TY - JOUR
AU - Osorio Buitron,, Carolina
AB - Abstract We examine the factors explaining the increase in gross and net income inequality in advanced economies since the 1980s. Our results support the view that globalization, technological progress, financial deregulation and lower top marginal tax rates are associated with higher inequality, and we find that the relation between the decline in union density and the rise in top decile income shares—a phenomenon which labour economists have long been discussing—is widespread across advanced economies. The influence of union density on top income shares appears to be causal, as evidenced by our instrumental variable estimates and the inclusion of potentially omitted variables. 1. Introduction The rise of inequality in advanced economies, and the growing concentration of incomes at the top of the distribution, has become a concern for economists and policymakers. While some degree of inequality can increase efficiency by strengthening incentives to work and invest, a strand of the economic literature argues that higher inequality tends to be associated with lower and less sustainable growth (Perotti, 1992, 1993, 1996; Alesina and Rodrik, 1994; Persson and Tabellini, 1994; Benabou, 1996; Aghion et al., 1999; Temple, 1999, 2005; Burtless, 2003; Berg et al., 2012). Moreover, a rising concentration of income at the top of the distribution can be welfare reducing, if it enables top earners to manipulate the economic and political system in their favour (Stiglitz, 2012). Traditional explanations for the rise in (wage) inequality have been technological progress and globalization (e.g., Gabaix and Landier, 2008) and, more recently, the role of institutional changes, such as financial deregulation and lower top marginal personal income tax rates (Philippon and Reshef, 2012, 2013; Alvaredo et al., 2013; Bivens and Mishel, 2013; Piketty and Stantcheva, 2014), as well as labour market institutions (e.g. DiNardo et al., 1996; Lee, 1999; Teulings, 2003; Card et al., 2004).1 In this paper, we provide evidence that all these factors have played an important role in advanced economies, with the erosion of labour market institutions—in particular, union density—contributing to the rise of both gross and net income inequality, notably at the top of the income distribution. The rise in top income shares is possibly supported by the weakening of labour market institutions, as the latter reduces the bargaining power of average wage earners relative to capital owners and top earners (Gomez and Tzioumis, 2006, 2011; Duenhaupt, 2012; Piketty and Stantcheva, 2014). Moreover, weaker labour market institutions can limit workers’ influence on redistributive policies, thus contributing to the rise of net income inequality. Our analysis uses data for 20 advanced economies over the period 1980–2011. We consider two measures of income inequality: the income share of the top 10% earners, which is less affected by top coding issues; and the Gini coefficient, which is more sensitive to changes at the middle and bottom of the income distribution. We also take into account Gini coefficients of both gross and net income to assess the effects on inequality through redistribution. The methodology relies on cross-country panel regressions. Our main results can be summarized as follows. First, we find that—in addition to the role played by technology, globalization, financial liberalization, top marginal personal income tax rates, and common global trends in increasing inequality—the decline in union density (the proportion of union members in the workforce) is strongly associated with the rise of top earners’ income shares in advanced economies. The influence of union density on top income shares appears to be largely causal, as evidenced by our instrumental variable estimates. The instruments for union density capture the fact that, although unionization tends to decline in periods of high unemployment, the effect is weaker in countries where unemployment benefits are managed by unions (i.e. the Ghent system), or where collective bargaining is more centralized. In addition, we conduct a variety of robustness checks to ensure that the result is not driven by omitted variables (such as changes in elected governments and social norms, sectoral employment shifts, and rising education levels), that could both reduce unionization and increase inequality. The magnitude of the effect of union density is large: on average, across our sample countries, the decline in union density explains about 40% of the 5 percentage point average increase in the top 10% income shares. Second, we find some empirical evidence that unions also affect income redistribution, likely through their influence on public policy: for a given level of gross income inequality, a decline in union density is associated with an increase in the Gini of net income. As regards to other labour market institutions, there is some evidence that reductions in the minimum wage relative to the median wage are related to significant increases in overall inequality measured by the Gini. Our novel panel dataset estimates confirm what many labour economists have been long saying is at work; namely, that the weakening of labour market institutions is the proximate cause of rising inequality, especially at the top which is often neglected in published work. The negative relationship between unionization and top earners’ income share across countries could underscore two potential mechanisms. First, de-unionization reduces workers’ bargaining power, leading to an increase in the income share of corporate managers and shareholders. Second, weaker unions reduce workers’ influence on corporate decisions, including those related to top executive compensation. The remainder of the paper is organized as follows. In Section 2, we present a brief review of the literature. In Section 3, we present some stylized facts about inequality and labour market institutions. In Section 4, we discuss the results from the empirical analysis. Finally, Section 5 concludes. 2. Conceptual framework and literature review Although various measures indicate that inequality in advanced economies has risen considerably since the 1980s, the literature has recently focused on the rise of top income shares based on tax returns data (Dew-Becker and Gordon, 2005; Piketty and Saez, 2006; Piketty and Stantcheva, 2014). The rise of inequality in high-income countries is mostly explained by the upper half of the distribution, reflecting rising income differentials between the 9th and 5th deciles, and a stable ratio between the 5th and 1st deciles of earners. These developments imply that the Gini coefficient may not be sufficient to assess the evolution of inequality in advanced economies, as the statistic is more affected by the under-reporting of top incomes than the tax returns data used to construct top income shares (Alvaredo, 2011; Burkhauser et al., 2012).2 Notwithstanding this limitation, the Gini effectively captures changes below the 9th decile of the distribution, as it gauges the average difference in income between any two individuals randomly chosen from the income distribution. Further, since the Gini is available for both gross and net incomes, it provides an indirect measure of the effectiveness of redistributive policies. Therefore, we consider Gini coefficients and top income shares in our analysis. Explanations for the rise of inequality in the developed world either focus on market-driven forces or institutional changes. According to the market forces hypothesis, the increase in inequality reflects the rising demand for skills at the top, owing to skill-biased technological change (SBTC) and globalization (e.g., Gabaix and Landier, 2008). Since there is little policymakers can do to reverse these global trends, and since inequality has risen at different speeds across advanced economies—which are similarly affected by technological change and globalization, the role of institutional changes matters. Substantial changes in labour market institutions have taken place over the past 30 years. This is the focus of our paper, which emphasizes the role played by changes in union density—the labour market institution that we find to be more robustly associated with inequality. Figure 1 provides a schematic illustration of the channels through which union density can affect inequality. Union density impacts the distribution of earnings at the bottom and middle of the distribution. This takes place through its effects on the dispersion of wages, unemployment, and redistribution, as well as through mechanisms that affect workers’ bargaining power and representativeness—which impact the income share of top earners vis-à-vis the rest of the population. These channels are discussed below in detail. Fig. 1 Open in new tabDownload slide Conceptual framework: channels through which labour market institutions affect income distribution. Sources: Authors. Fig. 1 Open in new tabDownload slide Conceptual framework: channels through which labour market institutions affect income distribution. Sources: Authors. 2.1 Wage distribution The literature provides much evidence that labour market institutions reduce inequality of the wage distribution (see recent surveys in Betcherman, 2012; Kierzenkowski and Koske, 2012). Unions are found to have an equalizing impact on the distribution of labour compensation (Card, 1996, 2001; DiNardo et al., 1996; Card et al., 2004). In addition, the literature has established that a higher minimum wage reduces wage inequality (DiNardo et al., 1996; Lee, 1999; Teulings, 2003).3 2.2 Unemployment rate While some labour market institutions may, indeed, lead to lower wage inequality, they may also increase unemployment, leading to higher gross income inequality. The empirical evidence for the potential trade-off between wage inequality and unemployment is inconclusive. There is robust empirical support for the hypothesis that labour market institutions increase wage equality, while the evidence of their impact on unemployment is less robust. Studies have found that changes in union density have modest adverse effects on unemployment (Freeman, 2000; Baker et al., 2004; OECD, 2006; Howell et al., 2007; Betcherman, 2012). Unlike most of the literature on inequality and labour market institutions, we focus on income rather than wage inequality. Therefore, our analysis implicitly takes account of the effects of labour market institutions on unemployment. 2.3 Redistribution Some labour market institutions can also play a role in the process of redistribution of market income and contribute to reduce net income inequality. In particular, strong unions can play an important role in the determination of redistributive policies, as evidenced by their contribution to the introduction of fundamental social and labour rights (Betcherman, 2012). Further, strong unions can induce policymakers to engage in more redistribution, by mobilizing workers to vote for parties that promise to redistribute income or by leading all political parties to engage in more redistribution (Korpi, 2006).4 2.4 Bargaining power and influence of average wage earners The weakening of unions could increase top income shares by reducing the bargaining power of average wage earners. Theoretical models suggest that a decline in the bargaining power of workers relative to capital owners reduces the labour income share (Pissarides, 1990; Layard et al., 1991). Since capital incomes tend to be highly concentrated, a higher capital income share is likely associated with increased top income shares. The Organisation for Economic Co-operation and Development (OECD) estimates that, in the mid-2000s, the average Gini for capital income was 0.84 in advanced economies, compared with 0.36 and 0.58 for wages and self-employment income, respectively.5 Further, weaker unions can reduce workers’ influence on corporate decisions that benefit top earners.6 By contrast, where unions are strong, firms are more likely to engage in consultations with worker representatives, allowing them to have some influence over the size and structure of top executive compensation (Sjöberg, 2009; McCall and Percheski, 2010). Indeed, strong unions can bring pressure on the firm when executive salaries are perceived as excessive, by voicing fairness concerns and threatening industrial disruption. Boards are also likely to be more cautious in setting executive compensation when unions are strong, since high levels of executive compensation could be perceived as a sign of strong firm financial health and lead to increased wage demands (Jensen and Murphy, 1990, Gomez and Tzioumis, 2006, 2011). Finally, Acemoglu and Robinson (2013) highlight a political economy channel, by which the weakening of unions could change the political equilibrium in favour of already-dominant groups, enabling them to influence the economic and political system in their favour. However, an alternative interpretation of the negative relationship between union density and top income shares—one that is not based on changes in bargaining power of unions—could be that weaker unions increase the productivity of top executives by giving them more managerial freedom, which justifies their higher remuneration. A few empirical studies have looked at the relationship between union density and top income shares. Gomez and Tzioumis (2006, 2011) find that union presence is significantly associated with lower levels of CEO compensation in a panel of publicly listed US firms. DiNardo et al. (1997) find a negative correlation between executive compensation and unionization in cross-sectional data, but less of a relationship between changes in unionization and the growth of executive compensation over time. Finally, focusing on the effects of the political system on inequality, Scheve and Stasavage (2009) and Volscho and Kelly (2012) find a negative effect of union density on top income shares for a panel of countries and the United States, respectively. Our study builds on this literature by providing novel panel-data estimates of the relationship between unionization and top income shares, and establishing causality through the use of an instrumentation strategy and the inclusion of a large number of possibly omitted variables. 