Recent research shows that financial activities expand the wage gap between affluent and middle-class workers in advanced industrial societies. Furthermore, a well-established literature indicates that differences in labor institutions may be responsible for cross-national variations of income inequality in developed countries. Surprisingly, there is no empirical research examining whether the positive association between financialization and income inequality is conditioned by differences in wage coordination. We contribute to this comparative income inequality literature by testing the claim that wage-setting institutions suppress the inequality-widening effects of finance in advanced industrial societies. To test this contention, we compile an unbalanced panel dataset of 20 developed economies during the years 1988 to 2009. According to our results, financial activities are a robust positive predictor of 90–50 inequality. More importantly, the interaction of financialization and wage coordination returns significant negative associations with the dependent variable. These results are found to be consistent across different estimation techniques and numerous regression parameters.

## INTRODUCTION

Few scholars would disagree with the contention that income inequality expanded in many advanced industrial societies over the last 40 years. Perhaps more importantly, this recent increase of inequality is largely driven by the growth of wages at the top of the income distribution, while the middle- and working-class continue to fall further behind. Given these trends in the national distribution of earnings, as summarized in Figure 1, numerous explanations were formulated over the years to explain this transformation of social stratification across the developed world. Specifically, while some argue that this growth of income inequality is attributable to deindustrialization and economic globalization, others stress the role of liberalization and skill-biased technological change (Brookman, Chang, and Rennie 2007; Goldstein 2012; Kalleberg 2011; Krippner 2011; Kuttner 1995; Kwon 2016; Mandel 1996; Osterman 2008).

Figure 1.

90–50 Earnings Inequality in Advanced Industrial Societies, 1975 to 2010

Figure 1.

90–50 Earnings Inequality in Advanced Industrial Societies, 1975 to 2010

Surprisingly, it is only somewhat recently that researchers started to explore how financialization impacts distributional patterns in developed economies. Financialization, defined as the socioeconomic transformation in which financial activities become the dominant mode of profit accumulation, represents one of the most important structural transformations in advanced economies (Arrighi 1994; Davis 2009; Krippner 2011). More importantly, a growing field of research shows that the increasing prominence of financial activities in the operation of the domestic economy exacerbates income inequality both within the United States and across a panel of advanced industrial societies (Assa 2012; Goldstein 2012; Kus 2012; Lin and Tomaskovic-Devey 2013; Tomaskovic-Devey and Lin 2011; Volscho and Kelly 2012). That is, many studies indicate that financialization's widening of income inequality is largely due to its tendency to expand the wages of wealthy households and top earners at the direct expense of the middle- and working-class (Bell and van Reenen 2013; Denk and Cournede 2015; Godechot 2012; Kwon and Roberts 2015; Nau 2013).

However, although this work helps expand our understanding of finance's distributional effect, there is much less research on how cross-national differences in labor institutions could dampen financialization's tendency to widen income inequality. This dearth of research is surprising when considering that there exists a robust literature on how differences in labor institutions can shape the national distribution of earnings (Alderson and Nielsen 2002; Mahler 2004; Pontusson, Rueda, and Way 2002; Rueda and Pontusson 2000; Wallerstein 1999). This topic is especially important given that many developed countries across the world experienced a decline of their wage-setting systems over the past few decades (Glyn 2006; Howell 2006; Schneider and Paunescu 2011; Streeck 2009).

Given these observations, we attempt to fill this gap in the literature by examining whether the relationship between financialization and income inequality is conditioned by coordinated and centralized wage-setting institutions. According to the literature, countries with strong wage-setting institutions tend to induce strategic partnerships between employers and labor, generating an environment where firms are more likely to invest in production and expand the wages of workers. In contrast, countries with weak wage-coordination systems lack these strategic partnerships, meaning that firms are more likely to reduce investments in production in favor of a finance-based model of corporate profitability. If these arguments are correct, wage coordination should dampen the inequality-producing effect of financialization (Hall and Gingerich 2009; Hall and Soskice 2001; Thelen 2014).

To test these arguments, we compile an unbalanced panel dataset for 20 advanced industrial societies for the years 1988 to 2009. In particular, we use interaction effects to explore whether the positive association between finance and income inequality is reduced at higher levels of wage coordination. To reiterate, the financialization of a national economy is largely defined by the extent to which financial activities become the principal mode of profit accumulation (Arrighi 1994; Davis 2009; Krippner 2011). Following this definition, we compile an index of financialization that combines various measures of financial activity in the national economy. We also collect information for a well-established measure of wage coordination to interact with financialization, allowing us to test our main hypothesis.

## FINANCIALIZATION, WAGE COORDINATION, AND INCOME INEQUALITY

An emergent field of research indicates that various measures of financial activity are positively associated with income inequality (Assa 2012; Kus 2012; Lin and Tomaskovic-Devey 2013; Tomaskovic-Devey and Lin 2011). Although this research is still ongoing, there are three main observations in this literature. First, financial activities expand the earnings of upper-level management and those in the financial sector while depressing the wages of rank-and-file workers (Goldstein 2012; Tomaskovic-Devey and Lin 2011). Second, finance incentivizes executives to augment returns to shareholders via anti-union activities, mass layoffs, lower wages, and investment in labor-saving technologies (Fligstein and Shin 2007; Kus 2012; Lin and Tomaskovic-Devey 2013). And third, financialization increases the importance of investment income, thereby enlarging the rents received by affluent households and other actors (Nau 2013; Volscho and Kelly 2012). In this way, financial activities shape distributional outcomes by expanding the wages of top earners at the expense of the middle- and working-class (Bell and van Reenen 2013; Denk and Cournede 2015; Godechot 2012; Kwon and Roberts 2015). Accordingly, we hypothesize the following:

H1: Financialization is positively associated with wage inequality.

According to the literature, the rise of finance in developed economies is largely due to the deregulation and reregulation of the financial industry (Vogel 1996; Weiss 2010). And while many focus on these processes of deregulation and reregulation in countries with weak wage-coordination systems, such as the United States (Davis 2009; Osterman 1999; Useem 1993), others stress that financial liberalization also takes place in countries with strong wage-coordination systems, such as Belgium and Finland (Ebbinghaus 2015). This is further supported by Figure 2, which shows that there does not seem to be a clear relationship between a country's level of wage coordination and financialization. Some even suggest using the term “varieties of liberalization” to describe how liberal economic policies are being adopted by all developed economies around the world. However, it is important to note that the specific form of liberalization ultimately adopted tends to be shaped by the labor institutions of these economies (Thelen 2014).

Figure 2.

Box Plot of Financialization and Wage Coordination in Advanced Industrial Societies, 1988 to 2009

Figure 2.

Box Plot of Financialization and Wage Coordination in Advanced Industrial Societies, 1988 to 2009

In light of these observations, how may wage-setting institutions mitigate the link between financialization and income inequality?1 It is important to note that there is currently only one study which examines this question, by exploring whether unionization rates and formal employment protections suppress the connection between finance and inequality (Darcillon 2016). However, there are no studies we are aware of that examine whether wage-setting institutions suppress the inequality-widening effects of financialization. This dearth of research is surprising, considering that wage coordination may be a salient factor in cross-national differences in income inequality (Alderson and Nielsen 2002; Mahler 2004; Pontusson, Rueda, and Way 2002; Rueda and Pontusson 2000; Wallerstein 1999). According to Wallerstein (1999: 673–76), centralized and coordinated wage-setting institutions reduce income inequality through three general channels: economic, political, and ideological. Economically, in countries with decentralized wage bargaining, inefficiencies in wage and employment allocation between unionized and non-unionized sectors are likely to occur due to the asymmetry of labor market power. In countries with centralized wage-setting, this misallocation is less likely because wage agreement might impose wage schedules that approximate the competitive allocation of employment and wages. Politically, centralized wage-setting alters the influence of different groups, namely low-wage labor. Ideologically, centralized wage-setting promotes the egalitarian distribution of wages; the more workers are incorporated into the process, the further this norm of wage egalitarianism will permeate. Given these observations, it may be useful to explore how variations in wage-setting institutions explain cross-national differences in the effect of finance on wage inequality.

A good starting point is the comparative capitalism literature. A major theme in this literature is that differences in labor institutions induce variation in corporate strategy and employment relations, which in turn shape distributional outcomes in the economy (Rueda and Pontusson 2000; Hall and Gingerich 2009; Hall and Soskice 2001). Specifically, the process of wage-setting varies across countries along a general spectrum from, on the one hand, bilateral bargaining between an individual employer and an individual employee to, on the other hand, bargaining between peak union associations, government agencies, and business confederations (Wallerstein 1999: 672). In countries with more centralized wage-setting, wage variation is mitigated through the enactment of sector- or industry-level wage agreements. In contrast, countries with decentralized wage-setting maintain more wage variation, given the asymmetrical bargaining power between employers and employees.

We argue that the inequality-widening effects of financialization are suppressed by wage-setting institutions because these institutions alter corporate strategy and the wage structures of firms and sectors. First, centralized and coordinated wage-setting institutions suppress the effect of financialization on wage inequality by increasing the compensation of rank-and-file workers at the expense of top earners. Second, wage-setting institutions politically empower production-based workers, which restricts the ability of managers and executives to divest from these workers as firms engage in more financial activity. Finally, wage-setting institutions are important for promoting forms of “stakeholder capitalism,” where capital, labor, and the state actively negotiate wage agreements that address the needs and interests of all parties.

