Trade, Technology, and Transitions: Trampolines or Safety Nets for Displaced Workers?

Trade, Technology, and Transitions: Trampolines or Safety Nets for Displaced Workers? Abstract In the past several decades, the developed world has experienced significant labour market dislocations caused by international trade, technology, and other factors. While economic nationalism has risen in response to these challenges, technology is typically a more important factor than trade as a cause of these dislocations. Further, trade-related responses often impose additional costs on consumers through higher prices and on downstream industries that utilize inputs from protected sectors. Thus, the article argues that effective use of labour market adjustment policies (LMAPs) is a preferable approach to protectionist policies in addressing labour market adjustment costs. After laying out a spectrum of passive and active labour market policies, the article then goes on to provide a comparative evaluation of LMAPs in the USA, Canada, select Nordic and continental countries in Europe, and Australia. The article’s comparative evaluation suggests that the Nordic model, Germany, and Australia provide the most compelling utilization of LMAPs, while the USA lags behind other countries in our sample in relative resources devoted to LMAPs. However, recent trends in these jurisdictions suggest some degree of convergence on an ‘activation’ paradigm that utilizes incentive reinforcement and benefit conditionality in triggering participation in active labour market programmes. I. INTRODUCTION In many parts of the developed world, economic nationalism is on the rise, reflecting what The Economist magazine1 describes as a new political divide where political contests are not left versus right but open versus closed (drawbridges down or drawbridges up), implying a significant challenge to the broad free trade consensus that has prevailed in most of the developed world since the end of World War II. In this article, we provide a comparative evaluation of alternative labour market adjustment policies (LMAPs), arrayed along a spectrum from passive labour market policies (PLMPs) to active labour market policies (ALMPs), in response to labour market disruptions caused by international trade, technology, and other factors. We argue that such LMAPs are a preferable approach to the drawbridges up approach. To set the context, we set out some general propositions that we believe command a substantial consensus amongst trade policy scholars. First, the international trade law regime, especially the multilateral trade law regime beginning with the inception of the GATT in 1947 and superseded by the World Trade Organization in 1995, has always been sensitive to labour and related adjustment costs induced by trade liberalization. As the first author has argued elsewhere,2 the multilateral regime incorporated three important features designed to moderate adjustment costs: gradualism, reciprocity, and reversibility. With respect to gradualism, tariff and related restrictions on trade have been typically reduced gradually both within and across negotiating rounds. With respect to reciprocity, by structuring trade liberalization negotiations around the principle of reciprocity, export-oriented industries were often able to be enlisted as a political counterweight to import-impacted sectors and in addition provided the potential for moderating transition costs experienced by contracting import-impacted sectors through the reallocation of resources, including labour, over time to expanding export-oriented industries. With respect to reversibility, safeguard provisions built into the original GATT and maintained to the present-day permit importing countries facing an unexpectedly large import surge to reimpose tariffs or other import restrictions on a temporary basis to moderate the impact of these imports on domestic industries and their workforces, although unfortunately the Appellate Body of the WTO has interpreted the safeguard provisions so restrictively that invocation of these provisions has become highly problematic.3 Second, while it is indisputable that trade liberalization may cause job displacement in particular import-impacted sectors, the impact of trade liberalization on total employment in a developed country’s economy is much less clear.4 As Douglas Irwin argues, in a masterful review of both the theory and empirical evidence, ‘this is because the impact of trade on the total number of jobs in an economy is best approximated as zero’.5 Trade theory does not claim that trade liberalization will necessarily increase the total number of jobs in an economy, but rather that labour productivity will increase as countries focus on sectors in which they enjoy a comparative advantage. While there are net market losses in certain regions or sectors, the impact of trade has been found to be positive for overall employment and real wages.6 Third, import competition accounts for a relatively small percentage of total job displacements in most developed countries’ economies relative to technology, shifts in consumer preferences, new sources of domestic competition, etc. While the USA has experienced a decline of approximately six million manufacturing jobs since 2000, international trade, including the recent dramatic growth of imports from China, explains 20–25% of these job losses7 with the balance largely attributable to technologically induced increases in productivity in the manufacturing sector. Fourth, empirical evidence shows that protectionist responses to trade-induced labour market dislocations typically cost consumers in higher prices several multiples of the value of the jobs preserved in the protected industry. For example, in 2009 the Obama Administration imposed safeguard measures on car and truck tires imported from China resulting in an estimated saving of 1200 jobs in the US tire industry, amounting to $900,000 in consumer costs per job saved per year and resulting in the loss of an estimated 3731 jobs in other sectors of the economy.8 A high percentage of developed country imports are of intermediate goods that are inputs into the production of other goods and services, so that protectionist measures directed at these imports raise costs for downstream industries that use the imports as inputs and often result in much larger job losses in these downstream industries relative to jobs preserved in the protected sectors. However, it is commonly argued that some forms of contemporary international trade may constitute a form of ‘unfair’ trade or unfair competition9 and hence economic dislocations induced by unfair trade may be normatively unacceptable and justify trade restrictions or sanctions. First, trade deficits are often cited as evidence of unbalanced trade commitments. However, such concerns are often misplaced, with empirical evidence revealing that trade deficits often rise with a booming domestic economy where consumers have more resources to spend on imports. Moreover, as Irwin points out, if a foreign country is exporting more than it imports from another country, the overall balance of payments is divided between the current account and the financial account, which includes all portfolio and direct investments.10 Foreign countries can either use their export earnings to purchase imports or to invest the surplus in acquiring assets or investments in the country with which they are running a trade surplus (for example, China’s investment in US treasury bills and other securities). Thus, if the USA were to invoke protectionist measures to reduce the trade deficit, then capital inflows from abroad would necessarily have to fall and domestic investment would have to be financed by domestic savings, implying higher interest rates, which would reduce the number of jobs created by business investment, with the net effect on domestic employment unclear. Beyond the issue of trade deficits, unfair trade concerns often focus on claims that foreign exporters are deliberately manipulating their currency to induce an undervaluation and hence render their exports cheaper and imports more costly. It is difficult to evaluate as an analytical matter the ‘true’ value of a country’s currency, but in any event in the case of the principal target of such complaints in recent years, i.e., China, it now seems widely accepted that China has allowed its currency (the renminbi) to appreciate significantly in recent years and that it may no longer be significantly undervalued. A related complaint of unfair trade concerns foreign subsidization of exports, which renders them artificially competitive in importing countries’ markets, and leads to job displacement in these markets. This is again a difficult complaint to evaluate analytically, especially in the case of economies in transition from command to market economies (like China’s), with large numbers of state-owned enterprises and extensive use of regulated or directed pricing,11 but as with currency manipulation, it is a complaint that cannot be dismissed out of hand and is subject to detailed disciplines in the WTO’s Subsidies and Countervailing Measures Agreement. However, there are legitimate concerns over the appropriateness and enforceability of a number of these disciplines, nor do they address at all an increasingly salient issue of tax and subsidy competition among host jurisdictions to attract foreign direct investment. A further complaint is that foreign exporters often ‘dump’ goods in the markets of importing countries at lower prices than prevail in their home markets. While anti-dumping actions are often initiated by importing countries on this or related bases, beyond the very narrow category of cases involving predatory pricing by foreign exporters, most dumping cases lack any coherent economic rationale and simply reflect geographic price discrimination.12 Nevertheless, anti-dumping actions have proliferated around the world in recent years and have often become the protectionist instrument of choice, despite their lack of any convincing economic rationale in most cases, and dominate by orders of magnitude the two other principal trade remedy regimes: safeguards and countervailing duties which have more plausible economic rationales. A further claim of unfair trade relates to foreign countries improperly appropriating the intellectual property of firms based in other countries through either lax laws or lax enforcement thereof, hence conferring on them an artificial comparative advantage in international trade and unfairly prejudicing these firms and their workforces. While the Trade-Related Intellectual Property Agreement (TRIPs) negotiated during the Uruguay Round seeks to address many of these concerns, there may still be legitimate concerns relating to the lax enforcement of these commitments. Moreover, TRIPS does not apply to the controversial issue of technology-sharing agreements imposed on foreign investors as a condition of investment in host countries. A yet further set of ‘unfair’ trade complaints relate to lax labour or environmental laws in foreign countries that enable them to reduce costs unfairly, rendering their products artificially competitive in international trade. With respect to concerns over lax labour standards, while it may well be the case that insisting that all trading countries adhere to core labour standards such as those promulgated by the International Labour Organization (ILO) in its 1998 Declaration on Fundamental Principles and Rights at Work, i.e., freedom of association; the elimination of forced labour; the elimination of child labour; and the elimination of discrimination in employment (and perhaps also basic workplace safety standards), is normatively defensible on non-economic rationales, it is important to recognize (as the ILO does) that lower wage levels in some foreign countries (especially developing countries) are not a source of unfair trade but indeed a key source of their comparative advantage (due account being taken of differences in labour productivity across countries).13 Similarly, lax environmental standards in some foreign countries may be a legitimate concern for other countries where these involve transboundary externalities or threats to the global commons (such as climate change) and justify trade sanctions,14 but where their effects are purely local it is not nearly as obvious that other countries have any legitimate basis for insisting on adherence to their own environmental standards as a precondition to international trade. Yet further unfair trade concerns focus on the challenges facing small, less developed countries in competing in an open international trading system as a result of weakly developed industrial and financial sectors and the continuing predominance of traditional (albeit arguably inefficient) forms of agricultural production, thus arguably justifying dispensations from normal trade disciplines to advance more dynamic conceptions of latent comparative advantage and to moderate transition costs, given limited fiscal capacity to compensate for these directly.15 Where do these propositions leave us in terms of addressing legitimate concerns on, both equity and efficiency grounds, over the labour market dislocations induced by international trade, assuming that unfair trade concerns are not legitimately engaged? First, it is often impossible to disentangle job displacement (or indeed rising levels of inequality and stagnant middle-class incomes) caused by trade from job displacement (and rising inequality and stagnant incomes) caused by technology and other factors, so that trade-related (i.e. protectionist) policy responses are likely in most cases to be misdirected. Second, even if it were practically possible and ethically defensible to separate out trade-induced labour market dislocations from dislocations caused by other factors, trade-related responses are an enormously costly LMAP, both in terms of additional costs to domestic consumers from higher prices and hence reduced spending in other sectors of the economy, and in terms of higher costs and hence often greater job displacement in downstream industries that utilize the inputs of protected sectors. Hence, we argue in this article that most developed countries, in responding to labour market dislocations, should place a higher priority on domestic LMAPs, the effectiveness of which we seek to evaluate from comparative experience in the balance of this article in addressing the different implications of cyclical, frictional, and structural unemployment. While as a matter of political economy well-designed LMAPs may not fully mute demands for protectionism by trade-impacted sectors, they provide a principled basis for other constituencies to resist these demands by demonstrating a commitment to taking seriously the losers from economic transitions more generally. II. A MENU OF LMAPS There are various rationales for government intervention in labour markets, including bargaining failure, information failure, externalities, and equity principles.16 A laissez-faire approach to training would assume that employers are willing to pay for training that is firm-specific, and that employees are willing to pay for the acquisition of general human capital.17 However, in practice, employers and employees are often unable to make optimal decisions about investments in training and it would be unwarranted to assume that appropriate LMAPs should be left entirely to private parties. The menu of LMAPs available to government entails a spectrum of passive to active LMAPs, each with inherent strengths and limitations, depending on the causes of unemployment. There have been growing concerns in many developed countries in recent years over stagnant lower and middle incomes, rising inequality, and the dramatic increase in non-standard, precarious employment,18 which implicate a wide range of public policies for reducing inequalities in market and post-market incomes.19 However, in this article we focus more narrowly on job displacement from standard and non-standard employment, and how displaced workers can most effectively be supported using both passive and active LMAPs. We survey this menu of policies briefly below, before turning to an evaluation in the next section of comparative experience in designing the optimal combination of policies on this menu. A. PLMPs PLMPs can be described as ‘safety nets’ that helps minimize losses to workers through income support, using policy options such as unemployment insurance (UI), government subsidies to struggling businesses or regions, and wage-fixing policies. Because these benefits were originally developed to provide temporary assistance to the cyclically unemployed, PLMPs are ineffective in addressing structural unemployment (that is, they target only the symptom of unemployment rather than the cause).20 In addition, unemployment benefits tend to induce disincentive effects that increase the incidence and duration of unemployment. Recognizing this, the trend in OECD countries has been to shift resources away from PLMPs to more active policies.21 Still, two other passive policies have attracted significant recent attention. The first is wage insurance, which provides a temporary wage supplement to displaced workers who are compelled to move to lower-paying jobs. The second is universal basic income (UBI), a fixed and unconditional cash payment with clawbacks based on earned income, given to all citizens in a given economy. UBI has gained traction across the ideological spectrum in recent years as a potential solution to the problem of automation and technology displacing workers. While proponents argue that this would allow people to invest time and money in education or training, critics argue that people would be less motivated to work.22 UBI is likely to be too coarse-grained as a stand-alone response to address the widely varying circumstances of individuals who are unemployed, but new empirical evidence is likely to be soon forthcoming, with basic income pilots currently in place in countries such as Finland, Canada, India, and Namibia. B. ALMPs In contrast to a ‘safety net’, ALMPs act as a ‘trampoline’ to help those displaced from the labour market find or acquire new opportunities for re-entering it.23 These policies target the unemployed by reference to their income earning potential, and have been promoted by international agencies and adopted across the OECD over the past three decades.24 While evaluations of ALMPs have been mixed, studies clearly show that policies that are well designed and well targeted have a positive impact on participants’ employability and earnings.25 Over the past decade, ALMPs have become integrated with ‘activation’ strategies that recognize important interactions between the generosity of passive benefits, the size and mix of ALMP spending, and incentive measures such as attaching benefit eligibility to certain conditions.26 In this vein, some analyses have conceptualized ALMPs as dichotomous: ‘offensive’ workfare, which emphasizes improving skills and human capital investment, and ‘defensive’ workfare, which operates through work incentives, benefit conditionality, and the use of sanctions.27 The array of ALMPs can be grouped more finely into four broad categories.28 These categories, exhibiting defensive to offensive characteristics in this order, are incentive reinforcement, demand-side policies, employment assistance, and upskilling. 1. Incentive reinforcement This category refers to measures that strengthen the motivation for benefit recipients to take up existing job opportunities or to participate in programmes that will strengthen the recipient’s earnings capacity. This can be achieved by making passive benefits conditional upon behavioural requirements such as participation in programmes that strengthen individuals’ productivity or work capacity, through in-work benefits, or through the use of benefit sanctions (in the case of non-compliance).29 This has perhaps been best represented by the United States’ welfare-to-work reform in the 1990s. 2. Demand-side policies While ALMPs have tended to focus on supply-side factors (such as employment assistance and upskilling), demand-side policies (i.e. job creation and work experience programmes) are receiving increasing attention as they help to ensure that there are employment opportunities available to begin with. Other demand-side policies include programmes to finance business start-ups, and employment incentives such as targeted wage or hiring subsidies. Studies have found that employment incentives and direct job creation have positive impacts for disadvantaged groups if well targeted.30 However, public employment programmes have tended to result in temporary ‘make-work’ activities that have little effect on the future employability of participants.31 3. Employment assistance Employment assistance facilitates re-entry into labour-market participation, and includes counseling, placement services, and job-search programmes. Job search assistance is the main policy in this category, which aims to connect employers with the unemployed more efficiently. However, because these programmes help individuals find jobs for which they are already qualified, this policy option is appropriate only for frictional unemployment and not for structural unemployment. Job relocation assistance may be more effective. Policymakers have sometimes neglected barriers to geographic mobility that may prevent a qualified worker from becoming re-employed. Mobility rates have been especially problematic in the USA, where workers in depressed local markets tend to either drop out of the labour force or remain unemployed instead of relocating.32 Job relocation assistance such as relocation subsidy programmes would help alleviate these problems by providing individuals with incentives to overcome constraints on mobility. Another method of assistance is trade adjustment assistance (TAA), which has figured prominently in the USA, with mixed results. TAA programmes generally involve a mix of PLMPs and ALMPs by providing income support, skills training, and job-search assistance to workers who have lost jobs as a result of international trade. 4. Upskilling Upskilling uses education and training mechanisms to offer a ‘second chance’ to people whose skills have become obsolete, and is ideally designed to target structural unemployment. While there is a large literature devoted to the evaluation of classroom and on-the-job training programmes, a more recent concept of lifelong learning is emerging, reflecting the adaptability that workers require today to address the challenge of increasingly rapid labour market and technological changes. The private sector has already begun responding to this need. Companies such as Coursera, Udacity, and LinkedIn offer massive open online courses (MOOCs) that embrace an employment-focused business model. However, MOOCs are currently geared towards those with university degrees, and are not easily accessible to low-skilled workers who may lack time, money, and motivation.33 Unless government intervenes to level the playing field, lifelong learning may do little to reduce inequality. III. COMPARATIVE EXPERIENCE A comparative evaluation of LMAPs across developed countries reflects the fact that different jurisdictions have traditionally favoured different mixes of policy options. These differences have been the focus of an important line of research and have at times been generalized into three models.34 The ‘Anglo-Saxon’ model exhibits low employment protection, low UI, and the use of ‘defensive’ workfare (such as work incentives and benefit conditionality), which in theory leads to high job flow, short unemployment duration, and low unemployment. The ‘Nordic’ model—or ‘flexicurity’ model—is based on a medium to high degree of employment protection, generous but conditional UI, and strong ALMPs, thus allowing for effective re-employment and low unemployment. The Nordic model emphasizes near universal benefits and operates within broader welfare provisions that include universal health insurance, public education, maternity leave, and childcare. As a result, while both the Anglo-Saxon and Nordic models have historically been relatively successful, the Anglo-Saxon model is also associated with significantly more inequality than the Nordic model. The third model, the ‘continental’ model, is based on high job protection, generous UI, and limited ALMPs, and has been widely judged to be an ineffective method of reallocation of labour and of combating unemployment, as we discuss further below. While these groupings are to some extent stylized (for example, Germany has moved away from a continental model to a more Nordic model), these models are helpful for comparing and assessing different policy approaches and also draw on Esping-Andersen’s analysis of the ‘liberal’ welfare state, the ‘social-democratic’ welfare state, and the ‘corporatist’ welfare state.35 For example, the liberal welfare state, typically exemplified by the USA and Canada (although we argue that Canada is a middle ground between a US-styled liberal regime and a more interventionist European regime), is market-oriented and exhibits modest benefits and strict entitlement rules catered to low-income state dependents, which is consistent with the Anglo-Saxon model. The social-democratic welfare state exhibits de-commidifying and universalistic programmes that promote an equality of high standards for all strata, consistent with the Nordic model (exhibited by the Nordic countries). Finally, the corporatist welfare state displaces the market as a provider of welfare and aims to preserve occupational groups and status differentials. This can be seen in the continental model and especially in France, where historically rigid job protection laws have created an insider–outside problem (which will be later discussed). The chart below summarizes the models and their characteristics, and places the countries that will be surveyed along a spectrum according to their fit within each model. The following section will compare on-the-ground LMAP experiences in the USA, Canada, and selected countries in Europe, with a brief additional reference to recent Australian experience. Model Characteristics Esping-Andersen’s welfare state regime type Countries surveyed in this article Anglo-Saxon Low employment protection Low UI Emphasis on ‘defensive’ workfare, i.e. work incentives, benefit conditionality, and the use of sanctions Liberal USA Canada Denmark Sweden Finland Netherlands Germany France Nordic (‘flexicurity’) Flexibility in the labour market (i.e. relaxed hiring and firing rules) combined with high income and employment security (i.e. generous UI and strong ALMPs) Emphasis on ‘offensive’ workfare, i.e. human capital investment and upskilling, with a movement towards ‘defensive’ workfare Near universal unemployment benefits Social-democratic Continental High employment protection Generous UI Limited ALMPs Corporatist Model Characteristics Esping-Andersen’s welfare state regime type Countries surveyed in this article Anglo-Saxon Low employment protection Low UI Emphasis on ‘defensive’ workfare, i.e. work incentives, benefit conditionality, and the use of sanctions Liberal USA Canada Denmark Sweden Finland Netherlands Germany France Nordic (‘flexicurity’) Flexibility in the labour market (i.e. relaxed hiring and firing rules) combined with high income and employment security (i.e. generous UI and strong ALMPs) Emphasis on ‘offensive’ workfare, i.e. human capital investment and upskilling, with a movement towards ‘defensive’ workfare Near universal unemployment benefits Social-democratic Continental High employment protection Generous UI Limited ALMPs Corporatist Model Characteristics Esping-Andersen’s welfare state regime type Countries surveyed in this article Anglo-Saxon Low employment protection Low UI Emphasis on ‘defensive’ workfare, i.e. work incentives, benefit conditionality, and the use of sanctions Liberal USA Canada Denmark Sweden Finland Netherlands Germany France Nordic (‘flexicurity’) Flexibility in the labour market (i.e. relaxed hiring and firing rules) combined with high income and employment security (i.e. generous UI and strong ALMPs) Emphasis on ‘offensive’ workfare, i.e. human capital investment and upskilling, with a movement towards ‘defensive’ workfare Near universal unemployment benefits Social-democratic Continental High employment protection Generous UI Limited ALMPs Corporatist Model Characteristics Esping-Andersen’s welfare state regime type Countries surveyed in this article Anglo-Saxon Low employment protection Low UI Emphasis on ‘defensive’ workfare, i.e. work incentives, benefit conditionality, and the use of sanctions Liberal USA Canada Denmark Sweden Finland Netherlands Germany France Nordic (‘flexicurity’) Flexibility in the labour market (i.e. relaxed hiring and firing rules) combined with high income and employment security (i.e. generous UI and strong ALMPs) Emphasis on ‘offensive’ workfare, i.e. human capital investment and upskilling, with a movement towards ‘defensive’ workfare Near universal unemployment benefits Social-democratic Continental High employment protection Generous UI Limited ALMPs Corporatist A. United States The US labour market has seen significant recovery from the Great Recession—unemployment rates fell from 9.9% in November 2009 to 4.1% in November 2017 (more than one percentage point lower than the post-war unemployment average),36 indicating that the US economy is near full employment. However, this figure does not provide much relief for the many middle-class American workers facing rising inequality, real wage stagnation, a dramatic rise in non-standard employment, and limited job opportunities for those without a college degree.37 To illustrate, the wages of the top 1% of income earners have grown 138% since 1979 while wages for the bottom 90% grew a mere 15%. In fact, middle-wage workers’ hourly wage has increased only 6% between 1979 and 2013, and low-wage workers’ hourly wage declined 5% during that same period.38 Since the end of the 1990s, long-term unemployment has been rising and labour force participation has been declining, likely due to the perception that jobs are difficult to find.39 While it is understandable why an economic nationalist agenda designed to protect and restore the jobs of American workers resonated with many voters in the 2016 US presidential elections, protectionist measures are mostly an ineffective response (as we argue in our introductory section). The central problem is that the country has failed to invest in effective social programmes and LMAPs to cushion the working class through the adjustment process. Current US labour market supports are both inadequate and ineffective, with several dozen uncoordinated federal job-retraining programmes in addition to numerous state-level programmes (although corporate training programmes are increasing).40 In addition, the US