ideas have nonetheless resonated over the ensuing years in the writing of Duesenberry (1949), Leibenstein (1950) and Galbraith (1958) at the middle of the past century and Schor (1998) and Frank (1997) at the century's close. We investigate the importance of Veblen effects in the determination of work hours, namely the manner in which a desire to emulate the consumption standards of the rich may influence an individual's allocation of time between labour and leisure. Veblen effects are derived from a class of social‐comparison‐based utility functions on which there is a growing literature and some empirical evidence.1Clark and Oswald (1996) for example found that the satisfaction levels reported by British workers (in the British Household Panel Survey) vary inversely with the wage levels of peers. Neumark and Postlewaite (1998), using data from the US NLSY, studied the labour supply decisions of relatives, finding some evidence that women whose sister's husband had a higher income than their own husband were more likely to be employed. These studies provide some support for comparison‐based utility functions, but do not test Veblen effects directly. An explicitly Veblen‐inspired study by Schor (1998) using a US sample asked respondents how their ‘financial status’ compared to that of those in their reference group (primarily co‐workers and friends). While a majority of her sample responded that they personally did not feel pressure to ‘keep up with the Joneses’, Schor found that, independently of the effects of annual and permanent income and other standard regressors, those whose financial status was below their reference group saved significantly less than those who were better off than their reference group. Interestingly, those who watched TV more saved less, conditional on the other regressors. Our model of the choice of work hours, presented in the next section, captures Veblen effects by taking account of the influence of the consumption of the well‐to‐do on the marginal utility of own consumption of the less well‐off. The main result is that work hours are increasing in the degree of income inequality. We then use data on average annual work hours in ten countries over the period 1963–98, along with data on inequality of income to explore this hypothesis. Inequality is a predictor of work hours in both OLS and fixed‐effects estimates; its effects are large, and estimates are robust across a variety of specifications. This result is consistent with the hypothesis that social comparisons are upwards to a richer reference group and is inconsistent with the alternative hypothesis that social comparisons are downward‐looking, people's consumption and work choices reflecting a desire to distance themselves from a poorer reference group. We then address an alternative interpretation in which a positive relationship between work hours and inequality is due to the incentive effects of the latter (Bell and Freeman, 2001). In the penultimate Section we consider some of the normative implications of Veblen effects, identifying a class of policies which can implement a social welfare optimum: included are subsidies for the leisure of the rich and a graduated consumption tax (but not a flat consumption tax).2 1. Veblen Effects on Work Hours Veblen held that consumption is motivated by a desire for social standing as well as for the enjoyment of the goods and services per se the proximate ground for expenditure in excess of what is required for physical comfort is … a desire to live up to the conventional standard of decency…(1934, p. 81) His key idea (quoted at the outset) was that the best‐off members of a community –‘the leisure class’– establish the standards for the rest. But why is it the consumption of the leisure class that is emulated rather than their leisure? Veblen's response was that under modern conditions consumption is a more visible form of display. The exigencies of the modern industrial system frequently place individuals and households in juxtaposition between whom there is little contact in any other sense than juxtaposition. One's neighbours, mechanically speaking, often are socially not one's neighbours, or even acquaintances; and still their transient good opinion has a high degree of utility. The only practicable means of impressing one's pecuniary ability on these unsympathetic observers of one's everyday life is an unremitting demonstration of the ability to pay. (1934, p. 71) Veblen's ideas are thus a precursor to the contemporary theory of costly signalling of otherwise unobservable qualities initiated in economics by Spence (1973) and in biology by Zahavi (1975).3 The following model embodies the two propositions underlying Veblen's account, namely that people compare consumption (or wealth) but not leisure, and that they refer upwards, choosing their work and spending activities in order to be more like a higher income group, rather than seeking social distance from lower income groups. Suppose individuals differ in some trait that influences hourly wages and that they choose their hours of work (h) to maximise a utility function, the arguments of which are leisure (which we normalise as 1 − h) and what we term effective consumption, c* defined as their own consumption level (c) minus a constant v (for Veblen) times the consumption level of some higher income reference group (c∼). The individual's reference group might be the very rich, or it might be an intermediate group. The reference group's rank in the income distribution is taken as exogenous, as is the Veblen constant v. It may be convenient to think of each individual as belonging to a homogeneous income class, each member of which takes the next highest income class as its reference group (the richest class have no reference group). Together, the reference group and v measure the nature and intensity of the relevant social comparisons. Individuals do not save, so c = wh, where w is the wage rate. Thus for some individual not in the richest group we have (1) where u is increasing and concave in its first argument and decreasing and convex in the second. Leisure and consumption are complements so uc*h < 0. The effect of increased consumption by members of the reference group thus is both to lower the utility of the individual and to raise the marginal utility of effective consumption. The individual will choose hours to be h*, namely that which equates the marginal rate of substitution between leisure and effective consumption to the wage rate.4 We can now consider the