The Driver of Workplace Alienation or the Cost of Effective Stewardship? The Consequences of Wage Gap for Corporate Performance

Abstract

Relying on cross-country panel data, the paper investigates the possible repercussions of salary gap for employee productivity and corporate financial performance. Our empirical findings corroborate the presence of a negative tail effect of wage gap on productivity and employee morale. While worsening employee turnover and productivity, and increasing the chances of workplace controversies, high salary gap is found to be associated with a more efficient cost structure and higher profitability. Our evidence suggests that extreme salary gap may be curbed by targeted internal policies favoring internal promotion and career development, unionization, employee and managerial training. The composition of the board’s remuneration committee appears to play but a minor role in shaping the scale of salary gap. The results of the study are in line with equity aversion theory and suggest that extreme wage inequality may impede firms’ growth with spillover effects observable at the macro-level. Targeted policies may be necessary to counter the negative repercussions of high compensation disparities as within-firm mechanisms appear insufficient to mitigate them.

1. Introduction

With income disparities growing across all continents and CEO compensations enormously outpacing the average employee salaries, an increasing attention is drawn to the possible economic repercussions of extreme inequality. Numerous reports regarding the wage gap between top executives and average employees raised a popular outcry, with the debate frequently being politicized and constraining the policymakers to justify the status quo. Ideological clichés commonly shape the debate: top executives frequently allude to merit-based rewards and equality of opportunities, while their opponents point to the potentially destructive impact of income inequality on employees’ motivation and workplace discipline.

While undoubtedly incentivizing competition, the remuneration disparities may potentially undermine group cohesion. The latter may bear negative ramifications for the overall group performance [1]. The key praxeological problem consists in striking the right balance between competition and cohesion with the wage gap being an important instrument in customizing the incentives. Extreme inequality may instigate conflict and cause the inadequately remunerated group members to engage in retaliatory action, while compression of wage disparity between the top and average earners may contribute to the alignment of incentives of employees and management [2].

The persistent two-sided view of salary disparities stems from the ambiguous understanding of the driving causes thereof. To a certain extent, a reasonable wage disparity may be regarded as a cost of effective stewardship or a reward for a better performance [3]. On the other hand, when reaching outsized proportions, it may embody the outcome of rank-order tournaments [4], whereby wage is determined solely by the employees’ position within the corporate hierarchy with the windfalls disproportionately accruing to the entrenched top executives. If ‘rent’ maximization becomes the decisive determinant of executive pay [5], the inequitable distribution of output lacking any link to employee performance may undermine incentives to increase individual productivity and engender shirking. The latter may manifest itself through sluggish output dynamics, higher employee turnover, and intense workplace controversies.

The present paper attempts to verify the impact of wage disparities on corporate performance and employee behavior with particular attention drawn to the cases of the extreme wage gap. The key purpose of the study is to weigh the benefits of pay inequity against its potential negative ramifications by focusing on the following two aspects thereof. First, we intend to verify whether wage gap contributes to the erosion of group cohesion by worsening employee morale, disincentivizing productivity improvement and encouraging workplace controversies. Secondly, we quantify the impact of wage disparities on corporate performance in order to ascertain or disprove the presence of additional costs associated with maintaining disproportionate remunerations.

The problem undertaken in the study is of significant importance from the standpoint of strategic management at the micro-level and socioeconomic policies at the macro-level. The rising wage disparities may exercise significant long-term effects on employee performance, thus impeding the growth of companies. Rampant inequality may even engender organized employee disobedience actions in the form of strikes, wage controversies, and collective legal action, all of which may stifle corporate growth unless kept in check. At the broader societal level, growing wage disparities may threaten social cohesion, accelerate social stratification, reduce social mobility and produce atomization, all of which may impair the implementation of effective socio-economic policies. Economic polarization also inevitably leads to political polarization. Overall, increasing wage disparities observed across all developed countries may threaten their long-term sustainable development prospects.

The study relies on a cross-country firm-level panel database of 1709 companies observed between 2003 and 2018. We compiled data regarding the financial performance of the studied companies, employee productivity, salary disparity, internal workplace policies (including unionization, promotion and career development practices), corporate governance settings and controversies regarding CEO/employee remuneration.

Table 1. Definitions of the variables.