exhibits several distinct features that pose special policy challenges. First, the USA is a traditional user of PLMPs and is characterized by low levels of investment in ALMPs. Second, the TAA programme has taken on a central role in American policy but has largely failed to produce positive results. Third, the country faces significant policy complications induced by federalism and regulatory barriers, leading to inconsistent treatment of workers across the nation and a declining interstate mobility rate. 1. Low levels of investment into ALMPs The USA has historically been classified as a ‘liberal’ welfare state that provides few forms of social protection, marginal income benefits, and a narrower range of social services compared with most other developed countries.41 The average public expenditure for OECD countries in 2015 (as a % of GDP) was 0.78 for PLMPs and 0.53 for ALMPs, while the USA spent substantially less on both at 0.18 for PLMPs and 0.10 for ALMPs.42 In fact, the USA has viewed ALMPs as having a minimal role today in the labour market adjustment process—of all OECD countries, the USA was one of the countries with the lowest public expenditure on ALMPs according to the data from 2012.43 As a nation that relies more heavily on passive policies, UI is the primary safety net for unemployed workers. UI is compulsory and mandated by federal legislation, but the proportion of the unemployed who receive UI benefits has declined since the 1980s, partly due to much tighter eligibility rules such as decreased income cutoffs (by counting social security income and pension benefits) and stricter enforcement of work search requirements.44 While these eligibility requirements may help combat disincentive effects of UI benefits, the recipiency rate, or the percentage of insured unemployed out of total unemployed, was on average only 27% in 2015 and 2016 (compared with 43% in 1980).45 This problem will continue to grow in today’s emerging world of part-time, contract, and gig workers who are excluded from UI benefits. The duration and level of UI benefits have also declined during the same period—today, the maximum duration of benefits is 26 weeks and the average income replacement rate was around 40% in 2016,46 down from 39 weeks in the 1970s and a 50% income replacement rate between the 1970s to 1990s. Still, UI has played an important role in the American economy, and was estimated to have kept five million people out of poverty during the Great Recession when the duration of benefits was temporarily extended.47 2. Job training programmes: the federal workforce system and the central role of the TAA There is a growing disconnect between two narratives surrounding the American blue-collar work force: on the one hand, workers claim that jobs are going overseas or becoming automated, and on the other hand, employers have been insisting that they have been unable to fill skilled blue-collar jobs. According to a 2015 report by the Manufacturing Institute, seven out of ten manufacturing executives faced a shortage of workers with adequate technological skills.48 Job training seems to be the obvious answer to bridge the gap between these narratives,49 but federal government retraining programmes have failed to provide an integrated national scheme. Instead, there is a fragmented mix of federal and state-level schemes that have been called a ‘dizzying constellation’ of programmes that has become ‘increasingly byzantine’.50 Today, the US’s primary federal employment assistance and training programme is the Workforce Innovation and Opportunity Act (WIOA), which supersedes a complex body of federal legislation. Another key piece of legislation is the Trade Adjustment Assistance Reauthorization Act of 2015, which updates the TAA programme. In the following section, a brief history of the WIOA’s precursors is discussed to provide some perspective on America’s changing employment policy, followed by an evaluation of the TAA, the latter of which is unique among developed countries for the central political role that it has played. (i) Evolution of federal employment training programmes The first federally funded training and employment assistance programme in the USA was the Works Progress Administration (WPA), created in 1935 as part of the New Deal to alleviate the unemployment crisis during the Great Depression. During this period, public projects were the focus and training occurred only if necessary in small and fragmented programmes. In the 1970s, the government enacted the Comprehensive Employment and Training Program (CETA) that placed these programmes under the control of local governments, thus marking a trend towards decentralizing employment training. In 1982, CETA was succeeded by the Job Training Partnership Act (JTPA) and by 1994, there were 154 federal programmes run by 14 agencies, many of which were redundant.51 The Workforce Investment Act (WIA) of 1998 was a major structural overhaul that replaced the JTPA in an attempt to consolidate employment-training programmes into a single funding stream and further decentralized training to the states.52 However, the system remained fragmented and difficult to navigate, with different entry points and eligibility rules that made participation and choice difficult.53 By 2009, there were 47 programmes administered by nine federal agencies, a substantial decrease from 1994. However, almost all programmes overlapped with at least one other programme, and were criticized as a ‘maze of confusing training programmes’ by former President Obama.54 Perhaps of more concern, a 2011 report by the Government Accountability Office (GAO) found that ‘little is known about the effectiveness of most programmes’ as only 5 of 47 programmes had conducted impact evaluations. These five impact studies found that benefits to unemployed workers were small, inconclusive, or non-existent.55 In response to some of these concerns, the WIOA was enacted in 2014 by bipartisan majorities in Congress, and represented a significant reform of the federal workforce system. Recognizing that there was a limited evidence base for ALMPs, the WIOA seeks to improve accountability and transparency by requiring measurement and public reporting of a programme’s performance (such as employment and earning outcomes) by education and training providers. Notably, it mandates that programmes be evaluated at least every four years by independent third parties.56 The WIOA also requires states to submit a single unified strategic plan to better align workforce development programmes and help ensure that job seekers have seamless access to career services, education, and training. Other reforms include strengthened funding for work-based learning that focuses on youth programme services and apprenticeships, and the use of pay-for-performance contracts to incentivize higher quality training.57 With Final Rules that became effective July 2016, the WIOA promises some effective changes that will be evaluated on a regular basis. (ii) Central role of the TAA Although TAA programmes exist in other jurisdictions (the EU and South Korea introduced TAA programmes in 2007), the USA is unique for the central political role that its TAA has played since the 1960s. For example, in the fiscal year 2016, Congress appropriated $861m for the TAA, allocating $450m for training and reemployment services and $411m for income support.58 The EU’s programme has an annual budget of only $170m, while South Korea’s small programme provides support to firms rather than to workers.59 Under the current TAA, which was reauthorized by the TAA Reauthorization Act in 2015, displaced workers who can establish that foreign trade contributed importantly to their job loss can access training and reemployment services, weekly income support, wage insurance, and a health coverage tax credit. While the TAA has seen both expansions and budget cuts throughout its life, it has remained an integral part of the US’s trade policy for over half a century. The theory underlying the TAA programme is that while trade liberalization increases overall economic welfare, domestic costs fall disproportionately on certain sections of the workforce who suffer dislocation due to foreign competition or off-shoring. The perceived appropriate response is to compensate the losers (who are often older, have less education, and hold shorter job tenure60) with more robust benefits and services than would be available to a worker displaced for other reasons. The TAA was first created by the Kennedy Administration’s Trade Expansion Act of 1962, and has since historically been reauthorized alongside trade liberalization policies. While the programme offered income benefits, vocational education, and moving expense payments, the eligibility criteria were very difficult to meet until they were further expanded by the Trade Act of 1974. Still, between 1975 and 1981, only 4% of TAA-eligible workers had actually entered training programmes.61 In 2002, the Bush Administration expanded eligibility criteria and benefits once again, offering a wage insurance programme for the first time as well as a health coverage tax credit. The current TAA offers TAA-certified workers 50 years and older who are reemployed at a lower wage a cash payment equal to a 50% of the difference between the worker’s trade-affected job’s wage and the new wage, to a maximum of $10,000 over two years.62 However, workers receiving wage insurance cannot also receive income support from the TAA, which is provided to participants enrolled in training within 26 weeks of job separation or TAA certification. Training is paid for directly to the service provider or through a voucher system that allows participants to choose from a range of approved programmes, and participants can have 90% of their job search and relocation expenses covered, up to $1250 respectively.63 The TAA has been difficult to evaluate due to the lack of data—prior to the new WIOA performance-reporting requirement, no ongoing assessments were required. Recent evaluations have produced mixed to negative results. A 2012 report found that the net cost to participants was $26,837 due to lower wages than in the matched comparison group,64 and while the wage loss is likely due to participants engaging in training programmes rather than working, the training programmes appear to have flaws as well. A 2012 evaluation by the US Department of Labor (DOL) found that 85% of those who undertook TAA-funded training received a certificate or degree, but only 37% of them were working in that field four years later. In addition, those who received training waited an average of over seven months from the time they filed for UI before obtaining training placement.65 The result is that the TAA programme has become ‘token help for workers hurt by import competition’,66 and help for Americans has come largely from other sources. Autor, Dorn, and Hanson estimated that between 1990 and 2007, only 6% of additional government payments associated with Chinese import competition came from TAA or UI. The bulk of these payments came from other sources, with 32% coming from disability or retirement insurance, 26% from federal-government income assistance, and 32% from extra medical spending.67 An evaluation prepared for the DOL found that without considering the benefits stemming from the possibility that the TAA reduces resistance to free trade, the benefits of the programme as it operated under the 2002 amendments were negative.68 Commentators have also questioned the underlying rationale of giving special assistance only to trade-affected workers. Empirically, displaced workers in trade-sensitive industries are often similar to workers displaced in other industries, and there is no compelling economic or ethical basis for why these workers should be treated differently. Further, while the TAA may have helped reduce opposition to trade liberalization, there has been no evidence since the 1970s that it has been able to ‘buy off’ labour groups opposed to trade.69 3. Complications produced by federalism Another distinctive feature of the USA is its federal system and regional diversity, which produce complications for incentivizing job training, for the administration of federal programmes, and for internal labour market mobility. First, states often have weak incentives to provide training to rectify structural unemployment since they do not want to invest resources to train workers to move out of their own jurisdiction and reduce the local tax base. Second, while UI, the TAA, and the WIOA are federal programmes that remain federally funded, they are all administered at the state and local levels, resulting in striking disparities and differing treatment of workers across the country. For example, while federal legislation determines what forms of employment are covered by UI, each state establishes its own rules governing scale of benefits, eligibility, and duration of benefits. The recipiency and income replacement rates vary dramatically—recipiency rates in 2016 ranged from a low of 10% in Florida to a high of 51% in Massachusetts, and income replacement rates ranged from a low of 32% in Alaska to a high of 56% in Hawaii.70 Similarly, the TAA’s income support programme also reflects this kind of variation. In 2015, the highest maximum weekly benefit for a worker with no dependents was $698 in Massachusetts, while the lowest weekly benefit was $240 in Arizona.71 Finally, the WIOA is based on the assumption that authority should be decentralized so that states and local agencies can adapt employment and training programmes to their own economic and political circumstances. As a result, one-stop centres are operated locally, with the result that there are significant differences in how job seekers access and receive its programmes.72 Third, states and local governments have created many legal barriers that make interstate mobility difficult, such as land use laws that restrict housing construction, occupational licensing rules, conditions on public pensions, and variations in the standards for receiving public benefits.73 Currently, there are 1100 different jobs subject to licensing requirements in at least one state (ranging from doctors to manicurists, upholsterers, and funeral attendants)74 and while licensing is typically justified on grounds of consumer protection, it raises the incomes of the highest paid earners by up to 24% (compared with 3.6–5.1% for the lowest paid earners) while preventing lower-skilled workers from providing even basic services.75 More importantly, licensing requirements are often arbitrary, as evidenced by the dramatic inconsistency in licensing requirements across states. Excessive occupational licensing is a form of protectionism that slows economic growth and interferes with entrepreneurship and innovation, and was estimated to have created a deadweight loss accounting for a $40 billion loss of output.76 Better solutions include the use of inspections rather than mandatory licenses, or the use of optional qualifications and titles to allow consumers to decide for themselves whether these are valuable signals of quality, which is especially effective in today’s information sharing world.77 Other barriers to entry such as price inflation and land use restriction in opportunity-rich markets also create a sticky internal labour market, making it difficult to correct for regional labour market issues. Economists have estimated lost labour market gains (in the form of higher wages that could be earned), and found that the US GDP could be 13.5% larger if localities eased land-use restraints alone.78 Current workplace benefits (i.e. healthcare and retirement plans) are tied to jobs rather than to employees themselves, making it more difficult for workers to leave their jobs to seek better work opportunities. One in every six Americans currently live in an economically distressed community where job prospects continue to shrink79 and yet the percentage of Americans who move to a different state each year has fallen by half since the 1990s.80 This is problematic as it discourages agglomeration economies and prevents job seekers from securing available employment opportunities in other regions, which is especially significant in the face of growing regional inequalities. In addition, declining mobility rates contribute to greater inequality—mobility rates are especially low among the disadvantaged and the non-college educated, and specifically from distressed places to high-wage (but often higher housing cost) regions.81 Because states lack the incentive to address broader national labour market needs, policy change at the federal level is needed to encourage mobility without displacing the useful aspects of local control in a geographically large and economically complex country. While the TAA provides a relocation allowance, it is capped at $1250 and fewer than 1% of programme participants use it.82 A new federal policy could provide larger incentives to relocate, such as providing individuals beyond the TAA with more generous relocation subsidies, or providing tax incentives to states and cities that reduce entry constraints such as restrictive housing and occupational licensing policies. A portable system of employment benefits should also be created that ties benefits to individuals rather than to jobs, and should be expanded to include part-time and contingent workers to encourage a more flexible labour market.83 B. Canada The USA and Canada share many economic and social similarities. Historically one another’s largest trading partners, both countries have favoured a safety-net approach of LMAPs and employ a federal system to govern large and regionally diverse countries. However, they also maintain important differences in their governance, economies, financial, and labour markets. Canada’s economy was less affected by the 2008–9 financial crisis than most other OECD countries, and because Canada’s GDP is concentrated in natural resource-based sectors, the labour market performed well due to a boom in natural resource exports. On the other hand, Canada’s unemployment rate, which was 5.9% in November 2017, remains higher than that of the USA and the more recent fall in global commodity prices has led to lower rates of economic growth.84 In contrast to the USA, Canada has an explicit constitutional protection of personal mobility rights including the right to pursue the gaining of a livelihood in any province.85 Therefore, laws that discriminate against out-of-province residents may be classified as a breach of their mobility rights.86 However, Canadian courts have held that the right to pursue the gaining of livelihood in any province does not create a right to work, nor does it include the right for workers to establish themselves professionally anywhere in the country regardless of qualifications.87 Internal labour mobility rates are around the same as in the USA, if not slightly lower. In 2011, the interprovincial migration rate was 0.7% in Canada and 1.6% in the USA, while the intra-provincial migration rate was 3.4% in Canada compared with 2% in the USA.88 Significant disparities in regional unemployment rates continue to persist, suggesting that barriers to mobility remain. The Canadian government has taken steps to reduce these barriers including promoting interprovincial agreements that enable workers certified in one province/territory to be certified in another one.89 Structural changes are likely to continue in Canada with a decline in manufacturing jobs and a move towards service-producing industries and high-skilled jobs.90 Despite these structural changes, Canada has taken a different stance from the USA, embracing a continuing commitment to trade liberalization and liberal immigration policies rather than protectionism. In addition, while there has been political tension in federal–provincial relations over the administration of labour market training, the development of LMAPs has not been as politically charged as in the USA. Three distinctive features of Canadian policy will be examined. First, Canada is often regarded as ‘in between’ the less interventionist US-style regime and more interventionist European-style regimes,91 and has introduced more extensive active policy measures in recent years. Second, Canada has managed to pursue a period of intensified trade liberalization in the absence of a TAA programme. Third, the administration of ALMPs is decentralized under the federal system to the provinces and territories, which has resulted in flexible, regionally sensitive policies but a lack of a unified national policy framework. 1. The ‘middle ground’ As recalled from the previous section, the average public expenditure for OECD countries in 2015 (as a % of GDP) was 0.78 for PLMPs and 0.53 for ALMPs. Canada’s expenditure is relatively low by international standards, at 0.62 for PLMPs and 0.24 for ALMPs, but remains significantly higher than expenditures in the USA (0.18 and 0.10 respectively).92 Canada has a strong record of LMAPs and announced increased funding for LMAP interventions in its recent federal budgets. Previous LMAP experiments include TAA and wage insurance programmes, and in 2007 the federal government launched the Targeted Initiative for Older Workers (TIOW) that provides unemployed, older workers (ages 55–64 years) with group-based skills training.93 In 2017, the Ontario government launched a UBI pilot that provides almost $17,000 per year for a single person (in the form of cash payment with clawbacks based on earned income)94 and, recognizing the rise of non-standard employment, enacted equal pay for equal work provisions to ensure that part-time employees are paid proportionately the same as full-time employees if they are doing the same job.95 The backbone of Canada’s LMAP system and its principal safety net and workforce development programme is the Employment Insurance (EI) programme.96 The EI programme provides financial benefits as well as ALMPS known as Employment Benefits and Support Measures (EBSMs) that are funded through a payroll tax. While EI financial benefits are not as generous as those in Nordic countries, the coverage rates, recipiency rates, and benefit levels are higher than those in the USA.97 Regular financial benefits are available to displaced workers who have lost their jobs through no fault of their own, have been without work or pay for at least seven days, have accumulated between 420 and 700 hours of insurable employment in the 52 weeks prior to the claim, and are actively seeking employment during the entire claim period. Critics argue that the EI system has become weaker—tightened eligibility requirements and changes in the labour market such as increasing non-standard employment (i.e. self-employment and short-term jobs), have meant that the proportion of the unemployed eligible to receive EI benefits dropped from 83% in 1989 to 42% in 1997.98 Rates have not recovered since—in 2015–16, 43.8% of unemployed workers were eligible to receive regular EI benefits, with an average weekly benefit amount of $443.99 EBSMs are the second component of the EI programme and comprise four categories: targeted wage subsidies, skills development programmes, self-employment assistance for those who wish to start their own business, and job creation partnerships that provide short-term employment. While these programmes are federally funded through the EI fund, they are administered by the provinces and territories through federal-provincial Labour Market Development Agreements (LMDAs). A 2015–16 assessment found that 178,556 EI clients returned to employment following an intervention and calculated that the returns to employment were $1.15b (in the amount of unpaid EI financial benefits).100 However, while support measures such as counseling and resume writing workshops are available to all Canadians, employment benefits under the aforementioned categories are available only for EI-insured clients, meaning that only about 40% of the unemployed can access these services, and even fewer use them. Out of the 1.37m unemployed in 2012, only 18,815 claimants received training.101 As a result, the federal government entered into Labour Market Agreements (LMAs) with nine provinces and territories to fund skills development and training for non-eligible groups. These were renewed in 2014 as part of the Canada Job Fund Agreement and include the provision of up to $15,000 per person for training costs.102 In addition, there are relatively robust provincial programmes that provide financial incentives to employers who provide on-the-job training, especially in apprenticeships. For example, employers in Ontario can receive a $1000 taxable cash grant when hiring an apprentice, and wage subsidies and tax credits are given to employers who provide youth internships and training.103 However, most of these supports are in place for apprentices in skilled trades or for youth, and more could be done to target adults and older workers. 2. Intensified trade liberalization In contrast to recent US protectionist trends, Canada has embarked on a period of intensified trade liberalization under both Conservative and Liberal governments. In 2008, Canada only had four free trade agreements (FTAs) but has since brought into force nine more agreements and signed the Comprehensive and Progressive Agreement for Trans-Pacific Partnership (CPTPP) with 10 other countries in March 2018. However, there remain some trade adjustment concerns. The recently ratified Canada-European Union Comprehensive Economic and Trade Agreement (CETA) is estimated to increase Canada’s GDP by 0.77%, but it also provoked criticism that the expanded import quota for European cheese and enhanced IP rights in the pharmaceutical sector would hurt Canadian dairy farmers and pharmaceutical companies, respectively raising the question of whether the reintroduction of trade adjustment packages and compensation would be appropriate. In a persuasive paper, Lysenko and Schwartz104 argue that Canada does not need a TAA programme for several reasons that highlight the differences between the American and Canadian approaches to FTAs. First, a TAA programme is not necessary to overcome political opposition to trade agreements, whereas in the USA, the TAA was needed to ensure Congressional support for trade liberalization commitments.105 As explained below, consensus in Canada has instead been achieved through ad hoc federal compensation to different provinces and industries. Second, because Canada has already experienced substantial trade liberalization and has historically implemented trade liberalization commitments incrementally, adjustment costs from future FTAs will likely be small. Third, Canada has recognized that determining eligibility for a TAA programme is arbitrary in practice. While Canada had introduced a TAA programme in 1967, it was terminated in 1987 due to the practical difficulty of determining which workers had been injured by trade.106 The federal government’s ad hoc compensation policies warrant further comment. In recent years, Canadian trade policy has seen provinces become important political players that are willing to block FTAs given that certain trade issues fall under provincial jurisdiction. During the CETA negotiations, the federal government promised compensatory transfers to provinces as a way of reaching consensus where there was an issue of conflict. For example, Newfoundland was promised $280m in return for removing minimum local processing rules in its fishing industry and Ontario sought compensation for any negative effects of CETA, particularly for additional costs entailed for the pharmaceutical industry, cheese makers, winemakers, and distilleries.107 While this approach has been effective in addressing opposition, this assistance is provided primarily for political reasons and still implies an inequitable treatment of firms and workers in that some groups receive special benefits, but others who may be negatively harmed by FTAs receive none.108 A preferable policy approach would seek to aid all workers displaced by structural changes regardless of the cause of unemployment by providing strong ALMPs. 3. Decentralization of ALMPs As in the USA, the implementation of ALMPs in Canada is complicated by its federal system that has seen the federal government devolve labour market policy to the provinces and territories to mitigate political tensions. Because the Canadian constitution provides that the federal government is responsible for the economy while the provinces are responsible for education (thus training falls under both jurisdictions), disputes have arisen with regards to how training should be administered.109 These jurisdictional conflicts were resolved in 1996 when the Chrétien government devolved responsibility for training to provincial and territorial governments under the country’s EI scheme, thus paving the way for ‘collaborative federalism’,110 which has shaped labour market policy in Canada since. Under this scheme, the federal government transfers EI funds to provinces and territories to administer their own labour market policies to EI-clients under the previously mentioned LMDAs. The federal government retains control over the EI fund, delivers financial benefits, determines eligibility, and retains jurisdiction to develop and deliver ALMPs for non-EI clients (such as youth, persons with disabilities, Indigenous peoples, older workers, and recent immigrants). Decentralization has allowed for flexible and regionally sensitive policies and programmes that are adapted to local concerns. Despite the ‘administrative asymmetry’ that results from differing policies across the country, proponents argue that asymmetry is ‘built into federalism’ and is needed to address Canada’s regional diversity.111 However, because the LMDAs were negotiated individually with the provinces and territories, decentralization has developed without any coherent and unified policy framework for displaced workers.112 Critics have argued that this policy fragmentation has resulted in uneven programme quality across the country and federal programmes that are poorly integrated with provincial programmes. Bramwell suggests that because ALMPs are now de facto a provincial responsibility, Canada should decouple access to training programmes from EI eligibility and devolve responsibility and funding to provincial governments for training of displaced non-EI clients.113 C. Europe In contrast to the USA and Canada, which exhibit characteristics of the Anglo-Saxon model (i.e. relatively low employment protection and low unemployment benefits), Europe has seen the dominance of two models: a Nordic model and a continental model. While the Anglo-Saxon model emphasizes a ‘liberal activation’ framework or ‘defensive’ workfare (stronger work incentives, benefit conditionality, and the use of sanctions), the Nordic model emphasizes ‘universalistic activation’ or ‘offensive’ workfare such as human capital investment and upskilling.114 In past decades, the Nordic model has been held out as a model example for labour market policy, with strong ALMP supports for displaced workers and near universal unemployment benefits. While Nordic countries such as Sweden and Denmark have made major cuts in active measures and unemployment benefits in the past decade, they continue to spend the highest amount on ALMPs and PLMPs relative to other OECD countries. In contrast, continental Europe has placed more emphasis on employment protection than on actual labour-market participation.115 In this section, a brief discussion on intra-EU mobility will be followed by an analysis of the Nordic and continental models and the emerging trend towards defensive workfare. 