The “varieties of capitalism” perspective provides a useful framework for differentiating between types of wage-setting institutions and illustrating how these institutions shape corporate strategy and industrial relations in advanced industrial societies.2 In coordinated market economies (CMEs), economic institutions are designed to promote cooperation and shared governance between capital, labor, and the state. Accordingly, wage-setting in CMEs is conducted by active negotiations between peak associations over sector- or industry-level agreements on issues of pay, working conditions, and even employment levels in one or more industries (Hall and Soskice 2001). As a result, these wage-setting institutions generate dense networks of association that facilitate strategic interactions rather than “pure” market competition among economic actors (Hall and Gingerich 2009). In this environment, executive compensation is suppressed because firms are less likely to reward management for increasing shareholder value through short-term investment and reducing labor costs. Instead, executives are expected to make investments in long-term assets and firm-specific human capital to maintain high levels of productivity (Thelen 2014). Accordingly, non-financial firms in countries with centralized and coordinated wage-setting institutions are more likely to reinvest in labor by raising the wages of the productive labor force, especially since corporate interests are tied to maintaining strategic partnerships with peak labor associations. In this way, higher wage coordination preserves the wages of rank-and-file workers at the expense of managerial and executive compensation, facilitating a more equitable distribution of market resources (Fiss and Zajac 2004; Hall and Gingerich 2009).

An example of this “stakeholder model” of capitalism is Germany, which relies heavily on wage-coordination systems to align decision-making with the interest of peak business associations and union confederations, and restricts the ability of shareholders to dominate corporate strategy (Piketty 2014). Accordingly, the financial system is primarily utilized as a source of capital for long-term investment. Referred to as the Hausbank principle, firms secure financing via state-run banks instead of capital markets, allowing the formation of a highly coordinated system of control that deincentivizes short-term financial investments. In such an environment, upper-level management possesses a higher degree of independence from external investors and is rewarded for expanding the productivity of the labor force. Scholars thus point out that firms in Germany remain committed to the preservation of these coordinated financial systems and stakeholder decision-making despite the expansion of the financial sector (Haves, Vitols, and Wilke 2014). Furthermore, although it is true that executives use the language of shareholder value when interacting with foreign investors, in practice, German firms avoid the short-term investment strategies used by their counterparts in the United States and implement long-term investment strategies that improve the productive capacity of the firm (Fiss and Zajac 2004). Overall, the German case illustrates how firms in countries with centralized and coordinated wage-setting institutions utilize a “stakeholder” corporate strategy despite the ever-growing prominence of the financial sector.

In contrast, liberal market economies (LMEs) induce employers and labor to coordinate economic activities via competitive market processes instead of formalized systems of wage coordination. Given the presence of these weak labor institutions, the relations of production are defined as short-term and arm's-length relationships which necessitate formal contracting, with wages that are largely determined by decentralized bilateral negotiations between individual employers and employees. Accordingly, executives and managers view employees as a variable cost of production rather than stakeholders of the company. This perception of labor induces executives and managers to lower wages to improve the cost-efficiency of production and the short-term valuation of company shares. This “shareholder-value governance” became prominent in firms with the rise of large equity markets, diffusion of shareholding, and weakening of business regulations. As a result, executives and managers were increasingly forced to operate under a condition of heightened financial scrutiny that demands short-term profitability over long-term investment in labor (Hall and Gingerich 2009). In this environment, executives’ compensation is directly tied to their willingness to concentrate on investments that increase firm valuations for the benefit of shareholders, at the expense of rank-and-file workers (Davis 2009; Fligstein and Shin 2007; Goldstein 2012). While the compensation of executives, managers, and financial workers soared, wages for production-based workers declined (Lin and Tomaskovic-Devey 2013; Tomaskovic-Devey and Lin 2011).

The prototypical example of “shareholder capitalism” is the United States. As more businesses became reliant on financial activity for their profits, top executives of non-financial firms were increasingly held to a new performance metric that provided monetary incentives for increasing stock prices rather than firm productivity (Cappelli et al. 1997; Fligstein 2001; Fligstein and Shin 2004). In light of this new compensation structure, corporate leadership engaged in a variety of strategies to reduce production costs and maximize returns for investors, such as reducing the wages of production-based workers (Cappelli 2000; Stearns and Allan 1996; Zuckerman 2000). Thus, over the last 30 years, firms reoriented their investment strategies to boost short-term returns for shareholders at the expense of long-term assets and productivity (Fligstein 1990, 2001; Fligstein and Shin 2007), generating a rise of low-wage, precarious, and contingent labor (Davis 2009; Kalleberg 2011). Accordingly, as the financial sector expanded and become more centralized in the U.S. economy, the wages of rank-and-file workers steadily declined, while executive compensation grew exponentially. Overall, the U.S. case illustrates how firms in countries with decentralized wage-setting institutions increasingly utilize a shareholder corporate strategy with the increasing financialization of economic activity.

What should be clear from this discussion is that wage-setting institutions retain the capacity to moderate the distributional effect of financial activity in advanced industrial societies. Countries with strong wage-coordination systems encourage strategic negotiations between economic actors, inducing a production-based strategy of accumulation that accounts for the needs of all stakeholders. The result is a more mutually beneficial relationship between employers and labor which maintains the wages of middle- and working-class workers while suppressing executive and managerial pay. Furthermore, not only do countries with strong wage coordination incentivize firms to engage in long-term investments in production, highly coordinated labor institutions also provide workers the organizational capacity to engage in strikes, lockouts, and other forms of collective action to shape corporate strategy (Wallerstein 1999). This, of course, is in contrast to countries with weak wage-coordination systems, which lack a formalized mechanism that allows employers and labor to negotiate corporate policy. Firms are consequently more likely to divest from production to pursue a finance-based strategy of accumulation when a country's formal system of wage coordination is weak or virtually nonexistent. All in all, given the observations outlined in this section, we argue that wage coordination should suppress the positive association between financialization and income inequality by expanding the wages of rank-and-file workers and minimizing the incomes of executives, investors, and managers. Based on this theoretical discussion we expect the following:

H2: The association between financialization and wage inequality is suppressed by wage coordination.

## DATA AND METHODS

### Data Sources and Variables

Several cross-national studies of income inequality make use of the popular Gini coefficient (Assa 2012; Kus 2012; Huber and Stephens 2014; Lee, Kim, and Shim 2011). Although this measure is useful for estimating the overall level of income inequality, it tends to be overly sensitive to shifts in the middle of the income distribution, concealing changes in the top and bottom (Atkinson 1975; Hey and Lambert 1980). This characteristic of the Gini coefficient is important since the current investigation is mostly concerned with the growing income gap between top earners and the middle class. Furthermore, the evidence shows that financial activities should be directly connected to the growing difference of income between the top vis-à-vis the middle and/or bottom (Kwon and Roberts 2015; Nau 2013; Tomaskovic-Devey and Lin 2011; Volscho and Kelly 2012).3 In light of these observations, we use 90–50 income inequality as the dependent variable of this investigation.4 This measure is calculated by dividing the pretax earnings of those in the top 90th percentile by the earnings of those in the 50th percentile. This information is from the OECD database (http://www.oecd.org/statistics/).

The first independent variable of interest in the current study is the financialization index. This measure combines three covariates of financial activity that are measured as a percentage of GDP: (1) total value of stocks traded captures the overall trade in the domestic securities market; (2) market capitalization of public firms serves as a potential estimate of the “weight” of finance-based firms in the national economy; and (3) private-sector credit is a proxy for the expansion of domestic credit availability. These measures are used by many other cross-national studies of financialization (e.g. Darcillon 2016; Kus 2012; Kwon and Roberts 2015). As in previous research, each measure is z-score standardized and summed to generate an overall index of financialization following the formula:

$financializationit=Sit−X¯SσS+Mit−X¯MσM+Dit−X¯DσD$

where i denotes the country, t the observed year, $X¯$ is the mean, S represents stocks traded, M is the market capitalization of firms, and D is domestic credit. We perform a principal components analysis to ensure the reliability of this index. This analysis shows that the three variables tested produce a single factor solution with an eigenvalue of 2.68. This first factor returned loadings of .92 (stocks traded), .69 (domestic credit), and .90 (market capitalization of traded companies). Furthermore, the Cronbach's alpha is within an acceptable range, at .75. The data necessary to construct this index are from the aforementioned OECD database.

The next independent variable of interest is wage coordination (Kenworthy 2001). This measure estimates the extent to which wage-setting is centralized through peak associations and whether agreements are modeled in similar industries and sectors of the domestic economy. This covariate is measured on a five-point ordinal scale, with higher scores denoting more coordination. Specifically, the category of 1 is given to those countries that maintain a fragmented wage-bargaining system with agreements that are confined to individual firms, while the category of 5 is reserved for those countries with wage bargaining that occurs through peak associations as well as a centralized industry-level bargaining system coordinated by a union confederation. This information is from the ICTWSS database (http://www.uva-aias.net/nl/ictwss/). More details on the coding of this measure can be seen in Appendix A.

There are three sets of control variables included in the regression models. The first group of controls are closely associated with the effect of economic globalization on inequality. Southern import penetration is the inflow of manufactured goods (SITC Rev 1. 5–8) from developing countries as a proportion of GDP. This information is from UN Comtrade (http://comtrade.un.org/db/). Foreign direct investment outflow calculates the total outward flow of direct investment divided by GDP. This variable is from UNCTAD (http://unctad.org). Net migration is the inflow of migrants from foreign countries; these data are from the aforementioned OECD database. This measure is calculated as a ratio of 1,000. All three measures of economic globalization are expected to be positively associated with the dependent variable.