3. Descriptive facts The data used throughout our analysis is comprised of inequality measures, labour market institution variables, and indicators that measure other determinants of inequality. We use measures of income inequality, as opposed to wage inequality. While the use of the latter is commonplace in the literature, we focus on the distribution of incomes, which allows us to account implicitly for the potential effects of market-driven forces and institutional changes on unemployment. Further, we consider various measures of inequality, as each statistic has limitations and, jointly, the various measures allow for a comprehensive analysis of the evolution of inequality. The most commonly used measure of inequality, the Gini coefficient, relies on household income surveys data and tends to be sensitive to transfers at the centre of the distribution. We use a recently compiled cross-country dataset (Solt, 2009, The Standardized World Income Inequality Database, henceforth SWIID), where the definition of income is pre-tax and pre-transfer income for the computation of gross income Gini coefficients, and post-tax, post-transfer income for the computation of net income Gini coefficients. Therefore, we are able to examine the effects of labour market institutions on both the distribution of gross income and its redistribution. A drawback of the Gini coefficient, however, is that it fails to capture the effects of rising top income shares (see Alvaredo, 2011; Burkhauser et al., 2012). Thus, we also consider top income share statistics (Alvaredo et al., 2015, The World Top Incomes Database), which rely on tax records data, but which are only available for gross income. See the online Appendix, for further details on the definition and source of all the variables used in this paper. Measures of inequality based on Gini coefficients of gross and net income have increased substantially since 1980 in most of the developed world (Fig. 2). At the same time, the (gross) income shares of the top 10% and top 1% earners have grown continuously, indicating that the rise of inequality in advanced economies has been driven by both a greater dispersion of incomes between the top 10% and bottom 90% income groups, as well as a more uneven distribution of incomes within the top decile. Fig. 2 Open in new tabDownload slide Evolution of inequality measures in advanced economies, 1980–2011. Notes: Advanced Economies = Australia, Austria, Belgium, Canada, Denmark, Finland, France, Germany, Ireland, Italy, Japan, Netherlands, New Zealand, Norway, Portugal, Spain, Sweden, Switzerland, United Kingdom, and United States. Simple average. For the top 10 income share, Austria, Belgium, Finland, Portugal, and United Kingdom are excluded due to missing data over part of the 1980–2011 period. Sources: World Top Incomes Database; and SWIID (v. 4.0). Fig. 2 Open in new tabDownload slide Evolution of inequality measures in advanced economies, 1980–2011. Notes: Advanced Economies = Australia, Austria, Belgium, Canada, Denmark, Finland, France, Germany, Ireland, Italy, Japan, Netherlands, New Zealand, Norway, Portugal, Spain, Sweden, Switzerland, United Kingdom, and United States. Simple average. For the top 10 income share, Austria, Belgium, Finland, Portugal, and United Kingdom are excluded due to missing data over part of the 1980–2011 period. Sources: World Top Incomes Database; and SWIID (v. 4.0). Since the mid-1990s, however, the behaviour of the Gini and top income share statistics has started to diverge, as the former increased at a significantly slower pace than the latter. The disconnect between these two measures likely reflects the fact that incomes have become so concentrated at the top that Gini measures have not captured it well. To illustrate, we calculate the increase in the Gini of gross income implied by the evolution of top income shares since the mid-1990s. For this counterfactual analysis, we assume a two-class economy: the top 10% and bottom 90% earners. Further, we suppose that inequality within the bottom 90% of the population remained constant since 1995, which is broadly consistent with the available evidence.7 Our results suggest that, given the evolution of the top 10% income share, the Gini should have increased by at least 2.3 percentage points since 1995—1.3 percentage points above the observed change.8 The increase in gross income inequality is, on average, larger than in net inequality, and the cross-country dispersion of changes in net income inequality is smaller than for the gross income measures. This suggests that redistributive policies have been somewhat successful. Indeed, Ostry et al. (2014) find that, among advanced economies, higher inequality is associated with higher redistribution. However, the fact that net income inequality has increased in almost every year since 1980 indicates that the transfer and tax system has not kept pace with the rise in inequality. As with measures of income inequality, changes in the distribution of earnings indicate that inequality has risen, owing largely to a concentration of earnings at the top of the distribution. Gross earnings differentials between the 9th and 5th deciles of the distribution have increased over four times as much as the differential between the 5th and 1st deciles (Fig. 3). Moreover, data from the Luxembourg Income Study on net income shares indicate that income shares of the top 10% earners have increased at the expense of all other income groups. While there is some country heterogeneity, the increase in top income shares since the 1980s appears to be a pervasive phenomenon. Fig. 3 Open in new tabDownload slide Distributional changes in advanced economies, 1980–2011. Notes: (1) Earnings-dispersion measures–- ratio of 9th to 1st, 9th to 5th, and 5th to 1st – where ninth, fifth (or median) and first deciles are upper-earnings decile limits, unless otherwise indicated, of gross earnings of full-time dependent employees. Advanced economies include Australia, Denmark, Finland, Italy, Japan, New Zealand, Sweden, the United Kingdom, and the United States. Simple average. (2) Shares of disposable income by decile using Luxembourg Income Study data. Varying years for countries, including Australia, Austria, Belgium, Canada, Denmark, Finland, France, Germany, Ireland, Italy, Netherlands, Norway, Spain, Sweden, Switzerland, the United Kingdom and the United States. Sources: OECD Wages and Earnings Database; and LIS/New York Times Income Distribution Database (2014). Fig. 3 Open in new tabDownload slide Distributional changes in advanced economies, 1980–2011. Notes: (1) Earnings-dispersion measures–- ratio of 9th to 1st, 9th to 5th, and 5th to 1st – where ninth, fifth (or median) and first deciles are upper-earnings decile limits, unless otherwise indicated, of gross earnings of full-time dependent employees. Advanced economies include Australia, Denmark, Finland, Italy, Japan, New Zealand, Sweden, the United Kingdom, and the United States. Simple average. (2) Shares of disposable income by decile using Luxembourg Income Study data. Varying years for countries, including Australia, Austria, Belgium, Canada, Denmark, Finland, France, Germany, Ireland, Italy, Netherlands, Norway, Spain, Sweden, Switzerland, the United Kingdom and the United States. Sources: OECD Wages and Earnings Database; and LIS/New York Times Income Distribution Database (2014). At the same time, significant changes have taken place in labour market institutions—in particular, in union density, the focus of this paper. Figure 4 shows the evolution of this institution. Union density declined steadily in advanced economies including after the mid-1990s (Fig. 4). While most countries experienced a decline, there is substantial variation in the extent of the decline in unionization across countries. For instance, Anglo-Saxon countries experienced sizable declines, except for Canada which had a much smaller decline. In contrast, in most Nordic countries and Belgium, declines have been much more limited and unionization rates remain very high. Fig. 4 Open in new tabDownload slide Evolution of union density in advanced economies, 1980–2011. Source: OECD. Fig. 4 Open in new tabDownload slide Evolution of union density in advanced economies, 1980–2011. Source: OECD. Regarding the other institutional variables, there has been a clear trend decline in top marginal tax rates and continued reforms aimed at liberalizing financial markets until the mid-1990s. Descriptive statistics for all the variables included in our analysis are presented in Table 1.9 The variables’ variation comes largely from the within-country dimension, but for the inequality measures, union density and the minimum wage, both between and within-country variation, are important. Table 1 Descriptive statistics Variable . . Mean . Std. dev. . Min. . Max. . Observations . ln(top 10 %) Overall 3.45 0.16 2.94 3.84 N 478 Between 0.14 3.22 3.69 n 18 Within 0.08 2.95 3.79 T 27 ln(gini market) Overall 3.69 0.14 3.29 4.00 N 556 Between 0.10 3.47 3.85 n 20 Within 0.10 3.21 4.03 T 28 ln (gini net) Overall 3.33 0.15 2.94 3.63 N 556 Between 0.14 3.12 3.56 n 20 Within 0.08 2.84 3.49 T 28 ln (ICT K) Overall −3.61 1.09 −7.94 −1.60 N 556 Between 0.58 −4.78 −2.51 n 20 Within 0.96 −6.77 −1.28 T 28 ln(lag Y/N) Overall 10.19 0.25 9.39 10.80 N 556 Between 0.19 9.74 10.53 n 20 Within 0.16 9.81 10.53 T 28 ln(lag Y/N) * China X Overall 0.10 0.17 −0.35 0.93 N 556 Between 0.05 0.04 0.31 n 20 Within 0.16 −0.39 0.92 T 28 ln(fin. reform) Overall −0.23 0.32 −2.35 0.00 N 556 Between 0.15 −0.61 0.00 n 20 Within 0.28 −1.97 0.19 T 28 top tax Overall 0.49 0.14 0.12 0.93 N 556 Between 0.09 0.26 0.61 n 20 Within 0.10 0.12 0.85 T 28 union density Overall 0.38 0.21 0.08 0.84 N 556 Between 0.20 0.10 0.79 n 20 Within 0.06 0.23 0.68 T 28 minimum wage Overall 0.27 0.25 0.00 0.66 N 556 Between 0.25 0.00 0.60 n 20 Within 0.03 0.10 0.38 T 28 Variable . . Mean . Std. dev. . Min. . Max. . Observations . ln(top 10 %) Overall 3.45 0.16 2.94 3.84 N 478 Between 0.14 3.22 3.69 n 18 Within 0.08 2.95 3.79 T 27 ln(gini market) Overall 3.69 0.14 3.29 4.00 N 556 Between 0.10 3.47 3.85 n 20 Within 0.10 3.21 4.03 T 28 ln (gini net) Overall 3.33 0.15 2.94 3.63 N 556 Between 0.14 3.12 3.56 n 20 Within 0.08 2.84 3.49 T 28 ln (ICT K) Overall −3.61 1.09 −7.94 −1.60 N 556 Between 0.58 −4.78 −2.51 n 20 Within 0.96 −6.77 −1.28 T 28 ln(lag Y/N) Overall 10.19 0.25 9.39 10.80 N 556 Between 0.19 9.74 10.53 n 20 Within 0.16 9.81 10.53 T 28 ln(lag Y/N) * China X Overall 0.10 0.17 −0.35 0.93 N 556 Between 0.05 0.04 0.31 n 20 Within 0.16 −0.39 0.92 T 28 ln(fin. reform) Overall −0.23 0.32 −2.35 0.00 N 556 Between 0.15 −0.61 0.00 n 20 Within 0.28 −1.97 0.19 T 28 top tax Overall 0.49 0.14 0.12 0.93 N 556 Between 0.09 0.26 0.61 n 20 Within 0.10 0.12 0.85 T 28 union density Overall 0.38 0.21 0.08 0.84 N 556 Between 0.20 0.10 0.79 n 20 Within 0.06 0.23 0.68 T 28 minimum wage Overall 0.27 0.25 0.00 0.66 N 556 Between 0.25 0.00 0.60 n 20 Within 0.03 0.10 0.38 T 28 Source: Authors’ calculations. Open in new tab Table 1 Descriptive statistics Variable . . Mean . Std. dev. . Min. . Max. . Observations . ln(top 10 %) Overall 3.45 0.16 2.94 3.84 N 478 Between 0.14 3.22 3.69 n 18 Within 0.08 2.95 3.79 T 27 ln(gini market) Overall 3.69 0.14 3.29 4.00 N 556 Between 0.10 3.47 3.85 n 20 Within 0.10 3.21 4.03 T 28 ln (gini net) Overall 3.33 0.15 2.94 3.63 N 556 Between 0.14 3.12 3.56 n 20 Within 0.08 2.84 3.49 T 28 ln (ICT K) Overall −3.61 1.09 −7.94 −1.60 N 556 Between 0.58 −4.78 −2.51 n 20 Within 0.96 −6.77 −1.28 T 28 ln(lag Y/N) Overall 10.19 0.25 9.39 10.80 N 556 Between 0.19 9.74 10.53 n 20 Within 0.16 9.81 10.53 T 28 ln(lag Y/N) * China X Overall 0.10 0.17 −0.35 0.93 N 556 Between 0.05 0.04 0.31 n 20 Within 0.16 −0.39 0.92 T 28 ln(fin. reform) Overall −0.23 0.32 −2.35 0.00 N 556 Between 0.15 −0.61 0.00 n 20 Within 0.28 −1.97 0.19 T 28 top tax Overall 0.49 0.14 0.12 0.93 N 556 Between 0.09 0.26 0.61 n 20 Within 0.10 0.12 0.85 T 28 union density Overall 0.38 0.21 0.08 0.84 N 556 Between 0.20 0.10 0.79 n 20 Within 0.06 0.23 0.68 T 28 minimum wage Overall 0.27 0.25 0.00 0.66 N 556 Between 0.25 0.00 0.60 n 20 Within 0.03 0.10 0.38 T 28 Variable . . Mean . Std. dev. . Min. . Max. . Observations . ln(top 10 %) Overall 3.45 0.16 2.94 3.84 N 478 Between 0.14 3.22 3.69 n 18 Within 0.08 2.95 3.79 T 27 ln(gini market) Overall 3.69 0.14 3.29 4.00 N 556 Between 0.10 3.47 3.85 n 20 Within 0.10 3.21 4.03 T 28 ln (gini net) Overall 3.33 0.15 2.94 3.63 N 556 Between 0.14 3.12 3.56 n 20 Within 0.08 2.84 3.49 T 28 ln (ICT K) Overall −3.61 1.09 −7.94 −1.60 N 556 Between 0.58 −4.78 −2.51 n 20 Within 0.96 −6.77 −1.28 T 28 ln(lag Y/N) Overall 10.19 0.25 9.39 10.80 N 556 Between 0.19 9.74 10.53 n 20 Within 0.16 9.81 10.53 T 28 ln(lag Y/N) * China X Overall 0.10 0.17 −0.35 0.93 N 556 Between 0.05 0.04 0.31 n 20 Within 0.16 −0.39 0.92 T 28 ln(fin. reform) Overall −0.23 0.32 −2.35 0.00 N 556 Between 0.15 −0.61 0.00 n 20 Within 0.28 −1.97 0.19 T 28 top tax Overall 0.49 0.14 0.12 0.93 N 556 Between 0.09 0.26 0.61 n 20 Within 0.10 0.12 0.85 T 28 union density Overall 0.38 0.21 0.08 0.84 N 556 Between 0.20 0.10 0.79 n 20 Within 0.06 0.23 0.68 T 28 minimum wage Overall 0.27 0.25 0.00 0.66 N 556 Between 0.25 0.00 0.60 n 20 Within 0.03 0.10 0.38 T 28 Source: Authors’ calculations. Open in new tab To assess the relationship between the observed changes in inequality and its determinants, we first present simple correlations between these variables. Union density appears to have a much stronger correlation with the top 10% income share than that of other key determinants of inequality, such as SBTC, globalization, and top tax rates (Fig. 5). Moreover, there is a strong negative relation between the top 10% income share and union density, both within and across countries (Fig. 6). The Gini of gross income is also negatively related with union density, but the relationship is somewhat weaker and mostly present within countries. Fig. 5 Open in new tabDownload slide Top income shares and main determinants in advanced economies, 1980–2011. Notes: *** denotes significance at the 1% level, ** at the 5% level, and * at the 10% level. Advanced Economies = Australia, Canada, Denmark, Finland, France, Germany, Ireland, Italy, Japan, Netherlands, New Zealand, Norway, Portugal, Spain, Sweden, Switzerland, the United Kingdom, and the United States. Sources: World Top Incomes Database; Jorgenson and Vu (2011); IMF, Direction of Trade Statistics; and OECD. Fig. 5 Open in new tabDownload slide Top income shares and main determinants in advanced economies, 1980–2011. Notes: *** denotes significance at the 1% level, ** at the 5% level, and * at the 10% level. Advanced Economies = Australia, Canada, Denmark, Finland, France, Germany, Ireland, Italy, Japan, Netherlands, New Zealand, Norway, Portugal, Spain, Sweden, Switzerland, the United Kingdom, and the United States. Sources: World Top Incomes Database; Jorgenson and Vu (2011); IMF, Direction of Trade Statistics; and OECD. Fig. 6 Open in new tabDownload slide Gross income inequality measures and union density in advanced economies, 1980–2011. Notes: *** denotes significance at the 1% level, ** at the 5% level, and * at the 10% level. For top 10 income share, Advanced Economies = Australia, Canada, Denmark, Finland, France, Germany, Ireland, Italy, Japan, Netherlands, New Zealand, Norway, Portugal, Spain, Sweden, Switzerland, the United Kingdom, and the United States. For Gini of gross income, Advanced Economies = Australia, Austria, Belgium, Canada, Denmark, Finland, France, Germany, Ireland, Italy, Japan, Netherlands, New Zealand, Norway, Portugal, Spain, Sweden, Switzerland, the United Kingdom, and the United States. Sources: OECD; SWIID (v. 4.0); and World Top Incomes Database. Fig. 6 Open in new tabDownload slide Gross income inequality measures and union density in advanced economies, 1980–2011. Notes: *** denotes significance at the 1% level, ** at the 5% level, and * at the 10% level. For top 10 income share, Advanced Economies = Australia, Canada, Denmark, Finland, France, Germany, Ireland, Italy, Japan, Netherlands, New Zealand, Norway, Portugal, Spain, Sweden, Switzerland, the United Kingdom, and the United States. For Gini of gross income, Advanced Economies = Australia, Austria, Belgium, Canada, Denmark, Finland, France, Germany, Ireland, Italy, Japan, Netherlands, New Zealand, Norway, Portugal, Spain, Sweden, Switzerland, the United Kingdom, and the United States. Sources: OECD; SWIID (v. 4.0); and World Top Incomes Database. A similar exercise suggests a positive association between union density and redistribution: while the correlation between union density and the Gini coefficient of gross income is weak, its correlation with the Gini of net income is clearly negative (Fig. 7). Fig. 7 Open in new tabDownload slide Redistribution effect of unions in advanced economies, 1980–2011. Notes: *** denotes significance at the 1% level, ** at the 5% level, and * at the 10% level. Advanced Economies = Australia, Austria, Belgium, Canada, Denmark, Finland, France, Germany, Ireland, Italy, Japan, Netherlands, New Zealand, Norway, Portugal, Spain, Sweden, Switzerland, the United Kingdom, and the United States. Sources: OECD; and SWIID (v.4.0). Fig. 7 Open in new tabDownload slide Redistribution effect of unions in advanced economies, 1980–2011. Notes: *** denotes significance at the 1% level, ** at the 5% level, and * at the 10% level. Advanced Economies = Australia, Austria, Belgium, Canada, Denmark, Finland, France, Germany, Ireland, Italy, Japan, Netherlands, New Zealand, Norway, Portugal, Spain, Sweden, Switzerland, the United Kingdom, and the United States. Sources: OECD; and SWIID (v.4.0). 