Agency culture, constitutional provisions and entrepreneurship: a cross-country analysis

Agency culture, constitutional provisions and entrepreneurship: a cross-country analysis Abstract Substantial and systematic cross-country variation in entrepreneurship rates has been found in various studies. We attempt to explain such differences focusing on the interaction between institutional factors and population psychological characteristics. Constitutional provisions supporting economic freedom are our measure of the institutional context, whereas we proxy psychological characteristics with a country’s endowment of agency culture. We apply an IV-GMM treatment to deal with endogeneity to a data set comprising 86 countries over the period 2004–2013, and we control for de facto variables and other factors that are likely to influence entrepreneurship. Our results demonstrate that agency culture is indeed an important predictor of entrepreneurship and that this effect is moderated by constitutional provisions supporting economic freedom. In particular, the impact of agency culture on entrepreneurship becomes stronger as a country expands the constitutional protection of economic rights. 1. Introduction Cross-country comparison of industry dynamics and exploration of its determinants and consequences has traditionally attracted the interest of researchers in both industrial and developing countries (see Caves, 1998; and Bartelsman et al., 2009 for surveys). The results of this literature show that substantial and systematic differences in industry dynamics are generated also by country-specific institutional and cultural factors (see Bottazzi et al., 2010; Bartelsman et al., 2013; Niszczota, 2014). The aim of the present article is to study the interplay between the economic constitution of a country (institutional factor) and the macro-psychological traits of its population (cultural factor) in shaping cross-country differences in entrepreneurship rates. Our hypothesis is that constitutional protection of economic freedom may together create an institutional setting that favors the transformation of the innate agentic attitude of a country’s population into actual entrepreneurship. It follows from this assumption that differences in the constitutions and the endowment of agency culture, and also their interplay, may explain the cross-country variation in industry dynamics. We conduct our analysis using a sample comprising 86 countries over the period 2004–2013 (see list in Table 1). Table 1. List of countries by geographical area America: Argentina; Belize; Bolivia; Brazil; Canada; Chile; Colombia; Costa Rica; Dominican Republic; El Salvador; Guatemala; Jamaica; Mexico; Panama; Peru; the United States; Uruguay.  Europe: Albania; Armenia; Austria; Belgium; Bosnia and Herzegovina; Bulgaria; Croatia; Cyprus; Czech Republic; Denmark; Estonia; Finland; France; Germany; Greece; Hungary; Iceland; Ireland; Italy; Latvia; Lithuania; Luxembourg; Macedonia; Malta; Montenegro; The Netherlands; Norway; Poland; Portugal; Romania; Russia; Slovak Republic; Slovenia; Spain; Sweden; Switzerland; Ukraine; UK.  Africa: Algeria; Ethiopia; Ghana; Kenya; Mauritius; Morocco; Nigeria; South Africa; Uganda.  Asia: Afghanistan; Bangladesh; Brunei; Egypt; India; Indonesia; Israel; Japan; Jordan; Malaysia; Nepal; Oman; Pakistan; Philippines; Qatar; Singapore; South Korea; Sri Lanka; Thailand; Turkey; United Arab Emirates (UAE).  Oceania: Australia; New Zealand.  America: Argentina; Belize; Bolivia; Brazil; Canada; Chile; Colombia; Costa Rica; Dominican Republic; El Salvador; Guatemala; Jamaica; Mexico; Panama; Peru; the United States; Uruguay.  Europe: Albania; Armenia; Austria; Belgium; Bosnia and Herzegovina; Bulgaria; Croatia; Cyprus; Czech Republic; Denmark; Estonia; Finland; France; Germany; Greece; Hungary; Iceland; Ireland; Italy; Latvia; Lithuania; Luxembourg; Macedonia; Malta; Montenegro; The Netherlands; Norway; Poland; Portugal; Romania; Russia; Slovak Republic; Slovenia; Spain; Sweden; Switzerland; Ukraine; UK.  Africa: Algeria; Ethiopia; Ghana; Kenya; Mauritius; Morocco; Nigeria; South Africa; Uganda.  Asia: Afghanistan; Bangladesh; Brunei; Egypt; India; Indonesia; Israel; Japan; Jordan; Malaysia; Nepal; Oman; Pakistan; Philippines; Qatar; Singapore; South Korea; Sri Lanka; Thailand; Turkey; United Arab Emirates (UAE).  Oceania: Australia; New Zealand.  Table 1. List of countries by geographical area America: Argentina; Belize; Bolivia; Brazil; Canada; Chile; Colombia; Costa Rica; Dominican Republic; El Salvador; Guatemala; Jamaica; Mexico; Panama; Peru; the United States; Uruguay.  Europe: Albania; Armenia; Austria; Belgium; Bosnia and Herzegovina; Bulgaria; Croatia; Cyprus; Czech Republic; Denmark; Estonia; Finland; France; Germany; Greece; Hungary; Iceland; Ireland; Italy; Latvia; Lithuania; Luxembourg; Macedonia; Malta; Montenegro; The Netherlands; Norway; Poland; Portugal; Romania; Russia; Slovak Republic; Slovenia; Spain; Sweden; Switzerland; Ukraine; UK.  Africa: Algeria; Ethiopia; Ghana; Kenya; Mauritius; Morocco; Nigeria; South Africa; Uganda.  Asia: Afghanistan; Bangladesh; Brunei; Egypt; India; Indonesia; Israel; Japan; Jordan; Malaysia; Nepal; Oman; Pakistan; Philippines; Qatar; Singapore; South Korea; Sri Lanka; Thailand; Turkey; United Arab Emirates (UAE).  Oceania: Australia; New Zealand.  America: Argentina; Belize; Bolivia; Brazil; Canada; Chile; Colombia; Costa Rica; Dominican Republic; El Salvador; Guatemala; Jamaica; Mexico; Panama; Peru; the United States; Uruguay.  Europe: Albania; Armenia; Austria; Belgium; Bosnia and Herzegovina; Bulgaria; Croatia; Cyprus; Czech Republic; Denmark; Estonia; Finland; France; Germany; Greece; Hungary; Iceland; Ireland; Italy; Latvia; Lithuania; Luxembourg; Macedonia; Malta; Montenegro; The Netherlands; Norway; Poland; Portugal; Romania; Russia; Slovak Republic; Slovenia; Spain; Sweden; Switzerland; Ukraine; UK.  Africa: Algeria; Ethiopia; Ghana; Kenya; Mauritius; Morocco; Nigeria; South Africa; Uganda.  Asia: Afghanistan; Bangladesh; Brunei; Egypt; India; Indonesia; Israel; Japan; Jordan; Malaysia; Nepal; Oman; Pakistan; Philippines; Qatar; Singapore; South Korea; Sri Lanka; Thailand; Turkey; United Arab Emirates (UAE).  Oceania: Australia; New Zealand.  Taking a law and economics perspective, we focus on four principles stated in national constitutions: the right to conduct/establish a business, the right to free/competitive markets, the right to own property, and the independence of the judiciary organs. The first two principles have been proven in the law and economics literature to exert a significant impact on entrepreneurship (Carbonara et al., 2016). The right to own property (Besley and Ghatak, 2010) and the independence of the judiciary organs (Djankov et al., 2003, Chemin, 2009) are widely considered key factors in economic development. The use of constitutional provisions as proxy for institutional characteristics stems from the fact that constitutional laws represent hierarchically higher norms that cannot be opposed by ordinary laws and other rules (Kelsen, 1967). Thus, they represent the highest protection that a country can assign to rights. However, often the laws in the books remain unenforced (especially when they clash with social norms that are firmly embedded in culture: Carbonara et al., 2012, Acemoglu and Jackson, 2016). That is why de jure protection of legal rights does not necessarily imply a de facto protection, and we need to control for the actual implementation of the principles stated in the constitution, including measures of economic freedom based on rule of law, government size, regulatory efficiency, and market openness. The psychological literature has placed great importance on personality traits, arguing that regional differences in personality constitute a local culture that in turn influences regional entrepreneurship rates (cf., among others, Davidsson, 1995; Leutner et al., 2014; Obschonka et al., 2015; Stuetzer et al., 2016). Such a personality-based perspective on culture (Hofstede and McCrae, 2004) has enriched our understanding of the historical origins (Stuetzer et al., 2016) and economic effects of regional differences in an entrepreneurial culture (Davidsson, 1995; Steel et al., 2012; Rentfrow et al., 2013; Audretsch et al., 2017). To date, the law and economics and the psychological streams of literature have not been integrated in the explanation of the overall process of industry dynamics. There is good reason to suppose that combining them could prove profitable, with some studies hinting at the promise of such integrative perspectives. For example, Obschonka et al. (2015) examined the so-called “knowledge paradox”, which is the phenomenon whereby investments in resources for generating knowledge (e.g., education, diversity of industries) do not guarantee higher entrepreneurship rates; analyses revealed that knowledge resources are more likely to increase entrepreneurship rates in a region which also has a high number of residents with an entrepreneurship-prone personality. In other words, psychological (in this case, entrepreneurial personality) and institutional determinants (in this case, knowledge resources) may interact to yield better predictions about entrepreneurial activity than the additive effects of both determinants assessed in isolation. The article is organized as follows. Section 2 contains a review of the literature dealing with cross-national differences in entrepreneurship rates, and the impact of psychological traits and constitutional provisions on entrepreneurship. Section 3 presents the main research questions and theoretical explanations. Section 4 describes the data set. Section 5 presents the estimation model and econometric strategy. Section 6 discusses the findings of the empirical analysis. Finally, Section 7 offers some conclusions based on these findings. 2. Literature review In economics, agentic behavior is usually defined within a rational-choice perspective, assuming that agency is mainly characterized by the maximization of one’s own benefits: all actors are narrowly self-interested, all actors are boundedly rational, and agents are more risk averse than principals are (Bosse and Phillips, 2016). However, this approach neglects inter-individual psychological differences (i.e., personality characteristics motivating, guiding, and directing decisions and activities), which have been shown to predict a wide array of consequential life outcomes and economic behaviors, even when controlling for the effects of socio-economic status, demographic variables, and cognitive ability (Roberts et al., 2007). Psychological research points to some sort of psychological benefit for the individual (or the avoidance of negative, harmful states) if he or she can behave in accordance with his or her individual personality structure (Frey, 2008). Psychological theories suggest that peoples’ behavior can best be understood by an interplay between person variables (like personality) and the context (cf., among others, Lewin, 1935, 1951; Ajzen, 1985; Funder, 2006), which means that it might be worth analyzing carefully the interaction between psychological traits and constitutional provisions. The agentic perspective is widely regarded as a leading meta-theory of human behavior in psychology (Bandura, 2006), sociology (Elder, 1994), economics (Kihlstrom and Laffont, 1979), law (Parker, 2007), and management (Begley and Boyd, 1987). 2.1 Cross-country differences in entrepreneurship rates Some countries experience higher rates of new firm formation every year than other countries do (Carree et al., 2002; Santarelli and Vivarelli, 2007). From a theoretical viewpoint, two main explanations of this empirical regularity have pervaded the recent debate. On the one side, based on the observation that developed Western countries have become more entrepreneurial following globalization, Audretsch and Thurik (2000) hypothesize that such countries switch to new industries—such as software and biotechnology, in which small businesses and entrepreneurship are more important—only once they have lost their comparative advantage in large-scale manufacturing. Thus, high rates of new firm formation are typical of developed countries since the aftermath of the Information and Communication Technology revolution. This pattern might suggest that a country’s endowment of agency culture evolves in response to historical and institutional changes. On the other side, Galor and Michalopoulos (2012) suggest that entrepreneurial spirit has evolved non-monotonically in the course of history, through a Darwinian process. In the early stages of a country’s development, risk-tolerant entrepreneurial traits proved successful in promoting technological progress and economic development, whereas in mature stages of development, risk-averse traits prevailed, diminishing the growth potential of advanced economies. Thus, modern developed countries should experience lower rates of new firm formation. This latter approach implies that a country’s endowment of agency culture is linked to its stage of economic development rather than to time-invariant psychological features; hence, agency culture would tend to vanish as societies evolve, develop more complex institutional arrangements, and achieve higher levels of per capita income. Both of the views outlined above are consistent with the idea that the reason why countries with similar economic fundamentals differ in entrepreneurial activity may ultimately be found in cultural and institutional differences (Guiso et al., 2003; Stuetzer et al., 2016). The positions of Audretsch and Thurik (2000) and Galor and Michalopoulos (2012) can therefore be reconciled within a broader line of investigation, which spans from Diamond (1997) to Acemoglu and Robinson (2012) (cf. also Saxenian, 1994; Acemoglu et al., 2001; Autio et al., 2014). In fact, the empirical literature on the cross-country differences in start-up rates has provided several contributions which can be reconciled with each of the two explanations. Guiso et al. (2006) show that the cultural background of individuals plays a role in their decision to become entrepreneurs and therefore also shapes attitudes toward entrepreneurship at the region and country level. By the same token, Audretsch et al. (2017) posit the importance of culture as a primary determinant of variations in economic, political, and social phenomena across geographic space. They aggregate individual-level personality data to the level of each of the 3137 US counties to analyze the impact of a social and cultural imprinting on the rate of new firm formation at the county level. Wennekers et al. (2005) found a U-shaped relationship between a country’s start-up rate and a country’s level of economic development, with the impact of entrepreneurial dynamics on economic growth being smaller for developing countries. In contrast, Blanchflower (2000) showed that the overall trend of entrepreneurial activities does not follow the stages of a country’s development; rather, the trend shows a negative relationship with a country’s unemployment rate. Brück et al. (2011) found that entrepreneurship rates follow a history-dependent path and are subject to the influence of exogenous factors, with entrepreneurship rates being positively affected by extreme events such as natural disasters and terrorist attacks. Dealing with 85 countries between 2005 and 2014, Dheer (2017) has shown that an institutional setting that guarantees economic freedom affects the rate of entrepreneurial activity more in individualistic societies than in collectivistic ones. This is a clear indication that population psychological characteristics play a role in positively moderating the effect of pro-market institutional arrangements on entrepreneurship. Our study falls within the same line of investigation, although we use different measures for entrepreneurship and for economic freedom protection and we focus on different countries. A positive relationship between presence of an institutional framework able to promote economic freedom and various measures of entrepreneurship was found also by Bjørnskov and Foss (2008) and Nyström (2008). The above empirical evidence seems to suggest there is no single unique economic factor explaining cross-country differences in start-up rates. Differences might persist over time, regardless of a country’s level of economic development. Cultural differences might shape a country’s proneness toward entrepreneurial activity and are therefore a factor that needs consideration and integration. A country’s culture (e.g., its endowment of social capital; Guiso et al., 2008; or the national levels in personality traits, Hofstede and McCrae, 2004; Steel et al., 2012) interacts with and is shaped by the features and quality of a country’s institutions. Institutions are therefore an important aspect of a country’s profile and their impact extends from economic development (Guiso et al., 2003) to entrepreneurship (Acs et al., 2008; Carbonara et al. 2016). Culture and institutions are ultimately endogenous variables contributing to the wealth of countries. Extending the arguments proposed by Alesina and Giuliano (2015), we assume here that the psychological traits of a country’s population and the provisions contained in a country’s constitution are, respectively, aspects of a country’s culture and a country’s institutions. 2.2 Psychological agency and entrepreneurship The focus on an individual’s personal agency has long played a key role in seminal theorizing in the entrepreneurship literature (McClelland, 1961). In fact, Schumpeter himself (Schumpeter, 1911: 131, as translated from German in Santarelli and Pesciarelli, 1990), in the first German edition of his Theory of Economic Development (Schumpeter, 1934), stressed that entrepreneurs are “personalities who in se possess the rules of their actions” (for a detailed discussion of this issue, cf. Santarelli and Pesciarelli, 1990). Empirical entrepreneurship research has usually tried to capture such personal agency by focusing on an entrepreneur’s actual actions (Frese, 2009; Zhao et al., 2010; Hmieleski et al., 2015) or self-efficacy belief (Hechavarria et al., 2012; Wennberg et al., 2013). Here, we apply a novel approach to capture psychological agency, assessing it in terms of agentic personality traits (Digman, 1997). This approach is based on the leading and best researched model of personality traits, the Big Five model (John and Srivastava, 1999). This approach of assessing agency also allows us to draw from geographical approaches in the study of regional and national differences in these personality traits (Rentfrow et al., 2008; Steel et al., 2012). In psychology, Digman’s (1997) influential work on higher-order traits (or super traits) established that two Big Five traits, extraversion and openness to new experience, form a higher-order trait that can be labelled “psychological agency”. Drawing from that approach, we measure agency culture at the nation-level. Populations living in countries characterized by a high level of agency culture are highly active and assertive (components of extraversion), highly creative and open to change (components of openness to experience). Accordingly, aggregates of individual scores on traits are used as proxies for agency culture (for similar approaches to assess cultural dimensions, see Davidsson, 1995; Rentfrow et al., 2008, 2013; Steel et al., 2012; Stuetzer et al., 2016; Audretsch et al., 2017). In contrast to a purely rational-choice approach, the psychological approach defines agency by means of relatively stable personality traits that motivate, guide, and direct manifest individual agency. This psychological agency approach has largely been neglected in economic models of agency and entrepreneurship, despite the demonstrated importance of psychological models in economics (Borghans et al., 2008). In fact, psychological research has challenged the pure rational-choice view by pointing toward the relevance of “irrational” decision-making processes involving personality traits. A wide array of basic personality traits can have considerable influence on economic outcomes: for example, a recent study showed that entrepreneurial activity in the wake of the Great Recession of 2008–2009 was predicted better by regional personality differences than regional infrastructure parameters (“economic muscles”, such as human and financial capital) (Obschonka et al., 2016). 2.3 Entrepreneurship, agency culture, and the moderating effect of national constitutions Recent meta-analytic studies suggest significant relationships between personality, and both revealed preference for becoming an entrepreneur (latent entrepreneurship) (Zhao et al., 2010) and entrepreneurial performance after start-up (Brandstätter, 2011). However, there has been little investigation about the importance of personality as a predictor of the probability of actually being an entrepreneur (manifest entrepreneurship) (Grilo and Thurik, 2006; Baron and Baum, 2007; Audretsch et al. 2017). The level of agency culture in a country represents an important component of the overall cultural context within which entrepreneurial activity takes place. The relationship between a broader definition of culture—encompassing customary beliefs and values that are transmitted from generation to generation—and the likelihood of engaging in entrepreneurship has been explored by a line of investigation initiated by Guiso et al. (2006). Following the idea put forward by Glaeser et al. (2000), Guiso et al. (2006) assume that individuals who express trust have a comparative advantage in becoming entrepreneurs. Then, studying the impact of a measure of trust weighted for religious and ethnic background on the probability of a sample of individuals becoming entrepreneurs, Guiso et al. (2006) use an instrumental variable approach to confirm that trustworthy individuals will have a comparative advantage in becoming entrepreneurs. In the same way, defining a cultural variable along the dimension of individualism–collectivism, Gorodnichenko and Roland (2010) show that individualism has a dynamic advantage leading to a higher economic growth rate, whereas collectivism leads only to static efficiency gains. Then, a further element is added to the investigation initiated by Guiso et al. (2006): institutions (Alesina and Giuliano, 2015). Culture and institutions interact and evolve in a complementary way, both playing a role as determinants of the wealth of countries. Consequently, the same institutions may exert a different impact in different cultural contexts. Alesina and Giuliano (2015) identify a wide range of interactions between various types of political and legal institutions and various cultural traits, such as trust, family ties, generalized morality, and individualism. Lerner and Tåg (2013) show that institutional differences in the legal environment led to the later development of an active venture capital market in Sweden compared with the United States, where this source of external funding largely contributed to the emergence of clusters of innovative start-ups. The antinomy, individualism versus collectivism, is an important dimension of cultural variation across countries (Greif, 1994; Gorodnichenko and Roland, 2010; Alesina and Giuliano, 2015). Individualism is a trait that can make personal accomplishments more socially acceptable, so it is likely to be associated with a greater proneness to entrepreneurship. But for this cultural trait to result into actual action, the overall institutional setting should be proactive and remove the obstacles to the full display of individualism in the economic sphere. Consistent with this principle, within the broader field of law and economics, the public choice literature has emphasized (since the seminal contribution of Buchanan and Tullock, 1962) that a strong connection exists between a country’s economic performance and the main features of its constitution. Such a connection is likely not direct, but rather the result of the effectiveness of constitutions in shaping a country’s prevailing institutional arrangements (Melton et al., 2013; Carbonara et al., 2016). In relation to how constitutions may make it easier to turn individualism into actual entrepreneurial action, it is worth acknowledging that at least since France’s National Constituent Assembly passed the Déclaration des droits de l'homme et du citoyen in 1789, general recognition and protection of private property had been held to be universal and most constitutions started to protect property rights. The first empirical work on the impact of constitutions on economic performance dates back to the early 2000s; this work showed the positive impact jointly exerted by a presidential system and the majoritarian electoral rule on, among other things, total factor productivity and reduction of public expenditure (Persson and Tabellini, 2003). Research has also shown that direct democratic institutions affect fiscal policy and government efficiency (Blume et al., 2009). More recently, the optimal number of national representatives in relation to a country’s population size has been calculated (Auriol and Gary-Bobo, 2012); analyses suggest that an excessive number of national representatives are correlated with indicators of red tape and barriers to entrepreneurship. Moreover, the efficiency of the judiciary positively affects entrepreneurship. Constitutional provisions do play a role in making a judiciary system more efficient, for example, by stating that it must be independent from external influences. A well-functioning judiciary system facilitates access to finance and reduces the likelihood of contract breach (Chemin, 2009, 2012). The more a judiciary system is independent from the influence of both the other branches of government and partisan interests, the more judges are free to make impartial decisions based exclusively on fact and the rule of law. As a result, an independent and more efficient judiciary system may exert a direct impact on entrepreneurship, while it leaves the exit rate unaltered (Chemin, 2012; García-Posada and Mora-Sanguinetti, 2015). Highly skilled and better educated entrepreneurs take advantage of better access to justice (Lichand and Soares, 2014 on Brazilian data). Thus, reforms aimed at improving the efficiency of the judiciary may affect entrepreneurship positively among individuals with higher levels of education but not among those with lower educational levels. Education is a proxy for wealth, so this result seems to indicate that judicial change in Brazil pushed wealthier individuals toward entrepreneurship. Carbonara et al. (2016) show for 115 countries that constitutional provisions are the main institutional driver of entrepreneurship. Dealing with the endogeneity of constitutional rules, and controlling for de facto variables, they find that provisions about the right to conduct/establish a business, the right to strike, consumer protection, anti-corruption, and compulsory education promote higher rates of new firm formation. 