The second group of controls are taken from the literature on power-resource theory (Korpi 1983, 1989). Leftist share of government estimates the proportion of government seats in the possession of left-leaning political parties; this information is from Swank (2009). Union density calculates the total number of unionized workers as a proportion of the entire labor force; these data are from the OECD database. Industrial employment represents the proportion of labor employed in the manufacturing sector; these data are from the ILO database (http://www.ilo.org/ilostat/). Power-resource variables are anticipated to be negatively associated with 90–50 inequality. In addition to these covariates, we also include the unemployment rate, which estimates the proportion of the labor force that is without work but available and seeking employment. According to conventional economic wisdom, although power-resource variables reduce inequality, it is often at the expense of higher unemployment.

The third set of control variables are designed to capture the new economy's effect on income inequality. Tertiary school enrollment is the total number of students enrolled in tertiary educational institutions, as a proportion of the total age-relevant population. Many claim that the increasing need for non-routine cognitive skills expands the demand for and wages of highly educated employees, thereby increasing inequality (Autor and Dorn 2010; Autor, Levy, and Murnane 2003). Female labor participation is the total number of women in the labor force as a proportion of the female population. Past empirical research finds mixed results for female participation. Both variables are from the aforementioned OECD database.

The final control captures the effect of economic development on income inequality. According to Kuznets (1955), income inequality increases during the early stages of industrialization because a small segment of the national labor force is employed in the high-wage manufacturing sector vis-à-vis the low-wage agricultural sector. Yet as industrialization continues, income inequality decreases, as a larger segment of the labor force is absorbed into the higher-wage segments of the economy. Often referred to as the Kuznets inverted-U hypothesis, a number of scholars over the years tested and empirically confirmed these claims (Kwon 2014; Lee 2005; Nielsen 1994). We thus use GDP per capita and its squared term to control for the effect of development on income inequality. These data are from the OECD database.

### Statistical Techniques

To test our two main hypotheses, we compile a time-series cross-sectional (TSCS) dataset with information on 20 advanced industrial economies during the years 1988 to 2009.5 It is well documented that ordinary least squares (OLS) regression is inappropriate for TSCS data. Most importantly, OLS is susceptible to heterogeneity bias when analyzing TSCS data, because unobserved time-invariant unit-specific factors may be correlated with the observed covariates, resulting in biased estimates of the coefficients. In addition, TSCS data are typically hindered by issues of autocorrelation and heteroskedasticity, particularly when the units are countries. Simply put, OLS regression is not the best linear unbiased estimator when analyzing TSCS data.

To deal with these issues, social scientists often employ fixed effects models (FEM) or random effects models (REM). The advantage of FEM is that it controls for all unobservable time-invariant country-level effects, making this an ideal technique for eradicating spurious relationships. At the same time, FEM is unable to analyze time-invariant covariates, and slowly changing predictors often lose their statistical significance. The latter point is especially pertinent to our study since many national-level predictors of inequality change slowly over time. REM is a more efficient estimator of slowly changing variables. However, REM cannot control for within-country unobservables and tends to produce unreliable estimates when error terms are correlated with the regressors (Halaby 2004; Wooldridge 2002).

Given these considerations, we perform a range of preliminary diagnostics to determine the optimal statistical technique for our data. First, Hausman tests are used to determine whether heterogeneity bias is a concern in our dataset. These diagnostics reveal a significant difference in the coefficients produced by FEM and REM, which shows that the former is the preferred technique of choice. Second, Wooldridge tests show that every regression model calculated in the pending analysis contains a significant amount of autocorrelation. Third, preliminary tests also reveal substantial heteroskedasticity in all equations. In light of these preliminary diagnostics, we use Prais-Winsten regressions with panel-corrected standard errors and country-specific fixed effects (Beck and Katz 1995).

Prais-Winsten regression is ideal for the current study because it controls for temporal autocorrelation but retains all information by engaging in a rho-based transformation of the variables. This technique also allows researchers to control for heteroskedasticity of the error terms. Furthermore, Prais-Winsten is able to handle the inclusion of unit-specific fixed effects, negating all the time-invariant country-level unobservables, which often result in spurious relationships and inflated estimates of the coefficients. Thus, the pending analysis presents the following Prais-Winsten models:

1. $Yit=β0+αi+β1Xit−1+β2Wit−1+∑βkZit−1+εit$

2. $Yit=β0+αi+β1Xit−1+β2Wit−1+β3Xit−1*Wit−1+∑βkZit−1+εit$

where Y represents 90–50 inequality, i denotes the country, t is the observed year, α signifies a vector of country-specific intercepts, X is wage coordination, W is financialization, Z is a vector of control variables, the β's are coefficients, and ε is the error term. It is of note that the inclusion of fixed effects in Prais-Winsten regression inflates the R2 estimates, rendering them essentially uninterpretable.

We also perform a range of robustness checks. First, although we believe that the approach outlined above is the best technique for analyzing our data, we also make use of two popular alternative approaches to ensure that our results are robust across different estimation techniques: GLS random effects and OLS with robust-clustered standard errors. It is useful to test our models with these estimators given their popularity in the comparative income inequality literature (e.g. Alderson and Nielsen 2002; Huber and Stephens 2014).6 Second, given concerns over the potential non-stationary nature of 90–50 income inequality, we present robustness checks that include a lagged dependent variable in the right-hand side of the equation. We also estimate dynamic fixed effects error correction models to explore the short- and long-term effect of the independent variables.

Third, each model includes an indicator variable for the years 2000 to 2009 as a control for temporal effects on the dependent variable. However, alternative models are reported where we include yearly fixed effects in the regression. Fourth, the main regression models estimate the independent variables at time t − 1 to assuage potential concerns over temporal reverse causality. But we also report additional robustness checks with models that calculate the independents at time t. Fifth, an assessment of the Hadi procedure indicates that the United States may be overly influential in the regression models. Thus, robustness checks are performed by excluding the United States. Sixth, to ensure that the results are not an artifact of a particular measure of inequality, we present separate models using the Gini coefficient and the top 5% share of national income.

A few final notes. An assessment of the variance inflation factor indicates some collinearity between the right-hand-side variables when GDP per capita is included in the models (VIF = 12.80).7 Given this concern, we selectively include GDP per capita in a limited number of equations. To further assuage concerns of collinearity, Appendix B reproduces the main results by testing various combinations of the controls. The following predictors are positively skewed and thus converted to a natural log: unemployment rate, union density, southern import penetration, FDI outflow, financialization index, and GDP per capita. And finally, we mean-center wage coordination and the financialization index. This simplifies interpretation of the main effects of these covariates in the interaction models. The correlation matrix and summary statistics are available in Appendix C.

## RESULTS AND DISCUSSION

The results start in Table 1. According to Model 1, financialization is a positive and significant predictor of 90–50 income inequality. These findings support the first hypothesis of the study and are consistent with a bourgeoning field of research in the social sciences that indicates financial activities widen income inequality in advanced industrial societies (Assa 2012; Bell and van Reenen 2013; Denk and Cournede 2015; Godechot 2012; Goldstein 2012; Kus 2012; Kwon and Roberts 2015; Lin and Tomaskovic-Devey 2013; Nau 2013; Tomaskovic-Devey and Lin 2011; Volscho and Kelly 2012). Furthermore, given that we examine 90–50 inequality as our dependent variable, these results lend further credibility to the claim that financialization widens the gap between top- and middle-income earners in developed economies.

TABLE 1.
Fixed Effects Prais-Winsten Models: Financialization, Wage Coordination, and 90–50 Inequality
Model 1Model 2Model 3Model 4
Southern import penetration 0.006 0.001 −0.034 −0.033
(0.19) (0.04) (−1.05) (−1.01)
Outward foreign investment −0.007 −0.009 −0.010 −0.010
(−0.94) (−1.16) (−1.32) (−1.26)
Net migration 0.001 −0.003 −0.001 −0.003
(0.09) (−0.18) (−0.09) (−0.22)
Female labor participation −0.001 −0.001 −0.003* −0.004*
(−1.13) (−0.99) (−2.37) (−2.46)
Tertiary school enrollment 0.080** 0.094** 0.031 0.022
(3.67) (4.25) (1.41) (0.96)
Unemployment rate −0.015 −0.018 0.011 0.010
(−1.25) (−1.46) (0.86) (0.82)
Leftist party −0.000 −0.000 −0.000 −0.000
(−0.09) (−0.22) (−0.57) (−0.84)
Union density −0.134** −0.108** −0.108** −0.088*
(−3.43) (−2.73) (−2.93) (−2.27)
Industrial employment 0.002 0.003 0.005* 0.004+
(1.07) (1.50) (2.34) (1.88)
GDP per capita   0.366** 3.466
(5.46) (1.57)
GDP per capita (squared)    −0.147
(−1.40)
Wage coordination (WC) −0.000 0.095** 0.080** 0.071*
(−0.06) (3.18) (2.80) (2.43)
Financialization 0.074** 0.207** 0.130* 0.117*
(2.61) (4.00) (2.55) (2.28)
Financialization × WC  −0.042** −0.037** −0.032**
(−3.19) (−2.90) (−2.55)
2000–2009 0.004 0.004 −0.007 −0.007
(0.51) (0.52) (−0.95) (−0.95)
Constant 2.028** 2.193** −1.212+ −17.491
(7.22) (8.01) (−1.71) (−1.51)