effects of an exogenous increase in the wages of the richest group (raising c∼ relative to wh for every income class except the richest). Differentiating the individual's first order condition for the choice of work hours (and using the second order condition) we find that dh*/dc∼ has the sign of −(uc*c* + uc*h), which is positive. The effect of the larger gap between the consumption levels of the individual and the reference group is to reduce effective consumption wh − vc∼, and thus raise the marginal utility of consumption relative to the marginal utility of leisure, inducing an increase in the hours of work. The effect on work hours of variations in the Veblen constant have the same sign (dh*/dv > 0), reflecting an increase in the intensity of social comparison and perhaps capturing the negative effect of TV watching on saving in Schor's study. If, contra Veblen, the reference group were the poor (others seeking to distance themselves from the reference group), then an increase in inequality associated with a decline in the wages of the poorest group would induce a reduction in work hours of other higher income groups. Thus are able to test empirically whether the comparisons are upwards, to a rich reference group as in the Veblen hypothesis, or downwards to a poorer group from which others are seeking social distance. For individuals less well off than the reference group, a simple labour supply function may be derived that is additive in its conventional and Veblen effects. To see this, normalise the wage of the less well‐off to unity and suppose that all (the rich and the not‐so‐rich) share the following utility function (an example of (1) above). (2) In the absence of Veblen effects (v = 0, so c* = wh) each utility maximising individual would select h so as to equate the disutility of labour (δ) to the marginal benefit of labour via its contribution to consumption (1/h), thus setting h = 1/δ. With v > 0 the work hours of the rich are unaffected but those not in the reference group (with superscript n) will now set their work hours at (3) where the superscript r refers to the rich reference group. As can be seen from (3) those not in the reference group work more hours, as we would expect, the second additional term representing the Veblen effect. An extension of this model with many income groups each of which (except the richest) takes the next richest group as its reference group (footnotes 8 and 10) shows that a increase in consumption by the rich generates a downward cascade of Veblen effects, increasing work hours throughout the income distribution. One aspect of the model deserves comment, namely the assumption that individuals choose their hours of work. In a collective bargaining framework or an efficiency wage model, employers play a major role in setting work hours, and the relationship between individual preferences and observed hours may be considerably attenuated. Not surprisingly, a significant fraction of employees in the advanced economies would prefer hours different from what they have (Bell and Freeman, 2001). However in the studies reported, a majority preferred current pay with current hours (rather than more hours and more pay, or less hours and less pay) and Bell and Freeman report evidence that most European Community workers would prefer increases in pay (at the current hours) to decreases in hours (at the current total earnings) suggesting that they are close to the hours they would have chosen, even if the institutional setting allows no direct relationship between individual hours choices and outcomes. Böheim and Taylor (2004) report similar results using the British Household Panel Survey. This evidence that work hours respond to employee preferences may reflect the fact that employers and unions alike have an interest in taking account of employee preferences concerning hours of work (to maximise job rents and improve labour discipline, for example), even if this interest competes with tax and benefits arrangements which sometimes produce significant differences between actual and desired hours. As a result, individual preferences will affect observed work hours even in environments in which employees do not literally choose their work hours. A second comment on the model concerns its behavioural foundations. We do not suppose that people engage in a conscious optimising process in selecting their work hours. A more plausible view is that individuals have norms concerning the appropriate division of their time between family, friends, work, and other activities, and that these norms differ from group to group and evolve over time. Suppose this is the case, and that people simply seek to implement their ‘work hour norm’, occasionally updating this norm in response to two kinds of information: their perceptions of the subjective well‐being of others and the hours of work of others. A plausible model of this learning process would combine payoff‐based updating with conformism: that is, individuals adopt the norms of those in their social group perceived to be happier, but with a conformist bias towards adopting norms held by large numbers of their associates, independently of the associated utility levels (Bowles, 2004). Then the model just presented gives the payoff‐based aspect of the updating of the work hour norm. The main result is that the work hour norm typical of a given group (other than the richest) is increasing in the level of inequality but that the short‐run Veblen effect might be attenuated by conformist effects. 2. Work Hours and Inequality The importance of both social norms and labour market institutions in the determination of work hours suggests that it may be illuminating to study work hours averaged over individuals. We use data on average annual hours of work for ten advanced economies. The annual data for the ten countries presented in Figure 1 indicate substantial and growing differences between economies. The work year in Germany exceeded that in the US by 231 hours in 1960, and had fallen to 365 hours less than the US by 1998. Many countries show a decline in hours prior to the early 1980s followed by a levelling off or increase (in Sweden the work year fell by 388 hours over the first two decades and then increased by 128 hours over the next two decades) Fig. 1. Open in new tabDownload slide Work Hours Over TimeSource:OECD Labour Market Statistics Data Set (http://www1.oecd.org/scripts/cde/members/lfsdataauthenticate.asp) Average annual hours of work are the number of hours worked on average by