4. Empirical analysis We use panel regressions to test the effect of labour market institutions on inequality, controlling for contemporaneous changes in other determinants of inequality and potential omitted factors (through country- and time-fixed effects). 4.1 Baseline results We first estimate a simple model in which gross inequality measures (the top 10% income share and Gini of gross income) depend on labour market institutions and a vector of controls, which includes other determinants of inequality. The model is specified as follows: ln(Y)it=α1Xit+β1Zit+μ1i+θ1t+ϵ1,it (1) in which i denotes the country, and t the year. Y denotes either the top 10% income share, or the Gini of gross income. X includes labour market institution variables: these are union density and, in the case of the Gini of gross income, the ratio of the minimum wage to the median wage.10Z is a vector of controls, comprised of other determinants of inequality: technology (the share of information and communications technology capital in the total capital stock); globalization (the share of China in world exports interacted with the country’s lagged level of income per capita); financial reform (the index constructed by Abiad et al., 2008, which varies with changes in credit controls and reserve requirements, interest rate controls, entry barriers, state ownership, securities market policies, banking regulations, and capital account restrictions); and the top marginal personal income tax rate.11 Finally, μ and θ capture country- and time-fixed effects, respectively. To explain net income inequality, we estimate a model where the Gini of net income is assumed to depend on the Gini of gross income and a subset of variables that can affect redistribution; for example, union density and the top marginal tax rate. Ln(Gini Net)it=γLnGini Grossit+ α2Xit′+β2Zit′+μ2i+θ2t+ϵ2,it (2) where Gini Gross is defined as in equation (1) and X′ and Z′ denote, respectively, the subset of labour market and control variables that could influence redistribution. Determinants of inequality that do not enter equation (2) only affect the Gini of net income through their impact on the Gini of gross income. In other words, this specification assumes that technological progress, globalization, financial reform, and the minimum wage affect the Gini of gross income, but do not have an independent effect on redistribution. The sample consists of 18 countries, in the case of the top 10% income share, and 20 countries, for the Gini equations, and covers the period 1980–2011.12Table 2 reports results from a pooled OLS estimator, as well as a panel estimator with both country- and time-fixed effects. On the one hand, country-fixed effects capture countries’ time-invariant characteristics—including the specific methodology used to measure inequality—thereby isolating within-country changes and allowing for an assessment of the evolution of inequality over time. Time-fixed effects, on the other hand, capture how each country responds to common global phenomena. Standard errors are clustered at the country level, as disturbances are likely to be highly correlated within a country. Table 2 Level regressions Dependent variable (measure of inequality): . ln(Top 10%) . ln(Gini gross) . ln(Gini net) . . Pooled OLS . With FE . Pooled OLS . With FE . Pooled IV . IV with FE . . (1) . (2) . (3) . (4) . (5) . (6) . union density −0.47*** −0.46*** −0.12 −0.31** −0.48*** −0.29** (−5.457) (−2.928) (−1.052) (−2.759) (−7.304) (−2.374) ln(ICT K) 0.04 0.07 0.03 0.11*** (1.395) (1.469) (0.631) (5.664) ln (lag Y/N) −0.03 0.07 −0.22 −0.14 −0.08 0.31*** (−0.245) (0.470) (−1.304) (−0.768) (−1.084) (2.576) ln (lag Y/N) * China X 0.14 0.20*** −0.01 0.08 (1.635) (3.564) (−0.123) (1.710) China X −0.01 0.04 (−0.379) (0.873) ln(fin. reform) 0.06 0.05 0.15** 0.08*** (1.286) (1.175) (2.757) (3.113) top tax −0.14 −0.12 −0.12 −0.10 −0.20* −0.13** (−0.924) (−1.725) (−0.846) (−1.684) (−1.939) (−2.332) minimum wage −0.22** −0.39** (−2.276) (−2.519) ln(Gini gross) 0.38** 0.43 (1.989) (1.427) Observations 491 491 556 556 556 556 R-squared 0.570 0.685 0.342 0.687 0.663 0.480 country FE NO YES NO YES NO YES time FE NO YES NO YES NO YES Number of countries 18 18 20 20 20 20 Excluded instruments ln(ICT K) ln(lag Y/N)*China X China X ln(fin. reform) minimum wage ln(ICT K) ln(lag Y/N)*China X China X ln(fin. reform) minimum wage Underidentification test (H0: underidentified, p-value) 0.0481 0.116 Kleibergen-Paap rk Wald (F-statistic) 11.59 16.18 Hansen J-stat (H0: instruments are valid, p-value) 0.453 0.563 Dependent variable (measure of inequality): . ln(Top 10%) . ln(Gini gross) . ln(Gini net) . . Pooled OLS . With FE . Pooled OLS . With FE . Pooled IV . IV with FE . . (1) . (2) . (3) . (4) . (5) . (6) . union density −0.47*** −0.46*** −0.12 −0.31** −0.48*** −0.29** (−5.457) (−2.928) (−1.052) (−2.759) (−7.304) (−2.374) ln(ICT K) 0.04 0.07 0.03 0.11*** (1.395) (1.469) (0.631) (5.664) ln (lag Y/N) −0.03 0.07 −0.22 −0.14 −0.08 0.31*** (−0.245) (0.470) (−1.304) (−0.768) (−1.084) (2.576) ln (lag Y/N) * China X 0.14 0.20*** −0.01 0.08 (1.635) (3.564) (−0.123) (1.710) China X −0.01 0.04 (−0.379) (0.873) ln(fin. reform) 0.06 0.05 0.15** 0.08*** (1.286) (1.175) (2.757) (3.113) top tax −0.14 −0.12 −0.12 −0.10 −0.20* −0.13** (−0.924) (−1.725) (−0.846) (−1.684) (−1.939) (−2.332) minimum wage −0.22** −0.39** (−2.276) (−2.519) ln(Gini gross) 0.38** 0.43 (1.989) (1.427) Observations 491 491 556 556 556 556 R-squared 0.570 0.685 0.342 0.687 0.663 0.480 country FE NO YES NO YES NO YES time FE NO YES NO YES NO YES Number of countries 18 18 20 20 20 20 Excluded instruments ln(ICT K) ln(lag Y/N)*China X China X ln(fin. reform) minimum wage ln(ICT K) ln(lag Y/N)*China X China X ln(fin. reform) minimum wage Underidentification test (H0: underidentified, p-value) 0.0481 0.116 Kleibergen-Paap rk Wald (F-statistic) 11.59 16.18 Hansen J-stat (H0: instruments are valid, p-value) 0.453 0.563 Notes: Clustered t-statistics in parentheses; *** p < 0.01, ** p < 0.05, * p < 0.1. Source: Authors’ estimates. Open in new tab Table 2 Level regressions Dependent variable (measure of inequality): . ln(Top 10%) . ln(Gini gross) . ln(Gini net) . . Pooled OLS . With FE . Pooled OLS . With FE . Pooled IV . IV with FE . . (1) . (2) . (3) . (4) . (5) . (6) . union density −0.47*** −0.46*** −0.12 −0.31** −0.48*** −0.29** (−5.457) (−2.928) (−1.052) (−2.759) (−7.304) (−2.374) ln(ICT K) 0.04 0.07 0.03 0.11*** (1.395) (1.469) (0.631) (5.664) ln (lag Y/N) −0.03 0.07 −0.22 −0.14 −0.08 0.31*** (−0.245) (0.470) (−1.304) (−0.768) (−1.084) (2.576) ln (lag Y/N) * China X 0.14 0.20*** −0.01 0.08 (1.635) (3.564) (−0.123) (1.710) China X −0.01 0.04 (−0.379) (0.873) ln(fin. reform) 0.06 0.05 0.15** 0.08*** (1.286) (1.175) (2.757) (3.113) top tax −0.14 −0.12 −0.12 −0.10 −0.20* −0.13** (−0.924) (−1.725) (−0.846) (−1.684) (−1.939) (−2.332) minimum wage −0.22** −0.39** (−2.276) (−2.519) ln(Gini gross) 0.38** 0.43 (1.989) (1.427) Observations 491 491 556 556 556 556 R-squared 0.570 0.685 0.342 0.687 0.663 0.480 country FE NO YES NO YES NO YES time FE NO YES NO YES NO YES Number of countries 18 18 20 20 20 20 Excluded instruments ln(ICT K) ln(lag Y/N)*China X China X ln(fin. reform) minimum wage ln(ICT K) ln(lag Y/N)*China X China X ln(fin. reform) minimum wage Underidentification test (H0: underidentified, p-value) 0.0481 0.116 Kleibergen-Paap rk Wald (F-statistic) 11.59 16.18 Hansen J-stat (H0: instruments are valid, p-value) 0.453 0.563 Dependent variable (measure of inequality): . ln(Top 10%) . ln(Gini gross) . ln(Gini net) . . Pooled OLS . With FE . Pooled OLS . With FE . Pooled IV . IV with FE . . (1) . (2) . (3) . (4) . (5) . (6) . union density −0.47*** −0.46*** −0.12 −0.31** −0.48*** −0.29** (−5.457) (−2.928) (−1.052) (−2.759) (−7.304) (−2.374) ln(ICT K) 0.04 0.07 0.03 0.11*** (1.395) (1.469) (0.631) (5.664) ln (lag Y/N) −0.03 0.07 −0.22 −0.14 −0.08 0.31*** (−0.245) (0.470) (−1.304) (−0.768) (−1.084) (2.576) ln (lag Y/N) * China X 0.14 0.20*** −0.01 0.08 (1.635) (3.564) (−0.123) (1.710) China X −0.01 0.04 (−0.379) (0.873) ln(fin. reform) 0.06 0.05 0.15** 0.08*** (1.286) (1.175) (2.757) (3.113) top tax −0.14 −0.12 −0.12 −0.10 −0.20* −0.13** (−0.924) (−1.725) (−0.846) (−1.684) (−1.939) (−2.332) minimum wage −0.22** −0.39** (−2.276) (−2.519) ln(Gini gross) 0.38** 0.43 (1.989) (1.427) Observations 491 491 556 556 556 556 R-squared 0.570 0.685 0.342 0.687 0.663 0.480 country FE NO YES NO YES NO YES time FE NO YES NO YES NO YES Number of countries 18 18 20 20 20 20 Excluded instruments ln(ICT K) ln(lag Y/N)*China X China X ln(fin. reform) minimum wage ln(ICT K) ln(lag Y/N)*China X China X ln(fin. reform) minimum wage Underidentification test (H0: underidentified, p-value) 0.0481 0.116 Kleibergen-Paap rk Wald (F-statistic) 11.59 16.18 Hansen J-stat (H0: instruments are valid, p-value) 0.453 0.563 Notes: Clustered t-statistics in parentheses; *** p < 0.01, ** p < 0.05, * p < 0.1. Source: Authors’ estimates. Open in new tab In general, our estimations of gross and net income inequality validate the role of traditional determinants of income inequality highlighted in the literature—especially when inequality is measured by the Gini coefficient and when country- and time-fixed effects are included. We find that technological progress and globalization have a positive relation with inequality, and, among the institutional variables, the results confirm that lower top marginal tax rates and financial liberalization are associated with increased inequality. Our benchmark estimates of gross income inequality (Table 2, columns 1 and 2) indicate that the weakening of unions is significantly related to increases in the top 10% income share, even after taking account of the role played by traditional determinants. A 10 percentage point decline in union density is associated with a 5% increase in the top 10% income share. With most countries witnessing much larger declines in unionization over the period 1980–2011—some as large as 45 percentage points, as in the case of New Zealand—de-unionization has potentially had a non-trivial impact on the upper-tail inequality of advanced countries. The relation between union density and the Gini of gross income is weaker and only significant once we control for country- and time-fixed effects (Table 2, columns 3 and 4). The relation between union density and top income shares may appear surprising, as lower union density is believed to impact directly middle- and low-income earners, rather than top income shares. However, at the macroeconomic level, top income shares are mechanically influenced by what happens in the lower part of the income distribution (Volscho and Kelly, 2012; Bivens and Mishel, 2013). If de-unionization restrains earnings for middle- and low-income workers, this necessarily increases the income share of corporate managers and shareholders. At the other end of the income distribution, we find some evidence that the minimum wage is closely associated with the Gini coefficient of gross income. A 10 percentage point decline in the ratio of the minimum wage to the median wage is related to a 5% increase in the Gini coefficient of gross income. The Gini net equation is estimated with an instrumental variable approach, which implicitly assumes that technological progress, globalization, financial reform, and the minimum wage only affect the Gini of net income indirectly, through their effect on gross income inequality. Standard tests confirm that the instruments are valid. The results indicate that net income inequality is positively associated with gross inequality, with a coefficient less than 1 reflecting the impact of redistribution in reducing inequality. Further, lower union density is correlated with a higher Gini of net income for a given Gini of gross income (Table 2, columns 5 and 6). This result provides some support for the hypothesis that unions influence redistribution. Similarly, the evidence suggests that higher top marginal tax rates play a redistributive role. Table 3 presents our second set of baseline results where the regression for each inequality measure is estimated in first differences, first with pooled data and then with country-specific intercepts, which is equivalent to controlling for country-specific time trends in the level of inequality. While a first-difference estimation may exacerbate measurement error problems (Bound and Krueger, 1991), the coefficient of union density remains significant and of similar magnitude as in the level regressions for the top income share and the Gini of net income. For the Gini of gross income, however, the union density coefficient becomes small and insignificant, in line with the weaker level regressions’ results. In contrast, the negative relationship between the minimum wage and the Gini of gross income still holds in the first-difference model. Table 3 First-difference regressions Dependent variable (measure of inequality): . ln(Top 10%) . ln(Gini gross) . ln(Gini net) . . Pooled OLS . With FE . Pooled OLS . With FE . Pooled IV . IV with FE . . (1) . (2) . (3) . (4) . (5) . (6) . union density −0.43*** −0.40*** −0.02 0.03 −0.26*** −0.29** (−3.590) (−3.941) (−0.111) (0.167) (−2.663) (−2.463) ln(ICT K) 0.08 0.04 0.11*** 0.12** (1.428) (0.802) (2.966) (2.596) ln (lag Y/N) −0.12 −0.14 −0.16 −0.15 0.06 0.05 (−1.148) (−1.197) (−1.520) (−1.434) (0.847) (0.649) ln (lag Y/N)* China X 0.13* 0.15** 0.02 0.04 (2.106) (2.653) (0.372) (0.763) ln(fin. reform) −0.00 −0.01 0.05** 0.05** (−0.139) (−0.526) (2.138) (2.118) top tax −0.05* −0.05 −0.02 −0.02 0.01 0.01 (−1.748) (−1.594) (−0.498) (−0.503) (0.783) (1.002) minimum wage −0.32*** −0.34*** (−4.785) (−4.762) ln(Gini gross) 0.69*** 0.73*** (3.701) (4.031) Observations 473 473 536 536 536 536 R-squared 0.116 0.107 0.127 0.124 0.199 0.138 country FE NO YES NO YES NO YES Number of countries 18 18 20 20 20 20 Excluded instruments (in first differences) ln(ICT K) ln(lag Y/N)*China X China X ln(fin. reform) minimum wage ln(ICT K) ln(lag Y/N)*China X China X ln(fin. reform) minimum wage Underidentification test (H0: underidentified, p-value) 0.108 0.0419 Kleibergen-Paap rk Wald (F-statistic) 16.58 10.21 Hansen J-stat (H0: instruments are valid, p-value) 0.777 0.412 Dependent variable (measure of inequality): . ln(Top 10%) . ln(Gini gross) . ln(Gini net) . . Pooled OLS . With FE . Pooled OLS . With FE . Pooled IV . IV with FE . . (1) . (2) . (3) . (4) . (5) . (6) . union density −0.43*** −0.40*** −0.02 0.03 −0.26*** −0.29** (−3.590) (−3.941) (−0.111) (0.167) (−2.663) (−2.463) ln(ICT K) 0.08 0.04 0.11*** 0.12** (1.428) (0.802) (2.966) (2.596) ln (lag Y/N) −0.12 −0.14 −0.16 −0.15 0.06 0.05 (−1.148) (−1.197) (−1.520) (−1.434) (0.847) (0.649) ln (lag Y/N)* China X 0.13* 0.15** 0.02 0.04 (2.106) (2.653) (0.372) (0.763) ln(fin. reform) −0.00 −0.01 0.05** 0.05** (−0.139) (−0.526) (2.138) (2.118) top tax −0.05* −0.05 −0.02 −0.02 0.01 0.01 (−1.748) (−1.594) (−0.498) (−0.503) (0.783) (1.002) minimum wage −0.32*** −0.34*** (−4.785) (−4.762) ln(Gini gross) 0.69*** 0.73*** (3.701) (4.031) Observations 473 473 536 536 536 536 R-squared 0.116 0.107 0.127 0.124 0.199 0.138 country FE NO YES NO YES NO YES Number of countries 18 18 20 20 20 20 Excluded instruments (in first differences) ln(ICT K) ln(lag Y/N)*China X China X ln(fin. reform) minimum wage ln(ICT K) ln(lag Y/N)*China X China X ln(fin. reform) minimum wage Underidentification test (H0: underidentified, p-value) 0.108 0.0419 Kleibergen-Paap rk Wald (F-statistic) 16.58 10.21 Hansen J-stat (H0: instruments are valid, p-value) 0.777 0.412 Notes: Clustered t-statistics in parentheses; *** p < 0.01, ** p < 0.05, * p < 0.1. Source: Authors’ estimates. Open in new tab Table 3 First-difference regressions Dependent variable (measure of inequality): . ln(Top 10%) . ln(Gini gross) . ln(Gini net) . . Pooled OLS . With FE . Pooled OLS . With FE . Pooled IV . IV with FE . . (1) . (2) . (3) . (4) . (5) . (6) . union density −0.43*** −0.40*** −0.02 0.03 −0.26*** −0.29** (−3.590) (−3.941) (−0.111) (0.167) (−2.663) (−2.463) ln(ICT K) 