Wald chi2 6727.65 7198.53 10169.42 7247.96
R2 .975 .975 .977 .977
BIC −961.93 −978.90 −1006.79 −1008.20
N 322 322 322 322
Countries 20 20 20 20
Years 1988–2009 1988–2009 1988–2009 1988–2009
Model 1Model 2Model 3Model 4
Southern import penetration 0.006 0.001 −0.034 −0.033
(0.19) (0.04) (−1.05) (−1.01)
Outward foreign investment −0.007 −0.009 −0.010 −0.010
(−0.94) (−1.16) (−1.32) (−1.26)
Net migration 0.001 −0.003 −0.001 −0.003
(0.09) (−0.18) (−0.09) (−0.22)
Female labor participation −0.001 −0.001 −0.003* −0.004*
(−1.13) (−0.99) (−2.37) (−2.46)
Tertiary school enrollment 0.080** 0.094** 0.031 0.022
(3.67) (4.25) (1.41) (0.96)
Unemployment rate −0.015 −0.018 0.011 0.010
(−1.25) (−1.46) (0.86) (0.82)
Leftist party −0.000 −0.000 −0.000 −0.000
(−0.09) (−0.22) (−0.57) (−0.84)
Union density −0.134** −0.108** −0.108** −0.088*
(−3.43) (−2.73) (−2.93) (−2.27)
Industrial employment 0.002 0.003 0.005* 0.004+
(1.07) (1.50) (2.34) (1.88)
GDP per capita   0.366** 3.466
(5.46) (1.57)
GDP per capita (squared)    −0.147
(−1.40)
Wage coordination (WC) −0.000 0.095** 0.080** 0.071*
(−0.06) (3.18) (2.80) (2.43)
Financialization 0.074** 0.207** 0.130* 0.117*
(2.61) (4.00) (2.55) (2.28)
Financialization × WC  −0.042** −0.037** −0.032**
(−3.19) (−2.90) (−2.55)
2000–2009 0.004 0.004 −0.007 −0.007
(0.51) (0.52) (−0.95) (−0.95)
Constant 2.028** 2.193** −1.212+ −17.491
(7.22) (8.01) (−1.71) (−1.51)

Wald chi2 6727.65 7198.53 10169.42 7247.96
R2 .975 .975 .977 .977
BIC −961.93 −978.90 −1006.79 −1008.20
N 322 322 322 322
Countries 20 20 20 20
Years 1988–2009 1988–2009 1988–2009 1988–2009

+p < .10, *p < .05, **p < .01

Notes. t-values in parentheses. Independent variables are measured at t − 1. Reports robust panel-corrected standard errors. Country-level dummies included in regressions but not reported to save space. Bayesian information criterion (BIC) calculated using fixed effects OLS regression with robust standard errors.

In terms of the controls, although a number of covariates are inconstantly significant in various models, only union density is a negative robust predictor across all models reported in Table 1. Perhaps the most intriguing finding is that wage coordination is not significantly associated with 90–50 inequality. However, this null finding is not necessarily incompatible with the empirical literature, as some scholars find that wage coordination shares a somewhat ambiguous association with income inequality (e.g. Bradley et al. 2003; Golden and Londregan 2006; Huber and Stephens 2014). In addition, it is important to reiterate at this point that FEM is a less efficient estimator of slowly changing covariates, such as wage coordination, than REM. It is thus possible that these non-significant findings may be attributable to the nature of the estimators used in these particular specifications.8

The remaining models in Table 1 test this study's main contention. In these equations, we regress 90–50 inequality on wage coordination, financialization, and an interaction term of the latter two predictors. According to Model 2, the interaction of the financialization index and wage coordination is a significant negative predictor of the dependent variable. In Model 3, we include GDP per capita to control for the effect of development on inequality, but the main results remain unaltered. The inclusion of this covariate in the specification allows us to control for those changes that could be associated with the rise of finance in advanced industrial economies. Model 4 introduces a quadratic term for GDP per capita to control for the proposed inverted-U relationship between development and income inequality (Kuznets 1955). However, the interaction of financialization and wage coordination retains its robust negative association. Recent empirical research shows that income inequality may no longer display an inverted-U relationship with GDP per capita (Kwon 2014). Yet we include a quadratic term for this variable to demonstrate the robustness of our main results.

These models contribute a novel set of findings to the literature. That is, although previous research focuses on how unionization rates and labor protections diminish the positive association between finance and inequality (Darcillon 2016), we add to this line of research by showing that national wage-coordination systems may produce similar effects. Furthermore, it is noteworthy that the effects of finance, wage coordination, and the interaction of these covariates are reduced when GDP per capita is included in the regression model. This is to be expected since GDP per capita may be considered a proxy for many of the recent socioeconomic trends, such as the transition to the knowledge economy, which should be highly correlated with the rise of financialization in developed economies. As mentioned in previous sections, this is demonstrated by the presence of collinearity between the right-hand-side variables when GDP per capita is included in the regression models. Nevertheless, although GDP per capita somewhat diminished the overall effect of the covariates of interest, the interaction of financialization and wage coordination remains a robust predictor of income inequality when GDP per capita is included in the equation.

We now report a set of robustness checks, in Tables 2 and 3. In these models, we do not include GDP per capita due to the aforementioned collinearity concerns. Starting with Models 5 and 6, in these equations we analyze the interaction effects via two different analytic techniques—GLS random effects and OLS with robust clustered standard errors—to ensure that the findings are not merely an artifact of a particular statistical approach. These models are important since researchers in the cross-national inequality literature use different statistical techniques to analyze their data. We also retest the results using techniques that are specifically designed to control for non-stationary variables. We do this in two ways. First, Model 7 includes a lagged measure of the dependent variable in the right-hand side of the regression. Second, Model 8 summarizes findings for a dynamic fixed effects error correction model. For brevity, we present only the long-run effects. According to these results, the interaction of financialization and wage coordination retains its positive and significant link with 90–50 income inequality. That is, regardless of the technique used, wage coordination seems to reduce the inequality-producing effect of financialization.

TABLE 2.
Robustness Checks: Alternative Regression Techniques
GLS Random EffectsOLS Robust ClusterLagged DependentFE Error Correction
Model 5Model 6Model 7Model 8
Southern import penetration −0.056+ −0.056+ 0.003 −0.062
(−1.80) (−1.69) (0.16) (−0.68)
Outward foreign investment −0.020+ −0.020+ −0.015* −0.045
(−1.67) (−1.76) (−2.06) (−1.21)
Net migration −0.005 −0.005 −0.028* −0.059
(−0.31) (−0.26) (−2.06) (−1.37)
Female labor participation −0.001 −0.001 0.000 0.002
(−1.31) (−0.75) (0.45) (0.72)
Tertiary school enrollment 0.112** 0.112** 0.055** 0.143**
(5.62) (3.79) (3.85) (2.99)
Unemployment rate −0.024* −0.024 −0.006 −0.012
(−2.10) (−1.07) (−0.84) (−0.41)
Leftist party −0.000 −0.000 0.000 0.000
(−0.20) (−0.12) (1.34) (0.92)
Union density −0.097** −0.097* −0.025 −0.079
(−3.30) (−2.04) (−1.07) (−0.96)
Industrial employment 0.006** 0.006* 0.005** 0.011**
(2.84) (2.08) (3.21) (2.66)
Wage coordination (WC) 0.153** 0.153** 0.083** 0.235*
(4.60) (3.22) (3.87) (2.35)
Financialization 0.353** 0.353** 0.177** 0.500**
(6.60) (5.00) (4.68) (4.98)
Financialization × WC −0.067** −0.067** −0.036** −0.104*
(−4.60) (−3.19) (−3.79) (−2.44)
90-50 inequality   0.645**
(12.23)
Error correction rate    −0.357**
(−4.40)
2000–2009 0.005 0.005 −0.001 0.005
(0.59) (0.34) (−0.23) (0.34)
Constant 2.410** 2.410** 0.818** 0.871**
(10.28) (6.20) (4.05) (3.80)
Wald Chi2 9511.45  37686.53
R2 .970 .970 .983 .649
N 322 322 315 315
Countries 20 20 20 20
Years 1988–2009 1988–2009 1988–2009 1988–2009
GLS Random EffectsOLS Robust ClusterLagged DependentFE Error Correction
Model 5Model 6Model 7Model 8
Southern import penetration −0.056+ −0.056+ 0.003 −0.062
(−1.80) (−1.69) (0.16) (−0.68)
Outward foreign investment −0.020+ −0.020+ −0.015* −0.045
(−1.67) (−1.76) (−2.06) (−1.21)
Net migration −0.005 −0.005 −0.028* −0.059
(−0.31) (−0.26) (−2.06) (−1.37)
Female labor participation −0.001 −0.001 0.000 0.002
(−1.31) (−0.75) (0.45) (0.72)
Tertiary school enrollment 0.112** 0.112** 0.055** 0.143**
(5.62) (3.79) (3.85) (2.99)
Unemployment rate −0.024* −0.024 −0.006 −0.012
(−2.10) (−1.07) (−0.84) (−0.41)
Leftist party −0.000 −0.000 0.000 0.000
(−0.20) (−0.12) (1.34) (0.92)
Union density −0.097** −0.097* −0.025 −0.079
(−3.30) (−2.04) (−1.07) (−0.96)
Industrial employment 0.006** 0.006* 0.005** 0.011**
(2.84) (2.08) (3.21) (2.66)
Wage coordination (WC) 0.153** 0.153** 0.083** 0.235*
(4.60) (3.22) (3.87) (2.35)
Financialization 0.353** 0.353** 0.177** 0.500**
(6.60) (5.00) (4.68) (4.98)
Financialization × WC −0.067** −0.067** −0.036** −0.104*
(−4.60) (−3.19) (−3.79) (−2.44)
90-50 inequality   0.645**
(12.23)
Error correction rate    −0.357**
(−4.40)
2000–2009 0.005 0.005 −0.001 0.005
(0.59) (0.34) (−0.23) (0.34)
Constant 2.410** 2.410** 0.818** 0.871**
(10.28) (6.20) (4.05) (3.80)
Wald Chi2 9511.45  37686.53
R2 .970 .970 .983 .649
N 322 322 315 315
Countries 20 20 20 20
Years 1988–2009 1988–2009 1988–2009 1988–2009

+p < .10, *p < .05, **p < .01

Notes. t-values in parentheses. Independent variables are measured at t − 1. Reports robust panel-corrected standard errors. Country-level dummies included in regressions but not reported to save space.