persons for total employment (average employment over the year) Refer to Data Appendix for detailed definition. Fig. 1. Open in new tabDownload slide Work Hours Over TimeSource:OECD Labour Market Statistics Data Set (http://www1.oecd.org/scripts/cde/members/lfsdataauthenticate.asp) Average annual hours of work are the number of hours worked on average by persons for total employment (average employment over the year) Refer to Data Appendix for detailed definition. Because the reference group for Veblen effects is the rich, we chose a measure of income inequality that is sensitive to upper incomes, namely the ratio of the highest earnings in 90th percentile (that dividing the 90th from the 91st percentile) to the highest earnings in the 50th percentile. We also present estimates using two alternative measures of inequality, the Gini coefficient of after‐tax incomes from the Luxemburg Income Study and a Theil index of inter‐industry wage differences. Figure 2 presents the percentile data along with the annual hours, as well as the country means for these variables. The simple correlation (r = 0.66) is substantial, but as we will see, it arises in part from covarying influences on hours and inequality. Fig. 2. Open in new tabDownload slide Earnings Inequality and Average Annual Work Hours Source:OECD Labour Market Statistics Data Set (http://www1.oecd.org/scripts/cde/members/lfsdataauthenticate.asp) 
Earnings Inequality (percentile ratio) is based on gross earnings of full‐time workers and is the ratio of earnings at 90th percentile to the median earnings (as described in the text). Refer to Data Appendix for detailed definition. 
Square blocks represent country average of each country and diamond shaped points are annual data. Fig. 2. Open in new tabDownload slide Earnings Inequality and Average Annual Work Hours Source:OECD Labour Market Statistics Data Set (http://www1.oecd.org/scripts/cde/members/lfsdataauthenticate.asp) 
Earnings Inequality (percentile ratio) is based on gross earnings of full‐time workers and is the ratio of earnings at 90th percentile to the median earnings (as described in the text). Refer to Data Appendix for detailed definition. 
Square blocks represent country average of each country and diamond shaped points are annual data. We therefore estimate a more complete model. (4) where hit is the natural logarithm of work hours in country i in time t, g is the measure of inequality, xit is a vector of other possible exogenous influences on hours (with c its vector of estimated coefficients), λi is a country fixed effect, δt is a year fixed effect, and μit is an error term. The country fixed effects will take account of cultural and institutional differences and other country‐specific unobserved influences on hours. Among the x‐variables we considered union density (to capture possible time‐varying institutional differences), real gross domestic product per capita (to measure possible influences of income levels on consumption and leisure preferences) and real manufacturing wages (to capture conventional labour supply effects). The latter two were expressed in common units using purchasing power parity conversions. Because hours vary cyclically in response to labour demand rather than to individual labour supply decisions, we also include a measure of aggregate unemployment. To account for changes in the gender composition of the workforce we include the women as a fraction of employment. We included year fixed effects to capture the possible influences of changes in preferences (or other determinants of work hours) possibly reflecting the diffusion of what Inglehart (1977) terms ‘post materialist values’. However, extensive experimentation with the available measures of ‘post materialist values’ did not reveal any systematic results. Finally, we used a measure of government expenditures relative to gross domestic product. The variable proved insignificant while having no appreciable effect on the results reported in table 1. Table 1 Estimates of the Relationship Between Work Hours and Inequality . I . II . III . IV . Constant 9.635 7.833 10.279 9.878 (16.95) (12.16) (30.18) (16.27) Percentile earnings ratio 0.177 0.126 (4.81) (2.95) GINI coefficient (after‐tax income) 0.030 (2.22) Inter‐industry earnings inequality 0.023 0.020 (5.74) (2.81) Ln(real wage) −0.021 −0.041 −0.055 −0.017 (−0.69) (−2.56) (−7.47) (−0.51) Ln(real GDP per capita) −0.234 −0.065 −0.256 −0.243 (−3.70) (−0.98) (−7.30) (−3.57) Union density 0.023 0.002 0.002 0.018 (3.60) (0.30) (0.64) (2.65) Unemployment rate −0.005 −0.008 −0.005 −0.005 (−5.17) (−5.67) (−6.25) (−4.34) Female proportion in employment −0.094 0.038 −0.070 −0.106 (−3.82) (1.17) (−4.35) (−4.18) Country and year fixed effects Yes Yes Yes Yes Observations 155 89 240 143 Adjusted R‐squared  0.958  0.979  0.967  0.961 . I . II . III . IV . Constant 9.635 7.833 10.279 9.878 (16.95) (12.16) (30.18) (16.27) Percentile earnings ratio 0.177 0.126 (4.81) (2.95) GINI coefficient (after‐tax income) 0.030 (2.22) Inter‐industry earnings inequality 0.023 0.020 (5.74) (2.81) Ln(real wage) −0.021 −0.041 −0.055 −0.017 (−0.69) (−2.56) (−7.47) (−0.51) Ln(real GDP per capita) −0.234 −0.065 −0.256 −0.243 (−3.70) (−0.98) (−7.30) (−3.57) Union density 0.023 0.002 0.002 0.018 (3.60) (0.30) (0.64) (2.65) Unemployment rate −0.005 −0.008 −0.005 −0.005 (−5.17) (−5.67) (−6.25) (−4.34) Female proportion in employment −0.094 0.038 −0.070 −0.106 (−3.82) (1.17) (−4.35) (−4.18) Country and year fixed effects Yes Yes Yes Yes Observations 155 89 240 143 Adjusted R‐squared  0.958  0.979  0.967  0.961 Note: The dependent variable is ln(Average annual work hours). The number of observations is limited by the fact that our inequality measures are not available for all years. Open in new tab Table 1 Estimates of the Relationship Between Work Hours and Inequality . I . II . III . IV . Constant 9.635 7.833 10.279 9.878 (16.95) (12.16) (30.18) (16.27) Percentile earnings ratio 0.177 0.126 (4.81) (2.95) GINI coefficient (after‐tax income) 0.030 (2.22) Inter‐industry earnings inequality 0.023 0.020 (5.74) (2.81) Ln(real wage) −0.021 −0.041 −0.055 −0.017 (−0.69) (−2.56) (−7.47) (−0.51) Ln(real GDP per capita) −0.234 −0.065 −0.256 −0.243 (−3.70) (−0.98) (−7.30) (−3.57) Union density 0.023 0.002 0.002 0.018 (3.60) (0.30) (0.64) (2.65) Unemployment rate −0.005 −0.008 −0.005 −0.005 (−5.17) (−5.67) (−6.25) (−4.34) Female proportion in employment −0.094 0.038 −0.070 −0.106 (−3.82) (1.17) (−4.35) (−4.18) Country and year fixed effects Yes Yes Yes Yes Observations 