1. Labour mobility within the EU As in the USA, the UK has seen a surge in protectionist rhetoric, as reflected when the UK voted to leave the EU with a 51.9% majority in the 2016 Brexit referendum, in which free movement of people within the EU was a central campaign issue. British–EU negotiations on Brexit began in June 2017 and have been strained over the issue of free movement within the EU, which has significant implications for intra-EU labour mobility. Under EU law, the principle of freedom of movement of workers is enshrined in Article 45 of the Treaty for the Functioning of the European Union (TFEU), which gives migrant workers who hold nationality in at least one EU Member State the right to work and reside freely within the EU, as well as freedom from direct and indirect discrimination based on nationality with regards to employment, remuneration, and other conditions of work and employment.116 Articles 20 and 21 of the TFEU further grant the right of EU citizenship to every person holding the nationality of a Member State, and gives all EU citizens the right to move and freely reside within the EU.117 The Brexit debate has brought freedom of movement and intra-EU mobility to the forefront, but in fact, despite the EU’s expansive legal framework, EU workers are not taking extensive advantage of EU mobility rights, and mobility rates have remained low. This has been a concern for EU policymakers because there have been high disparities in unemployment rates across Europe since the financial crisis in 2008. For example, the 2016 unemployment rate was 19.7% in Spain, 10.1% in France, and 4.1% in Germany.118 In 2014, only around 0.4% of the EU population moved to work in a different EU country, which is a lower cross-border mobility rate than Canada’s interprovincial labour mobility rate of around 0.8% and the US’s interstate mobility rate of around 2.3% in that same year.119 In addition, movement is concentrated towards a small number of countries. The main countries of residence of EU-28 movers were Germany (2.7m), the UK (2.1m), Spain (1.4m), Italy (1.1m), France, and Switzerland (around 950K each)—together, these six countries accounted for 75% of EU-28 movers in 2014.120 While labour mobility accounts for approximately 50% of long-run adjustment to region-specific labour demand shocks in both Europe and the USA, this process has been shown to take twice as long in Europe (10 years) than in the USA (5 years).121 This is likely due in part to greater labour market rigidities typically found in continental and Southern European countries. Labour mobility is also a less important adjustment mechanism in Europe than in the USA, most likely due to cultural, linguistic, and institutional differences across Europe that act as barriers to labour mobility. While there has been a rise in the role of migration in Europe (and a corresponding fall in interstate migration in USA), more could be done to facilitate labour mobility across Europe. The Bologna Process has helped to facilitate portability between universities (through transferable credits), but other policies that harmonize education standards, improve language education, and promote portability of pension and social security rights would enhance labour mobility.122 However, these sorts of policies will likely engender resistance as Brexit negotiations continue, which thus far suggest that free movement between the UK and EU will end in 2019 once the UK formally leaves the EU. 2. Nordic model The Nordic model is upheld as a model example of LMAPs for good reason, as Nordic countries have tended to perform better on unemployment and employment measures. Three features will be highlighted. First, Nordic labour market policies are unique due to their heavy investments in ALMPs and the influence of the Ghent system, whereby unemployment benefits are administered by trade-union-linked funds under a voluntary scheme. Second, Nordic countries are known as the best examples of the ‘flexicurity’ policy scheme, which the European Commission (EC) has promoted in the EU since 2006. Third, while Nordic countries have maintained a distinctive social policy approach, developments and trends in the past two decades have pointed towards a convergence with the ‘work first’ principle that characterizes traditions in the USA and Canada.123 (i) Emphasis on ALMPs and the Ghent system Of the OECD countries, Denmark spends by far the highest amount on ALMPs with 2.05% of its GDP dedicated to ALMPs in 2015 (recalling that the average for OECD countries is 0.53%), and a substantial amount on PLMPs, at 1.28%. ALMPs emphasize early intervention, in which unemployed individuals must enter a regime of mandatory activities after one month that include interviews, counseling, and monitoring of active job seeking. After six to nine months, job seekers must take part in job training and labour market education.124 Thus, ALMPs have an upskilling effect but also a motivation effect. After Denmark, Sweden currently invests the most in ALMPs across OECD countries, spending 1.27% of its GDP on active programmes in 2015.125 ALMPs were in fact first pioneered in Sweden in the 1950s by two trade-union economists under the Rehn–Meidner model, which sought to improve the match between the supply and demand of labour and to sustain full employment in an evolving economy by focusing on vocational training programmes.126 Since their inception in the 1950s, ALMPs have been a central feature of Swedish economic policy and are widely accepted within Sweden as the best means of protecting its citizens from unemployment and poverty.127 In addition to a tradition of ALMPs, another unique aspect of the Nordic model is the use of the Ghent system. The Ghent system was introduced in Belgium in 1901, and consists of a voluntary UI system administered by private agencies linked to trade unions that are funded largely by the state.128 While this system was previously more prevalent than compulsory UI schemes (i.e. UI contributions that are automatically deducted from the payroll), today the Ghent system remains in place only in Denmark, Sweden, and Finland. Under the Ghent system, individuals choose to pay a membership fee to a UI fund in return for unemployment benefits. Although an individual does not need to be part of a trade union to be a member of a UI fund, there seems to be a positive impact of voluntary UI on the strength of trade unions, and trade union density in these three Nordic countries is distinctively high compared with other OECD countries. For example, 67.7% of Swedish employees were members of a trade union in 2013, and collective bargaining agreements cover 90% of employees in Sweden.129 Trade union density in Denmark was 66.8% in 2013 and 69% in Finland (compared with 10.8% in the USA and 27.1% in Canada). As a result, trade unions exert considerable influence over important labour market regulations and play a large role in ensuring that the labour market functions effectively. While voluntary UI schemes may theoretically lead to a problem of adverse selection, this has not been borne out in reality. The coverage rate is near universal in Denmark and Sweden, with 90% of the Danish workforce covered in 2014130 and 85% of the Swedish workforce,131 which are significantly higher than in OECD countries with compulsory UI schemes. This is likely because the low direct costs of membership and high unemployment benefits give a clear incentive for joining a voluntary UI system.132 The monthly membership fee is around 440 DKK (US$70) in Denmark and SEK 352 (US$43) in Sweden, and can be offset against personal taxes. Benefits are generous, with a benefit duration of two years in Denmark (it was lowered from four years to two years in 2010) up to a maximum of 90% of the unemployed worker’s previous income. In Sweden, members of UI funds can receive a benefit period of 300 insurance days that pays 80% of an unemployed worker’s previous income for the first 200 days and 70% of previous income for the remaining days.133 For those who do not fulfill the membership condition, a basic insurance amount (amounting to 320 SEK per day, or approximately US$40) is provided under certain conditions.134 In addition, it is important to note that, unlike in Canada and the USA, the function of UI is not to provide a social safety net, but rather to act as a work incentive, as individuals must participate in ALMPs and be prepared to resettle in regions with labour demand. Thus, UI is seen as compensation for the activity of job seeking rather than for unemployment itself,135 and passive and active LMAPs are closely linked. (ii) Flexicurity ‘Flexicurity’ is a welfare state model that came to the forefront in the mid-2000s when it was promoted by the EC as its flagship policy, and emphasizes flexibility in the labour market (i.e. relaxed hiring and firing rules) combined with security (i.e. income and employment security in the form of generous unemployment benefits and ALMPs).136 In this model, job security and employment security are two distinct concepts. Whereas job security refers to a job with a single employer, employment security refers to the potential to hold jobs with multiple employers (i.e. employability).137 Flexicurity thus emphasizes low job security but high employment security by providing a menu of passive and active LMAPs including social protection (through a social safety net), training, and education. Flexible labour markets have several advantages—they increase responsiveness of wages, promote labour mobility across sectors, and allow for firms to be more efficient and competitive in response to economic changes. While the flexicurity model was first developed as a labour market policy in the Nordic countries, most notably in the Netherlands and in Denmark, flexicurity has subtly different meanings in the Netherlands, in Denmark, and for the EC. The term ‘flexicurity’ was first used in the Netherlands during discussions that resulted in the Flexibility and Security Act of 1999, which promoted atypical and flexible forms of employment while also providing for protection of workers on such contracts. These laws aimed for the ‘normalization of atypical work while preserving flexibility in the labour market’,138 and have provided security to non-standard workers in the last two decades by applying the pro rata temporis principle to part-time workers not just to wage and working conditions, but also to social security entitlements.139 As a result, contractual diversity in terms of part-time, full-time, and agency work is very high. In 2006, 46% of the employed Dutch labour force worked part-time (compared with 18% in EU-27), and part-time work is regarded as a normal and desirable type of employment, as indicated by the high number of part-time workers. These types of policies will likely become increasingly important in other countries experiencing sharp increases in non-standard employment. Whereas the Dutch model views ‘flexibility’ within the flexicurity model as flexible types of employment (i.e. atypical and non-standard employment), the Danish model defines ‘flexibility’ as ease of job mobility. Today, flexicurity has become synonymous with Danish flexicurity, often referred to as the ‘golden triangle’—it combines flexible employment (high rates of job mobility), a strong social safety net (high income security), and comprehensive ALMPs (high employment security).140 Denmark’s flexicurity framework evolved following the 1980s, during which time the country was experiencing a long period of high unemployment, with an 8.16% unemployment rate throughout that decade.141 To address this problem, Denmark undertook a series of labour market reforms in the 1990s that emphasized activation policies and implemented its flexicurity model. These labour market reforms were described as a ‘job miracle’, with unemployment falling from 11% in 1993 to 3.4% in 2008142 (unemployment rates increased to 7.5% in the years following the financial crisis and was 6.2% in 2016).143 Job mobility within Denmark is high by international comparisons, most likely due to a low level of job protection (consistent with its flexicurity policy), combined with a willingness among workers to move given Denmark’s strong income and employment security. Due to Denmark and the Netherlands’ impressive labour market performance, the EC has promoted flexicurity as a strategy to enhance flexibility and security in the labour market (consistent with the Danish model) and to address the problem of atypical employment (in line with the Dutch model). Whereas US-style flexibility assumes that social welfare will be increased by the benefits of flexible labour markets through the ‘trickle down’ effect, flexicurity assumes the need for policies to address income and employment security issues and provides both a safety net and a trampoline for workers moving through the labour market.144 While flexicurity is a rather vague concept, Peter Auer, formerly of the International Labour Organization, used a clustering analysis to define ‘flexicurity’ versus ‘non-flexicurity’ countries using variables as proxies for flexibility and security (i.e. working time flexibility, average employment tenure, and strictness of employment protection legislation).145 Consistent with the differences between the Nordic and continental models, Denmark, the Netherlands, Sweden, Finland, and Austria were classified as ‘flexicurity’ countries while France, Belgium, Germany, Italy, Spain, Greece, and Portugal were classified as ‘non-flexicurity’. Ireland and the UK were found to have high flexibility but low employment security, consistent with the Anglo-Saxon model. Auer found that the flexicurity countries performed better than the non-flexicurity countries on unemployment and employment rates, although flexibility countries also had the highest increase in unemployment during the 2008–2009 crisis period.146 However, strong causality between flexicurity and labour market performance has not yet been firmly established and major challenges arise in attempting to transplant flexicurity polices into countries with different political traditions and institutions. (iii) Towards activation Sweden first developed ALMPs in the 1950s, and since then all OECD countries have developed some measure of ALMPs.147 Within the menu of ALMP options available, different policies have been emphasized during different decades, and have been generally grouped into three periods. The first period was marked by skill market shortages in the 1950–60s, and ALMP use in this period was dominated by upskilling to provide skilled workers to meet this labour demand. The second period followed the oil shock of 1973–75, which saw high rates of unemployment across Europe. Many countries began using occupation ALMPs (i.e. providing temporary work in the public sector) with the goal of keeping workers engaged to slow the deterioration of human capital linked with unemployment.148 In the mid-1990s, the OECD and EC advocated a new ‘activation’ paradigm that focused on encouraging and facilitating the re-entry of unemployed individuals into the labour market, thus marking the third and most recent period of ALMP use. The activation paradigm acknowledged that there were interactions between UI systems and ALMPs, such that benefit conditionality makes a difference in terms of labour market adjustment programmes effectiveness.149 Thus, this mixes ‘offensive’ workfare, which emphasizes human capital investment, with ‘defensive’ workfare’, which emphasizes work incentives and benefit conditionality. Denmark became a pioneer of activation ALMPs and other countries such as Sweden, the UK, and Germany soon followed suit, and since the late 1990s, emphasis has been placed on more incentive reinforcement and employment assistance measures (i.e. counseling, placement services, and job-search programmes).150 This reorientation of the activation paradigm has provided some evidence of a trend towards a degree of convergence in labour market policies between North American and Nordic models. The Anglo-Saxon framework, particularly in the USA and Canada, has traditionally focused on a ‘work first’ principle consistent with liberal traditions in North America, in that benefit levels are lower, more restricted, and subject to conditions.151 While Nordic countries such as Sweden and Denmark are still noted welfare leaders, both countries have made major recent cuts in its benefits and instituted stricter work obligations, consistent with a converging trend towards activation policies and defensive workfare.152 For example, Denmark implemented several reforms to its UI system in the mid to late 1990s such as a shorter benefits duration period, strengthened eligibility rules, and mandatory participation in training. In Sweden, it was previously possible to requalify for UI benefits by enrolling in a training programme. However, this was criticized for generating a perpetual cycle between UI recipiency and training programmes, and new rules in 2001 deemed that this requalification could only take place once.153 In 2006, Sweden reformed its UI system once again, tightening eligibility conditions, lowering benefit levels, and raising monthly UI contributions more than three-fold, from SEK 100 (US$12) to SEK 352 (US$43).154 In addition to the strengthening of work incentives for benefit recipients, ALMP emphasis has shifted to employment assistance (i.e. job subsidies, counseling, and job search programmes) and incentive reinforcement (i.e. tax credits, benefit reductions, and benefit conditionality) accompanied by significant expenditure cuts to training ALMPs.155 These measures create stronger incentives and pressure on the unemployed to take up employment, and are consistent with the ‘liberal’ type of activation or defensive workfare strategy, thus pointing towards a growing convergence of policy objectives and instruments.156 Nonetheless, it can be argued that despite these policy changes, the character of the Nordic welfare state and labour market policies remain distinct, maintaining relatively high investments in ALMPs and PLMPS as well as relatively more emphasis on equality and universality of benefits coverage. 3. Continental model The continental model has traditionally been focused on strong job protection through robust employment protection laws, combined with relatively generous compulsory UI schemes and a limited emphasis on ALMPs. This model has often prevented countries like France and Germany from combating high unemployment rates, but reforms since the early 2000s have changed the character of the continental model. First, the newly elected French President, Emmanuel Macron, has promised sweeping reforms of a French labour market that has been plagued by a stubborn unemployment rate of between 8% and 10% over the last decade. Second, Germany instituted major reforms between 2003 and 2005 that have given rise to a so-called ‘German model’. Third, both of these reforms also make use of activation factors, pointing to a convergent trend towards defensive workfare. (i) Labour market rigidity France remains one of the best-known examples of labour market rigidity due to its stringent job protection laws. Economists have frequently argued that this creates an insider–outsider problem, since it protects privileged ‘insiders’ (permanent and full-time employed workers) at the expense of ‘outsiders’ (part-time workers and the unemployed) who are excluded from active or full participation in the labour market. Such a dual labour market can undermine human capital development, economic growth, and overall welfare. This problem has been well documented in France, where labour codes make dismissal procedures costly and lengthy, thus creating rigid labour markets that discourage employers from hiring permanent staff and also make it difficult for the economy to adjust to shocks. In addition, since France’s 35-hour workweek was approved by the socialist government in 2000 (down from 39 hours, after which time overtime pay is required), employers have argued that this has pushed up labour costs and made France less competitive.157 Previous French presidents have attempted to liberalize the French labour market in a bid to reduce the country’s high unemployment rate, but have been met with intense resistance from trade unions. President Macron came to power in June 2017 on a campaign promise to implement sweeping labour market reforms to ease labour regulations, create a more flexible labour market, and create jobs. Proposed reforms included capping the compensation for employees who are laid off for no justifiable reason, reducing the period during which employees can commence legal challenges after being made redundant, and making it easier for companies to fire employees for ‘economic reasons’.158 In September 2017, Macron signed decrees implementing some of these reforms, thus pushing France closer to the Danish ‘flexicurity’ model by giving employers greater discretion in firing workers and more confidence in hiring. However, Danish reformers also had the fiscal means to implement employment security measures such as extensive training programmes, and it remains to be seen whether Macron’s stimulus plan of 50 billion Euros (which accounts for about 2% of France’s GDP) will provide sufficient assistance for workers.159 (ii) Rise of the ‘German model’ While France is currently attempting to deregulate its labour market, Germany moved away from a rigid labour market more than one decade ago. In the 1990s, Germany was dubbed the ‘sick man of Europe’, unable to shake its high unemployment and weak economic growth, which were due to high levels of job protection, high labour costs, strict regulation of labour market, and ineffective LMAPs.160 PLMPs consisted of an unlimited payment duration of UI benefits and the highest replacement rates out of OECD countries for the long-term unemployed, thus providing weak incentives for the unemployed to take up new jobs. In terms of ALMPs, job creation programmes were the main tools used, with little emphasis on job search assistance and monitoring, and no system in place to assess programme effectiveness. From 2003 to 2005, Germany underwent a significant labour market policy overhaul, known as the Hartz reforms. Implemented in four waves during those years, the Hartz reforms sought to achieve three objectives: (i) restructure the federal employment agency; (ii) activate the unemployed by significantly reducing long-term unemployment benefits both in amount and duration; and (iii) stimulate labour demand by deregulating the labour market.161 The government also specifically tied the reforms to a programme evaluation mandate to analyse the effectiveness of ALMP programmes. After these reforms, the unemployment rate declined steadily, falling from 11.3% in 2005 to 4.1% in 2016,162 earning Germany the title of the ‘North Star’ of labour market policies and reforms.163 This success has largely been attributed to the increased flexibility of the German labour market that has allowed small and medium businesses to fire more easily, thus lowering the risk of hiring workers. In addition, studies have consistently shown that cuts in unemployment benefits significantly reduced unemployment.164 However, while Germany has undoubtedly been a success story in reducing unemployment, wage and income inequality have risen in the same period along with an increase in atypical forms of employment.165 Today, Germany uses a compulsory UI system with a one-year benefit duration (or 18 months for those over the age of 55 years). Replacement rates are capped at 60% of previous earnings, and the claimant must be looking for and available to work.166 In terms of ALMPs, Germany functions similarly to Canada by allocating a lump sum of money to regional governments and allowing local offices to decide how to budget between different ALMP measures. Prior to the Hartz reforms, these local offices did not provide training measures themselves, but rather contracted them out to private providers using a tendering process.167 However, competition was limited because in reality the contracts were awarded on the basis of familiarity with the provider. The Hartz reforms introduced quasi-market mechanisms through a system of training vouchers with the goal of increasing transparency, consumer choice, and competition.168 While these goals have been accomplished to some degree, clients often lack sufficient information to make informed choices, and competition has been attenuated because of an insufficient number of providers. As a result, it has been argued that a decentralized contracting-out system with an emphasis on performance measurement would better achieve those aims.169 While strictly outside of the ambit of our basic geographic comparative frame of reference, Australia is worthy of mention as it provides a similar example to Germany’s training voucher system, and is often cited as a model for market forces in employment training and placement.170 While other countries such as the Netherlands have also privatized their employment reintegration services, Australia is unique among OECD countries as its employment services are delivered by over 100 providers in a competitive quasi-market financed by service fees, employment outcome payments, and a special fund.171 Australia was one of the first OECD countries to use a quasi-market framework in 1998 when it began delivering employment services through contracts (using a tendering process) with community providers. The underlying assumption is that contracting out services will create competition to foster greater efficiency, better quality of service, and reduce costs.172 While contracting out raises concerns that providers might seek to ‘cream-skim’ easier-to-place workers, this problem is addressed by scaling outcome fees such that providers receive higher fees for individuals that are more difficult to place.173 In addition, the government publishes Star Ratings, which measures provider performance based on how many paid placements and paid outcomes were achieved. These ratings inform participants and are also used to eliminate poorly performing providers in subsequent tendering rounds. In 2009, further reforms through the Job Services Australia model emphasized employment outcomes by making providers responsible for organizing work-experience activities and enforcing a minimum hours of participation in its programmes. Since 2009, Australia has experienced some of the highest employment rates (i.e. the extent to which people available to work are being employed) among G7 and OECD G20 countries, in part due to effective and structured monitoring of its labour market programmes and its quasi-market delivery regime.174 IV. LESSONS FROM THE COMPARATIVE EXPERIENCE LMAPs have played an important role in the political economy, often helping to mitigate political opposition to technological innovation and trade openness. However, the effectiveness of these policies has been met with mixed degrees of success. It would be naïve to assume that some ideal model of LMAPs that has emerged in a particular country can be transplanted effectively to other countries with very different economies, geographies, histories, labour relations, and political systems. For example, the flexicurity model operates within broader ‘social-democratic’ welfare provisions that emphasize universal and portable benefits, which would be more difficult to implement in the context of the American ‘liberal’ political economy. In addition, research has suggested that differing degrees of trust and confidence between institutional actors (i.e. business, labour, and government) affect the success of adjustment programmes.175 The history of efforts to promote institutional reform in developing countries over the past twenty or thirty years by external donors promoting transplantation of ‘best practices’ from other jurisdictions, which efforts have met with mixed to poor results, counsels against such an assumption.176 Despite this important caveat against uncritical transplantation of policies and institutions from one country to another, perhaps the most striking conclusion to emerge from our comparative review of LMAPs in the previous section is a marked convergence towards some general principles, while recognizing that important differences in these policies are likely to persist from one country to another. We believe that comparative experience suggests that two opposing or outlier labour market policies have not well withstood the test of time and experience. First, the austere policy model for responses to job displacement that has historically prevailed in the USA, i.e., PLMPs principally exemplified by very limited forms of UI in terms of income coverage and length of benefits, and minimally supported by ALMPs, i.e., job counseling, job search, and job training (albeit with more extensive benefits for workers laid off as a result of international trade) has left many displaced workers bearing an unacceptable share of the burden of the costs of displacement. This is inequitable in terms of the disproportionate burdens that such workers bear, inefficient in under-utilizing potentially productive human capital, and politically perverse in intensifying opposition to trade, technology, and other factors affecting the demand for labour in developed economies. Second, the opposing model, most strongly exemplified by that which historically prevailed in France, entails extremely high levels of job protection and, in the event of job displacement, high levels of largely unconditional UI benefits in terms of duration and income replacement. These policies have tended to create a dual labour market that exhibits sharply different treatment between ‘insiders’ (workers with full-time employment) and ‘outsiders’ (workers with part-time employment or job-seekers who have been unable to secure any employment at all). The so-called Nordic model of labour market policies, exemplified by policies that have historically prevailed in countries such as Sweden, Denmark, and the Netherlands, is characterized by generous PLMPs, in particular generous UI benefits in terms of duration and income replacement, complemented by generous ALMPs, and has often been held out as a compelling alternative