Variable	Definition
Firm-level fundamentals
COGS	Costs of goods sold scaled by the contemporaneous value of total revenue
EBIT.MARGIN	Ratio of EBIT to total revenue
EBITDA.MARGIN	Ratio of EBITDA to total revenue
NET.PROFIT.M	Ratio of net profit after taxes excluding extraordinary items to total revenue
PROD.EBITDA	Natural logarithm of the ratio of EBITDA to the number of employees
PROD.REVENUE	Natural logarithm of the ratio of total revenue to the number of full-time employees
SGA.EXP	Sum of sales, general and administrative expenses, labor and related expenses, and advertising expenses divided by total revenue
Firm-level controls
ASSET.TANG	Ratio of Net property, plant and equipment to total assets
CAPEX	Ratio of capital expenditures to total assets
CASH.RESERVES	Ratio of cash and short-term investments to total assets
DEBT.RATIO	Ratio of total debt to total assets
DIV.PAYOUT	Dividend payout ratio
SIZE	Natural logarithm of total assets
Corporate governance and internal policies
COMP.C.INDEP	Dummy variable equal to one if the percentage of independent board members on the board compensation committee is lower than 50%
COMP.C.NONEX	Dummy variable equal to one if the percentage of nonexecutive board members on the board compensation committee is lower than 50%
E.TURNOVER	Percentage of employee turnover reported by the company during the given period
EX.COMP.CONTR	Dummy variable equal to one if during the given year, the company was covered by media because of controversies related to the remuneration of the executive or board members
GAP	Wage gap calculated as the natural logarithm of the ratio of CEO’s (or the highest) salary within the company to the average salary and benefits
GAP.HIGH.X	Dummy variable equal to one if the salary gap (defined by the variable GAP) within the given company is higher than the Xth percentile for the research sample
GAP.LOW.X	Dummy variable equal to one if the salary gap (defined by the variable GAP) within the given company is below the Xth percentile for the research sample
GAP.LOW.X.Y	Dummy variable equal to one if the salary gap (defined by the variable GAP) within the given company is between the Xth and Yth percentiles for the research sample
INT.PROMOTION.D	Dummy variable equal to one if the company’s reports contain a policy statement favoring internal promotion of employees/promotion from within
MGM.DEPARTURE	Dummy variable equal to one if during the given year, an important member of the supervisory or management board announced voluntary departure (apart from retirement) or was displaced
MGM.TRAINING.D	Dummy variable equal to one if the company claims to organize regular staff trainings for its management
POL.CAREER.DEV	Dummy variable equal to one if the company claims to implement an internal policy aimed at improving/facilitating the career development of its employees
POL.DIVERS.OPP	Dummy variable equal to one if the company claims to implement an internal policy aimed at driving diversity and equal career opportunities for its employees
UNION.DUMMY	Dummy variable equal to one if the percentage of company’s employees represented by an independent labor union or covered by collective bargaining agreements is higher or equal than 50%, zero otherwise
WAGE.CONTR	Dummy variable equal to one if during the given year, the company was covered by media because of controversies related to or initiated by the company’s employees, contractors or suppliers (including wage, layoff or working condition controversies)
Country-level controls
GDP.GROWTH	GDP growth rate reported by the World Bank
GINI	Gini coefficient

presents the complete list of variables used for quantitative analysis.

In the first stage of the study, we verify whether the scale of wage disparities impacts employee incentives. The existing theoretical framework suggests that inequality may exercise both positive and negative influence on employee behavior. On one hand, inequity aversion may motivate employees to retaliate against the top earners by engaging in economically suboptimal behavior such as shirking and voluntary displacement [6]. The outcomes of ultimatum pay bargaining, which promote or perpetuate immoderate inequality, may be perceived as a violation of the tenets of cooperative game and precipitate employees to quit or reduce their output [7]. If inequity aversion prevails, wage gap should reduce employee productivity, increase staff turnover and potentially cause instances of workplace disobedience. Alternatively, wage disparities could be perceived as a potential reward in a competitive process. The promise of future payoffs may motivate employees to exercise a diligent effort aimed at maximizing the chances of upward movement within the corporate hierarchy. Within this framework, only the rank within the hierarchy counts, thereby potentially diminishing the role of merit-based remuneration tied to individual productivity. If ultimately the competitive motive prevails, the empirical analysis should reveal a positive impact of wage gap on employee productivity and turnover.

Our results are consistent with the inequity aversion theory. Wage disparities measured as a relation of CEO remuneration to that of an average employee are documented to persistently exercise a negative impact on employee productivity and turnover. We separately diagnose and analyze the presence of the tail effect in the studied relationships. We find the negative effect of wage gap to be the most pronounced within the top quartile/decile of the wage inequality distribution, where the scale of inequality grows exponentially. In turn, within the bottom decile/quartile of wage gap distribution, employee productivity is found to be higher than that in the remainder of the sample, while staff turnover is consistently lower than in other firms. The discovered nonlinearities represent a strong argument against extreme cases of wage inequity, which appear to bear detrimental consequences to employee morale and workplace incentives. Further advancing our argument, we check whether wage gap contributes to the probability of workplace controversies. Our dataset tracks yearly newsfeed regarding on-the-job disputes initiated by employees and reported by media outlets. The scale or frequency of such controversies is disregarded for such information was difficult to operationalize. However, we clearly distinguish between disputes aimed at improving working conditions and increasing minimum/average salary on one hand, and at curbing disproportionate remuneration/bonuses of top executives on the other. Salary disparity is found to be a significant independent predictor of workplace controversies regarding both (1) employee remuneration and working conditions; and (2) top executive remuneration.