0.08 0.04 0.11*** 0.12** (1.428) (0.802) (2.966) (2.596) ln (lag Y/N) −0.12 −0.14 −0.16 −0.15 0.06 0.05 (−1.148) (−1.197) (−1.520) (−1.434) (0.847) (0.649) ln (lag Y/N)* China X 0.13* 0.15** 0.02 0.04 (2.106) (2.653) (0.372) (0.763) ln(fin. reform) −0.00 −0.01 0.05** 0.05** (−0.139) (−0.526) (2.138) (2.118) top tax −0.05* −0.05 −0.02 −0.02 0.01 0.01 (−1.748) (−1.594) (−0.498) (−0.503) (0.783) (1.002) minimum wage −0.32*** −0.34*** (−4.785) (−4.762) ln(Gini gross) 0.69*** 0.73*** (3.701) (4.031) Observations 473 473 536 536 536 536 R-squared 0.116 0.107 0.127 0.124 0.199 0.138 country FE NO YES NO YES NO YES Number of countries 18 18 20 20 20 20 Excluded instruments (in first differences) ln(ICT K) ln(lag Y/N)*China X China X ln(fin. reform) minimum wage ln(ICT K) ln(lag Y/N)*China X China X ln(fin. reform) minimum wage Underidentification test (H0: underidentified, p-value) 0.108 0.0419 Kleibergen-Paap rk Wald (F-statistic) 16.58 10.21 Hansen J-stat (H0: instruments are valid, p-value) 0.777 0.412 Dependent variable (measure of inequality): . ln(Top 10%) . ln(Gini gross) . ln(Gini net) . . Pooled OLS . With FE . Pooled OLS . With FE . Pooled IV . IV with FE . . (1) . (2) . (3) . (4) . (5) . (6) . union density −0.43*** −0.40*** −0.02 0.03 −0.26*** −0.29** (−3.590) (−3.941) (−0.111) (0.167) (−2.663) (−2.463) ln(ICT K) 0.08 0.04 0.11*** 0.12** (1.428) (0.802) (2.966) (2.596) ln (lag Y/N) −0.12 −0.14 −0.16 −0.15 0.06 0.05 (−1.148) (−1.197) (−1.520) (−1.434) (0.847) (0.649) ln (lag Y/N)* China X 0.13* 0.15** 0.02 0.04 (2.106) (2.653) (0.372) (0.763) ln(fin. reform) −0.00 −0.01 0.05** 0.05** (−0.139) (−0.526) (2.138) (2.118) top tax −0.05* −0.05 −0.02 −0.02 0.01 0.01 (−1.748) (−1.594) (−0.498) (−0.503) (0.783) (1.002) minimum wage −0.32*** −0.34*** (−4.785) (−4.762) ln(Gini gross) 0.69*** 0.73*** (3.701) (4.031) Observations 473 473 536 536 536 536 R-squared 0.116 0.107 0.127 0.124 0.199 0.138 country FE NO YES NO YES NO YES Number of countries 18 18 20 20 20 20 Excluded instruments (in first differences) ln(ICT K) ln(lag Y/N)*China X China X ln(fin. reform) minimum wage ln(ICT K) ln(lag Y/N)*China X China X ln(fin. reform) minimum wage Underidentification test (H0: underidentified, p-value) 0.108 0.0419 Kleibergen-Paap rk Wald (F-statistic) 16.58 10.21 Hansen J-stat (H0: instruments are valid, p-value) 0.777 0.412 Notes: Clustered t-statistics in parentheses; *** p < 0.01, ** p < 0.05, * p < 0.1. Source: Authors’ estimates. Open in new tab The finding of a solid negative relation between unionization and top income shares in a cross-national panel data setting is the novel aspect of our work, and the rest of the paper focuses on demonstrating its robustness. We examine the time series properties of our data, and address potential issues of non-stationarity and heterogeneous coefficients across countries. Subsequently, we focus on addressing potential endogeneity bias related to omitted variables that could both lower unionization and increase inequality, as well as reverse causality. To establish the effect of union density on the top 10% income share as a causal link, we use an instrumental variable approach. 4.2 Time series properties of the data The panel structure of our data and the relatively long time-dimension of most panels (29 years) raise several potential issues. First, several variables could be non-stationary. Second, the assumption of homogeneity of parameters across countries could be inappropriate, both for the intercept and the slope parameters. Finally, there is a risk that common factors generate unobserved dependence between cross-sectional units (Pesaran, 2006, 2007). Therefore, our baseline result could be spurious, including because of autocorrelated error terms, which bias OLS estimates. To rule out this possibility and establish our findings more robustly, we address each of the aforementioned time series issues, focusing only on the 16 countries for which sufficiently long time series are available (Ireland and the United Kingdom are excluded). The variables in our model are first tested for unit roots. We use the IPS panel unit root test (Im et al., 2003), which does not require the panel dataset to be strictly balanced. Table 4 reports the results for each variable in both levels and first-differences. The results suggest that some variables—the top 10% income share, income per capita, China’s world export share, and union density—have unit roots and are integrated of order 1. Table 4 Panel unit root tests, 1980–2011 (16 countries) . Level . First-difference . Diagnostic . ln(top 10 share) 1.82 −9.67*** I(1) ln (ICT K) −6.87*** −2.75*** I(0) ln(lag Y/N) 2.17 −3.63*** I(1) ln(lag Y/N) * China X 10.16 −4.25*** I(1) China X 18.92 −6.01*** I(1) ln(fin. reform) −4.99*** −10.97*** I(0) top tax −2.02** −9.64*** I(0) union density −0.97 −6.96*** I(1) . Level . First-difference . Diagnostic . ln(top 10 share) 1.82 −9.67*** I(1) ln (ICT K) −6.87*** −2.75*** I(0) ln(lag Y/N) 2.17 −3.63*** I(1) ln(lag Y/N) * China X 10.16 −4.25*** I(1) China X 18.92 −6.01*** I(1) ln(fin. reform) −4.99*** −10.97*** I(0) top tax −2.02** −9.64*** I(0) union density −0.97 −6.96*** I(1) Notes: The table reports the Im et al. (2003) unit root tests. Significance levels at the 10%, 5%, and 1% levels are indicated by *, **, and ***, respectively. The null hypothesis of a unit root is rejected if the test statistic is significant. Source: Authors’ estimates. Open in new tab Table 4 Panel unit root tests, 1980–2011 (16 countries) . Level . First-difference . Diagnostic . ln(top 10 share) 1.82 −9.67*** I(1) ln (ICT K) −6.87*** −2.75*** I(0) ln(lag Y/N) 2.17 −3.63*** I(1) ln(lag Y/N) * China X 10.16 −4.25*** I(1) China X 18.92 −6.01*** I(1) ln(fin. reform) −4.99*** −10.97*** I(0) top tax −2.02** −9.64*** I(0) union density −0.97 −6.96*** I(1) . Level . First-difference . Diagnostic . ln(top 10 share) 1.82 −9.67*** I(1) ln (ICT K) −6.87*** −2.75*** I(0) ln(lag Y/N) 2.17 −3.63*** I(1) ln(lag Y/N) * China X 10.16 −4.25*** I(1) China X 18.92 −6.01*** I(1) ln(fin. reform) −4.99*** −10.97*** I(0) top tax −2.02** −9.64*** I(0) union density −0.97 −6.96*** I(1) Notes: The table reports the Im et al. (2003) unit root tests. Significance levels at the 10%, 5%, and 1% levels are indicated by *, **, and ***, respectively. The null hypothesis of a unit root is rejected if the test statistic is significant. Source: Authors’ estimates. Open in new tab All the variables in first-difference are stationary, however, suggesting that the first-difference estimations presented in Table 3 are not spurious. Since both the top income share (the dependent variable) and several regressors are integrated of order 1, we test for the existence of a cointegrating relationship between these variables, using panel cointegration tests developed by Pedroni (1999, 2004). The results suggest that, in general, the hypothesis of no cointegration among the main variables of interest can be rejected (Table 5). In other words, while several variables have a unit root, their linear combination does not. Table 5 Panel cointegration tests, 1980–2010 (16 countries) H0: no cointegration among ln(top 10 share), ln (ICT K), ln(lag Y/N) * China X, ln(lag Y/N), China X, ln(fin. reform), top tax, union density . Mean-group tests Pooled tests Modified Phillips Perron t 3.48*** 4.28*** 2.64*** 4.89*** −15.21*** 3.09*** 1.63* 3.41*** Phillips Perron t −2.01** −1.15 −2.87*** −2.57*** −1.96** −1.16 −3.40*** −3.29*** Augmented Dickey- Fuller t −1.66** −1.27 −1.33* −2.42*** −1.78** −1.47* −1.878** −3.69*** Country FE NO YES YES YES NO YES YES YES Time FE NO NO YES NO NO NO YES NO Time trend NO NO NO YES NO NO NO YES H0: no cointegration among ln(top 10 share), ln (ICT K), ln(lag Y/N) * China X, ln(lag Y/N), China X, ln(fin. reform), top tax, union density . Mean-group tests Pooled tests Modified Phillips Perron t 3.48*** 4.28*** 2.64*** 4.89*** −15.21*** 3.09*** 1.63* 3.41*** Phillips Perron t −2.01** −1.15 −2.87*** −2.57*** −1.96** −1.16 −3.40*** −3.29*** Augmented Dickey- Fuller t −1.66** −1.27 −1.33* −2.42*** −1.78** −1.47* −1.878** −3.69*** Country FE NO YES YES YES NO YES YES YES Time FE NO NO YES NO NO NO YES NO Time trend NO NO NO YES NO NO NO YES Notes: Significance levels at the 10%, 5%, and 1% levels are indicated by *, **, and ***, respectively. The null hypothesis of no cointegration is rejected if the test statistic is significant. Mean-group tests test for the presence of country-specific cointegrating vectors, while the pooled tests test for the same cointegrated vector across countries. The tests that include time FE do not include China X, because it is not a country-specific variable. Source: Authors’ estimates. Open in new tab Table 5 Panel cointegration tests, 1980–2010 (16 countries) H0: no cointegration among ln(top 10 share), ln (ICT K), ln(lag Y/N) * China X, ln(lag Y/N), China X, ln(fin. reform), top tax, union density . Mean-group tests Pooled tests Modified Phillips Perron t 3.48*** 4.28*** 2.64*** 4.89*** −15.21*** 3.09*** 1.63* 3.41*** Phillips Perron t −2.01** −1.15 −2.87*** −2.57*** −1.96** −1.16 −3.40*** −3.29*** Augmented Dickey- Fuller t −1.66** −1.27 −1.33* −2.42*** −1.78** −1.47* −1.878** −3.69*** Country FE NO YES YES YES NO YES YES YES Time FE NO NO YES NO NO NO YES NO Time trend NO NO NO YES NO NO NO YES H0: no cointegration among ln(top 10 share), ln (ICT K), ln(lag Y/N) * China X, ln(lag Y/N), China X, ln(fin. reform), top tax, union density . Mean-group tests Pooled tests Modified Phillips Perron t 3.48*** 4.28*** 2.64*** 4.89*** −15.21*** 3.09*** 1.63* 3.41*** Phillips Perron t −2.01** −1.15 −2.87*** −2.57*** −1.96** −1.16 −3.40*** −3.29*** Augmented Dickey- Fuller t −1.66** −1.27 −1.33* −2.42*** −1.78** −1.47* −1.878** −3.69*** Country FE NO YES YES YES NO YES YES YES Time FE NO NO YES NO NO NO YES NO Time trend NO NO NO YES NO NO NO YES Notes: Significance levels at the 10%, 5%, and 1% levels are indicated by *, **, and ***, respectively. The null hypothesis of no cointegration is rejected if the test statistic is significant. Mean-group tests test for the presence of country-specific cointegrating vectors, while the pooled tests test for the same cointegrated vector across countries. The tests that include time FE do not include China X, because it is not a country-specific variable. Source: Authors’ estimates. Open in new tab In the presence of cointegration, the residual of a fitted error-correction model is stationary. The model in equation (1) is thus re-parameterized into the following error correction equation: dln(Y)it=φ1ln(Y)it-1-μ1i-α1Xit-β1Zit+δ1dXit+ω1dZit+ϵ1,it (3) The error-correction term represents the long-run level relationship between the cointegrated variables (or equilibrium), while the first-differences of the regressors capture short-run dynamics, which are influenced by deviations from the long-run equilibrium. This framework allows us to use techniques developed to estimate country-specific parameters. Table 6 shows, respectively, the mean-group (MG) estimator, which relies on estimating country-specific time-series regressions and averaging the coefficients (see Pesaran and Smith, 1995); the pooled mean-group (PMG) estimator, which allows the intercept, short-run coefficients, and error variances to differ across countries, while constraining long-run coefficients to be equal across countries (see Pesaran et al., 1999); the pooled dynamic-fixed effect (DFE) estimator with clustered standard errors, where only the intercepts are allowed to be country-specific; and a DFE estimator with time-fixed effects. Table 6 Dynamic models Dependent variable (measure of inequality): . ln(Top 10%) . . Mean group . Pooled mean group . Dynamic FE . Dynamic and year FE . . (1) . (2) . (3) . (4) . error correction term −0.77*** −0.32*** −0.21*** −0.20*** (−8.077) (−3.627) (−4.642) (−4.281) union density −1.32* −0.43*** −0.35*** −0.26* (−1.770) (−5.427) (−2.652) (−1.731) ln(ICT K) 0.08** 0.00 0.11*** 0.09** (2.295) (0.027) (3.991) (2.063) ln (lag Y/N) −0.49** 0.16*** −0.16 −0.31* (−2.148) (2.610) (−0.910) (−1.745) ln (lag Y/N)* China X −0.43 −0.08* 0.15 0.13 (−1.545) (−1.685) (1.497) (1.176) China X 0.12 −0.01 −0.08*** (0.742) (−0.863) (−2.721) top tax −0.35** −0.18*** −0.19** −0.22*** (−2.223) (−6.295) (−2.156) (−2.916) ln(fin. reform) 0.16 −0.00 −0.04 −0.03 (0.605) (−0.201) (−1.399) (−0.726) short-term Δ union density 0.21 −0.72** −0.31** −0.33** (0.394) (−2.362) (−2.187) (−2.255) Δ ln(ICT K) −0.03 0.02 −0.00 0.03 (−0.568) (0.419) (−0.008) (0.463) Δ ln (lag Y/N) 0.20 −0.10 −0.08 −0.04 (1.043) (−1.286) (−1.025) (−0.303) Δ ln (lag Y/N)* China X 0.36 −0.06 0.13** 0.18*** (1.508) (−0.508) (2.178) (2.872) Δ China X −0.12 0.01 −0.01 (−0.798) (0.317) (−0.297) Δ top tax 0.02 −0.11 0.01 0.01 (0.265) (−1.256) (0.422) (0.298) Δ ln(fin. reform) −0.14 −0.03 −0.01 −0.01 (−0.715) (−0.765) (−0.891) (−0.839) Observations 443 443 443 443 Dependent variable (measure of inequality): . ln(Top 10%) . . Mean group . Pooled mean group . Dynamic FE . Dynamic and year FE . . (1) . (2) . (3) . (4) . error correction term −0.77*** −0.32*** −0.21*** −0.20*** (−8.077) (−3.627) (−4.642) (−4.281) union density −1.32* −0.43*** −0.35*** −0.26* (−1.770) (−5.427) (−2.652) (−1.731) ln(ICT K) 0.08** 0.00 0.11*** 0.09** (2.295) (0.027) (3.991) (2.063) ln (lag Y/N) −0.49** 0.16*** −0.16 −0.31* (−2.148) (2.610) (−0.910) (−1.745) ln (lag Y/N)* China X −0.43 −0.08* 0.15 0.13 (−1.545) (−1.685) (1.497) (1.176) China X 0.12 −0.01 −0.08*** (0.742) (−0.863) (−2.721) top tax −0.35** −0.18*** −0.19** −0.22*** (−2.223) (−6.295) (−2.156) (−2.916) ln(fin. reform) 0.16 −0.00 −0.04 −0.03 (0.605) (−0.201) (−1.399) (−0.726) short-term Δ union density 0.21 −0.72** −0.31** −0.33** (0.394) (−2.362) (−2.187) (−2.255) Δ ln(ICT K) −0.03 0.02 −0.00 0.03 (−0.568) (0.419) (−0.008) (0.463) Δ ln (lag Y/N) 0.20 −0.10 −0.08 −0.04 (1.043) (−1.286) (−1.025) (−0.303) Δ ln (lag Y/N)* China X 0.36 −0.06 0.13** 0.18*** (1.508) (−0.508) (2.178) (2.872) Δ China X −0.12 0.01 −0.01 (−0.798) (0.317) (−0.297) Δ top tax 0.02 −0.11 0.01 0.01 (0.265) (−1.256) (0.422) (0.298) Δ ln(fin. reform) −0.14 −0.03 −0.01 −0.01 (−0.715) (−0.765) (−0.891) (−0.839) Observations 443 443 443 443 Notes: z-statistics in parentheses (clustered for DFE estimates); *** p < 0.01, ** p < 0.05, * p < 0.1; Δ denotes first-difference. Source: Authors’ estimates. Open in new tab Table 6 Dynamic models Dependent variable (measure of inequality): . ln(Top 10%) . . Mean group . Pooled mean group . Dynamic FE . Dynamic and year FE . . (1) . (2) . (3) . (4) . error correction term −0.77*** −0.32*** −0.21*** −0.20*** (−8.077) (−3.627) (−4.642) (−4.281) union density −1.32* −0.43*** −0.35*** −0.26* (−1.770) (−5.427) (−2.652) (−1.731) ln(ICT K) 0.08** 0.00 0.11*** 0.09** (2.295) (0.027) (3.991) (2.063) ln (lag Y/N) −0.49** 0.16*** −0.16 −0.31* (−2.148) (2.610) (−0.910) (−1.745) ln (lag Y/N)* China X −0.43 −0.08* 0.15 0.13 (−1.545) (−1.685) (1.497) (1.176) China X 0.12 −0.01 −0.08*** (0.742) (−0.863) (−2.721) top tax −0.35** −0.18*** −0.19** −0.22*** (−2.223) (−6.295) (−2.156) (−2.916) ln(fin. reform) 0.16 −0.00 −0.04 −0.03 (0.605) (−0.201) (−1.399) (−0.726) short-term Δ union density 0.21 −0.72** −0.31** −0.33** (0.394) (−2.362) (−2.187) (−2.255) Δ ln(ICT K) −0.03 0.02 −0.00 0.03 (−0.568) (0.419) (−0.008) (0.463) Δ ln (lag Y/N) 0.20 −0.10 −0.08 −0.04 (1.043) (−1.286) (−1.025) (−0.303) Δ ln (lag Y/N)* China X 0.36 −0.06 0.13** 0.18*** (1.508) (−0.508) (2.178) (2.872) Δ China X −0.12 0.01 −0.01 (−0.798) (0.317) (−0.297) Δ top tax 0.02 −0.11 0.01 0.01 (0.265) (−1.256) (0.422) (0.298) Δ ln(fin. reform) −0.14 −0.03 −0.01 −0.01 (−0.715) (−0.765) (−0.891) (−0.839) Observations 443 443 443 443 Dependent variable (measure of inequality): . ln(Top 10%) . . Mean group . Pooled mean group . Dynamic FE . Dynamic and year FE . . (1) . (2) . (3) . (4) . error correction term −0.77*** −0.32*** −0.21*** −0.20*** (−8.077) (−3.627) (−4.642) (−4.281) union density −1.32* −0.43*** −0.35*** −0.26* (−1.770) (−5.427) (−2.652) (−1.731) ln(ICT K) 0.08** 0.00 0.11*** 