TABLE 3.
Robustness Checks: Testing for Different Effects of Time, Potentially Influential Cases, and Alternative Measures of Income Inequality
Independents Measured at tTime-Specific Fixed EffectsExcluding United StatesMarket Gini CoefficientTop 5% Share of Income
Model 9Model 10Model 11Model 12Model 13
Southern import penetration −0.043 −0.159 −0.008 1.254 0.961
(−1.38) (−3.65) (−0.26) (1.40) (0.61)
Outward foreign investment 0.004 −0.007 −0.008 −0.029 −0.424
(0.52) (−0.84) (−0.96) (−0.08) (−0.70)
Net migration 0.045* −0.015 0.002 0.391 −0.215
(2.56) (−0.88) (0.11) (0.89) (0.56)
Female labor participation −0.002+ −0.004 −0.000 0.002 −0.152**
(−1.78) (−3.30) (−0.36) (0.05) (−3.42)
Tertiary school enrollment 0.063** −0.008 0.116** 3.476** 2.252**
(2.86) (−0.34) (4.99) (5.03) (2.88)
Unemployment rate −0.018 −0.030* −0.019 1.868** 0.204
(−1.51) (−2.45) (−1.51) (4.31) (0.33)
Leftist party −0.003 0.000 −0.000 −0.009 −0.010
(−0.88) (0.39) (−0.08) (−0.75) (−0.95)
Union density −0.112** −0.070* −0.101* −3.552* −0.880
(−2.64) (−2.04) (−2.54) (−2.29) (−0.60)
Industrial employment −0.003 0.011** 0.006* 0.005 −0.128
(−1.25) (4.50) (2.53) (0.07) (−1.62)
Wage coordination (WC) 0.059* 0.051+ 0.054 1.803+ 2.797*
(2.00) (1.72) (1.61) (1.95) (1.99)
Financialization 0.156** 0.142** 0.124* 2.567+ 7.564**
(3.03) (2.67) (2.02) (1.83) (3.77)
Financialization × WC −0.027* −0.035* −0.048* −0.838* −1.408*
(−2.11) (−2.21) (−2.32) (−2.09) (−2.38)
2000–2009 −0.000  0.005 −0.286 0.232
(−0.10)  (0.61) (−1.07) (0.66)
Constant 2.272** 2.816** 1.812** 35.051** 32.487**
(8.25) (11.23) (6.16) (3.85) (3.90)
Wald Chi2 8330.83 10115.70 5541.63 1078.00 9511.45
R2 .978 .978 .974 .919 .970
N 318 322 300 396 235
Countries 20 20 19 20 13
Years 1988–2009 1988–2009 1988–2009 1988–2009 1988–2009
Independents Measured at tTime-Specific Fixed EffectsExcluding United StatesMarket Gini CoefficientTop 5% Share of Income
Model 9Model 10Model 11Model 12Model 13
Southern import penetration −0.043 −0.159 −0.008 1.254 0.961
(−1.38) (−3.65) (−0.26) (1.40) (0.61)
Outward foreign investment 0.004 −0.007 −0.008 −0.029 −0.424
(0.52) (−0.84) (−0.96) (−0.08) (−0.70)
Net migration 0.045* −0.015 0.002 0.391 −0.215
(2.56) (−0.88) (0.11) (0.89) (0.56)
Female labor participation −0.002+ −0.004 −0.000 0.002 −0.152**
(−1.78) (−3.30) (−0.36) (0.05) (−3.42)
Tertiary school enrollment 0.063** −0.008 0.116** 3.476** 2.252**
(2.86) (−0.34) (4.99) (5.03) (2.88)
Unemployment rate −0.018 −0.030* −0.019 1.868** 0.204
(−1.51) (−2.45) (−1.51) (4.31) (0.33)
Leftist party −0.003 0.000 −0.000 −0.009 −0.010
(−0.88) (0.39) (−0.08) (−0.75) (−0.95)
Union density −0.112** −0.070* −0.101* −3.552* −0.880
(−2.64) (−2.04) (−2.54) (−2.29) (−0.60)
Industrial employment −0.003 0.011** 0.006* 0.005 −0.128
(−1.25) (4.50) (2.53) (0.07) (−1.62)
Wage coordination (WC) 0.059* 0.051+ 0.054 1.803+ 2.797*
(2.00) (1.72) (1.61) (1.95) (1.99)
Financialization 0.156** 0.142** 0.124* 2.567+ 7.564**
(3.03) (2.67) (2.02) (1.83) (3.77)
Financialization × WC −0.027* −0.035* −0.048* −0.838* −1.408*
(−2.11) (−2.21) (−2.32) (−2.09) (−2.38)
2000–2009 −0.000  0.005 −0.286 0.232
(−0.10)  (0.61) (−1.07) (0.66)
Constant 2.272** 2.816** 1.812** 35.051** 32.487**
(8.25) (11.23) (6.16) (3.85) (3.90)
Wald Chi2 8330.83 10115.70 5541.63 1078.00 9511.45
R2 .978 .978 .974 .919 .970
N 318 322 300 396 235
Countries 20 20 19 20 13
Years 1988–2009 1988–2009 1988–2009 1988–2009 1988–2009

+p < .10, *p < .05, **p < .01

Notes.t-values in parentheses. Independent variables are measured at t − 1. Reports robust panel-corrected standard errors. Country-level dummies included in regressions but not reported to save space.

The next set of robustness checks reanalyze the main interaction results by dealing with potential temporal effects on the dependent variable in different ways. Starting with Model 9, in which the independent variables are measured at time t as opposed to t − 1, the financialization and wage-coordination interaction term retains its robust negative association with the dependent variable. In Model 10, when time-specific fixed effects are included in the specification in lieu of the 2000–2009 dummy indicator included in the main model, the interaction of finance and wage coordination remains a negative and significant predictor of 90–50 income inequality. Thus, what these tests show is that regardless of whether we measure the independent variables at time t or t − 1, or include a dummy indicator for decade or annual time-specific fixed effects, wage coordination seems to significantly reduce the inequality-producing effect of financialization.

In Model 11, we remove the United States from the analysis, as the Hadi procedure identifies this country's observations as potential overly influential. Although removing the United States reduces the significance of the finance and wage-coordination interaction term vis-à-vis the main interaction model, this covariate continues to share a robust negative relationship with the dependent variable.9 In Models 12 and 13, we replace 90–50 income inequality with the pretax measure of the Gini coefficient and the top 5% share of national income, respectively. The former measure is particularly useful as the closest study to the current investigation uses the Gini coefficient to test whether unionization and labor protections condition the link between finance and inequality (Darcillon 2016). However, even when we replace our dependent variable with these alternative measures, the results remain unchanged.

In summary, financialization shares a significant positive association with 90–50 inequality in advanced industrial societies. More importantly, and in line with our contention, the positive connection between financial activities and income inequality is significantly lower at increasing levels of wage coordination. According to our findings, this relationship is robust across different analytic techniques (fixed effects, random effects, robust clustered standard errors, lagged dependent variable, fixed effects error correction models) and a number of alternative regression parameters (lagged and non-lagged variables, time-specific fixed effects versus decade-specific dummies, including and excluding the United States). In short, wage-setting institutions help reduce the inequality-producing effect of financialization.

## CONCLUSIONS

According to a rapidly expanding field of research, the growing centrality of financial activity in advanced industrial societies is a major driving force of inequality over the past few decades (Assa 2012; Bell and van Reenen 2013; Denk and Cournede 2015; Godechot 2012; Goldstein 2012; Kus 2012; Kwon and Roberts 2015; Lin and Tomaskovic-Devey 2013; Nau 2013; Tomaskovic-Devey and Lin 2011; Volscho and Kelly 2012). However, although many scholars recognize that between-country differences in labor institutions may account for some of the cross-national variation in income inequality (Alderson and Nielsen 2002; Mahler 2004; Pontusson, Rueda, and Way 2002; Rueda and Pontusson 2000; Wallerstein 1999), no studies explore whether wage coordination conditions the distributional effect of financialization.

We thus contribute to this literature by testing whether wage coordination moderates the positive association between financialization and income inequality. Specifically, while previous work examines how unionization and labor protections may mitigate the positive link between finance and inequality (Darcillon 2016), we explore whether wage coordination produces a similar effect. According to our results, the interaction of financialization with wage coordination is significant and negatively associated with the dependent variable. Furthermore, these results are robust across different estimation techniques and a variety of regression parameters. To further clarify these results, we use information from Model 2 to graph the estimated effect of financialization at different levels of wage coordination. As seen in Figure 3, what becomes clear from these results is that financialization's overall effect on income inequality decreases at higher levels of wage coordination. Specifically, when calculating finance's effect at the observed minimum and maximum levels of wage coordination, 90–50 inequality decreases by 0.16, or approximately 0.72 standard deviations. When thus taken as a whole, our findings indicate that wage coordination helps offset the inequality-producing effect of financial activity in advanced industrial societies.

