Elsevier

Review of Economic Dynamics

Volume 31, January 2019, Pages 305-325
Review of Economic Dynamics

Heterogeneity, selection and labor market disparities

https://doi.org/10.1016/j.red.2018.08.002Get rights and content

Abstract

We propose a model in which differences in socioeconomic and labor market outcomes between ex-ante identical countries can be generated as multiple equilibria sustained by different beliefs on the value of effort for finding jobs. To do so, we study the incentive to improve ability in a model where heterogeneous firms and workers interact in a labor market characterized by matching frictions and costly screening. When effort in improving ability raises both the mean and the variance of the resulting ability distribution, a complementarity between workers' choices and firms' hiring strategies can give rise to multiple equilibria. In the high-effort equilibrium, heterogeneity in ability is larger and induces firms to screen more intensively workers, thereby confirming the belief that effort is important for finding good jobs. In the low-effort equilibrium, ability is less dispersed and firms screen less intensively, which confirms the belief that effort is not so important. The model has novel implications for wage inequality, the distribution of firm characteristics, productivity, sorting patterns between firms and workers, and unemployment rates that can help explain observed differences across countries.

Introduction

Countries at similar stages of development differ markedly in a number of socioeconomic indicators. For instance, wage inequality, labor productivity, school attainment and employment rates are all higher in the United States than in Southern Europe. The population of active firms differs too, with a relatively larger number of small and less productive firms in the latter group of countries. While understanding these differences is important both from a positive and a normative standpoint, their origin remains largely an open question. One strand of literature attributes them to distortions, but typically does not explain how they arose in the first place.1 Another strand of literature emphasizes the role of cultural values.2 Yet, the mechanisms through which values and beliefs translate into different economic outcomes, even in places that started with similar conditions, are still poorly understood.

The objective of this paper is to show that significant differences in socioeconomic and labor market outcomes can emerge as alternative equilibria sustained by different, and yet rational, beliefs on the role played by ability and effort in determining individual economic success. We will argue that the mechanism we identify has implications for wage inequality, the distribution of firm characteristics, sorting patterns between firms and workers, and unemployment rates that can help to explain the cross-country variation observed in the data.

To this end, we study the incentives to invest in ability in a model where heterogeneous firms and workers interact in a labor market with matching frictions. Ability is unobservable, but firms can use a screening technology to select the best workers. As in Helpman et al. (2010), the combination of these features yields realistic distributions of firms and wages. We then allow workers to invest costly effort to improve their ability under the realistic assumption that effort and exogenous talent are complementary.3 The latter feature implies that exerting effort increases average ability, but also its dispersion in the population, and introduces a novel complementarity between firms' and workers' strategies.4 On the one hand, the returns to screening are higher when ability is more dispersed, i.e., when effort is high. On the other hand, investing effort pays out more when firms screen workers more intensively.

The main result of the paper is to show that this complementarity can give rise to two equilibria. In the high-effort equilibrium, heterogeneity in ability is higher and this induces firms to be more selective when hiring workers. In turn, this makes ability, and hence effort, more important for finding good jobs, thereby confirming the initial belief. In the low-effort equilibrium, instead, since ability is less dispersed, firms screen less intensively and hence the probability of finding jobs depends more on luck rather than merit, which confirms the initial belief on the low value of effort. Relative to the alternative scenario, in the high-effort equilibrium ability is higher and more dispersed, firms are more productive, and a stronger sorting pattern between firms and workers generates more inequality among firms. Wage inequality is also typically higher. Our aim is to show that this mechanism can replicate several salient differences observed between countries such as the United States, Italy and Spain.5

First, regarding perceptions, the existing evidence suggests that Americans believe in individual merit, work ethic and competition more than Southern Europeans. For instance, according to the 1981–2000 World Values Survey, 26.4% of Americans strongly agree with the statement that “hard work brings success”, against a share of 14.6% in Italy and 12.2% in Spain. Those who instead strongly believe that success “is a matter of luck and connections” represent 2.3%, 8.9% and 7.8% of respondents in the three countries, respectively. Similarly, 43.3% of Americans think that “hard work is an important quality that a child should learn”, against 26.8% in Italy. More broadly, 29.6% of Americans strongly believe that “competition is good”, as opposed to 19.2% of Italians and 15.6% of Spaniards.