At the same time, we find no evidence suggesting that wage gap may contribute to the retention of corporate executives. While undoubtedly being a symptom of managerial entrenchment [8], wage disparities are found to exercise no significant impact on the probability of top management departures. The study covers both executive and non-executive/supervisory management. Our findings generally corroborate Oyer’s theory [9] of non-incentive remuneration, whereby the executives’ pay is not explicitly tied to any operational/strategic performance measures, which are endogenously impacted by managerial decisions. At the same time, we find no evidence of wage gap impacting the odds of non-executive board members’ departure. The latter finding may suggest that supervisory boards do not attempt to curb the widening wage gap by increasing voluntary turnover.

Next, we verify whether wage gap may inflict deadweight costs on firms. The study documents a generally positive impact of the scale of wage gap on the accounting performance measures. We find that firms with a larger salary disparity between top and average earners exhibit higher profitability ratios and have a better cost structure. In several instances, we report the presence of a tail effect of wage inequality on corporate performance with the top quartile/decile of firms in wage gap distribution recording better profitability and lower cost ratios. The positive impact of inequity on cost efficiency is present both in case of production and sales/general/administrative expenses. These findings allow us to state that, at least to a certain extent, remuneration disparities may be regarded as a price paid by corporate stakeholders for promoting efficiency, yet possibly at the expense of employee discontent.

In view of the discovered double-sided impact of inequities on corporate performance, we also focus on the drivers of wage disparities with the aim of identifying the mechanisms/internal policy actions, which could potentially remedy the negative effects of wage gap on employee incentives without unnecessarily burdening the corporate performance scorecard.

Similarly to [10], we find that wage gaps are driven upwards within relatively larger and better prospering companies, where CEOs enjoy a wider discretionary decision-making power. Tangible manufacturing businesses are more likely to face a higher wage gap. In line with prior studies [11], we find that a higher degree of employee unionization is evidenced to constrain wage inequities. The corporate governance settings are found to play a negligible role in shaping wage disparities. The shares of nonexecutive and independent directors on board are found to exhibit only univariate negative association with the scale of wage gap. In turn, the implementation of internal policies aimed at facilitating career development (e.g., training) and internal promotion of employees were found to endogenously reduce wage disparities. Interestingly, the same policies were found to decrease employee turnover, suggesting that they may play a crucial role in alleviating the negative consequences of the growing inequities.

The findings reported in the paper contribute to the empirical literature in the domain of industrial organization by inquiring into the repercussions of wage disparities for employee incentives and corporate performance scorecards. In contrast with previous studies (e.g., [12]), we show that wage gap negatively impacts employee productivity, staff turnover and the probability of workplace controversies. Importantly, we emphasize the presence of a strong tail effect with the firms reporting extreme wage gaps being more exposed to the negative consequences of inequity. At the same time, the paper exposes the bright side of wage inequality by reporting its beneficial impact on accounting performance of the studied firms. Ultimately, we show that corporate governance settings do not allow curbing wage disparities. Instead, the specially designed corporate policies favoring internal promotion and employee development may positively impact employee incentives and put a restraint on the widening wage gap.

The remainder of the paper is organized as follows. In the next section, we present a concise literature review and derive our research questions therefrom. Section 3 summarizes the applied research methodology and discusses empirical results. The final section concludes.

2. Literature Review and Research Questions

While it is conventionally argued that CEO remuneration should be primarily performance-based, the sharp increase in wage inequities during the recent decades suggests that the intrafirm decision making may be impacted by the rent-seeking behavior on the part of top executives [13]. Whenever the shareholder control dissipates to the point where the supervisory board effectively falls under the influence of top executives, the CEOs’ compensation packages may grow independently from corporate performance whether measured by accounting or market indicators [14]. While the rent-maximizing behavior of CEOs represents a serious governance issue, it remains unclear whether the widening wage gap bears any consequences for employee behavior.

Inequity aversion theory suggests that as employees start perceiving the CEO’s pay as excessively high, they may undertake discretionary remedial action aimed at decreasing the scale of inequity [15]. Disproportionate remuneration may be regarded as a breach of the rule of a cooperative game [16], thus triggering punitive actions, which despite reducing individual utility, serve the goal of reverting to a more equal repartition of output [6,17]. The egalitarian considerations have also been shown to override the reciprocity principle [18], suggesting that even when an individual is well-off and is generally benefiting from the interaction, inequality may cause misalignment of incentives thereby entailing a possible retaliatory action aimed at ameliorating inequity. Hence, even when well-paid, employees may be primarily concerned with their relative payoff and may, therefore, exhibit discontent with the prevailing inequity. The punitive actions undertaken by employees in response to the widening wage gap may take the form of disobedience, strikes, workplace controversies, shirking manifested through productivity decline and quitting.