155 89 240 143 Adjusted R‐squared  0.958  0.979  0.967  0.961 . I . II . III . IV . Constant 9.635 7.833 10.279 9.878 (16.95) (12.16) (30.18) (16.27) Percentile earnings ratio 0.177 0.126 (4.81) (2.95) GINI coefficient (after‐tax income) 0.030 (2.22) Inter‐industry earnings inequality 0.023 0.020 (5.74) (2.81) Ln(real wage) −0.021 −0.041 −0.055 −0.017 (−0.69) (−2.56) (−7.47) (−0.51) Ln(real GDP per capita) −0.234 −0.065 −0.256 −0.243 (−3.70) (−0.98) (−7.30) (−3.57) Union density 0.023 0.002 0.002 0.018 (3.60) (0.30) (0.64) (2.65) Unemployment rate −0.005 −0.008 −0.005 −0.005 (−5.17) (−5.67) (−6.25) (−4.34) Female proportion in employment −0.094 0.038 −0.070 −0.106 (−3.82) (1.17) (−4.35) (−4.18) Country and year fixed effects Yes Yes Yes Yes Observations 155 89 240 143 Adjusted R‐squared  0.958  0.979  0.967  0.961 Note: The dependent variable is ln(Average annual work hours). The number of observations is limited by the fact that our inequality measures are not available for all years. Open in new tab We treat g as exogenous. A more adequate approach would take g and h to be jointly determined.5 A plausible exogenous instrument for g proved impossible to find. Thus our results could capture the effect of exogenous shifts in labour supply on the degree of inequality. To test this possibility we use both contemporaneous and one year lagged inequality measures on the right‐hand side of (4). The results with lagged inequality measures (presented in Table 2) are virtually identical to those with contemporaneous regressions. Our view that the endogeneity problem is not accounting for our results is supported by a companion study of the US (Park, 2004) in which the labour force participation of wives of full time full year working men covaries with measures of income inequality among men of similar age and of the same locality. Because it is unlikely that the labour force decisions of wives affected income inequality among men (especially during the period studied, 1969–79) it appears inequality was the cause of increased work hours rather than the converse. Table 2 Alternative Estimates of the Veblen Effect . P90/50 . GINI . THEIL . Using BLS (manufacturing) hours 0.090 0.042 0.033 (2.47) (2.66) (7.60) Using one‐year lagged inequality 0.170 0.049 0.027 (4.45) (3.50) (5.90) Without country fixed effect 0.528 1.015 0.066 (9.30) (7.20) (11.36) . P90/50 . GINI . THEIL . Using BLS (manufacturing) hours 0.090 0.042 0.033 (2.47) (2.66) (7.60) Using one‐year lagged inequality 0.170 0.049 0.027 (4.45) (3.50) (5.90) Without country fixed effect 0.528 1.015 0.066 (9.30) (7.20) (11.36) Note: The dependent variable is ln(Average annual work hours). Open in new tab Table 2 Alternative Estimates of the Veblen Effect . P90/50 . GINI . THEIL . Using BLS (manufacturing) hours 0.090 0.042 0.033 (2.47) (2.66) (7.60) Using one‐year lagged inequality 0.170 0.049 0.027 (4.45) (3.50) (5.90) Without country fixed effect 0.528 1.015 0.066 (9.30) (7.20) (11.36) . P90/50 . GINI . THEIL . Using BLS (manufacturing) hours 0.090 0.042 0.033 (2.47) (2.66) (7.60) Using one‐year lagged inequality 0.170 0.049 0.027 (4.45) (3.50) (5.90) Without country fixed effect 0.528 1.015 0.066 (9.30) (7.20) (11.36) Note: The dependent variable is ln(Average annual work hours). Open in new tab Our estimates appear in Table 1. Our preferred estimate (I) as well as alternative estimates using other measures of inequality (II) and (III) indicate significant positive effects of inequality on work hours. Moreover, these effects are large. A standard deviation change in 90/50 percentile ratio, Gini, and Theil, is associated with a predicted increase in annual hours of 3.4%, 2.2% and 1.8% respectively. Taken literally this means that the difference in the US and Swedish percentile ratio in 1992 accounts for 59% of the difference between the hours of work in the two countries. The estimates also suggest a small (and in the preferred estimate, not significant) negative labour supply elasticity consistent with other estimates of labour supply functions and with the derived labour supply function above (3). The unemployment rate has the predicted coefficient, as does the female proportion in employment. In OLS estimates (not shown) Union Density had a large and statistically significant negative coefficient; but in these country fixed‐effects equations its coefficient is small and positive, suggesting that our country fixed effects may be capturing some of the institutional differences associated with the degree of unionisation. The specific country effects across all of the equations indicate major differences among the countries due to idiosyncratic effects of time invariant cultural, institutional and other country differences uncorrelated with the regressors. Sweden and Norway are similar in their short work year while the English‐speaking countries are distinct and not significantly different from one another in their long work hours; the remainder of the continental countries occupy a middle ground with Belgium closest to the Nordic pattern. The country‐effect difference between the English speaking and the Nordic group is about 295 hours per year, indicating large idiosyncratic effects presumably due to cultural, political, and other differences. We estimated the same fixed‐effects equations as in Table 1, but using as our dependent variable the natural logarithm of the US Bureau of Labour Statistics series on average annual hours of manufacturing workers. This series may provide a more accurate measure of hours (but for a more limited portion of the population.) The results in Table 2, which cover the same countries and time period, show that the coefficients of our three inequality measures are highly significant, and of approximately the same magnitudes as those using the OECD labour hours series. Table 2 also presents the estimated coefficients of one‐year lagged inequality measures. The results with lagged inequality measures are very similar to or slightly stronger than those with contemporaneous regressions, suggesting that our results are not driven by the endogenous relation between work hours and inequality. Lastly, we show the coefficients of inequality measures for a specification without the country fixed effects (but with the year fixed effects). As expected, the estimates of the Veblen effect are considerably larger but these are likely to be upward biased because of the co‐variation of both hours and inequality with time‐invariant country‐specific differences, the effects of which are captured in our fixed‐effects estimates. 