to either the USA or French models. However, in recent years even these policies have undergone significant modification, with a scaling back of passive labour market benefits, and increased reliance on conditionality attaching to such benefits that require displaced workers to participate successively in job search assistance programmes and combinations of classroom and on-the-job training in the event of continuing unemployment. Germany’s labour market policies have also increasingly converged on this model, with a readjustment in the balance between ‘defensive workfare’ and ‘offensive workfare-oriented’ policies. A striking feature of this increasing convergence in policies is an emphasis on high employment security but relatively low job security, or as it is sometimes called, ‘flexicurity’. These policies have sought to provide displaced workers with the knowledge that despite flexible hiring and firing practices, there is a robust system consisting of both a safety net and a trampoline in place, i.e. a system to provide generous income support and training programmes to displaced workers. Flexicurity in the Netherlands has also sought to normalize atypical work while preserving flexibility in labour markets and have provided security to non-standard workers by applying the pro-rata temporis principle to part-time workers not just to wage and working conditions but also to social security entitlements, resulting in high levels of contractual diversity in terms of part-time, full-time, and agency work.177 As we noted in the prior section, in 2006, 46% of the employed Dutch labour force worked part-time and part-time work is regarded as a normal and desirable type of employment. In such a model, receipt of UI benefits would be conditional, after a relatively short period of receipt, on participation in a job counselling and search programme and in the event that this proves unsuccessful in identifying alternative employment within a relatively short period, receipt of further UI benefits would be conditional on participation in a classroom or on-the-job training programme (or a combination of the two). However, participation in these two programmes would not be restricted to recipients of UI benefits but would be available to other unemployed workers who do not qualify for unemployment insurance, recognizing that in many countries a relatively large (and increasing) percentage of unemployed workers do not qualify for unemployment insurance.178 A related concern is that unemployed workers who do not qualify for UI benefits may not be able to support themselves or their families while in job assistance programmes. While Canada, the USA, and most countries in Europe have highly developed financial assistance schemes for higher education or training, most of these programmes are limited to students and young adults. Thus, there is a need for a comprehensive income-contingent loan programme to support older adults who would otherwise be unable to participate in job assistance programmes. Comparative experience, particularly recent German and Australian experience, strongly suggests that the delivery of these programmes should not be an exclusive government monopoly but be provided on a competitive basis by a broad range of public, non-profit, and for-profit providers, with substantial scope for choice by job seekers among competing service providers, with such choices informed by appropriate prior job counselling services. Such a diversity of providers accords recognition to differences in employment prospects by sector, region, and characteristics of job seekers. Some form of government accreditation of service providers is likely desirable to weed out clearly fraudulent or incompetent providers and to establish some basic conditions that would need to be met to qualify for full or partial reimbursement of fees from government sources (in effect, a kind of voucher programme). These conditions would establish on the public record providers’ fee schedules and reimbursement rates, and would require periodic public disclosure of specified performance metrics, e.g., completion rates and post-completion employment experience of participants, which would be subject to periodic audits for accuracy. While out-of-pocket expenses incurred by participants would be largely borne by government, living expenses for participants for the duration of longer programmes (e.g. remedial education or long-term vocational training), might be financed or underwritten by government on an income-contingent loan basis. In the delivery of such programmes, there are many virtues in geographically large and regionally diverse countries, such as the USA, Canada, and Australia (which are all federal states), for delivery to be decentralized to subnational or local levels of government that are likely to be better attuned to local labour market needs than central governments. However, there is an off-setting risk that partial or parochial perspectives may excessively influence local decision-making, particularly when realistic future job opportunities may entail job seekers moving out of the jurisdiction, hence justifying a role for central governments in attaching conditions to financial assistance, or negotiating agreements with subnational levels of government, governing the basic contours of the services provided to ensure consistency with national labour market needs. With respect to the financing of these programmes, there is a case for moving away from primary reliance on UI premiums or payroll taxes (given the declining percentage of the unemployed receiving UI benefits) and greater reliance on some form of broad-based national consumption tax, with refundable tax credits for low-income earners, given that consumers are the principal beneficiaries of lower product and service costs from trade and technology. There is a clear role for governments at all levels in more systematically collecting (and frequently updating) intelligence on short- to medium-term forecasted employment needs of a representative range of employers by sector and region which would be a key input into the services provided by decentralized service providers (and accredited by government) and to choices by individual job seekers among programme providers. While local government agencies might continue to administer unemployment insurance, they would largely be confined to the role of an information-clearing house on the availability of other service providers in the public, non-profit, and for-profit sectors. Moreover, a striking lesson that emerges from our comparative survey is the importance of having governments commit to systematic, periodic independent programme evaluations for cost-effectiveness, which have often been haphazard and crude in the past, with programmes sustained more by faith than facts.179 Finally, investing in education and training increases the competitiveness of the economy by allowing workers to better respond to economic change. As the Independent Task Force on the Future of the US Workforce has put it, a change in thinking is needed such that education and work are viewed as two closely linked activities rather than as separate and distinct.180 Companies can provide some of this training, but it is difficult for smaller firms to support retraining and government help may be necessary for countries to maintain a competitive edge. While strictly outside the remit of this article, we are attracted to a model based on the Singapore Skills Future Initiative that would allocate to every individual of working age a modest annual voucher (of, e.g., $500) to be spent on any of a wide array of accredited training courses, thus signifying a collective commitment to lifelong learning, or what Yuval Levin calls in a recent book, ‘a mobility agenda’.181 This commitment will become especially important as trade and technology continue to change the labour market landscape, rendering well-designed and targeted LMAPs a high priority and a superior response to protectionist policies in addressing today’s costs from job displacement and economic transitions. Funding Project was funded out of a general University of Toronto research grant. Footnotes 1 ‘The New Political Divide’, The Economist, 30 July 2016, www.economist.com (visited 12 May 2018). 2 Michael Trebilcock, Dealing with Losers: The Political Economy of Policy Transitions (Toronto: Oxford University Press, 2014) at chapter 5. 3 See Michael Trebilcock, Advanced Introduction to International Trade Law (UK: Edward Elgar Publishing, 2015) at chapter 8. 4 For further discussion on the employment effects of trade, see ILO, ‘Trade and Employment: From Myths to Facts’, 4 October 2011, http://www.ilo.org/employment/areas/trade-and-employment/WCMS_162297/lang--en/index.htm (visited 12 May 2018). 5 Douglas Irwin, Free Trade Under Fire, 4th ed (Princeton: Princeton University Press, 2015) at 115 [Irwin]. 6 WTO, ‘World Trade Report 2017: Trade, Technology and Jobs’ (Geneva: World Trade Organization, 2017), https://www.wto.org/english/res_e/publications_e/wtr17_e.htm (visited 12 May 2018), at 109 [WTO Report 2017]. 7 David Autor, David Dorn and Gordon Hanson, ‘The China Shock: Learning from Labour Market Adjustment to Large Changes in Trade’, Annual Review of Economics (2016), 205–40. 8 Gary Hufbauer and Sean Lowry, ‘US Tire Tariffs: Saving Few Jobs at High Cost’, Peterson Institute for International Economics Policy Brief No. PB 12-9 (2012). 9 Jagdish Bhagwati and Robert Hudec (eds), Fair Trade and Harmonization: Prerequisites for Free Trade? (Cambridge: MIT Press, 1996) vol. 2. 10 Irwin, above n 5, at 162. 11 Mark Wu, ‘The China Inc. Challenge to Global Trade Governance’, 57 Harvard International Law Journal (2016), 261. 12 See Michael Trebilcock, Advanced Introduction to International Trade (Cheltenham, UK/Northampton, MA: Edward Elgar, 2015), chapter 6. 13 See Peter Singer, One World Now: The Ethics of Globalization (New Haven: Yale University Press, 2016), at 107–8. 14 Ibid, at chapter 2. 15 See Justin Lin, The Quest for Prosperity: How Developing Economies Can Take Off (Princeton: Princeton University Press, 2012) and Joseph Stiglitz, Globalization and Its Discontents Revisited: Anti-Globalization in the Era of Trump (New York: WW Norton, 2017). 16 For discussion on some of these rationales, see Ronald Daniels and Michael Trebilcock, Rethinking the Welfare State (Oxon: Routledge, 2005), chapter 10 [Daniels]. 17 Ibid. 18 For further discussion on policies needed to address the rise of non-standard forms of employment, see ILO, ‘Non-standard employment around the world: Understanding challenges, shaping prospects’ (2016), http://www.ilo.org/global/publications/books/WCMS_534326/lang--en/index.htm (visited 12 May 2018). 19 For reviews of many of these issues see, e.g.: Anthony Atkinson, Inequality: What Can Be Done? (Cambridge: Harvard University Press, 2015); Richard Reeves, Dream Hoarders: How the American Upper Middle Class is Leaving Everyone Else in the Dust, Why That is a Problem, and What To Do About It (Washington: Brookings Institution Press, 2017). 20 Daniels, above n 16, at 203. 21 Werner Eichhorst and Regina Konle-Seidl, ‘Evaluating Labour Market Policy’, 3 Institute for the Study of Labor (2016). 22 Robert S. Rycroft, The American Middle Class: An Economic Encyclopedia of Progress and Poverty (Santa Barbara: ABC CLIO, 2017), at 314. 23 Daniels, above n 16, at 203. 24 Giuliano Bonoli, ‘The Political Economy of Active Labor-Market Policy’ 38(4) Politics and Society 435 (2010), at 436 [Bonoli]. 25 OECD, OECD Employment Outlook 2015 (Paris: OECD, 2015), https://www.oecd-ilibrary.org/employment/oecd-employment-outlook-2015_empl_outlook-2015-en (visited 12 May 2018), at 107 [OECD Employment Outlook 2015]. 26 John P. Martin, ‘Activation and Active Labour Market Policies in OECD Countries: Stylized Facts and Evidence on their Effectiveness’, IZA Policy Paper No. 84 (2014) [Martin]. 27 Bonoli, above n 24, at 5–6. 28 Three of these categories (incentive reinforcement, employment assistance, and upskilling) are drawn from Bonoli, above n 24. 29 OECD, Activation Strategies for Stronger and More Inclusive Labour Markets in G20 Countries (Paris: OECD, 2013), 8 OECD, https://www.oecd.org/els/emp/G20-2013ReportActivation.pdfpdf (visited 12 May 2018). 30 OECD Employment Outlook 2015, above n 25, at 140. 31 Daniels, above n 16, at 206. 32 David Schleicher, ‘Stuck! The Law and Economics of Residential Stability’, Yale Law Journal (2017), 127 [Schleicher]. 33 Schleicher, above n 32, at 15. 34 Olivier J. Blanchard, Florence Jaumotte and Prakash Loungani, ‘Labor Market Policies and IMF Advice in Advanced Economies During the Great Recession’, 3(2) IZA Journal of Labor Policy (2014), at 5. 35 Gosta Esping-Andersen, The Three Worlds of Welfare Capitalism (Cambridge: Polity Press, 1990). 36 United States Bureau of Labour Statistics, https://www.bls.gov/ (visited 12 May 2018). 37 For further discussion on inequality and policies to tackle poverty, see Anthony Atkinson, Inequality: What Can Be Done? (Cambridge, MA: Harvard University Press, 2015). 38 Lawrence Mishel, Elise Gould and Josh Bivens, ‘Wage Stagnation in Nine Charts’, Economic Policy Institute (2015), http://www.epi.org/publication/charting-wage-stagnation/ (visited 12 May 2018). 39 Stijn Broecke, Shruti Singh and Paul Swaim, ‘The State of the North American Labour Market’ (Report prepared for the 2016 North American Leaders Summit, 2016) (Paris: OECD Directorate for Employment, Labour and Social Affairs, 2016), at 57 [Broecke]. 40 For discussion regarding the rise of corporate training programs, see Josh Bersin, ‘Spending on Corporate Training Soars: Employee Capabilities Now a Priority’, Forbes, 4 February 2014, https://www.forbes.com/sites/joshbersin/2014/02/04/the-recovery-arrives-corporate-training-spend-skyrockets (visited 12 May 2018). 41 Gregg M. Olsen, ‘Labour Market Policy in the United States, Canada and Sweden: Addressing the Issue of Convergence’, 42(4) Social Policy and Administration 323 (2008), at 326 [Olsen]. 42 OECD, OECD Employment Outlook 2017 (Paris: OECD, 2017), http://www.oecd.org/els/oecd-employment-outlook-19991266.htm (visited 12 May 2018), at 216 [OECD Employment Outlook 2017]. 43 Ibid, at 218. 44 Olsen, above n 41, at 329. 45 ‘Unemployment Insurance Chartbook’, United States Department of Labor, https://oui.doleta.gov/unemploy/chartbook.asp (visited 12 May 2018). 46 US Replacement Rates taken from US Department of Labor: the replacement ratio was 0.399, and represented the ratio of a weighted average weekly benefit amount (WBA)/weighted average of (normal hourly wage × 40 hours), https://oui.doleta.gov/unemploy/ui_replacement_rates.asp (visited 12 May 2018). 47 Judith M. Conti, ‘Unemployment Insurance: An Overview of the Challenges and Strengths of Today’s System’ (Hearing before the United States Congress Human Resources Subcommittee, 7 September 2016) (2016) National Employment Law Project, https://waysandmeans.house.gov/wp-content/uploads/2016/09/20160907HR-Testimony-Conti.pdf (visited 12 May 2018), at 2. 48 Carrick et al., ‘The Skills Gaps in U.S. Manufacturing 2015 and Beyond’, The Manufacturing Institute and Deloitte LLP (2015), at 3. 49 The other obvious answer is to raise wages, although that can disrupt internal wage structures. 50 Ruth Graham, ‘The Retraining Paradox’, The New York Times, 23 February 2017, https://www.nytimes.com/2017/02/23/magazine/retraining-jobs-unemployment.html [Graham] (visited 12 May 2018). 51 Helena Worthen, ‘The Workforce Investment Act and the Labor Movement’, 6(3) WorkingUSA 70 (2002–3), at 72. 52 David E. Balducchi, Randall W. Eberts and Christopher J. O'Leary, ‘Experience and Prospects for Labor Exchange Policy’, in David E. Balducchi, Randall W. Eberts and Christopher J. O'Leary (eds), Labor Exchange Policy in the United States (Kalamazoo: W E Upjohn Institute, 2004), at 252. 53 Lawrence F. Katz, ‘America’s Jobs Challenges and the Continuing Role of the US Department of Labor’, 67 Industrial and Labour Relations Review 578 (2014), at 581. 54 Graham, above n 50. 55 US Government Accountability Office, Report to Congressional Requesters, ‘Multiple Employment and Training Programs: Providing Information on Colocating Services and Consolidating Administrative Structures Could Promote Efficiencies’ (2011), http://www.gao.gov/products/GAO-11-92 (visited 12 May 2018), at 11. 56 ‘WIOA Overview’, United States Department of Labor, https://www.doleta.gov/WIOA/Overview.cfm [WIOA Overview] (visited 12 May 2018). 57 WIOA Overview, above n 56. 58 Benjamin Collins, ‘Trade Adjustments Assistance for Workers and the TAA Reauthorization Act of 2015’ (CRS Report Prepared for the Members and Committees of Congress, 14 September 2016) (Washington: Congressional Research Service, 2016), at 3 [Collins]. 59 Dmitry Lysenko and Saul Schwartz, ‘Does Canada Need Trade Adjustment Assistance?’, 57 IRPP Study (2015), at 4 [Lysenko]. 60 Irwin, above n 5, at 146. 61 Edward Alden, Failure to Adjust: How Americans Got Left Behind in the Global Economy (Lanham: Rowman & Littlefield, 2017), at 120 [Alden]. 62 Collins, above n 58, at 9. 63 Ibid, at 8. 64 Sarah Dolfin and Peter Z. Schochet, The Benefits and Costs of the Trade Adjustment Assistance (TAA) Program under the 2002 Amendments (Princeton, NJ: Mathematica Policy Research, 2012), at 58. 65 Ronald Daniels and Michael Trebilcock, ‘A Better Way to Help US Victims of Free Trade’, Bloomberg (2016), https://www.bloomberg.com/view/articles/2016-12-02/a-better-way-to-help-u-s-victims-of-free-trade [Bloomberg] (visited 12 May 2018). 66 Alden, above n 61, at 113. 67 ‘America’s Programme to Help Trade’s Losers Needs Fixing’, The Economist, 1 July 2017, at 67, citing David Autor, David Dorn and Gordon Hanson, ‘The China Syndrome: Local Labor Market Effects of Import Competition in the United States’, 103(6) American Economic Review 2121 (2013), at 2148. 68 Ronald D’Amico and Peter Z. Schochet, ‘The Evaluation of the Trade Adjustment Assistance Program: A Synthesis of Major Findings’ (Final Report prepared as part of the Evaluation of Trade Adjustment Assistance Program for the US Department of Labor Employment and Training Administration Office of Policy Development and Research, December 2012) (Washington: US Department of Labor Employment and Training Administration, 2012), at iv. 69 Irwin, above n 5, at 150. 70 Recipiency rates pulled from ‘Unemployment Insurance Chartbook’, United States Department of Labor, https://oui.doleta.gov/unemploy/chartbook.asp (visited 12 May 2018); income replacement based on a weighted average of: weekly benefit amount / (normal hourly wage × 40 hours) in the year 2016 from ‘UI Replacement Rates Report’, United States Department of Labor, https://oui.doleta.gov/unemploy/ui_replacement_rates.asp (visited 12 May 2018). 71 Collins, above n 58, at 8. 72 Carolyn J. Heinrich, ‘Workforce Development in the United States: Changing Public and Private Roles and Program Effectiveness’, in Douglas Besharov and Douglas Call (eds), Labour Activation in a Time of High Unemployment: Encouraging Work While Preserving the Social Safety-Net (Oxford: Oxford University Press, 2018), at 2. 73 Schleicher, above n 32, at 25. 74 Brink Lindsey and Steven M. Teles, The Captured Economy (New York: Oxford University Press, 2017), at 92. 75 Morris M. Kleiner and Evgeny Vorotnikov, ‘Analyzing Occupational Licensing among the States’, 52(2) Journal of Regulatory Economics 132 (2017), at 150. 76 Morris M. Kleiner, Licensing Occupations: Ensuring Quality or Restricting Competition? (Kalamazoo, MI: W.E Upjohn Institute for Employment Research, 2006), at 115. 77 ‘America Should Get Rid of Oppressive Job Licensing’, The Economist, 17 February 2018, www.economist.com. 78 Schleicher, above n 32, at 18. 79 Engler et al., ‘The Work Ahead: Machines, Skills, and US Leadership in the Twenty-First Century’ (2018) Council on Foreign Relations, Independent Task Force Report No. 76, at 24 [Task Force Report]. 80 ‘Left behind’, The Economist, 21 October 2017, at 20. 81 Ibid, at 2. 82 Bloomberg, above n 65. 83 For further discussion of policy recommendations regarding the changing nature of work in the USA, see Task Force Report, above n 79. 84 ‘Chapter 1: Labour Market Context: 2015/2016 Employment Insurance Monitoring and Assessment Report’, Canadian Employment Insurance Commission, https://www.canada.ca/en/employment-social-development/programs/ei/ei-list/reports/monitoring2016/chapter1.html (visited 12 May 2018). 85 Canadian Charter of Rights and Freedoms, Part I of the Constitution Act, 1982, being Schedule B to the Canada Act 1982 (UK), 1982, c 11, s 6. 86 Peter Hogg, Constitutional Law of Canada, 5th ed, supp vol. 2 (Toronto: Edmond Montgomery, 2016), at 46. 87 Ibid, at 46–47. 88 Department of Finance Canada, Jobs Report: The State of the Canadian Labour Market (Ottawa: Department of Finance Canada, 2014), at 28. 89 Broecke, above n 39, at 31. 90 Ibid. 91 Morley Gunderson, ‘Changes in the Labour Market and the Nature of Employment in Western Countries’, in Rethinking Workplace Regulation: Beyond the Standard Contract of Employment (New York: Russell Sage Foundation, 2013). 92 OECD Employment Outlook 2017, above n 42, at 216. 93 Broecke, above n 39, at 23. 94 ‘Ontario Basic Income Pilot’ (24 April 2017), Government of Ontario, https://news.ontario.ca/mcss/en/2017/04/ontarios-basic-income-pilot.html (visited 12 May 2018). 95 Government of Ontario, News Release, ‘Proposed Changes to Ontario’s Employment and Labour Laws’ (30 May 2017), https://news.ontario.ca/mol/en/2017/05/proposed-changes-to-ontarios-employment-and-labour-laws.html (visited 12 May 2018). 96 To clarify, the Employment Insurance program is Canada’s unemployment insurance program. It was renamed ‘Employment Insurance’ from ‘Unemployment Insurance’ in 1996 to better reflect the program’s goal of promoting employment in the labour force. 97 Olsen, above n 41, at 328. 98 Applied Research Branch, Strategic Policy, Human Resources, Development Canada, An Analysis of Employment Insurance Benefit Coverage (Ottawa: Development Canada, 1998), at 13. 99 ‘Table 1, Employment Insurance Coverage Survey’ (2015), Statistics Canada, http://www.statcan.gc.ca/daily-quotidien/161116/t001b-eng.htm (visited 12 May 2018). 100 ‘Employment Insurance Service Quality Review Report: Making Citizens Central’ (2015–16), Employment Insurance Program, https://www.canada.ca/en/employment-social-development/programs/ei/ei-list/reports/service-quality-review-report.html (visited 12 May 2018), at chapter 3. 101 ‘Number of Employment Insurance Beneficiaries by Type of Income Benefit’ (21 February 2013), Statistics Canada, http://www.statcan.gc.ca/tables-tableaux/sum-som/l01/cst01/labor13-eng.htm (visited 12 May 2018). 102 Government of Canada, ‘Canada Job Fund Agreements: ESDC’ (14 January 2016), https://www.canada.ca/en/employment-social-development/programs/training-agreements/cjf.html (visited 12 May 2018). 103 ‘Wage subsidies and tax credits for employers’ (30 January 2017), Canada Business Ontario, http://www.cbo-eco.ca/en/index.cfm/financing/government-loans-and-grants/wage-subsidies-and-tax-credits-for-employers/ (visited 12 May 2018). 104 Lysenko, above n 59, at 23. 105 Ibid, at 10. 106 Lysenko, above n 59, at 10. 107 Ibid, at 5. 108 Ibid, at 27. 109 R. Albert Berry, Labor Market Policies in Canada and Latin America: Challenges of the New Millennium (New York: Springer Science+Business Media, 2001), at 253. 110 Allison Bramwell, Training Policy for the 21st Century: Decentralization and Workforce Development Programs for Unemployed Working-Age Adults in Canada (Toronto: Mowat Centre for Policy Innovation, 2011), at 4 [Bramwell]. 111 Ibid, at 5. 112 Ibid, at 2. 113 For more detailed discussion of Bramwell’s recommendations for reform, see 16 to 21 of Bramwell, above n 110. 114 Bonoli, above n 24, at 439. 115 Ibid. 116 Consolidated Version of the Treaty on the Functioning of the European Union, Art. 45, 2008 OJ C 115/47. Note that Art. 45(3) stipulates that the rights encompassed under the free movement of workers may be limited on grounds of public policy, public security, or public health. In addition, under 45(4), freedom of movement does not apply to employment in the public service. 117 Elena Fries-Tersch, Tugce Tugran and Harriet Bradley, 2016 Annual Report on Intra-EU Labour Mobility, 2nd ed. (Brussels: European Commission, 2017), at 16 [2016 Intra-EU Labour Mobility Report]. 118 OECD Employment Outlook 2017, above n 42, at 190: ‘Table A, Harmonised unemployment rates in OECD countries’. 119 2016 Intra-EU Labour Mobility Report, above n 117, at 43. 120 Ibid, at 11. 121 Robert C. M. Beyer and Frank Smets. ‘Labour Market Adjustments and Migration in Europe and the United States: How Different’, Economic Policy 643 (2015), at 647 [Beyer]. 122 Beyer, above n 121, at 669. 123 Olsen, above n 41, at 331. 124 Per Kongshoj Madsen, ‘Flexicurity in Danish: A Model for Labour Market Reform in Europe?’ 43(2) Intereconomics 74 (2008), at 75. 125 OECD Employment Outlook 2017, above n 42, at 216: ‘Table Q, ‘Public expenditure and participant stocks in labour market programmes in OECD countries, 2014 and 2015’. 126 Bonoli, above n 24, at 444. 127 Olsen, above n 41, at 332. 128 Jochen Clasen and Elke Heins, ‘Voluntary Unemployment Insurance and Trade Union Membership: Investigating the Connections in Denmark and Sweden’, 37(3) Journal of Social Policy 433 (2008), at 434 [Clasen]. 129 Hulya Ulku and Silvia Muzi, ‘Labor Market Regulations and Outcomes in Sweden’ (2015) World Bank Group Development Economics Global Indicators Group Working Paper No. 7229, at 6. 130 OECD, ‘Denmark’ in Benefits and Wages (Paris: OECD, 2014), at 2. 131 Olsen, above n 41, at 330. 132 Clasen, above n 128, at 7. 133 Mattias Bengtsson, Transformation of Labour Market Policies in Nordic Countries: Towards a Regime Shift in Sweden and Denmark? (Gothenburg: University of Gothenburg, Department of Sociology and Work Science, 2012), at 10 [Bengtsson]. 134 ‘Unemployment’, Nordic Social Insurance Portal, http://www.nordsoc.org/en/Sweden/Unemployment1 (visited 12 May 2018). 135 Bengtsson, above n 133, at 10. 136 Sonja Bekker and Ton Wilthagan, ‘Flexicurity: A European Approach to Labour Market Policy’, March/April Intereconomics 68 (2008), at 68 [Bekker]. 137 Peter Auer, ‘What’s in a Name? The Rise (and Fall?) of Flexicurity’, 52(3) Journal of Industrial Relations 371 (2010), at 380 [Auer]. 138 Bekker, above n 136, at 71. 139 Ibid. 140 Maarten Keune, ‘Flexicurity: A Contested Concept at the Core of the European Labour Market Debate’, March/April Intereconomics 92 (2008), at 94 [Keune]. 141 Bengtsson, above n 133, at 6. 142 Ibid, at 7. 143 OECD Employment Outlook 2017, above n 42, at 190: ‘Table A, Harmonised unemployment rates in OECD countries’. 144 Auer, above n 137, at 373. 145 Ibid, at 374. 146 Ibid, at 376–77. 147 OECD Employment Outlook 2017, above n 42, at 216: ‘Table Q of OECD Employment Outlook 2017 to review the level of ALMP expenditure in OECD countries.’ 148 Bonoli, above n 24, at 446. 149 Martin, above n 26, at 4. 150 Bonoli, above n 24, at 449. 151 Olsen, above n 41, at 331. 152 Bengtsson, above n 133, at 17. 153 Olsen, above n 41, at 331. 154 Clasen, above n 128, at 15. 155 Bengtsson, above n 133, at 18. 156 Olsen, above n 41, at 331. 157 Hugh Schofield, ‘Labour Law Reform: Hollande’s Last Throw of the Dice’, The BBC, 15 March 2016, http://www.bbc.co.uk/news/world-europe-35803001 (visited 12 May 2018). 158 Chloe Farand, ‘Emmanuel Macron Kicks off Controversial Attempt to Reform France’s Labour Laws’, The Independent, 28 June 2017, http://www.independent.co.uk/news/world/europe/emmanuel-macron-france-president-launches-controversial-labour-market-reforms-labour-law-a7813171.html (visited 12 May 2018). 159 Philip Rathgeb and Fabio Wolkenstein, ‘Third-Way à la Français: What do Macron’s Reforms Involve and How Likely are They to Succeed?’, London School of Economics, http://blogs.lse.ac.uk/europpblog/207/06/23/third-way-francaise-macron-reforms/ (visited 12 May 2018). 160 Ulf Rinne. ‘Is Germany the North Star of Labor Market Policy?’ IZA Discussion Paper No. 7260, at 1 [Rinne]. 161 Ibid, at 4. 162 OECD Employment Outlook 2017, above n 42, at 190: ‘Table A, Harmonised unemployment rates in OECD countries’. 163 Rinne, above n 160, at 1. 164 Niklas Engbom, Enrica Detragiache and Faezeh Raei, ‘The German Labour Market Reforms and Post-Unemployment Earnings’, International Monetary Fund Working Paper No. WP/15/162 (2015), at 10. 165 Werner Eichhorst, ‘Non-Standard Employment across Occupations in Germany: The Role of Replaceability and Labour Market Flexibility’, IZA Discussion Paper No. 7662 (2013), at 2. 166 OECD, ‘Germany’ in Benefits and Wages (Paris: OECD, 2014), at 2. 167 Rüdiger Wapler, Daniel Werner and Katja Wolf, ‘Active Labour-Market Policies in Germany: Do Regional Labour Markets Benefit?’, Institute for Employment Research IAB Discussion Paper No. 28/2014 (2014), at 12. 168 Oliver Bruttel, ‘Delivering Active Labour Market Policy through Vouchers: Experiences with Training Vouchers in Germany’, 71(3) International Rev of Administrative Sciences 391 (2005), at 396 [Bruttel]. 169 Ibid, at 402. 170 Ludo Struyven and Geert Steurs, ‘Design and Redesign of a Quasi-Market for the Reintegration of Jobseekers: Empirical Evidence from Australia and the Netherlands’, 15(3) Journal of European Social Policy 211 (2005), at 212. 171 OECD, Activating Jobseekers: How Australia Does It (Paris: OECD, 2012), http://www.oecd.org/els/emp/activatingjobseekershowaustraliadoesit.htm (visited 12 May 2018), at 13 [How Australia Does It]. 172 Ibid, at 62. 173 Ibid, at 76. 174 How Australia Does It, above n 171, at 13. 175 WTO Report 2017, above n 6, at 10. 176 Michael Trebilcock and Mariana Mota Prado, Advanced Introduction to Law and Development (Toronto: Edward Elgar Publishing, 2014), at chapter 3. 177 See also: Jordann Thirgood, ‘What if You Could Take It With You?’ (27 July 2017), Mowat Centre, https://mowatcentre.ca/what-if-you-could-take-it-with-you/ (visited 12 May 2018). 178 Bramwell, above n 110. 179 While effective government evaluations are lacking, there have been many academic evaluations of ALMPs in OECD countries, including (i) James Heckman, Robert Lalonde and Jeffrey Smith, ‘The Economics and Econometrics of Active Labor Market Programs’ in David Card and Orley Ashenfelter (eds), Handbook of Labor Economics, vol. 3 (Amsterdam: Elsevier Science, 1999); and (ii) David Card, Jochen Kluve and Andrea Weber, ‘Active Labour Market Policy Evaluations: A Meta-Analysis’, 120 Economic Journal (2010) F452 to F477. 180 Task Force Report, above n 79, at 35. 181 Yuval Levin, The Fractured Republic: Renewing America’s Social Contract in an Age of Individualism (New York: Basic Books, 2016). © The Author(s) 2018. Published by Oxford University Press. All rights reserved. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of International Economic Law Oxford University Press

Trade, Technology, and Transitions: Trampolines or Safety Nets for Displaced Workers?

Loading next page...