The validity of inequity aversion theory is supported by a number of empirical studies including those relying on natural experiments. Notable examples include an examination of wage disparities between public and private-sector jobs [19]. Due to relatively lower wages and lower career mobility, the public sector across many developed countries is chronically understaffed and suffers from inadequate access to the pool of available labor. Importantly, because of relatively slower career progress [20], tournament incentives are also evidenced to be weaker in the public sector, thereby further strengthening the sense of perceived inequities. Another natural experiment took place during the COVID-19 pandemic: executives were widely expected to take a pay cut [21] in order to demonstrate solidarity with employees in difficult times even though managing the firm through the turmoil of the pandemic could potentially justify a compensation premium.

A narrower strand of literature in the domain of economic incentives [22] argues that employees tend to compare themselves to the average rather than the top earners when evaluating their relative well-being. By this token, the ratio of CEO pay to the average rank-and-file employee’s salary would matter less for the assessment of perceived level of inequity than dispersion around the mean. However, the public outcry over top executives’ pay [23] and the increasing level of wage inequity manifesting through organized employee actions suggest that top earners remain one of the key benchmarks when assessing the degree of perceived salary gap in the corporate sector. In fact, in survey studies [24,25], people tend to overestimate the share of income going to the top earners possibly because of the perceived growing wage gap. The negative publicity of extreme wage disparities may be partially assuaged through non-disclosure of wages [26]; however, information leakages and reports by non-governmental agencies eventually bring the disparities to light. The negative publicity may be further strengthened by evidence of compensation being partially attributable to non-performance-based factors [27,28].

While potentially engendering employee discontent through infringement on innate egalitarian principles, wage gap is frequently referenced as a necessary precondition for workplace competition. Ref. [29] note that only a large disparity in prizes may induce individuals to actively engage in competitive tournaments. A significant inequality of outcomes may thus be essential for inducing the competing individuals to invest an effort aimed at payoff maximization. A reduction in payoff differential accompanied with mitigation of potential losses under negative tournament outcomes may induce a uniform reduction in the level of effort on the part of tournament participants [29]. Therefore, if individual productivity is set to be the key criterion of workplace promotion with the wage gap being perceived as a reward for the best performing participant, an increase in salary disparities should induce employees to exercise a better effort in ameliorating their performance ceteris paribus. It is worth noting, however, that the necessary precondition for the wage gap to exert an incentivizing effect on employee behavior is that employees have direct control over the factors that underlie the outcome of the tournament. The tournament effect has some grounding in the extant empirical literature, e.g., [30]: firms with larger wage disparities (driver primarily by contingent remuneration in the form of employee stock options) tend to perform better in some aspects.

The opposing viewpoint, e.g., [31], posits that extreme wage gaps do not necessarily lead to higher engagement on the part of top earners. While a high compensation may encourage executives to be more proactive and risk-taking in their roles, thus possibly increasing the amount of value-added they produce, there is a certain saturation point, after which additional compensation does not allow for any gains in productivity. Having exhausted their time and effort, executives may be unable to put in any more effort. Thus, caps on executive compensation set at a specific point of maximum productivity would be economically justified.

Inequity aversion and tournament effects work in opposite directions with the ultimate repercussions of wage gap being determined by the relative strengths of the two. Therefore, wage gap may have either a positive or negative impact on employee incentives depending on the prevailing effect. Our first research question is thus formulated as follows:

RQ1: Does wage gap exercise a positive or a negative impact on employee incentives?

Despite the importance of the problem in question, there are relatively few empirical studies attempting to elucidate it and frequently yielding ambiguous results. Supporting the tenets of inequity aversion theory, Ref. [32] documents a negative impact of wage disparities on the voluntary turnover of senior management. Vice-presidents are found to be more likely to quit if their remuneration is lower than that of their peers within and outside the firm. Ref. [11] report no negative effect of wage gap on employee incentives. Supporting the predictions of tournament theory, salary disparities are found to improve productivity by supposedly instigating tournament competition within smaller and less unionized firms. Ref. [33] attempted to merge the two effects within one model and predicted an inverted U-shaped relationship between salary gap and employee productivity. If the latter relationship holds, one may infer the existence of an optimal level of wage gap, which maximizes employee productivity. A positive correlation is commonly observed between executive compensation and firm performance [34]. The principal problem of such studies resides in the lack of possibility to reliably establish the presence and direction of causality. In fact, some studies find that higher variable compensation (based on equity instruments such as options) and contingent compensation may create wrong incentives by increasing the likelihood of accounting irregularities, whose purpose may sometimes be a maximization of contingent remuneration [35]. Firms, where CEOs receive a disproportionate share of compensation in the form of options, tend to initiate larger share buybacks [36], possibly in order to increase share price and maximize the variable component of their own compensation. This argument is strengthened by the fact that the variable component of executive compensation is more frequently tied to the short-term performance of firms’ stock rather than to the fulfilment of strategic goals [37]. Therefore, the evidence on the relationship between executive pay and firms’ performance is mixed [38,39].