0.09** (2.295) (0.027) (3.991) (2.063) ln (lag Y/N) −0.49** 0.16*** −0.16 −0.31* (−2.148) (2.610) (−0.910) (−1.745) ln (lag Y/N)* China X −0.43 −0.08* 0.15 0.13 (−1.545) (−1.685) (1.497) (1.176) China X 0.12 −0.01 −0.08*** (0.742) (−0.863) (−2.721) top tax −0.35** −0.18*** −0.19** −0.22*** (−2.223) (−6.295) (−2.156) (−2.916) ln(fin. reform) 0.16 −0.00 −0.04 −0.03 (0.605) (−0.201) (−1.399) (−0.726) short-term Δ union density 0.21 −0.72** −0.31** −0.33** (0.394) (−2.362) (−2.187) (−2.255) Δ ln(ICT K) −0.03 0.02 −0.00 0.03 (−0.568) (0.419) (−0.008) (0.463) Δ ln (lag Y/N) 0.20 −0.10 −0.08 −0.04 (1.043) (−1.286) (−1.025) (−0.303) Δ ln (lag Y/N)* China X 0.36 −0.06 0.13** 0.18*** (1.508) (−0.508) (2.178) (2.872) Δ China X −0.12 0.01 −0.01 (−0.798) (0.317) (−0.297) Δ top tax 0.02 −0.11 0.01 0.01 (0.265) (−1.256) (0.422) (0.298) Δ ln(fin. reform) −0.14 −0.03 −0.01 −0.01 (−0.715) (−0.765) (−0.891) (−0.839) Observations 443 443 443 443 Notes: z-statistics in parentheses (clustered for DFE estimates); *** p < 0.01, ** p < 0.05, * p < 0.1; Δ denotes first-difference. Source: Authors’ estimates. Open in new tab The estimates confirm the existence of a significant long-term negative relationship between union density and the top 10% income share, and (except in the MG estimates) a short-term negative relationship between union density and the top 10% income share. The long-run coefficient on union density is broadly of the same magnitude (albeit smaller) for the PMG and DFE estimations. The MG estimation differs the most, with a much more negative long-term effect of union density (albeit only significant at the 10% level), a much faster error-correcting speed of adjustment, and a non-significant (positive) short-term effect of union density on the top 10% income share. While MG is the most flexible estimation, the ‘pooling’ across countries in the PMG and DFE models yields more efficient and consistent estimates when the restrictions they impose are true. A Hausman test confirms that the difference in coefficients between the MG and PMG models is not systematic (the chi2 statistic is 4.29, with a p-value of 0.75); hence, the PMG estimator is preferred. Similarly, a comparison of the MG and DFE model indicates that the DFE model is consistent and efficient, and hence preferred (the chi2 statistic is 0.03, with a p-value of 1). This suggests that the risk of bias resulting from the endogeneity between the error term and the lagged dependent variable in the DFE model is small. We also test for the possibility that common factors generate unobserved cross-sectional dependence between countries. The cross-sectional dependence statistic proposed by Pesaran (2015) is −0.45, with a p-value of 0.65 (in the MG model) failing to reject the null hypothesis of weak cross-sectional dependence. In other words, there is no evidence of strong cross-sectional dependence.13 4.3 Omitted variables Next, we test the robustness of our result to factors that have possibly been omitted and that could be correlated with unionization and inequality. These include the economic cycle and banking crises, other labour market institutions, political factors and social norms, sectoral employment shifts, and increases in education levels.14 We test for robustness by re-estimating both the level equation and the DFE model with the additional control variables. Tables 7 and 8 report the estimates. Table 7 Level regressions with additional controls Dependent variable (measure of inequality): . ln(Top 10%) . Additional controls: . Cyclical conditions . Other LMIs . Political factors . Social values . Sectoral shifts . Financial sector . Education . . (1) . (2) . (3) . (4) . (5) . (6) . (7) . union density −0.46*** −0.44*** −0.44** −0.44* −0.67*** −0.73** −0.45** (−2.921) (−4.564) (−2.736) (−1.912) (−3.610) (−2.792) (−2.818) ln(ICT K) 0.07 0.03 0.06 0.09*** 0.04 0.04 0.07 (1.482) (0.705) (1.246) (3.265) (0.837) (0.808) (1.510) ln (lag Y/N) 0.09 0.15 0.06 −0.13 −0.04 −0.02 0.06 (0.531) (0.890) (0.361) (−0.855) (−0.236) (−0.115) (0.368) ln (lag Y/N)* China X 0.20*** 0.13*** 0.19*** 0.11** 0.21*** 0.10 0.20*** (3.428) (3.343) (3.779) (2.686) (4.281) (1.140) (3.559) ln(fin. reform) 0.05 0.01 0.01 0.05 0.05 0.05 0.05 (1.076) (0.345) (0.279) (0.628) (1.273) (1.241) (1.214) top tax −0.12* −0.08 −0.09 0.14* −0.12* −0.16*** −0.12 (−1.752) (−1.289) (−1.319) (1.883) (−1.933) (−3.690) (−1.633) bank crisis −0.02 (−1.169) output gap −0.00 (−0.516) excessive coll. cov. 0.31** (2.469) unempl. benefits 0.09 (1.052) EPL reg. contracts 0.02 (0.598) EPL temp. contracts −0.01 (−0.957) minimum wage −0.04 (−0.116) political right 0.03* (2.100) political left 0.03* (1.954) social pref. −0.01 (−0.520) share ind. emp. 0.01 (1.054) share serv. emp. 0.01 (1.040) share hours fin. 5.49 (1.393) higher educ. 0.00 (0.371) Observations 491 478 458 256 472 367 491 R-squared 0.686 0.722 0.679 0.612 0.703 0.776 0.686 Countries 18 18 17 18 18 15 18 country FE YES YES YES YES YES YES YES time FE YES YES YES YES YES YES YES Dependent variable (measure of inequality): . ln(Top 10%) . Additional controls: . Cyclical conditions . Other LMIs . Political factors . Social values . Sectoral shifts . Financial sector . Education . . (1) . (2) . (3) . (4) . (5) . (6) . (7) . union density −0.46*** −0.44*** −0.44** −0.44* −0.67*** −0.73** −0.45** (−2.921) (−4.564) (−2.736) (−1.912) (−3.610) (−2.792) (−2.818) ln(ICT K) 0.07 0.03 0.06 0.09*** 0.04 0.04 0.07 (1.482) (0.705) (1.246) (3.265) (0.837) (0.808) (1.510) ln (lag Y/N) 0.09 0.15 0.06 −0.13 −0.04 −0.02 0.06 (0.531) (0.890) (0.361) (−0.855) (−0.236) (−0.115) (0.368) ln (lag Y/N)* China X 0.20*** 0.13*** 0.19*** 0.11** 0.21*** 0.10 0.20*** (3.428) (3.343) (3.779) (2.686) (4.281) (1.140) (3.559) ln(fin. reform) 0.05 0.01 0.01 0.05 0.05 0.05 0.05 (1.076) (0.345) (0.279) (0.628) (1.273) (1.241) (1.214) top tax −0.12* −0.08 −0.09 0.14* −0.12* −0.16*** −0.12 (−1.752) (−1.289) (−1.319) (1.883) (−1.933) (−3.690) (−1.633) bank crisis −0.02 (−1.169) output gap −0.00 (−0.516) excessive coll. cov. 0.31** (2.469) unempl. benefits 0.09 (1.052) EPL reg. contracts 0.02 (0.598) EPL temp. contracts −0.01 (−0.957) minimum wage −0.04 (−0.116) political right 0.03* (2.100) political left 0.03* (1.954) social pref. −0.01 (−0.520) share ind. emp. 0.01 (1.054) share serv. emp. 0.01 (1.040) share hours fin. 5.49 (1.393) higher educ. 0.00 (0.371) Observations 491 478 458 256 472 367 491 R-squared 0.686 0.722 0.679 0.612 0.703 0.776 0.686 Countries 18 18 17 18 18 15 18 country FE YES YES YES YES YES YES YES time FE YES YES YES YES YES YES YES Notes: Clustered t-statistics in parentheses; *** p < 0.01, ** p < 0.05, * p < 0.1. Source: Authors’ estimates. Open in new tab Table 7 Level regressions with additional controls Dependent variable (measure of inequality): . ln(Top 10%) . Additional controls: . Cyclical conditions . Other LMIs . Political factors . Social values . Sectoral shifts . Financial sector . Education . . (1) . (2) . (3) . (4) . (5) . (6) . (7) . union density −0.46*** −0.44*** −0.44** −0.44* −0.67*** −0.73** −0.45** (−2.921) (−4.564) (−2.736) (−1.912) (−3.610) (−2.792) (−2.818) ln(ICT K) 0.07 0.03 0.06 0.09*** 0.04 0.04 0.07 (1.482) (0.705) (1.246) (3.265) (0.837) (0.808) (1.510) ln (lag Y/N) 0.09 0.15 0.06 −0.13 −0.04 −0.02 0.06 (0.531) (0.890) (0.361) (−0.855) (−0.236) (−0.115) (0.368) ln (lag Y/N)* China X 0.20*** 0.13*** 0.19*** 0.11** 0.21*** 0.10 0.20*** (3.428) (3.343) (3.779) (2.686) (4.281) (1.140) (3.559) ln(fin. reform) 0.05 0.01 0.01 0.05 0.05 0.05 0.05 (1.076) (0.345) (0.279) (0.628) (1.273) (1.241) (1.214) top tax −0.12* −0.08 −0.09 0.14* −0.12* −0.16*** −0.12 (−1.752) (−1.289) (−1.319) (1.883) (−1.933) (−3.690) (−1.633) bank crisis −0.02 (−1.169) output gap −0.00 (−0.516) excessive coll. cov. 0.31** (2.469) unempl. benefits 0.09 (1.052) EPL reg. contracts 0.02 (0.598) EPL temp. contracts −0.01 (−0.957) minimum wage −0.04 (−0.116) political right 0.03* (2.100) political left 0.03* (1.954) social pref. −0.01 (−0.520) share ind. emp. 0.01 (1.054) share serv. emp. 0.01 (1.040) share hours fin. 5.49 (1.393) higher educ. 0.00 (0.371) Observations 491 478 458 256 472 367 491 R-squared 0.686 0.722 0.679 0.612 0.703 0.776 0.686 Countries 18 18 17 18 18 15 18 country FE YES YES YES YES YES YES YES time FE YES YES YES YES YES YES YES Dependent variable (measure of inequality): . ln(Top 10%) . Additional controls: . Cyclical conditions . Other LMIs . Political factors . Social values . Sectoral shifts . Financial sector . Education . . (1) . (2) . (3) . (4) . (5) . (6) . (7) . union density −0.46*** −0.44*** −0.44** −0.44* −0.67*** −0.73** −0.45** (−2.921) (−4.564) (−2.736) (−1.912) (−3.610) (−2.792) (−2.818) ln(ICT K) 0.07 0.03 0.06 0.09*** 0.04 0.04 0.07 (1.482) (0.705) (1.246) (3.265) (0.837) (0.808) (1.510) ln (lag Y/N) 0.09 0.15 0.06 −0.13 −0.04 −0.02 0.06 (0.531) (0.890) (0.361) (−0.855) (−0.236) (−0.115) (0.368) ln (lag Y/N)* China X 0.20*** 0.13*** 0.19*** 0.11** 0.21*** 0.10 0.20*** (3.428) (3.343) (3.779) (2.686) (4.281) (1.140) (3.559) ln(fin. reform) 0.05 0.01 0.01 0.05 0.05 0.05 0.05 (1.076) (0.345) (0.279) (0.628) (1.273) (1.241) (1.214) top tax −0.12* −0.08 −0.09 0.14* −0.12* −0.16*** −0.12 (−1.752) (−1.289) (−1.319) (1.883) (−1.933) (−3.690) (−1.633) bank crisis −0.02 (−1.169) output gap −0.00 (−0.516) excessive coll. cov. 0.31** (2.469) unempl. benefits 0.09 (1.052) EPL reg. contracts 0.02 (0.598) EPL temp. contracts −0.01 (−0.957) minimum wage −0.04 (−0.116) political right 0.03* (2.100) political left 0.03* (1.954) social pref. −0.01 (−0.520) share ind. emp. 0.01 (1.054) share serv. emp. 0.01 (1.040) share hours fin. 5.49 (1.393) higher educ. 0.00 (0.371) Observations 491 478 458 256 472 367 491 R-squared 0.686 0.722 0.679 0.612 0.703 0.776 0.686 Countries 18 18 17 18 18 15 18 country FE YES YES YES YES YES YES YES time FE YES YES YES YES YES YES YES Notes: Clustered t-statistics in parentheses; *** p < 0.01, ** p < 0.05, * p < 0.1. Source: Authors’ estimates. Open in new tab Table 8 Dynamic FE model with additional controls Dependent variable (measure of inequality): . ln(Top 10%) . . Cyclical conditions . Other LMIs . Political factors . Social values . Sectoral shifts . Financial sector . Higher education . . (1) . (2) . (3) . (4) . (5) . (6) . (7) . error correction term −0.21*** −0.22*** −0.21*** −0.49*** −0.20*** −0.17*** −0.21*** (−4.556) (−3.904) (−3.824) (−5.198) (−4.269) (−5.290) (−4.800) union density −0.35*** −0.29** −0.38*** −0.47 −0.30** −0.43** −0.32** (−2.607) (−2.519) (−2.775) (−1.338) (−2.137) (−2.168) (−2.330) ln(ICT K) 0.10*** 0.08*** 0.11*** 0.10** 0.11*** 0.09** 0.10*** (3.242) (2.750) (3.922) (2.481) (3.126) (1.981) (3.813) ln (lag Y/N) −0.12 −0.00 −0.11 −0.30* −0.14 −0.18 −0.18 (−0.511) (−0.008) (−0.567) (−1.803) (−0.772) (−1.011) (−1.039) ln (lag Y/N)* China X 0.14 0.12 0.16* −0.14 0.10 0.17 0.13 (1.219) (1.062) (1.699) (−1.404) (1.134) (1.304) (1.460) China X −0.07*** −0.07** −0.09*** −0.01 −0.08*** −0.04 −0.08*** (−2.652) (−2.507) (−3.716) (−0.814) (−2.821) (−1.018) (−2.656) top tax −0.20** −0.16* −0.21** 0.03 −0.17* −0.24*** −0.16* (−2.059) (−1.942) (−2.165) (0.389) (−1.946) (−2.980) (−1.865) ln(fin. reform) −0.04 −0.05 −0.06* −0.03 −0.01 −0.05 −0.03 (−1.214) (−1.538) (−1.799) (−0.374) (−0.295) (−1.364) (−1.082) bank crisis −0.04 (−0.592) output gap −0.00 (−0.144) excessive coll. cov. 0.25*** (2.604) unempl. benefits −0.01 (−0.170) EPL reg. contracts 0.04 (1.211) EPL temp. contracts −0.02* (−1.896) minimum wage 0.25 (1.040) political right 0.03** (2.245) political left 0.03 (1.353) social pref. −0.01 (−0.562) share ind. emp. −0.01 (−1.395) share serv. emp. −0.01 (−0.995) share hours fin. 6.24 (1.340) higher educ. 0.00 (0.816) short-term Δ union density −0.30** −0.29* −0.27* 0.04 −0.45*** −0.15 −0.30** (−2.016) (−1.840) (−1.860) (0.155) (−2.583) (−0.936) (−2.115) Δ ln(ICT K) −0.00 0.01 0.00 −0.02 −0.01 −0.01 −0.00 (−0.164) (0.409) (0.100) (−0.324) (−0.379) (−0.939) (−0.138) Δ ln (lag Y/N) −0.09 −0.10 −0.09 0.01 −0.07 0.01 −0.08 (−0.791) (−1.206) (−1.156) (0.038) (−1.224) (0.188) (−1.006) Δ ln (lag Y/N)* China X 0.14** 0.13*** 0.13** 0.61 0.14** 0.08 0.13** (2.208) (2.625) (2.237) (1.521) (2.369) (0.934) (2.142) Δ China X −0.01 −0.00 −0.00 −0.08 −0.00 0.01 −0.00 (−0.764) (−0.193) (−0.296) (−1.367) (−0.154) (0.321) (−0.203) Δ top tax 0.01 0.02 0.01 0.10 0.01 −0.01 0.01 (0.468) (0.525) (0.325) (1.530) (0.184) (−0.360) (0.213) Δ ln(fin. reform) −0.01 −0.01 −0.01 0.02 −0.00 0.01 −0.01 (−0.782) (−0.489) (−0.723) (0.502) (−0.060) (0.587) (−1.004) Δ bank crisis 0.01 (0.950) Δ output gap 0.00 (0.395) Δ excessive coll. cov. −0.03 (−0.378) Δ unempl. Benefits 0.05** (2.446) Δ EPL reg. contracts −0.00 (−0.112) Δ EPL temp. contracts −0.00 (−0.135) Δ minimum wage −0.28* (−1.866) Δ political right −0.00 (−0.278) Δ political left −0.01*** (−3.497) Δ social pref. −0.01 (−0.133) Δ share ind. emp. −0.00 (−0.650) Δ share serv. emp. −0.00 (−0.142) Δ share hours fin. −2.18 (−1.176) Δ higher educ. −0.00** (−2.059) Dependent variable (measure of inequality): . ln(Top 10%) . . Cyclical conditions . Other LMIs . Political factors . Social values . Sectoral shifts . Financial sector . Higher education . . (1) . (2) . (3) . (4) . (5) . (6) . (7) . error correction term −0.21*** −0.22*** −0.21*** −0.49*** −0.20*** −0.17*** −0.21*** (−4.556) (−3.904) (−3.824) (−5.198) (−4.269) (−5.290) (−4.800) union density −0.35*** −0.29** −0.38*** −0.47 −0.30** −0.43** −0.32** (−2.607) (−2.519) (−2.775) (−1.338) (−2.137) (−2.168) (−2.330) ln(ICT K) 0.10*** 0.08*** 0.11*** 0.10** 0.11*** 0.09** 0.10*** (3.242) (2.750) (3.922) (2.481) (3.126) (1.981) (3.813) ln (lag Y/N) −0.12 −0.00 −0.11 −0.30* −0.14 −0.18 −0.18 (−0.511) (−0.008) (−0.567) (−1.803) (−0.772) (−1.011) (−1.039) ln (lag Y/N)* China X 0.14 0.12 0.16* −0.14 0.10 0.17 0.13 (1.219) (1.062) (1.699) (−1.404) (1.134) (1.304) (1.460) China X −0.07*** −0.07** −0.09*** −0.01 −0.08*** −0.04 −0.08*** (−2.652) (−2.507) (−3.716) (−0.814) (−2.821) (−1.018) (−2.656) top tax −0.20** −0.16* −0.21** 0.03 −0.17* −0.24*** −0.16* (−2.059) (−1.942) (−2.165) (0.389) (−1.946) (−2.980) (−1.865) ln(fin. reform) −0.04 −0.05 −0.06* −0.03 −0.01 −0.05 −0.03 (−1.214) (−1.538) (−1.799) (−0.374) (−0.295) (−1.364) (−1.082) bank crisis −0.04 (−0.592) output gap −0.00 (−0.144) excessive coll. cov. 0.25*** (2.604) unempl. benefits −0.01 (−0.170) EPL reg. contracts 0.04 (1.211) EPL temp. contracts −0.02* (−1.896) minimum wage 0.25 (1.040) political right 0.03** (2.245) political left 0.03 (1.353) social pref. −0.01 (−0.562) share ind. emp. −0.01 (−1.395) share serv. emp. −0.01 (−0.995) share hours fin. 6.24 (1.340) higher educ. 0.00 (0.816) short-term Δ union density −0.30** −0.29* −0.27* 0.04 −0.45*** −0.15 −0.30** (−2.016) (−1.840) (−1.860) (0.155) (−2.583) (−0.936) (−2.115) Δ ln(ICT K) −0.00 0.01 0.00 −0.02 −0.01 −0.01 −0.00 (−0.164) (0.409) (0.100) (−0.324) (−0.379) (−0.939) (−0.138) Δ ln (lag Y/N) −0.09 −0.10 −0.09 0.01 −0.07 0.01 −0.08 (−0.791) (−1.206) (−1.156) (0.038) (−1.224) (0.188) (−1.006) Δ ln (lag Y/N)* China X 0.14** 0.13*** 0.13** 0.61 0.14** 0.08 0.13** (2.208) (2.625) (2.237) (1.521) (2.369) (0.934) (2.142) Δ China X −0.01 −0.00 −0.00 −0.08 −0.00 0.01 −0.00 (−0.764) (−0.193) (−0.296) (−1.367) (−0.154) (0.321) (−0.203) Δ top tax 0.01 0.02 0.01 0.10 0.01 −0.01 0.01 (0.468) (0.525) (0.325) (1.530) (0.184) (−0.360) (0.213) Δ ln(fin. reform) −0.01 −0.01 −0.01 0.02 −0.00 0.01 −0.01 (−0.782) (−0.489) (−0.723) (0.502) (−0.060) (0.587) (−1.004) Δ bank crisis 0.01 (0.950) Δ output gap 0.00 (0.395) Δ excessive coll. cov. −0.03 (−0.378) Δ unempl. Benefits 0.05** (2.446) Δ EPL reg. contracts −0.00 (−0.112) Δ EPL temp. contracts −0.00 (−0.135) Δ minimum wage −0.28* (−1.866) Δ political right −0.00 (−0.278) Δ political left −0.01*** (−3.497) Δ social pref. −0.01 (−0.133) Δ share ind. emp. −0.00 (−0.650) Δ share serv. emp. −0.00 (−0.142) Δ share hours fin. −2.18 (−1.176) Δ higher educ. −0.00** (−2.059) Notes: Clustered z-statistics in parentheses; ***p < 0.01, ** p < 0.05, * p < 0.1; Δ denotes first difference. Source: Authors’ estimates. Open in new tab Table 8 