3. Main hypotheses In what follows, we extend the investigation of the relationship between culture and institutions in the context of entrepreneurship by exploring the interaction between agency culture and the aspect of legal institutions represented by the provisions supporting economic freedom that are contained in national constitutions. We put a special focus on the interplay between agency culture and these aspects of the legal institutions within a country. To quantify psychological differences in agency, we apply the personality-based approach to culture (Rentfrow et al., 2008), which aggregates individual-level personality traits to estimate local cultural differences (Stuetzer et al., 2016). This approach has delivered promising findings in research predicting regional outcomes, including social, economic, political, and health outcomes (Rentfrow et al., 2013; Jokela et al., 2015). The basic idea underlying this research is that regional personality differences constitute the pillars of the local culture, affecting the developmental trajectories of whole regions (Hofstede and McCrae, 2004). In psychological science, there is broad consensus that the five-factor model of personality is the best-established, validated, and cross-culturally valid model of personality (cf., among many others, Digman, 1997; Benet-Martinez and John, 1998; John and Srivastava, 1999; Zhao and Seibert, 2006; Lang et al., 2011; Gebauer et al., 2014a; Vedel, 2014). The Big Five personality traits constituting this five-factor model are extraversion, conscientiousness, openness, agreeableness, and neuroticism. The Big Five traits can be further summarized in the form of higher order “super” traits (Wiggins, 2003). Based on analyses of child, adolescent, and adult samples, Digman (1997) established two higher-order “super” traits: α (consisting of conscientiousness, agreeableness, and neuroticism) and β (consisting of extraversion and openness). α can be described as a dimension encapsulating themes of communion, and β can be described as a dimension encapsulating themes of agency (Wiggins, 1991). The β super trait also includes such traits as superiority striving, individuation, personal growth, self-actualization, achieving status, and power motivation (Digman, 1997). These traits are associated with both agency and entrepreneurial behavior (Zhao and Seibert, 2006) making β a good candidate for indexing psychological agency in a way that is relevant for entrepreneurship. The study of such super factors (e.g., agency and communion) has received considerable attention in recent years (Blackburn et al., 2004; DeYoung, 2006; Abele and Wojciszke, 2007; Vecchione and Alessandri, 2013; Gebauer et al., 2014b), but this trend has not been mirrored in economic research, which has remained focused on narrower personality traits or profiles (Borghans et al., 2008; Stuetzer et al., 2016). With the aim to combine the views that institutional factors and population psychological characteristics are drivers of new firm formation, in line with the approach established by Dheer (2017), we aim to test the degree to which the combination of agency culture and pro-market constitutional framework combine to predict entrepreneurial activity. Thus, extending Carbonara et al. (2016), we take the constitutional protection granted to some principles relevant for economic activity and their de facto implementation (cf. also Carlsson et al., 2009; Czarnitzki et al., 2016) as proxies for the institutional determinants of entrepreneurship. Consistent with the bounded agency approach, we predict that boundary conditions (in the form of national constitutional framework) will shape the overall impact of agency culture on a country’s proneness toward entrepreneurial activity. Accordingly, pro-entrepreneurship constitutions cannot stimulate new business formation across countries as expected if people in those countries are not sufficiently proactive and innovative to exploit the benefits of the created favourable constitutional environment. Broadening the perspective followed in the previous empirical literature—from Blau (1987) to Acs et al. (2009)—and extending previous findings by Carbonara et al. (2016), we take into account the possible moderating effect of a specific aspect of the institutional setting, represented by the provisions contained in a country’s constitution. Particularly, we focus on two central hypotheses: H1a:The level of agency culture and the presence in the constitutions of provisions supporting economic freedom predict a country’s level of entrepreneurial activity. H1b:Following the bounded agency perspective, the constitutional environment moderates the effect of agency culture on entrepreneurship. 4. Data 4.1 Dependent variable 4.1.1 New business density Using data from the World Bank Group Entrepreneurship Database, we construct a measure of new business density, given by the number of new business registrations (private, formal sector companies with limited liability) in every year in each country per 1000 residents aged 15–64 years over the period 2004–2013. Our dependent variable is a standard measure of the total start-up activity in 86 countries (Table 1). It is a measure of entrepreneurship that follows a labor market approach (Audretsch and Fritsch, 1994): all firms are the result of individual actions, and new entrepreneurs are individuals who had previously or have been interested in having a dependent job, who exploit their knowledge of production processes and market features to switch to independent work (Santarelli and Sterlacchini, 1994; Gries and Naudé, 2011). Accordingly, each individual in the labor pool is considered a potential entrepreneur, with the capability to set up his or her own business. We believe that this measure of entrepreneurship is best suited to study the impact of cultural and institutional factors on entrepreneurship because we are in fact focusing on how individual private initiative is fostered or jeopardized by culture and institutions. Other available and commonly used measures include the ratio of new entrants on existing firms, adopted in the so-called ecological approach (Tag et al., 2016), and the number of business owners per labor force (Acs et al., 2009). However, such measures are less interesting for our purposes. The ecological approach measures new start-up activity relative to existing entrepreneurship, thus capturing only one component of the overall process. The percentage of the self-employed is more suited for a study on occupational choices, whereas here we are more interested in a story of entrepreneurial success. 4.2 Independent variables 4.2.1 Agency culture We utilize personality data collected by the ongoing, global Gosling–Potter Internet project (Gosling et al., 2004; see also Gebauer et al., 2015; Rentfrow et al., 2013, 2015). The project collects personality data via a noncommercial Internet website, which can be reached through several channels (e.g., search engines, unsolicited links on other Web pages). People voluntarily participate in this study by responding to items on a standard Big Five personality questionnaire (in English, German, Spanish, or Dutch) using a five-point Likert scale (1 = disagree strongly, 5 = agree strongly); as an incentive, participants receive a personality evaluation based on their responses. Participants also provide responses to questions on several sociodemographic variables, and report their state of residence. This database has yielded numerous publications relating personality traits to various aspects of human behavior. Its validity is supported by the number and quality of publications, mainly in the field of psychology, which have used data from this large-scale Internet project. Of most relevance to the current work, smaller versions of this data set have been successfully employed in cross-cultural studies (Bleidorn et al., 2013; Gebauer et al., 2015). For a list of published studies using the database, see http://www.thebigfiveproject.com/published-papers/). To estimate cross-country differences in agency culture, we use data from all respondents who completed the questionnaire from the start of the project in December 1998 until 2015. In total, N = 7092, 784 respondents are included in this data set. The number of respondents in each country ranges between 1008 (Ethiopia) and 4,275,860 (the United States). Table 2 provides an overview over the sample sizes in each country under study. Country-level agency scores were derived in two steps. In the first step, participants’ extraversion and openness scores were computed and these were averaged to yield an agency score at the individual level. In the second step, individuals’ scores were aggregated within country, yielding country-level scores for agency culture. Table 2. Sample size of the individual-level personality data set for each country Country  N  Country  N  Country  N  Country  N  Afghanistan  1172  Denmark  19,074  Kenya  6985  Portugal  8334  Albania  2288  Dominican R.  6222  Korea (South)  9960  Qatar  2064  Algeria  1074  Egypt  9075  Latvia  1440  Romania  13,055  Argentina  88,211  El Salvador  3682  Lithuania  2277  Russia  3624  Armenia  1084  Estonia  2459  Luxembourg  1116  Serbia and Montenegro  5665  Australia  195,857  Ethiopia  1008  Macedonia  1121  Singapore  59,119  Austria  27,143  Finland  23,526  Malaysia  39,606  Slovak Republic  1691  Bangladesh  3482  France  18,502  Malta  1590  Slovenia  3095  Belgium  43,692  Germany  186,848  Mauritius  1706  South Africa  26,039  Belize  1025  Ghana  1949  Mexico  136,305  Spain  135,048  Bolivia  6115  Greece  10,982  Morocco  1346  Sri Lanka  3958  Bosnia and Herzegovina  1371  Guatemala  5635  Nepal  2142  Sweden  46,828  Brazil  26,538  Hungary  3746  The Netherlands  163,472  Switzerland  36,741  Brunei  1211  India  114,500  New Zealand  43,167  Thailand  8501  Bulgaria  3610  Indonesia  15,199  Nigeria  7033  Turkey  5298  Canada  371,882  Iceland  2520  Norway  42,859  UAE  14,907  Chile  44,552  Ireland  41,257  Oman  1068  Uganda  1377  Colombia  34,905  Israel  7426  Pakistan  27,498  Ukraine  1081  Costa Rica  6712  Italy  13,831  Panama  2938  UK  438,854  Croatia  6920  Jamaica  4199  Perù  23,056  The United States  4,275,860  Cyprus  2307  Japan  10,232  Philippines  91,638  Uruguay  6351  Czech Republic  3566  Jordan  2431  Poland  7951      Country  N  Country  N  Country  N  Country  N  Afghanistan  1172  Denmark  19,074  Kenya  6985  Portugal  8334  Albania  2288  Dominican R.  6222  Korea (South)  9960  Qatar  2064  Algeria  1074  Egypt  9075  Latvia  1440  Romania  13,055  Argentina  88,211  El Salvador  3682  Lithuania  2277  Russia  3624  Armenia  1084  Estonia  2459  Luxembourg  1116  Serbia and Montenegro  5665  Australia  195,857  Ethiopia  1008  Macedonia  1121  Singapore  59,119  Austria  27,143  Finland  23,526  Malaysia  39,606  Slovak Republic  1691  Bangladesh  3482  France  18,502  Malta  1590  Slovenia  3095  Belgium  43,692  Germany  186,848  Mauritius  1706  South Africa  26,039  Belize  1025  Ghana  1949  Mexico  136,305  Spain  135,048  Bolivia  6115  Greece  10,982  Morocco  1346  Sri Lanka  3958  Bosnia and Herzegovina  1371  Guatemala  5635  Nepal  2142  Sweden  46,828  Brazil  26,538  Hungary  3746  The Netherlands  163,472  Switzerland  36,741  Brunei  1211  India  114,500  New Zealand  43,167  Thailand  8501  Bulgaria  3610  Indonesia  15,199  Nigeria  7033  Turkey  5298  Canada  371,882  Iceland  2520  Norway  42,859  UAE  14,907  Chile  44,552  Ireland  41,257  Oman  1068  Uganda  1377  Colombia  34,905  Israel  7426  Pakistan  27,498  Ukraine  1081  Costa Rica  6712  Italy  13,831  Panama  2938  UK  438,854  Croatia  6920  Jamaica  4199  Perù  23,056  The United States  4,275,860  Cyprus  2307  Japan  10,232  Philippines  91,638  Uruguay  6351  Czech Republic  3566  Jordan  2431  Poland  7951      Table 2. Sample size of the individual-level personality data set for each country Country  N  Country  N  Country  N  Country  N  Afghanistan  1172  Denmark  19,074  Kenya  6985  Portugal  8334  Albania  2288  Dominican R.  6222  Korea (South)  9960  Qatar  2064  Algeria  1074  Egypt  9075  Latvia  1440  Romania  13,055  Argentina  88,211  El Salvador  3682  Lithuania  2277  Russia  3624  Armenia  1084  Estonia  2459  Luxembourg  1116  Serbia and Montenegro  5665  Australia  195,857  Ethiopia  1008  Macedonia  1121  Singapore  59,119  Austria  27,143  Finland  23,526  Malaysia  39,606  Slovak Republic  1691  Bangladesh  3482  France  18,502  Malta  1590  Slovenia  3095  Belgium  43,692  Germany  186,848  Mauritius  1706  South Africa  26,039  Belize  1025  Ghana  1949  Mexico  136,305  Spain  135,048  Bolivia  6115  Greece  10,982  Morocco  1346  Sri Lanka  3958  Bosnia and Herzegovina  1371  Guatemala  5635  Nepal  2142  Sweden  46,828  Brazil  26,538  Hungary  3746  The Netherlands  163,472  Switzerland  36,741  Brunei  1211  India  114,500  New Zealand  43,167  Thailand  8501  Bulgaria  3610  Indonesia  15,199  Nigeria  7033  Turkey  5298  Canada  371,882  Iceland  2520  Norway  42,859  UAE  14,907  Chile  44,552  Ireland  41,257  Oman  1068  Uganda  1377  Colombia  34,905  Israel  7426  Pakistan  27,498  Ukraine  1081  Costa Rica  6712  Italy  13,831  Panama  2938  UK  438,854  Croatia  6920  Jamaica  4199  Perù  23,056  The United States  4,275,860  Cyprus  2307  Japan  10,232  Philippines  91,638  Uruguay  6351  Czech Republic  3566  Jordan  2431  Poland  7951      Country  N  Country  N  Country  N  Country  N  Afghanistan  1172  Denmark  19,074  Kenya  6985  Portugal  8334  Albania  2288  Dominican R.  6222  Korea (South)  9960  Qatar  2064  Algeria  1074  Egypt  9075  Latvia  1440  Romania  13,055  Argentina  88,211  El Salvador  3682  Lithuania  2277  Russia  3624  Armenia  1084  Estonia  2459  Luxembourg  1116  Serbia and Montenegro  5665  Australia  195,857  Ethiopia  1008  Macedonia  1121  Singapore  59,119  Austria  27,143  Finland  23,526  Malaysia  39,606  Slovak Republic  1691  Bangladesh  3482  France  18,502  Malta  1590  Slovenia  3095  Belgium  43,692  Germany  186,848  Mauritius  1706  South Africa  26,039  Belize  1025  Ghana  1949  Mexico  136,305  Spain  135,048  Bolivia  6115  Greece  10,982  Morocco  1346  Sri Lanka  3958  Bosnia and Herzegovina  1371  Guatemala  5635  Nepal  2142  Sweden  46,828  Brazil  26,538  Hungary  3746  The Netherlands  163,472  Switzerland  36,741  Brunei  1211  India  114,500  New Zealand  43,167  Thailand  8501  Bulgaria  3610  Indonesia  15,199  Nigeria  7033  Turkey  5298  Canada  371,882  Iceland  2520  Norway  42,859  UAE  14,907  Chile  44,552  Ireland  41,257  Oman  1068  Uganda  1377  Colombia  34,905  Israel  7426  Pakistan  27,498  Ukraine  1081  Costa Rica  6712  Italy  13,831  Panama  2938  UK  438,854  Croatia  6920  Jamaica  4199  Perù  23,056  The United States  4,275,860  Cyprus  2307  Japan  10,232  Philippines  91,638  Uruguay  6351  Czech Republic  3566  Jordan  2431  Poland  7951      4.2.2 Constitutional protection As noted in Section 2.1 above, several provisions contained in national constitutions affect the dynamics of entrepreneurship. Information about constitutional provisions is drawn from the Comparative Constitutions Project: A Cross-National Historical Data set of Written Constitutions (henceforth CCP) (Elkins et al., 2009), an archive of data on the features of written constitutions for most countries since 1789. We focus on four provisions that represent how a constitution can protect the market mechanism, to derive an integrated variable by summing up: right to own property, right to conduct/establish a business, right to free/competitive markets, and independence of the judiciary organs. This Constitutional protection variable ranges from 0 (constitution not mentioning any of the four provisions) to 4 (constitution mentions all four provisions). Constitutional provisions represent the pillars of a country’s legal and institutional framework, and they should be enforced by “lower” laws, which are hierarchically subordinate to constitutions. Often, in fact, lawmakers enact new rules at the constitutional level as a commitment device to guarantee their application (Kelsen, 1967). For example, legal reforms increasing the protection of investors’ rights—and therefore consistent with constitutional protection of the free market—might lead to lower use of control enhancing mechanisms and ultimately create conditions more favorable to the emergence of a corporate economy dominated by widely held corporations (Cuomo et al., 2013). However, to control for their de facto implementation and to measure whether and to what extent “higher” constitutional norms are enforced by the legal and institutional framework, and effectively protect economic freedom, we need to measure the functioning of the market mechanism. For this purpose, we use the Index of Economic Freedom calculated by the Heritage Foundation, (http://www.heritage.org/index/). The index measures economic freedom based on four broad categories, each of which includes three or four types of economic freedom (in parentheses): rule of law (property rights, government integrity, judicial effectiveness), government size (government spending, tax burden, fiscal health), regulatory efficiency (business freedom, labor freedom, monetary freedom), and open markets (trade freedom, investment freedom, financial freedom). Each of the factors shaping the four broad categories is graded on a scale from 0 to 100, and a country’s score is obtained by averaging the resulting 12 values with equal weight given to each. 4.3 Control variables To control for the general economic foundations of each country, we consider the following set of control variables. To capture the wealth of countries and labor market characteristics, we use gross domestic product (GDP) per capita and the percentage of residents aged 15 years or more who are part of the labor force. Other control variables are electric consumption (in Kwh) per capita, as a proxy of the business cycle, and mobile cellular subscription per 100 residents as a proxy of the quality of the infrastructures. Table 3 presents variable descriptions and summary statistics (mean, standard deviation, minimum, and maximum) for all variables included in the analysis, and Table 4 presents the corresponding correlation matrix. Table 3. List of variables and their descriptive statistics: standard deviation is decomposed into between and within components Variable description    Code  Mean  Standard deviation  Minimum  Maximum  New Business density  Overall  Busdensity  5.606  4.899  0.0201  39.699  Between  4.683  0.027  25.826  Within  1.516  −2.981  19.479  Constitutional protection  Overall  Constprot  2.296  1.135  0  4  Between  1.122  0  4  Within  0.202  −0.108  5.091  Agency culture  Overall  Agency  3.477  0.0738  3.283  3.681  Between  0.0736  3.283  3.669  Within  0.0083  3.387  3.557  Economic freedom  Overall  Ecofreedom  63.851  10.652  21.7  89.7  Between  10.533  28.29  87.99  Within  1.864  56.89  73.091  GDP per capita (log)  Overall  Ln GDPcapita  9.307  1.077  5.855  11.212  Between  0.981  6.631  11.133  Within  0.454  6.664  11.604  Labor force participation rate (% of total population +15 years old)    Laborforce  61.921  9.685  37.1  87.7  Between    9.232  39.86  87.2  Within    3.051  41.62  72.08  Electric consumption per capita (Kwh)    Electriccon  5100  6210  30.4  54,799  Between  6135  32.57  43,751  Within  1120  −10,663  16,147  Mobile cellular subscription per 100 people  Overall  Mobilesup  93.814  39.913  0.2  217  Between  30.305  8.345  156.8  Within  26.127  −1.986  181.31  Variable description    Code  Mean  Standard deviation  Minimum  Maximum  New Business density  Overall  Busdensity  5.606  4.899  0.0201  39.699  Between  4.683  0.027  25.826  Within  1.516  −2.981  19.479  Constitutional protection  Overall  Constprot  2.296  1.135  0  4  Between  1.122  0  4  Within  0.202  −0.108  5.091  Agency culture  Overall  Agency  3.477  0.0738  3.283  3.681  Between  0.0736  3.283  3.669  Within  0.0083  3.387  3.557  Economic freedom  Overall  Ecofreedom  63.851  10.652  21.7  89.7  Between  10.533  28.29  87.99  Within  1.864  56.89  73.091  GDP per capita (log)  Overall  Ln GDPcapita  9.307  1.077  5.855  11.212  Between  0.981  6.631  11.133  Within  0.454  6.664  11.604  Labor force participation rate (% of total population +15 years old)    Laborforce  61.921  9.685  37.1  87.7  Between    9.232  39.86  87.2  Within    3.051  41.62  72.08  Electric consumption per capita (Kwh)    Electriccon  5100  6210  30.4  54,799  Between  6135  32.57  43,751  Within  1120  −10,663  16,147  Mobile cellular subscription per 100 people  Overall  Mobilesup  93.814  39.913  0.2  217  Between  30.305  8.345  156.8  Within  26.127  −1.986  181.31  Table 3. List of variables and their descriptive statistics: standard deviation is decomposed into between and within components Variable description    Code  Mean  Standard deviation  Minimum  Maximum  New Business density  Overall  Busdensity  5.606  4.899  0.0201  39.699  Between  4.683  0.027  25.826  Within  1.516  −2.981  19.479  Constitutional protection  Overall  Constprot  2.296  1.135  0  4  Between  1.122  0  4  Within  0.202  −0.108  5.091  Agency culture  Overall  Agency  3.477  0.0738  3.283  3.681  Between  0.0736  3.283  3.669  Within  0.0083  3.387  3.557  Economic freedom  Overall  Ecofreedom  63.851  10.652  21.7  89.7  Between  10.533  28.29  87.99  Within  1.864  56.89  73.091  GDP per capita (log)  Overall  Ln GDPcapita  9.307  1.077  5.855  11.212  Between  0.981  6.631  11.133  Within  0.454  6.664  11.604  Labor force participation rate (% of total population +15 years old)    Laborforce  61.921  9.685  37.1  87.7  Between    9.232  39.86  87.2  Within    3.051  41.62  72.08  Electric consumption per capita (Kwh)    Electriccon  5100  6210  30.4  54,799  Between  6135  32.57  43,751  Within  1120  −10,663  16,147  Mobile cellular subscription per 100 people  Overall  Mobilesup  93.814  39.913  0.2  217  Between  30.305  8.345  156.8  Within  26.127  −1.986  181.31  Variable description    Code  Mean  Standard deviation  Minimum  Maximum  New Business density  Overall  Busdensity  5.606  4.899  0.0201  39.699  Between  4.683  0.027  25.826  Within  1.516  −2.981  19.479  Constitutional protection  Overall  Constprot  2.296  1.135  0  4  Between  1.122  0  4  Within  0.202  −0.108  5.091  Agency culture  Overall  Agency  3.477  0.0738  3.283  3.681  Between  0.0736  3.283  3.669  Within  0.0083  3.387  3.557  Economic freedom  Overall  Ecofreedom  63.851  10.652  21.7  89.7  Between  10.533  28.29  87.99  Within  1.864  56.89  73.091  GDP per capita (log)  Overall  Ln GDPcapita  9.307  1.077  5.855  11.212  Between  0.981  6.631  11.133  Within  0.454  6.664  11.604  Labor force participation rate (% of total population +15 years old)    Laborforce  61.921  9.685  37.1  87.7  Between    9.232  39.86  87.2  Within    3.051  41.62  72.08  Electric consumption per capita (Kwh)    Electriccon  5100  6210  30.4  54,799  Between  6135  32.57  43,751  Within  1120  −10,663  16,147  Mobile cellular subscription per 100 people  Overall  Mobilesup  93.814  39.913  0.2  217  Between  30.305  8.345  156.8  Within  26.127  −1.986  181.31  Table 4. Pairwise correlation matrix (86 countries: average values) Variables  Business density  Consprot  Agency culture  Economic freedom  GDP per capita  Labor force  Electric consump  Mobile subscript  Busdensity  1.0000                Constprot  0.2537*  1.0000              Agency culture  0.2286*  0.2929*  1.0000            Ecofreedom  0.4087*  −0.0938*  −0.0120  1.0000          GDP capita  0.3218*  −0.0026  0.1395*  0.5881*  1.0000        Labor force  0.0892*  −0.0794  −0.1204*  0.0482  −0.0252  1.0000      Electconsum  0.2736*  −0.1627*  −0.0712  0.4574*  0.5637*  0.2294*  1.0000    Mobile subscript  0.2990*  0.0959*  0.0327  0.3813*  0.5021*  −0.0160  0.3423*  1.0000  Variables  Business density  Consprot  Agency culture  Economic freedom  GDP per capita  Labor force  Electric consump  Mobile subscript  Busdensity  1.0000                Constprot  0.2537*  1.0000              Agency culture  0.2286*  0.2929*  1.0000            Ecofreedom  0.4087*  −0.0938*  −0.0120  1.0000          GDP capita  0.3218*  −0.0026  0.1395*  0.5881*  1.0000        Labor force  0.0892*  −0.0794  −0.1204*  0.0482  −0.0252  1.0000      Electconsum  0.2736*  −0.1627*  −0.0712  0.4574*  0.5637*  0.2294*  1.0000    Mobile subscript  0.2990*  0.0959*  0.0327  0.3813*  0.5021*  −0.0160  0.3423*  1.0000  * : Significant at 1% level. Table 4. Pairwise correlation matrix (86 countries: average values) Variables  Business density  Consprot  Agency culture  Economic freedom  GDP per capita  Labor force  Electric consump  Mobile subscript  Busdensity  1.0000                Constprot  0.2537*  1.0000              Agency culture  0.2286*  0.2929*  1.0000            Ecofreedom  0.4087*  −0.0938*  −0.0120  1.0000          GDP capita  0.3218*  −0.0026  0.1395*  0.5881*  1.0000        Labor force  0.0892*  −0.0794  −0.1204*  0.0482  −0.0252  1.0000      Electconsum  0.2736*  −0.1627*  −0.0712  0.4574*  0.5637*  0.2294*  1.0000    Mobile subscript  0.2990*  0.0959*  0.0327  0.3813*  0.5021*  −0.0160  0.3423*  1.0000  Variables  Business density  Consprot  Agency culture  Economic freedom  GDP per capita  Labor force  Electric consump  Mobile subscript  Busdensity  1.0000                Constprot  0.2537*  1.0000              Agency culture  0.2286*  0.2929*  1.0000            Ecofreedom  0.4087*  −0.0938*  −0.0120  1.0000          GDP capita  0.3218*  −0.0026  0.1395*  0.5881*  1.0000        Labor force  0.0892*  −0.0794  −0.1204*  0.0482  −0.0252  1.0000      Electconsum  0.2736*  −0.1627*  −0.0712  0.4574*  0.5637*  0.2294*  1.0000    Mobile subscript  0.2990*  0.0959*  0.0327  0.3813*  0.5021*  −0.0160  0.3423*  1.0000  * : Significant at 1% level. 