3. Other Explanations The fact that inequality predicts work hours is consistent with the Veblen effects proposed at the outset but there are other consistent explanations. Bell and Freeman (2001) have suggested that inequality induces longer work hours because those who work longer hours attain a higher percentile rank in the wage distribution at the workplace and an increase in rank implies greater wage gains the more unequal is the wage distribution. They provide convincing evidence for this effect: in the US and Germany wage inequality within detailed occupation/industry cells is positively correlated with work hours for those working more 35 hours per week and longer. Discriminating empirically between this incentive‐based account and the social comparisons interpretation offered here may be impossible, and it is very likely that both incentive and Veblen effects are at work. However, we are not persuaded that the Bell and Freeman model accounts for the relationship apparent in Figure 2 and Table 1. First, Bell and Freeman treat long hours as an effective signal of a difficult to observe quality likely to result in promotion. While this is true for young lawyers as in the account by Landers et al. (1996), we think it more likely that hard work when on the job (that is, effort, not hours) is a more common way to move up. Second, the fact that their inequality‐hours relationship is much weaker (in both the US and Germany) for all workers (rather than just those working full time or more) is not easy to reconcile with their model. Finally Bell (1998) found that black workers in the US in 1990 are more responsive to measures of earnings inequality among blacks only. Bell suggests that this may be because the black‐only distribution is a better indicator of the gains to working longer hours (but points out that it is not easy to explain why this would be so). A more parsimonious explanation might be that the relevant reference group for black workers is other black workers and their response to measures of black‐only inequality is picking up a Veblen effect. These caveats about the Bell‐Freeman interpretation are far from decisive, however. It would be valuable to see if the evidence for Veblen effects is robust when using a measure of inequality that could not plausibly be related to the incentive effects they stress. Two measures accomplish this. First, the previously‐mentioned study (Park, 2004) showing that wives’ labour force participation covaries with male income inequality cannot plausibly be capturing the incentive effects, unless we have a reason to believe that having a wife with a job has positive implication on a husband's promotion. Second, the most plausible measure of inequality for the incentive effects view would be within firm or within industry inequality, of the type Bell and Freeman used. The reason is that if workers are putting in extra hours to impress their employer, it is the firm's wage structure that is providing the incentive, not the level of inequality within other firms, and less still, the difference in average wages between firms. (Employers in other firms have no way of knowing how many hours a worker puts in.) Thus the Theil index of inter‐industry average wage inequality provides such a test. The fact that this measure of inequality is a significant predictor of work hours (equation III in Table 1) suggests that the Veblen effects model captures some of the causal mechanisms at work, for this measure could not possibly be capturing the Bell‐Freeman incentive effects. Notice (equation IV) that the estimate of its coefficient is reduced only marginally by the addition of the percentile ratio to the equation, suggesting that the estimated effect on work hours in equation III is not primarily due to the correlation of the Theil index with other measures of inequality that may be picking up incentive effects modelled by Bell and Freeman. A second alternative interpretation of the inequality‐hours relationship is that the acceleration of skill‐intensive technical change over the last two decades may have increased inequality and at the same time increased hours of work. Freeman (2002) for example, found that in the US those using computers or the Internet at work put in longer hours, and we know from Krueger (1993) that computer use has raised the economic returns to schooling. Taken together, these two facts suggest that an exogenous increase in computer use may account for a positive correlation between hours and earnings inequality. We do not think this accounts for our results, however, because when we split our time period (at 1983) using the Theil index (the only measure on which we have sufficiently long time series to do this) we find that its estimated coefficient in the early period is almost twice that in the later period.6 4. Consumption Inequality as a Public Bad If Veblen effects of the type modelled here are important, there may be a case for public policies to limit consumption on the conventional grounds that it generates social costs not accounted in the private calculations of the consumer. Frank (1997) and others have proposed a tax exemption for savings on just this grounds.7 Veblen effects are an example of this class of consumption externalities, but with two special characteristics. First, note that the usual consumption externalities are symmetrical (my consumption reduces the well being of the Jones’ I am trying to keep up with, just as theirs reduces mine). But Veblen effects are asymmetrical: if the Jones’ are richer than me, they do not care about my consumption but instead are trying to keep up with some even richer reference group. Thus Veblen effects cascade downward through the income distribution, with the richest group inflicting subjective costs on the next group, whose emulation of the consumption of the rich then augments its own consumption level, thus passing additional subjective costs to the groups further down. A second difference is that the influence of a reference group in the Veblen‐inspired model may be substantially independent of its size, so a relatively small number of well‐off but visible consumers may constitute the reference consumption standard for a much larger number of less well‐off individuals. In this case their consumption decisions may inflict subjective costs on large numbers of less well‐off individuals. For both reasons – the asymmetry of the effects and the differing sizes of various ranks in the income distribution – an