Figure 3.

Estimated Effect of Financialization and Wage Coordination, 1988 to 2009

Figure 3.

Estimated Effect of Financialization and Wage Coordination, 1988 to 2009

The findings of the current study provide multiple viable avenues for future research. First, our analysis draws heavily from the varieties of capitalism framework, by concentrating specifically on this literature's discussion of how divergent labor institutions in CMEs and LMEs may condition the impact of financialization (Hall and Gingerich 2009; Hall and Soskice 2001; Hall and Thelen 2014; Rueda and Pontusson 2000; Thelen 2014). Yet the current investigation was more concerned with this literature's discussion of wage coordination than with its treatment of the CME–LME distinction. Thus, in light of our finding that labor institutions condition the distributional impact of financial activities, it may be useful to identify other divergent characteristic of CMEs and LMEs that may affect cross-national variations in income inequality.

Another potential avenue of future scholarship is on the relationship between wage coordination and income inequality. Although prior research reveals a significant and negative association between various measures of wage-setting and inequality (Alderson and Nielsen 2002; Mahler 2004; Pontusson, Rueda, and Way 2002; Rueda and Pontusson 2000; Wallerstein 1999), wage coordination is not a significant predictor in the regression models reported in this study. This lack of evidence is not necessarily surprising, especially when considering other researchers’ finding that wage coordination returns ambiguous associations with income inequality (e.g. Bradley et al. 2003; Huber and Stephens 2014). Nevertheless, given these observations, it would be interesting for future studies to explore whether the use of different measures of inequality and/or potential time-contingent effects of wage coordination are responsible for this apparent disconnect.

Finally, and related to the previous point, it is interesting to note that in stark contrast to conventional wisdom, many of our regression models suggest that wage coordination shares a positive association with 90–50 inequality when financialization is at zero. This phenomenon is further shown by the estimated effect of wage coordination as shown in Figure 3. According to recent scholarship in the varieties of capitalism literature, while it is common for scholars to equate wage-setting institutions with equity-producing institutions, wage-coordination systems were adopted in many CMEs as a way to generate social solidarity and control wage inflation, not as a way to reduce income inequality (Thelen 2014). What this means in many CMEs is that wage coordination serves as an institutionalized mechanism which allows employers to shape national economic policy and obtain labor's blessing on corporate strategy (Avdagic 2010; Baccaro and Lim 2007; Culpepper and Regan 2014). In light of these observations, it may be interesting for future studies to explore the specific circumstances under which wage-setting may, or may not, reduce income inequality.

In closing, while top-income earners in rich countries were able to use financial activities to augment their earnings over the past few decades, this was at the expense of middle- and low-income earners in advanced industrial societies. However, we show that national labor institutions may help condition the inequality-producing tendency of financialization. Moving forward, it will be interesting to see whether the middle- and working-class are able to strengthen national wage-coordination systems, or whether the corporate community continues to weaken these labor institutions.