5. Model development For the purposes of our empirical analysis, we opted for a parsimonious specification, hypothesizing the following structural model:   Yit=α+βCit+γAi+δCit*Ai+θXit+εit, where Yit denotes new business density of country i in year t; Cit is an indicator of constitutional protection of economic freedom and rights in country i and year t; Ai represents a measure of agency culture for country i;  δCit*Ai is the interaction between the constitutional and the psychological variable; Xit is a set of other control variables; and εit is the usual error term. The interaction term is added to the model for testing the hypotheses that the impact exerted by the relationship between endowment of agency culture and the strength of constitutional protection was different for different levels of agency and constitutional protection. The Breusch–Pagan test indicates the presence of heteroskedasticity.1 The White’s method of correcting for heteroskedastic errors should then be applied. The Wooldridge test for autocorrelation in panel data also indicates the presence of serial correlation in our data set2. The Hausman test reveals the existence of an endogeneity problem for our constitutional variable3. Constitution is likely to be endogenous because economies are not exogenously endowed with the institutions and incentives that make up their entrepreneurial environment, but rather institutions are determined endogenously, perhaps influenced by the history, geographical features, and level of entrepreneurship in an economy. The presence of heteroskedasticity, serial correlation, and endogeneity in our data set deserves careful treatment in choosing an appropriate estimation model. On the one hand, robust pooled ordinary least squares (OLS) estimation fails to give unbiased and efficient estimators, and instrumental variable two-staged least square (2SLS) could be a wise choice. On the other hand, our data incur the problem of heteroskedasticity, so we apply the IV generalized method of moments (GMM) technique, which gives more reliable and consistent estimation results (Baum and Schaffer, 2003). The IV-GMM treatment requires the availability and validity of exogenous instruments that are correlated with the independent variables for which endogeneity has been detected, but that are uncorrelated with the measure of new business density. We adopt two instrumental variables: the distance from the equator used by Hall and Jones (1999) and the predicted trade share of an economy constructed by Frankel and Romer (1996). The underidentification test and the Sargan test to detect the relevance and validity of our IVs do support our approach, and thus our choice of instruments is plausible (see Table 6). As robustness checks, we estimate three extra models. We estimate a dynamic Blundell–Bond (Blundell and Bond, 1998) GMM model, including the lagged dependent variable to consider the potential effect of the business cycle and the lagged value of constitutional protection to account for institutional change. This model allows for a low-order moving average correlation in the idiosyncratic errors and is well suited to deal with the low variance in the process of constitutional change, with the time-invariant nature of the agency culture variable, and with the relatively small longitudinal length of the data set (only one decade). Moreover, to account for unobserved country effects across time, the third and the fourth models are country-fixed-effects OLS and generalized least squares (GLS). 6. Empirical results and discussion 6.1 Regression results We start by estimating regressions with Agency culture as the main and only explanatory variable. The results from the dynamic Blundell–Bond (Blundell and Bond, 1998) GMM, the country-fixed-effects OLS, and the country-fixed-effects GLS models presented in Table 5 show a positive and highly statistically significant impact of stronger agency culture on our measure of entrepreneurship: the first part of H1a is therefore supported. Higher labor force participation rate and better infrastructures (i.e., more widespread adoption of mobile cellular phones) are also associated with higher levels of new business density. Table 5. Agency culture and entrepreneurship Variables  Dynamic Blundell–Bond (Blundell and Bond, 1998) GMM  Country-FE with robust SEs  GLS-RE with robust SEs  New Business Density, t−1  0.592***      (0.033)  Agency culture  60.630***  85.985***  61.376***  (5.433)  (15.421)  (13.351)  Economic freedom  0.127***  −0.016  0.027  (0.043)  (0.031)  (0.032)  GDP per capita  −0.625  −1.125  −0.915  (1.841)  (1.701)  (1.763)  GDP per capita squared  0.042  0.059  0.050  (0.101)  (0.101)  (0.097)  Labor force participation  0.005  0.0687***  0.077***  (0.022)  (0.032)  (0.032)  Electric power consumption  −0.001  −0.001***  −0.001***  (0.001)  (0.001)  (0.003)  Mobile cellular subscription  0.002  0.11***  0.011***  (0.003)  (0.004)  (0.004)  Intercept  −215.11***  −291.13***  −210.47***  (20.103)  (53.87)  (47.03)    F-test    33.33***    Wald statistics χ2  2212.50***    40.40***  Observations  774  860  860  Variables  Dynamic Blundell–Bond (Blundell and Bond, 1998) GMM  Country-FE with robust SEs  GLS-RE with robust SEs  New Business Density, t−1  0.592***      (0.033)  Agency culture  60.630***  85.985***  61.376***  (5.433)  (15.421)  (13.351)  Economic freedom  0.127***  −0.016  0.027  (0.043)  (0.031)  (0.032)  GDP per capita  −0.625  −1.125  −0.915  (1.841)  (1.701)  (1.763)  GDP per capita squared  0.042  0.059  0.050  (0.101)  (0.101)  (0.097)  Labor force participation  0.005  0.0687***  0.077***  (0.022)  (0.032)  (0.032)  Electric power consumption  −0.001  −0.001***  −0.001***  (0.001)  (0.001)  (0.003)  Mobile cellular subscription  0.002  0.11***  0.011***  (0.003)  (0.004)  (0.004)  Intercept  −215.11***  −291.13***  −210.47***  (20.103)  (53.87)  (47.03)    F-test    33.33***    Wald statistics χ2  2212.50***    40.40***  Observations  774  860  860  Standard errors in brackets. ***: Significant at 10% level. Table 5. Agency culture and entrepreneurship Variables  Dynamic Blundell–Bond (Blundell and Bond, 1998) GMM  Country-FE with robust SEs  GLS-RE with robust SEs  New Business Density, t−1  0.592***      (0.033)  Agency culture  60.630***  85.985***  61.376***  (5.433)  (15.421)  (13.351)  Economic freedom  0.127***  −0.016  0.027  (0.043)  (0.031)  (0.032)  GDP per capita  −0.625  −1.125  −0.915  (1.841)  (1.701)  (1.763)  GDP per capita squared  0.042  0.059  0.050  (0.101)  (0.101)  (0.097)  Labor force participation  0.005  0.0687***  0.077***  (0.022)  (0.032)  (0.032)  Electric power consumption  −0.001  −0.001***  −0.001***  (0.001)  (0.001)  (0.003)  Mobile cellular subscription  0.002  0.11***  0.011***  (0.003)  (0.004)  (0.004)  Intercept  −215.11***  −291.13***  −210.47***  (20.103)  (53.87)  (47.03)    F-test    33.33***    Wald statistics χ2  2212.50***    40.40***  Observations  774  860  860  Variables  Dynamic Blundell–Bond (Blundell and Bond, 1998) GMM  Country-FE with robust SEs  GLS-RE with robust SEs  New Business Density, t−1  0.592***      (0.033)  Agency culture  60.630***  85.985***  61.376***  (5.433)  (15.421)  (13.351)  Economic freedom  0.127***  −0.016  0.027  (0.043)  (0.031)  (0.032)  GDP per capita  −0.625  −1.125  −0.915  (1.841)  (1.701)  (1.763)  GDP per capita squared  0.042  0.059  0.050  (0.101)  (0.101)  (0.097)  Labor force participation  0.005  0.0687***  0.077***  (0.022)  (0.032)  (0.032)  Electric power consumption  −0.001  −0.001***  −0.001***  (0.001)  (0.001)  (0.003)  Mobile cellular subscription  0.002  0.11***  0.011***  (0.003)  (0.004)  (0.004)  Intercept  −215.11***  −291.13***  −210.47***  (20.103)  (53.87)  (47.03)    F-test    33.33***    Wald statistics χ2  2212.50***    40.40***  Observations  774  860  860  Standard errors in brackets. ***: Significant at 10% level. We then turn to the discussion of the general model (Table 6). As far as the main variables of interest (Agency culture and Constitutional protection) are concerned, the results of the estimates show a consistent pattern across the IV-GMM, the dynamic Blundell–Bond (Blundell and Bond, 1998) GMM, the country-fixed-effects OLS, and the country-fixed-effects GLS models. In Table 6 we present three specifications for each of the four methodological treatments: the first specification controls for the effect of constitutional protection only (Columns 1, 4, and 7); the second specification considers both constitutional protection and psychological agency culture (Columns 2, 5, and 8); and the third specification takes into account their interaction effect as well (Columns 3, 6, and 9). Our data incur the problem of heteroskedasticity, serial autocorrelation, and endogeneity of constitutional protection, so the static IV-GMM model with robust SEs is the most appropriate estimation method; thus, we base our interpretation on the results of this model. Results from the other model specifications are also traced for the purpose of comparison. Table 6. Agency culture, constitutional protection, and entrepreneurship Dependent variable: new business density   Variables  IV-GMM with robust SEsa   Dynamic Blundell–Bond (Blundell and Bond, 1998) GMMb   Country-FE with robust SEs   GLS-RE with rob SEs  (1)  (2)  (3)  (4)  (5)  (6)  (7)  (8)  (9)  New business density, t − 1        0.707***  0.611***  0.633***          (0.031)  (0.031)  (0.031)  Constitutional protection  0.773**  0.635  −2.955***  2.627***  2.146***  0.217  2.185***  2.264***  9.174  −6.084***  (0.346)  (0.401)  (1.059)  (0.327)  (0.302)  (0.385)  (0.605)  (0.606)  (27.515)  (1.937)  Constitutional protection, t − 1        0.106  −0.295  0.51          (0.411)  (0.376)  (0.395)  Agency culture    6.583**  −1.59**    39.354***  18.098***    86.766***  94.795***  −5.329  (3.577)  (0.645)    (4.705)  (5.774)    (15.753)  (35.701)  (16.535)  Constitutional protection * psychological agency      8.884***      0.581***      −2.008  18.089***  (3.049)      (0.089)      (7.998)  (5.566)  Economic freedom  0.225***  0.214***  0.222***  0.149***  0.094***  0.131***  0.030  0.012  0.011  0.043  (0.02)  (0.022)  (0.019)  (0.041)  (0.038)  (0.038)  (0.034)  (0.028)  (0.028)  (0.031)  GDP per capita  5.26***  4.876***  2.598  1.603  0.644  1.235  0.848  1.458  1.484  0.984  (2.21)  (2.06)  (1.838)  (1.893)  (1.721)  (1.736)  (1.814)  (1.654)  (1.65)  (1.825)  GDP per capita squared  −0.301***  −0.282***  −0.156  0.106  0.044  0.073  0.039  0.071  0.073  0.051  (0.122)  (0.114)  (0.102)  (0.104)  (0.095)  (0.095)  (0.099)  (0.091)  (0.091)  (0.1003)  Labor force participation rate  0.005  0.012  0.008  0.022  0.002  0.008  0.061**  0.055**  0.055**  0.062**  (0.015)  (0.017)  (0.012)  (0.022)  (0.021)  (0.021)  (0.027)  (0.026)  (0.026)  (0.027)  Electric power consumption per capita (Kwh)  0.0001***  0.0001***  0.0001***  0.0001***  0.0001***  0.000  0.0002***  0.0002***  0.0002***  0.0001**  (0.000)  (0.000)  (0.000)  (0.0001)  (0.000)  (0.000)  (0.000)  (0.000)  (0.000)  (0.000)  Mobile cellular subscription per 100 people  0.0105***  0.011***  0.008**  0.001  0.001  0.001  0.008**  0.007**  0.007**  0.007**  (0.004)  (0.004)  (0.0039)  (0.002)  (0.002)  (0.002)  (0.003)  (0.003)  (0.003)  (0.003)  Intercept  −35.49***  −55.95***  31.358  −6.980  −142.94***  −70.531***  3.524  −29.65***  −32.414***  16.500  (9.783)  (17.21)  (24.284)  (8.681)  (18.164)  (21.477)  (8.996)  (5.499)  (12.345)  (58.165)    F test  51.03***  44.02***  44.46***        15.61***  40.93***  40.46***    Wald statistic χ2(2)        2095.9***  2627.6***  2631.8***        46.61***  Under-identification test χ2(2)  72.44***  105.3***  64.64***                Over-identification test χ2(1)  4.926  4.618  0.815                Endogeneity test χ2(1)  2.78*  3.006*  6.759***                Observations  860  860  860  774  774  774  860  860  860  860  Dependent variable: new business density   Variables  IV-GMM with robust SEsa   Dynamic Blundell–Bond (Blundell and Bond, 1998) GMMb   Country-FE with robust SEs   GLS-RE with rob SEs  (1)  (2)  (3)  (4)  (5)  (6)  (7)  (8)  (9)  New business density, t − 1        0.707***  0.611***  0.633***          (0.031)  (0.031)  (0.031)  Constitutional protection  0.773**  0.635  −2.955***  2.627***  2.146***  0.217  2.185***  2.264***  9.174  −6.084***  (0.346)  (0.401)  (1.059)  (0.327)  (0.302)  (0.385)  (0.605)  (0.606)  (27.515)  (1.937)  Constitutional protection, t − 1        0.106  −0.295  0.51          (0.411)  (0.376)  (0.395)  Agency culture    6.583**  −1.59**    39.354***  18.098***    86.766***  94.795***  −5.329  (3.577)  (0.645)    (4.705)  (5.774)    (15.753)  (35.701)  (16.535)  Constitutional protection * psychological agency      8.884***      0.581***      −2.008  18.089***  (3.049)      (0.089)      (7.998)  (5.566)  Economic freedom  0.225***  0.214***  0.222***  0.149***  0.094***  0.131***  0.030  0.012  0.011  0.043  (0.02)  (0.022)  (0.019)  (0.041)  (0.038)  (0.038)  (0.034)  (0.028)  (0.028)  (0.031)  GDP per capita  5.26***  4.876***  2.598  1.603  0.644  1.235  0.848  1.458  1.484  0.984  (2.21)  (2.06)  (1.838)  (1.893)  (1.721)  (1.736)  (1.814)  (1.654)  (1.65)  (1.825)  GDP per capita squared  −0.301***  −0.282***  −0.156  0.106  0.044  0.073  0.039  0.071  0.073  0.051  (0.122)  (0.114)  (0.102)  (0.104)  (0.095)  (0.095)  (0.099)  (0.091)  (0.091)  (0.1003)  Labor force participation rate  0.005  0.012  0.008  0.022  0.002  0.008  0.061**  0.055**  0.055**  0.062**  (0.015)  (0.017)  (0.012)  (0.022)  (0.021)  (0.021)  (0.027)  (0.026)  (0.026)  (0.027)  Electric power consumption per capita (Kwh)  0.0001***  0.0001***  0.0001***  0.0001***  0.0001***  0.000  0.0002***  0.0002***  0.0002***  0.0001**  (0.000)  (0.000)  (0.000)  (0.0001)  (0.000)  (0.000)  (0.000)  (0.000)  (0.000)  (0.000)  Mobile cellular subscription per 100 people  0.0105***  0.011***  0.008**  0.001  0.001  0.001  0.008**  0.007**  0.007**  0.007**  (0.004)  (0.004)  (0.0039)  (0.002)  (0.002)  (0.002)  (0.003)  (0.003)  (0.003)  (0.003)  Intercept  −35.49***  −55.95***  31.358  −6.980  −142.94***  −70.531***  3.524  −29.65***  −32.414***  16.500  (9.783)  (17.21)  (24.284)  (8.681)  (18.164)  (21.477)  (8.996)  (5.499)  (12.345)  (58.165)    F test  51.03***  44.02***  44.46***        15.61***  40.93***  40.46***    Wald statistic χ2(2)        2095.9***  2627.6***  2631.8***        46.61***  Under-identification test χ2(2)  72.44***  105.3***  64.64***                Over-identification test χ2(1)  4.926  4.618  0.815                Endogeneity test χ2(1)  2.78*  3.006*  6.759***                Observations  860  860  860  774  774  774  860  860  860  860  *, **, *** Note: : Significant at 10%, 5%, 1% significant level, respectively. (1), (4), (7): Constitutional protection is controlled; (2), (5), (8): constitutional protection and psychological ‘agency’ trait are controlled; (3), (6), (9): constitutional protection, psychological ‘agency’ trait, and their interaction are controlled. a Since at least one of the two instruments should vary over time (i.e., the trade share of the economy), the IV GMM model could overcome the time invariance in the regressors of interest. b Constitutional protection is treated as an endogenous variable. The first lagged value of constitutional protection is used as IV. Table 6. Agency culture, constitutional protection, and entrepreneurship Dependent variable: new business density   Variables  IV-GMM with robust SEsa   Dynamic Blundell–Bond (Blundell and Bond, 1998) GMMb   Country-FE with robust SEs   GLS-RE with rob SEs  (1)  (2)  (3)  (4)  (5)  (6)  (7)  (8)  (9)  New business density, t − 1        0.707***  0.611***  0.633***          (0.031)  (0.031)  (0.031)  Constitutional protection  0.773**  0.635  −2.955***  2.627***  2.146***  0.217  2.185***  2.264***  9.174  −6.084***  (0.346)  (0.401)  (1.059)  (0.327)  (0.302)  (0.385)  (0.605)  (0.606)  (27.515)  (1.937)  Constitutional protection, t − 1        0.106  −0.295  0.51          (0.411)  (0.376)  (0.395)  Agency culture    6.583**  −1.59**    39.354***  18.098***    86.766***  94.795***  −5.329  (3.577)  (0.645)    (4.705)  (5.774)    (15.753)  (35.701)  (16.535)  Constitutional protection * psychological agency      8.884***      0.581***      −2.008  18.089***  (3.049)      (0.089)      (7.998)  (5.566)  Economic freedom  0.225***  0.214***  0.222***  0.149***  0.094***  0.131***  0.030  0.012  0.011  0.043  (0.02)  (0.022)  (0.019)  (0.041)  (0.038)  (0.038)  (0.034)  (0.028)  (0.028)  (0.031)  GDP per capita  5.26***  4.876***  2.598  1.603  0.644  1.235  0.848  1.458  1.484  0.984  (2.21)  (2.06)  (1.838)  (1.893)  (1.721)  (1.736)  (1.814)  (1.654)  (1.65)  (1.825)  GDP per capita squared  −0.301***  −0.282***  −0.156  0.106  0.044  0.073  0.039  0.071  0.073  0.051  (0.122)  (0.114)  (0.102)  (0.104)  (0.095)  (0.095)  (0.099)  (0.091)  (0.091)  (0.1003)  Labor force participation rate  0.005  0.012  0.008  0.022  0.002  0.008  0.061**  0.055**  0.055**  0.062**  (0.015)  (0.017)  (0.012)  (0.022)  (0.021)  (0.021)  (0.027)  (0.026)  (0.026)  (0.027)  Electric power consumption per capita (Kwh)  0.0001***  0.0001***  0.0001***  0.0001***  0.0001***  0.000  0.0002***  0.0002***  0.0002***  0.0001**  (0.000)  (0.000)  (0.000)  (0.0001)  (0.000)  (0.000)  (0.000)  (0.000)  (0.000)  (0.000)  Mobile cellular subscription per 100 people  0.0105***  0.011***  0.008**  0.001  0.001  0.001  0.008**  0.007**  0.007**  0.007**  (0.004)  (0.004)  (0.0039)  (0.002)  (0.002)  (0.002)  (0.003)  (0.003)  (0.003)  (0.003)  Intercept  −35.49***  −55.95***  31.358  −6.980  −142.94***  −70.531***  3.524  −29.65***  −32.414***  16.500  (9.783)  (17.21)  (24.284)  (8.681)  (18.164)  (21.477)  (8.996)  (5.499)  (12.345)  (58.165)    F test  51.03***  44.02***  44.46***        15.61***  40.93***  40.46***    Wald statistic χ2(2)        2095.9***  2627.6***  2631.8***        46.61***  Under-identification test χ2(2)  72.44***  105.3***  64.64***                Over-identification test χ2(1)  4.926  4.618  0.815                Endogeneity test χ2(1)  2.78*  3.006*  6.759***                Observations  860  860  860  774  774  774  860  860  860  860  Dependent variable: new business density   Variables  IV-GMM with robust SEsa   Dynamic Blundell–Bond (Blundell and Bond, 1998) GMMb   Country-FE with robust SEs   GLS-RE with rob SEs  (1)  (2)  (3)  (4)  (5)  (6)  (7)  (8)  (9)  New business density, t − 1        0.707***  0.611***  0.633***          (0.031)  (0.031)  (0.031)  Constitutional protection  0.773**  0.635  −2.955***  2.627***  2.146***  0.217  2.185***  2.264***  9.174  −6.084***  (0.346)  (0.401)  (1.059)  (0.327)  (0.302)  (0.385)  (0.605)  (0.606)  (27.515)  (1.937)  Constitutional protection, t − 1        0.106  −0.295  0.51          (0.411)  (0.376)  (0.395)  Agency culture    6.583**  −1.59**    39.354***  18.098***    86.766***  94.795***  −5.329  (3.577)  (0.645)    (4.705)  (5.774)    (15.753)  (35.701)  (16.535)  Constitutional protection * psychological agency      8.884***      0.581***      −2.008  18.089***  (3.049)      (0.089)      (7.998)  (5.566)  Economic freedom  0.225***  0.214***  0.222***  0.149***  0.094***  0.131***  0.030  0.012  0.011  0.043  (0.02)  (0.022)  (0.019)  (0.041)  (0.038)  (0.038)  (0.034)  (0.028)  (0.028)  (0.031)  GDP per capita  5.26***  4.876***  2.598  1.603  0.644  1.235  0.848  1.458  1.484  0.984  (2.21)  (2.06)  (1.838)  (1.893)  (1.721)  (1.736)  (1.814)  (1.654)  (1.65)  (1.825)  GDP per capita squared  −0.301***  −0.282***  −0.156  0.106  0.044  0.073  0.039  0.071  0.073  0.051  (0.122)  (0.114)  (0.102)  (0.104)  (0.095)  (0.095)  (0.099)  (0.091)  (0.091)  (0.1003)  Labor force participation rate  0.005  0.012  0.008  0.022  0.002  0.008  0.061**  0.055**  0.055**  0.062**  (0.015)  (0.017)  (0.012)  (0.022)  (0.021)  (0.021)  (0.027)  (0.026)  (0.026)  (0.027)  Electric power consumption per capita (Kwh)  0.0001***  0.0001***  0.0001***  0.0001***  0.0001***  0.000  0.0002***  0.0002***  0.0002***  0.0001**  (0.000)  (0.000)  (0.000)  (0.0001)  (0.000)  (0.000)  (0.000)  (0.000)  (0.000)  (0.000)  Mobile cellular subscription per 100 people  0.0105***  0.011***  0.008**  0.001  0.001  0.001  0.008**  0.007**  0.007**  0.007**  (0.004)  (0.004)  (0.0039)  (0.002)  (0.002)  (0.002)  (0.003)  (0.003)  (0.003)  (0.003)  Intercept  −35.49***  −55.95***  31.358  −6.980  −142.94***  −70.531***  3.524  −29.65***  −32.414***  16.500  (9.783)  (17.21)  (24.284)  (8.681)  (18.164)  (21.477)  (8.996)  (5.499)  (12.345)  (58.165)    F test  51.03***  44.02***  44.46***        15.61***  40.93***  40.46***    Wald statistic χ2(2)        2095.9***  2627.6***  2631.8***        46.61***  Under-identification test χ2(2)  72.44***  105.3***  64.64***                Over-identification test χ2(1)  4.926  4.618  0.815                Endogeneity test χ2(1)  2.78*  3.006*  6.759***                Observations  860  860  860  774  774  774  860  860  860  860  *, **, *** Note: : Significant at 10%, 5%, 1% significant level, respectively. (1), (4), (7): Constitutional protection is controlled; (2), (5), (8): constitutional protection and psychological ‘agency’ trait are controlled; (3), (6), (9): constitutional protection, psychological ‘agency’ trait, and their interaction are controlled. a Since at least one of the two instruments should vary over time (i.e., the trade share of the economy), the IV GMM model could overcome the time invariance in the regressors of interest. b Constitutional protection is treated as an endogenous variable. The first lagged value of constitutional protection is used as IV. Our findings show that constitutional protection of economic freedom plays a key role in generating the wide variation in entrepreneurship across countries (Columns 1, 4 and 7). The second part of H1a is therefore confirmed. Consistent with the findings of Bell et al. (2008) and Parker (2009: Ch. 15) on legal rules, we too find that when institutions support economic freedom, entrepreneurship is stronger. With respect to diffusion in each country’s population of the agency culture (Columns 2, 5 and 8), in all estimates the coefficient is largely positive and statistically significant at a 99% confidence level. A stronger agency culture is thus associated with a higher level of new business density, and corresponds to a greater propensity of individuals to create their own entrepreneurial ventures. Unsurprisingly, countries with high agency culture foster the development of a more dynamic entrepreneurial climate. Our findings therefore suggest a possible way of promoting the propensity to become entrepreneurs in the overall population through training activities for elating agentic characteristics, such as creativity, risk-taking propensity, and entrepreneurial proactiveness. However, when the impact of the psychological “agency” trait is taken into account, the institutional factor, despite having positive influence, loses its statistical significance on new business density in the IV-GMM while maintaining it in all other model specifications. Looking at the interaction between agency culture and constitutional protection (Columns 3, 6 and 9), consistent with the bounded agency approach, and with our H1b, the effect of agency culture on entrepreneurship is indeed bounded by the characteristics of the national constitutions. These two factors do combine in such a way that the effect of high agency culture is strengthened in countries with a pro-entrepreneurship constitution. Similarly, the constitutional protection of economic freedom leverages the entrepreneurial intention of countries with a great endowment of agency culture. A healthy business environment facilitates nascent entrepreneurs to discover and exploit entrepreneurial opportunities. The positive effect of constitutional protection and agency culture is also corroborated by the positive and statistically significant coefficient of our de facto measure, the Index of Economic Freedom, in both GMM estimates and by a positive, albeit insignificant, effect in the other model specifications. The impact of economic freedom is in line with the effect of constitutional variables promoting economic rights. A free and competitive market creates a favorable business environment and a level playing field to both incumbent and nascent entrepreneurs. With respect to control variables, there are several findings of note. First, in the IV-GMM, GDP per capita is positive and statistically significant, whereas its squared value is negative and statistically significant. The positive and negative signs are maintained in all other model specifications, although statistical significance disappears. This pattern is consistent with the findings from many empirical studies (Koellinger and Thurik, 2012) and confirms that, on the aggregate level, GDP cycles do predict the entrepreneurial cycle, although the relationship between GDP per capita and entrepreneurship is quadratic. In general, high GDP per capita reflects stronger demands, which leads to an abundance of emerging entrepreneurial opportunities, and in turn induces new entries to capture such opportunities (Santarelli and Tran, 2013). Second, a higher labor force participation rate is conducive to more entrepreneurship. Provided that a large fraction of new entrepreneurs is usually represented by individuals previously involved in paid employment (Storey and Jones, 1987), this last finding is straightforward. Third, countries with high electric power consumption per capita and mobile cellular subscription are more likely to enjoy a higher rate of new firm formation. 6.2 Illustrating the interaction effect between agency culture and constitutional protection Based on estimation of the baseline equation by GLS random-effects (RE) with a robust standard errors model (whose estimation results are presented in the final column of Table 6), we used the command “margins” in Stata 14 to estimate the margins of responses for specified values of covariates and present the results as a table. Finally, to draw the interaction plot we used the command “marginsplot” to graph the results of the “margins” command. Plots were also constructed at each of the five specified values of the “constitutional protection” variable from 0 to 4. Figure 1 presents the interaction plots of constitutional protection and agency culture, with the aim of capturing the moderating effect of constitutional provisions on psychological agency on a country’s proneness toward entrepreneurship. Figure 1. View largeDownload slide Interaction plot of constitutional protection and agency culture. Figure 1. View largeDownload slide Interaction plot of constitutional protection and agency culture. The plot on the right combines all five interaction lines, whereas the five smaller plots on the left separate interaction lines for each value of the constitutional protection variable. The combined interaction plot on the right has five lines representing the indirect effect of agency culture on new business density at five different values of constitutional protection (from 0 to 4). For ease of comparison, the interaction plot on the left separates these five interaction lines into five charts. Obviously, if the interaction is not significant, the plotted lines should be parallel, which is clearly not the case here. If a country does not have any constitutional provision supporting economic freedom (constprot = 0), agency culture is negatively associated with new business density. No matter how proactive and innovative its citizens are, they are just simply unmotivated to set up new ventures in an environment where they cannot enjoy free competition or the freedom to set up business and own their property. Second, when the country increases its constitutional protection of economic freedom (constprot increases from 1 to 3), the line representing the relationship between agency culture and entrepreneurship moves up gradually, while the impact of agency turns from negative to slightly positive. Countries with higher psychological agency find themselves more entrepreneurial with a higher rate of self-employment over time when they start to apply constitutional provision protecting economic freedom. Finally, the effect of psychological agency turns out significantly positive when countries possess a high level of constitutional protection (constprot = 4). These results suggest that high agentic economies enjoying a constitutional protection of economic freedom are particularly powerful cradles, nurturing entrepreneurial activity. Creative and innovative people are more motivated to exploit their ideas in a transparent and healthy business environment in which they do not need to care about bribes or corruption. Thus, if governments find that the majority of the country’s population has high levels of agency, they could help to capitalize on this entrepreneurial trait by creating a healthy institutional environment supporting free competition, business and property rights, and an independent judicial system. These steps would significantly foster a dynamic entrepreneurial sector within the economy. 