Dynamic FE model with additional controls Dependent variable (measure of inequality): . ln(Top 10%) . . Cyclical conditions . Other LMIs . Political factors . Social values . Sectoral shifts . Financial sector . Higher education . . (1) . (2) . (3) . (4) . (5) . (6) . (7) . error correction term −0.21*** −0.22*** −0.21*** −0.49*** −0.20*** −0.17*** −0.21*** (−4.556) (−3.904) (−3.824) (−5.198) (−4.269) (−5.290) (−4.800) union density −0.35*** −0.29** −0.38*** −0.47 −0.30** −0.43** −0.32** (−2.607) (−2.519) (−2.775) (−1.338) (−2.137) (−2.168) (−2.330) ln(ICT K) 0.10*** 0.08*** 0.11*** 0.10** 0.11*** 0.09** 0.10*** (3.242) (2.750) (3.922) (2.481) (3.126) (1.981) (3.813) ln (lag Y/N) −0.12 −0.00 −0.11 −0.30* −0.14 −0.18 −0.18 (−0.511) (−0.008) (−0.567) (−1.803) (−0.772) (−1.011) (−1.039) ln (lag Y/N)* China X 0.14 0.12 0.16* −0.14 0.10 0.17 0.13 (1.219) (1.062) (1.699) (−1.404) (1.134) (1.304) (1.460) China X −0.07*** −0.07** −0.09*** −0.01 −0.08*** −0.04 −0.08*** (−2.652) (−2.507) (−3.716) (−0.814) (−2.821) (−1.018) (−2.656) top tax −0.20** −0.16* −0.21** 0.03 −0.17* −0.24*** −0.16* (−2.059) (−1.942) (−2.165) (0.389) (−1.946) (−2.980) (−1.865) ln(fin. reform) −0.04 −0.05 −0.06* −0.03 −0.01 −0.05 −0.03 (−1.214) (−1.538) (−1.799) (−0.374) (−0.295) (−1.364) (−1.082) bank crisis −0.04 (−0.592) output gap −0.00 (−0.144) excessive coll. cov. 0.25*** (2.604) unempl. benefits −0.01 (−0.170) EPL reg. contracts 0.04 (1.211) EPL temp. contracts −0.02* (−1.896) minimum wage 0.25 (1.040) political right 0.03** (2.245) political left 0.03 (1.353) social pref. −0.01 (−0.562) share ind. emp. −0.01 (−1.395) share serv. emp. −0.01 (−0.995) share hours fin. 6.24 (1.340) higher educ. 0.00 (0.816) short-term Δ union density −0.30** −0.29* −0.27* 0.04 −0.45*** −0.15 −0.30** (−2.016) (−1.840) (−1.860) (0.155) (−2.583) (−0.936) (−2.115) Δ ln(ICT K) −0.00 0.01 0.00 −0.02 −0.01 −0.01 −0.00 (−0.164) (0.409) (0.100) (−0.324) (−0.379) (−0.939) (−0.138) Δ ln (lag Y/N) −0.09 −0.10 −0.09 0.01 −0.07 0.01 −0.08 (−0.791) (−1.206) (−1.156) (0.038) (−1.224) (0.188) (−1.006) Δ ln (lag Y/N)* China X 0.14** 0.13*** 0.13** 0.61 0.14** 0.08 0.13** (2.208) (2.625) (2.237) (1.521) (2.369) (0.934) (2.142) Δ China X −0.01 −0.00 −0.00 −0.08 −0.00 0.01 −0.00 (−0.764) (−0.193) (−0.296) (−1.367) (−0.154) (0.321) (−0.203) Δ top tax 0.01 0.02 0.01 0.10 0.01 −0.01 0.01 (0.468) (0.525) (0.325) (1.530) (0.184) (−0.360) (0.213) Δ ln(fin. reform) −0.01 −0.01 −0.01 0.02 −0.00 0.01 −0.01 (−0.782) (−0.489) (−0.723) (0.502) (−0.060) (0.587) (−1.004) Δ bank crisis 0.01 (0.950) Δ output gap 0.00 (0.395) Δ excessive coll. cov. −0.03 (−0.378) Δ unempl. Benefits 0.05** (2.446) Δ EPL reg. contracts −0.00 (−0.112) Δ EPL temp. contracts −0.00 (−0.135) Δ minimum wage −0.28* (−1.866) Δ political right −0.00 (−0.278) Δ political left −0.01*** (−3.497) Δ social pref. −0.01 (−0.133) Δ share ind. emp. −0.00 (−0.650) Δ share serv. emp. −0.00 (−0.142) Δ share hours fin. −2.18 (−1.176) Δ higher educ. −0.00** (−2.059) Dependent variable (measure of inequality): . ln(Top 10%) . . Cyclical conditions . Other LMIs . Political factors . Social values . Sectoral shifts . Financial sector . Higher education . . (1) . (2) . (3) . (4) . (5) . (6) . (7) . error correction term −0.21*** −0.22*** −0.21*** −0.49*** −0.20*** −0.17*** −0.21*** (−4.556) (−3.904) (−3.824) (−5.198) (−4.269) (−5.290) (−4.800) union density −0.35*** −0.29** −0.38*** −0.47 −0.30** −0.43** −0.32** (−2.607) (−2.519) (−2.775) (−1.338) (−2.137) (−2.168) (−2.330) ln(ICT K) 0.10*** 0.08*** 0.11*** 0.10** 0.11*** 0.09** 0.10*** (3.242) (2.750) (3.922) (2.481) (3.126) (1.981) (3.813) ln (lag Y/N) −0.12 −0.00 −0.11 −0.30* −0.14 −0.18 −0.18 (−0.511) (−0.008) (−0.567) (−1.803) (−0.772) (−1.011) (−1.039) ln (lag Y/N)* China X 0.14 0.12 0.16* −0.14 0.10 0.17 0.13 (1.219) (1.062) (1.699) (−1.404) (1.134) (1.304) (1.460) China X −0.07*** −0.07** −0.09*** −0.01 −0.08*** −0.04 −0.08*** (−2.652) (−2.507) (−3.716) (−0.814) (−2.821) (−1.018) (−2.656) top tax −0.20** −0.16* −0.21** 0.03 −0.17* −0.24*** −0.16* (−2.059) (−1.942) (−2.165) (0.389) (−1.946) (−2.980) (−1.865) ln(fin. reform) −0.04 −0.05 −0.06* −0.03 −0.01 −0.05 −0.03 (−1.214) (−1.538) (−1.799) (−0.374) (−0.295) (−1.364) (−1.082) bank crisis −0.04 (−0.592) output gap −0.00 (−0.144) excessive coll. cov. 0.25*** (2.604) unempl. benefits −0.01 (−0.170) EPL reg. contracts 0.04 (1.211) EPL temp. contracts −0.02* (−1.896) minimum wage 0.25 (1.040) political right 0.03** (2.245) political left 0.03 (1.353) social pref. −0.01 (−0.562) share ind. emp. −0.01 (−1.395) share serv. emp. −0.01 (−0.995) share hours fin. 6.24 (1.340) higher educ. 0.00 (0.816) short-term Δ union density −0.30** −0.29* −0.27* 0.04 −0.45*** −0.15 −0.30** (−2.016) (−1.840) (−1.860) (0.155) (−2.583) (−0.936) (−2.115) Δ ln(ICT K) −0.00 0.01 0.00 −0.02 −0.01 −0.01 −0.00 (−0.164) (0.409) (0.100) (−0.324) (−0.379) (−0.939) (−0.138) Δ ln (lag Y/N) −0.09 −0.10 −0.09 0.01 −0.07 0.01 −0.08 (−0.791) (−1.206) (−1.156) (0.038) (−1.224) (0.188) (−1.006) Δ ln (lag Y/N)* China X 0.14** 0.13*** 0.13** 0.61 0.14** 0.08 0.13** (2.208) (2.625) (2.237) (1.521) (2.369) (0.934) (2.142) Δ China X −0.01 −0.00 −0.00 −0.08 −0.00 0.01 −0.00 (−0.764) (−0.193) (−0.296) (−1.367) (−0.154) (0.321) (−0.203) Δ top tax 0.01 0.02 0.01 0.10 0.01 −0.01 0.01 (0.468) (0.525) (0.325) (1.530) (0.184) (−0.360) (0.213) Δ ln(fin. reform) −0.01 −0.01 −0.01 0.02 −0.00 0.01 −0.01 (−0.782) (−0.489) (−0.723) (0.502) (−0.060) (0.587) (−1.004) Δ bank crisis 0.01 (0.950) Δ output gap 0.00 (0.395) Δ excessive coll. cov. −0.03 (−0.378) Δ unempl. Benefits 0.05** (2.446) Δ EPL reg. contracts −0.00 (−0.112) Δ EPL temp. contracts −0.00 (−0.135) Δ minimum wage −0.28* (−1.866) Δ political right −0.00 (−0.278) Δ political left −0.01*** (−3.497) Δ social pref. −0.01 (−0.133) Δ share ind. emp. −0.00 (−0.650) Δ share serv. emp. −0.00 (−0.142) Δ share hours fin. −2.18 (−1.176) Δ higher educ. −0.00** (−2.059) Notes: Clustered z-statistics in parentheses; ***p < 0.01, ** p < 0.05, * p < 0.1; Δ denotes first difference. Source: Authors’ estimates. Open in new tab 4.3.1 Cyclical and crisis factors Recessions and economic crises could lead to both higher inequality and lower union density as unemployment rises. However, some crises—especially banking crises—could, instead, reduce the top 10% income share as the sharp fall in equity prices disproportionately affects the wealthy. Controlling for the output gap and a banking crisis dummy variable does not affect the result for union density (column 1). Moreover, these two additional control variables are not statistically significant. 4.3.2 Other labour market institutions A possible caveat to the negative relation between union density and inequality is that union density need not equate to union strength. Owing to extension agreements, collective bargaining coverage (the proportion of workers covered by collective agreements) can be very high and union density very low. Two noteworthy examples are France and Spain. While higher collective bargaining coverage may increase the incomes of the population covered by the agreements, an excess coverage of collective agreements (relative to union density) may also raise unemployment, thereby leading to higher inequality. That is, since unions may not internalize the effects of their wage demands on the whole workforce, these may become excessive. Indeed, Bouis et al. (2012) find that increases in the excess coverage of collective bargaining, defined as the difference between the share of workers covered by collective agreements and the share of workers that are members of a union, lead to higher unemployment. In addition, the strong relation we find between union density and inequality could capture the impact of other labour market institutions, either because these institutions are jointly determined by underlying social preferences, or because unions are active in defending institutions that protect workers. We therefore re-estimate our benchmark model adding excess collective bargaining coverage as an explanatory variable, and taking into account the role played by other labour market institutions: the generosity of unemployment benefits (the average gross income replacement rate in the first year of unemployment), the stringency of employment protection for regular and temporary contracts, and the ratio of the minimum wage to the median wage (column 2). The results indicate that the coefficients of union density in the top 10% income share are broadly unaffected. Interestingly, there is some evidence that excess collective bargaining coverage is positively associated with the top 10% income share. Thus, for countries where collective agreements coverage exceeds the extent of unionization considerably, inequality may rise. Overall, however, the relation between inequality and excess collective bargaining coverage does not survive all the robustness tests discussed in this paper. The other control labour market institutions are not statistically significant. 4.3.3 Political factors and social norms From the late 1970s to the mid-1990s, there was a shift towards elected governments that favoured the implementation of more market-driven policies, which could have led to a decline in union density and an increase in inequality. However, controlling for the political orientation of elected governments does not alter the union density result (column 3). Moreover, the coefficient on union density remains unaffected by the inclusion of a measure of social preferences regarding inequality (column 4).15 Changes in the top marginal tax rate could also be seen as an indirect measure of social preferences with respect to inequality. Controlling for this variable in the baseline regression, however, did not affect the union density result. 4.3.4 Sectoral shift Another major shift that occurred under the pressure of globalization was the de-industrialization of advanced economies. This phenomenon led to a decline in union density (as the services sector outgrew the industrial sector and industry workers tend to be more unionized) and a rise in inequality (as industry workers struggled to find alternative high-paying jobs in services).16 Controls for employment shares in industry and services actually increase the magnitude and significance of the coefficient on union density (column 5). Further, in the top 10% income share equation, the coefficients on the share of employment in the industry sector and in services are insignificant. 4.3.5 Role of employment in finance The financial sector has grown in importance in many countries. Its unionization rate is one of the lowest of all sectors; and compensation in this sector relative to the rest of the economy has been growing fast. Thus, we control for the share of hours worked in finance relative to the total economy (column 6). Once more, the coefficient on union density is robust, even increasing in magnitude and significance. The coefficient on the share of finance is positive, but not statistically significant.17 4.3.6 Higher education During the period 1980–2011, education levels in advanced economies improved considerably, reflected in a rise of the share of workers with at least some post-secondary education. More formal schooling leads workers towards jobs and/or sectors in which the incentives to organize unions are weaker (Acemoglu et al., 2001) while, at the same time, they are generating larger income differentials. However, we do not find evidence of a significant relation between the share of workers with higher education and the top income share, while our result on union density remains unaffected (column 7). 4.4 Instrumental variable estimation Finally, we attempt to establish causality using an instrumental variable strategy for union density. This requires setting out a model of what we think determines union density. Several factors have been advanced to explain the decline in unionization. These include SBTC and globalization, which are in our set of control variables for inequality. SBTC increases the outside option of skilled workers and weakens their incentives to join unions (Acemoglu et al., 2001). Globalization increases competition and reduces rents that could be appropriated by unions, reducing the benefit of joining a union. Other possible determinants of union density have been discussed in Section 4.3 on omitted variables, such as changes in sectoral employment shares, the political orientation of the government, and the level of education of the workforce. Finally, some institutional variables could also have a significant impact on union density. Wallerstein and Western (2000) find that countries where the unemployment insurance system is managed by unions (Ghent systems), or where collective bargaining is more centralized, tended to experience smaller declines in unionization rates in periods of poor economic conditions and high unemployment. Indeed, Fig. 4 showed that most Nordic countries and Belgium—which have a more centralized collective bargaining system and the Ghent system of unemployment benefits—experienced only limited declines in unionization rates. While this list of determinants of union density is not exhaustive, it provides us with a solid basis for our instrumentation strategy. In addition to the control variables, which are automatically included in the first stage estimation, it suggests a number of excluded instruments—specifically, the employment shares in industry and in services (which did not affect inequality), an indicator variable for countries with a Ghent system of unemployment benefits, and an indicator variable for the degree of centralization of collective bargaining. The latter two instruments are interacted with the five-year lagged OECD unemployment rate to capture their differentiated effect across periods of unemployment, and to distinguish them from the country-fixed effects (since they are for the most part time-invariant).18 We use the OECD average unemployment rate and lag it by five years to reduce the risk of endogeneity. The measure of centralization of collective bargaining is from Visser (2013), while the Ghent system is measured by a variable taking the value of 1 for the sample countries with a Ghent system; namely, Denmark, Finland, Sweden, and Belgium (the latter has a joint system, where funds are distributed by both the government and labour unions). Finally, we also control for the five-year lagged own unemployment rate in order to control for cyclical factors. We estimate equation (1) in levels with country- and year-fixed effects using instrumental variable estimation. Next, we test the robustness of the result by adding additional control variables that could influence both union density and inequality; namely, the political orientation of the government and the education level of the workforce. Table 9 reports the first-stage results and Table 10 presents the second-stage results. The first-stage estimates indicate that most of our excluded instruments have a significant effect on union density. Employment shares in industry and services both increase significantly union density relative to the employment share in agriculture, but their effect is not significantly different from each other. This likely reflects the fact that some service subsectors—such as the public sector—are highly unionized. The estimates also confirm that having a high level of centralization of collective bargaining, or a Ghent system of unemployment benefits, helps preserve unionization rates significantly in periods of high unemployment. Among control variables, technological progress is found to be significantly associated with a decline in union density, albeit less significantly so when controlling for education. Financial liberalization is also found to reduce union density significantly, possibly supporting the hypothesis that financialization of the economy has reduced the bargaining power of labour to the benefit of capital owners. The interpretation of the coefficient on the globalization variable—measured