A number of empirical studies, e.g., [40], demonstrate that employee productivity is contingent upon the adequacy of compensation mechanisms and correlation between employees’ productivity and pay. A perceived insufficiency or inadequacy of compensation mechanisms may cause incidents of shirking or increased staff turnover, all of which may have negative ramifications for the business. In particular, the quality of the services or products offered by the business may deteriorate, while operational security may degrade because of increased incidents of negligence or organized employee actions.

We contribute to the ongoing debate regarding the impact of wage gap on employee incentives by adding the following elements to our empirical analysis. First, in addition to exploring the productivity effects of wage gap, we inquire into the possible nonlinearities of the productivity–pay relationship. Secondly, we explore the influence of wage inequity on staff turnover. In contrast to the study by [41], we supplement the data on senior management departures with the conventional measures of rank-and-file employee turnover. Thirdly, we document the associative links between the scale of wage disparities and the odds of workplace controversies. The suggested analytical design should provide a more comprehensive insight into the potential distortionary impact of wage gap on employee incentives.

Despite possibly generating employee discontent, the widening wage gap is frequently justified by the need to retain and appropriately reward scarce talents of the top management [42]. Since CEOs are held accountable for the financial performance of the firms, it is reasonable to assume that their effort will ultimately be centered on the optimization thereof. The most successful CEOs should subsequently see their pay increase in response to the improvement of the bottom line. Executive compensation may also be boosted through managers’ entrepreneurial activities [43,44] and their ability to adapt to challenging operational environments. Yet, the evidence on the relationship between wage gap and corporate performance is scarce and ambiguous.

Ref. [45] demonstrates that CEO talents are not in fact significantly dispersed with their managerial capabilities being relatively homogenous. However, the slight differences in stewardship talents are further amplified manifold by the firm size and by the increasing average firm size in the overall economy. Therefore, the widening pay gap may in fact reflect the economywide trend for consolidation, which endows top managers with increasing amounts of resources. Alternatively, it might well be the case that high remuneration is set to constrain the risk-seeking behavior of CEOs, who may engage in excessively risky projects if their salary is linked to corporate performance [46].

Opposing the merit-based view of pay gap, Ref. [5] signals that executive remunerations may in fact be unrelated to the performance scorecard but rather embody the inefficiency of the corporate governance systems. Benefiting from lack of oversight on the part of a dispersed shareholder base and of an insufficiently diligent board, top managers may conceal the growing disproportion between the dynamics of their bonuses and corporate performance. This type of behavior may be qualified as rent extraction [13]. Consequently, unjustified remuneration premia may contribute to the deterioration of corporate performance and cause a redistribution of value to the top management at the expense of shareholders. Simultaneously, wage gaps may inflict additional costs on the firm by engendering employee discontent and causing productivity declines.

In the presence of multiple explanations of the pay–performance relationship, we formulate the following research question:

RQ2: How does wage gap influence corporate performance?

Kaplan and Rauh [47] justify the rising inequality by the scale of technological change, which enabled talented people to implement their ideas on a global scale thereby significantly increasing the potential payoff. The wage inequities may thus represent the price the firm has to pay for attracting and retaining the most talented executives, which are more likely to make a positive contribution to the firm’s performance. The study by [48] documents that the rising wage gap is rather the consequence of the markets putting a higher value on the managerial capabilities of top executives. The firms that reported a higher internal wage inequity were shown to exhibit better operating performance and higher shareholder returns. The discovered effect was shown to be more pronounced in the more competitive industries where retention of top executives is more likely to be a policy priority. Similar results were reported by [11], who found a positive impact of wage inequities on firms’ operating performance. The rank-and-file employees are not always benefiting from the performance improvement with the wage gap frequently being fueled by a reduction or stagnation of the incomes of low-paid workers [49].