7. Conclusions Inspired by the central role that the concept of (psychological) agency plays in seminal theorizing in entrepreneurship (Schumpeter, 1911, 1934; McClelland, 1961), the present study is to our knowledge, the first systematic attempt to examine the effect of agency culture on national entrepreneurship rates, with a special focus on an important contextual moderator—formal institutions such as constitutional provisions relevant for entrepreneurship. In general, our findings reveal that: (i) a greater endowment of agency culture is associated with a country’s higher willingness or intention to start a business; (ii) constitutional protection of economic freedom plays a key role in generating the observed wide variation in entrepreneurship across countries, by exerting a moderating effect on how a certain endowment of agency culture influences a country’s proneness toward entrepreneurship; (iii) when institutions do support economic freedom, as denoted by higher values of the Index of Economic Freedom, entrepreneurship is stronger. In sum, there is an interaction between constitutional and legal protection of economic freedom on the one side and the presence in the country of a large fraction of individuals characterized by an agentic personality on the other side. In particular, there seems to be a benefit from instituting stronger protections of economic freedom, and such benefit is the stronger the higher the level of agency culture characterizing the country. Footnotes 1 χ2(1) = 142.17; P-value = 0.0000. 2 F(1, 85) = 15.944; P-value = 0.0001. 3 χ2(2) = 12.982; P-value = 0.0015. Acknowledgments The authors thank Tom Ginsburg for providing us with the data from the Comparative Constitutions Project, Jerg Gutmann and an anonymous reviewer for helpful suggestions. Previous versions of this article have been presented at the Workshop on The Future of Small Business Economics in Utrecht (February 10, 2017), the Midterm Law and Economics Conference in Ghent (February 17, 2017), and the 34th Annual Conference of the European Association of Law and Economics (London, September 14–16, 2017). E.S. gratefully acknowledges financial support from Saarland University within the framework of the program “EXIST-Culture of Entrepreneurship—The Entrepreneurship University” arranged by the German Federal Ministry of Economics and Technology. References Abele A. E., Wojciszke B. ( 2007), ‘ Agency and communion from the perspective of self versus others,’ Journal of Personality and Social Psychology , 93( 5), 751– 763. Google Scholar CrossRef Search ADS   Acemoglu D., Jackson M. O. ( 2016), ‘Social norms and the enforcement of laws,’ MIT Department of Economics Working Paper No. 14-16. Acemoglu D., Johnson S., Robinson J. A. ( 2001), ‘ The colonial origins of comparative development: an empirical investigation,’ American Economic Review , 91( 5), 1369– 1401. Google Scholar CrossRef Search ADS   Acemoglu D., Robinson J. A. ( 2012), Why Nations Fail . Crown Business: New York, NY. Acs Z. J., Audretsch D. B., Braunerhjelm P., Carlsson B. ( 2009), ‘ The knowledge spillover theory of entrepreneurship,’ Small Business Economics , 32( 1), 15– 30. Google Scholar CrossRef Search ADS   Acs Z. J., Desai S., Hessels J. ( 2008), ‘ Entrepreneurship, economic development and institutions,’ Small Business Economics , 31( 3), 219– 234. Google Scholar CrossRef Search ADS   Ajzen I. ( 1985), ‘From intentions to actions: a theory of planned behavior,’ in Kuhl J., Beckman J. (eds), Action-Control: From Cognition to Behavior . Springer: Heidelberg, pp. 11– 39. Alesina A., Giuliano P. ( 2015), ‘ Culture and institutions,’ Journal of Economic Literature , 53( 4), 898– 944. Google Scholar CrossRef Search ADS   Audretsch D. B., Fritsch M. ( 1994), ‘ The geography of firm births in Germany,’ Regional Studies , 28( 4), 359– 365. Google Scholar CrossRef Search ADS   Audretsch D. B., Obschonka M., Gosling S. D., Potter J. ( 2017), ‘ A new perspective on entrepreneurial regions: linking cultural identity with latent and manifest entrepreneurship,’ Small Business Economics , 48( 3), 681– 697. Google Scholar CrossRef Search ADS   Audretsch D. B., Thurik A. R. ( 2000), ‘ Capitalism and democracy in the 21st century: from the managed to the entrepreneurial economy,’ Journal of Evolutionary Economics , 10( 1–2), 17– 34. Google Scholar CrossRef Search ADS   Auriol E., Gary-Bobo R. J. ( 2012), ‘ On the optimal number of representatives,’ Public Choice , 153( 3–4), 419– 445. Google Scholar CrossRef Search ADS   Autio E., Kenney M., Mustar P., Siegel D., Wright M. ( 2014), ‘ Entrepreneurial innovation: the importance of context,’ Research Policy , 43( 7), 1097– 1108. Google Scholar CrossRef Search ADS   Bandura A. ( 2006), ‘ Toward a psychology of human agency,’ Perspectives on Psychological Science , 1( 2), 164– 180. Google Scholar CrossRef Search ADS   Baron R. A., Baum J. R. ( 2007), ‘Research gains: benefits of closer links between I/O psychology and entrepreneurship,’ in Baum J. R., Frese M., Baron R. A. (eds), The Psychology of Entrepreneurship . Erlbaum: Mahwah, NJ, pp. 347– 373. Bartelsman E., Haltiwanger J., Scarpetta S. ( 2009), ‘Measuring and analyzing cross-country differences in firm dynamics,’ in Dunne T., Jensen J. B., Roberts M. J. (eds), Producer Dynamics: New Evidence from Micro Data . University of Chicago Press for NBER: Chicago, pp. 15– 79. Google Scholar CrossRef Search ADS   Bartelsman E., Haltiwanger J., Scarpetta S. ( 2013), ‘ Cross-country differences in productivity: the role of allocation and selection,’ American Economic Review , 103( 1), 305– 334. Google Scholar CrossRef Search ADS   Baum C. F., Schaffer M. E. ( 2003), ‘ Instrumental variables and GMM: estimation and testing,’ Stata Journal , 3, 1– 31. Begley T. M., Boyd D. P. ( 1987), ‘ Psychological characteristics associated with performance in entrepreneurial firms and smaller businesses,’ Journal of Business Venturing , 2( 1), 79– 93. Google Scholar CrossRef Search ADS   Bell R. G., Moore C. B., Al-Shammari H. A. ( 2008), ‘ Country of origin and foreign IPO legitimacy,’ Entrepreneurship Theory and Practice , 32( 1), 185– 202. Google Scholar CrossRef Search ADS   Benet-Martinez V., John O. P. ( 1998), ‘ Los Cinco grandes across cultures and ethnic groups: multitrait-multimethod analyses of the big five in Spanish and English,’ Journal of Personality and Social Psychology , 75, 729– 750. Google Scholar CrossRef Search ADS   Besley T., Ghatak M. ( 2010), ‘Property rights and economic development,’ in Rodrik D., Rosenzweig M. (eds), Handbook of Development Economics , Vol. 5. Elsevier: North Holland, The Netherlands, pp. 4525– 4595. Google Scholar CrossRef Search ADS   Bjørnskov C., Foss N. J. ( 2008), ‘ Economic freedom and entrepreneurial activity: some cross-country evidence,’ Public Choice , 134( 3–4), 307– 328. Google Scholar CrossRef Search ADS   Blackburn R., Renwick S. J., Donnelly J. P., Logan C. ( 2004), ‘ Big five or big two? Superordinate factors in the NEO five factor inventory and the antisocial personality questionnaire,’ Personality and Individual Differences , 37( 5), 957– 970. Google Scholar CrossRef Search ADS   Blanchflower D. G. ( 2000), ‘ Self-employment in OECD countries,’ Labor Economics , 7( 5), 471– 505. Google Scholar CrossRef Search ADS   Blau D. M. ( 1987), ‘ A time-series analysis of self-employment in the United States,’ Journal of Political Economy , 95( 3), 445– 467. Google Scholar CrossRef Search ADS   Bleidorn W., Klimstra T. A., Denissen J. J. A., Rentfrow P. J., Potter J., Gosling S. D. ( 2013), ‘ Personality maturation around the world: a cross-cultural examination of social-investment theory,’ Psychological Science , 24( 12), 2530– 2540. Google Scholar CrossRef Search ADS   Blume L., Müller J., Voigt S., Wolf C. ( 2009), ‘ The economic effects of constitutions: replicating—and extending—Persson and Tabellini,’ Public Choice , 139( 1–2), 197– 225. Google Scholar CrossRef Search ADS   Blundell R., Bond S. ( 1998), ‘ Initial conditions and moment restrictions in dynamic panel data models,’ Journal of Econometrics , 87( 1), 115– 143. Google Scholar CrossRef Search ADS   Borghans L., Duckworth A. L., Heckman J. J., Ter Weel B. ( 2008), ‘ The economics and psychology of personality traits,’ Journal of Human Resources , 43( 4), 972– 1059. Bosse D. A., Phillips R. A. ( 2016), ‘ Agency theory and bounded self-interest,’ Academy of Management Review , 41( 2), 276– 297. Google Scholar CrossRef Search ADS   Bottazzi G., Dosi G., Jacoby N., Secchi A., Tamagni F. ( 2010), ‘ Corporate performances and market selection: some comparative evidence,’ Industrial and Corporate Change , 19( 6), 1953– 1996. Google Scholar CrossRef Search ADS   Brandstätter H. ( 2011), ‘ Personality aspects of entrepreneurship: a look at five meta-analyses,’ Personality and Individual Differences , 51( 3), 222– 230. Google Scholar CrossRef Search ADS   Brück T., Lussá F., Tavares J. A. ( 2011), ‘ Entrepreneurship: the role of extreme events,’ European Journal of Political Economy , 27, S78– S88. Google Scholar CrossRef Search ADS   Buchanan J., Tullock G. ( 1962), The Calculus of Consent—Logical Foundations of Constitutional Democracy . University of Michigan Press: Ann Arbor. Google Scholar CrossRef Search ADS   Carbonara E., Parisi F., von Wangenheim G. ( 2012), ‘ Unjust laws and illegal norms,’ International Review of Law and Economics , 32( 3), 285– 299. Google Scholar CrossRef Search ADS   Carbonara E., Santarelli E., Tran H. T. ( 2016), ‘ De Jure determinants of new firm formation: how the pillars of constitutions influence entrepreneurship,’ Small Business Economics , 47( 1), 139– 162. Google Scholar CrossRef Search ADS   Carlsson B., Acs Z. J., Audretsch D. B., Braunerhjelm P. ( 2009), ‘ Knowledge creation, and economic growth: a historical review,’ Industrial and Corporate Change , 18( 6), 1193– 1229. Google Scholar CrossRef Search ADS   Carree M., van Stel A., Thurik R., Wennekers S. ( 2002), ‘ Economic development and business ownership: an analysis using data of 23 OECD countries in the period 1976-1996,’ Small Business Economics , 19( 3), 271– 290. Google Scholar CrossRef Search ADS   Caves R. E. ( 1998), ‘ Industrial organization and new findings on the turnover and mobility of firms,’ Journal of Economic Literature , 36, 1947– 1982. Chemin M. ( 2009), ‘ The impact of the judiciary on entrepreneurship: evaluation of Pakistan’s access to justice programme,’ Journal of Public Economics , 93( 1–2), 114– 125. Google Scholar CrossRef Search ADS   Chemin M. ( 2012), ‘ Does court speed shape economic activity? Evidence from a court reform in India,’ Journal of Law, Economics, and Organization , 28( 3), 460– 485. Google Scholar CrossRef Search ADS   Cuomo F., Zattoni A., Valentini G. ( 2013), ‘ The effects of legal reforms on the ownership structure of listed companies,’ Industrial and Corporate Change , 22( 2), 427– 458. Google Scholar CrossRef Search ADS   Czarnitzki D., Doherr T., Hussinger K., Schliessler P., Toole A. A. ( 2016), ‘ Knowledge creates markets: the influence of entrepreneurial support and patent rights on academic entrepreneurship,’ European Economic Review , 86, 131– 146. Google Scholar CrossRef Search ADS   Davidsson P. ( 1995), ‘ Culture, structure and regional levels of entrepreneurship,’ Entrepreneurship and Regional Development , 7( 1), 41– 62. Google Scholar CrossRef Search ADS   DeYoung C. G. ( 2006), ‘ Higher-order factors of the big five in a multi-informant sample,’ Journal of Personality and Social Psychology , 91( 6), 1138– 1151. Google Scholar CrossRef Search ADS   Dheer R. J. S. ( 2017), ‘ Cross-national differences in entrepreneurial activity: role of culture and institutional factors,’ Small Business Economics , 48( 4), 813– 842. Google Scholar CrossRef Search ADS   Diamond J. ( 1997), Guns, Germs, and Steel. The Fates of Human Societies . W.W. Norton & Company: New York, NY; London. Digman J. M. ( 1997), ‘ Higher-order factors of the big five,’ Journal of Personality and Social Psychology , 73( 6), 1246– 56. Google Scholar CrossRef Search ADS   Djankov S., La Porta R., Lopez-De-Silanes F., Shleifer A. ( 2003), ‘ Courts,’ The Quarterly Journal of Economics , 118( 2), 453– 517. Google Scholar CrossRef Search ADS   Elder G. H. ( 1994), ‘ Time, human agency, and social change: perspectives on the life course,’ Social Psychology Quarterly , 57( 1), 4– 15. Google Scholar CrossRef Search ADS   Elkins Z., Ginsburg T., Melton J. ( 2009), The Comparative Constitutions Project: A Cross-National Historical Dataset of Written Constitutions . Mimeo: Chicago. Frankel J. A., Romer D. ( 1996), ‘Trade and growth: an empirical investigation,’ NBER Working paper series #5476. Frese M. ( 2009), ‘ Towards a psychology of entrepreneurship—an action theory perspective,’ Foundations and Trends in Entrepreneurship , 5( 6), 437– 496. Google Scholar CrossRef Search ADS   Frey B. S. ( 2008), Happiness: A Revolution in Economics . MIT Press Books: Cambridge, MA. Google Scholar CrossRef Search ADS   Funder D. C. ( 2006), ‘ Towards a resolution of the personality triad: persons, situations, and behaviors,’ Journal of Research in Personality , 40( 1), 21– 34. Google Scholar CrossRef Search ADS   Galor O., Michalopoulos S. ( 2012), ‘ Evolution and the growth process: natural selection of entrepreneurial traits,’ Journal of Economic Theory , 147( 2), 759– 780. Google Scholar CrossRef Search ADS   García-Posada M., Mora-Sanguinetti J. S. ( 2015), ‘ Entrepreneurship and enforcement institutions: disaggregated evidence for Spain,’ European Journal of Law and Economics , 40( 1), 49– 74. Google Scholar CrossRef Search ADS   Gebauer J. E., Bleidorn W., Gosling S. D., Rentfrow P. J., Lamb M. E., Potter J. ( 2014), ‘ Cross-cultural variations in big five relationships with religiosity: a sociocultural motives perspective,’ Journal of Personality and Social Psychology , 107( 6), 1064– 1091. Google Scholar CrossRef Search ADS   Gebauer J. E., Sedikides C., Lüdtke O., Neberich W. ( 2014), ‘ Agency‐communion and interest in prosocial behavior: social motives for assimilation and contrast explain sociocultural inconsistencies,’ Journal of Personality , 82( 5), 452– 466. Google Scholar CrossRef Search ADS   Gebauer J. E., Sedikides C., Wagner J., Bleidorn W., Rentfrow P. J., Potter J., Gosling S. D. ( 2015), ‘ Cultural norm fulfilment, interpersonal belonging, or getting ahead? A large-scale cross-cultural test of three perspectives on the function of self-esteem,’ Journal of Personality and Social Psychology , 109( 3), 526– 548. Google Scholar CrossRef Search ADS   Glaeser E. L., Laibson D. I., Scheinkman J. A., Soutter C. L. ( 2000), ‘ Measuring Trust,’ Quarterly Journal of Economics , 115( 3), 811– 846, Google Scholar CrossRef Search ADS   Gorodnichenko Y., Roland G. ( 2010), ‘Culture, institutions and the wealth of nations,’ NBER Working Paper No. 16368. Gosling S. D., Vazire S., Srivastava S., John O. P. ( 2004), ‘ Should we trust web-based studies? A comparative analysis of six preconceptions about internet questionnaires,’ American Psychologist , 59( 2), 93– 104. Google Scholar CrossRef Search ADS   Greif A. ( 1994), ‘ Cultural beliefs and the organization of society: a historical and theoretical reflection on collectivist and individualist societies,’ Journal of Political Economy , 102( 5), 912– 950. Google Scholar CrossRef Search ADS   Gries T., Naudé W. ( 2011), ‘ Entrepreneurship and human development: a capability approach,’ Journal of Public Economics , 95( 3–4), 216– 224. Google Scholar CrossRef Search ADS   Grilo I., Thurik A. R. ( 2006), ‘Entrepreneurship in the old and new Europe,’ in Santarelli E. (ed.), Entrepreneurship, Growth, and Innovation. The Dynamics of Firms and Industries , Springer: New York, NY, pp. 75– 103. Google Scholar CrossRef Search ADS   Guiso L., Sapienza P., Zingales L. ( 2003), ‘ People’s Opium? Religion and economic attitudes,’ Journal of Monetary Economics , 50( 1), 225– 282. Google Scholar CrossRef Search ADS   Guiso L., Sapienza P., Zingales L. ( 2006), ‘ Does culture affect economic outcomes?,’ Journal of Economic Perspectives , 20( 2), 23– 48. Google Scholar CrossRef Search ADS   Guiso L., Sapienza P., Zingales L. ( 2008), ‘ Alfred Marshall lecture—social capital as good culture,’ Journal of the European Economic Association , 6( 2–3), 295– 320. Google Scholar CrossRef Search ADS   Hall R. E., Jones C. I. ( 1999), ‘ Why do some countries produce so much more output per worker than others?,’ Quarterly Journal of Economics , 114( 1), 83– 116. Google Scholar CrossRef Search ADS   Hechavarria D. M., Renko M., Matthews C. H. ( 2012), ‘ The nascent entrepreneurship hub: goals, entrepreneurial self-efficacy and start-up outcomes,’ Small Business Economics , 39( 3), 685– 701. Google Scholar CrossRef Search ADS   Hmieleski K. M., Carr J. C., Baron R. A. ( 2015), ‘ Integrating discovery and creation perspectives of entrepreneurial action: the relative roles of founding CEO human capital, social capital, and psychological capital in contexts of risk versus uncertainty,’ Strategic Entrepreneurship Journal , 9( 4), 289– 312. Google Scholar CrossRef Search ADS   Hofstede G., McCrae R. R. ( 2004), ‘ Personality and culture revisited: linking traits and dimensions of culture,’ Cross-Cultural Research , 38( 1), 52– 88. Google Scholar CrossRef Search ADS   John O. P., Srivastava S. ( 1999), ‘The big five trait taxonomy: history, measurement, and theoretical perspectives,’ in Pervin L. A., John O. P. (eds), Handbook of Personality: Theory and Research . Guilford Press: New York, NY, pp. 102– 138. Jokela M., Bleidorn W., Lamb M. E., Gosling S. D., Rentfrow P. J. ( 2015), ‘ Geographically varying associations between personality and life satisfaction in the London Metropolitan area,’ Proceedings of the National Academy of Sciences , 112( 3), 725– 730. Google Scholar CrossRef Search ADS   Kelsen H. ( 1967), Pure Theory of Law . University of California Press: Berkeley; Los Angeles. Kihlstrom R. E., Laffont J. J. ( 1979), ‘ A general equilibrium entrepreneurial theory of firm formation based on risk aversion,’ Journal of Political Economy , 87( 4), 719– 748. Google Scholar CrossRef Search ADS   Koellinger P. D., Thurik A. R. ( 2012), ‘ Entrepreneurship and the business cycle,’ Review of Economics and Statistics , 94( 4), 1143– 1156. Google Scholar CrossRef Search ADS   Lang F. R., John D., Lüdtke O., Schupp J., Wagner G. G. ( 2011), ‘ Short assessment of the big five: robust across survey methods except telephone interviewing,’ Behavior Research Methods , 43( 2), 548– 567. Google Scholar CrossRef Search ADS   Lerner J., Tåg J. ( 2013), ‘ Institutions and venture capital,’ Industrial and Corporate Change , 22( 1), 153– 182. Google Scholar CrossRef Search ADS   Leutner F., Ahmetoglu G., Akhtar R., Chamorro-Premuzic T. ( 2014), ‘ The relationship between the entrepreneurial personality and the big five personality traits,’ Personality and Individual Differences , 63, 58– 63. Google Scholar CrossRef Search ADS   Lewin K. ( 1935), A Dynamic Theory of Personality . McGraw-Hill: New York, NY. Lewin K. ( 1951), Field Theory in Social Science . McGraw-Hill: New York, NY. Lichand G., Soares R. R. ( 2014), ‘ Access to justice and entrepreneurship: evidence from Brazil’s special civil tribunals,’ Journal of Law and Economics , 57( 2), 459– 499.0022-2186/2014/5702-0014. Google Scholar CrossRef Search ADS   McClelland D. ( 1961), The Achieving Society . Van Nostrand: Princeton, NJ. Google Scholar CrossRef Search ADS   Melton J., Elkins Z., Ginsburg T., Leetaru K. ( 2013). ‘ On the interpretability of law: lessons from the decoding of national constitutions,’ British Journal of Political Science , 43( 02), 399– 423. Google Scholar CrossRef Search ADS   Niszczota P. ( 2014), ‘ Cross-country differences in personality and the foreign bias in international equity portfolios,’ European Journal of Finance , 20( 10), 934– 956. Google Scholar CrossRef Search ADS   Nyström K. ( 2008), ‘ The institutions of economic freedom and entrepreneurship: evidence from panel data,’ Public Choice , 136( 3–4), 269– 282. Google Scholar CrossRef Search ADS   Obschonka M., Stuetzer M., Audretsch D. B., Rentfrow P. J., Potter J., Gosling S. D. ( 2016), ‘ Macropsychological factors predict regional economic resilience during a major economic crisis,’ Social Psychological and Personality Science , 7( 2), 95– 104. Google Scholar CrossRef Search ADS   Obschonka M., Stuetzer M., Gosling S. D., Rentfrow P. J., Lamb M. E., Potter J., Audretsch D. B., Waldorp L. J. ( 2015), ‘ Entrepreneurial regions: do macro-psychological cultural characteristics of regions help solve the “knowledge paradox” of economics?,’ PLoS One , 10( 6), e0129332.10.1371/journal.pone.0129332, June 22. Google Scholar CrossRef Search ADS   Parker S. C. ( 2007), ‘ Law and the economics of entrepreneurship,’ Comparative Labor Law and Policy Journal , 28, 695– 715. Parker S. C. ( 2009), The Economics of Entrepreneurship . Cambridge University Press: Cambridge, MA. Google Scholar CrossRef Search ADS   Persson T., Tabellini G. ( 2003), The Economic Effects of Constitutions . The MIT Press: Cambridge, MA. Rentfrow P. J., Gosling S. D., Jokela M., Stillwell D. J., Kosinski M., Potter J. ( 2013), ‘ Divided we stand: three psychological regions of the United States and their political, economic, social, and health correlates,’ Journal of Personality and Social Psychology , 105( 6), 996– 1012. Google Scholar CrossRef Search ADS   Rentfrow P. J., Gosling S. D., Potter J. ( 2008), ‘ A theory of the emergence, persistence, and expression of geographic variation in psychological characteristics,’ Perspectives on Psychological Science , 3( 5), 339– 369. Google Scholar CrossRef Search ADS   Rentfrow P. J., Jokela M., Lamb M. E. ( 2015), ‘ Regional personality differences in Great Britain,’ R. D. Latzman, PLoS One , 10( 3), e0122245. Google Scholar CrossRef Search ADS   Roberts B. W., Kuncel N. R., Shiner R., Caspi A., Goldberg L. R. ( 2007), ‘ The power of personality: the comparative validity of personality traits, socioeconomic status, and cognitive ability for predicting important life outcomes,’ Perspectives on Psychological Science , 2( 4), 313– 345. Google Scholar CrossRef Search ADS   Santarelli E., Pesciarelli E. ( 1990), ‘ The emergence of a vision: the development of schumpeter’s theory of entrepreneurship,’ History of Political Economy , 22( 4), 677– 696. Google Scholar CrossRef Search ADS   Santarelli E., Sterlacchini A. ( 1994), ‘ New firm formation in Italian industry: 1975-89,’ Small Business Economics , 6( 2), 95– 106. Google Scholar CrossRef Search ADS   Santarelli E., Tran H. T. ( 2013), ‘ The interplay of human capital and social capital in shaping entrepreneurial performance: the case of Vietnam,’ Small Business Economics , 40( 2), 435– 458. Google Scholar CrossRef Search ADS   Santarelli E., Vivarelli M. ( 2007), ‘ Entrepreneurship and the process of firm’s entry, survival and growth,’ Industrial and Corporate Change , 16( 3), 455– 488. Google Scholar CrossRef Search ADS   Saxenian A. ( 1994), Regional Advantage . Harvard University Press: Cambridge, MA. Schumpeter J. A. ( 1911), Theorie der Wirtschaftlichen Entwicklung . Duncker & Humblot: Leipzig. Schumpeter J. A. ( 1934), Theory of Economic Development . Harvard University Press: Cambridge, MA. Steel G. D., Rinne T., Fairweather J. ( 2012), ‘ Personality, nations, and innovations: relationships between the big five personality traits and national innovation scores,’ Cross-Cultural Research , 46( 1), 3– 30. Google Scholar CrossRef Search ADS   Storey D. J., Jones A. M. ( 1987), ‘ New firm formation: a labor market approach to industrial entry,’ Scottish Journal of Political Economy , 34( 1), 37– 51. Google Scholar CrossRef Search ADS   Stuetzer M., Obschonka M., Audretsch D. B., Wyrwich M., Rentfrow P. J., Coombes M., Shaw-Taylor L., Satchell M. ( 2016), ‘Industry structure, entrepreneurship, and culture: an empirical analysis using historical coalfields,’ European Economic Review , 86, 52– 72. Google Scholar CrossRef Search ADS   Tag J., Astebro T., Thompson P. ( 2016), ‘ Hierarchies and entrepreneurship,’ European Economic Review , 89, 129– 147. Google Scholar CrossRef Search ADS   Vecchione M., Alessandri G. ( 2013), ‘ Alpha and beta traits and egoistic and moralistic self‐enhancement: a point of convergence between two research traditions,’ Journal of Personality , 81( 1), 39– 48. Google Scholar CrossRef Search ADS   Vedel A. ( 2014), ‘ The big five and tertiary academic performance: a systematic review and meta-analysis,’ Personality and Individual Differences , 71, 66– 76. Google Scholar CrossRef Search ADS   Wennberg K., Pathak S., Autio E. ( 2013), ‘ How culture moulds the effects of self-efficacy and fear of failure on entrepreneurship,’ Entrepreneurship Theory and Practice , 25( 9–10), 756– 780. Wennekers S., van Wennekers A., Thurik R., Reynolds P. ( 2005), ‘ Nascent entrepreneurship and the level of economic development,’ Small Business Economics , 24( 3), 311– 321. Google Scholar CrossRef Search ADS   Wiggins J. S. ( 1991), ‘Agency and communion as conceptual coordinates for the understanding and measurement of interpersonal behavior,’ in Grove W. M., Cicchetti D. (eds), Thinking Clearly About Psychology. Vol. 2: Personality and Psychopathology . University of Minnesota Press: Minneapolis, pp. 89– 113. Wiggins J. S. ( 2003), Paradigms of Personality Assessment . Guilford: New York, NY. Zhao H., Seibert S. E. ( 2006), ‘ The big five personality dimensions and entrepreneurial status: a meta-analytical review,’ Journal of Applied Psychology , 91( 2), 259– 271. Google Scholar CrossRef Search ADS   Zhao H., Seibert S. E., Lumpkin G. T. ( 2010), ‘ The relationship of personality to entrepreneurial intentions and performance: a meta-analytic review,’ Journal of Management , 36, 381– 404. Google Scholar CrossRef Search ADS   © The Author 2017. Published by Oxford University Press on behalf of Associazione ICC. 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 Industrial and Corporate Change Oxford University Press

Agency culture, constitutional provisions and entrepreneurship: a cross-country analysis

Loading next page...