Second, these beliefs come together with significant differences in investment in education. Available data on the quality and quantity of schooling indicate that Americans attach a higher value to education than people from Southern Europe. For instance, in 2010 the working-age population with tertiary schooling was 41% in the United States against 15% in Italy and 32% in Spain (OECD, 2013).6 Investment in education, both private and total, is also higher in the United States. For instance, total expenditure on tertiary education as a percentage of GDP is 2.8%, 1% and 1.3% in the three countries respectively. Regarding outcomes, U.S. students outperform those from Italy and Spain in all major international comparisons, but also exhibit more dispersion in the results. For example, the standard deviation of IALS test scores is about 22% higher in the United Sates than in Italy (Cebreros, 2014).7 Finally, the United States also score higher than Souther European countries in reported measures of discipline at school, which may be a proxy for effort (OECD, 2010a). However, effort in acquiring human capital is notoriously difficult to observe. For instance, Hamermesh and Donald (2008) show that college GPAs have small and mostly non-significant effects on earnings. The high value attached to education in the United States may also be reflected in the fierce competition for admission and the high tuition fees of top schools. Yet, the quality of the long tail of non-top institutions is hard to assess for employers without investing resources.

Third, the differential value attached to education and effort is also reflected in measures of wage inequality and other labor market outcomes. In particular, the college premium relative to the earnings of workers with secondary education is higher than 1.7 in the United States against 1.5 in Italy and 1.4 in Spain (OECD, 2013). Broader measures of wage inequality display similar patterns. For instance, the variance of the logarithm of hourly wages in 2006 is 0.38 in the United States against 0.17 in the other two countries.8 Even after controlling for workers' characteristics, the variance of the logarithms of residual wages is 0.26 in the United States and 0.12 in Italy.9 Unemployment is also lower in the United States, especially for skilled workers. For example, the unemployment rate of U.S. college graduate is about half of that of the total workforce, while in Italy it is about 70% of the mean (OECD, 2013). As a result, the different composition of the workforce alone contributes to generate a significantly lower unemployment rate in the former country.

Fourth, there are also large cross-country differences in firm-level outcomes. Available data suggest U.S. firms to be on average bigger and more productive, and their size distribution to be more dispersed than their European counterparts. For example, the standard deviation of log sales among manufacturing firms is 0.66 in the United States, 0.53 in Italy and 0.49 in Spain.10 Interestingly, there is also evidence that American markets are more selective: for example, the survival rate for new firms is about 10% lower in the United States than in Italy (Bartelsman et al., 2009). Regarding the covariance between size and productivity, Bartelsman et al. (2013) find that, within the typical U.S. manufacturing industry, labor productivity is almost 50% higher than it would be if employment was allocated randomly and that this measure of allocative efficiency is much lower on average in European countries. Finally, U.S. labor productivity, measured as GDP per hour worked in 2006, is 21% higher than in Italy and 38% higher than in Spain (OECD Data).

Fifth, existing data suggest that American firms value selecting talent more. From their survey on managerial practices around the world, Bloom and Van Reenen (2010) build a synthetic measure of how strongly firms value selection, based on the answers to questions on the importance of attracting and keeping talented people to the company. In their sample of 17 countries, U.S. firms have the highest average score, while Italian firms have the lowest one.11 Moreover, consistently with our hypothesis on differences in hiring strategies, only 13% of U.S. workers claim to have found their job through personal contacts against 25.5% in Italy and 45% in Spain (Pellizzari, 2010).

To our knowledge, our theory is the first to be able to match all these observations without referring to exogenous differences in preferences and/or institutions. Despite this being a remarkable result, it is important to stress that we do not believe the multiplicity of equilibria identified in this paper to be the only or even the most important source of these socioeconomic differences. Rather, our theory illustrates a simple and yet powerful mechanism through which large differences in economic outcomes can arise even when countries have access to the same technologies and share similar market and political institutions. The success at replicating some of the salient differences between the two sides of the Atlantic makes us more confident that the model is capturing real-world phenomena. In particular, we present numerical exercises suggesting that multiple equilibria can account for a significant part of the observed differences between Italy and the United States. Moreover, given that labor markets are often segmented regionally, we believe that our model can be useful for understanding disparities in firm- and labor-market outcomes between regions of the same country, such as the North and South of Italy, which share the same broad institutions and policies.12