Several studies point to the possible negative impact of wage gap on corporate performance. The agency problem aggravated by the inefficiency of corporate oversight mechanisms may produce excessive executive compensations, which may exercise a negative impact on shareholder value and financial scorecard [50]. The empirical results from the nonprofit institutions, where the agency problems may be particularly acute, corroborate this intuition: the wage gap is reported to be negatively associated with the quality of corporate governance and operating performance [51]. The negative effects of wage disparities observed at the firm level may translate into wider macroeconomic effects such as slower economic growth, lower investments and slower salary growth [52,53]. Therefore, it is crucial to understand the transmission channels through which the said negative impact materializes at the level of individual economic agents. Deszo et al. [54] also demonstrate that growing salary gap is associated with deteriorating tax compliance.

Contributing to the ongoing discussion of the performance–pay relationship, we supplement the analysis of the conventional operating performance measures with a study of the impact of wage gap on firms’ cost structure. We attempt to verify whether a higher CEO-to-employee pay multiple is associated with a more effective cost management and hence with a higher operational profitability. The ultimate goal of the study is to compare the possible benefits stemming from efficient stewardship by highly remunerated executives against the implicit costs inflicted by employee resentment and misaligned incentives.

Finally, having established and quantified the impact of wage inequities on employee incentives and corporate bottom line, we intend to verify whether specific internal organizational settings may contribute to the scale of pay gap. We depart from the assumption that wage gap is inherently shaped by discretionary managerial decisions, which influence the bargaining power of top management with respect to other employees [11], supervisory bodies [55] and unions [56,57]. Further building upon this premise, we formulate the following research question:

RQ3: What are the organizational determinants of wage gap?

In addition to exploring the impact of corporate governance and employee unionization on the scale of wage gap, we study the role which employee policies may play in aggravating/alleviating salary inequities and employee incentives. Our analysis starts with internal promotions and career development policies. During the last few decades, the share of CEOs, who are recruited outside the companies they are supposed to manage, has increased significantly [58]. Therefore, as the external employment opportunities of top executives are expanding, their bargaining power may increase and induce firms to reward CEOs even for economywide performance effects, which CEOs do not directly control [59,60]. Additionally, external candidates may negotiate a premium to their remuneration stemming from the risks related to steering an unfamiliar company. A consistent implementation of a corporate policy favoring internal promotion and assuring a competitive internal selection process may thus curb the scale of wage inequity and contribute to the alignment of incentives of employees with those of the top management. Similar results may be achieved by enforcing a commitment to employee development and by providing regular in-house employee trainings. Both may positively contribute to the enhancement of the internal competitive process and assure a more active employee participation in the internal rank tournaments [4], which in turn may reduce the disincentivizing impact of wage gap. We also examine the possible role of capture of the board’s remuneration committee on the level of wage disparities. We collect the data on remuneration committees’ composition and analyze its impact on the scale of wage gap.

By exploring the organizational determinants of wage gap, we intend to come up with guidelines regarding the alleviation of the possible negative repercussions of the growing wage disparities. Since salary inequities may bear both positive and negative consequences for corporate performance, decision makers may be particularly interested in elaborating and implementing remedial tools allowing to channel wage disparities towards maximization of the benefits accruing to the key stakeholders.

3. Dataset

The present study relies on a novel firm-level database covering 1709 firms observed during the period between 2003 and 2018. The principal source of firm-level financial and governance data (including executive compensation) is Refinitiv Eikon. The essential problem we had to cope with at the stage of data collection was the limited availability of publicly available records of CEO remunerations. Only several countries mandate compulsory disclosure of top executive salaries and average employee wages. Other firms, which are not constrained to disclose the wage disparities by the local regulatory framework, may provide the data regarding the wage gap upon their discretionary decision. Due to the sensitive nature of the analyzed data, the final research sample is limited to forty countries. The geographical distribution of the sample is as follows: 54% of firms come from North and South America, where the data on executive compensation is the most available; 24.3% of firms are from Europe; 11.2% are from Asia; 7.2% represent Australia and New Zealand; 3.4% are from Africa.

The sample selection was dictated by the study’s focus on employee incentives, productivity and industrial organization. Therefore, we completely excluded financial firms from the research sample. Had we not done so, the majority of firms in our sample would have been from the financial sector, thereby potentially skewing the results, especially in productivity measurement. Many firms, which disclosed the CEO’s pay, did not provide the average employee remuneration, which caused them to be ejected from the research sample.

Initially, we analyzed the CEO-to-employee salary relation for every firm in the sample. Having screened the sample structure and the descriptive statistics for salary gap, we noticed that the distribution of the newly created variable was non-normal. Within the highest-ranked quartile of the distribution, the value of salary gap grew exponentially to levels significantly above the population average (see Figure 1). The CEO-to-employee salary ratio rapidly increases from 154 times in the 70th percentile to 340 and 1309 in the 80th and 90th percentiles, respectively. In view of the fact, that these cases of extreme wage disparities have not previously been closely studied, we decided to verify, whether the top quartile of firms exhibited any inherent empirically observable features. In turn, in order to eliminate the potentially distortionary impact of outliers on the econometric results, we normalized the CEO-to-employee relation by calculating the natural logarithm thereof, thereby creating the key explanatory variable GAP reflecting the scale of wage disparities within a given company.