 
/lp/ou_press/agency-culture-constitutional-provisions-and-entrepreneurship-a-cross-i7500C2vZ0
Publisher
Oxford University Press
Copyright
© The Author 2017. Published by Oxford University Press on behalf of Associazione ICC. All rights reserved.
ISSN
0960-6491
eISSN
1464-3650
D.O.I.
10.1093/icc/dtx047
Publisher site
See Article on Publisher Site

Abstract

Abstract Substantial and systematic cross-country variation in entrepreneurship rates has been found in various studies. We attempt to explain such differences focusing on the interaction between institutional factors and population psychological characteristics. Constitutional provisions supporting economic freedom are our measure of the institutional context, whereas we proxy psychological characteristics with a country’s endowment of agency culture. We apply an IV-GMM treatment to deal with endogeneity to a data set comprising 86 countries over the period 2004–2013, and we control for de facto variables and other factors that are likely to influence entrepreneurship. Our results demonstrate that agency culture is indeed an important predictor of entrepreneurship and that this effect is moderated by constitutional provisions supporting economic freedom. In particular, the impact of agency culture on entrepreneurship becomes stronger as a country expands the constitutional protection of economic rights. 1. Introduction Cross-country comparison of industry dynamics and exploration of its determinants and consequences has traditionally attracted the interest of researchers in both industrial and developing countries (see Caves, 1998; and Bartelsman et al., 2009 for surveys). The results of this literature show that substantial and systematic differences in industry dynamics are generated also by country-specific institutional and cultural factors (see Bottazzi et al., 2010; Bartelsman et al., 2013; Niszczota, 2014). The aim of the present article is to study the interplay between the economic constitution of a country (institutional factor) and the macro-psychological traits of its population (cultural factor) in shaping cross-country differences in entrepreneurship rates. Our hypothesis is that constitutional protection of economic freedom may together create an institutional setting that favors the transformation of the innate agentic attitude of a country’s population into actual entrepreneurship. It follows from this assumption that differences in the constitutions and the endowment of agency culture, and also their interplay, may explain the cross-country variation in industry dynamics. We conduct our analysis using a sample comprising 86 countries over the period 2004–2013 (see list in Table 1). Table 1. List of countries by geographical area America: Argentina; Belize; Bolivia; Brazil; Canada; Chile; Colombia; Costa Rica; Dominican Republic; El Salvador; Guatemala; Jamaica; Mexico; Panama; Peru; the United States; Uruguay.  Europe: Albania; Armenia; Austria; Belgium; Bosnia and Herzegovina; Bulgaria; Croatia; Cyprus; Czech Republic; Denmark; Estonia; Finland; France; Germany; Greece; Hungary; Iceland; Ireland; Italy; Latvia; Lithuania; Luxembourg; Macedonia; Malta; Montenegro; The Netherlands; Norway; Poland; Portugal; Romania; Russia; Slovak Republic; Slovenia; Spain; Sweden; Switzerland; Ukraine; UK.  Africa: Algeria; Ethiopia; Ghana; Kenya; Mauritius; Morocco; Nigeria; South Africa; Uganda.  Asia: Afghanistan; Bangladesh; Brunei; Egypt; India; Indonesia; Israel; Japan; Jordan; Malaysia; Nepal; Oman; Pakistan; Philippines; Qatar; Singapore; South Korea; Sri Lanka; Thailand; Turkey; United Arab Emirates (UAE).  Oceania: Australia; New Zealand.  America: Argentina; Belize; Bolivia; Brazil; Canada; Chile; Colombia; Costa Rica; Dominican Republic; El Salvador; Guatemala; Jamaica; Mexico; Panama; Peru; the United States; Uruguay.  Europe: Albania; Armenia; Austria; Belgium; Bosnia and Herzegovina; Bulgaria; Croatia; Cyprus; Czech Republic; Denmark; Estonia; Finland; France; Germany; Greece; Hungary; Iceland; Ireland; Italy; Latvia; Lithuania; Luxembourg; Macedonia; Malta; Montenegro; The Netherlands; Norway; Poland; Portugal; Romania; Russia; Slovak Republic; Slovenia; Spain; Sweden; Switzerland; Ukraine; UK.  Africa: Algeria; Ethiopia; Ghana; Kenya; Mauritius; Morocco; Nigeria; South Africa; Uganda.  Asia: Afghanistan; Bangladesh; Brunei; Egypt; India; Indonesia; Israel; Japan; Jordan; Malaysia; Nepal; Oman; Pakistan; Philippines; Qatar; Singapore; South Korea; Sri Lanka; Thailand; Turkey; United Arab Emirates (UAE).  Oceania: Australia; New Zealand.  Table 1. List of countries by geographical area America: Argentina; Belize; Bolivia; Brazil; Canada; Chile; Colombia; Costa Rica; Dominican Republic; El Salvador; Guatemala; Jamaica; Mexico; Panama; Peru; the United States; Uruguay.  Europe: Albania; Armenia; Austria; Belgium; Bosnia and Herzegovina; Bulgaria; Croatia; Cyprus; Czech Republic; Denmark; Estonia; Finland; France; Germany; Greece; Hungary; Iceland; Ireland; Italy; Latvia; Lithuania; Luxembourg; Macedonia; Malta; Montenegro; The Netherlands; Norway; Poland; Portugal; Romania; Russia; Slovak Republic; Slovenia; Spain; Sweden; Switzerland; Ukraine; UK.  Africa: Algeria; Ethiopia; Ghana; Kenya; Mauritius; Morocco; Nigeria; South Africa; Uganda.  Asia: Afghanistan; Bangladesh; Brunei; Egypt; India; Indonesia; Israel; Japan; Jordan; Malaysia; Nepal; Oman; Pakistan; Philippines; Qatar; Singapore; South Korea; Sri Lanka; Thailand; Turkey; United Arab Emirates (UAE).  Oceania: Australia; New Zealand.  America: Argentina; Belize; Bolivia; Brazil; Canada; Chile; Colombia; Costa Rica; Dominican Republic; El Salvador; Guatemala; Jamaica; Mexico; Panama; Peru; the United States; Uruguay.  Europe: Albania; Armenia; Austria; Belgium; Bosnia and Herzegovina; Bulgaria; Croatia; Cyprus; Czech Republic; Denmark; Estonia; Finland; France; Germany; Greece; Hungary; Iceland; Ireland; Italy; Latvia; Lithuania; Luxembourg; Macedonia; Malta; Montenegro; The Netherlands; Norway; Poland; Portugal; Romania; Russia; Slovak Republic; Slovenia; Spain; Sweden; Switzerland; Ukraine; UK.  Africa: Algeria; Ethiopia; Ghana; Kenya; Mauritius; Morocco; Nigeria; South Africa; Uganda.  Asia: Afghanistan; Bangladesh; Brunei; Egypt; India; Indonesia; Israel; Japan; Jordan; Malaysia; Nepal; Oman; Pakistan; Philippines; Qatar; Singapore; South Korea; Sri Lanka; Thailand; Turkey; United Arab Emirates (UAE).  Oceania: Australia; New Zealand.  Taking a law and economics perspective, we focus on four principles stated in national constitutions: the right to conduct/establish a business, the right to free/competitive markets, the right to own property, and the independence of the judiciary organs. The first two principles have been proven in the law and economics literature to exert a significant impact on entrepreneurship (Carbonara et al., 2016). The right to own property (Besley and Ghatak, 2010) and the independence of the judiciary organs (Djankov et al., 2003, Chemin, 2009) are widely considered key factors in economic development. The use of constitutional provisions as proxy for institutional characteristics stems from the fact that constitutional laws represent hierarchically higher norms that cannot be opposed by ordinary laws and other rules (Kelsen, 1967). Thus, they represent the highest protection that a country can assign to rights. However, often the laws in the books remain unenforced (especially when they clash with social norms that are firmly embedded in culture: Carbonara et al., 2012, Acemoglu and Jackson, 2016). That is why de jure protection of legal rights does not necessarily imply a de facto protection, and we need to control for the actual implementation of the principles stated in the constitution, including measures of economic freedom based on rule of law, government size, regulatory efficiency, and market openness. The psychological literature has placed great importance on personality traits, arguing that regional differences in personality constitute a local culture that in turn influences regional entrepreneurship rates (cf., among others, Davidsson, 1995; Leutner et al., 2014; Obschonka et al., 2015; Stuetzer et al., 2016). Such a personality-based perspective on culture (Hofstede and McCrae, 2004) has enriched our understanding of the historical origins (Stuetzer et al., 2016) and economic effects of regional differences in an entrepreneurial culture (Davidsson, 1995; Steel et al., 2012; Rentfrow et al., 2013; Audretsch et al., 2017). To date, the law and economics and the psychological streams of literature have not been integrated in the explanation of the overall process of industry dynamics. There is good reason to suppose that combining them could prove profitable, with some studies hinting at the promise of such integrative perspectives. For example, Obschonka et al. (2015) examined the so-called “knowledge paradox”, which is the phenomenon whereby investments in resources for generating knowledge (e.g., education, diversity of industries) do not guarantee higher entrepreneurship rates; analyses revealed that knowledge resources are more likely to increase entrepreneurship rates in a region which also has a high number of residents with an entrepreneurship-prone personality. In other words, psychological (in this case, entrepreneurial personality) and institutional determinants (in this case, knowledge resources) may interact to yield better predictions about entrepreneurial activity than the additive effects of both determinants assessed in isolation. The article is organized as follows. Section 2 contains a review of the literature dealing with cross-national differences in entrepreneurship rates, and the impact of psychological traits and constitutional provisions on entrepreneurship. Section 3 presents the main research questions and theoretical explanations. Section 4 describes the data set. Section 5 presents the estimation model and econometric strategy. Section 6 discusses the findings of the empirical analysis. Finally, Section 7 offers some conclusions based on these findings. 2. Literature review In economics, agentic behavior is usually defined within a rational-choice perspective, assuming that agency is mainly characterized by the maximization of one’s own benefits: all actors are narrowly self-interested, all actors are boundedly rational, and agents are more risk averse than principals are (Bosse and Phillips, 2016). However, this approach neglects inter-individual psychological differences (i.e., personality characteristics motivating, guiding, and directing decisions and activities), which have been shown to predict a wide array of consequential life outcomes and economic behaviors, even when controlling for the effects of socio-economic status, demographic variables, and cognitive ability (Roberts et al., 2007). Psychological research points to some sort of psychological benefit for the individual (or the avoidance of negative, harmful states) if he or she can behave in accordance with his or her individual personality structure (Frey, 2008). Psychological theories suggest that peoples’ behavior can best be understood by an interplay between person variables (like personality) and the context (cf., among others, Lewin, 1935, 1951; Ajzen, 1985; Funder, 2006), which means that it might be worth analyzing carefully the interaction between psychological traits and constitutional provisions. The agentic perspective is widely regarded as a leading meta-theory of human behavior in psychology (Bandura, 2006), sociology (Elder, 1994), economics (Kihlstrom and Laffont, 1979), law (Parker, 2007), and management (Begley and Boyd, 1987). 2.1 Cross-country differences in entrepreneurship rates Some countries experience higher rates of new firm formation every year than other countries do (Carree et al., 2002; Santarelli and Vivarelli, 2007). From a theoretical viewpoint, two main explanations of this empirical regularity have pervaded the recent debate. On the one side, based on the observation that developed Western countries have become more entrepreneurial following globalization, Audretsch and Thurik (2000) hypothesize that such countries switch to new industries—such as software and biotechnology, in which small businesses and entrepreneurship are more important—only once they have lost their comparative advantage in large-scale manufacturing. Thus, high rates of new firm formation are typical of developed countries since the aftermath of the Information and Communication Technology revolution. This pattern might suggest that a country’s endowment of agency culture evolves in response to historical and institutional changes. On the other side, Galor and Michalopoulos (2012) suggest that entrepreneurial spirit has evolved non-monotonically in the course of history, through a Darwinian process. In the early stages of a country’s development, risk-tolerant entrepreneurial traits proved successful in promoting technological progress and economic development, whereas in mature stages of development, risk-averse traits prevailed, diminishing the growth potential of advanced economies. Thus, modern developed countries should experience lower rates of new firm formation. This latter approach implies that a country’s endowment of agency culture is linked to its stage of economic development rather than to time-invariant psychological features; hence, agency culture would tend to vanish as societies evolve, develop more complex institutional arrangements, and achieve higher levels of per capita income. Both of the views outlined above are consistent with the idea that the reason why countries with similar economic fundamentals differ in entrepreneurial activity may ultimately be found in cultural and institutional differences (Guiso et al., 2003; Stuetzer et al., 2016). The positions of Audretsch and Thurik (2000) and Galor and Michalopoulos (2012) can therefore be reconciled within a broader line of investigation, which spans from Diamond (1997) to Acemoglu and Robinson (2012) (cf. also Saxenian, 1994; Acemoglu et al., 2001; Autio et al., 2014). In fact, the empirical literature on the cross-country differences in start-up rates has provided several contributions which can be reconciled with each of the two explanations. Guiso et al. (2006) show that the cultural background of individuals plays a role in their decision to become entrepreneurs and therefore also shapes attitudes toward entrepreneurship at the region and country level. By the same token, Audretsch et al. (2017) posit the importance of culture as a primary determinant of variations in economic, political, and social phenomena across geographic space. They aggregate individual-level personality data to the level of each of the 3137 US counties to analyze the impact of a social and cultural imprinting on the rate of new firm formation at the county level. Wennekers et al. (2005) found a U-shaped relationship between a country’s start-up rate and a country’s level of economic development, with the impact of entrepreneurial dynamics on economic growth being smaller for developing countries. In contrast, Blanchflower (2000) showed that the overall trend of entrepreneurial activities does not follow the stages of a country’s development; rather, the trend shows a negative relationship with a country’s unemployment rate. Brück et al. (2011) found that entrepreneurship rates follow a history-dependent path and are subject to the influence of exogenous factors, with entrepreneurship rates being positively affected by extreme events such as natural disasters and terrorist attacks. Dealing with 85 countries between 2005 and 2014, Dheer (2017) has shown that an institutional setting that guarantees economic freedom affects the rate of entrepreneurial activity more in individualistic societies than in collectivistic ones. This is a clear indication that population psychological characteristics play a role in positively moderating the effect of pro-market institutional arrangements on entrepreneurship. Our study falls within the same line of investigation, although we use different measures for entrepreneurship and for economic freedom protection and we focus on different countries. A positive relationship between presence of an institutional framework able to promote economic freedom and various measures of entrepreneurship was found also by Bjørnskov and Foss (2008) and Nyström (2008). The above empirical evidence seems to suggest there is no single unique economic factor explaining cross-country differences in start-up rates. Differences might persist over time, regardless of a country’s level of economic development. Cultural differences might shape a country’s proneness toward entrepreneurial activity and are therefore a factor that needs consideration and integration. A country’s culture (e.g., its endowment of social capital; Guiso et al., 2008; or the national levels in personality traits, Hofstede and McCrae, 2004; Steel et al., 2012) interacts with and is shaped by the features and quality of a country’s institutions. Institutions are therefore an important aspect of a country’s profile and their impact extends from economic development (Guiso et al., 2003) to entrepreneurship (Acs et al., 2008; Carbonara et al. 2016). Culture and institutions are ultimately endogenous variables contributing to the wealth of countries. Extending the arguments proposed by Alesina and Giuliano (2015), we assume here that the psychological traits of a country’s population and the provisions contained in a country’s constitution are, respectively, aspects of a country’s culture and a country’s institutions. 2.2 Psychological agency and entrepreneurship The focus on an individual’s personal agency has long played a key role in seminal theorizing in the entrepreneurship literature (McClelland, 1961). In fact, Schumpeter himself (Schumpeter, 1911: 131, as translated from German in Santarelli and Pesciarelli, 1990), in the first German edition of his Theory of Economic Development (Schumpeter, 1934), stressed that entrepreneurs are “personalities who in se possess the rules of their actions” (for a detailed discussion of this issue, cf. Santarelli and Pesciarelli, 1990). Empirical entrepreneurship research has usually tried to capture such personal agency by focusing on an entrepreneur’s actual actions (Frese, 2009; Zhao et al., 2010; Hmieleski et al., 2015) or self-efficacy belief (Hechavarria et al., 2012; Wennberg et al., 2013). Here, we apply a novel approach to capture psychological agency, assessing it in terms of agentic personality traits (Digman, 1997). This approach is based on the leading and best researched model of personality traits, the Big Five model (John and Srivastava, 1999). This approach of assessing agency also allows us to draw from geographical approaches in the study of regional and national differences in these personality traits (Rentfrow et al., 2008; Steel et al., 2012). In psychology, Digman’s (1997) influential work on higher-order traits (or super traits) established that two Big Five traits, extraversion and openness to new experience, form a higher-order trait that can be labelled “psychological agency”. Drawing from that approach, we measure agency culture at the nation-level. Populations living in countries characterized by a high level of agency culture are highly active and assertive (components of extraversion), highly creative and open to change (components of openness to experience). Accordingly, aggregates of individual scores on traits are used as proxies for agency culture (for similar approaches to assess cultural dimensions, see Davidsson, 1995; Rentfrow et al., 2008, 2013; Steel et al., 2012; Stuetzer et al., 2016; Audretsch et al., 2017). In contrast to a purely rational-choice approach, the psychological approach defines agency by means of relatively stable personality traits that motivate, guide, and direct manifest individual agency. This psychological agency approach has largely been neglected in economic models of agency and entrepreneurship, despite the demonstrated importance of psychological models in economics (Borghans et al., 2008). In fact, psychological research has challenged the pure rational-choice view by pointing toward the relevance of “irrational” decision-making processes involving personality traits. A wide array of basic personality traits can have considerable influence on economic outcomes: for example, a recent study showed that entrepreneurial activity in the wake of the Great Recession of 2008–2009 was predicted better by regional personality differences than regional infrastructure parameters (“economic muscles”, such as human and financial capital) (Obschonka et al., 2016). 2.3 Entrepreneurship, agency culture, and the moderating effect of national constitutions Recent meta-analytic studies suggest significant relationships between personality, and both revealed preference for becoming an entrepreneur (latent entrepreneurship) (Zhao et al., 2010) and entrepreneurial performance after start-up (Brandstätter, 2011). However, there has been little investigation about the importance of personality as a predictor of the probability of actually being an entrepreneur (manifest entrepreneurship) (Grilo and Thurik, 2006; Baron and Baum, 2007; Audretsch et al. 2017). The level of agency culture in a country represents an important component of the overall cultural context within which entrepreneurial activity takes place. The relationship between a broader definition of culture—encompassing customary beliefs and values that are transmitted from generation to generation—and the likelihood of engaging in entrepreneurship has been explored by a line of investigation initiated by Guiso et al. (2006). Following the idea put forward by Glaeser et al. (2000), Guiso et al. (2006) assume that individuals who express trust have a comparative advantage in becoming entrepreneurs. Then, studying the impact of a measure of trust weighted for religious and ethnic background on the probability of a sample of individuals becoming entrepreneurs, Guiso et al. (2006) use an instrumental variable approach to confirm that trustworthy individuals will have a comparative advantage in becoming entrepreneurs. In the same way, defining a cultural variable along the dimension of individualism–collectivism, Gorodnichenko and Roland (2010) show that individualism has a dynamic advantage leading to a higher economic growth rate, whereas collectivism leads only to static efficiency gains. Then, a further element is added to the investigation initiated by Guiso et al. (2006): institutions (Alesina and Giuliano, 2015). Culture and institutions interact and evolve in a complementary way, both playing a role as determinants of the wealth of countries. Consequently, the same institutions may exert a different impact in different cultural contexts. Alesina and Giuliano (2015) identify a wide range of interactions between various types of political and legal institutions and various cultural traits, such as trust, family ties, generalized morality, and individualism. Lerner and Tåg (2013) show that institutional differences in the legal environment led to the later development of an active venture capital market in Sweden compared with the United States, where this source of external funding largely contributed to the emergence of clusters of innovative start-ups. The antinomy, individualism versus collectivism, is an important dimension of cultural variation across countries (Greif, 1994; Gorodnichenko and Roland, 2010; Alesina and Giuliano, 2015). Individualism is a trait that can make personal accomplishments more socially acceptable, so it is likely to be associated with a greater proneness to entrepreneurship. But for this cultural trait to result into actual action, the overall institutional setting should be proactive and remove the obstacles to the full display of individualism in the economic sphere. Consistent with this principle, within the broader field of law and economics, the public choice literature has emphasized (since the seminal contribution of Buchanan and Tullock, 1962) that a strong connection exists between a country’s economic performance and the main features of its constitution. Such a connection is likely not direct, but rather the result of the effectiveness of constitutions in shaping a country’s prevailing institutional arrangements (Melton et al., 2013; Carbonara et al., 2016). In relation to how constitutions may make it easier to turn individualism into actual entrepreneurial action, it is worth acknowledging that at least since France’s National Constituent Assembly passed the Déclaration des droits de l'homme et du citoyen in 1789, general recognition and protection of private property had been held to be universal and most constitutions started to protect property rights. The first empirical work on the impact of constitutions on economic performance dates back to the early 2000s; this work showed the positive impact jointly exerted by a presidential system and the majoritarian electoral rule on, among other things, total factor productivity and reduction of public expenditure (Persson and Tabellini, 2003). Research has also shown that direct democratic institutions affect fiscal policy and government efficiency (Blume et al., 2009). More recently, the optimal number of national representatives in relation to a country’s population size has been calculated (Auriol and Gary-Bobo, 2012); analyses suggest that an excessive number of national representatives are correlated with indicators of red tape and barriers to entrepreneurship. Moreover, the efficiency of the judiciary positively affects entrepreneurship. Constitutional provisions do play a role in making a judiciary system more efficient, for example, by stating that it must be independent from external influences. A well-functioning judiciary system facilitates access to finance and reduces the likelihood of contract breach (Chemin, 2009, 2012). The more a judiciary system is independent from the influence of both the other branches of government and partisan interests, the more judges are free to make impartial decisions based exclusively on fact and the rule of law. As a result, an independent and more efficient judiciary system may exert a direct impact on entrepreneurship, while it leaves the exit rate unaltered (Chemin, 2012; García-Posada and Mora-Sanguinetti, 2015). Highly skilled and better educated entrepreneurs take advantage of better access to justice (Lichand and Soares, 2014 on Brazilian data). Thus, reforms aimed at improving the efficiency of the judiciary may affect entrepreneurship positively among individuals with higher levels of education but not among those with lower educational levels. Education is a proxy for wealth, so this result seems to indicate that judicial change in Brazil pushed wealthier individuals toward entrepreneurship. Carbonara et al. (2016) show for 115 countries that constitutional provisions are the main institutional driver of entrepreneurship. Dealing with the endogeneity of constitutional rules, and controlling for de facto variables, they find that provisions about the right to conduct/establish a business, the right to strike, consumer protection, anti-corruption, and compulsory education promote higher rates of new firm formation. 