appropriate policy response to Veblen effects may be a progressive consumption tax rather than the flat consumption tax implied by symmetrical consumption externalities. To see why this is true take a simple two‐class society in which there are a number (normalised to unity) of well‐off individuals indicated (as above) by the superscript r, and a larger number, n, of less well‐off people. As our point is to clarify the logic of policies to correct Veblen effects rather than to advocate particular policies, we will retain our simplifying assumptions (including that there is no saving). We also set the wage of the less well‐off at unity. Using the utility function (2) and the resulting labour supply function (3), suppose that a social planner wished to know what level of work hours of both groups would maximise the sum of utilities in this society, ω, where (5) The planner would know that in the social optimum the consumption of the well‐off will be less than under private optimisation, and because there are no savings, the only way to accomplish this is to reduce the work hours of the well‐off. As the work hours of the lower group generate no externalities (they are the reference group for no one) the planner would simply vary hr to maximise ω, using (3) to take account of the endogenous response of hn to the planner's chosen level of hr. While private optimisation induces the rich to equate the marginal contribution of work to (private) consumption utility (1/hr) to the (private) disutility of labour (δ), social welfare optimisation requires (6) where the first term on the right is the private cost (disutility of labour) experienced by the rich and the second is the sum of the marginal social cost imposed on those attempting to emulate the well‐to‐do. The aggregate‐welfare maximising level of work hours of the rich is thus given by (7) which shows that the welfare optimum requires the rich to work less than 1/δ by a proportional amount nvwr which is equal to the sum of the loss in effective consumption imposed on the lower income group. The required change in the work hours of the rich is proportional to both the relative size of the two income groups and to their wage rates.8 As the social optimum requires a change in the labour‐leisure allocations of the higher income reference group but not of the lower income group, the social planner will not introduce an across the board consumption tax (applying to both groups). A well‐designed policy will target the consumption of the rich specifically, as it is this which generates the negative externalities. From (6) we see that the implied reduction in the work hours of the rich could be implemented by policies that enhance their marginal utility of leisure (or what is equivalent, increasing their marginal disutility of labour) by a proportional amount δnvwr.9 Suppose the social planner's only instrument is a linear tax on the consumption of the well‐off. The particular utility function used here implies that the tax will not affect the labour hours they perform, so a tax at rate τ will reduce the consumption of the reference group by the same rate. Assuming that the tax revenues, when spent, yield a per dollar contribution to aggregate welfare of β, the planner will vary τ to maximise: (8) The resulting optimal tax rate τ* is given (using (3) and hr = 1/δ) by (9) This requires that the tax rate be selected to equate the marginal benefits of additional taxes (that is the reduced Veblen effects for the less well‐off plus the expenditure benefits, shown on the left‐hand side of (9)) to the marginal costs (in reduced consumption) to the well‐off (the right‐hand side). Setting β = 0 so as to abstract from the expenditure related benefits of the tax policy), and assuming that nvwr < 1, (10) As expected, the optimal tax is increasing in the relative size of the less well‐off group, the size of the Veblen effect and the relative wages of the better‐off group.10 5. Conclusion We have shown that increased inequality induces people to work longer hours and have also provided evidence that the underlying cause is the Veblen effect of the consumption of the rich on the behaviour of those less well off. The effects are large enough to invite attention from policy makers. The design of policies to attenuate possible market failures arising from Veblen effects requires attention to considerations wholly absent above, including their effects on savings, distributional impacts and political viability (the public might not favour subsidising wilderness retreats for the well‐off, even if, as the leisure subsidy example requires, they were inconspicuous!) We will not address these issues here. It is clear, however, that policies designed to discourage consumption per se (such as the flat consumption tax advocated by many) are not optimally designed to address Veblen effects. The reason is that where Veblen effects are important, the social cost imposed by consumption depends on who is doing it, on the structure of reference groups (who cares about whom) and the size of the hierarchically ordered reference groups. The consumption of those who, like the well‐to‐do, are directly or indirectly reference models for many would ideally be taxed at a higher rate than the consumption of those who are models to none or to few.11 Such a policy would be doubly attractive as it as it would enhance the welfare of the less well‐off by limiting the downward cascade of welfare‐reducing Veblen effects while funding valued social projects or allowing the reduction of other incentive‐distorting forms of taxation. As (10) shows, the richer and smaller is the reference group, the higher is the tax rate that maximises total social utility. For well‐known reasons, policies that raise average living standards while favouring the less well‐off should be attractive to vote‐maximising political parties and candidates. Specific taxes on high‐end consumption items have occasionally been advocated and the village of Mamaroneck, New York, even placed a limit on house size specifically to curb Veblen effects (Foderano, 2001). But Veblen‐inspired policies are a rarity in both academic and policy circles.12 Trends in work hours have responded to other influences. Over a century ago Veblen thought that conspicuous consumption would increase in importance The means of communication and the mobility of the population now expose the individual to the observation of many persons who have no other means of judging his reputability than the display of goods... the present