The explained variables used in the empirical analysis of employee incentives include the conventional operating performance measures, e.g., return on assets or net profit margin, as well as productivity measures. The latter include the natural logarithm of the value of sales per employee [11] and EBITDA per employee. The latter should primarily capture the impact of cost structure, which is directly influenced by employees’ decisions and actions, on firms’ bottom line. The financial data were collected from Thomson Reuters Database.

The data regarding employee turnover were sparsely available. Predominantly, firms are not obliged to disclose these figures and may only choose to do so. Only a fraction of sampled firms published these figures thereby possibly exposing the study to self-selection bias. Similar studies focusing on the problem of wage gap also suffer from this limitation and acknowledge it appropriately. Despite the limited data availability, the number of observations (2529 firm-years representing 461 firms) is still large enough to allow for valid statistical inference. The turnover figures concern only regular full-time employees. In order to study the impact of wage gap on the turnover of senior management, we collect data regarding the departures of members of executive and supervisory boards. The resulting binary variable (MGM. DEPARTURE) takes on the value of one if during a year the given company reported a voluntary termination of either a member of supervisory or executive board. The number of top executives leaving the firm during a given year was not taken into consideration for the resulting variable would not manifest sufficient heterogeneity due to the rarity of top management departures.

The data regarding employee disobedience was organized in the form of two binary variables reflecting the instances of (1) employee wage and working condition controversies (WAGE.CONTR), and (2) executive remuneration controversies (EX.COMP.CONTR). The variables take on the value of one if, during a given year, the firm was covered in the media because of the above-mentioned controversies.

Finally, in order to study the organizational determinants of wage gap, we collected the data regarding internal corporate policies disclosed in CSR reports. In particular, we were interested in whether the analyzed companies implemented separate targeted policies favoring (1) internal promotion and (2) career development of employees as well as (3) assuring diversity and equal opportunity. As a result, we obtained three binary variables reflecting the declaration of the company’s commitment to the implementation of the above enumerated internal policies. We would like to emphasize that these variables only reflect the firms’ declarations in the CSR reports. The actual state of implementation of these policies may not be assessed accurately. Similarly, we collect data regarding the quality of corporate governance and the internal level of employee unionization. In particular, we gather data regarding the share of non-executive and independent members on the remuneration committee of the supervisory board and the percentage of unionized employees. Relying on the nominal variables, we binary-code firms with the highest level of unionization and the highest quality of corporate governance (UNION.DUMMY, COMP.C.INDEP, COMP.C.NONEX).

4. Methodology

When studying the impact of wage gap on corporate performance, we rely on static panel regression analysis with random effects. The baseline econometric specification underlying the analysis may be stated as follows:

$$C{P}_{it}={\beta }_0+{\beta }_{1}GA{P}_{it}+{\beta }^{\prime }CONTRO{L}_{it}+{\epsilon }_{it}$$

(1)

where $C{P}_{it}$—corporate performance indicators including employee productivity, operating performance measures, and employee turnover reported for the ith company in year t; $GA{P}_{it}$—wage gap measured as the natural logarithm of CEO-to-employee ratio; $CONTRO{L}_{it}$—the vector of control variables.

In order to study the possible non-linearities of the performance–inequity relationship, we create a set of dummy variables, which encode firms representing specific subsamples of the wage gap distribution. Variables GAP.LOW.X are binary variables encoding the subsample of firms situated below the Xth percentile in the distribution of wage gap. In contrast, variables GAP.HIGH.Y encode the subsamples of firms situated above the Yth percentile of the wage gap distribution. The suggested design of binary variables should allow for identification of intra-sample differences between high-inequality and low-inequality firms, which in turn may either corroborate or disprove the presence of a tail effect in the performance–inequity relationship.

The vector of control variables includes firm size (SIZE) approximated by natural logarithm of firms’ total assets, level of indebtedness (DEBT.RATIO) measured as debt-to-assets ratio, liquidity (CASH.RESERVES) measured as ratio of cash and short-term investments to total assets, asset tangibility (ASSET.TANG) defined as ratio of Net property, plant and equipment to total assets, profit margins, dividend payout ratio (DIV.PAYOUT) and investment demand (CAPEX) specified as a ratio of capital expenditures to the contemporaneous value of total assets.

Table 2. Descriptive statistics for selected variables.