3. Main hypotheses In what follows, we extend the investigation of the relationship between culture and institutions in the context of entrepreneurship by exploring the interaction between agency culture and the aspect of legal institutions represented by the provisions supporting economic freedom that are contained in national constitutions. We put a special focus on the interplay between agency culture and these aspects of the legal institutions within a country. To quantify psychological differences in agency, we apply the personality-based approach to culture (Rentfrow et al., 2008), which aggregates individual-level personality traits to estimate local cultural differences (Stuetzer et al., 2016). This approach has delivered promising findings in research predicting regional outcomes, including social, economic, political, and health outcomes (Rentfrow et al., 2013; Jokela et al., 2015). The basic idea underlying this research is that regional personality differences constitute the pillars of the local culture, affecting the developmental trajectories of whole regions (Hofstede and McCrae, 2004). In psychological science, there is broad consensus that the five-factor model of personality is the best-established, validated, and cross-culturally valid model of personality (cf., among many others, Digman, 1997; Benet-Martinez and John, 1998; John and Srivastava, 1999; Zhao and Seibert, 2006; Lang et al., 2011; Gebauer et al., 2014a; Vedel, 2014). The Big Five personality traits constituting this five-factor model are extraversion, conscientiousness, openness, agreeableness, and neuroticism. The Big Five traits can be further summarized in the form of higher order “super” traits (Wiggins, 2003). Based on analyses of child, adolescent, and adult samples, Digman (1997) established two higher-order “super” traits: α (consisting of conscientiousness, agreeableness, and neuroticism) and β (consisting of extraversion and openness). α can be described as a dimension encapsulating themes of communion, and β can be described as a dimension encapsulating themes of agency (Wiggins, 1991). The β super trait also includes such traits as superiority striving, individuation, personal growth, self-actualization, achieving status, and power motivation (Digman, 1997). These traits are associated with both agency and entrepreneurial behavior (Zhao and Seibert, 2006) making β a good candidate for indexing psychological agency in a way that is relevant for entrepreneurship. The study of such super factors (e.g., agency and communion) has received considerable attention in recent years (Blackburn et al., 2004; DeYoung, 2006; Abele and Wojciszke, 2007; Vecchione and Alessandri, 2013; Gebauer et al., 2014b), but this trend has not been mirrored in economic research, which has remained focused on narrower personality traits or profiles (Borghans et al., 2008; Stuetzer et al., 2016). With the aim to combine the views that institutional factors and population psychological characteristics are drivers of new firm formation, in line with the approach established by Dheer (2017), we aim to test the degree to which the combination of agency culture and pro-market constitutional framework combine to predict entrepreneurial activity. Thus, extending Carbonara et al. (2016), we take the constitutional protection granted to some principles relevant for economic activity and their de facto implementation (cf. also Carlsson et al., 2009; Czarnitzki et al., 2016) as proxies for the institutional determinants of entrepreneurship. Consistent with the bounded agency approach, we predict that boundary conditions (in the form of national constitutional framework) will shape the overall impact of agency culture on a country’s proneness toward entrepreneurial activity. Accordingly, pro-entrepreneurship constitutions cannot stimulate new business formation across countries as expected if people in those countries are not sufficiently proactive and innovative to exploit the benefits of the created favourable constitutional environment. Broadening the perspective followed in the previous empirical literature—from Blau (1987) to Acs et al. (2009)—and extending previous findings by Carbonara et al. (2016), we take into account the possible moderating effect of a specific aspect of the institutional setting, represented by the provisions contained in a country’s constitution. Particularly, we focus on two central hypotheses: H1a:The level of agency culture and the presence in the constitutions of provisions supporting economic freedom predict a country’s level of entrepreneurial activity. H1b:Following the bounded agency perspective, the constitutional environment moderates the effect of agency culture on entrepreneurship. 4. Data 4.1 Dependent variable 4.1.1 New business density Using data from the World Bank Group Entrepreneurship Database, we construct a measure of new business density, given by the number of new business registrations (private, formal sector companies with limited liability) in every year in each country per 1000 residents aged 15–64 years over the period 2004–2013. Our dependent variable is a standard measure of the total start-up activity in 86 countries (Table 1). It is a measure of entrepreneurship that follows a labor market approach (Audretsch and Fritsch, 1994): all firms are the result of individual actions, and new entrepreneurs are individuals who had previously or have been interested in having a dependent job, who exploit their knowledge of production processes and market features to switch to independent work (Santarelli and Sterlacchini, 1994; Gries and Naudé, 2011). Accordingly, each individual in the labor pool is considered a potential entrepreneur, with the capability to set up his or her own business. We believe that this measure of entrepreneurship is best suited to study the impact of cultural and institutional factors on entrepreneurship because we are in fact focusing on how individual private initiative is fostered or jeopardized by culture and institutions. Other available and commonly used measures include the ratio of new entrants on existing firms, adopted in the so-called ecological approach (Tag et al., 2016), and the number of business owners per labor force (Acs et al., 2009). However, such measures are less interesting for our purposes. The ecological approach measures new start-up activity relative to existing entrepreneurship, thus capturing only one component of the overall process. The percentage of the self-employed is more suited for a study on occupational choices, whereas here we are more interested in a story of entrepreneurial success. 4.2 Independent variables 4.2.1 Agency culture We utilize personality data collected by the ongoing, global Gosling–Potter Internet project (Gosling et al., 2004; see also Gebauer et al., 2015; Rentfrow et al., 2013, 2015). The project collects personality data via a noncommercial Internet website, which can be reached through several channels (e.g., search engines, unsolicited links on other Web pages). People voluntarily participate in this study by responding to items on a standard Big Five personality questionnaire (in English, German, Spanish, or Dutch) using a five-point Likert scale (1 = disagree strongly, 5 = agree strongly); as an incentive, participants receive a personality evaluation based on their responses. Participants also provide responses to questions on several sociodemographic variables, and report their state of residence. This database has yielded numerous publications relating personality traits to various aspects of human behavior. Its validity is supported by the number and quality of publications, mainly in the field of psychology, which have used data from this large-scale Internet project. Of most relevance to the current work, smaller versions of this data set have been successfully employed in cross-cultural studies (Bleidorn et al., 2013; Gebauer et al., 2015). For a list of published studies using the database, see http://www.thebigfiveproject.com/published-papers/). To estimate cross-country differences in agency culture, we use data from all respondents who completed the questionnaire from the start of the project in December 1998 until 2015. In total, N = 7092, 784 respondents are included in this data set. The number of respondents in each country ranges between 1008 (Ethiopia) and 4,275,860 (the United States). Table 2 provides an overview over the sample sizes in each country under study. Country-level agency scores were derived in two steps. In the first step, participants’ extraversion and openness scores were computed and these were averaged to yield an agency score at the individual level. In the second step, individuals’ scores were aggregated within country, yielding country-level scores for agency culture. Table 2. Sample size of the individual-level personality data set for each country Country  N  Country  N  Country  N  Country  N  Afghanistan  1172  Denmark  19,074  Kenya  6985  Portugal  8334  Albania  2288  Dominican R.  6222  Korea (South)  9960  Qatar  2064  Algeria  1074  Egypt  9075  Latvia  1440  Romania  13,055  Argentina  88,211  El Salvador  3682  Lithuania  2277  Russia  3624  Armenia  1084  Estonia  2459  Luxembourg  1116  Serbia and Montenegro  5665  Australia  195,857  Ethiopia  1008  Macedonia  1121  Singapore  59,119  Austria  27,143  Finland  23,526  Malaysia  39,606  Slovak Republic  1691  Bangladesh  3482  France  18,502  Malta  1590  Slovenia  3095  Belgium  43,692  Germany  186,848  Mauritius  1706  South Africa  26,039  Belize  1025  Ghana  1949  Mexico  136,305  Spain  135,048  Bolivia  6115  Greece  10,982  Morocco  1346  Sri Lanka  3958  Bosnia and Herzegovina  1371  Guatemala  5635  Nepal  2142  Sweden  46,828  Brazil  26,538  Hungary  3746  The Netherlands  163,472  Switzerland  36,741  Brunei  1211  India  114,500  New Zealand  43,167  Thailand  8501  Bulgaria  3610  Indonesia  15,199  Nigeria  7033  Turkey  5298  Canada  371,882  Iceland  2520  Norway  42,859  UAE  14,907  Chile  44,552  Ireland  41,257  Oman  1068  Uganda  1377  Colombia  34,905  Israel  7426  Pakistan  27,498  Ukraine  1081  Costa Rica  6712  Italy  13,831  Panama  2938  UK  438,854  Croatia  6920  Jamaica  4199  Perù  23,056  The United States  4,275,860  Cyprus  2307  Japan  10,232  Philippines  91,638  Uruguay  6351  Czech Republic  3566  Jordan  2431  Poland  7951      Country  N  Country  N  Country  N  Country  N  Afghanistan  1172  Denmark  19,074  Kenya  6985  Portugal  8334  Albania  2288  Dominican R.  6222  Korea (South)  9960  Qatar  2064  Algeria  1074  Egypt  9075  Latvia  1440  Romania  13,055  Argentina  88,211  El Salvador  3682  Lithuania  2277  Russia  3624  Armenia  1084  Estonia  2459  Luxembourg  1116  Serbia and Montenegro  5665  Australia  195,857  Ethiopia  1008  Macedonia  1121  Singapore  59,119  Austria  27,143  Finland  23,526  Malaysia  39,606  Slovak Republic  1691  Bangladesh  3482  France  18,502  Malta  1590  Slovenia  3095  Belgium  43,692  Germany  186,848  Mauritius  1706  South Africa  26,039  Belize  1025  Ghana  1949  Mexico  136,305  Spain  135,048  Bolivia  6115  Greece  10,982  Morocco  1346  Sri Lanka  3958  Bosnia and Herzegovina  1371  Guatemala  5635  Nepal  2142  Sweden  46,828  Brazil  26,538  Hungary  3746  The Netherlands  163,472  Switzerland  36,741  Brunei  1211  India  114,500  New Zealand  43,167  Thailand  8501  Bulgaria  3610  Indonesia  15,199  Nigeria  7033  Turkey  5298  Canada  371,882  Iceland  2520  Norway  42,859  UAE  14,907  Chile  44,552  Ireland  41,257  Oman  1068  Uganda  1377  Colombia  34,905  Israel  7426  Pakistan  27,498  Ukraine  1081  Costa Rica  6712  Italy  13,831  Panama  2938  UK  438,854  Croatia  6920  Jamaica  4199  Perù  23,056  The United States  4,275,860  Cyprus  2307  Japan  10,232  Philippines  91,638  Uruguay  6351  Czech Republic  3566  Jordan  2431  Poland  7951      Table 2. Sample size of the individual-level personality data set for each country Country  N  Country  N  Country  N  Country  N  Afghanistan  1172  Denmark  19,074  Kenya  6985  Portugal  8334  Albania  2288  Dominican R.  6222  Korea (South)  9960  Qatar  2064  Algeria  1074  Egypt  9075  Latvia  1440  Romania  13,055  Argentina  88,211  El Salvador  3682  Lithuania  2277  Russia  3624  Armenia  1084  Estonia  2459  Luxembourg  1116  Serbia and Montenegro  5665  Australia  195,857  Ethiopia  1008  Macedonia  1121  Singapore  59,119  Austria  27,143  Finland  23,526  Malaysia  39,606  Slovak Republic  1691  Bangladesh  3482  France  18,502  Malta  1590  Slovenia  3095  Belgium  43,692  Germany  186,848  Mauritius  1706  South Africa  26,039  Belize  1025  Ghana  1949  Mexico  136,305  Spain  135,048  Bolivia  6115  Greece  10,982  Morocco  1346  Sri Lanka  3958  Bosnia and Herzegovina  1371  Guatemala  5635  Nepal  2142  Sweden  46,828  Brazil  26,538  Hungary  3746  The Netherlands  163,472  Switzerland  36,741  Brunei  1211  India  114,500  New Zealand  43,167  Thailand  8501  Bulgaria  3610  Indonesia  15,199  Nigeria  7033  Turkey  5298  Canada  371,882  Iceland  2520  Norway  42,859  UAE  14,907  Chile  44,552  Ireland  41,257  Oman  1068  Uganda  1377  Colombia  34,905  Israel  7426  Pakistan  27,498  Ukraine  1081  Costa Rica  6712  Italy  13,831  Panama  2938  UK  438,854  Croatia  6920  Jamaica  4199  Perù  23,056  The United States  4,275,860  Cyprus  2307  Japan  10,232  Philippines  91,638  Uruguay  6351  Czech Republic  3566  Jordan  2431  Poland  7951      Country  N  Country  N  Country  N  Country  N  Afghanistan  1172  Denmark  19,074  Kenya  6985  Portugal  8334  Albania  2288  Dominican R.  6222  Korea (South)  9960  Qatar  2064  Algeria  1074  Egypt  9075  Latvia  1440  Romania  13,055  Argentina  88,211  El Salvador  3682  Lithuania  2277  Russia  3624  Armenia  1084  Estonia  2459  Luxembourg  1116  Serbia and Montenegro  5665  Australia  195,857  Ethiopia  1008  Macedonia  1121  Singapore  59,119  Austria  27,143  Finland  23,526  Malaysia  39,606  Slovak Republic  1691  Bangladesh  3482  France  18,502  Malta  1590  Slovenia  3095  Belgium  43,692  Germany  186,848  Mauritius  1706  South Africa  26,039  Belize  1025  Ghana  1949  Mexico  136,305  Spain  135,048  Bolivia  6115  Greece  10,982  Morocco  1346  Sri Lanka  3958  Bosnia and Herzegovina  1371  Guatemala  5635  Nepal  2142  Sweden  46,828  Brazil  26,538  Hungary  3746  The Netherlands  163,472  Switzerland  36,741  Brunei  1211  India  114,500  New Zealand  43,167  Thailand  8501  Bulgaria  3610  Indonesia  15,199  Nigeria  7033  Turkey  5298  Canada  371,882  Iceland  2520  Norway  42,859  UAE  14,907  Chile  44,552  Ireland  41,257  Oman  1068  Uganda  1377  Colombia  34,905  Israel  7426  Pakistan  27,498  Ukraine  1081  Costa Rica  6712  Italy  13,831  Panama  2938  UK  438,854  Croatia  6920  Jamaica  4199  Perù  23,056  The United States  4,275,860  Cyprus  2307  Japan  10,232  Philippines  91,638  Uruguay  6351  Czech Republic  3566  Jordan  2431  Poland  7951      4.2.2 Constitutional protection As noted in Section 2.1 above, several provisions contained in national constitutions affect the dynamics of entrepreneurship. Information about constitutional provisions is drawn from the Comparative Constitutions Project: A Cross-National Historical Data set of Written Constitutions (henceforth CCP) (Elkins et al., 2009), an archive of data on the features of written constitutions for most countries since 1789. We focus on four provisions that represent how a constitution can protect the market mechanism, to derive an integrated variable by summing up: right to own property, right to conduct/establish a business, right to free/competitive markets, and independence of the judiciary organs. This Constitutional protection variable ranges from 0 (constitution not mentioning any of the four provisions) to 4 (constitution mentions all four provisions). Constitutional provisions represent the pillars of a country’s legal and institutional framework, and they should be enforced by “lower” laws, which are hierarchically subordinate to constitutions. Often, in fact, lawmakers enact new rules at the constitutional level as a commitment device to guarantee their application (Kelsen, 1967). For example, legal reforms increasing the protection of investors’ rights—and therefore consistent with constitutional protection of the free market—might lead to lower use of control enhancing mechanisms and ultimately create conditions more favorable to the emergence of a corporate economy dominated by widely held corporations (Cuomo et al., 2013). However, to control for their de facto implementation and to measure whether and to what extent “higher” constitutional norms are enforced by the legal and institutional framework, and effectively protect economic freedom, we need to measure the functioning of the market mechanism. For this purpose, we use the Index of Economic Freedom calculated by the Heritage Foundation, (http://www.heritage.org/index/). The index measures economic freedom based on four broad categories, each of which includes three or four types of economic freedom (in parentheses): rule of law (property rights, government integrity, judicial effectiveness), government size (government spending, tax burden, fiscal health), regulatory efficiency (business freedom, labor freedom, monetary freedom), and open markets (trade freedom, investment freedom, financial freedom). Each of the factors shaping the four broad categories is graded on a scale from 0 to 100, and a country’s score is obtained by averaging the resulting 12 values with equal weight given to each. 4.3 Control variables To control for the general economic foundations of each country, we consider the following set of control variables. To capture the wealth of countries and labor market characteristics, we use gross domestic product (GDP) per capita and the percentage of residents aged 15 years or more who are part of the labor force. Other control variables are electric consumption (in Kwh) per capita, as a proxy of the business cycle, and mobile cellular subscription per 100 residents as a proxy of the quality of the infrastructures. Table 3 presents variable descriptions and summary statistics (mean, standard deviation, minimum, and maximum) for all variables included in the analysis, and Table 4 presents the corresponding correlation matrix. Table 3. List of variables and their descriptive statistics: standard deviation is decomposed into between and within components Variable description    Code  Mean  Standard deviation  Minimum  Maximum  New Business density  Overall  Busdensity  5.606  4.899  0.0201  39.699  Between  4.683  0.027  25.826  Within  1.516  −2.981  19.479  Constitutional protection  Overall  Constprot  2.296  1.135  0  4  Between  1.122  0  4  Within  0.202  −0.108  5.091  Agency culture  Overall  Agency  3.477  0.0738  3.283  3.681  Between  0.0736  3.283  3.669  Within  0.0083  3.387  3.557  Economic freedom  Overall  Ecofreedom  63.851  10.652  21.7  89.7  Between  10.533  28.29  87.99  Within  1.864  56.89  73.091  GDP per capita (log)  Overall  Ln GDPcapita  9.307  1.077  5.855  11.212  Between  0.981  6.631  11.133  Within  0.454  6.664  11.604  Labor force participation rate (% of total population +15 years old)    Laborforce  61.921  9.685  37.1  87.7  Between    9.232  39.86  87.2  Within    3.051  41.62  72.08  Electric consumption per capita (Kwh)    Electriccon  5100  6210  30.4  54,799  Between  6135  32.57  43,751  Within  1120  −10,663  16,147  Mobile cellular subscription per 100 people  Overall  Mobilesup  93.814  39.913  0.2  217  Between  30.305  8.345  156.8  Within  26.127  −1.986  181.31  Variable description    Code  Mean  Standard deviation  Minimum  Maximum  New Business density  Overall  Busdensity  5.606  4.899  0.0201  39.699  Between  4.683  0.027  25.826  Within  1.516  −2.981  19.479  Constitutional protection  Overall  Constprot  2.296  1.135  0  4  Between  1.122  0  4  Within  0.202  −0.108  5.091  Agency culture  Overall  Agency  3.477  0.0738  3.283  3.681  Between  0.0736  3.283  3.669  Within  0.0083  3.387  3.557  Economic freedom  Overall  Ecofreedom  63.851  10.652  21.7  89.7  Between  10.533  28.29  87.99  Within  1.864  56.89  73.091  GDP per capita (log)  Overall  Ln GDPcapita  9.307  1.077  5.855  11.212  Between  0.981  6.631  11.133  Within  0.454  6.664  11.604  Labor force participation rate (% of total population +15 years old)    Laborforce  61.921  9.685  37.1  87.7  Between    9.232  39.86  87.2  Within    3.051  41.62  72.08  Electric consumption per capita (Kwh)    Electriccon  5100  6210  30.4  54,799  Between  6135  32.57  43,751  Within  1120  −10,663  16,147  Mobile cellular subscription per 100 people  Overall  Mobilesup  93.814  39.913  0.2  217  Between  30.305  8.345  156.8  Within  26.127  −1.986  181.31  Table 3. List of variables and their descriptive statistics: standard deviation is decomposed into between and within components Variable description    Code  Mean  Standard deviation  Minimum  Maximum  New Business density  Overall  Busdensity  5.606  4.899  0.0201  39.699  Between  4.683  0.027  25.826  Within  1.516  −2.981  19.479  Constitutional protection  Overall  Constprot  2.296  1.135  0  4  Between  1.122  0  4  Within  0.202  −0.108  5.091  Agency culture  Overall  Agency  3.477  0.0738  3.283  3.681  Between  0.0736  3.283  3.669  Within  0.0083  3.387  3.557  Economic freedom  Overall  Ecofreedom  63.851  10.652  21.7  89.7  Between  10.533  28.29  87.99  Within  1.864  56.89  73.091  GDP per capita (log)  Overall  Ln GDPcapita  9.307  1.077  5.855  11.212  Between  0.981  6.631  11.133  Within  0.454  6.664  11.604  Labor force participation rate (% of total population +15 years old)    Laborforce  61.921  9.685  37.1  87.7  Between    9.232  39.86  87.2  Within    3.051  41.62  72.08  Electric consumption per capita (Kwh)    Electriccon  5100  6210  30.4  54,799  Between  6135  32.57  43,751  Within  1120  −10,663  16,147  Mobile cellular subscription per 100 people  Overall  Mobilesup  93.814  39.913  0.2  217  Between  30.305  8.345  156.8  Within  26.127  −1.986  181.31  Variable description    Code  Mean  Standard deviation  Minimum  Maximum  New Business density  Overall  Busdensity  5.606  4.899  0.0201  39.699  Between  4.683  0.027  25.826  Within  1.516  −2.981  19.479  Constitutional protection  Overall  Constprot  2.296  1.135  0  4  Between  1.122  0  4  Within  0.202  −0.108  5.091  Agency culture  Overall  Agency  3.477  0.0738  3.283  3.681  Between  0.0736  3.283  3.669  Within  0.0083  3.387  3.557  Economic freedom  Overall  Ecofreedom  63.851  10.652  21.7  89.7  Between  10.533  28.29  87.99  Within  1.864  56.89  73.091  GDP per capita (log)  Overall  Ln GDPcapita  9.307  1.077  5.855  11.212  Between  0.981  6.631  11.133  Within  0.454  6.664  11.604  Labor force participation rate (% of total population +15 years old)    Laborforce  61.921  9.685  37.1  87.7  Between    9.232  39.86  87.2  Within    3.051  41.62  72.08  Electric consumption per capita (Kwh)    Electriccon  5100  6210  30.4  54,799  Between  6135  32.57  43,751  Within  1120  −10,663  16,147  Mobile cellular subscription per 100 people  Overall  Mobilesup  93.814  39.913  0.2  217  Between  30.305  8.345  156.8  Within  26.127  −1.986  181.31  Table 4. Pairwise correlation matrix (86 countries: average values) Variables  Business density  Consprot  Agency culture  Economic freedom  GDP per capita  Labor force  Electric consump  Mobile subscript  Busdensity  1.0000                Constprot  0.2537*  1.0000              Agency culture  0.2286*  0.2929*  1.0000            Ecofreedom  0.4087*  −0.0938*  −0.0120  1.0000          GDP capita  0.3218*  −0.0026  0.1395*  0.5881*  1.0000        Labor force  0.0892*  −0.0794  −0.1204*  0.0482  −0.0252  1.0000      Electconsum  0.2736*  −0.1627*  −0.0712  0.4574*  0.5637*  0.2294*  1.0000    Mobile subscript  0.2990*  0.0959*  0.0327  0.3813*  0.5021*  −0.0160  0.3423*  1.0000  Variables  Business density  Consprot  Agency culture  Economic freedom  GDP per capita  Labor force  Electric consump  Mobile subscript  Busdensity  1.0000                Constprot  0.2537*  1.0000              Agency culture  0.2286*  0.2929*  1.0000            Ecofreedom  0.4087*  −0.0938*  −0.0120  1.0000          GDP capita  0.3218*  −0.0026  0.1395*  0.5881*  1.0000        Labor force  0.0892*  −0.0794  −0.1204*  0.0482  −0.0252  1.0000      Electconsum  0.2736*  −0.1627*  −0.0712  0.4574*  0.5637*  0.2294*  1.0000    Mobile subscript  0.2990*  0.0959*  0.0327  0.3813*  0.5021*  −0.0160  0.3423*  1.0000  * : Significant at 1% level. Table 4. Pairwise correlation matrix (86 countries: average values) Variables  Business density  Consprot  Agency culture  Economic freedom  GDP per capita  Labor force  Electric consump  Mobile subscript  Busdensity  1.0000                Constprot  0.2537*  1.0000              Agency culture  0.2286*  0.2929*  1.0000            Ecofreedom  0.4087*  −0.0938*  −0.0120  1.0000          GDP capita  0.3218*  −0.0026  0.1395*  0.5881*  1.0000        Labor force  0.0892*  −0.0794  −0.1204*  0.0482  −0.0252  1.0000      Electconsum  0.2736*  −0.1627*  −0.0712  0.4574*  0.5637*  0.2294*  1.0000    Mobile subscript  0.2990*  0.0959*  0.0327  0.3813*  0.5021*  −0.0160  0.3423*  1.0000  Variables  Business density  Consprot  Agency culture  Economic freedom  GDP per capita  Labor force  Electric consump  Mobile subscript  Busdensity  1.0000                Constprot  0.2537*  1.0000              Agency culture  0.2286*  0.2929*  1.0000            Ecofreedom  0.4087*  −0.0938*  −0.0120  1.0000          GDP capita  0.3218*  −0.0026  0.1395*  0.5881*  1.0000        Labor force  0.0892*  −0.0794  −0.1204*  0.0482  −0.0252  1.0000      Electconsum  0.2736*  −0.1627*  −0.0712  0.4574*  0.5637*  0.2294*  1.0000    Mobile subscript  0.2990*  0.0959*  0.0327  0.3813*  0.5021*  −0.0160  0.3423*  1.0000  * : Significant at 1% level. 