as China’s share in world exports interacted by initial income level—is less clear: it suggests that the effect of the Chinese shock on union density has been more positive (or less negative) for higher income countries. The education variable has a significantly positive coefficient, contrary to expectations. Other variables (top tax rate and political orientation of the government) are not significant. Table 9 IV regressions (first stage) Dependent variable: . union density . Additional controls: . . political factors . political factors and higher educ. . . (1) . (2) . (3) . ln(ICT K) −0.025** −0.020* −0.018 (−2.17) (−1.93) (−1.52) ln (lag Y/N) −0.061 −0.109** −0.110** (−1.22) (−2.01) (−2.00) ln (lag Y/N) * China X 0.084*** 0.097*** 0.100*** (4.32) (5.05) (5.12) top tax 0.007 0.014 0.015 (0.35) (0.62) (0.68) ln(fin. reform) −0.083*** −0.074*** −0.068*** (−4.62) (−4.33) (−4.07) political right 0.006 0.010 (0.69) (1.13) political left 0.010 0.013 (1.18) (1.58) higher educ. 0.002** (2.13) L5. OECD unempl. * Ghent 0.009*** 0.007*** 0.009*** (3.53) (2.79) (3.31) L5. OECD unempl. * cent.coll.barg. 0.002*** 0.002*** 0.002*** (3.71) (3.92) (3.75) L5. unempl. −0.000 −0.000 −0.000 (−0.03) (−0.73) (−0.51) share ind. emp. 0.011*** 0.012*** 0.011*** (3.98) (4.50) (4.30) share serv. emp. 0.011*** 0.014*** 0.013*** (3.80) (4.61) (4.44) Observations 471 438 438 R-squared 0.629 0.627 0.629 Countries 18 17 17 country FE YES YES YES time FE YES YES YES Dependent variable: . union density . Additional controls: . . political factors . political factors and higher educ. . . (1) . (2) . (3) . ln(ICT K) −0.025** −0.020* −0.018 (−2.17) (−1.93) (−1.52) ln (lag Y/N) −0.061 −0.109** −0.110** (−1.22) (−2.01) (−2.00) ln (lag Y/N) * China X 0.084*** 0.097*** 0.100*** (4.32) (5.05) (5.12) top tax 0.007 0.014 0.015 (0.35) (0.62) (0.68) ln(fin. reform) −0.083*** −0.074*** −0.068*** (−4.62) (−4.33) (−4.07) political right 0.006 0.010 (0.69) (1.13) political left 0.010 0.013 (1.18) (1.58) higher educ. 0.002** (2.13) L5. OECD unempl. * Ghent 0.009*** 0.007*** 0.009*** (3.53) (2.79) (3.31) L5. OECD unempl. * cent.coll.barg. 0.002*** 0.002*** 0.002*** (3.71) (3.92) (3.75) L5. unempl. −0.000 −0.000 −0.000 (−0.03) (−0.73) (−0.51) share ind. emp. 0.011*** 0.012*** 0.011*** (3.98) (4.50) (4.30) share serv. emp. 0.011*** 0.014*** 0.013*** (3.80) (4.61) (4.44) Observations 471 438 438 R-squared 0.629 0.627 0.629 Countries 18 17 17 country FE YES YES YES time FE YES YES YES Notes: Robust z-statistics in parentheses; *** p < 0.01, ** p < 0.05, * p < 0.1. Source: Authors’ estimates. Open in new tab Table 9 IV regressions (first stage) Dependent variable: . union density . Additional controls: . . political factors . political factors and higher educ. . . (1) . (2) . (3) . ln(ICT K) −0.025** −0.020* −0.018 (−2.17) (−1.93) (−1.52) ln (lag Y/N) −0.061 −0.109** −0.110** (−1.22) (−2.01) (−2.00) ln (lag Y/N) * China X 0.084*** 0.097*** 0.100*** (4.32) (5.05) (5.12) top tax 0.007 0.014 0.015 (0.35) (0.62) (0.68) ln(fin. reform) −0.083*** −0.074*** −0.068*** (−4.62) (−4.33) (−4.07) political right 0.006 0.010 (0.69) (1.13) political left 0.010 0.013 (1.18) (1.58) higher educ. 0.002** (2.13) L5. OECD unempl. * Ghent 0.009*** 0.007*** 0.009*** (3.53) (2.79) (3.31) L5. OECD unempl. * cent.coll.barg. 0.002*** 0.002*** 0.002*** (3.71) (3.92) (3.75) L5. unempl. −0.000 −0.000 −0.000 (−0.03) (−0.73) (−0.51) share ind. emp. 0.011*** 0.012*** 0.011*** (3.98) (4.50) (4.30) share serv. emp. 0.011*** 0.014*** 0.013*** (3.80) (4.61) (4.44) Observations 471 438 438 R-squared 0.629 0.627 0.629 Countries 18 17 17 country FE YES YES YES time FE YES YES YES Dependent variable: . union density . Additional controls: . . political factors . political factors and higher educ. . . (1) . (2) . (3) . ln(ICT K) −0.025** −0.020* −0.018 (−2.17) (−1.93) (−1.52) ln (lag Y/N) −0.061 −0.109** −0.110** (−1.22) (−2.01) (−2.00) ln (lag Y/N) * China X 0.084*** 0.097*** 0.100*** (4.32) (5.05) (5.12) top tax 0.007 0.014 0.015 (0.35) (0.62) (0.68) ln(fin. reform) −0.083*** −0.074*** −0.068*** (−4.62) (−4.33) (−4.07) political right 0.006 0.010 (0.69) (1.13) political left 0.010 0.013 (1.18) (1.58) higher educ. 0.002** (2.13) L5. OECD unempl. * Ghent 0.009*** 0.007*** 0.009*** (3.53) (2.79) (3.31) L5. OECD unempl. * cent.coll.barg. 0.002*** 0.002*** 0.002*** (3.71) (3.92) (3.75) L5. unempl. −0.000 −0.000 −0.000 (−0.03) (−0.73) (−0.51) share ind. emp. 0.011*** 0.012*** 0.011*** (3.98) (4.50) (4.30) share serv. emp. 0.011*** 0.014*** 0.013*** (3.80) (4.61) (4.44) Observations 471 438 438 R-squared 0.629 0.627 0.629 Countries 18 17 17 country FE YES YES YES time FE YES YES YES Notes: Robust z-statistics in parentheses; *** p < 0.01, ** p < 0.05, * p < 0.1. Source: Authors’ estimates. Open in new tab Table 10 IV regressions (second stage) Dependent variable (measure of inequality): . ln(Top 10%) . Additional controls: . . political factors . political factors and higher educ. . . (1) . (2) . (3) . union density −0.41*** −0.47*** −0.50*** (−3.101) (−3.683) (−3.964) ln(ICT K) 0.05*** 0.05*** 0.05*** (3.293) (3.090) (3.490) ln (lag Y/N) 0.03 0.01 −0.00 (0.470) (0.242) (-0.050) ln (lag Y/N)* China X 0.18*** 0.18*** 0.19*** (5.516) (5.821) (5.876) top tax −0.12*** −0.10*** −0.09*** (−4.568) (−3.219) (−3.187) ln(fin. reform) 0.08*** 0.03 0.04* (3.069) (1.339) (1.665) political right 0.03*** 0.04*** (2.926) (3.534) political left 0.03** 0.04*** (2.490) (3.058) higher educ. 0.00** (2.421) Observations 471 438 438 R-squared 0.690 0.681 0.687 Number of countries 18 17 17 country FE YES YES YES time FE YES YES YES Excluded instruments L5. OECD unempl. * Ghent, L5. OECD unempl. * cent.coll. barg., L5. unempl., share ind. emp., share serv. emp. Underidentification test (H0: underidentified, p-value) 1.24e-08 6.78e-08 4.85e-08 Kleibergen-Paap rk Wald (F-statistic) 12.57 11.96 12.59 Hansen J-stat (H0: instruments are valid, p-value) 0.304 0.393 0.488 Dependent variable (measure of inequality): . ln(Top 10%) . Additional controls: . . political factors . political factors and higher educ. . . (1) . (2) . (3) . union density −0.41*** −0.47*** −0.50*** (−3.101) (−3.683) (−3.964) ln(ICT K) 0.05*** 0.05*** 0.05*** (3.293) (3.090) (3.490) ln (lag Y/N) 0.03 0.01 −0.00 (0.470) (0.242) (-0.050) ln (lag Y/N)* China X 0.18*** 0.18*** 0.19*** (5.516) (5.821) (5.876) top tax −0.12*** −0.10*** −0.09*** (−4.568) (−3.219) (−3.187) ln(fin. reform) 0.08*** 0.03 0.04* (3.069) (1.339) (1.665) political right 0.03*** 0.04*** (2.926) (3.534) political left 0.03** 0.04*** (2.490) (3.058) higher educ. 0.00** (2.421) Observations 471 438 438 R-squared 0.690 0.681 0.687 Number of countries 18 17 17 country FE YES YES YES time FE YES YES YES Excluded instruments L5. OECD unempl. * Ghent, L5. OECD unempl. * cent.coll. barg., L5. unempl., share ind. emp., share serv. emp. Underidentification test (H0: underidentified, p-value) 1.24e-08 6.78e-08 4.85e-08 Kleibergen-Paap rk Wald (F-statistic) 12.57 11.96 12.59 Hansen J-stat (H0: instruments are valid, p-value) 0.304 0.393 0.488 Notes: Robust z-statistics in parentheses; *** p < 0.01, ** p < 0.05, * p < 0.1. Source: Authors’ estimates. Open in new tab Table 10 IV regressions (second stage) Dependent variable (measure of inequality): . ln(Top 10%) . Additional controls: . . political factors . political factors and higher educ. . . (1) . (2) . (3) . union density −0.41*** −0.47*** −0.50*** (−3.101) (−3.683) (−3.964) ln(ICT K) 0.05*** 0.05*** 0.05*** (3.293) (3.090) (3.490) ln (lag Y/N) 0.03 0.01 −0.00 (0.470) (0.242) (-0.050) ln (lag Y/N)* China X 0.18*** 0.18*** 0.19*** (5.516) (5.821) (5.876) top tax −0.12*** −0.10*** −0.09*** (−4.568) (−3.219) (−3.187) ln(fin. reform) 0.08*** 0.03 0.04* (3.069) (1.339) (1.665) political right 0.03*** 0.04*** (2.926) (3.534) political left 0.03** 0.04*** (2.490) (3.058) higher educ. 0.00** (2.421) Observations 471 438 438 R-squared 0.690 0.681 0.687 Number of countries 18 17 17 country FE YES YES YES time FE YES YES YES Excluded instruments L5. OECD unempl. * Ghent, L5. OECD unempl. * cent.coll. barg., L5. unempl., share ind. emp., share serv. emp. Underidentification test (H0: underidentified, p-value) 1.24e-08 6.78e-08 4.85e-08 Kleibergen-Paap rk Wald (F-statistic) 12.57 11.96 12.59 Hansen J-stat (H0: instruments are valid, p-value) 0.304 0.393 0.488 Dependent variable (measure of inequality): . ln(Top 10%) . Additional controls: . . political factors . political factors and higher educ. . . (1) . (2) . (3) . union density −0.41*** −0.47*** −0.50*** (−3.101) (−3.683) (−3.964) ln(ICT K) 0.05*** 0.05*** 0.05*** (3.293) (3.090) (3.490) ln (lag Y/N) 0.03 0.01 −0.00 (0.470) (0.242) (-0.050) ln (lag Y/N)* China X 0.18*** 0.18*** 0.19*** (5.516) (5.821) (5.876) top tax −0.12*** −0.10*** −0.09*** (−4.568) (−3.219) (−3.187) ln(fin. reform) 0.08*** 0.03 0.04* (3.069) (1.339) (1.665) political right 0.03*** 0.04*** (2.926) (3.534) political left 0.03** 0.04*** (2.490) (3.058) higher educ. 0.00** (2.421) Observations 471 438 438 R-squared 0.690 0.681 0.687 Number of countries 18 17 17 country FE YES YES YES time FE YES YES YES Excluded instruments L5. OECD unempl. * Ghent, L5. OECD unempl. * cent.coll. barg., L5. unempl., share ind. emp., share serv. emp. Underidentification test (H0: underidentified, p-value) 1.24e-08 6.78e-08 4.85e-08 Kleibergen-Paap rk Wald (F-statistic) 12.57 11.96 12.59 Hansen J-stat (H0: instruments are valid, p-value) 0.304 0.393 0.488 Notes: Robust z-statistics in parentheses; *** p < 0.01, ** p < 0.05, * p < 0.1. Source: Authors’ estimates. Open in new tab The instruments satisfy tests of validity. Statistical tests confirm that instruments are strongly and positively correlated with union density. The under-identification test rejects the null hypothesis that the equation is under-identified at the 1% level (see Table 10). In addition, tests also suggest that the instruments are not weak (the Kleibergen-Paap rk Wald F statistic is 12.57). The Hansen statistic fails to reject the null hypothesis that the instruments are uncorrelated with the error term. Turning to the second stage, the coefficient of union density in the top 10% income share regression remains similar in magnitude to the baseline estimates and statistically significant at the 1% level, suggesting a largely causal effect of union density on top income shares. In addition, combined with the first-stage estimates, our result suggests that some of the factors we examine in this paper and which have been highlighted as important determinants of inequality in the literature—such as technological progress and financial liberalization—indirectly increased top income shares through their negative effect on union density. The results suggest that the influence of union density on the top 10% income shares is largely causal. Moreover, while the decline in union density may have been partly driven by factors that also increase inequality, the evidence suggests that de-unionization has an independent effect on the rise in the top 10% income shares that is economically significant. 5. Concluding remarks Our analysis shows that the rise of inequality in most advanced economies has been driven by the upper part of the income distribution, owing largely to the increase in income shares of the top 10% earners. We find strong evidence that, in addition to traditional explanations related to higher inequality (such as globalization, technological progress, financial deregulation and lower top marginal tax rates), the erosion of labour market institutions in the advanced economies plays an important role. In particular, our results show that the weakening of unions contributed significantly to the rise of top earners’ income shares. We also find some evidence that weaker unions may result in less income redistribution and that eroding minimum wages increase the Gini of gross income. The strong negative link we find between unions and top income shares across advanced economies accords with the growing perception that labour market institutions, such as trade unions, are not only relevant for low- and middle-income earners. If de-unionization restrains earnings at the middle and bottom of the distribution, income shares of top earners necessarily rise. Other mechanisms include the positive effect of weaker unions on the share of capital income—which tends to be more concentrated than labour income—and the fact that lower union density may reduce workers’ influence on corporate decisions, including those related to top executive compensation. We plan to test the importance of these channels in future work. Another important area for future research is assessing the welfare implications brought about by the weakening of labour market institutions. While the rise in top earners’ income shares could reflect a relative increase in their productivity (good inequality), top earners’ compensation may be larger than is justified by their contribution to the economy’s output, reflecting rent extraction (bad inequality). In the latter case, there would be grounds for policy action. However, such an analysis should also take account of possible trade-offs with policy objectives other than reducing inequality, such as macroeconomic stability, competitiveness, growth, and employment. Supplementary material Supplementary material is available on the OUP website. These are the data and replication files and the online Appendix. Acknowledgments We thank Olivier Blanchard, Rupa Duttagupta, Romain Duval, Hamid Faruqee, Prakash Loungani, Suresh Naidu, Maurice Obstfeld, Jonathan Ostry, Emil Stavrev, Sebastian Weber, and colleagues in the International Monetary Fund (IMF) departments for their insightful comments and suggestions. We are extremely grateful to Chanpheng Fizzarotti and Ava Yeabin Hong, for their superb and invaluable research assistance, and to Gabi Ionescu, for editorial assistance. The views expressed in this paper are those of the authors and do not necessarily represent the views of the IMF, its Executive Board, or IMF management. All remaining errors are ours. Footnotes 1 It should be noted that Card and DiNardo (2002) find that skilled-bias technological change (SBTC) cannot possibly be the only cause behind the rising US wage inequality in the 1980s and 1990s. 2 This problem is largely explained by under-coverage of highest income earners and top coding. 3 A few studies investigated the role of the wage bargaining process. Most, but not all, studies find that centralized wage setting also narrows the distribution of wages (OECD, 2012). 4 In comparative welfare-state research, theories link class-based political power with income distribution. 5 Similarly, Piketty and Saez (2006) and Fräßdorf et al. (2008) show that the share of capital income is larger at the top of the distribution. Moreover, since the 1980s, the concentration of capital income has increased more than that of labor income in 75% of OECD countries (OECD, 2012). 6 Lemieux et al. (2009) find that the increased prevalence of performance-related pay, which may have been partly facilitated by de-unionization, explains one-quarter of the rise of wage inequality among men in the United States and nearly all of the top-end growth in wage dispersion. 7 We constructed a Gini coefficient for the bottom 90% of the population based on the Luxembourg Income Study data on net income shares. The results indicate that net income inequality below the 90th percentile has not risen since the mid-1990s. This implies that the Gini of gross income below the 90th percentile likely rose and, thus, that our counterfactual estimate of the evolution of the Gini since 1995 is a lower-bound estimate. These calculations are available on request. 8 Similarly, Atkinson et al. (2009) argue that US household survey data, which does not measure top incomes, fail to capture approximately half of the increase in overall inequality. 