 
/lp/ou_press/trade-technology-and-transitions-trampolines-or-safety-nets-for-JJ7h5Mul00
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press. All rights reserved.
ISSN
1369-3034
eISSN
1464-3758
D.O.I.
10.1093/jiel/jgy023
Publisher site
See Article on Publisher Site

Abstract

Abstract In the past several decades, the developed world has experienced significant labour market dislocations caused by international trade, technology, and other factors. While economic nationalism has risen in response to these challenges, technology is typically a more important factor than trade as a cause of these dislocations. Further, trade-related responses often impose additional costs on consumers through higher prices and on downstream industries that utilize inputs from protected sectors. Thus, the article argues that effective use of labour market adjustment policies (LMAPs) is a preferable approach to protectionist policies in addressing labour market adjustment costs. After laying out a spectrum of passive and active labour market policies, the article then goes on to provide a comparative evaluation of LMAPs in the USA, Canada, select Nordic and continental countries in Europe, and Australia. The article’s comparative evaluation suggests that the Nordic model, Germany, and Australia provide the most compelling utilization of LMAPs, while the USA lags behind other countries in our sample in relative resources devoted to LMAPs. However, recent trends in these jurisdictions suggest some degree of convergence on an ‘activation’ paradigm that utilizes incentive reinforcement and benefit conditionality in triggering participation in active labour market programmes. I. INTRODUCTION In many parts of the developed world, economic nationalism is on the rise, reflecting what The Economist magazine1 describes as a new political divide where political contests are not left versus right but open versus closed (drawbridges down or drawbridges up), implying a significant challenge to the broad free trade consensus that has prevailed in most of the developed world since the end of World War II. In this article, we provide a comparative evaluation of alternative labour market adjustment policies (LMAPs), arrayed along a spectrum from passive labour market policies (PLMPs) to active labour market policies (ALMPs), in response to labour market disruptions caused by international trade, technology, and other factors. We argue that such LMAPs are a preferable approach to the drawbridges up approach. To set the context, we set out some general propositions that we believe command a substantial consensus amongst trade policy scholars. First, the international trade law regime, especially the multilateral trade law regime beginning with the inception of the GATT in 1947 and superseded by the World Trade Organization in 1995, has always been sensitive to labour and related adjustment costs induced by trade liberalization. As the first author has argued elsewhere,2 the multilateral regime incorporated three important features designed to moderate adjustment costs: gradualism, reciprocity, and reversibility. With respect to gradualism, tariff and related restrictions on trade have been typically reduced gradually both within and across negotiating rounds. With respect to reciprocity, by structuring trade liberalization negotiations around the principle of reciprocity, export-oriented industries were often able to be enlisted as a political counterweight to import-impacted sectors and in addition provided the potential for moderating transition costs experienced by contracting import-impacted sectors through the reallocation of resources, including labour, over time to expanding export-oriented industries. With respect to reversibility, safeguard provisions built into the original GATT and maintained to the present-day permit importing countries facing an unexpectedly large import surge to reimpose tariffs or other import restrictions on a temporary basis to moderate the impact of these imports on domestic industries and their workforces, although unfortunately the Appellate Body of the WTO has interpreted the safeguard provisions so restrictively that invocation of these provisions has become highly problematic.3 Second, while it is indisputable that trade liberalization may cause job displacement in particular import-impacted sectors, the impact of trade liberalization on total employment in a developed country’s economy is much less clear.4 As Douglas Irwin argues, in a masterful review of both the theory and empirical evidence, ‘this is because the impact of trade on the total number of jobs in an economy is best approximated as zero’.5 Trade theory does not claim that trade liberalization will necessarily increase the total number of jobs in an economy, but rather that labour productivity will increase as countries focus on sectors in which they enjoy a comparative advantage. While there are net market losses in certain regions or sectors, the impact of trade has been found to be positive for overall employment and real wages.6 Third, import competition accounts for a relatively small percentage of total job displacements in most developed countries’ economies relative to technology, shifts in consumer preferences, new sources of domestic competition, etc. While the USA has experienced a decline of approximately six million manufacturing jobs since 2000, international trade, including the recent dramatic growth of imports from China, explains 20–25% of these job losses7 with the balance largely attributable to technologically induced increases in productivity in the manufacturing sector. Fourth, empirical evidence shows that protectionist responses to trade-induced labour market dislocations typically cost consumers in higher prices several multiples of the value of the jobs preserved in the protected industry. For example, in 2009 the Obama Administration imposed safeguard measures on car and truck tires imported from China resulting in an estimated saving of 1200 jobs in the US tire industry, amounting to $900,000 in consumer costs per job saved per year and resulting in the loss of an estimated 3731 jobs in other sectors of the economy.8 A high percentage of developed country imports are of intermediate goods that are inputs into the production of other goods and services, so that protectionist measures directed at these imports raise costs for downstream industries that use the imports as inputs and often result in much larger job losses in these downstream industries relative to jobs preserved in the protected sectors. However, it is commonly argued that some forms of contemporary international trade may constitute a form of ‘unfair’ trade or unfair competition9 and hence economic dislocations induced by unfair trade may be normatively unacceptable and justify trade restrictions or sanctions. First, trade deficits are often cited as evidence of unbalanced trade commitments. However, such concerns are often misplaced, with empirical evidence revealing that trade deficits often rise with a booming domestic economy where consumers have more resources to spend on imports. Moreover, as Irwin points out, if a foreign country is exporting more than it imports from another country, the overall balance of payments is divided between the current account and the financial account, which includes all portfolio and direct investments.10 Foreign countries can either use their export earnings to purchase imports or to invest the surplus in acquiring assets or investments in the country with which they are running a trade surplus (for example, China’s investment in US treasury bills and other securities). Thus, if the USA were to invoke protectionist measures to reduce the trade deficit, then capital inflows from abroad would necessarily have to fall and domestic investment would have to be financed by domestic savings, implying higher interest rates, which would reduce the number of jobs created by business investment, with the net effect on domestic employment unclear. Beyond the issue of trade deficits, unfair trade concerns often focus on claims that foreign exporters are deliberately manipulating their currency to induce an undervaluation and hence render their exports cheaper and imports more costly. It is difficult to evaluate as an analytical matter the ‘true’ value of a country’s currency, but in any event in the case of the principal target of such complaints in recent years, i.e., China, it now seems widely accepted that China has allowed its currency (the renminbi) to appreciate significantly in recent years and that it may no longer be significantly undervalued. A related complaint of unfair trade concerns foreign subsidization of exports, which renders them artificially competitive in importing countries’ markets, and leads to job displacement in these markets. This is again a difficult complaint to evaluate analytically, especially in the case of economies in transition from command to market economies (like China’s), with large numbers of state-owned enterprises and extensive use of regulated or directed pricing,11 but as with currency manipulation, it is a complaint that cannot be dismissed out of hand and is subject to detailed disciplines in the WTO’s Subsidies and Countervailing Measures Agreement. However, there are legitimate concerns over the appropriateness and enforceability of a number of these disciplines, nor do they address at all an increasingly salient issue of tax and subsidy competition among host jurisdictions to attract foreign direct investment. A further complaint is that foreign exporters often ‘dump’ goods in the markets of importing countries at lower prices than prevail in their home markets. While anti-dumping actions are often initiated by importing countries on this or related bases, beyond the very narrow category of cases involving predatory pricing by foreign exporters, most dumping cases lack any coherent economic rationale and simply reflect geographic price discrimination.12 Nevertheless, anti-dumping actions have proliferated around the world in recent years and have often become the protectionist instrument of choice, despite their lack of any convincing economic rationale in most cases, and dominate by orders of magnitude the two other principal trade remedy regimes: safeguards and countervailing duties which have more plausible economic rationales. A further claim of unfair trade relates to foreign countries improperly appropriating the intellectual property of firms based in other countries through either lax laws or lax enforcement thereof, hence conferring on them an artificial comparative advantage in international trade and unfairly prejudicing these firms and their workforces. While the Trade-Related Intellectual Property Agreement (TRIPs) negotiated during the Uruguay Round seeks to address many of these concerns, there may still be legitimate concerns relating to the lax enforcement of these commitments. Moreover, TRIPS does not apply to the controversial issue of technology-sharing agreements imposed on foreign investors as a condition of investment in host countries. A yet further set of ‘unfair’ trade complaints relate to lax labour or environmental laws in foreign countries that enable them to reduce costs unfairly, rendering their products artificially competitive in international trade. With respect to concerns over lax labour standards, while it may well be the case that insisting that all trading countries adhere to core labour standards such as those promulgated by the International Labour Organization (ILO) in its 1998 Declaration on Fundamental Principles and Rights at Work, i.e., freedom of association; the elimination of forced labour; the elimination of child labour; and the elimination of discrimination in employment (and perhaps also basic workplace safety standards), is normatively defensible on non-economic rationales, it is important to recognize (as the ILO does) that lower wage levels in some foreign countries (especially developing countries) are not a source of unfair trade but indeed a key source of their comparative advantage (due account being taken of differences in labour productivity across countries).13 Similarly, lax environmental standards in some foreign countries may be a legitimate concern for other countries where these involve transboundary externalities or threats to the global commons (such as climate change) and justify trade sanctions,14 but where their effects are purely local it is not nearly as obvious that other countries have any legitimate basis for insisting on adherence to their own environmental standards as a precondition to international trade. Yet further unfair trade concerns focus on the challenges facing small, less developed countries in competing in an open international trading system as a result of weakly developed industrial and financial sectors and the continuing predominance of traditional (albeit arguably inefficient) forms of agricultural production, thus arguably justifying dispensations from normal trade disciplines to advance more dynamic conceptions of latent comparative advantage and to moderate transition costs, given limited fiscal capacity to compensate for these directly.15 Where do these propositions leave us in terms of addressing legitimate concerns on, both equity and efficiency grounds, over the labour market dislocations induced by international trade, assuming that unfair trade concerns are not legitimately engaged? First, it is often impossible to disentangle job displacement (or indeed rising levels of inequality and stagnant middle-class incomes) caused by trade from job displacement (and rising inequality and stagnant incomes) caused by technology and other factors, so that trade-related (i.e. protectionist) policy responses are likely in most cases to be misdirected. Second, even if it were practically possible and ethically defensible to separate out trade-induced labour market dislocations from dislocations caused by other factors, trade-related responses are an enormously costly LMAP, both in terms of additional costs to domestic consumers from higher prices and hence reduced spending in other sectors of the economy, and in terms of higher costs and hence often greater job displacement in downstream industries that utilize the inputs of protected sectors. Hence, we argue in this article that most developed countries, in responding to labour market dislocations, should place a higher priority on domestic LMAPs, the effectiveness of which we seek to evaluate from comparative experience in the balance of this article in addressing the different implications of cyclical, frictional, and structural unemployment. While as a matter of political economy well-designed LMAPs may not fully mute demands for protectionism by trade-impacted sectors, they provide a principled basis for other constituencies to resist these demands by demonstrating a commitment to taking seriously the losers from economic transitions more generally. II. A MENU OF LMAPS There are various rationales for government intervention in labour markets, including bargaining failure, information failure, externalities, and equity principles.16 A laissez-faire approach to training would assume that employers are willing to pay for training that is firm-specific, and that employees are willing to pay for the acquisition of general human capital.17 However, in practice, employers and employees are often unable to make optimal decisions about investments in training and it would be unwarranted to assume that appropriate LMAPs should be left entirely to private parties. The menu of LMAPs available to government entails a spectrum of passive to active LMAPs, each with inherent strengths and limitations, depending on the causes of unemployment. There have been growing concerns in many developed countries in recent years over stagnant lower and middle incomes, rising inequality, and the dramatic increase in non-standard, precarious employment,18 which implicate a wide range of public policies for reducing inequalities in market and post-market incomes.19 However, in this article we focus more narrowly on job displacement from standard and non-standard employment, and how displaced workers can most effectively be supported using both passive and active LMAPs. We survey this menu of policies briefly below, before turning to an evaluation in the next section of comparative experience in designing the optimal combination of policies on this menu. A. PLMPs PLMPs can be described as ‘safety nets’ that helps minimize losses to workers through income support, using policy options such as unemployment insurance (UI), government subsidies to struggling businesses or regions, and wage-fixing policies. Because these benefits were originally developed to provide temporary assistance to the cyclically unemployed, PLMPs are ineffective in addressing structural unemployment (that is, they target only the symptom of unemployment rather than the cause).20 In addition, unemployment benefits tend to induce disincentive effects that increase the incidence and duration of unemployment. Recognizing this, the trend in OECD countries has been to shift resources away from PLMPs to more active policies.21 Still, two other passive policies have attracted significant recent attention. The first is wage insurance, which provides a temporary wage supplement to displaced workers who are compelled to move to lower-paying jobs. The second is universal basic income (UBI), a fixed and unconditional cash payment with clawbacks based on earned income, given to all citizens in a given economy. UBI has gained traction across the ideological spectrum in recent years as a potential solution to the problem of automation and technology displacing workers. While proponents argue that this would allow people to invest time and money in education or training, critics argue that people would be less motivated to work.22 UBI is likely to be too coarse-grained as a stand-alone response to address the widely varying circumstances of individuals who are unemployed, but new empirical evidence is likely to be soon forthcoming, with basic income pilots currently in place in countries such as Finland, Canada, India, and Namibia. B. ALMPs In contrast to a ‘safety net’, ALMPs act as a ‘trampoline’ to help those displaced from the labour market find or acquire new opportunities for re-entering it.23 These policies target the unemployed by reference to their income earning potential, and have been promoted by international agencies and adopted across the OECD over the past three decades.24 While evaluations of ALMPs have been mixed, studies clearly show that policies that are well designed and well targeted have a positive impact on participants’ employability and earnings.25 Over the past decade, ALMPs have become integrated with ‘activation’ strategies that recognize important interactions between the generosity of passive benefits, the size and mix of ALMP spending, and incentive measures such as attaching benefit eligibility to certain conditions.26 In this vein, some analyses have conceptualized ALMPs as dichotomous: ‘offensive’ workfare, which emphasizes improving skills and human capital investment, and ‘defensive’ workfare, which operates through work incentives, benefit conditionality, and the use of sanctions.27 The array of ALMPs can be grouped more finely into four broad categories.28 These categories, exhibiting defensive to offensive characteristics in this order, are incentive reinforcement, demand-side policies, employment assistance, and upskilling. 1. Incentive reinforcement This category refers to measures that strengthen the motivation for benefit recipients to take up existing job opportunities or to participate in programmes that will strengthen the recipient’s earnings capacity. This can be achieved by making passive benefits conditional upon behavioural requirements such as participation in programmes that strengthen individuals’ productivity or work capacity, through in-work benefits, or through the use of benefit sanctions (in the case of non-compliance).29 This has perhaps been best represented by the United States’ welfare-to-work reform in the 1990s. 2. Demand-side policies While ALMPs have tended to focus on supply-side factors (such as employment assistance and upskilling), demand-side policies (i.e. job creation and work experience programmes) are receiving increasing attention as they help to ensure that there are employment opportunities available to begin with. Other demand-side policies include programmes to finance business start-ups, and employment incentives such as targeted wage or hiring subsidies. Studies have found that employment incentives and direct job creation have positive impacts for disadvantaged groups if well targeted.30 However, public employment programmes have tended to result in temporary ‘make-work’ activities that have little effect on the future employability of participants.31 3. Employment assistance Employment assistance facilitates re-entry into labour-market participation, and includes counseling, placement services, and job-search programmes. Job search assistance is the main policy in this category, which aims to connect employers with the unemployed more efficiently. However, because these programmes help individuals find jobs for which they are already qualified, this policy option is appropriate only for frictional unemployment and not for structural unemployment. Job relocation assistance may be more effective. Policymakers have sometimes neglected barriers to geographic mobility that may prevent a qualified worker from becoming re-employed. Mobility rates have been especially problematic in the USA, where workers in depressed local markets tend to either drop out of the labour force or remain unemployed instead of relocating.32 Job relocation assistance such as relocation subsidy programmes would help alleviate these problems by providing individuals with incentives to overcome constraints on mobility. Another method of assistance is trade adjustment assistance (TAA), which has figured prominently in the USA, with mixed results. TAA programmes generally involve a mix of PLMPs and ALMPs by providing income support, skills training, and job-search assistance to workers who have lost jobs as a result of international trade. 4. Upskilling Upskilling uses education and training mechanisms to offer a ‘second chance’ to people whose skills have become obsolete, and is ideally designed to target structural unemployment. While there is a large literature devoted to the evaluation of classroom and on-the-job training programmes, a more recent concept of lifelong learning is emerging, reflecting the adaptability that workers require today to address the challenge of increasingly rapid labour market and technological changes. The private sector has already begun responding to this need. Companies such as Coursera, Udacity, and LinkedIn offer massive open online courses (MOOCs) that embrace an employment-focused business model. However, MOOCs are currently geared towards those with university degrees, and are not easily accessible to low-skilled workers who may lack time, money, and motivation.33 Unless government intervenes to level the playing field, lifelong learning may do little to reduce inequality. III. COMPARATIVE EXPERIENCE A comparative evaluation of LMAPs across developed countries reflects the fact that different jurisdictions have traditionally favoured different mixes of policy options. These differences have been the focus of an important line of research and have at times been generalized into three models.34 The ‘Anglo-Saxon’ model exhibits low employment protection, low UI, and the use of ‘defensive’ workfare (such as work incentives and benefit conditionality), which in theory leads to high job flow, short unemployment duration, and low unemployment. The ‘Nordic’ model—or ‘flexicurity’ model—is based on a medium to high degree of employment protection, generous but conditional UI, and strong ALMPs, thus allowing for effective re-employment and low unemployment. The Nordic model emphasizes near universal benefits and operates within broader welfare provisions that include universal health insurance, public education, maternity leave, and childcare. As a result, while both the Anglo-Saxon and Nordic models have historically been relatively successful, the Anglo-Saxon model is also associated with significantly more inequality than the Nordic model. The third model, the ‘continental’ model, is based on high job protection, generous UI, and limited ALMPs, and has been widely judged to be an ineffective method of reallocation of labour and of combating unemployment, as we discuss further below. While these groupings are to some extent stylized (for example, Germany has moved away from a continental model to a more Nordic model), these models are helpful for comparing and assessing different policy approaches and also draw on Esping-Andersen’s analysis of the ‘liberal’ welfare state, the ‘social-democratic’ welfare state, and the ‘corporatist’ welfare state.35 For example, the liberal welfare state, typically exemplified by the USA and Canada (although we argue that Canada is a middle ground between a US-styled liberal regime and a more interventionist European regime), is market-oriented and exhibits modest benefits and strict entitlement rules catered to low-income state dependents, which is consistent with the Anglo-Saxon model. The social-democratic welfare state exhibits de-commidifying and universalistic programmes that promote an equality of high standards for all strata, consistent with the Nordic model (exhibited by the Nordic countries). Finally, the corporatist welfare state displaces the market as a provider of welfare and aims to preserve occupational groups and status differentials. This can be seen in the continental model and especially in France, where historically rigid job protection laws have created an insider–outside problem (which will be later discussed). The chart below summarizes the models and their characteristics, and places the countries that will be surveyed along a spectrum according to their fit within each model. The following section will compare on-the-ground LMAP experiences in the USA, Canada, and selected countries in Europe, with a brief additional reference to recent Australian experience. Model Characteristics Esping-Andersen’s welfare state regime type Countries surveyed in this article Anglo-Saxon Low employment protection Low UI Emphasis on ‘defensive’ workfare, i.e. work incentives, benefit conditionality, and the use of sanctions Liberal USA Canada Denmark Sweden Finland Netherlands Germany France Nordic (‘flexicurity’) Flexibility in the labour market (i.e. relaxed hiring and firing rules) combined with high income and employment security (i.e. generous UI and strong ALMPs) Emphasis on ‘offensive’ workfare, i.e. human capital investment and upskilling, with a movement towards ‘defensive’ workfare Near universal unemployment benefits Social-democratic Continental High employment protection Generous UI Limited ALMPs Corporatist Model Characteristics Esping-Andersen’s welfare state regime type Countries surveyed in this article Anglo-Saxon Low employment protection Low UI Emphasis on ‘defensive’ workfare, i.e. work incentives, benefit conditionality, and the use of sanctions Liberal USA Canada Denmark Sweden Finland Netherlands Germany France Nordic (‘flexicurity’) Flexibility in the labour market (i.e. relaxed hiring and firing rules) combined with high income and employment security (i.e. generous UI and strong ALMPs) Emphasis on ‘offensive’ workfare, i.e. human capital investment and upskilling, with a movement towards ‘defensive’ workfare Near universal unemployment benefits Social-democratic Continental High employment protection Generous UI Limited ALMPs Corporatist Model Characteristics Esping-Andersen’s welfare state regime type Countries surveyed in this article Anglo-Saxon Low employment protection Low UI Emphasis on ‘defensive’ workfare, i.e. work incentives, benefit conditionality, and the use of sanctions Liberal USA Canada Denmark Sweden Finland Netherlands Germany France Nordic (‘flexicurity’) Flexibility in the labour market (i.e. relaxed hiring and firing rules) combined with high income and employment security (i.e. generous UI and strong ALMPs) Emphasis on ‘offensive’ workfare, i.e. human capital investment and upskilling, with a movement towards ‘defensive’ workfare Near universal unemployment benefits Social-democratic Continental High employment protection Generous UI Limited ALMPs Corporatist Model Characteristics Esping-Andersen’s welfare state regime type Countries surveyed in this article Anglo-Saxon Low employment protection Low UI Emphasis on ‘defensive’ workfare, i.e. work incentives, benefit conditionality, and the use of sanctions Liberal USA Canada Denmark Sweden Finland Netherlands Germany France Nordic (‘flexicurity’) Flexibility in the labour market (i.e. relaxed hiring and firing rules) combined with high income and employment security (i.e. generous UI and strong ALMPs) Emphasis on ‘offensive’ workfare, i.e. human capital investment and upskilling, with a movement towards ‘defensive’ workfare Near universal unemployment benefits Social-democratic Continental High employment protection Generous UI Limited ALMPs Corporatist A. United States The US labour market has seen significant recovery from the Great Recession—unemployment rates fell from 9.9% in November 2009 to 4.1% in November 2017 (more than one percentage point lower than the post-war unemployment average),36 indicating that the US economy is near full employment. However, this figure does not provide much relief for the many middle-class American workers facing rising inequality, real wage stagnation, a dramatic rise in non-standard employment, and limited job opportunities for those without a college degree.37 To illustrate, the wages of the top 1% of income earners have grown 138% since 1979 while wages for the bottom 90% grew a mere 15%. In fact, middle-wage workers’ hourly wage has increased only 6% between 1979 and 2013, and low-wage workers’ hourly wage declined 5% during that same period.38 Since the end of the 1990s, long-term unemployment has been rising and labour force participation has been declining, likely due to the perception that jobs are difficult to find.39 While it is understandable why an economic nationalist agenda designed to protect and restore the jobs of American workers resonated with many voters in the 2016 US presidential elections, protectionist measures are mostly an ineffective response (as we argue in our introductory section). The central problem is that the country has failed to invest in effective social programmes and LMAPs to cushion the working class through the adjustment process. Current US labour market supports are both inadequate and ineffective, with several dozen uncoordinated federal job-retraining programmes in addition to numerous state-level