Our paper is related to several lines of research. First, it contributes to a set of papers that study the role of social beliefs in explaining the main differences in economic performance and inequality observed between the United States, Europe and other countries. Several important contributions show how alternative sets of beliefs can sustain equilibria with high and low levels of inequality. In Benabou (2000), Alesina and Angeletos (2005), Hassler et al. (2005) and Benabou and Tirole (2006), this happens through the endogenous determination of the political support for redistributive policies; in Piketty (1998) through a status motive. In other papers multiple equilibria arise through endogenous preference formation (e.g., Francois and Zabojnik, 2005, Doepke and Zilibotti, 2014). Differently from these works, we focus on a complementarity between workers' effort decisions and the hiring strategies of heterogeneous firms. This approach seems well-suited for our aim of studying especially differences in the distribution of wages, workers and firms.

The paper is also related to the large literature on the role of human capital, broadly defined, for economic development. Several contributions have shown how multiple equilibria and poverty traps can arise in the presence of increasing returns due to human capital externalities (e.g., Azariadis and Drazen, 1990), non-convexities coupled with credit frictions (e.g., Galor and Zeira, 1993), or a complementarity between talent and technological change (e.g., Hassler and Rodriguez Mora, 2000). Differently from these works, technological increasing returns to human capital or credit frictions are not needed in our approach to generate multiple equilibria. Moreover, none of the above mentioned papers examines the interaction between workers and firm heterogeneity.

Closer to our spirit, Acemoglu (1996) and Burdett and Smith (2002) show that human capital externalities may arise naturally when labor markets are characterized by search frictions. Similarly to our model, agents choose human capital depending on their job prospects and firms choose jobs depending on the average human capital of the workforce.13 Differently from our framework, however, these papers abstract from firm heterogeneity and selection through screening. The importance of the allocation of talent is stressed by many papers, including Acemoglu (1995), Hsieh et al. (2013) and Bonfiglioli and Gancia (2014a).14 None of them, however, studies its interplay with the hiring strategies of heterogeneous firms which is at the core of our theory.

Finally, the paper builds on the literature on wage inequality in models with imperfect labor markets and firm heterogeneity. Acemoglu (1997) shows how search frictions à la Mortensen and Pissarides (1994) can generate and shape wage inequality.15 Lagos (2006) and Marimon and Zilibotti (1999) study how different shocks and policies may affect aggregate outcomes and wage inequality in labor markets with matching frictions. Helpman et al., 2008, Helpman et al., 2010 combine search frictions, firm heterogeneity (as in Melitz, 2003) and worker heterogeneity to study wage dispersion, wage-size premia and unemployment in open and closed economy. Our model builds on these frameworks by adding an endogenous ability distribution and by exploring how the novel equilibrium multiplicity that arises can help explain some of the observed cross-county differences in the distribution of wages, firm characteristics and unemployment rates.

The rest of the paper is organized as follows. In Section 2 we lay down the model and derive the conditions for equilibrium multiplicity. In Section 3 we compare labor market outcomes, firms and welfare across equilibria. In Section 4 we explore the quantitative implications of the model by comparing numerical simulations to data for the United States and Italy. Section 5 concludes.

Section snippets

The model

We build a model where heterogeneous firms and workers meet in a labor market characterized by matching frictions along the lines of Helpman et al. (2010). For ease of comparison, we borrow their notation whenever possible. Firms are matched randomly with workers of unknown ability although they can use a screening technology to select them. The profitability of screening is proportional to the heterogeneity among workers. Moreover, ability is relatively more beneficial for more productive

Comparing equilibria

In this section, we compare the predictions of the model for a number of variables of interest in the two equilibria. In what follows, we use again the subindexes 1 and 0 to denote the equilibrium with high and low effort, respectively, and we state explicitly the variables that are functions of k, when needed to avoid confusion. Since some comparisons are ambiguous, in the next Section we complement the analysis with numerical examples under plausible parametrizations.

Comparing labor market outcomes: numerical examples

In this section, we complement the qualitative comparison between equilibria presented in Section 3 with some numerical examples. The goal is twofold. First, given that the model predictions for some outcomes are potentially ambiguous, it is useful to explore them using plausible parameter values. Second, we would like to have a sense of how much of the observed cross-country differences in economic outcomes can be accounted for by our theory. Since we have already discussed the effects of

Conclusions

We have proposed a model that explains disparities in several economic and labor market outcomes across similar countries based on multiple equilibria sustained by different beliefs on the value of effort and ability. In particular, when effort raises the dispersion of workers' ability and firms have access to a costly screening technology, two equilibria arise: in the “American” equilibrium, workers expect firms to screen more intensively and hence invest effort to improve their job prospects.