5. Model development For the purposes of our empirical analysis, we opted for a parsimonious specification, hypothesizing the following structural model:   Yit=α+βCit+γAi+δCit*Ai+θXit+εit, where Yit denotes new business density of country i in year t; Cit is an indicator of constitutional protection of economic freedom and rights in country i and year t; Ai represents a measure of agency culture for country i;  δCit*Ai is the interaction between the constitutional and the psychological variable; Xit is a set of other control variables; and εit is the usual error term. The interaction term is added to the model for testing the hypotheses that the impact exerted by the relationship between endowment of agency culture and the strength of constitutional protection was different for different levels of agency and constitutional protection. The Breusch–Pagan test indicates the presence of heteroskedasticity.1 The White’s method of correcting for heteroskedastic errors should then be applied. The Wooldridge test for autocorrelation in panel data also indicates the presence of serial correlation in our data set2. The Hausman test reveals the existence of an endogeneity problem for our constitutional variable3. Constitution is likely to be endogenous because economies are not exogenously endowed with the institutions and incentives that make up their entrepreneurial environment, but rather institutions are determined endogenously, perhaps influenced by the history, geographical features, and level of entrepreneurship in an economy. The presence of heteroskedasticity, serial correlation, and endogeneity in our data set deserves careful treatment in choosing an appropriate estimation model. On the one hand, robust pooled ordinary least squares (OLS) estimation fails to give unbiased and efficient estimators, and instrumental variable two-staged least square (2SLS) could be a wise choice. On the other hand, our data incur the problem of heteroskedasticity, so we apply the IV generalized method of moments (GMM) technique, which gives more reliable and consistent estimation results (Baum and Schaffer, 2003). The IV-GMM treatment requires the availability and validity of exogenous instruments that are correlated with the independent variables for which endogeneity has been detected, but that are uncorrelated with the measure of new business density. We adopt two instrumental variables: the distance from the equator used by Hall and Jones (1999) and the predicted trade share of an economy constructed by Frankel and Romer (1996). The underidentification test and the Sargan test to detect the relevance and validity of our IVs do support our approach, and thus our choice of instruments is plausible (see Table 6). As robustness checks, we estimate three extra models. We estimate a dynamic Blundell–Bond (Blundell and Bond, 1998) GMM model, including the lagged dependent variable to consider the potential effect of the business cycle and the lagged value of constitutional protection to account for institutional change. This model allows for a low-order moving average correlation in the idiosyncratic errors and is well suited to deal with the low variance in the process of constitutional change, with the time-invariant nature of the agency culture variable, and with the relatively small longitudinal length of the data set (only one decade). Moreover, to account for unobserved country effects across time, the third and the fourth models are country-fixed-effects OLS and generalized least squares (GLS). 6. Empirical results and discussion 6.1 Regression results We start by estimating regressions with Agency culture as the main and only explanatory variable. The results from the dynamic Blundell–Bond (Blundell and Bond, 1998) GMM, the country-fixed-effects OLS, and the country-fixed-effects GLS models presented in Table 5 show a positive and highly statistically significant impact of stronger agency culture on our measure of entrepreneurship: the first part of H1a is therefore supported. Higher labor force participation rate and better infrastructures (i.e., more widespread adoption of mobile cellular phones) are also associated with higher levels of new business density. Table 5. Agency culture and entrepreneurship Variables  Dynamic Blundell–Bond (Blundell and Bond, 1998) GMM  Country-FE with robust SEs  GLS-RE with robust SEs  New Business Density, t−1  0.592***      (0.033)  Agency culture  60.630***  85.985***  61.376***  (5.433)  (15.421)  (13.351)  Economic freedom  0.127***  −0.016  0.027  (0.043)  (0.031)  (0.032)  GDP per capita  −0.625  −1.125  −0.915  (1.841)  (1.701)  (1.763)  GDP per capita squared  0.042  0.059  0.050  (0.101)  (0.101)  (0.097)  Labor force participation  0.005  0.0687***  0.077***  (0.022)  (0.032)  (0.032)  Electric power consumption  −0.001  −0.001***  −0.001***  (0.001)  (0.001)  (0.003)  Mobile cellular subscription  0.002  0.11***  0.011***  (0.003)  (0.004)  (0.004)  Intercept  −215.11***  −291.13***  −210.47***  (20.103)  (53.87)  (47.03)    F-test    33.33***    Wald statistics χ2  2212.50***    40.40***  Observations  774  860  860  Variables  Dynamic Blundell–Bond (Blundell and Bond, 1998) GMM  Country-FE with robust SEs  GLS-RE with robust SEs  New Business Density, t−1  0.592***      (0.033)  Agency culture  60.630***  85.985***  61.376***  (5.433)  (15.421)  (13.351)  Economic freedom  0.127***  −0.016  0.027  (0.043)  (0.031)  (0.032)  GDP per capita  −0.625  −1.125  −0.915  (1.841)  (1.701)  (1.763)  GDP per capita squared  0.042  0.059  0.050  (0.101)  (0.101)  (0.097)  Labor force participation  0.005  0.0687***  0.077***  (0.022)  (0.032)  (0.032)  Electric power consumption  −0.001  −0.001***  −0.001***  (0.001)  (0.001)  (0.003)  Mobile cellular subscription  0.002  0.11***  0.011***  (0.003)  (0.004)  (0.004)  Intercept  −215.11***  −291.13***  −210.47***  (20.103)  (53.87)  (47.03)    F-test    33.33***    Wald statistics χ2  2212.50***    40.40***  Observations  774  860  860  Standard errors in brackets. ***: Significant at 10% level. Table 5. Agency culture and entrepreneurship Variables  Dynamic Blundell–Bond (Blundell and Bond, 1998) GMM  Country-FE with robust SEs  GLS-RE with robust SEs  New Business Density, t−1  0.592***      (0.033)  Agency culture  60.630***  85.985***  61.376***  (5.433)  (15.421)  (13.351)  Economic freedom  0.127***  −0.016  0.027  (0.043)  (0.031)  (0.032)  GDP per capita  −0.625  −1.125  −0.915  (1.841)  (1.701)  (1.763)  GDP per capita squared  0.042  0.059  0.050  (0.101)  (0.101)  (0.097)  Labor force participation  0.005  0.0687***  0.077***  (0.022)  (0.032)  (0.032)  Electric power consumption  −0.001  −0.001***  −0.001***  (0.001)  (0.001)  (0.003)  Mobile cellular subscription  0.002  0.11***  0.011***  (0.003)  (0.004)  (0.004)  Intercept  −215.11***  −291.13***  −210.47***  (20.103)  (53.87)  (47.03)    F-test    33.33***    Wald statistics χ2  2212.50***    40.40***  Observations  774  860  860  Variables  Dynamic Blundell–Bond (Blundell and Bond, 1998) GMM  Country-FE with robust SEs  GLS-RE with robust SEs  New Business Density, t−1  0.592***      (0.033)  Agency culture  60.630***  85.985***  61.376***  (5.433)  (15.421)  (13.351)  Economic freedom  0.127***  −0.016  0.027  (0.043)  (0.031)  (0.032)  GDP per capita  −0.625  −1.125  −0.915  (1.841)  (1.701)  (1.763)  GDP per capita squared  0.042  0.059  0.050  (0.101)  (0.101)  (0.097)  Labor force participation  0.005  0.0687***  0.077***  (0.022)  (0.032)  (0.032)  Electric power consumption  −0.001  −0.001***  −0.001***  (0.001)  (0.001)  (0.003)  Mobile cellular subscription  0.002  0.11***  0.011***  (0.003)  (0.004)  (0.004)  Intercept  −215.11***  −291.13***  −210.47***  (20.103)  (53.87)  (47.03)    F-test    33.33***    Wald statistics χ2  2212.50***    40.40***  Observations  774  860  860  Standard errors in brackets. ***: Significant at 10% level. We then turn to the discussion of the general model (Table 6). As far as the main variables of interest (Agency culture and Constitutional protection) are concerned, the results of the estimates show a consistent pattern across the IV-GMM, the dynamic Blundell–Bond (Blundell and Bond, 1998) GMM, the country-fixed-effects OLS, and the country-fixed-effects GLS models. In Table 6 we present three specifications for each of the four methodological treatments: the first specification controls for the effect of constitutional protection only (Columns 1, 4, and 7); the second specification considers both constitutional protection and psychological agency culture (Columns 2, 5, and 8); and the third specification takes into account their interaction effect as well (Columns 3, 6, and 9). Our data incur the problem of heteroskedasticity, serial autocorrelation, and endogeneity of constitutional protection, so the static IV-GMM model with robust SEs is the most appropriate estimation method; thus, we base our interpretation on the results of this model. Results from the other model specifications are also traced for the purpose of comparison. Table 6. Agency culture, constitutional protection, and entrepreneurship Dependent variable: new business density   Variables  IV-GMM with robust SEsa   Dynamic Blundell–Bond (Blundell and Bond, 1998) GMMb   Country-FE with robust SEs   GLS-RE with rob SEs  (1)  (2)  (3)  (4)  (5)  (6)  (7)  (8)  (9)  New business density, t − 1        0.707***  0.611***  0.633***          (0.031)  (0.031)  (0.031)  Constitutional protection  0.773**  0.635  −2.955***  2.627***  2.146***  0.217  2.185***  2.264***  9.174  −6.084***  (0.346)  (0.401)  (1.059)  (0.327)  (0.302)  (0.385)  (0.605)  (0.606)  (27.515)  (1.937)  Constitutional protection, t − 1        0.106  −0.295  0.51          (0.411)  (0.376)  (0.395)  Agency culture    6.583**  −1.59**    39.354***  18.098***    86.766***  94.795***  −5.329  (3.577)  (0.645)    (4.705)  (5.774)    (15.753)  (35.701)  (16.535)  Constitutional protection * psychological agency      8.884***      0.581***      −2.008  18.089***  (3.049)      (0.089)      (7.998)  (5.566)  Economic freedom  0.225***  0.214***  0.222***  0.149***  0.094***  0.131***  0.030  0.012  0.011  0.043  (0.02)  (0.022)  (0.019)  (0.041)  (0.038)  (0.038)  (0.034)  (0.028)  (0.028)  (0.031)  GDP per capita  5.26***  4.876***  2.598  1.603  0.644  1.235  0.848  1.458  1.484  0.984  (2.21)  (2.06)  (1.838)  (1.893)  (1.721)  (1.736)  (1.814)  (1.654)  (1.65)  (1.825)  GDP per capita squared  −0.301***  −0.282***  −0.156  0.106  0.044  0.073  0.039  0.071  0.073  0.051  (0.122)  (0.114)  (0.102)  (0.104)  (0.095)  (0.095)  (0.099)  (0.091)  (0.091)  (0.1003)  Labor force participation rate  0.005  0.012  0.008  0.022  0.002  0.008  0.061**  0.055**  0.055**  0.062**  (0.015)  (0.017)  (0.012)  (0.022)  (0.021)  (0.021)  (0.027)  (0.026)  (0.026)  (0.027)  Electric power consumption per capita (Kwh)  0.0001***  0.0001***  0.0001***  0.0001***  0.0001***  0.000  0.0002***  0.0002***  0.0002***  0.0001**  (0.000)  (0.000)  (0.000)  (0.0001)  (0.000)  (0.000)  (0.000)  (0.000)  (0.000)  (0.000)  Mobile cellular subscription per 100 people  0.0105***  0.011***  0.008**  0.001  0.001  0.001  0.008**  0.007**  0.007**  0.007**  (0.004)  (0.004)  (0.0039)  (0.002)  (0.002)  (0.002)  (0.003)  (0.003)  (0.003)  (0.003)  Intercept  −35.49***  −55.95***  31.358  −6.980  −142.94***  −70.531***  3.524  −29.65***  −32.414***  16.500  (9.783)  (17.21)  (24.284)  (8.681)  (18.164)  (21.477)  (8.996)  (5.499)  (12.345)  (58.165)    F test  51.03***  44.02***  44.46***        15.61***  40.93***  40.46***    Wald statistic χ2(2)        2095.9***  2627.6***  2631.8***        46.61***  Under-identification test χ2(2)  72.44***  105.3***  64.64***                Over-identification test χ2(1)  4.926  4.618  0.815                Endogeneity test χ2(1)  2.78*  3.006*  6.759***                Observations  860  860  860  774  774  774  860  860  860  860  Dependent variable: new business density   Variables  IV-GMM with robust SEsa   Dynamic Blundell–Bond (Blundell and Bond, 1998) GMMb   Country-FE with robust SEs   GLS-RE with rob SEs  (1)  (2)  (3)  (4)  (5)  (6)  (7)  (8)  (9)  New business density, t − 1        0.707***  0.611***  0.633***          (0.031)  (0.031)  (0.031)  Constitutional protection  0.773**  0.635  −2.955***  2.627***  2.146***  0.217  2.185***  2.264***  9.174  −6.084***  (0.346)  (0.401)  (1.059)  (0.327)  (0.302)  (0.385)  (0.605)  (0.606)  (27.515)  (1.937)  Constitutional protection, t − 1        0.106  −0.295  0.51          (0.411)  (0.376)  (0.395)  Agency culture    6.583**  −1.59**    39.354***  18.098***    86.766***  94.795***  −5.329  (3.577)  (0.645)    (4.705)  (5.774)    (15.753)  (35.701)  (16.535)  Constitutional protection * psychological agency      8.884***      0.581***      −2.008  18.089***  (3.049)      (0.089)      (7.998)  (5.566)  Economic freedom  0.225***  0.214***  0.222***  0.149***  0.094***  0.131***  0.030  0.012  0.011  0.043  (0.02)  (0.022)  (0.019)  (0.041)  (0.038)  (0.038)  (0.034)  (0.028)  (0.028)  (0.031)  GDP per capita  5.26***  4.876***  2.598  1.603  0.644  1.235  0.848  1.458  1.484  0.984  (2.21)  (2.06)  (1.838)  (1.893)  (1.721)  (1.736)  (1.814)  (1.654)  (1.65)  (1.825)  GDP per capita squared  −0.301***  −0.282***  −0.156  0.106  0.044  0.073  0.039  0.071  0.073  0.051  (0.122)  (0.114)  (0.102)  (0.104)  (0.095)  (0.095)  (0.099)  (0.091)  (0.091)  (0.1003)  Labor force participation rate  0.005  0.012  0.008  0.022  0.002  0.008  0.061**  0.055**  0.055**  0.062**  (0.015)  (0.017)  (0.012)  (0.022)  (0.021)  (0.021)  (0.027)  (0.026)  (0.026)  (0.027)  Electric power consumption per capita (Kwh)  0.0001***  0.0001***  0.0001***  0.0001***  0.0001***  0.000  0.0002***  0.0002***  0.0002***  0.0001**  (0.000)  (0.000)  (0.000)  (0.0001)  (0.000)  (0.000)  (0.000)  (0.000)  (0.000)  (0.000)  Mobile cellular subscription per 100 people  0.0105***  0.011***  0.008**  0.001  0.001  0.001  0.008**  0.007**  0.007**  0.007**  (0.004)  (0.004)  (0.0039)  (0.002)  (0.002)  (0.002)  (0.003)  (0.003)  (0.003)  (0.003)  Intercept  −35.49***  −55.95***  31.358  −6.980  −142.94***  −70.531***  3.524  −29.65***  −32.414***  16.500  (9.783)  (17.21)  (24.284)  (8.681)  (18.164)  (21.477)  (8.996)  (5.499)  (12.345)  (58.165)    F test  51.03***  44.02***  44.46***        15.61***  40.93***  40.46***    Wald statistic χ2(2)        2095.9***  2627.6***  2631.8***        46.61***  Under-identification test χ2(2)  72.44***  105.3***  64.64***                Over-identification test χ2(1)  4.926  4.618  0.815                Endogeneity test χ2(1)  2.78*  3.006*  6.759***                Observations  860  860  860  774  774  774  860  860  860  860  *, **, *** Note: : Significant at 10%, 5%, 1% significant level, respectively. (1), (4), (7): Constitutional protection is controlled; (2), (5), (8): constitutional protection and psychological ‘agency’ trait are controlled; (3), (6), (9): constitutional protection, psychological ‘agency’ trait, and their interaction are controlled. a Since at least one of the two instruments should vary over time (i.e., the trade share of the economy), the IV GMM model could overcome the time invariance in the regressors of interest. b Constitutional protection is treated as an endogenous variable. The first lagged value of constitutional protection is used as IV. Table 6. Agency culture, constitutional protection, and entrepreneurship Dependent variable: new business density   Variables  IV-GMM with robust SEsa   Dynamic Blundell–Bond (Blundell and Bond, 1998) GMMb   Country-FE with robust SEs   GLS-RE with rob SEs  (1)  (2)  (3)  (4)  (5)  (6)  (7)  (8)  (9)  New business density, t − 1        0.707***  0.611***  0.633***          (0.031)  (0.031)  (0.031)  Constitutional protection  0.773**  0.635  −2.955***  2.627***  2.146***  0.217  2.185***  2.264***  9.174  −6.084***  (0.346)  (0.401)  (1.059)  (0.327)  (0.302)  (0.385)  (0.605)  (0.606)  (27.515)  (1.937)  Constitutional protection, t − 1        0.106  −0.295  0.51          (0.411)  (0.376)  (0.395)  Agency culture    6.583**  −1.59**    39.354***  18.098***    86.766***  94.795***  −5.329  (3.577)  (0.645)    (4.705)  (5.774)    (15.753)  (35.701)  (16.535)  Constitutional protection * psychological agency      8.884***      0.581***      −2.008  18.089***  (3.049)      (0.089)      (7.998)  (5.566)  Economic freedom  0.225***  0.214***  0.222***  0.149***  0.094***  0.131***  0.030  0.012  0.011  0.043  (0.02)  (0.022)  (0.019)  (0.041)  (0.038)  (0.038)  (0.034)  (0.028)  (0.028)  (0.031)  GDP per capita  5.26***  4.876***  2.598  1.603  0.644  1.235  0.848  1.458  1.484  0.984  (2.21)  (2.06)  (1.838)  (1.893)  (1.721)  (1.736)  (1.814)  (1.654)  (1.65)  (1.825)  GDP per capita squared  −0.301***  −0.282***  −0.156  0.106  0.044  0.073  0.039  0.071  0.073  0.051  (0.122)  (0.114)  (0.102)  (0.104)  (0.095)  (0.095)  (0.099)  (0.091)  (0.091)  (0.1003)  Labor force participation rate  0.005  0.012  0.008  0.022  0.002  0.008  0.061**  0.055**  0.055**  0.062**  (0.015)  (0.017)  (0.012)  (0.022)  (0.021)  (0.021)  (0.027)  (0.026)  (0.026)  (0.027)  Electric power consumption per capita (Kwh)  0.0001***  0.0001***  0.0001***  0.0001***  0.0001***  0.000  0.0002***  0.0002***  0.0002***  0.0001**  (0.000)  (0.000)  (0.000)  (0.0001)  (0.000)  (0.000)  (0.000)  (0.000)  (0.000)  (0.000)  Mobile cellular subscription per 100 people  0.0105***  0.011***  0.008**  0.001  0.001  0.001  0.008**  0.007**  0.007**  0.007**  (0.004)  (0.004)  (0.0039)  (0.002)  (0.002)  (0.002)  (0.003)  (0.003)  (0.003)  (0.003)  Intercept  −35.49***  −55.95***  31.358  −6.980  −142.94***  −70.531***  3.524  −29.65***  −32.414***  16.500  (9.783)  (17.21)  (24.284)  (8.681)  (18.164)  (21.477)  (8.996)  (5.499)  (12.345)  (58.165)    F test  51.03***  44.02***  44.46***        15.61***  40.93***  40.46***    Wald statistic χ2(2)        2095.9***  2627.6***  2631.8***        46.61***  Under-identification test χ2(2)  72.44***  105.3***  64.64***                Over-identification test χ2(1)  4.926  4.618  0.815                Endogeneity test χ2(1)  2.78*  3.006*  6.759***                Observations  860  860  860  774  774  774  860  860  860  860  Dependent variable: new business density   Variables  IV-GMM with robust SEsa   Dynamic Blundell–Bond (Blundell and Bond, 1998) GMMb   Country-FE with robust SEs   GLS-RE with rob SEs  (1)  (2)  (3)  (4)  (5)  (6)  (7)  (8)  (9)  New business density, t − 1        0.707***  0.611***  0.633***          (0.031)  (0.031)  (0.031)  Constitutional protection  0.773**  0.635  −2.955***  2.627***  2.146***  0.217  2.185***  2.264***  9.174  −6.084***  (0.346)  (0.401)  (1.059)  (0.327)  (0.302)  (0.385)  (0.605)  (0.606)  (27.515)  (1.937)  Constitutional protection, t − 1        0.106  −0.295  0.51          (0.411)  (0.376)  (0.395)  Agency culture    6.583**  −1.59**    39.354***  18.098***    86.766***  94.795***  −5.329  (3.577)  (0.645)    (4.705)  (5.774)    (15.753)  (35.701)  (16.535)  Constitutional protection * psychological agency      8.884***      0.581***      −2.008  18.089***  (3.049)      (0.089)      (7.998)  (5.566)  Economic freedom  0.225***  0.214***  0.222***  0.149***  0.094***  0.131***  0.030  0.012  0.011  0.043  (0.02)  (0.022)  (0.019)  (0.041)  (0.038)  (0.038)  (0.034)  (0.028)  (0.028)  (0.031)  GDP per capita  5.26***  4.876***  2.598  1.603  0.644  1.235  0.848  1.458  1.484  0.984  (2.21)  (2.06)  (1.838)  (1.893)  (1.721)  (1.736)  (1.814)  (1.654)  (1.65)  (1.825)  GDP per capita squared  −0.301***  −0.282***  −0.156  0.106  0.044  0.073  0.039  0.071  0.073  0.051  (0.122)  (0.114)  (0.102)  (0.104)  (0.095)  (0.095)  (0.099)  (0.091)  (0.091)  (0.1003)  Labor force participation rate  0.005  0.012  0.008  0.022  0.002  0.008  0.061**  0.055**  0.055**  0.062**  (0.015)  (0.017)  (0.012)  (0.022)  (0.021)  (0.021)  (0.027)  (0.026)  (0.026)  (0.027)  Electric power consumption per capita (Kwh)  0.0001***  0.0001***  0.0001***  0.0001***  0.0001***  0.000  0.0002***  0.0002***  0.0002***  0.0001**  (0.000)  (0.000)  (0.000)  (0.0001)  (0.000)  (0.000)  (0.000)  (0.000)  (0.000)  (0.000)  Mobile cellular subscription per 100 people  0.0105***  0.011***  0.008**  0.001  0.001  0.001  0.008**  0.007**  0.007**  0.007**  (0.004)  (0.004)  (0.0039)  (0.002)  (0.002)  (0.002)  (0.003)  (0.003)  (0.003)  (0.003)  Intercept  −35.49***  −55.95***  31.358  −6.980  −142.94***  −70.531***  3.524  −29.65***  −32.414***  16.500  (9.783)  (17.21)  (24.284)  (8.681)  (18.164)  (21.477)  (8.996)  (5.499)  (12.345)  (58.165)    F test  51.03***  44.02***  44.46***        15.61***  40.93***  40.46***    Wald statistic χ2(2)        2095.9***  2627.6***  2631.8***        46.61***  Under-identification test χ2(2)  72.44***  105.3***  64.64***                Over-identification test χ2(1)  4.926  4.618  0.815                Endogeneity test χ2(1)  2.78*  3.006*  6.759***                Observations  860  860  860  774  774  774  860  860  860  860  *, **, *** Note: : Significant at 10%, 5%, 1% significant level, respectively. (1), (4), (7): Constitutional protection is controlled; (2), (5), (8): constitutional protection and psychological ‘agency’ trait are controlled; (3), (6), (9): constitutional protection, psychological ‘agency’ trait, and their interaction are controlled. a Since at least one of the two instruments should vary over time (i.e., the trade share of the economy), the IV GMM model could overcome the time invariance in the regressors of interest. b Constitutional protection is treated as an endogenous variable. The first lagged value of constitutional protection is used as IV. Our findings show that constitutional protection of economic freedom plays a key role in generating the wide variation in entrepreneurship across countries (Columns 1, 4 and 7). The second part of H1a is therefore confirmed. Consistent with the findings of Bell et al. (2008) and Parker (2009: Ch. 15) on legal rules, we too find that when institutions support economic freedom, entrepreneurship is stronger. With respect to diffusion in each country’s population of the agency culture (Columns 2, 5 and 8), in all estimates the coefficient is largely positive and statistically significant at a 99% confidence level. A stronger agency culture is thus associated with a higher level of new business density, and corresponds to a greater propensity of individuals to create their own entrepreneurial ventures. Unsurprisingly, countries with high agency culture foster the development of a more dynamic entrepreneurial climate. Our findings therefore suggest a possible way of promoting the propensity to become entrepreneurs in the overall population through training activities for elating agentic characteristics, such as creativity, risk-taking propensity, and entrepreneurial proactiveness. However, when the impact of the psychological “agency” trait is taken into account, the institutional factor, despite having positive influence, loses its statistical significance on new business density in the IV-GMM while maintaining it in all other model specifications. Looking at the interaction between agency culture and constitutional protection (Columns 3, 6 and 9), consistent with the bounded agency approach, and with our H1b, the effect of agency culture on entrepreneurship is indeed bounded by the characteristics of the national constitutions. These two factors do combine in such a way that the effect of high agency culture is strengthened in countries with a pro-entrepreneurship constitution. Similarly, the constitutional protection of economic freedom leverages the entrepreneurial intention of countries with a great endowment of agency culture. A healthy business environment facilitates nascent entrepreneurs to discover and exploit entrepreneurial opportunities. The positive effect of constitutional protection and agency culture is also corroborated by the positive and statistically significant coefficient of our de facto measure, the Index of Economic Freedom, in both GMM estimates and by a positive, albeit insignificant, effect in the other model specifications. The impact of economic freedom is in line with the effect of constitutional variables promoting economic rights. A free and competitive market creates a favorable business environment and a level playing field to both incumbent and nascent entrepreneurs. With respect to control variables, there are several findings of note. First, in the IV-GMM, GDP per capita is positive and statistically significant, whereas its squared value is negative and statistically significant. The positive and negative signs are maintained in all other model specifications, although statistical significance disappears. This pattern is consistent with the findings from many empirical studies (Koellinger and Thurik, 2012) and confirms that, on the aggregate level, GDP cycles do predict the entrepreneurial cycle, although the relationship between GDP per capita and entrepreneurship is quadratic. In general, high GDP per capita reflects stronger demands, which leads to an abundance of emerging entrepreneurial opportunities, and in turn induces new entries to capture such opportunities (Santarelli and Tran, 2013). Second, a higher labor force participation rate is conducive to more entrepreneurship. Provided that a large fraction of new entrepreneurs is usually represented by individuals previously involved in paid employment (Storey and Jones, 1987), this last finding is straightforward. Third, countries with high electric power consumption per capita and mobile cellular subscription are more likely to enjoy a higher rate of new firm formation. 6.2 Illustrating the interaction effect between agency culture and constitutional protection Based on estimation of the baseline equation by GLS random-effects (RE) with a robust standard errors model (whose estimation results are presented in the final column of Table 6), we used the command “margins” in Stata 14 to estimate the margins of responses for specified values of covariates and present the results as a table. Finally, to draw the interaction plot we used the command “marginsplot” to graph the results of the “margins” command. Plots were also constructed at each of the five specified values of the “constitutional protection” variable from 0 to 4. Figure 1 presents the interaction plots of constitutional protection and agency culture, with the aim of capturing the moderating effect of constitutional provisions on psychological agency on a country’s proneness toward entrepreneurship. Figure 1. View largeDownload slide Interaction plot of constitutional protection and agency culture. Figure 1. View largeDownload slide Interaction plot of constitutional protection and agency culture. The plot on the right combines all five interaction lines, whereas the five smaller plots on the left separate interaction lines for each value of the constitutional protection variable. The combined interaction plot on the right has five lines representing the indirect effect of agency culture on new business density at five different values of constitutional protection (from 0 to 4). For ease of comparison, the interaction plot on the left separates these five interaction lines into five charts. Obviously, if the interaction is not significant, the plotted lines should be parallel, which is clearly not the case here. If a country does not have any constitutional provision supporting economic freedom (constprot = 0), agency culture is negatively associated with new business density. No matter how proactive and innovative its citizens are, they are just simply unmotivated to set up new ventures in an environment where they cannot enjoy free competition or the freedom to set up business and own their property. Second, when the country increases its constitutional protection of economic freedom (constprot increases from 1 to 3), the line representing the relationship between agency culture and entrepreneurship moves up gradually, while the impact of agency turns from negative to slightly positive. Countries with higher psychological agency find themselves more entrepreneurial with a higher rate of self-employment over time when they start to apply constitutional provision protecting economic freedom. Finally, the effect of psychological agency turns out significantly positive when countries possess a high level of constitutional protection (constprot = 4). These results suggest that high agentic economies enjoying a constitutional protection of economic freedom are particularly powerful cradles, nurturing entrepreneurial activity. Creative and innovative people are more motivated to exploit their ideas in a transparent and healthy business environment in which they do not need to care about bribes or corruption. Thus, if governments find that the majority of the country’s population has high levels of agency, they could help to capitalize on this entrepreneurial trait by creating a healthy institutional environment supporting free competition, business and property rights, and an independent judicial system. These steps would significantly foster a dynamic entrepreneurial sector within the economy. 