programmes (although corporate training programmes are increasing).40 In addition, the US exhibits several distinct features that pose special policy challenges. First, the USA is a traditional user of PLMPs and is characterized by low levels of investment in ALMPs. Second, the TAA programme has taken on a central role in American policy but has largely failed to produce positive results. Third, the country faces significant policy complications induced by federalism and regulatory barriers, leading to inconsistent treatment of workers across the nation and a declining interstate mobility rate. 1. Low levels of investment into ALMPs The USA has historically been classified as a ‘liberal’ welfare state that provides few forms of social protection, marginal income benefits, and a narrower range of social services compared with most other developed countries.41 The average public expenditure for OECD countries in 2015 (as a % of GDP) was 0.78 for PLMPs and 0.53 for ALMPs, while the USA spent substantially less on both at 0.18 for PLMPs and 0.10 for ALMPs.42 In fact, the USA has viewed ALMPs as having a minimal role today in the labour market adjustment process—of all OECD countries, the USA was one of the countries with the lowest public expenditure on ALMPs according to the data from 2012.43 As a nation that relies more heavily on passive policies, UI is the primary safety net for unemployed workers. UI is compulsory and mandated by federal legislation, but the proportion of the unemployed who receive UI benefits has declined since the 1980s, partly due to much tighter eligibility rules such as decreased income cutoffs (by counting social security income and pension benefits) and stricter enforcement of work search requirements.44 While these eligibility requirements may help combat disincentive effects of UI benefits, the recipiency rate, or the percentage of insured unemployed out of total unemployed, was on average only 27% in 2015 and 2016 (compared with 43% in 1980).45 This problem will continue to grow in today’s emerging world of part-time, contract, and gig workers who are excluded from UI benefits. The duration and level of UI benefits have also declined during the same period—today, the maximum duration of benefits is 26 weeks and the average income replacement rate was around 40% in 2016,46 down from 39 weeks in the 1970s and a 50% income replacement rate between the 1970s to 1990s. Still, UI has played an important role in the American economy, and was estimated to have kept five million people out of poverty during the Great Recession when the duration of benefits was temporarily extended.47 2. Job training programmes: the federal workforce system and the central role of the TAA There is a growing disconnect between two narratives surrounding the American blue-collar work force: on the one hand, workers claim that jobs are going overseas or becoming automated, and on the other hand, employers have been insisting that they have been unable to fill skilled blue-collar jobs. According to a 2015 report by the Manufacturing Institute, seven out of ten manufacturing executives faced a shortage of workers with adequate technological skills.48 Job training seems to be the obvious answer to bridge the gap between these narratives,49 but federal government retraining programmes have failed to provide an integrated national scheme. Instead, there is a fragmented mix of federal and state-level schemes that have been called a ‘dizzying constellation’ of programmes that has become ‘increasingly byzantine’.50 Today, the US’s primary federal employment assistance and training programme is the Workforce Innovation and Opportunity Act (WIOA), which supersedes a complex body of federal legislation. Another key piece of legislation is the Trade Adjustment Assistance Reauthorization Act of 2015, which updates the TAA programme. In the following section, a brief history of the WIOA’s precursors is discussed to provide some perspective on America’s changing employment policy, followed by an evaluation of the TAA, the latter of which is unique among developed countries for the central political role that it has played. (i) Evolution of federal employment training programmes The first federally funded training and employment assistance programme in the USA was the Works Progress Administration (WPA), created in 1935 as part of the New Deal to alleviate the unemployment crisis during the Great Depression. During this period, public projects were the focus and training occurred only if necessary in small and fragmented programmes. In the 1970s, the government enacted the Comprehensive Employment and Training Program (CETA) that placed these programmes under the control of local governments, thus marking a trend towards decentralizing employment training. In 1982, CETA was succeeded by the Job Training Partnership Act (JTPA) and by 1994, there were 154 federal programmes run by 14 agencies, many of which were redundant.51 The Workforce Investment Act (WIA) of 1998 was a major structural overhaul that replaced the JTPA in an attempt to consolidate employment-training programmes into a single funding stream and further decentralized training to the states.52 However, the system remained fragmented and difficult to navigate, with different entry points and eligibility rules that made participation and choice difficult.53 By 2009, there were 47 programmes administered by nine federal agencies, a substantial decrease from 1994. However, almost all programmes overlapped with at least one other programme, and were criticized as a ‘maze of confusing training programmes’ by former President Obama.54 Perhaps of more concern, a 2011 report by the Government Accountability Office (GAO) found that ‘little is known about the effectiveness of most programmes’ as only 5 of 47 programmes had conducted impact evaluations. These five impact studies found that benefits to unemployed workers were small, inconclusive, or non-existent.55 In response to some of these concerns, the WIOA was enacted in 2014 by bipartisan majorities in Congress, and represented a significant reform of the federal workforce system. Recognizing that there was a limited evidence base for ALMPs, the WIOA seeks to improve accountability and transparency by requiring measurement and public reporting of a programme’s performance (such as employment and earning outcomes) by education and training providers. Notably, it mandates that programmes be evaluated at least every four years by independent third parties.56 The WIOA also requires states to submit a single unified strategic plan to better align workforce development programmes and help ensure that job seekers have seamless access to career services, education, and training. Other reforms include strengthened funding for work-based learning that focuses on youth programme services and apprenticeships, and the use of pay-for-performance contracts to incentivize higher quality training.57 With Final Rules that became effective July 2016, the WIOA promises some effective changes that will be evaluated on a regular basis. (ii) Central role of the TAA Although TAA programmes exist in other jurisdictions (the EU and South Korea introduced TAA programmes in 2007), the USA is unique for the central political role that its TAA has played since the 1960s. For example, in the fiscal year 2016, Congress appropriated $861m for the TAA, allocating $450m for training and reemployment services and $411m for income support.58 The EU’s programme has an annual budget of only $170m, while South Korea’s small programme provides support to firms rather than to workers.59 Under the current TAA, which was reauthorized by the TAA Reauthorization Act in 2015, displaced workers who can establish that foreign trade contributed importantly to their job loss can access training and reemployment services, weekly income support, wage insurance, and a health coverage tax credit. While the TAA has seen both expansions and budget cuts throughout its life, it has remained an integral part of the US’s trade policy for over half a century. The theory underlying the TAA programme is that while trade liberalization increases overall economic welfare, domestic costs fall disproportionately on certain sections of the workforce who suffer dislocation due to foreign competition or off-shoring. The perceived appropriate response is to compensate the losers (who are often older, have less education, and hold shorter job tenure60) with more robust benefits and services than would be available to a worker displaced for other reasons. The TAA was first created by the Kennedy Administration’s Trade Expansion Act of 1962, and has since historically been reauthorized alongside trade liberalization policies. While the programme offered income benefits, vocational education, and moving expense payments, the eligibility criteria were very difficult to meet until they were further expanded by the Trade Act of 1974. Still, between 1975 and 1981, only 4% of TAA-eligible workers had actually entered training programmes.61 In 2002, the Bush Administration expanded eligibility criteria and benefits once again, offering a wage insurance programme for the first time as well as a health coverage tax credit. The current TAA offers TAA-certified workers 50 years and older who are reemployed at a lower wage a cash payment equal to a 50% of the difference between the worker’s trade-affected job’s wage and the new wage, to a maximum of $10,000 over two years.62 However, workers receiving wage insurance cannot also receive income support from the TAA, which is provided to participants enrolled in training within 26 weeks of job separation or TAA certification. Training is paid for directly to the service provider or through a voucher system that allows participants to choose from a range of approved programmes, and participants can have 90% of their job search and relocation expenses covered, up to $1250 respectively.63 The TAA has been difficult to evaluate due to the lack of data—prior to the new WIOA performance-reporting requirement, no ongoing assessments were required. Recent evaluations have produced mixed to negative results. A 2012 report found that the net cost to participants was $26,837 due to lower wages than in the matched comparison group,64 and while the wage loss is likely due to participants engaging in training programmes rather than working, the training programmes appear to have flaws as well. A 2012 evaluation by the US Department of Labor (DOL) found that 85% of those who undertook TAA-funded training received a certificate or degree, but only 37% of them were working in that field four years later. In addition, those who received training waited an average of over seven months from the time they filed for UI before obtaining training placement.65 The result is that the TAA programme has become ‘token help for workers hurt by import competition’,66 and help for Americans has come largely from other sources. Autor, Dorn, and Hanson estimated that between 1990 and 2007, only 6% of additional government payments associated with Chinese import competition came from TAA or UI. The bulk of these payments came from other sources, with 32% coming from disability or retirement insurance, 26% from federal-government income assistance, and 32% from extra medical spending.67 An evaluation prepared for the DOL found that without considering the benefits stemming from the possibility that the TAA reduces resistance to free trade, the benefits of the programme as it operated under the 2002 amendments were negative.68 Commentators have also questioned the underlying rationale of giving special assistance only to trade-affected workers. Empirically, displaced workers in trade-sensitive industries are often similar to workers displaced in other industries, and there is no compelling economic or ethical basis for why these workers should be treated differently. Further, while the TAA may have helped reduce opposition to trade liberalization, there has been no evidence since the 1970s that it has been able to ‘buy off’ labour groups opposed to trade.69 3. Complications produced by federalism Another distinctive feature of the USA is its federal system and regional diversity, which produce complications for incentivizing job training, for the administration of federal programmes, and for internal labour market mobility. First, states often have weak incentives to provide training to rectify structural unemployment since they do not want to invest resources to train workers to move out of their own jurisdiction and reduce the local tax base. Second, while UI, the TAA, and the WIOA are federal programmes that remain federally funded, they are all administered at the state and local levels, resulting in striking disparities and differing treatment of workers across the country. For example, while federal legislation determines what forms of employment are covered by UI, each state establishes its own rules governing scale of benefits, eligibility, and duration of benefits. The recipiency and income replacement rates vary dramatically—recipiency rates in 2016 ranged from a low of 10% in Florida to a high of 51% in Massachusetts, and income replacement rates ranged from a low of 32% in Alaska to a high of 56% in Hawaii.70 Similarly, the TAA’s income support programme also reflects this kind of variation. In 2015, the highest maximum weekly benefit for a worker with no dependents was $698 in Massachusetts, while the lowest weekly benefit was $240 in Arizona.71 Finally, the WIOA is based on the assumption that authority should be decentralized so that states and local agencies can adapt employment and training programmes to their own economic and political circumstances. As a result, one-stop centres are operated locally, with the result that there are significant differences in how job seekers access and receive its programmes.72 Third, states and local governments have created many legal barriers that make interstate mobility difficult, such as land use laws that restrict housing construction, occupational licensing rules, conditions on public pensions, and variations in the standards for receiving public benefits.73 Currently, there are 1100 different jobs subject to licensing requirements in at least one state (ranging from doctors to manicurists, upholsterers, and funeral attendants)74 and while licensing is typically justified on grounds of consumer protection, it raises the incomes of the highest paid earners by up to 24% (compared with 3.6–5.1% for the lowest paid earners) while preventing lower-skilled workers from providing even basic services.75 More importantly, licensing requirements are often arbitrary, as evidenced by the dramatic inconsistency in licensing requirements across states. Excessive occupational licensing is a form of protectionism that slows economic growth and interferes with entrepreneurship and innovation, and was estimated to have created a deadweight loss accounting for a $40 billion loss of output.76 Better solutions include the use of inspections rather than mandatory licenses, or the use of optional qualifications and titles to allow consumers to decide for themselves whether these are valuable signals of quality, which is especially effective in today’s information sharing world.77 Other barriers to entry such as price inflation and land use restriction in opportunity-rich markets also create a sticky internal labour market, making it difficult to correct for regional labour market issues. Economists have estimated lost labour market gains (in the form of higher wages that could be earned), and found that the US GDP could be 13.5% larger if localities eased land-use restraints alone.78 Current workplace benefits (i.e. healthcare and retirement plans) are tied to jobs rather than to employees themselves, making it more difficult for workers to leave their jobs to seek better work opportunities. One in every six Americans currently live in an economically distressed community where job prospects continue to shrink79 and yet the percentage of Americans who move to a different state each year has fallen by half since the 1990s.80 This is problematic as it discourages agglomeration economies and prevents job seekers from securing available employment opportunities in other regions, which is especially significant in the face of growing regional inequalities. In addition, declining mobility rates contribute to greater inequality—mobility rates are especially low among the disadvantaged and the non-college educated, and specifically from distressed places to high-wage (but often higher housing cost) regions.81 Because states lack the incentive to address broader national labour market needs, policy change at the federal level is needed to encourage mobility without displacing the useful aspects of local control in a geographically large and economically complex country. While the TAA provides a relocation allowance, it is capped at $1250 and fewer than 1% of programme participants use it.82 A new federal policy could provide larger incentives to relocate, such as providing individuals beyond the TAA with more generous relocation subsidies, or providing tax incentives to states and cities that reduce entry constraints such as restrictive housing and occupational licensing policies. A portable system of employment benefits should also be created that ties benefits to individuals rather than to jobs, and should be expanded to include part-time and contingent workers to encourage a more flexible labour market.83 B. Canada The USA and Canada share many economic and social similarities. Historically one another’s largest trading partners, both countries have favoured a safety-net approach of LMAPs and employ a federal system to govern large and regionally diverse countries. However, they also maintain important differences in their governance, economies, financial, and labour markets. Canada’s economy was less affected by the 2008–9 financial crisis than most other OECD countries, and because Canada’s GDP is concentrated in natural resource-based sectors, the labour market performed well due to a boom in natural resource exports. On the other hand, Canada’s unemployment rate, which was 5.9% in November 2017, remains higher than that of the USA and the more recent fall in global commodity prices has led to lower rates of economic growth.84 In contrast to the USA, Canada has an explicit constitutional protection of personal mobility rights including the right to pursue the gaining of a livelihood in any province.85 Therefore, laws that discriminate against out-of-province residents may be classified as a breach of their mobility rights.86 However, Canadian courts have held that the right to pursue the gaining of livelihood in any province does not create a right to work, nor does it include the right for workers to establish themselves professionally anywhere in the country regardless of qualifications.87 Internal labour mobility rates are around the same as in the USA, if not slightly lower. In 2011, the interprovincial migration rate was 0.7% in Canada and 1.6% in the USA, while the intra-provincial migration rate was 3.4% in Canada compared with 2% in the USA.88 Significant disparities in regional unemployment rates continue to persist, suggesting that barriers to mobility remain. The Canadian government has taken steps to reduce these barriers including promoting interprovincial agreements that enable workers certified in one province/territory to be certified in another one.89 Structural changes are likely to continue in Canada with a decline in manufacturing jobs and a move towards service-producing industries and high-skilled jobs.90 Despite these structural changes, Canada has taken a different stance from the USA, embracing a continuing commitment to trade liberalization and liberal immigration policies rather than protectionism. In addition, while there has been political tension in federal–provincial relations over the administration of labour market training, the development of LMAPs has not been as politically charged as in the USA. Three distinctive features of Canadian policy will be examined. First, Canada is often regarded as ‘in between’ the less interventionist US-style regime and more interventionist European-style regimes,91 and has introduced more extensive active policy measures in recent years. Second, Canada has managed to pursue a period of intensified trade liberalization in the absence of a TAA programme. Third, the administration of ALMPs is decentralized under the federal system to the provinces and territories, which has resulted in flexible, regionally sensitive policies but a lack of a unified national policy framework. 1. The ‘middle ground’ As recalled from the previous section, the average public expenditure for OECD countries in 2015 (as a % of GDP) was 0.78 for PLMPs and 0.53 for ALMPs. Canada’s expenditure is relatively low by international standards, at 0.62 for PLMPs and 0.24 for ALMPs, but remains significantly higher than expenditures in the USA (0.18 and 0.10 respectively).92 Canada has a strong record of LMAPs and announced increased funding for LMAP interventions in its recent federal budgets. Previous LMAP experiments include TAA and wage insurance programmes, and in 2007 the federal government launched the Targeted Initiative for Older Workers (TIOW) that provides unemployed, older workers (ages 55–64 years) with group-based skills training.93 In 2017, the Ontario government launched a UBI pilot that provides almost $17,000 per year for a single person (in the form of cash payment with clawbacks based on earned income)94 and, recognizing the rise of non-standard employment, enacted equal pay for equal work provisions to ensure that part-time employees are paid proportionately the same as full-time employees if they are doing the same job.95 The backbone of Canada’s LMAP system and its principal safety net and workforce development programme is the Employment Insurance (EI) programme.96 The EI programme provides financial benefits as well as ALMPS known as Employment Benefits and Support Measures (EBSMs) that are funded through a payroll tax. While EI financial benefits are not as generous as those in Nordic countries, the coverage rates, recipiency rates, and benefit levels are higher than those in the USA.97 Regular financial benefits are available to displaced workers who have lost their jobs through no fault of their own, have been without work or pay for at least seven days, have accumulated between 420 and 700 hours of insurable employment in the 52 weeks prior to the claim, and are actively seeking employment during the entire claim period. Critics argue that the EI system has become weaker—tightened eligibility requirements and changes in the labour market such as increasing non-standard employment (i.e. self-employment and short-term jobs), have meant that the proportion of the unemployed eligible to receive EI benefits dropped from 83% in 1989 to 42% in 1997.98 Rates have not recovered since—in 2015–16, 43.8% of unemployed workers were eligible to receive regular EI benefits, with an average weekly benefit amount of $443.99 EBSMs are the second component of the EI programme and comprise four categories: targeted wage subsidies, skills development programmes, self-employment assistance for those who wish to start their own business, and job creation partnerships that provide short-term employment. While these programmes are federally funded through the EI fund, they are administered by the provinces and territories through federal-provincial Labour Market Development Agreements (LMDAs). A 2015–16 assessment found that 178,556 EI clients returned to employment following an intervention and calculated that the returns to employment were $1.15b (in the amount of unpaid EI financial benefits).100 However, while support measures such as counseling and resume writing workshops are available to all Canadians, employment benefits under the aforementioned categories are available only for EI-insured clients, meaning that only about 40% of the unemployed can access these services, and even fewer use them. Out of the 1.37m unemployed in 2012, only 18,815 claimants received training.101 As a result, the federal government entered into Labour Market Agreements (LMAs) with nine provinces and territories to fund skills development and training for non-eligible groups. These were renewed in 2014 as part of the Canada Job Fund Agreement and include the provision of up to $15,000 per person for training costs.102 In addition, there are relatively robust provincial programmes that provide financial incentives to employers who provide on-the-job training, especially in apprenticeships. For example, employers in Ontario can receive a $1000 taxable cash grant when hiring an apprentice, and wage subsidies and tax credits are given to employers who provide youth internships and training.103 However, most of these supports are in place for apprentices in skilled trades or for youth, and more could be done to target adults and older workers. 2. Intensified trade liberalization In contrast to recent US protectionist trends, Canada has embarked on a period of intensified trade liberalization under both Conservative and Liberal governments. In 2008, Canada only had four free trade agreements (FTAs) but has since brought into force nine more agreements and signed the Comprehensive and Progressive Agreement for Trans-Pacific Partnership (CPTPP) with 10 other countries in March 2018. However, there remain some trade adjustment concerns. The recently ratified Canada-European Union Comprehensive Economic and Trade Agreement (CETA) is estimated to increase Canada’s GDP by 0.77%, but it also provoked criticism that the expanded import quota for European cheese and enhanced IP rights in the pharmaceutical sector would hurt Canadian dairy farmers and pharmaceutical companies, respectively raising the question of whether the reintroduction of trade adjustment packages and compensation would be appropriate. In a persuasive paper, Lysenko and Schwartz104 argue that Canada does not need a TAA programme for several reasons that highlight the differences between the American and Canadian approaches to FTAs. First, a TAA programme is not necessary to overcome political opposition to trade agreements, whereas in the USA, the TAA was needed to ensure Congressional support for trade liberalization commitments.105 As explained below, consensus in Canada has instead been achieved through ad hoc federal compensation to different provinces and industries. Second, because Canada has already experienced substantial trade liberalization and has historically implemented trade liberalization commitments incrementally, adjustment costs from future FTAs will likely be small. Third, Canada has recognized that determining eligibility for a TAA programme is arbitrary in practice. While Canada had introduced a TAA programme in 1967, it was terminated in 1987 due to the practical difficulty of determining which workers had been injured by trade.106 The federal government’s ad hoc compensation policies warrant further comment. In recent years, Canadian trade policy has seen provinces become important political players that are willing to block FTAs given that certain trade issues fall under provincial jurisdiction. During the CETA negotiations, the federal government promised compensatory transfers to provinces as a way of reaching consensus where there was an issue of conflict. For example, Newfoundland was promised $280m in return for removing minimum local processing rules in its fishing industry and Ontario sought compensation for any negative effects of CETA, particularly for additional costs entailed for the pharmaceutical industry, cheese makers, winemakers, and distilleries.107 While this approach has been effective in addressing opposition, this assistance is provided primarily for political reasons and still implies an inequitable treatment of firms and workers in that some groups receive special benefits, but others who may be negatively harmed by FTAs receive none.108 A preferable policy approach would seek to aid all workers displaced by structural changes regardless of the cause of unemployment by providing strong ALMPs. 3. Decentralization of ALMPs As in the USA, the implementation of ALMPs in Canada is complicated by its federal system that has seen the federal government devolve labour market policy to the provinces and territories to mitigate political tensions. Because the Canadian constitution provides that the federal government is responsible for the economy while the provinces are responsible for education (thus training falls under both jurisdictions), disputes have arisen with regards to how training should be administered.109 These jurisdictional conflicts were resolved in 1996 when the Chrétien government devolved responsibility for training to provincial and territorial governments under the country’s EI scheme, thus paving the way for ‘collaborative federalism’,110 which has shaped labour market policy in Canada since. Under this scheme, the federal government transfers EI funds to provinces and territories to administer their own labour market policies to EI-clients under the previously mentioned LMDAs. The federal government retains control over the EI fund, delivers financial benefits, determines eligibility, and retains jurisdiction to develop and deliver ALMPs for non-EI clients (such as youth, persons with disabilities, Indigenous peoples, older workers, and recent immigrants). Decentralization has allowed for flexible and regionally sensitive policies and programmes that are adapted to local concerns. Despite the ‘administrative asymmetry’ that results from differing policies across the country, proponents argue that asymmetry is ‘built into federalism’ and is needed to address Canada’s regional diversity.111 However, because the LMDAs were negotiated individually with the provinces and territories, decentralization has developed without any coherent and unified policy framework for displaced workers.112 Critics have argued that this policy fragmentation has resulted in uneven programme quality across the country and federal programmes that are poorly integrated with provincial programmes. Bramwell suggests that because ALMPs are now de facto a provincial responsibility, Canada should decouple access to training programmes from EI eligibility and devolve responsibility and funding to provincial governments for training of displaced non-EI clients.113 C. Europe In contrast to the USA and Canada, which exhibit characteristics of the Anglo-Saxon model (i.e. relatively low employment protection and low unemployment benefits), Europe has seen the dominance of two models: a Nordic model and a continental model. While the Anglo-Saxon model emphasizes a ‘liberal activation’ framework or ‘defensive’ workfare (stronger work incentives, benefit conditionality, and the use of sanctions), the Nordic model emphasizes ‘universalistic activation’ or ‘offensive’ workfare such as human capital investment and upskilling.114 In past decades, the Nordic model has been held out as a model example for labour market policy, with strong ALMP supports for displaced workers and near universal unemployment benefits. While Nordic countries such as Sweden and Denmark have made major cuts in active measures and unemployment benefits in the past decade, they continue to spend the highest amount on ALMPs and PLMPs relative to other OECD countries. In contrast, continental Europe has placed more emphasis on employment protection than on actual labour-market participation.115 In this section, a brief discussion on intra-EU mobility will be followed by an analysis of the Nordic and continental models and the emerging trend towards defensive workfare. 