## REFERENCES

REFERENCES
Alderson, Arthur, and François Nielsen.
2002
.
“Globalization and the Great U-Turn: Income Inequality Trends in 16 OECD Countries.”
American Journal of Sociology
107
:
1244
99
.
Arrighi, Giovanni.
1994
.
The Long Twentieth Century: Money, Power, and the Origins of Our Times
.
New York
:
Verso
.
Assa, Jacob.
2012
.
“Financialization and its Consequences: The OECD Experience.”
Finance Research
1
:
35
39
.
Atkinson, Anthony B.
1975
.
The Economics of Inequality
.
Oxford
:
Oxford University Press
.
Autor, David H., and David Dorn.
2010
.
“Inequality and Specialization: The Growth of Low-Skill Service Jobs and the Polarization of the U.S. Labor Market.”
Working Paper,
Massachusetts Institute of Technology
.
Autor, David H., Frank Levy, and Richard J. Murnane.
2003
.
“The Skill Context of Recent Technological Change: An Empirical Exploration.”
Quarterly Journal of Economics
118
:
1279
1333
.
Avdagic, Sabina.
2010
.
“When Are Concerted Reforms Feasible? Explaining the Emergence of Social Pacts in Western Europe.”
Comparative Political Studies
43
:
628
57
.
Baccaro, Lucio, and Sang-Hoon Lim.
2007
.
“Social Pacts as Coalitions of the Weak and Moderate: Ireland, Italy and South Korea in Comparative Perspective.”
European Journal of Industrial Relations
13
:
27
46
.
Beck, Nathaniel, and Jonathan Katz.
1995
.
“What to Do (and Not to Do) with Time-Series Cross-Section Data.”
American Political Science Review
89
:
637
47
.
Bell, Brian, and John Van Reenen.
2013
.
“Extreme Wage Inequality: Pay at the Very Top.”
American Economic Review
103
(
3
):
153
57
.
Bradley, David, Evelyne Huber, Stephanie Moller, François Nielsen, and John D. Stephens.
2003
.
“Distribution and Redistribution in Postindustrial Democracies.”
World Politics
55
:
193
228
.
Brookman, Jeffrey, Saeyoung Chang, and Craig Rennie.
2007
.
“CEO Cash and Stock-Based Compensation Changes, Layoff Decisions, and Shareholder Value.”
Financial Review
42
:
99
119
.
Cappelli, Peter.
2000
.
“Examining the Incidence of Downsizing and Its Effect on Establishment Performance.”
Working Paper 7742,
National Bureau of Economic Research
,
Cambridge, MA
.
Cappelli, Peter, Laurie Bassai, Harry Katz, David Knoke, Paul Osterman, and Michael Useem.
1997
.
Change at Work
.
New York
:
Oxford University Press
.
Culpepper, Pepper D., and Aidan Regan.
2014
.
“Why Don't Governments Need Trade Unions Anymore? The Death of Social Pacts in Ireland and Italy.”
Socio-Economic Review
13
:
723
45
.
Darcillon, Thibault.
2016
.
“Do Interactions between Finance and Labour Market Institutions Affect the Income Distribution?”
Labour
30
(
3
):
235
57
.
Davis, Gerald F.
2009
.
Managed by the Markets: How Finance Re-Shaped America
.
Oxford
:
Oxford University Press
.
Denk, Oliver D., and Boris Cournede.
2015
.
“Finance and Income Inequality in OECD Countries.”
Working paper no. 1224, OECD Economics Department.
Ebbinghaus, Bernhard.
2015
.
“The Privatization and Marketization of Pensions in Europe: A Double Transformation Facing the Crisis.”
European Policy Analysis
1
:
56
73
.
Fiss, Peer C., and Edward J. Zajac.
2004
.
“The Diffusion of Ideas over Contested Terrain: The (Non)adoption of a Shareholder Value Orientation among German Firms.”
49
(
4
):
501
34
.
Fligstein, Neil.
1990
.
The Transformation of Corporate Control
.
Boston
:
Harvard University Press
.
Fligstein, Neil.
2001
.
The Architecture of Markets: An Economic Sociology of Twenty-First-Century Capitalist Societies
.
Princeton
:
Princeton University Press
.
Fligstein, Neil, and Taek-Jin Shin.
2004
. “The Shareholder-Value Society: A Review of the Changes in Working Conditions and Inequality in the U.S. 1976–2000.” Pp.
401
32
in
Social Inequality
, edited by K. Neckerman.
New York
:
Russell Sage Foundation
.
Fligstein, Neil, and Taek-Jin Shin.
2007
.
“Shareholder Value and the Transformation of the U.S. Economy, 1984–2000.”
Sociological Forum
22
:
399
424
.
Glyn, Andrew.
2006
.
Capitalism Unleashed: Finance, Globalization, and Welfare
.
Oxford
:
Oxford University Press
.
Godechot, Olivier.
2015
.
“Financialization Is Marketization! A Study of the Respective Impacts of Various Dimensions of Financialization on the Increase in Global Inequality.”
MaxPo Discussion Paper no. 15/3, Max Planck Sciences Po Center on Coping with Instability in Market Societies.
Golden, Miriam A., and John B. Londregan.
2006
.
“Centralization of Bargaining and Wage Inequality: A Correction of Wallerstein.”
American Journal of Political Science
50
(
1
):
208
13
.
2012
.
“Revenge of the Managers: Labor Cost-Cutting and the Paradoxical Resurgence of Managerialism in the Shareholder Value Era, 1984 to 2001.”
American Sociological Association
77
:
268
94
.
Halaby, Charles.
2004
“Panel Models in Sociological Research: Theory into Practice.”
Annual Review of Sociology
30
:
507
44
.
Hall, Peter A., and Daniel W. Gingerich.
2009
.
“Varieties of Capitalism and Institutional Complementarities in the Political Economy.”
British Journal of Political Science
39
:
449
82
.
Hall, Peter A., and David Soskice.
2001
.
Varieties of Capitalism: The Institutional Foundations of Comparative Advantage
.
Oxford
:
Oxford University Press
.
Hall, Peter A., and Kathleen Thelen.
2009
.
“Institutional Change in Varieties of Capitalism.”
Socio-Economic Review
7
:
7
34
Haves, Jacob, Sigurt Vitols, and Peter Wilke.
2014
. “Financialization and Ownership Change: Challenges for the German Model of Labour Relations.” Pp.
148
75
in
Financialization, New Investment Funds, and Labour: An International Comparison
, edited by H. Gospel, A. Pendleton, and S. Vitols.
Oxford
:
Oxford University Press
.
Hey, John D., and Peter J. Lambert.
1980
.
“Relative Deprivation and the Gini Coefficient: Comment.”
Quarterly Journal of Economics
95
:
567
73
.
Howell, Chris.
2006
.
“Varieties of Capitalism: And Then There Was One?”
Comparative Politics
36
(
1
):
103
24
.
Huber, Evelyne, and John D. Stephens.
2014
.
“Income Inequality and Redistribution in Post-Industrial Democracies: Demographic, Economic and Political Determinants.”
Socio-Economic Review
12
:
245
67
.
Kalleberg, Arne L.
2011
.
Good Jobs, Bad Jobs: The Rise of Polarized and Precarious Employment Systems in the United States, 1970–2000
.
New York
:
Russell Sage Foundation
.
Kenworthy, Lane.
2001
.
“Wage-Setting Measures: A Survey and Assessment.”
World Politics
54
:
57
98
.
Korpi, Walter.
1983
.
The Democratic Class Struggle
.
London
:
Routledge
.
Korpi, Walter.
1989
.
“Power, Politics, and State Autonomy in the Development of Social Citizenship.”
American Sociological Review
54
(
3
):
309
28
.
Krippner, Greta.
2011
.
Capitalizing on Crisis: The Political Origins of the Rise of Finance
.
Cambridge
:
Harvard University Press
.
Kus, Basak.
2012
.
“Financialisation and Income Inequality in OECD Nations: 1995–2007.”
Economic and Social Review
43
:
477
95
.
Kuttner, Robert.
1995
.
Everything for Sale: The Virtues and Limits of Markets
.
New York
:
Knopf
.
Kuznets, Simon.
1955
.
“Economic Growth and Income Inequality.”
American Economic Review
45
:
1
28
.
Kwon, Roy.
2014
.
“Employment Transitions and the Cycle of Income Inequality in Postindustrial Societies.”
International Journal of Comparative Sociology
55
:
404
28
.
Kwon, Roy.
2016
.
“Can We Have Our Cake and Eat It Too? Liberalization, Economic Growth, and Income Inequality in Advanced Industrial Societies.”
Social Forces
95
:
469
502
.
Kwon, Roy, and Anthony Roberts.
2015
.
“Financialization and Income Inequality in the New Economy: An Exploration of Finance's Conditional Effects.”
Sociology of Development
1
(
4
):
442
62
.
Lee, Cheol-Sung.
2005
.
“Income Inequality, Democracy, and Public Sector Size.”
American Sociological Review
70
:
158
81
.
Lee, Cheol-Sung, Young-Bum Kim, and Jae-Mahn Shim.
2011
.
“The Limit of Equality Projects, Public-Sector Expansion, Sectoral Conflicts, and Income Inequality in Postindustrial Economies.”
American Sociological Review
76
:
100
24
.
2013
.
“Financialization and U.S. Income Inequality, 1970–2008.”
American Journal of Sociology
118
:
1284
1329
.
Mahler, Vincent A.
2004
.
“Economic Globalization, Domestic Politics, and Income Inequality in the Developed Countries: A Cross-National Study.”
Comparative Political Studies
37
:
1025
53
.
Mandel, Michael J.
1996
.
The High-Risk Society: Peril and Promise in the New Economy
.
New York
:
.
Nau, Michael.
2013
.
“Economic Elites, Investments, and Income Inequality.”
Social Forces
92
:
437
61
.
Nielsen, François.
1994
.
“Income Inequality and Industrial Development: Dualism Revisited.”
American Sociological Review
59
:
654
77
.
Osterman, Paul.
1999
.
Securing Prosperity: The American Labor Market: How It Has Changed and What We Can Do about It
.
Princeton, NJ
:
Princeton University Press
.
Osterman, Paul.
2008
.
The Truth about Middle Managers: Who They Are, How They Think, Why They Matter
.
Cambridge
:
.
Piketty, Thomas.
2014
.
Capital in the Twenty-First Century
.
Cambridge
:
Harvard University Press
.
Pontusson, Jonas, David Rueda, and Christopher R. Way.
2002
.
“Comparative Political Economy of Wage Distribution: The Role of Partisanship and Labour Market Institutions.”
British Journal of Political Science
32
:
281
308
.
Rueda, David, and Jonas Pontusson.
2000
.
“Wage Inequality and Varieties of Capitalism.”
World Politics
52
:
350
83
.
Schneider, Martin R., and Mihai Paunescu.
2011
.
“Changing Varieties of Capitalism and Revealed Comparative Advantages from 1990 to 2005: A Test of the Hall and Soskice Claims.”
Socio-Economic Review
9
:
1
23
.
Stearns, Linda B., and Kenneth D. Allan.
1996
.
“Economic Behavior in Institutional Environments: The Corporate Merger Wave of the 1980s.”
American Sociological Review
61
:
699
718
.
Streeck, Wolfgang.
2009
.
Re-Forming Capitalism
.
Oxford
:
Oxford University Press
.
Swank, Duane.
2009
.
“Electoral, Legislative, and Government Strength of Political Parties by Ideological Group in Capitalist Democracies, 1950–2006: A Database.”
Retrieved June 2014 (http://www.marquette.edu/polisci/faculty_swank.shtml).
Thelen, Kathleen.
2014
.
Varieties of Liberalization and the New Politics of Social Solidarity
.
New York
:
Cambridge University Press
.
2011
.
“Income Dynamics, Economic Rents, and the Financialization of the U.S. Economy.”
American Sociological Review
76
:
538
59
.
Useem, Michael.
1993
.
Executive Defense: Shareholder Power and Corporate Reorganization
.
Cambridge
:
Harvard University Press
.
Vogel, Steven K.
1996
.
Freer Markets, More Rules: Regulatory Reform in Advanced Industrial Countries
.
New York
:
Cornell University Press
.
Volscho, Thomas W., and Nathan J. Kelly.
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
.
Wallerstein, Michael.
1999
.
“Wage-Setting Institutions and Pay Inequality in Advanced Industrial Societies.”
American Journal of Political Science
43
:
649
80
.
Weiss, Linda.
2010
. “The State in the Economy: Neoliberal or Neoactivist?” Pp.
183
210
in
The Oxford Handbook of Comparative Institutional Analysis
, edited by G. Morgan, J. Campbell, C. Crouch, O. Pedersen, and R. Whitley.
Oxford
:
Oxford University Press
.
Wooldridge, Jeffrey M.
2002
.
Econometric Analysis of Cross Section and Panel Data
.
Cambridge
:
MIT Press
.
Zuckerman, Ezra W.
2000
.
“Focusing the Corporate Product: Securities Analysts and De-diversification.”
45
:
591
619
.