7. Conclusions Inspired by the central role that the concept of (psychological) agency plays in seminal theorizing in entrepreneurship (Schumpeter, 1911, 1934; McClelland, 1961), the present study is to our knowledge, the first systematic attempt to examine the effect of agency culture on national entrepreneurship rates, with a special focus on an important contextual moderator—formal institutions such as constitutional provisions relevant for entrepreneurship. In general, our findings reveal that: (i) a greater endowment of agency culture is associated with a country’s higher willingness or intention to start a business; (ii) constitutional protection of economic freedom plays a key role in generating the observed wide variation in entrepreneurship across countries, by exerting a moderating effect on how a certain endowment of agency culture influences a country’s proneness toward entrepreneurship; (iii) when institutions do support economic freedom, as denoted by higher values of the Index of Economic Freedom, entrepreneurship is stronger. In sum, there is an interaction between constitutional and legal protection of economic freedom on the one side and the presence in the country of a large fraction of individuals characterized by an agentic personality on the other side. In particular, there seems to be a benefit from instituting stronger protections of economic freedom, and such benefit is the stronger the higher the level of agency culture characterizing the country. Footnotes 1 χ2(1) = 142.17; P-value = 0.0000. 2 F(1, 85) = 15.944; P-value = 0.0001. 3 χ2(2) = 12.982; P-value = 0.0015. Acknowledgments The authors thank Tom Ginsburg for providing us with the data from the Comparative Constitutions Project, Jerg Gutmann and an anonymous reviewer for helpful suggestions. Previous versions of this article have been presented at the Workshop on The Future of Small Business Economics in Utrecht (February 10, 2017), the Midterm Law and Economics Conference in Ghent (February 17, 2017), and the 34th Annual Conference of the European Association of Law and Economics (London, September 14–16, 2017). E.S. gratefully acknowledges financial support from Saarland University within the framework of the program “EXIST-Culture of Entrepreneurship—The Entrepreneurship University” arranged by the German Federal Ministry of Economics and Technology. References Abele A. E., Wojciszke B. ( 2007), ‘ Agency and communion from the perspective of self versus others,’ Journal of Personality and Social Psychology , 93( 5), 751– 763. Google Scholar CrossRef Search ADS   Acemoglu D., Jackson M. O. ( 2016), ‘Social norms and the enforcement of laws,’ MIT Department of Economics Working Paper No. 14-16. Acemoglu D., Johnson S., Robinson J. A. ( 2001), ‘ The colonial origins of comparative development: an empirical investigation,’ American Economic Review , 91( 5), 1369– 1401. Google Scholar CrossRef Search ADS   Acemoglu D., Robinson J. A. ( 2012), Why Nations Fail . Crown Business: New York, NY. Acs Z. J., Audretsch D. B., Braunerhjelm P., Carlsson B. ( 2009), ‘ The knowledge spillover theory of entrepreneurship,’ Small Business Economics , 32( 1), 15– 30. Google Scholar CrossRef Search ADS   Acs Z. J., Desai S., Hessels J. ( 2008), ‘ Entrepreneurship, economic development and institutions,’ Small Business Economics , 31( 3), 219– 234. Google Scholar CrossRef Search ADS   Ajzen I. ( 1985), ‘From intentions to actions: a theory of planned behavior,’ in Kuhl J., Beckman J. (eds), Action-Control: From Cognition to Behavior . Springer: Heidelberg, pp. 11– 39. Alesina A., Giuliano P. ( 2015), ‘ Culture and institutions,’ Journal of Economic Literature , 53( 4), 898– 944. Google Scholar CrossRef Search ADS   Audretsch D. B., Fritsch M. ( 1994), ‘ The geography of firm births in Germany,’ Regional Studies , 28( 4), 359– 365. Google Scholar CrossRef Search ADS   Audretsch D. B., Obschonka M., Gosling S. D., Potter J. ( 2017), ‘ A new perspective on entrepreneurial regions: linking cultural identity with latent and manifest entrepreneurship,’ Small Business Economics , 48( 3), 681– 697. Google Scholar CrossRef Search ADS   Audretsch D. B., Thurik A. R. ( 2000), ‘ Capitalism and democracy in the 21st century: from the managed to the entrepreneurial economy,’ Journal of Evolutionary Economics , 10( 1–2), 17– 34. Google Scholar CrossRef Search ADS   Auriol E., Gary-Bobo R. J. ( 2012), ‘ On the optimal number of representatives,’ Public Choice , 153( 3–4), 419– 445. Google Scholar CrossRef Search ADS   Autio E., Kenney M., Mustar P., Siegel D., Wright M. ( 2014), ‘ Entrepreneurial innovation: the importance of context,’ Research Policy , 43( 7), 1097– 1108. Google Scholar CrossRef Search ADS   Bandura A. ( 2006), ‘ Toward a psychology of human agency,’ Perspectives on Psychological Science , 1( 2), 164– 180. Google Scholar CrossRef Search ADS   Baron R. A., Baum J. R. ( 2007), ‘Research gains: benefits of closer links between I/O psychology and entrepreneurship,’ in Baum J. R., Frese M., Baron R. A. (eds), The Psychology of Entrepreneurship . Erlbaum: Mahwah, NJ, pp. 347– 373. Bartelsman E., Haltiwanger J., Scarpetta S. ( 2009), ‘Measuring and analyzing cross-country differences in firm dynamics,’ in Dunne T., Jensen J. B., Roberts M. J. (eds), Producer Dynamics: New Evidence from Micro Data . University of Chicago Press for NBER: Chicago, pp. 15– 79. Google Scholar CrossRef Search ADS   Bartelsman E., Haltiwanger J., Scarpetta S. ( 2013), ‘ Cross-country differences in productivity: the role of allocation and selection,’ American Economic Review , 103( 1), 305– 334. Google Scholar CrossRef Search ADS   Baum C. F., Schaffer M. E. ( 2003), ‘ Instrumental variables and GMM: estimation and testing,’ Stata Journal , 3, 1– 31. Begley T. M., Boyd D. P. ( 1987), ‘ Psychological characteristics associated with performance in entrepreneurial firms and smaller businesses,’ Journal of Business Venturing , 2( 1), 79– 93. Google Scholar CrossRef Search ADS   Bell R. G., Moore C. B., Al-Shammari H. A. ( 2008), ‘ Country of origin and foreign IPO legitimacy,’ Entrepreneurship Theory and Practice , 32( 1), 185– 202. Google Scholar CrossRef Search ADS   Benet-Martinez V., John O. P. ( 1998), ‘ Los Cinco grandes across cultures and ethnic groups: multitrait-multimethod analyses of the big five in Spanish and English,’ Journal of Personality and Social Psychology , 75, 729– 750. Google Scholar CrossRef Search ADS   Besley T., Ghatak M. ( 2010), ‘Property rights and economic development,’ in Rodrik D., Rosenzweig M. (eds), Handbook of Development Economics , Vol. 5. Elsevier: North Holland, The Netherlands, pp. 4525– 4595. Google Scholar CrossRef Search ADS   Bjørnskov C., Foss N. J. ( 2008), ‘ Economic freedom and entrepreneurial activity: some cross-country evidence,’ Public Choice , 134( 3–4), 307– 328. Google Scholar CrossRef Search ADS   Blackburn R., Renwick S. J., Donnelly J. P., Logan C. ( 2004), ‘ Big five or big two? Superordinate factors in the NEO five factor inventory and the antisocial personality questionnaire,’ Personality and Individual Differences , 37( 5), 957– 970. Google Scholar CrossRef Search ADS   Blanchflower D. G. ( 2000), ‘ Self-employment in OECD countries,’ Labor Economics , 7( 5), 471– 505. Google Scholar CrossRef Search ADS   Blau D. M. ( 1987), ‘ A time-series analysis of self-employment in the United States,’ Journal of Political Economy , 95( 3), 445– 467. Google Scholar CrossRef Search ADS   Bleidorn W., Klimstra T. A., Denissen J. J. A., Rentfrow P. J., Potter J., Gosling S. D. ( 2013), ‘ Personality maturation around the world: a cross-cultural examination of social-investment theory,’ Psychological Science , 24( 12), 2530– 2540. Google Scholar CrossRef Search ADS   Blume L., Müller J., Voigt S., Wolf C. ( 2009), ‘ The economic effects of constitutions: replicating—and extending—Persson and Tabellini,’ Public Choice , 139( 1–2), 197– 225. Google Scholar CrossRef Search ADS   Blundell R., Bond S. ( 1998), ‘ Initial conditions and moment restrictions in dynamic panel data models,’ Journal of Econometrics , 87( 1), 115– 143. Google Scholar CrossRef Search ADS   Borghans L., Duckworth A. L., Heckman J. J., Ter Weel B. ( 2008), ‘ The economics and psychology of personality traits,’ Journal of Human Resources , 43( 4), 972– 1059. Bosse D. A., Phillips R. A. ( 2016), ‘ Agency theory and bounded self-interest,’ Academy of Management Review , 41( 2), 276– 297. Google Scholar CrossRef Search ADS   Bottazzi G., Dosi G., Jacoby N., Secchi A., Tamagni F. ( 2010), ‘ Corporate performances and market selection: some comparative evidence,’ Industrial and Corporate Change , 19( 6), 1953– 1996. Google Scholar CrossRef Search ADS   Brandstätter H. ( 2011), ‘ Personality aspects of entrepreneurship: a look at five meta-analyses,’ Personality and Individual Differences , 51( 3), 222– 230. Google Scholar CrossRef Search ADS   Brück T., Lussá F., Tavares J. A. ( 2011), ‘ Entrepreneurship: the role of extreme events,’ European Journal of Political Economy , 27, S78– S88. Google Scholar CrossRef Search ADS   Buchanan J., Tullock G. ( 1962), The Calculus of Consent—Logical Foundations of Constitutional Democracy . University of Michigan Press: Ann Arbor. Google Scholar CrossRef Search ADS   Carbonara E., Parisi F., von Wangenheim G. ( 2012), ‘ Unjust laws and illegal norms,’ International Review of Law and Economics , 32( 3), 285– 299. Google Scholar CrossRef Search ADS   Carbonara E., Santarelli E., Tran H. T. ( 2016), ‘ De Jure determinants of new firm formation: how the pillars of constitutions influence entrepreneurship,’ Small Business Economics , 47( 1), 139– 162. Google Scholar CrossRef Search ADS   Carlsson B., Acs Z. J., Audretsch D. B., Braunerhjelm P. ( 2009), ‘ Knowledge creation, and economic growth: a historical review,’ Industrial and Corporate Change , 18( 6), 1193– 1229. Google Scholar CrossRef Search ADS   Carree M., van Stel A., Thurik R., Wennekers S. ( 2002), ‘ Economic development and business ownership: an analysis using data of 23 OECD countries in the period 1976-1996,’ Small Business Economics , 19( 3), 271– 290. Google Scholar CrossRef Search ADS   Caves R. E. ( 1998), ‘ Industrial organization and new findings on the turnover and mobility of firms,’ Journal of Economic Literature , 36, 1947– 1982. Chemin M. ( 2009), ‘ The impact of the judiciary on entrepreneurship: evaluation of Pakistan’s access to justice programme,’ Journal of Public Economics , 93( 1–2), 114– 125. Google Scholar CrossRef Search ADS   Chemin M. ( 2012), ‘ Does court speed shape economic activity? Evidence from a court reform in India,’ Journal of Law, Economics, and Organization , 28( 3), 460– 485. Google Scholar CrossRef Search ADS   Cuomo F., Zattoni A., Valentini G. ( 2013), ‘ The effects of legal reforms on the ownership structure of listed companies,’ Industrial and Corporate Change , 22( 2), 427– 458. Google Scholar CrossRef Search ADS   Czarnitzki D., Doherr T., Hussinger K., Schliessler P., Toole A. A. ( 2016), ‘ Knowledge creates markets: the influence of entrepreneurial support and patent rights on academic entrepreneurship,’ European Economic Review , 86, 131– 146. Google Scholar CrossRef Search ADS   Davidsson P. ( 1995), ‘ Culture, structure and regional levels of entrepreneurship,’ Entrepreneurship and Regional Development , 7( 1), 41– 62. Google Scholar CrossRef Search ADS   DeYoung C. G. ( 2006), ‘ Higher-order factors of the big five in a multi-informant sample,’ Journal of Personality and Social Psychology , 91( 6), 1138– 1151. Google Scholar CrossRef Search ADS   Dheer R. J. S. ( 2017), ‘ Cross-national differences in entrepreneurial activity: role of culture and institutional factors,’ Small Business Economics , 48( 4), 813– 842. Google Scholar CrossRef Search ADS   Diamond J. ( 1997), Guns, Germs, and Steel. The Fates of Human Societies . W.W. Norton & Company: New York, NY; London. Digman J. M. ( 1997), ‘ Higher-order factors of the big five,’ Journal of Personality and Social Psychology , 73( 6), 1246– 56. Google Scholar CrossRef Search ADS   Djankov S., La Porta R., Lopez-De-Silanes F., Shleifer A. ( 2003), ‘ Courts,’ The Quarterly Journal of Economics , 118( 2), 453– 517. Google Scholar CrossRef Search ADS   Elder G. H. ( 1994), ‘ Time, human agency, and social change: perspectives on the life course,’ Social Psychology Quarterly , 57( 1), 4– 15. Google Scholar CrossRef Search ADS   Elkins Z., Ginsburg T., Melton J. ( 2009), The Comparative Constitutions Project: A Cross-National Historical Dataset of Written Constitutions . Mimeo: Chicago. Frankel J. A., Romer D. ( 1996), ‘Trade and growth: an empirical investigation,’ NBER Working paper series #5476. Frese M. ( 2009), ‘ Towards a psychology of entrepreneurship—an action theory perspective,’ Foundations and Trends in Entrepreneurship , 5( 6), 437– 496. Google Scholar CrossRef Search ADS   Frey B. S. ( 2008), Happiness: A Revolution in Economics . MIT Press Books: Cambridge, MA. Google Scholar CrossRef Search ADS   Funder D. C. ( 2006), ‘ Towards a resolution of the personality triad: persons, situations, and behaviors,’ Journal of Research in Personality , 40( 1), 21– 34. Google Scholar CrossRef Search ADS   Galor O., Michalopoulos S. ( 2012), ‘ Evolution and the growth process: natural selection of entrepreneurial traits,’ Journal of Economic Theory , 147( 2), 759– 780. Google Scholar CrossRef Search ADS   García-Posada M., Mora-Sanguinetti J. S. ( 2015), ‘ Entrepreneurship and enforcement institutions: disaggregated evidence for Spain,’ European Journal of Law and Economics , 40( 1), 49– 74. Google Scholar CrossRef Search ADS   Gebauer J. E., Bleidorn W., Gosling S. D., Rentfrow P. J., Lamb M. E., Potter J. ( 2014), ‘ Cross-cultural variations in big five relationships with religiosity: a sociocultural motives perspective,’ Journal of Personality and Social Psychology , 107( 6), 1064– 1091. Google Scholar CrossRef Search ADS   Gebauer J. E., Sedikides C., Lüdtke O., Neberich W. ( 2014), ‘ Agency‐communion and interest in prosocial behavior: social motives for assimilation and contrast explain sociocultural inconsistencies,’ Journal of Personality , 82( 5), 452– 466. Google Scholar CrossRef Search ADS   Gebauer J. E., Sedikides C., Wagner J., Bleidorn W., Rentfrow P. J., Potter J., Gosling S. D. ( 2015), ‘ Cultural norm fulfilment, interpersonal belonging, or getting ahead? A large-scale cross-cultural test of three perspectives on the function of self-esteem,’ Journal of Personality and Social Psychology , 109( 3), 526– 548. Google Scholar CrossRef Search ADS   Glaeser E. L., Laibson D. I., Scheinkman J. A., Soutter C. L. ( 2000), ‘ Measuring Trust,’ Quarterly Journal of Economics , 115( 3), 811– 846, Google Scholar CrossRef Search ADS   Gorodnichenko Y., Roland G. ( 2010), ‘Culture, institutions and the wealth of nations,’ NBER Working Paper No. 16368. Gosling S. D., Vazire S., Srivastava S., John O. P. ( 2004), ‘ Should we trust web-based studies? A comparative analysis of six preconceptions about internet questionnaires,’ American Psychologist , 59( 2), 93– 104. Google Scholar CrossRef Search ADS   Greif A. ( 1994), ‘ Cultural beliefs and the organization of society: a historical and theoretical reflection on collectivist and individualist societies,’ Journal of Political Economy , 102( 5), 912– 950. Google Scholar CrossRef Search ADS   Gries T., Naudé W. ( 2011), ‘ Entrepreneurship and human development: a capability approach,’ Journal of Public Economics , 95( 3–4), 216– 224. Google Scholar CrossRef Search ADS   Grilo I., Thurik A. R. ( 2006), ‘Entrepreneurship in the old and new Europe,’ in Santarelli E. (ed.), Entrepreneurship, Growth, and Innovation. The Dynamics of Firms and Industries , Springer: New York, NY, pp. 75– 103. Google Scholar CrossRef Search ADS   Guiso L., Sapienza P., Zingales L. ( 2003), ‘ People’s Opium? Religion and economic attitudes,’ Journal of Monetary Economics , 50( 1), 225– 282. Google Scholar CrossRef Search ADS   Guiso L., Sapienza P., Zingales L. ( 2006), ‘ Does culture affect economic outcomes?,’ Journal of Economic Perspectives , 20( 2), 23– 48. Google Scholar CrossRef Search ADS   Guiso L., Sapienza P., Zingales L. ( 2008), ‘ Alfred Marshall lecture—social capital as good culture,’ Journal of the European Economic Association , 6( 2–3), 295– 320. Google Scholar CrossRef Search ADS   Hall R. E., Jones C. I. ( 1999), ‘ Why do some countries produce so much more output per worker than others?,’ Quarterly Journal of Economics , 114( 1), 83– 116. Google Scholar CrossRef Search ADS   Hechavarria D. M., Renko M., Matthews C. H. ( 2012), ‘ The nascent entrepreneurship hub: goals, entrepreneurial self-efficacy and start-up outcomes,’ Small Business Economics , 39( 3), 685– 701. Google Scholar CrossRef Search ADS   Hmieleski K. M., Carr J. C., Baron R. A. ( 2015), ‘ Integrating discovery and creation perspectives of entrepreneurial action: the relative roles of founding CEO human capital, social capital, and psychological capital in contexts of risk versus uncertainty,’ Strategic Entrepreneurship Journal , 9( 4), 289– 312. Google Scholar CrossRef Search ADS   Hofstede G., McCrae R. R. ( 2004), ‘ Personality and culture revisited: linking traits and dimensions of culture,’ Cross-Cultural Research , 38( 1), 52– 88. Google Scholar CrossRef Search ADS   John O. P., Srivastava S. ( 1999), ‘The big five trait taxonomy: history, measurement, and theoretical perspectives,’ in Pervin L. A., John O. P. (eds), Handbook of Personality: Theory and Research . Guilford Press: New York, NY, pp. 102– 138. Jokela M., Bleidorn W., Lamb M. E., Gosling S. D., Rentfrow P. J. ( 2015), ‘ Geographically varying associations between personality and life satisfaction in the London Metropolitan area,’ Proceedings of the National Academy of Sciences , 112( 3), 725– 730. Google Scholar CrossRef Search ADS   Kelsen H. ( 1967), Pure Theory of Law . University of California Press: Berkeley; Los Angeles. Kihlstrom R. E., Laffont J. J. ( 1979), ‘ A general equilibrium entrepreneurial theory of firm formation based on risk aversion,’ Journal of Political Economy , 87( 4), 719– 748. Google Scholar CrossRef Search ADS   Koellinger P. D., Thurik A. R. ( 2012), ‘ Entrepreneurship and the business cycle,’ Review of Economics and Statistics , 94( 4), 1143– 1156. Google Scholar CrossRef Search ADS   Lang F. R., John D., Lüdtke O., Schupp J., Wagner G. G. ( 2011), ‘ Short assessment of the big five: robust across survey methods except telephone interviewing,’ Behavior Research Methods , 43( 2), 548– 567. Google Scholar CrossRef Search ADS   Lerner J., Tåg J. ( 2013), ‘ Institutions and venture capital,’ Industrial and Corporate Change , 22( 1), 153– 182. Google Scholar CrossRef Search ADS   Leutner F., Ahmetoglu G., Akhtar R., Chamorro-Premuzic T. ( 2014), ‘ The relationship between the entrepreneurial personality and the big five personality traits,’ Personality and Individual Differences , 63, 58– 63. Google Scholar CrossRef Search ADS   Lewin K. ( 1935), A Dynamic Theory of Personality . McGraw-Hill: New York, NY. Lewin K. ( 1951), Field Theory in Social Science . McGraw-Hill: New York, NY. Lichand G., Soares R. R. ( 2014), ‘ Access to justice and entrepreneurship: evidence from Brazil’s special civil tribunals,’ Journal of Law and Economics , 57( 2), 459– 499.0022-2186/2014/5702-0014. Google Scholar CrossRef Search ADS   McClelland D. ( 1961), The Achieving Society . Van Nostrand: Princeton, NJ. Google Scholar CrossRef Search ADS   Melton J., Elkins Z., Ginsburg T., Leetaru K. ( 2013). ‘ On the interpretability of law: lessons from the decoding of national constitutions,’ British Journal of Political Science , 43( 02), 399– 423. Google Scholar CrossRef Search ADS   Niszczota P. ( 2014), ‘ Cross-country differences in personality and the foreign bias in international equity portfolios,’ European Journal of Finance , 20( 10), 934– 956. Google Scholar CrossRef Search ADS   Nyström K. ( 2008), ‘ The institutions of economic freedom and entrepreneurship: evidence from panel data,’ Public Choice , 136( 3–4), 269– 282. Google Scholar CrossRef Search ADS   Obschonka M., Stuetzer M., Audretsch D. B., Rentfrow P. J., Potter J., Gosling S. D. ( 2016), ‘ Macropsychological factors predict regional economic resilience during a major economic crisis,’ Social Psychological and Personality Science , 7( 2), 95– 104. Google Scholar CrossRef Search ADS   Obschonka M., Stuetzer M., Gosling S. D., Rentfrow P. J., Lamb M. E., Potter J., Audretsch D. B., Waldorp L. J. ( 2015), ‘ Entrepreneurial regions: do macro-psychological cultural characteristics of regions help solve the “knowledge paradox” of economics?,’ PLoS One , 10( 6), e0129332.10.1371/journal.pone.0129332, June 22. Google Scholar CrossRef Search ADS   Parker S. C. ( 2007), ‘ Law and the economics of entrepreneurship,’ Comparative Labor Law and Policy Journal , 28, 695– 715. Parker S. C. ( 2009), The Economics of Entrepreneurship . Cambridge University Press: Cambridge, MA. Google Scholar CrossRef Search ADS   Persson T., Tabellini G. ( 2003), The Economic Effects of Constitutions . The MIT Press: Cambridge, MA. Rentfrow P. J., Gosling S. D., Jokela M., Stillwell D. J., Kosinski M., Potter J. ( 2013), ‘ Divided we stand: three psychological regions of the United States and their political, economic, social, and health correlates,’ Journal of Personality and Social Psychology , 105( 6), 996– 1012. Google Scholar CrossRef Search ADS   Rentfrow P. J., Gosling S. D., Potter J. ( 2008), ‘ A theory of the emergence, persistence, and expression of geographic variation in psychological characteristics,’ Perspectives on Psychological Science , 3( 5), 339– 369. Google Scholar CrossRef Search ADS   Rentfrow P. J., Jokela M., Lamb M. E. ( 2015), ‘ Regional personality differences in Great Britain,’ R. D. Latzman, PLoS One , 10( 3), e0122245. Google Scholar CrossRef Search ADS   Roberts B. W., Kuncel N. R., Shiner R., Caspi A., Goldberg L. R. ( 2007), ‘ The power of personality: the comparative validity of personality traits, socioeconomic status, and cognitive ability for predicting important life outcomes,’ Perspectives on Psychological Science , 2( 4), 313– 345. Google Scholar CrossRef Search ADS   Santarelli E., Pesciarelli E. ( 1990), ‘ The emergence of a vision: the development of schumpeter’s theory of entrepreneurship,’ History of Political Economy , 22( 4), 677– 696. Google Scholar CrossRef Search ADS   Santarelli E., Sterlacchini A. ( 1994), ‘ New firm formation in Italian industry: 1975-89,’ Small Business Economics , 6( 2), 95– 106. Google Scholar CrossRef Search ADS   Santarelli E., Tran H. T. ( 2013), ‘ The interplay of human capital and social capital in shaping entrepreneurial performance: the case of Vietnam,’ Small Business Economics , 40( 2), 435– 458. Google Scholar CrossRef Search ADS   Santarelli E., Vivarelli M. ( 2007), ‘ Entrepreneurship and the process of firm’s entry, survival and growth,’ Industrial and Corporate Change , 16( 3), 455– 488. Google Scholar CrossRef Search ADS   Saxenian A. ( 1994), Regional Advantage . Harvard University Press: Cambridge, MA. Schumpeter J. A. ( 1911), Theorie der Wirtschaftlichen Entwicklung . Duncker & Humblot: Leipzig. Schumpeter J. A. ( 1934), Theory of Economic Development . Harvard University Press: Cambridge, MA. Steel G. D., Rinne T., Fairweather J. ( 2012), ‘ Personality, nations, and innovations: relationships between the big five personality traits and national innovation scores,’ Cross-Cultural Research , 46( 1), 3– 30. Google Scholar CrossRef Search ADS   Storey D. J., Jones A. M. ( 1987), ‘ New firm formation: a labor market approach to industrial entry,’ Scottish Journal of Political Economy , 34( 1), 37– 51. Google Scholar CrossRef Search ADS   Stuetzer M., Obschonka M., Audretsch D. B., Wyrwich M., Rentfrow P. J., Coombes M., Shaw-Taylor L., Satchell M. ( 2016), ‘Industry structure, entrepreneurship, and culture: an empirical analysis using historical coalfields,’ European Economic Review , 86, 52– 72. Google Scholar CrossRef Search ADS   Tag J., Astebro T., Thompson P. ( 2016), ‘ Hierarchies and entrepreneurship,’ European Economic Review , 89, 129– 147. Google Scholar CrossRef Search ADS   Vecchione M., Alessandri G. ( 2013), ‘ Alpha and beta traits and egoistic and moralistic self‐enhancement: a point of convergence between two research traditions,’ Journal of Personality , 81( 1), 39– 48. Google Scholar CrossRef Search ADS   Vedel A. ( 2014), ‘ The big five and tertiary academic performance: a systematic review and meta-analysis,’ Personality and Individual Differences , 71, 66– 76. Google Scholar CrossRef Search ADS   Wennberg K., Pathak S., Autio E. ( 2013), ‘ How culture moulds the effects of self-efficacy and fear of failure on entrepreneurship,’ Entrepreneurship Theory and Practice , 25( 9–10), 756– 780. Wennekers S., van Wennekers A., Thurik R., Reynolds P. ( 2005), ‘ Nascent entrepreneurship and the level of economic development,’ Small Business Economics , 24( 3), 311– 321. Google Scholar CrossRef Search ADS   Wiggins J. S. ( 1991), ‘Agency and communion as conceptual coordinates for the understanding and measurement of interpersonal behavior,’ in Grove W. M., Cicchetti D. (eds), Thinking Clearly About Psychology. Vol. 2: Personality and Psychopathology . University of Minnesota Press: Minneapolis, pp. 89– 113. Wiggins J. S. ( 2003), Paradigms of Personality Assessment . Guilford: New York, NY. Zhao H., Seibert S. E. ( 2006), ‘ The big five personality dimensions and entrepreneurial status: a meta-analytical review,’ Journal of Applied Psychology , 91( 2), 259– 271. Google Scholar CrossRef Search ADS   Zhao H., Seibert S. E., Lumpkin G. T. ( 2010), ‘ The relationship of personality to entrepreneurial intentions and performance: a meta-analytic review,’ Journal of Management , 36, 381– 404. Google Scholar CrossRef Search ADS   © The Author 2017. Published by Oxford University Press on behalf of Associazione ICC. 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

Industrial and Corporate ChangeOxford University Press

Published: Dec 5, 2017

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