1. Labour mobility within the EU As in the USA, the UK has seen a surge in protectionist rhetoric, as reflected when the UK voted to leave the EU with a 51.9% majority in the 2016 Brexit referendum, in which free movement of people within the EU was a central campaign issue. British–EU negotiations on Brexit began in June 2017 and have been strained over the issue of free movement within the EU, which has significant implications for intra-EU labour mobility. Under EU law, the principle of freedom of movement of workers is enshrined in Article 45 of the Treaty for the Functioning of the European Union (TFEU), which gives migrant workers who hold nationality in at least one EU Member State the right to work and reside freely within the EU, as well as freedom from direct and indirect discrimination based on nationality with regards to employment, remuneration, and other conditions of work and employment.116 Articles 20 and 21 of the TFEU further grant the right of EU citizenship to every person holding the nationality of a Member State, and gives all EU citizens the right to move and freely reside within the EU.117 The Brexit debate has brought freedom of movement and intra-EU mobility to the forefront, but in fact, despite the EU’s expansive legal framework, EU workers are not taking extensive advantage of EU mobility rights, and mobility rates have remained low. This has been a concern for EU policymakers because there have been high disparities in unemployment rates across Europe since the financial crisis in 2008. For example, the 2016 unemployment rate was 19.7% in Spain, 10.1% in France, and 4.1% in Germany.118 In 2014, only around 0.4% of the EU population moved to work in a different EU country, which is a lower cross-border mobility rate than Canada’s interprovincial labour mobility rate of around 0.8% and the US’s interstate mobility rate of around 2.3% in that same year.119 In addition, movement is concentrated towards a small number of countries. The main countries of residence of EU-28 movers were Germany (2.7m), the UK (2.1m), Spain (1.4m), Italy (1.1m), France, and Switzerland (around 950K each)—together, these six countries accounted for 75% of EU-28 movers in 2014.120 While labour mobility accounts for approximately 50% of long-run adjustment to region-specific labour demand shocks in both Europe and the USA, this process has been shown to take twice as long in Europe (10 years) than in the USA (5 years).121 This is likely due in part to greater labour market rigidities typically found in continental and Southern European countries. Labour mobility is also a less important adjustment mechanism in Europe than in the USA, most likely due to cultural, linguistic, and institutional differences across Europe that act as barriers to labour mobility. While there has been a rise in the role of migration in Europe (and a corresponding fall in interstate migration in USA), more could be done to facilitate labour mobility across Europe. The Bologna Process has helped to facilitate portability between universities (through transferable credits), but other policies that harmonize education standards, improve language education, and promote portability of pension and social security rights would enhance labour mobility.122 However, these sorts of policies will likely engender resistance as Brexit negotiations continue, which thus far suggest that free movement between the UK and EU will end in 2019 once the UK formally leaves the EU. 2. Nordic model The Nordic model is upheld as a model example of LMAPs for good reason, as Nordic countries have tended to perform better on unemployment and employment measures. Three features will be highlighted. First, Nordic labour market policies are unique due to their heavy investments in ALMPs and the influence of the Ghent system, whereby unemployment benefits are administered by trade-union-linked funds under a voluntary scheme. Second, Nordic countries are known as the best examples of the ‘flexicurity’ policy scheme, which the European Commission (EC) has promoted in the EU since 2006. Third, while Nordic countries have maintained a distinctive social policy approach, developments and trends in the past two decades have pointed towards a convergence with the ‘work first’ principle that characterizes traditions in the USA and Canada.123 (i) Emphasis on ALMPs and the Ghent system Of the OECD countries, Denmark spends by far the highest amount on ALMPs with 2.05% of its GDP dedicated to ALMPs in 2015 (recalling that the average for OECD countries is 0.53%), and a substantial amount on PLMPs, at 1.28%. ALMPs emphasize early intervention, in which unemployed individuals must enter a regime of mandatory activities after one month that include interviews, counseling, and monitoring of active job seeking. After six to nine months, job seekers must take part in job training and labour market education.124 Thus, ALMPs have an upskilling effect but also a motivation effect. After Denmark, Sweden currently invests the most in ALMPs across OECD countries, spending 1.27% of its GDP on active programmes in 2015.125 ALMPs were in fact first pioneered in Sweden in the 1950s by two trade-union economists under the Rehn–Meidner model, which sought to improve the match between the supply and demand of labour and to sustain full employment in an evolving economy by focusing on vocational training programmes.126 Since their inception in the 1950s, ALMPs have been a central feature of Swedish economic policy and are widely accepted within Sweden as the best means of protecting its citizens from unemployment and poverty.127 In addition to a tradition of ALMPs, another unique aspect of the Nordic model is the use of the Ghent system. The Ghent system was introduced in Belgium in 1901, and consists of a voluntary UI system administered by private agencies linked to trade unions that are funded largely by the state.128 While this system was previously more prevalent than compulsory UI schemes (i.e. UI contributions that are automatically deducted from the payroll), today the Ghent system remains in place only in Denmark, Sweden, and Finland. Under the Ghent system, individuals choose to pay a membership fee to a UI fund in return for unemployment benefits. Although an individual does not need to be part of a trade union to be a member of a UI fund, there seems to be a positive impact of voluntary UI on the strength of trade unions, and trade union density in these three Nordic countries is distinctively high compared with other OECD countries. For example, 67.7% of Swedish employees were members of a trade union in 2013, and collective bargaining agreements cover 90% of employees in Sweden.129 Trade union density in Denmark was 66.8% in 2013 and 69% in Finland (compared with 10.8% in the USA and 27.1% in Canada). As a result, trade unions exert considerable influence over important labour market regulations and play a large role in ensuring that the labour market functions effectively. While voluntary UI schemes may theoretically lead to a problem of adverse selection, this has not been borne out in reality. The coverage rate is near universal in Denmark and Sweden, with 90% of the Danish workforce covered in 2014130 and 85% of the Swedish workforce,131 which are significantly higher than in OECD countries with compulsory UI schemes. This is likely because the low direct costs of membership and high unemployment benefits give a clear incentive for joining a voluntary UI system.132 The monthly membership fee is around 440 DKK (US$70) in Denmark and SEK 352 (US$43) in Sweden, and can be offset against personal taxes. Benefits are generous, with a benefit duration of two years in Denmark (it was lowered from four years to two years in 2010) up to a maximum of 90% of the unemployed worker’s previous income. In Sweden, members of UI funds can receive a benefit period of 300 insurance days that pays 80% of an unemployed worker’s previous income for the first 200 days and 70% of previous income for the remaining days.133 For those who do not fulfill the membership condition, a basic insurance amount (amounting to 320 SEK per day, or approximately US$40) is provided under certain conditions.134 In addition, it is important to note that, unlike in Canada and the USA, the function of UI is not to provide a social safety net, but rather to act as a work incentive, as individuals must participate in ALMPs and be prepared to resettle in regions with labour demand. Thus, UI is seen as compensation for the activity of job seeking rather than for unemployment itself,135 and passive and active LMAPs are closely linked. (ii) Flexicurity ‘Flexicurity’ is a welfare state model that came to the forefront in the mid-2000s when it was promoted by the EC as its flagship policy, and emphasizes flexibility in the labour market (i.e. relaxed hiring and firing rules) combined with security (i.e. income and employment security in the form of generous unemployment benefits and ALMPs).136 In this model, job security and employment security are two distinct concepts. Whereas job security refers to a job with a single employer, employment security refers to the potential to hold jobs with multiple employers (i.e. employability).137 Flexicurity thus emphasizes low job security but high employment security by providing a menu of passive and active LMAPs including social protection (through a social safety net), training, and education. Flexible labour markets have several advantages—they increase responsiveness of wages, promote labour mobility across sectors, and allow for firms to be more efficient and competitive in response to economic changes. While the flexicurity model was first developed as a labour market policy in the Nordic countries, most notably in the Netherlands and in Denmark, flexicurity has subtly different meanings in the Netherlands, in Denmark, and for the EC. The term ‘flexicurity’ was first used in the Netherlands during discussions that resulted in the Flexibility and Security Act of 1999, which promoted atypical and flexible forms of employment while also providing for protection of workers on such contracts. These laws aimed for the ‘normalization of atypical work while preserving flexibility in the labour market’,138 and have provided security to non-standard workers in the last two decades by applying the pro rata temporis principle to part-time workers not just to wage and working conditions, but also to social security entitlements.139 As a result, contractual diversity in terms of part-time, full-time, and agency work is very high. In 2006, 46% of the employed Dutch labour force worked part-time (compared with 18% in EU-27), and part-time work is regarded as a normal and desirable type of employment, as indicated by the high number of part-time workers. These types of policies will likely become increasingly important in other countries experiencing sharp increases in non-standard employment. Whereas the Dutch model views ‘flexibility’ within the flexicurity model as flexible types of employment (i.e. atypical and non-standard employment), the Danish model defines ‘flexibility’ as ease of job mobility. Today, flexicurity has become synonymous with Danish flexicurity, often referred to as the ‘golden triangle’—it combines flexible employment (high rates of job mobility), a strong social safety net (high income security), and comprehensive ALMPs (high employment security).140 Denmark’s flexicurity framework evolved following the 1980s, during which time the country was experiencing a long period of high unemployment, with an 8.16% unemployment rate throughout that decade.141 To address this problem, Denmark undertook a series of labour market reforms in the 1990s that emphasized activation policies and implemented its flexicurity model. These labour market reforms were described as a ‘job miracle’, with unemployment falling from 11% in 1993 to 3.4% in 2008142 (unemployment rates increased to 7.5% in the years following the financial crisis and was 6.2% in 2016).143 Job mobility within Denmark is high by international comparisons, most likely due to a low level of job protection (consistent with its flexicurity policy), combined with a willingness among workers to move given Denmark’s strong income and employment security. Due to Denmark and the Netherlands’ impressive labour market performance, the EC has promoted flexicurity as a strategy to enhance flexibility and security in the labour market (consistent with the Danish model) and to address the problem of atypical employment (in line with the Dutch model). Whereas US-style flexibility assumes that social welfare will be increased by the benefits of flexible labour markets through the ‘trickle down’ effect, flexicurity assumes the need for policies to address income and employment security issues and provides both a safety net and a trampoline for workers moving through the labour market.144 While flexicurity is a rather vague concept, Peter Auer, formerly of the International Labour Organization, used a clustering analysis to define ‘flexicurity’ versus ‘non-flexicurity’ countries using variables as proxies for flexibility and security (i.e. working time flexibility, average employment tenure, and strictness of employment protection legislation).145 Consistent with the differences between the Nordic and continental models, Denmark, the Netherlands, Sweden, Finland, and Austria were classified as ‘flexicurity’ countries while France, Belgium, Germany, Italy, Spain, Greece, and Portugal were classified as ‘non-flexicurity’. Ireland and the UK were found to have high flexibility but low employment security, consistent with the Anglo-Saxon model. Auer found that the flexicurity countries performed better than the non-flexicurity countries on unemployment and employment rates, although flexibility countries also had the highest increase in unemployment during the 2008–2009 crisis period.146 However, strong causality between flexicurity and labour market performance has not yet been firmly established and major challenges arise in attempting to transplant flexicurity polices into countries with different political traditions and institutions. (iii) Towards activation Sweden first developed ALMPs in the 1950s, and since then all OECD countries have developed some measure of ALMPs.147 Within the menu of ALMP options available, different policies have been emphasized during different decades, and have been generally grouped into three periods. The first period was marked by skill market shortages in the 1950–60s, and ALMP use in this period was dominated by upskilling to provide skilled workers to meet this labour demand. The second period followed the oil shock of 1973–75, which saw high rates of unemployment across Europe. Many countries began using occupation ALMPs (i.e. providing temporary work in the public sector) with the goal of keeping workers engaged to slow the deterioration of human capital linked with unemployment.148 In the mid-1990s, the OECD and EC advocated a new ‘activation’ paradigm that focused on encouraging and facilitating the re-entry of unemployed individuals into the labour market, thus marking the third and most recent period of ALMP use. The activation paradigm acknowledged that there were interactions between UI systems and ALMPs, such that benefit conditionality makes a difference in terms of labour market adjustment programmes effectiveness.149 Thus, this mixes ‘offensive’ workfare, which emphasizes human capital investment, with ‘defensive’ workfare’, which emphasizes work incentives and benefit conditionality. Denmark became a pioneer of activation ALMPs and other countries such as Sweden, the UK, and Germany soon followed suit, and since the late 1990s, emphasis has been placed on more incentive reinforcement and employment assistance measures (i.e. counseling, placement services, and job-search programmes).150 This reorientation of the activation paradigm has provided some evidence of a trend towards a degree of convergence in labour market policies between North American and Nordic models. The Anglo-Saxon framework, particularly in the USA and Canada, has traditionally focused on a ‘work first’ principle consistent with liberal traditions in North America, in that benefit levels are lower, more restricted, and subject to conditions.151 While Nordic countries such as Sweden and Denmark are still noted welfare leaders, both countries have made major recent cuts in its benefits and instituted stricter work obligations, consistent with a converging trend towards activation policies and defensive workfare.152 For example, Denmark implemented several reforms to its UI system in the mid to late 1990s such as a shorter benefits duration period, strengthened eligibility rules, and mandatory participation in training. In Sweden, it was previously possible to requalify for UI benefits by enrolling in a training programme. However, this was criticized for generating a perpetual cycle between UI recipiency and training programmes, and new rules in 2001 deemed that this requalification could only take place once.153 In 2006, Sweden reformed its UI system once again, tightening eligibility conditions, lowering benefit levels, and raising monthly UI contributions more than three-fold, from SEK 100 (US$12) to SEK 352 (US$43).154 In addition to the strengthening of work incentives for benefit recipients, ALMP emphasis has shifted to employment assistance (i.e. job subsidies, counseling, and job search programmes) and incentive reinforcement (i.e. tax credits, benefit reductions, and benefit conditionality) accompanied by significant expenditure cuts to training ALMPs.155 These measures create stronger incentives and pressure on the unemployed to take up employment, and are consistent with the ‘liberal’ type of activation or defensive workfare strategy, thus pointing towards a growing convergence of policy objectives and instruments.156 Nonetheless, it can be argued that despite these policy changes, the character of the Nordic welfare state and labour market policies remain distinct, maintaining relatively high investments in ALMPs and PLMPS as well as relatively more emphasis on equality and universality of benefits coverage. 3. Continental model The continental model has traditionally been focused on strong job protection through robust employment protection laws, combined with relatively generous compulsory UI schemes and a limited emphasis on ALMPs. This model has often prevented countries like France and Germany from combating high unemployment rates, but reforms since the early 2000s have changed the character of the continental model. First, the newly elected French President, Emmanuel Macron, has promised sweeping reforms of a French labour market that has been plagued by a stubborn unemployment rate of between 8% and 10% over the last decade. Second, Germany instituted major reforms between 2003 and 2005 that have given rise to a so-called ‘German model’. Third, both of these reforms also make use of activation factors, pointing to a convergent trend towards defensive workfare. (i) Labour market rigidity France remains one of the best-known examples of labour market rigidity due to its stringent job protection laws. Economists have frequently argued that this creates an insider–outsider problem, since it protects privileged ‘insiders’ (permanent and full-time employed workers) at the expense of ‘outsiders’ (part-time workers and the unemployed) who are excluded from active or full participation in the labour market. Such a dual labour market can undermine human capital development, economic growth, and overall welfare. This problem has been well documented in France, where labour codes make dismissal procedures costly and lengthy, thus creating rigid labour markets that discourage employers from hiring permanent staff and also make it difficult for the economy to adjust to shocks. In addition, since France’s 35-hour workweek was approved by the socialist government in 2000 (down from 39 hours, after which time overtime pay is required), employers have argued that this has pushed up labour costs and made France less competitive.157 Previous French presidents have attempted to liberalize the French labour market in a bid to reduce the country’s high unemployment rate, but have been met with intense resistance from trade unions. President Macron came to power in June 2017 on a campaign promise to implement sweeping labour market reforms to ease labour regulations, create a more flexible labour market, and create jobs. Proposed reforms included capping the compensation for employees who are laid off for no justifiable reason, reducing the period during which employees can commence legal challenges after being made redundant, and making it easier for companies to fire employees for ‘economic reasons’.158 In September 2017, Macron signed decrees implementing some of these reforms, thus pushing France closer to the Danish ‘flexicurity’ model by giving employers greater discretion in firing workers and more confidence in hiring. However, Danish reformers also had the fiscal means to implement employment security measures such as extensive training programmes, and it remains to be seen whether Macron’s stimulus plan of 50 billion Euros (which accounts for about 2% of France’s GDP) will provide sufficient assistance for workers.159 (ii) Rise of the ‘German model’ While France is currently attempting to deregulate its labour market, Germany moved away from a rigid labour market more than one decade ago. In the 1990s, Germany was dubbed the ‘sick man of Europe’, unable to shake its high unemployment and weak economic growth, which were due to high levels of job protection, high labour costs, strict regulation of labour market, and ineffective LMAPs.160 PLMPs consisted of an unlimited payment duration of UI benefits and the highest replacement rates out of OECD countries for the long-term unemployed, thus providing weak incentives for the unemployed to take up new jobs. In terms of ALMPs, job creation programmes were the main tools used, with little emphasis on job search assistance and monitoring, and no system in place to assess programme effectiveness. From 2003 to 2005, Germany underwent a significant labour market policy overhaul, known as the Hartz reforms. Implemented in four waves during those years, the Hartz reforms sought to achieve three objectives: (i) restructure the federal employment agency; (ii) activate the unemployed by significantly reducing long-term unemployment benefits both in amount and duration; and (iii) stimulate labour demand by deregulating the labour market.161 The government also specifically tied the reforms to a programme evaluation mandate to analyse the effectiveness of ALMP programmes. After these reforms, the unemployment rate declined steadily, falling from 11.3% in 2005 to 4.1% in 2016,162 earning Germany the title of the ‘North Star’ of labour market policies and reforms.163 This success has largely been attributed to the increased flexibility of the German labour market that has allowed small and medium businesses to fire more easily, thus lowering the risk of hiring workers. In addition, studies have consistently shown that cuts in unemployment benefits significantly reduced unemployment.164 However, while Germany has undoubtedly been a success story in reducing unemployment, wage and income inequality have risen in the same period along with an increase in atypical forms of employment.165 Today, Germany uses a compulsory UI system with a one-year benefit duration (or 18 months for those over the age of 55 years). Replacement rates are capped at 60% of previous earnings, and the claimant must be looking for and available to work.166 In terms of ALMPs, Germany functions similarly to Canada by allocating a lump sum of money to regional governments and allowing local offices to decide how to budget between different ALMP measures. Prior to the Hartz reforms, these local offices did not provide training measures themselves, but rather contracted them out to private providers using a tendering process.167 However, competition was limited because in reality the contracts were awarded on the basis of familiarity with the provider. The Hartz reforms introduced quasi-market mechanisms through a system of training vouchers with the goal of increasing transparency, consumer choice, and competition.168 While these goals have been accomplished to some degree, clients often lack sufficient information to make informed choices, and competition has been attenuated because of an insufficient number of providers. As a result, it has been argued that a decentralized contracting-out system with an emphasis on performance measurement would better achieve those aims.169 While strictly outside of the ambit of our basic geographic comparative frame of reference, Australia is worthy of mention as it provides a similar example to Germany’s training voucher system, and is often cited as a model for market forces in employment training and placement.170 While other countries such as the Netherlands have also privatized their employment reintegration services, Australia is unique among OECD countries as its employment services are delivered by over 100 providers in a competitive quasi-market financed by service fees, employment outcome payments, and a special fund.171 Australia was one of the first OECD countries to use a quasi-market framework in 1998 when it began delivering employment services through contracts (using a tendering process) with community providers. The underlying assumption is that contracting out services will create competition to foster greater efficiency, better quality of service, and reduce costs.172 While contracting out raises concerns that providers might seek to ‘cream-skim’ easier-to-place workers, this problem is addressed by scaling outcome fees such that providers receive higher fees for individuals that are more difficult to place.173 In addition, the government publishes Star Ratings, which measures provider performance based on how many paid placements and paid outcomes were achieved. These ratings inform participants and are also used to eliminate poorly performing providers in subsequent tendering rounds. In 2009, further reforms through the Job Services Australia model emphasized employment outcomes by making providers responsible for organizing work-experience activities and enforcing a minimum hours of participation in its programmes. Since 2009, Australia has experienced some of the highest employment rates (i.e. the extent to which people available to work are being employed) among G7 and OECD G20 countries, in part due to effective and structured monitoring of its labour market programmes and its quasi-market delivery regime.174 IV. LESSONS FROM THE COMPARATIVE EXPERIENCE LMAPs have played an important role in the political economy, often helping to mitigate political opposition to technological innovation and trade openness. However, the effectiveness of these policies has been met with mixed degrees of success. It would be naïve to assume that some ideal model of LMAPs that has emerged in a particular country can be transplanted effectively to other countries with very different economies, geographies, histories, labour relations, and political systems. For example, the flexicurity model operates within broader ‘social-democratic’ welfare provisions that emphasize universal and portable benefits, which would be more difficult to implement in the context of the American ‘liberal’ political economy. In addition, research has suggested that differing degrees of trust and confidence between institutional actors (i.e. business, labour, and government) affect the success of adjustment programmes.175 The history of efforts to promote institutional reform in developing countries over the past twenty or thirty years by external donors promoting transplantation of ‘best practices’ from other jurisdictions, which efforts have met with mixed to poor results, counsels against such an assumption.176 Despite this important caveat against uncritical transplantation of policies and institutions from one country to another, perhaps the most striking conclusion to emerge from our comparative review of LMAPs in the previous section is a marked convergence towards some general principles, while recognizing that important differences in these policies are likely to persist from one country to another. We believe that comparative experience suggests that two opposing or outlier labour market policies have not well withstood the test of time and experience. First, the austere policy model for responses to job displacement that has historically prevailed in the USA, i.e., PLMPs principally exemplified by very limited forms of UI in terms of income coverage and length of benefits, and minimally supported by ALMPs, i.e., job counseling, job search, and job training (albeit with more extensive benefits for workers laid off as a result of international trade) has left many displaced workers bearing an unacceptable share of the burden of the costs of displacement. This is inequitable in terms of the disproportionate burdens that such workers bear, inefficient in under-utilizing potentially productive human capital, and politically perverse in intensifying opposition to trade, technology, and other factors affecting the demand for labour in developed economies. Second, the opposing model, most strongly exemplified by that which historically prevailed in France, entails extremely high levels of job protection and, in the event of job displacement, high levels of largely unconditional UI benefits in terms of duration and income replacement. These policies have tended to create a dual labour market that exhibits sharply different treatment between ‘insiders’ (workers with full-time employment) and ‘outsiders’ (workers with part-time employment or job-seekers who have been unable to secure any employment at all). The so-called Nordic model of labour market policies, exemplified by policies that have historically prevailed in countries such as Sweden, Denmark, and the Netherlands, is characterized by generous PLMPs, in particular generous UI benefits in terms of duration and income replacement, complemented by generous ALMPs, and has