9 The summary statistics reported for inequality’s determinants are based on the largest sample, for which Gini index data are available. Summary statistics based on a shorter sample, covering the countries and years for which the top 10% income shares data are available, are reported in the online Appendix. 10 We show in robustness tests that the ratio of the minimum wage to the median wage is not a robust determinant of the top income share. 11 Technology is measured by the capital stock share of information and communications technology, which is the key general-purpose technological innovation over the sample period considered. The share of China in world exports proxies for the growing competition from low-cost foreign labor. Its effect is allowed to vary as a function of the income level of the country. The financial reform index is available to 2005. For 2006–2010, the financial reform index is assumed to remain constant at its 2005 value. Preliminary work on an update of the database suggests that this is a reasonable assumption for the most relevant components of the index. 12 The size of the sample changes; this reflects the data availability of inequality measures and explanatory variables. The sample size used for each estimation maximizes the number of observations, given the available data. The main results are robust to using the same sample across inequality measures (see the online Appendix), as well as excluding the global financial crisis period and estimating over the period 1981–2005. 13 Similar tests for the PMG and dynamic FE models also fail to reject the null hypothesis of weak cross-sectional dependence. 14 Although not reported here, we also confirmed the robustness of the effect of union density on the top 10% income shares to additional control variables, including: (i) demographic variables (female participation rate, immigration rate, population aging); (ii) public indebtedness that may have forced governments to cut social spending; (iii) the start of the European Monetary Union, which may have limited government’s scope for discretionary policies; (iv) the ratio of private sector credit to GDP; (v) traditional measures of trade openness; and (vi) a measure of capital account liberalization. 15 Social preferences regarding inequality are measured by the mean country sentiment to the statement ‘We need larger income differences as incentives for individual effort,’ from the World Values Surveys and European Values Surveys. 16 A large body of research has established that de-industrialization has lowered unionization. 17 The empirical finance literature, however, finds a negative link between broad proxies for financial development (such as private sector credit to GDP) and inequality measured by the Gini coefficient (see Beck et al., 2007). 18 We are thankful to Suresh Naidu for suggesting these instruments. References Abiad A. , Detragiache E. , Tressel T. ( 2008 ) A new database of financial reforms. IMF, Working Paper, No. 08/266, Washington DC. Acemoglu D. , Aghion P. , Violante G.L. ( 2001 ) Deunionisation, technical change and inequality, Carnegie-Rochester Conference Series on Public Policy , 55 , 229 – 64 . Google Scholar Crossref Search ADS WorldCat Acemoglu D. , Robinson J.A. ( 2013 ) Economics versus politics: pitfalls of policy advice, Journal of Economic Perspectives , 27 , 173 – 92 . Google Scholar Crossref Search ADS WorldCat Aghion P. , Caroli E. , García-Peñalosa C. ( 1999 ) Inequality and economic growth: the perspective of the new growth theories, Journal of Economic Literature , 37 , 1615 – 60 . Google Scholar Crossref Search ADS WorldCat Alesina A. , Rodrik D. ( 1994 ) Distributive politics and economic growth, The Quarterly Journal of Economics , 109 , 465 – 90 . Google Scholar Crossref Search ADS WorldCat Alvaredo F. ( 2011 ) A note on the relationship between top income shares and the Gini coefficient, Economics Letters , 110 , 274 – 7 . Google Scholar Crossref Search ADS WorldCat Alvaredo F. , Atkinson A.B. , Piketty T. , Saez E. ( 2013 ) The top 1 percent in international and historical perspective, Journal of Economic Perspectives , 27 , 3 – 20 . Google Scholar Crossref Search ADS WorldCat Alvaredo F. , Atkinson A.B. , Piketty T. , Saez E. ( 2015 ) The world top incomes database, http://topincomes.parisschoolofeconomics.eu/ (last accessed 11 March 2015). Atkinson A.B. , Piketty T. , Saez E. ( 2009 ) Top incomes in the long run of history. NBER, Working Paper, No. w15408, Cambridge, MA. Baker D. , Glyn A. , Howell D.R. , Schmitt J. ( 2004 ) Labor market institutions and unemployment: assessment of the cross-country evidence, in Howell D.R. (ed.) Fighting Unemployment: The Limits of Free Market Orthodoxy , Oxford University Press , New York, NY . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Beck T. , Demirguc-Kunt A. , Levine R. ( 2007 ) Finance, inequality, and the poor, Journal of Economic Growth , 12 , 27 – 49 . Google Scholar Crossref Search ADS WorldCat Benabou R. ( 1996 ) Inequality and growth, NBER Macroeconomics Annual , 11 , 11 – 74 . Google Scholar Crossref Search ADS WorldCat Berg A. , Ostry J. , Zettelmeyer J. ( 2012 ) What makes growth sustained? Journal of Development Economics , 98 , 149 – 66 . Google Scholar Crossref Search ADS WorldCat Betcherman G. ( 2012 ) Labor market institutions: a review of the literature. World Bank Policy Research, Working Paper, No. 6276, Washington, DC. Bivens J. , Mishel L. ( 2013 ) The pay of corporate executives and financial professionals as evidence of rents in top 1 percent incomes, The Journal of Economic Perspectives , 27 , 57 – 77 . Google Scholar Crossref Search ADS WorldCat Bouis R. , Causa O. , Demmou L. , Duval R. , Zdzienicka A. ( 2012 ) The short-term effects of structural reforms: an empirical analysis. OECD Economics Department, Working Papers, No. 949, Paris. Bound J. , Krueger A.B. ( 1991 ) The extent of measurement error in longitudinal earnings data: Do two wrongs make a right? Journal of Labor Economics , 9 , 1 – 24 . Google Scholar Crossref Search ADS WorldCat Burkhauser R.V. , Feng S. , Jenkins S.P. , Larrimore J. ( 2012 ) Recent trends in top income shares in the United States: reconciling estimates from March CPS and IRS tax return data, Review of Economics and Statistics , 94 , 371 – 88 . Google Scholar Crossref Search ADS WorldCat Burtless G. ( 2003 ) Has widening inequality promoted or retarded U.S. Growth? Canadian Public Policy. Analyse de Politiques , XXIX, S185 – 201 . OpenURL Placeholder Text WorldCat Card D. ( 1996 ) The effect of unions on the structure of wages: a longitudinal analysis, Econometrica: Journal of the Econometric Society , 64 , 957 – 79 . Google Scholar Crossref Search ADS WorldCat Card D. ( 2001 ) The effect of unions on wage inequality in the US labor market, Industrial and Labor Relations Review , 54 , 296 – 315 . Google Scholar Crossref Search ADS WorldCat Card D. , DiNardo J.E. ( 2002 ) Skill-based technological change and rising wage inequality: some problems and puzzles, Journal of Labor Economics , 20 , 733 – 83 . Google Scholar Crossref Search ADS WorldCat Card D. , Lemieux T. , Riddell W.C. ( 2004 ) Unions and wage inequality, Journal of Labor Research , 25 , 519 – 59 . Google Scholar Crossref Search ADS WorldCat Dew-Becker I. , Gordon R.J. ( 2005 ) Where did the productivity growth go? Inflation dynamics and the distribution of income. NBER, Working Paper, No. w11842, Cambridge, MA. DiNardo J. , Fortin N. , Lemieux T. ( 1996 ) Labor market institutions and the distribution of wages, 1973–1992: a semiparametric approach, Econometrica , 64 , 1001 – 44 . Google Scholar Crossref Search ADS WorldCat DiNardo J. , Hallock K. , Pischke J.S. ( 1997 ) Unions and managerial pay. NBER, Working Paper, No. w6318, Cambridge, MA. Duenhaupt P. ( 2012 ) Financialization and the rentier income share–evidence from the USA and Germany, International Review of Applied Economics , 26 , 465 – 87 . Google Scholar Crossref Search ADS WorldCat Fräßdorf A. , Grabka M. , Schwarze J. ( 2008 ) The impact of household capital income on income inequality: a factor decomposition analysis for Great Britain, Germany and the USA. Society for the Study of Economic Inequality, Working Paper, No. 89, Berlin. Freeman R.B. ( 2000 ) Single peaked vs. diversified capitalism: the relation between economic institutions and outcomes. NBER, Working Paper, No. w7556, Cambridge, MA. Gabaix X. , Landier A. ( 2008 ) Why has CEO pay increased so much? Quarterly Journal of Economics , 123 , 49 – 100 . Google Scholar Crossref Search ADS WorldCat Gomez R. , Tzioumis K. (2006, 2011 ) What do unions do to executive compensation? CEP, Discussion Paper, No. CEPDP0720, London. Howell D.R. , Baker D. , Glyn A. , Schmitt J. ( 2007 ) Are protective labor market institutions at the root of unemployment? A critical review of the evidence , Capitalism and Society , 2 , 1 – 73 . Google Scholar Crossref Search ADS WorldCat Im K.S. , Pesaran M.H. , Shin Y. ( 2003 ) Testing for unit roots in heterogeneous panels, Journal of Econometrics , 115 , 53 – 74 . Google Scholar Crossref Search ADS WorldCat Jensen M.C. , Murphy K.J. ( 1990 ) Performance pay and top-management incentives Journal of political economy , 98 , 225 – 64 . Google Scholar Crossref Search ADS WorldCat Kierzenkowski R. , Koske I. ( 2012 ) Less income inequality and more growth—are they compatible? Part 8, the drivers of labor income inequality—a literature review. OECD Economics Department, Working Papers, No. 931, Paris. Korpi W. ( 2006 ) Power resources and employer-centered approaches in explanations of welfare states and varieties of capitalism, World Politics , 58 , 167 – 206 . Google Scholar Crossref Search ADS WorldCat Layard R. , Nickell S. , Jackman R. ( 1991 ) Unemployment: Macroeconomic Performance and the Labor Market , Oxford University Press , Oxford . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Lee D.S. ( 1999 ) Wage inequality in the United States during the 1980s: rising dispersion or falling minimum wage? Quarterly Journal of Economics , 114 , 977 – 1023 . Google Scholar Crossref Search ADS WorldCat Lemieux T. , MacLeod W.B. , Parent D. ( 2009 ) Performance pay and wage inequality, Quarterly Journal of Economics , 124 , 1 – 49 . Google Scholar Crossref Search ADS WorldCat McCall L. , Percheski C. ( 2010 ) Income inequality: new trends and research directions, Annual Review of Sociology , 36 , 329 – 47 . Google Scholar Crossref Search ADS WorldCat Organisation for Economic Co-operation and Development (OECD). ( 2006 ) OECD Employment Outlook , OECD Publishing , Paris . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Organisation for Economic Co-operation and Development (OECD). ( 2012 ) OECD Employment Outlook , OECD Publishing , Paris . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Ostry J.D. , Berg A. , Tsangarides C.G. ( 2014 ) Redistribution, inequality and growth. IMF, Staff Discussion Note, No. 14/02, Washington, DC. Pedroni P. ( 1999 ) Critical values for cointegration tests in heterogeneous panels with multiple regressors, Oxford Bulletin of Economics and Statistics , 61 , 653 – 70 . Google Scholar Crossref Search ADS WorldCat Pedroni P. ( 2004 ) Panel cointegration: asymptotic and finite sample properties of pooled time series tests with an application to the PPP hypothesis, Econometric Theory , 20 , 597 – 625 . Google Scholar Crossref Search ADS WorldCat Perotti R. ( 1992 ) Fiscal policy, income distribution, and growth. Columbia University, Working Paper, No. 636, New York, NY. Perotti R. ( 1993 ) Political equilibrium, income distribution, and growth , Review of Economic Studies , 60 , 755 – 76 . Google Scholar Crossref Search ADS WorldCat Perotti R. ( 1996 ) Growth, income distribution, and democracy: what the data say, Journal of Economic Growth , 1 , 149 – 87 . Google Scholar Crossref Search ADS WorldCat Persson T. , Tabellini G. ( 1994 ) Is inequality harmful for growth? American Economic Review , 84 , 600 – 21 . OpenURL Placeholder Text WorldCat Pesaran M.H. ( 2006 ) Estimation and inference in large heterogeneous panels with a multifactor error structure, Econometrica , 74 , 967 – 1012 . Google Scholar Crossref Search ADS WorldCat Pesaran M.H. ( 2007 ) A simple panel unit root test in the presence of cross‐section dependence, Journal of Applied Econometrics , 22 , 265 – 312 . Google Scholar Crossref Search ADS WorldCat Pesaran M.H. ( 2015 ) Testing weak cross-sectional dependence in large panels, Econometric Reviews , 34 , 1089 – 117 . Google Scholar Crossref Search ADS WorldCat Pesaran M.H. , Shin Y. , Smith R.P. ( 1999 ) Pooled mean group estimation of dynamic heterogeneous panels, Journal of the American Statistical Association , 94 , 621 – 34 . Google Scholar Crossref Search ADS WorldCat Pesaran M.H. , Smith R. ( 1995 ) Estimating long-run relationships from dynamic heterogeneous panels, Journal of Econometrics , 68 , 79 – 113 . Google Scholar Crossref Search ADS WorldCat Philippon T. , Reshef A. ( 2012 ) Capital in the US financial industry: 1909–2006, The Quarterly Journal of Economics , 127 , 1551 – 609 . Google Scholar Crossref Search ADS WorldCat Philippon T. , Reshef A. ( 2013 ) An international look at the growth of modern finance, The Journal of Economic Perspectives , 27 , 73 – 96 . Google Scholar Crossref Search ADS WorldCat Piketty T. , Saez E. ( 2006 ) The evolution of top incomes: a historical and international perspective, American Economic Review , 96 , 200 – 5 . Google Scholar Crossref Search ADS WorldCat Piketty T. , Stantcheva S. ( 2014 ) Optimal taxation of top labor incomes: a tale of three elasticities, American Economic Journal: Economic Policy , 6 , 230 – 71 . Google Scholar Crossref Search ADS WorldCat Pissarides C.A. ( 1990 ) Equilibrium Unemployment Theory , MIT Press , Cambridge, MA . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Scheve K. , Stasavage D. ( 2009 ) Institutions, partisanship, and inequality in the long run, World Politics , 61 , 215 – 53 . Google Scholar Crossref Search ADS WorldCat Sjöberg O. ( 2009 ) Corporate governance and earnings inequality in the OECD countries 1979–2000, European Sociological Review , 25 , 519 – 33 . Google Scholar Crossref Search ADS WorldCat Solt F. ( 2009 ) Standardizing the World Income Inequality Database, Social Science Quarterly , 90 , 231 – 42 . Google Scholar Crossref Search ADS WorldCat Stiglitz J. ( 2012 ) The Price of Inequality: How Today’s Divided Society Endangers Our Future , W.W. Norton , New York, NY . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Temple J. ( 1999 ) The new growth evidence, Journal of Economic Literature , 37 , 112 – 56 . Google Scholar Crossref Search ADS WorldCat Temple J. ( 2005 ) Growth and wage inequality, Bulletin of Economic Research , 57 , 145 – 69 . Google Scholar Crossref Search ADS WorldCat Teulings C.N. ( 2003 ) The contribution of minimum wages to increasing wage inequality, The Economic Journal , 113 , 801 – 33 . Google Scholar Crossref Search ADS WorldCat Visser J. ( 2013 ) ICTWSS: Database on Institutional Characteristics of Trade Unions, Wage Setting, State Intervention and Social Pacts in 34 Countries between 1960 and 2007, Amsterdam Institute for Advanced Labor Studies , University of Amsterdam , Amsterdam . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Volscho T.W. , Kelly N.J. ( 2012 ) The rise of the super-rich: power resources, taxes, financial markets, and the dynamics of the top 1 percent, 1949 to 2008, American Sociological Review , 77 , 679 – 99 . Google Scholar Crossref Search ADS WorldCat Wallerstein M. , Western B. ( 2000 ) Unions in decline? What has changed and why, Annual Review of Political Science , 3 , 355 – 77 . Google Scholar Crossref Search ADS WorldCat © Oxford University Press 2019. 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/open_access/funder_policies/chorus/standard_publication_model)
TI - Inequality: traditional drivers and the role of union power
JF - Oxford Economic Papers
DO - 10.1093/oep/gpz024
DA - 2020-01-01
UR - https://www.deepdyve.com/lp/oxford-university-press/inequality-traditional-drivers-and-the-role-of-union-power-064r8keR7M
SP - 25
VL - 72
IS - 1
DP - DeepDyve
ER -