trend of the development is in the direction of heightening the utility of conspicuous consumption as compared with leisure. (1934, pp. 71–2) The description seems almost contemporary, and could well apply across nations, as large cosmopolitan elements in many populations now take their consumption standards from the well to do in New York, Milan or Tokyo rather than their domestic exemplars of style and respectability. But since he wrote, leisure has not been crowded out by consumption, conspicuous or otherwise. Indeed in the nations on which data are available, since 1870 work hours have declined substantially, by roughly 50% in continental Europe and by about a third in the English‐speaking nations (Huberman, 2004a,b). It seems plausible, and consistent with our estimates for a much shorter period, that among the causes included the sustained increase in per capita income, the increase in women's labour force participation, and the very long term decline in the income share of the top income earners, in many countries extending from the early part of the twentieth century until well into its final quarter.13 Footnotes 1 " See Bagwell and Bernheim (1996), Layard (1980), Frey and Stutzer (2002), van Praag (1993), Sen (1983), Hirsch (1976), Scitovsky (1976) and Easterlin (1974). Frank (1997), Cole et al. (1995) and Clark and Oswald (1996) provide extensive additional references to the empirical literature. By comparison to the economic literature, the relevant sociological and social psychological literature is extensive and venerable: Homans (1961) and Festinger (1957) are influential contributions. 2 " Corneo and Olivier (1997) analyse optimal taxation in Veblen‐inspired model of an indivisible conspicuous consumption good with both snobbish and conformist consumers. As in the model below, the tax implications of the Veblen effects they model depend on the number of consumers. 3 " See the works cited in Gintis et al. (2002). 4 " If the utility function is Cobb‐Douglas in leisure and effective consumption (with a the coefficient of c* and 1 − a the coefficient of (1 − h)) then the choice of hours is such that with the increased hours indicated by the second term on the right hand side representing the Veblen effect (if v = 0, h = a). 5 " We have (in reduced forms): g = g(h; z) and h = h(g; k) where z and k are exogenous influences. We would like to estimate the partial effect on h of an exogenous shift in g that is hg. What we observe, however, are intersections of these two functions, inferences from which (unless gh = 0) will over or underestimate the true effect. If g varies inversely with h– increases in work hours of those in the middle of the earnings distribution attenuating inequality – we underestimate the true effect, and conversely. 6 " Both estimates are smaller than the estimate in Table 1 and are only marginally significant, suggesting that inequality may explain much of the distinct nature of the two periods evident in Figure 1, while providing a weaker account of the within‐period movements. 7 " Among others, Boskin and Sheshinski (1978), Ireland (1994) and Oswald (1983) have made similar proposals. 8 " Were there m members of a third (poorer) class for whom the next highest income class is the reference group, with a wage rate wo and hours of work ho, a tedious calculation shows that A comparison with (7) shows that the optimal work hours of the rich are further reduced by the consideration of additional poorer classes. 9 " This could be accomplished, for example by subsidising the leisure activities of the rich. Under these conditions the rich would maximise and their private optimisation would give the first order condition (6) thus implementing (7). 10 " If there exists a third, poorer class, as defined in the previous footnote, and the intermediate class is taxed at the rate τm < 1, the optimal tax on the consumption of the rich increases to (10') to take account of the indirect Veblen effects (via increased work and consumption by the middle group) on the well‐being of the poorest group (the increase in τ* varying positively with the relative size of the poorer class and inversely with its wage). 11 " A government that sought to increase output (rather than maximising the sum of utilities) could mobilise Veblen effects by shifting the tax burden from the rich to the less well‐off, thereby inducing higher levels of work hours among the latter. 12 " This may be due to their seemingly punitive stance towards the well‐off. It is easy to see, however, that one could design a Pareto‐improving policy to attenuate negative consumption spillovers. Let the rich have a utility function that values some form of ‘inconspicuous consumption’ that does not stimulate emulation, in addition to both conspicuous consumption and leisure. Suppose that instead of a tax on the consumption of the rich, a restriction on their work hours was introduced. Because in the pre‐restriction allocation they selected their work hours to maximise their utility (choosing hr so that dur/dhr = 0), a sufficiently small reduction in their work hours would have only second‐order effects on their utility while conferring a cascade of first‐order benefits on those below them in the social comparison ranking. Thus there exists some Pareto improving hours restriction on the rich, accompanied by a transfer of income to the rich from the rest, with the proviso that it must be spent on inconspicuous consumption. 13 " See Dell (2003) on Germany and the data on France, UK and the US presented in Piketty and Saez (2003) References Bagwell , L. S. and Bernheim , D. ( 1996 ). ‘Veblen effects in a theory of conspicuous consumption’ , American Economic Review , vol. 86 , pp. 349 – 73 . OpenURL Placeholder Text WorldCat Bell , L. ( 1998 ). ‘Differences in work hours and hours preferences by race in the U.S.’ , Review of Social Economy , vol. 56 ( 4 ), pp. 481 – 501 . Google Scholar Crossref Search ADS WorldCat Bell , L. and Freeman , R. ( 2001 ). ‘The incentive for working hard: explaining hours worked differences in the US and Germany’ , Labour Economics , vol. 8 ( 2 ), pp. 181 – 202 . Google Scholar Crossref Search ADS WorldCat Böheim , R. and Taylor , M. ( 2004 ). ‘Actual and preferred working hours’ , British Journal of Industrial Relations , vol. 42 ( 1 ), pp. 149 – 66 . Google Scholar Crossref Search ADS WorldCat Boskin , M. and Sheshinski , E. ( 1978 ). ‘Optimal redistributive taxation and when individual welfare depends on relative income’ , Quarterly Journal of Economics , vol. 92 , pp. 589 – 601 . Google Scholar Crossref Search ADS WorldCat Bowles , S. ( 2004 ). Microeconomics: Behavior, Institutions, and Evolution , Princeton: Princeton University Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Clark , A.E. and Oswald , A. ( 1996 ). ‘Satisfaction and comparison income’ , Journal of Public Economics , vol. 61 ( 3 ), pp. 359 – 81 . Google Scholar Crossref Search ADS WorldCat Cole , H. , Mailath , G. and Postlewaite , A. ( 1995 ). ‘Incorporating concern for relative wealth into economic models’ , Federal Reserve Bank of Minneapolis Quarterly Review , vol. 19 ( 3 ), pp. 12 – 21 . OpenURL Placeholder Text WorldCat Corneo , G. and Olivier , J. ( 1997 ). ‘Conspicuous consumption, snobbism, and conformism’ , Journal of Public Economics , vol. 66 , pp. 55 – 71 . Google Scholar Crossref Search ADS WorldCat Dell , F. ( 2003 ). ‘Top incomes in Germany over the twentieth century: 1891–1995’ , CEPREMAP: Paris . Duesenberry , J. ( 1949 ). Income, Saving, and the Theory of Consumer Behavior , Cambridge, MA: Harvard University Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Easterlin , R. ( 1974 ). ‘Does economic growth improve the human lot? Some empirical evidence’, in Nations and Households in Economic Growth: Essays in Honor of Moses Abramovitz , New York: Academic Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Festinger , L. ( 1957 ). A Theory of Cognitive Dissonance , Stanford: Stanford University Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Foderano , L. ( 2001 ). ‘In suburbs, they're cracking down on the Joneses’ , New York Times , March 19. OpenURL Placeholder Text WorldCat Frank , R. ( 1997 ). ‘The frame of referance as a public good’ , Economic Journal , vol. 107 ( 445 ), pp. 1832 – 47 . Google Scholar Crossref Search ADS WorldCat Freeman , R. ( 2002 ). ‘The labour market in the new information economy’ , Oxford Review of Economic Policy , vol. 18 ( 3 ), pp. 288 – 305 . Google Scholar Crossref Search ADS WorldCat Frey , B. and Stutzer , A. ( 2002 ). Happiness & Economics: How the Economy and Institutions Affect Human‐Well‐Being , Princeton: Princeton University Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Galbraith , J. K. ( 1958 ). The Affluent Society , Boston: Houghton Mifflin Company . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Gintis , H. , Eric , S. and Bowles , S. ( 2002 ). ‘Costly signaling and cooperation’ , Journal of Theoretical Biology , vol. 213 ( 1 ), pp. 103 – 19 . Google Scholar Crossref Search ADS WorldCat Hirsch , F. ( 1976 ). Social Limits to Growth , Cambridge, MA: Harvard University Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Homans , G. ( 1961 ). Social Behavior: Its Elementary Forms , New York: Harcourt Brace . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Huberman , M. ( 2004a ). ‘Workdays, workweeks, and annual hours, new international estimates 1870–2000’ , mimeo , Universite de Montreal. Huberman , M. ( 2004b ). ‘Working hours of the world unite: new international evidence of worktime, 1870–1913’ , Journal of Economic History , vol. 64 , pp. 964 – 1001 . Google Scholar Crossref Search ADS WorldCat Inglehart , R. ( 1977 ). The Silent Revolution: Changing Values and Political Styles Among Western Publics , Princeton: Princeton University Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Ireland , N. ( 1994 ). ‘On limiting the market for status signals’ , Journal of Public Economics , vol. 53 ( 1 ), pp. 91 – 110 . Google Scholar Crossref Search ADS WorldCat Krueger , A. ( 1993 ). ‘How computers have changed the wage structure’ , Quarterly Journal of Economics , vol. 108 ( 1 ), pp. 33 – 60 . Google Scholar Crossref Search ADS WorldCat Landers , R. , Rebitzer , J. and Taylor , L. ( 1996 ). ‘Rat race redux: adverse selection in the determination of work hours in law firms’ , American Economic Review , vol. 86 ( 3 ), pp. 329 – 48 . OpenURL Placeholder Text WorldCat Layard , R. ( 1980 ). ‘Human satisfactions and public policy’ , Economic Journal , vol. 90 ( 360 ), pp. 737 – 50 . Google Scholar Crossref Search ADS WorldCat Leibenstein , H. ( 1950 ). ‘Bandwagon, snob, and Veblen effects in the theory of consumers’ demand’ , Quarterly Journal of Economics , vol. 64 ( 2 ), pp. 183 – 207 . Google Scholar Crossref Search ADS WorldCat Neumark , D and Postlewaite , A. ( 1998 ). ‘Relative income concerns and the rise in married women's employment’ , Journal of Public Economics , vol. 70 , pp. 157 – 83 . Google Scholar Crossref Search ADS WorldCat Oswald , A. ( 1983 ). ‘Altruism, jealousy, and the theory of optimal nonlinear income taxation’ , Journal of Public Economics , vol. 20 , pp. 77 – 87 . Google Scholar Crossref Search ADS WorldCat Park , Y. ( 2004 ). ‘Veblen effects on labour supply: male earnings inequality increases women's labour force participation’ , Department of Economics , Connecticut College. Piketty , T. and Saez , E. ( 2003 ). ‘Income inequality in the United States: 1913–1998’ , Quarterly Journal of Economics , vol. 118 , pp. 1 – 39 . Google Scholar Crossref Search ADS WorldCat Schor , J. ( 1998 ). The Overspent American: Upscaling, Downshifting, and the New Consumer , New York: Basic Books . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Scitovsky , T. ( 1976 ). The Joyless Economy: An Inquiry into Human Satisfaction and Consumer Dissatisfaction , New York: Oxford University Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Sen , A. ( 1983 ). ‘Poor, relatively speaking’ , Oxford Economic Papers , vol. 35 , pp. 153 – 67 . Google Scholar Crossref Search ADS WorldCat Spence , M. ( 1973 ). ‘Job market signaling’ , Quarterly Journal of Economics , vol. 90 , pp. 225 – 43 . OpenURL Placeholder Text WorldCat van Praag , B. ( 1993 ). ‘The relativity of the welfare concept’, in ( M. Nussbaum and A. Sen, eds.), The Quality of Life , pp. 363 – 92 , Oxford: Clarendon Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Veblen , T. ( 1934 ). The Theory of the Leisure Class , New York: Modern Library . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Zahavi , A. ( 1975 ). ‘Mate selection–a selection for handicap’ , Journal of Theoretical Biology , vol. 53 , pp. 205 – 14 . Google Scholar Crossref Search ADS PubMed WorldCat Author notes " Thanks to Michael Ash, Giacomo Corneo, James Rebitzer, Juliet Schor, Alois Stutzer and Elisabeth Wood and two anonymous referees for comments, Bridget Longridge for research assistance, and the MacArthur Foundation and the Behavioral Sciences Program of the Santa Fe Institute for financial support. © Royal Economic Society 2005.