Acknowledgements

We thank Matthias Doepke (the Editor), two anonymous Referees, Teodora Borota, Jordi Galí, Oleg Itskhoki, Kiminori Matsuyama, Ferdinando Monte, Giacomo Ponzetto and seminar participants at the 2014 Annual Meeting of the SED, the Workshop on Economic Integration and Labor Markets at CREST, the 2014 CEPR ESSIM, 2014 Annual Conference of the Royal Economic Society, IAE-CSIC, UPF, the Barcelona Summer Forum and the EEA Annual Meeting 2015. We also thank Javier Quintana for excellent research

References (65)

  • Diego Restuccia et al.

    Policy distortions and aggregate productivity with heterogeneous establishments

    Review of Economic Dynamics

    (2008)
  • Daron Acemoglu

    A microfoundation for social increasing returns in human capital accumulation

    The Quarterly Journal of Economics

    (1996)
  • Daron Acemoglu

    Matching, heterogeneity, and the evolution of income distribution

    Journal of Economic Growth

    (1997)
  • Alberto Alesina et al.

    Family values and the regulation of labor

    Journal of the European Economic Association

    (2015)
  • Alberto Alesina et al.

    Fairness and redistribution

    The American Economic Review

    (2005)
  • David H. Autor et al.

    Trends in U.S. Wage inequality: revising the revisionists

    Review of Economics and Statistics

    (2008)
  • Costas Azariadis et al.

    Threshold externalities in economic development

    The Quarterly Journal of Economics

    (1990)
  • Oriana Bandiera et al.

    Social incentives in the workplace

    The Review of Economic Studies

    (2010)
  • Eric Bartelsman et al.

    Employment protection, technology choice, and worker allocation

    International Economic Review

    (2016)
  • Eric Bartelsman et al.

    Measuring and analyzing cross country differences in firm dynamics

  • Eric Bartelsman et al.

    Cross country differences in productivity: the role of allocative efficiency

    The American Economic Review

    (2013)
  • Roland Benabou

    Unequal societies: income distribution and the social contract

    The American Economic Review

    (2000)
  • Roland Benabou et al.

    Belief in a just world and redistributive politics

    The Quarterly Journal of Economics

    (2006)
  • Chris Bidner et al.

    Cultivating trust: norms, institutions and the implications of scale

    Economic Journal

    (2012)
  • Nicholas Bloom et al.

    Why do management practices differ across firms and countries?

    The Journal of Economic Perspectives

    (2010)
  • Lawrence E. Blume

    The dynamics of statistical discrimination

    Economic Journal

    (2006)
  • Matilde Bombardini et al.

    Skill dispersion and trade flows

    The American Economic Review

    (2012)
  • Alessandra Bonfiglioli et al.

    Betting on exports: trade and endogenous heterogeneity

    Economic Journal

    (2018)
  • Alessandra Bonfiglioli et al.

    Trade, finance and endogenous firm heterogeneity

    Journal of the European Economic Association

    (2018)
  • Alessandra Bonfiglioli et al.

    Heterogeneity, Selection and Labor Market Disparities

    (2014)
  • Francesco Caselli et al.

    Dynastic management

    Economic Inquiry

    (2013)
  • Alfonso Cebreros

    Labor Heterogeneity and the Pattern of Trade

    (2018)
  • Cited by (6)

    • Concentration in international markets: Evidence from US imports

      2021, Journal of Monetary Economics
      Citation Excerpt :

      However, our results also show that firms are growing more and more unequal, a finding that resonates with recent evidence using very different data.23 Some possible explanations for this widespread trend may include changes in innovation strategies (e.g., Benhabib et al. (2017); Bonfiglioli et al. (2018, 2019); Dhingra (2013); König et al. (2016)), stronger sorting between firms, suppliers and workers (e.g., Bonfiglioli and Gancia (2019); Song et al. (2019)) or the uneven adoption of automation technologies (e.g., Acemoglu et al. (2020); Bonfiglioli et al. (2020a) and Hubmer and Restrepo (2021)). In turn, unequal growth at the firm level can potentially have adverse effects on labor market outcomes and the distribution of income.

    View full text