## NOTES

NOTES
1.
We make no claim that wage-setting institutions mediate the effect of financialization, since there is little evidence that countries with strong wage-setting institutions have less financialization.
2.
It is important to note that this study is not concerned with exploring differences between coordinated and liberal market economies. Rather, varieties of capitalism is a useful framework for how to conceptualize differences in wage coordination in the context of corporate strategy during a period of financialization.
3.
We also test all reported models using 90–10 inequality as suggested by some scholars (e.g. Godechot 2015). However, although the main interaction terms were significant, these covariates were only able to surpass the less conventional p < .10 significance threshold in most models. Thus, we use 90–50 inequality as it is more consistent with the purpose of this study and it returns much more stable and significant results.
4.
This approach of analyzing the 90–50 wage ratio is also consistent with the research on wage coordination and inequality (e.g. Mahler 2004; Rueda and Pontusson 2000; Wallerstein 1999).
5.
The following countries and years are included in this study: Australia 1988–2009, Austria 2004–2009, Belgium 1999–2009, Canada 1997–2000, Denmark 1988–2009, Finland 1988–2009, France 1988–2009, Germany 1991–2009, Greece 2004–2008, Ireland 1995–2009, Italy 1988–2009, Japan 1988–2009, Netherlands 1988–2005, New Zealand 1988–2009, Norway 1997–2006, Portugal 2004–2009, Spain 2004–2009, Sweden 1988–2009, United Kingdom 1988–2007, and United States 1988–2009.
6.
Most scholars are able to justify the use of these approaches given that FEM measures within-country variations of inequality, while REM calculates between-country variations of inequality. Although this may be true, using REM produces biased estimates, for the reasons outlined in prior sections of this article.
7.
This reported range of VIFs does not take into consideration models with interaction effects since their inclusion artificially inflates the VIF scores.
8.
In an unreported set of tests, we find that wage coordination does indeed return a robust negative association with 90–50 inequality in a random effects model.
9.
We also iteratively remove each country from the analysis to see whether a particular country is driving the results for the interaction of finance and wage coordination. Regardless of the country that is removed from the analysis, the results remain unchanged.
APPENDIX A.
Coding of Wage-Coordination Variable
 5 Centralized bargaining by peak association(s), with or without government involvement, and/or government imposition of wage schedule/freeze, with peace obligationInformal centralization of industry-level bargaining by a powerful and monopolistic union confederationExtensive, regularized pattern-setting and highly synchronized bargaining coupled with coordination of bargaining by influential large firms 4 Centralized bargaining by peak association(s), with or without government involvement, and/or government imposition of wage schedule/freeze, without peace obligationInformal (intra-associational and/or inter-associational) centralization of industry- and firm-level bargaining by peak associations (both sides)Extensive, regularized pattern-setting coupled with high degree of union concentration 3 Informal (intra-associational and/or inter-associational) centralization of industry- and firm-level bargaining by peak associations (one side, or only some unions) with or without government participationIndustry-level bargaining with irregular and uncertain pattern-setting and only moderate union concentrationGovernment arbitration or intervention 2 Mixed industry- and firm-level bargaining, with no or little pattern bargaining and relatively weak elements of government coordination through the setting of basic pay rates (statutory minimum wage) or wage indexation (example: France most years) 1 Fragmented wage bargaining, confined largely to individual firms or plants
 5 Centralized bargaining by peak association(s), with or without government involvement, and/or government imposition of wage schedule/freeze, with peace obligationInformal centralization of industry-level bargaining by a powerful and monopolistic union confederationExtensive, regularized pattern-setting and highly synchronized bargaining coupled with coordination of bargaining by influential large firms 4 Centralized bargaining by peak association(s), with or without government involvement, and/or government imposition of wage schedule/freeze, without peace obligationInformal (intra-associational and/or inter-associational) centralization of industry- and firm-level bargaining by peak associations (both sides)Extensive, regularized pattern-setting coupled with high degree of union concentration 3 Informal (intra-associational and/or inter-associational) centralization of industry- and firm-level bargaining by peak associations (one side, or only some unions) with or without government participationIndustry-level bargaining with irregular and uncertain pattern-setting and only moderate union concentrationGovernment arbitration or intervention 2 Mixed industry- and firm-level bargaining, with no or little pattern bargaining and relatively weak elements of government coordination through the setting of basic pay rates (statutory minimum wage) or wage indexation (example: France most years) 1 Fragmented wage bargaining, confined largely to individual firms or plants
APPENDIX B.
Fixed Effects Prais-Winsten Models: Testing Different Combination of the Controls
 Southern import penetration 0.052 0.013 0.023 (1.62) (0.42) (0.74) Outward foreign investment −0.013 −0.010 −0.007 (−1.58) (−1.34) (−0.93) Net migration 0.013 −0.005 0.021 (0.71) (−0.29) (1.15) Female labor participation 0.000 −0.000 −0.001 (0.02) (−0.18) (−1.00) Tertiary school enrollment 0.099** 0.097 0.091** (5.31) (5.08) (4.31) Unemployment rate −0.030** −0.032** −0.017 (−2.69) (−2.77) (−1.42) Leftist party −0.000 −0.000 −0.000 (−0.12) (−0.11) (−0.30) Union density −0.142** −0.136** −0.113** (−4.11) (−3.91) (−2.87) Industrial employment −0.000 −0.001 0.003 (−0.50) (−0.61) (1.26) Wage coordination (WC) 0.102** 0.101** 0.106** 0.066* 0.110** 0.066* 0.091** (3.60) (3.49) (3.74) (2.37) (3.81) (2.35) (3.09) Financialization 0.263** 0.261** 0.207** 0.177** 0.220** 0.181** 0.190** (5.83) (5.52) (4.36) (3.95) (4.49) (3.83) (3.84) Financialization × WC −0.045** −0.044** −0.047** −0.028* −0.048** −0.028* −0.040** (−3.68) (−3.59) (−3.81) (−2.31) (−3.89) (−2.29) (−3.09) 2000–2009 0.023** 0.023** 0.001 0.007 0.002 0.007 0.003 (3.09) (3.09) (0.14) (0.96) (0.26) (0.98) (0.38) Constant 2.126* 1.923** 1.803** 2.518** 1.862** 2.323** 2.186** (2.12) (10.76) (12.82) (20.29) (9.05) (11.46) (9.00) Wald Chi2 7328.42 6489.62 6743.94 7136.59 6585.53 7110.11 7255.49 R2 .971 .972 .974 .973 .975 .973 .975 N 322 322 322 322 322 322 322 Countries 20 20 20 20 20 20 20 Years 1988–2009 1988–2009 1988–2009 1988–2009 1988–2009 1988–2009 1988–2009
 Southern import penetration 0.052 0.013 0.023 (1.62) (0.42) (0.74) Outward foreign investment −0.013 −0.010 −0.007 (−1.58) (−1.34) (−0.93) Net migration 0.013 −0.005 0.021 (0.71) (−0.29) (1.15) Female labor participation 0.000 −0.000 −0.001 (0.02) (−0.18) (−1.00) Tertiary school enrollment 0.099** 0.097 0.091** (5.31) (5.08) (4.31) Unemployment rate −0.030** −0.032** −0.017 (−2.69) (−2.77) (−1.42) Leftist party −0.000 −0.000 −0.000 (−0.12) (−0.11) (−0.30) Union density −0.142** −0.136** −0.113** (−4.11) (−3.91) (−2.87) Industrial employment −0.000 −0.001 0.003 (−0.50) (−0.61) (1.26) Wage coordination (WC) 0.102** 0.101** 0.106** 0.066* 0.110** 0.066* 0.091** (3.60) (3.49) (3.74) (2.37) (3.81) (2.35) (3.09) Financialization 0.263** 0.261** 0.207** 0.177** 0.220** 0.181** 0.190** (5.83) (5.52) (4.36) (3.95) (4.49) (3.83) (3.84) Financialization × WC −0.045** −0.044** −0.047** −0.028* −0.048** −0.028* −0.040** (−3.68) (−3.59) (−3.81) (−2.31) (−3.89) (−2.29) (−3.09) 2000–2009 0.023** 0.023** 0.001 0.007 0.002 0.007 0.003 (3.09) (3.09) (0.14) (0.96) (0.26) (0.98) (0.38) Constant 2.126* 1.923** 1.803** 2.518** 1.862** 2.323** 2.186** (2.12) (10.76) (12.82) (20.29) (9.05) (11.46) (9.00) Wald Chi2 7328.42 6489.62 6743.94 7136.59 6585.53 7110.11 7255.49 R2 .971 .972 .974 .973 .975 .973 .975 N 322 322 322 322 322 322 322 Countries 20 20 20 20 20 20 20 Years 1988–2009 1988–2009 1988–2009 1988–2009 1988–2009 1988–2009 1988–2009

+p < .10, *p < .05, **p < .01

Notes. t-values in parentheses. Independent variables are measured at t − 1. Reports robust panel-corrected standard errors. Country-level dummies included in regressions but not reported to save space.

APPENDIX C.
Correlation Matrix and Summary Statistics
1234567891011121314
1. 90–50 Inequality
2. Southern import pen. −.169
3. Outward foreign invest. −.032 .533
4. Net migration .463 −.294 −.101
5. Female labor participation .047 .139 .037 .017
6. Tertiary school enrollment .233 .076 .084 .253 .340
7. Unemployment rate −029 .063 −.039 −.015 −.337 .039
8. Leftist party −.320 .201 .093 −.422 .159 −.116 .162
9. Union density −.303 .349 .053 −.369 .051 −.012 .045 .193
10. Industrial employment −.060 −.227 −.178 −.061 −.436 −.618 .099 .032 .137
11. GDP per capita −.125 .335 .306 .077 .372 .279 −.433 −.232 .047 −.315
12. Wage coordination −384 .357 .228 −.407 −.078 −.165 −.121 .038 .393 .305 .339
13. Financialization .480 −.172 .202 .360 .298 .406 −.313 −.288 −.353 −.300 .336 −.162
14. 2000–2009 .239 .187 .200 .027 .152 .577 −.145 .031 −.056 −.339 .366 .050 .433
Minimum 1.45 2.32 1.69 5.17 41.1 2.86 0.44 0.00 2.02 15.3 9.46 −1.98 −0.40 0.00
Maximum 2.84 4.21 4.06 7.38 83.3 4.74 3.09 65.0 4.43 40.3 11.11 2.02 0.70 1.00
Mean 1.81 3.25 2.56 5.94 65.3 3.97 1.92 38.5 3.40 26.3 10.38 0.00 0.00 0.50
Std. dev. 0.22 0.30 0.27 0.37 9.5 0.30 0.43 16.3 0.58 4.49 0.38 1.29 0.21 0.50
1234567891011121314
1. 90–50 Inequality
2. Southern import pen. −.169
3. Outward foreign invest. −.032 .533
4. Net migration .463 −.294 −.101
5. Female labor participation .047 .139 .037 .017
6. Tertiary school enrollment .233 .076 .084 .253 .340
7. Unemployment rate −029 .063 −.039 −.015 −.337 .039
8. Leftist party −.320 .201 .093 −.422 .159 −.116 .162
9. Union density −.303 .349 .053 −.369 .051 −.012 .045 .193
10. Industrial employment −.060 −.227 −.178 −.061 −.436 −.618 .099 .032 .137
11. GDP per capita −.125 .335 .306 .077 .372 .279 −.433 −.232 .047 −.315
12. Wage coordination −384 .357 .228 −.407 −.078 −.165 −.121 .038 .393 .305 .339
13. Financialization .480 −.172 .202 .360 .298 .406 −.313 −.288 −.353 −.300 .336 −.162
14. 2000–2009 .239 .187 .200 .027 .152 .577 −.145 .031 −.056 −.339 .366 .050 .433
Minimum 1.45 2.32 1.69 5.17 41.1 2.86 0.44 0.00 2.02 15.3 9.46 −1.98 −0.40 0.00
Maximum 2.84 4.21 4.06 7.38 83.3 4.74 3.09 65.0 4.43 40.3 11.11 2.02 0.70 1.00
Mean 1.81 3.25 2.56 5.94 65.3 3.97 1.92 38.5 3.40 26.3 10.38 0.00 0.00 0.50
Std. dev. 0.22 0.30 0.27 0.37 9.5 0.30 0.43 16.3 0.58 4.49 0.38 1.29 0.21 0.50