often been held out as a compelling alternative to either the USA or French models. However, in recent years even these policies have undergone significant modification, with a scaling back of passive labour market benefits, and increased reliance on conditionality attaching to such benefits that require displaced workers to participate successively in job search assistance programmes and combinations of classroom and on-the-job training in the event of continuing unemployment. Germany’s labour market policies have also increasingly converged on this model, with a readjustment in the balance between ‘defensive workfare’ and ‘offensive workfare-oriented’ policies. A striking feature of this increasing convergence in policies is an emphasis on high employment security but relatively low job security, or as it is sometimes called, ‘flexicurity’. These policies have sought to provide displaced workers with the knowledge that despite flexible hiring and firing practices, there is a robust system consisting of both a safety net and a trampoline in place, i.e. a system to provide generous income support and training programmes to displaced workers. Flexicurity in the Netherlands has also sought to normalize atypical work while preserving flexibility in labour markets and have provided security to non-standard workers by applying the pro-rata temporis principle to part-time workers not just to wage and working conditions but also to social security entitlements, resulting in high levels of contractual diversity in terms of part-time, full-time, and agency work.177 As we noted in the prior section, in 2006, 46% of the employed Dutch labour force worked part-time and part-time work is regarded as a normal and desirable type of employment. In such a model, receipt of UI benefits would be conditional, after a relatively short period of receipt, on participation in a job counselling and search programme and in the event that this proves unsuccessful in identifying alternative employment within a relatively short period, receipt of further UI benefits would be conditional on participation in a classroom or on-the-job training programme (or a combination of the two). However, participation in these two programmes would not be restricted to recipients of UI benefits but would be available to other unemployed workers who do not qualify for unemployment insurance, recognizing that in many countries a relatively large (and increasing) percentage of unemployed workers do not qualify for unemployment insurance.178 A related concern is that unemployed workers who do not qualify for UI benefits may not be able to support themselves or their families while in job assistance programmes. While Canada, the USA, and most countries in Europe have highly developed financial assistance schemes for higher education or training, most of these programmes are limited to students and young adults. Thus, there is a need for a comprehensive income-contingent loan programme to support older adults who would otherwise be unable to participate in job assistance programmes. Comparative experience, particularly recent German and Australian experience, strongly suggests that the delivery of these programmes should not be an exclusive government monopoly but be provided on a competitive basis by a broad range of public, non-profit, and for-profit providers, with substantial scope for choice by job seekers among competing service providers, with such choices informed by appropriate prior job counselling services. Such a diversity of providers accords recognition to differences in employment prospects by sector, region, and characteristics of job seekers. Some form of government accreditation of service providers is likely desirable to weed out clearly fraudulent or incompetent providers and to establish some basic conditions that would need to be met to qualify for full or partial reimbursement of fees from government sources (in effect, a kind of voucher programme). These conditions would establish on the public record providers’ fee schedules and reimbursement rates, and would require periodic public disclosure of specified performance metrics, e.g., completion rates and post-completion employment experience of participants, which would be subject to periodic audits for accuracy. While out-of-pocket expenses incurred by participants would be largely borne by government, living expenses for participants for the duration of longer programmes (e.g. remedial education or long-term vocational training), might be financed or underwritten by government on an income-contingent loan basis. In the delivery of such programmes, there are many virtues in geographically large and regionally diverse countries, such as the USA, Canada, and Australia (which are all federal states), for delivery to be decentralized to subnational or local levels of government that are likely to be better attuned to local labour market needs than central governments. However, there is an off-setting risk that partial or parochial perspectives may excessively influence local decision-making, particularly when realistic future job opportunities may entail job seekers moving out of the jurisdiction, hence justifying a role for central governments in attaching conditions to financial assistance, or negotiating agreements with subnational levels of government, governing the basic contours of the services provided to ensure consistency with national labour market needs. With respect to the financing of these programmes, there is a case for moving away from primary reliance on UI premiums or payroll taxes (given the declining percentage of the unemployed receiving UI benefits) and greater reliance on some form of broad-based national consumption tax, with refundable tax credits for low-income earners, given that consumers are the principal beneficiaries of lower product and service costs from trade and technology. There is a clear role for governments at all levels in more systematically collecting (and frequently updating) intelligence on short- to medium-term forecasted employment needs of a representative range of employers by sector and region which would be a key input into the services provided by decentralized service providers (and accredited by government) and to choices by individual job seekers among programme providers. While local government agencies might continue to administer unemployment insurance, they would largely be confined to the role of an information-clearing house on the availability of other service providers in the public, non-profit, and for-profit sectors. Moreover, a striking lesson that emerges from our comparative survey is the importance of having governments commit to systematic, periodic independent programme evaluations for cost-effectiveness, which have often been haphazard and crude in the past, with programmes sustained more by faith than facts.179 Finally, investing in education and training increases the competitiveness of the economy by allowing workers to better respond to economic change. As the Independent Task Force on the Future of the US Workforce has put it, a change in thinking is needed such that education and work are viewed as two closely linked activities rather than as separate and distinct.180 Companies can provide some of this training, but it is difficult for smaller firms to support retraining and government help may be necessary for countries to maintain a competitive edge. While strictly outside the remit of this article, we are attracted to a model based on the Singapore Skills Future Initiative that would allocate to every individual of working age a modest annual voucher (of, e.g., $500) to be spent on any of a wide array of accredited training courses, thus signifying a collective commitment to lifelong learning, or what Yuval Levin calls in a recent book, ‘a mobility agenda’.181 This commitment will become especially important as trade and technology continue to change the labour market landscape, rendering well-designed and targeted LMAPs a high priority and a superior response to protectionist policies in addressing today’s costs from job displacement and economic transitions. Funding Project was funded out of a general University of Toronto research grant. Footnotes 1 ‘The New Political Divide’, The Economist, 30 July 2016, www.economist.com (visited 12 May 2018). 2 Michael Trebilcock, Dealing with Losers: The Political Economy of Policy Transitions (Toronto: Oxford University Press, 2014) at chapter 5. 3 See Michael Trebilcock, Advanced Introduction to International Trade Law (UK: Edward Elgar Publishing, 2015) at chapter 8. 4 For further discussion on the employment effects of trade, see ILO, ‘Trade and Employment: From Myths to Facts’, 4 October 2011, http://www.ilo.org/employment/areas/trade-and-employment/WCMS_162297/lang--en/index.htm (visited 12 May 2018). 5 Douglas Irwin, Free Trade Under Fire, 4th ed (Princeton: Princeton University Press, 2015) at 115 [Irwin]. 6 WTO, ‘World Trade Report 2017: Trade, Technology and Jobs’ (Geneva: World Trade Organization, 2017), https://www.wto.org/english/res_e/publications_e/wtr17_e.htm (visited 12 May 2018), at 109 [WTO Report 2017]. 7 David Autor, David Dorn and Gordon Hanson, ‘The China Shock: Learning from Labour Market Adjustment to Large Changes in Trade’, Annual Review of Economics (2016), 205–40. 8 Gary Hufbauer and Sean Lowry, ‘US Tire Tariffs: Saving Few Jobs at High Cost’, Peterson Institute for International Economics Policy Brief No. PB 12-9 (2012). 9 Jagdish Bhagwati and Robert Hudec (eds), Fair Trade and Harmonization: Prerequisites for Free Trade? (Cambridge: MIT Press, 1996) vol. 2. 10 Irwin, above n 5, at 162. 11 Mark Wu, ‘The China Inc. Challenge to Global Trade Governance’, 57 Harvard International Law Journal (2016), 261. 12 See Michael Trebilcock, Advanced Introduction to International Trade (Cheltenham, UK/Northampton, MA: Edward Elgar, 2015), chapter 6. 13 See Peter Singer, One World Now: The Ethics of Globalization (New Haven: Yale University Press, 2016), at 107–8. 14 Ibid, at chapter 2. 15 See Justin Lin, The Quest for Prosperity: How Developing Economies Can Take Off (Princeton: Princeton University Press, 2012) and Joseph Stiglitz, Globalization and Its Discontents Revisited: Anti-Globalization in the Era of Trump (New York: WW Norton, 2017). 16 For discussion on some of these rationales, see Ronald Daniels and Michael Trebilcock, Rethinking the Welfare State (Oxon: Routledge, 2005), chapter 10 [Daniels]. 17 Ibid. 18 For further discussion on policies needed to address the rise of non-standard forms of employment, see ILO, ‘Non-standard employment around the world: Understanding challenges, shaping prospects’ (2016), http://www.ilo.org/global/publications/books/WCMS_534326/lang--en/index.htm (visited 12 May 2018). 19 For reviews of many of these issues see, e.g.: Anthony Atkinson, Inequality: What Can Be Done? (Cambridge: Harvard University Press, 2015); Richard Reeves, Dream Hoarders: How the American Upper Middle Class is Leaving Everyone Else in the Dust, Why That is a Problem, and What To Do About It (Washington: Brookings Institution Press, 2017). 20 Daniels, above n 16, at 203. 21 Werner Eichhorst and Regina Konle-Seidl, ‘Evaluating Labour Market Policy’, 3 Institute for the Study of Labor (2016). 22 Robert S. Rycroft, The American Middle Class: An Economic Encyclopedia of Progress and Poverty (Santa Barbara: ABC CLIO, 2017), at 314. 23 Daniels, above n 16, at 203. 24 Giuliano Bonoli, ‘The Political Economy of Active Labor-Market Policy’ 38(4) Politics and Society 435 (2010), at 436 [Bonoli]. 25 OECD, OECD Employment Outlook 2015 (Paris: OECD, 2015), https://www.oecd-ilibrary.org/employment/oecd-employment-outlook-2015_empl_outlook-2015-en (visited 12 May 2018), at 107 [OECD Employment Outlook 2015]. 26 John P. Martin, ‘Activation and Active Labour Market Policies in OECD Countries: Stylized Facts and Evidence on their Effectiveness’, IZA Policy Paper No. 84 (2014) [Martin]. 27 Bonoli, above n 24, at 5–6. 28 Three of these categories (incentive reinforcement, employment assistance, and upskilling) are drawn from Bonoli, above n 24. 29 OECD, Activation Strategies for Stronger and More Inclusive Labour Markets in G20 Countries (Paris: OECD, 2013), 8 OECD, https://www.oecd.org/els/emp/G20-2013ReportActivation.pdfpdf (visited 12 May 2018). 30 OECD Employment Outlook 2015, above n 25, at 140. 31 Daniels, above n 16, at 206. 32 David Schleicher, ‘Stuck! The Law and Economics of Residential Stability’, Yale Law Journal (2017), 127 [Schleicher]. 33 Schleicher, above n 32, at 15. 34 Olivier J. Blanchard, Florence Jaumotte and Prakash Loungani, ‘Labor Market Policies and IMF Advice in Advanced Economies During the Great Recession’, 3(2) IZA Journal of Labor Policy (2014), at 5. 35 Gosta Esping-Andersen, The Three Worlds of Welfare Capitalism (Cambridge: Polity Press, 1990). 36 United States Bureau of Labour Statistics, https://www.bls.gov/ (visited 12 May 2018). 37 For further discussion on inequality and policies to tackle poverty, see Anthony Atkinson, Inequality: What Can Be Done? (Cambridge, MA: Harvard University Press, 2015). 38 Lawrence Mishel, Elise Gould and Josh Bivens, ‘Wage Stagnation in Nine Charts’, Economic Policy Institute (2015), http://www.epi.org/publication/charting-wage-stagnation/ (visited 12 May 2018). 39 Stijn Broecke, Shruti Singh and Paul Swaim, ‘The State of the North American Labour Market’ (Report prepared for the 2016 North American Leaders Summit, 2016) (Paris: OECD Directorate for Employment, Labour and Social Affairs, 2016), at 57 [Broecke]. 40 For discussion regarding the rise of corporate training programs, see Josh Bersin, ‘Spending on Corporate Training Soars: Employee Capabilities Now a Priority’, Forbes, 4 February 2014, https://www.forbes.com/sites/joshbersin/2014/02/04/the-recovery-arrives-corporate-training-spend-skyrockets (visited 12 May 2018). 41 Gregg M. Olsen, ‘Labour Market Policy in the United States, Canada and Sweden: Addressing the Issue of Convergence’, 42(4) Social Policy and Administration 323 (2008), at 326 [Olsen]. 42 OECD, OECD Employment Outlook 2017 (Paris: OECD, 2017), http://www.oecd.org/els/oecd-employment-outlook-19991266.htm (visited 12 May 2018), at 216 [OECD Employment Outlook 2017]. 43 Ibid, at 218. 44 Olsen, above n 41, at 329. 45 ‘Unemployment Insurance Chartbook’, United States Department of Labor, https://oui.doleta.gov/unemploy/chartbook.asp (visited 12 May 2018). 46 US Replacement Rates taken from US Department of Labor: the replacement ratio was 0.399, and represented the ratio of a weighted average weekly benefit amount (WBA)/weighted average of (normal hourly wage × 40 hours), https://oui.doleta.gov/unemploy/ui_replacement_rates.asp (visited 12 May 2018). 47 Judith M. Conti, ‘Unemployment Insurance: An Overview of the Challenges and Strengths of Today’s System’ (Hearing before the United States Congress Human Resources Subcommittee, 7 September 2016) (2016) National Employment Law Project, https://waysandmeans.house.gov/wp-content/uploads/2016/09/20160907HR-Testimony-Conti.pdf (visited 12 May 2018), at 2. 48 Carrick et al., ‘The Skills Gaps in U.S. Manufacturing 2015 and Beyond’, The Manufacturing Institute and Deloitte LLP (2015), at 3. 49 The other obvious answer is to raise wages, although that can disrupt internal wage structures. 50 Ruth Graham, ‘The Retraining Paradox’, The New York Times, 23 February 2017, https://www.nytimes.com/2017/02/23/magazine/retraining-jobs-unemployment.html [Graham] (visited 12 May 2018). 51 Helena Worthen, ‘The Workforce Investment Act and the Labor Movement’, 6(3) WorkingUSA 70 (2002–3), at 72. 52 David E. Balducchi, Randall W. Eberts and Christopher J. O'Leary, ‘Experience and Prospects for Labor Exchange Policy’, in David E. Balducchi, Randall W. Eberts and Christopher J. O'Leary (eds), Labor Exchange Policy in the United States (Kalamazoo: W E Upjohn Institute, 2004), at 252. 53 Lawrence F. Katz, ‘America’s Jobs Challenges and the Continuing Role of the US Department of Labor’, 67 Industrial and Labour Relations Review 578 (2014), at 581. 54 Graham, above n 50. 55 US Government Accountability Office, Report to Congressional Requesters, ‘Multiple Employment and Training Programs: Providing Information on Colocating Services and Consolidating Administrative Structures Could Promote Efficiencies’ (2011), http://www.gao.gov/products/GAO-11-92 (visited 12 May 2018), at 11. 56 ‘WIOA Overview’, United States Department of Labor, https://www.doleta.gov/WIOA/Overview.cfm [WIOA Overview] (visited 12 May 2018). 57 WIOA Overview, above n 56. 58 Benjamin Collins, ‘Trade Adjustments Assistance for Workers and the TAA Reauthorization Act of 2015’ (CRS Report Prepared for the Members and Committees of Congress, 14 September 2016) (Washington: Congressional Research Service, 2016), at 3 [Collins]. 59 Dmitry Lysenko and Saul Schwartz, ‘Does Canada Need Trade Adjustment Assistance?’, 57 IRPP Study (2015), at 4 [Lysenko]. 60 Irwin, above n 5, at 146. 61 Edward Alden, Failure to Adjust: How Americans Got Left Behind in the Global Economy (Lanham: Rowman & Littlefield, 2017), at 120 [Alden]. 62 Collins, above n 58, at 9. 63 Ibid, at 8. 64 Sarah Dolfin and Peter Z. Schochet, The Benefits and Costs of the Trade Adjustment Assistance (TAA) Program under the 2002 Amendments (Princeton, NJ: Mathematica Policy Research, 2012), at 58. 65 Ronald Daniels and Michael Trebilcock, ‘A Better Way to Help US Victims of Free Trade’, Bloomberg (2016), https://www.bloomberg.com/view/articles/2016-12-02/a-better-way-to-help-u-s-victims-of-free-trade [Bloomberg] (visited 12 May 2018). 66 Alden, above n 61, at 113. 67 ‘America’s Programme to Help Trade’s Losers Needs Fixing’, The Economist, 1 July 2017, at 67, citing David Autor, David Dorn and Gordon Hanson, ‘The China Syndrome: Local Labor Market Effects of Import Competition in the United States’, 103(6) American Economic Review 2121 (2013), at 2148. 68 Ronald D’Amico and Peter Z. Schochet, ‘The Evaluation of the Trade Adjustment Assistance Program: A Synthesis of Major Findings’ (Final Report prepared as part of the Evaluation of Trade Adjustment Assistance Program for the US Department of Labor Employment and Training Administration Office of Policy Development and Research, December 2012) (Washington: US Department of Labor Employment and Training Administration, 2012), at iv. 69 Irwin, above n 5, at 150. 70 Recipiency rates pulled from ‘Unemployment Insurance Chartbook’, United States Department of Labor, https://oui.doleta.gov/unemploy/chartbook.asp (visited 12 May 2018); income replacement based on a weighted average of: weekly benefit amount / (normal hourly wage × 40 hours) in the year 2016 from ‘UI Replacement Rates Report’, United States Department of Labor, https://oui.doleta.gov/unemploy/ui_replacement_rates.asp (visited 12 May 2018). 71 Collins, above n 58, at 8. 72 Carolyn J. Heinrich, ‘Workforce Development in the United States: Changing Public and Private Roles and Program Effectiveness’, in Douglas Besharov and Douglas Call (eds), Labour Activation in a Time of High Unemployment: Encouraging Work While Preserving the Social Safety-Net (Oxford: Oxford University Press, 2018), at 2. 73 Schleicher, above n 32, at 25. 74 Brink Lindsey and Steven M. Teles, The Captured Economy (New York: Oxford University Press, 2017), at 92. 75 Morris M. Kleiner and Evgeny Vorotnikov, ‘Analyzing Occupational Licensing among the States’, 52(2) Journal of Regulatory Economics 132 (2017), at 150. 76 Morris M. Kleiner, Licensing Occupations: Ensuring Quality or Restricting Competition? (Kalamazoo, MI: W.E Upjohn Institute for Employment Research, 2006), at 115. 77 ‘America Should Get Rid of Oppressive Job Licensing’, The Economist, 17 February 2018, www.economist.com. 78 Schleicher, above n 32, at 18. 79 Engler et al., ‘The Work Ahead: Machines, Skills, and US Leadership in the Twenty-First Century’ (2018) Council on Foreign Relations, Independent Task Force Report No. 76, at 24 [Task Force Report]. 80 ‘Left behind’, The Economist, 21 October 2017, at 20. 81 Ibid, at 2. 82 Bloomberg, above n 65. 83 For further discussion of policy recommendations regarding the changing nature of work in the USA, see Task Force Report, above n 79. 84 ‘Chapter 1: Labour Market Context: 2015/2016 Employment Insurance Monitoring and Assessment Report’, Canadian Employment Insurance Commission, https://www.canada.ca/en/employment-social-development/programs/ei/ei-list/reports/monitoring2016/chapter1.html (visited 12 May 2018). 85 Canadian Charter of Rights and Freedoms, Part I of the Constitution Act, 1982, being Schedule B to the Canada Act 1982 (UK), 1982, c 11, s 6. 86 Peter Hogg, Constitutional Law of Canada, 5th ed, supp vol. 2 (Toronto: Edmond Montgomery, 2016), at 46. 87 Ibid, at 46–47. 88 Department of Finance Canada, Jobs Report: The State of the Canadian Labour Market (Ottawa: Department of Finance Canada, 2014), at 28. 89 Broecke, above n 39, at 31. 90 Ibid. 91 Morley Gunderson, ‘Changes in the Labour Market and the Nature of Employment in Western Countries’, in Rethinking Workplace Regulation: Beyond the Standard Contract of Employment (New York: Russell Sage Foundation, 2013). 92 OECD Employment Outlook 2017, above n 42, at 216. 93 Broecke, above n 39, at 23. 94 ‘Ontario Basic Income Pilot’ (24 April 2017), Government of Ontario, https://news.ontario.ca/mcss/en/2017/04/ontarios-basic-income-pilot.html (visited 12 May 2018). 95 Government of Ontario, News Release, ‘Proposed Changes to Ontario’s Employment and Labour Laws’ (30 May 2017), https://news.ontario.ca/mol/en/2017/05/proposed-changes-to-ontarios-employment-and-labour-laws.html (visited 12 May 2018). 96 To clarify, the Employment Insurance program is Canada’s unemployment insurance program. It was renamed ‘Employment Insurance’ from ‘Unemployment Insurance’ in 1996 to better reflect the program’s goal of promoting employment in the labour force. 97 Olsen, above n 41, at 328. 98 Applied Research Branch, Strategic Policy, Human Resources, Development Canada, An Analysis of Employment Insurance Benefit Coverage (Ottawa: Development Canada, 1998), at 13. 99 ‘Table 1, Employment Insurance Coverage Survey’ (2015), Statistics Canada, http://www.statcan.gc.ca/daily-quotidien/161116/t001b-eng.htm (visited 12 May 2018). 100 ‘Employment Insurance Service Quality Review Report: Making Citizens Central’ (2015–16), Employment Insurance Program, https://www.canada.ca/en/employment-social-development/programs/ei/ei-list/reports/service-quality-review-report.html (visited 12 May 2018), at chapter 3. 101 ‘Number of Employment Insurance Beneficiaries by Type of Income Benefit’ (21 February 2013), Statistics Canada, http://www.statcan.gc.ca/tables-tableaux/sum-som/l01/cst01/labor13-eng.htm (visited 12 May 2018). 102 Government of Canada, ‘Canada Job Fund Agreements: ESDC’ (14 January 2016), https://www.canada.ca/en/employment-social-development/programs/training-agreements/cjf.html (visited 12 May 2018). 103 ‘Wage subsidies and tax credits for employers’ (30 January 2017), Canada Business Ontario, http://www.cbo-eco.ca/en/index.cfm/financing/government-loans-and-grants/wage-subsidies-and-tax-credits-for-employers/ (visited 12 May 2018). 104 Lysenko, above n 59, at 23. 105 Ibid, at 10. 106 Lysenko, above n 59, at 10. 107 Ibid, at 5. 108 Ibid, at 27. 109 R. Albert Berry, Labor Market Policies in Canada and Latin America: Challenges of the New Millennium (New York: Springer Science+Business Media, 2001), at 253. 110 Allison Bramwell, Training Policy for the 21st Century: Decentralization and Workforce Development Programs for Unemployed Working-Age Adults in Canada (Toronto: Mowat Centre for Policy Innovation, 2011), at 4 [Bramwell]. 111 Ibid, at 5. 112 Ibid, at 2. 113 For more detailed discussion of Bramwell’s recommendations for reform, see 16 to 21 of Bramwell, above n 110. 114 Bonoli, above n 24, at 439. 115 Ibid. 116 Consolidated Version of the Treaty on the Functioning of the European Union, Art. 45, 2008 OJ C 115/47. Note that Art. 45(3) stipulates that the rights encompassed under the free movement of workers may be limited on grounds of public policy, public security, or public health. In addition, under 45(4), freedom of movement does not apply to employment in the public service. 117 Elena Fries-Tersch, Tugce Tugran and Harriet Bradley, 2016 Annual Report on Intra-EU Labour Mobility, 2nd ed. (Brussels: European Commission, 2017), at 16 [2016 Intra-EU Labour Mobility Report]. 118 OECD Employment Outlook 2017, above n 42, at 190: ‘Table A, Harmonised unemployment rates in OECD countries’. 119 2016 Intra-EU Labour Mobility Report, above n 117, at 43. 120 Ibid, at 11. 121 Robert C. M. Beyer and Frank Smets. ‘Labour Market Adjustments and Migration in Europe and the United States: How Different’, Economic Policy 643 (2015), at 647 [Beyer]. 122 Beyer, above n 121, at 669. 123 Olsen, above n 41, at 331. 124 Per Kongshoj Madsen, ‘Flexicurity in Danish: A Model for Labour Market Reform in Europe?’ 43(2) Intereconomics 74 (2008), at 75. 125 OECD Employment Outlook 2017, above n 42, at 216: ‘Table Q, ‘Public expenditure and participant stocks in labour market programmes in OECD countries, 2014 and 2015’. 126 Bonoli, above n 24, at 444. 127 Olsen, above n 41, at 332. 128 Jochen Clasen and Elke Heins, ‘Voluntary Unemployment Insurance and Trade Union Membership: Investigating the Connections in Denmark and Sweden’, 37(3) Journal of Social Policy 433 (2008), at 434 [Clasen]. 129 Hulya Ulku and Silvia Muzi, ‘Labor Market Regulations and Outcomes in Sweden’ (2015) World Bank Group Development Economics Global Indicators Group Working Paper No. 7229, at 6. 130 OECD, ‘Denmark’ in Benefits and Wages (Paris: OECD, 2014), at 2. 131 Olsen, above n 41, at 330. 132 Clasen, above n 128, at 7. 133 Mattias Bengtsson, Transformation of Labour Market Policies in Nordic Countries: Towards a Regime Shift in Sweden and Denmark? (Gothenburg: University of Gothenburg, Department of Sociology and Work Science, 2012), at 10 [Bengtsson]. 134 ‘Unemployment’, Nordic Social Insurance Portal, http://www.nordsoc.org/en/Sweden/Unemployment1 (visited 12 May 2018). 135 Bengtsson, above n 133, at 10. 136 Sonja Bekker and Ton Wilthagan, ‘Flexicurity: A European Approach to Labour Market Policy’, March/April Intereconomics 68 (2008), at 68 [Bekker]. 137 Peter Auer, ‘What’s in a Name? The Rise (and Fall?) of Flexicurity’, 52(3) Journal of Industrial Relations 371 (2010), at 380 [Auer]. 138 Bekker, above n 136, at 71. 139 Ibid. 140 Maarten Keune, ‘Flexicurity: A Contested Concept at the Core of the European Labour Market Debate’, March/April Intereconomics 92 (2008), at 94 [Keune]. 141 Bengtsson, above n 133, at 6. 142 Ibid, at 7. 143 OECD Employment Outlook 2017, above n 42, at 190: ‘Table A, Harmonised unemployment rates in OECD countries’. 144 Auer, above n 137, at 373. 145 Ibid, at 374. 146 Ibid, at 376–77. 147 OECD Employment Outlook 2017, above n 42, at 216: ‘Table Q of OECD Employment Outlook 2017 to review the level of ALMP expenditure in OECD countries.’ 148 Bonoli, above n 24, at 446. 149 Martin, above n 26, at 4. 150 Bonoli, above n 24, at 449. 151 Olsen, above n 41, at 331. 152 Bengtsson, above n 133, at 17. 153 Olsen, above n 41, at 331. 154 Clasen, above n 128, at 15. 155 Bengtsson, above n 133, at 18. 156 Olsen, above n 41, at 331. 157 Hugh Schofield, ‘Labour Law Reform: Hollande’s Last Throw of the Dice’, The BBC, 15 March 2016, http://www.bbc.co.uk/news/world-europe-35803001 (visited 12 May 2018). 158 Chloe Farand, ‘Emmanuel Macron Kicks off Controversial Attempt to Reform France’s Labour Laws’, The Independent, 28 June 2017, http://www.independent.co.uk/news/world/europe/emmanuel-macron-france-president-launches-controversial-labour-market-reforms-labour-law-a7813171.html (visited 12 May 2018). 159 Philip Rathgeb and Fabio Wolkenstein, ‘Third-Way à la Français: What do Macron’s Reforms Involve and How Likely are They to Succeed?’, London School of Economics, http://blogs.lse.ac.uk/europpblog/207/06/23/third-way-francaise-macron-reforms/ (visited 12 May 2018). 160 Ulf Rinne. ‘Is Germany the North Star of Labor Market Policy?’ IZA Discussion Paper No. 7260, at 1 [Rinne]. 161 Ibid, at 4. 162 OECD Employment Outlook 2017, above n 42, at 190: ‘Table A, Harmonised unemployment rates in OECD countries’. 163 Rinne, above n 160, at 1. 164 Niklas Engbom, Enrica Detragiache and Faezeh Raei, ‘The German Labour Market Reforms and Post-Unemployment Earnings’, International Monetary Fund Working Paper No. WP/15/162 (2015), at 10. 165 Werner Eichhorst, ‘Non-Standard Employment across Occupations in Germany: The Role of Replaceability and Labour Market Flexibility’, IZA Discussion Paper No. 7662 (2013), at 2. 166 OECD, ‘Germany’ in Benefits and Wages (Paris: OECD, 2014), at 2. 167 Rüdiger Wapler, Daniel Werner and Katja Wolf, ‘Active Labour-Market Policies in Germany: Do Regional Labour Markets Benefit?’, Institute for Employment Research IAB Discussion Paper No. 28/2014 (2014), at 12. 168 Oliver Bruttel, ‘Delivering Active Labour Market Policy through Vouchers: Experiences with Training Vouchers in Germany’, 71(3) International Rev of Administrative Sciences 391 (2005), at 396 [Bruttel]. 169 Ibid, at 402. 170 Ludo Struyven and Geert Steurs, ‘Design and Redesign of a Quasi-Market for the Reintegration of Jobseekers: Empirical Evidence from Australia and the Netherlands’, 15(3) Journal of European Social Policy 211 (2005), at 212. 171 OECD, Activating Jobseekers: How Australia Does It (Paris: OECD, 2012), http://www.oecd.org/els/emp/activatingjobseekershowaustraliadoesit.htm (visited 12 May 2018), at 13 [How Australia Does It]. 172 Ibid, at 62. 173 Ibid, at 76. 174 How Australia Does It, above n 171, at 13. 175 WTO Report 2017, above n 6, at 10. 176 Michael Trebilcock and Mariana Mota Prado, Advanced Introduction to Law and Development (Toronto: Edward Elgar Publishing, 2014), at chapter 3. 177 See also: Jordann Thirgood, ‘What if You Could Take It With You?’ (27 July 2017), Mowat Centre, https://mowatcentre.ca/what-if-you-could-take-it-with-you/ (visited 12 May 2018). 178 Bramwell, above n 110. 179 While effective government evaluations are lacking, there have been many academic evaluations of ALMPs in OECD countries, including (i) James Heckman, Robert Lalonde and Jeffrey Smith, ‘The Economics and Econometrics of Active Labor Market Programs’ in David Card and Orley Ashenfelter (eds), Handbook of Labor Economics, vol. 3 (Amsterdam: Elsevier Science, 1999); and (ii) David Card, Jochen Kluve and Andrea Weber, ‘Active Labour Market Policy Evaluations: A Meta-Analysis’, 120 Economic Journal (2010) F452 to F477. 180 Task Force Report, above n 79, at 35. 181 Yuval Levin, The Fractured Republic: Renewing America’s Social Contract in an Age of Individualism (New York: Basic Books, 2016). © The Author(s) 2018. Published by Oxford University Press. All rights reserved. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

Journal of International Economic LawOxford University Press

Published: Jun 5, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off