Variable	Mean	Median	St.Dev.	Min	Max
SIZE	22.1227	22.1363	1.6875	14.9795	27.5565
DEBT.RATIO	0.2590	0.2397	0.1903	0.0000	0.9960
CAPEX	0.0583	0.0425	0.0543	0.0002	0.2794
EBIT.MARGIN	0.1330	0.1220	0.2173	−0.8062	0.7450
EBITDA.MARGIN	0.2094	0.1843	0.2457	−0.9007	0.8273
NET.PROFIT.M	0.0841	0.0772	0.2368	−0.8728	0.9469
SGA.EXP	0.2225	0.1747	0.1910	0.0122	0.9148
CASH.RESERVES	0.1422	0.0866	0.1612	0.0005	0.8208
ASSET.TANG	0.3071	0.2277	0.2625	0.0006	0.9345
E.TURNOVER	0.1237	0.1010	0.0952	0.0000	0.9981
DIV.PAYOUT	0.3558	0.2553	0.3724	0.0000	1.9488

contains the descriptive statistics for selected variables. The country-level controls, which were included because of the cross-border composition of the sample, include GDP growth (GDP.GROWTH), which should reflect the stage of business cycle and Gini coefficient (GINI), which controls for the cross-country differences in income inequality. Each model also includes time and industry dummies.

In order to investigate the impact of wage gap on the probability of employee action, we run binary logit regressions of the following specification:

$$logit\left(p\right)={\beta }_0+{\beta }_{1}GA{P}_{it}+{\beta }^{\prime }CONTRO{L}_{it}$$

(2)

where the explained variables are binary-coding the occurrence of (1) wage and working condition controversies (WAGE.CONTR) initiated by employees; (2) executive compensation controversies (EX.COMP.CONTR); and (3) management departures (MGM.DEPARTURE).

Finally, the study of organizational determinants of wage gap relies on the following econometric specification:

$$GA{P}_{it}={\beta }_0+{\beta }_{1}PO{L}_{it}+{\beta }_{2}GO{V}_{it}+{\beta }^{\prime }CONTRO{L}_{it}+{\epsilon }_{it}$$

(3)

where $PO{L}_{it}$ represents a set of dummy variables describing the internally implemented corporate policies including (1) policy favoring internal promotion (INT.PROMOTION.D); (2) policy favoring career development of employees (POL.CAREER.DEV); (3) policy favoring diversity and equal opportunity at the workplace (POL.DIVERS.OPP); (4) policy of internal provision of on-the-job trainings to the company’s management (MGM.TRAINING); $GO{V}_{it}$ represents a set of dummy variables describing the internal corporate governance practices such as (1) COMP.C.INDEP encoding the fact of the company’s remuneration committee being in at least 50% staffed with independent directors; (2) COMP.C.NONEX reflecting the fact of the company’s remuneration committee being in at least 50% staffed with non-executive directors. The variables describing the composition of the board’s remuneration committee should reflect the possible degree of the committees’ capture by CEOs. Specification (3), similarly to specification (1), is tested using random-effect static panel regression analysis with time and industry dummies.

5. Empirical Results

5.1. The Impact of Wage Gap on Employee Incentives

In this subsection, we attempt to answer the first research question relying on the principal empirical findings obtained from the quantitative analysis.

Table 3. Specification (1)–The impact of wage gap on employee productivity.

Panel A. Dependent variable: natural logarithm of revenue per employee (PROD.REVENUE)
Model No	1		2		3		4		5
Constant	10.298	***	10.220	***	10.159	***	10.339	***	10.310	***
	(0.274)		(0.276)		(0.276)		(0.277)		(0.277)
SIZE	0.072	***	0.063	***	0.065	***	0.061	***	0.059	***
	(0.010)		(0.010)		(0.010)		(0.010)		(0.010)
DEBT.RATIO	−0.090	**	−0.109	***	−0.114	***	−0.109	***	−0.106	***
	(0.042)		(0.042)		(0.042)		(0.042)		(0.042)
CAPEX	1.068	***	1.072	***	1.089	***	1.079	***	1.085	***
	(0.122)		(0.122)		(0.122)		(0.122)		(0.122)
DIV.PAYOUT	−0.030	***	−0.026	***	−0.029	***	−0.028	***	−0.028	***
	(0.008)		(0.008)		(0.008)		(0.008)		(0.008)
ASSET.TANG	−0.333	***	−0.342	***	−0.345	***	−0.349	***	−0.342	***
	(0.053)		(0.053)		(0.053)		(0.053)		(0.053)
CASH.RESERVES	0.268	***	0.281	***	0.292	***	0.271	***	0.283	***
	(0.060)		(0.060)		(0.060)		(0.060)		(0.060)
GINI	0.024	***	0.024	***	0.024	***	0.024	***	0.024	***
	(0.002)		(0.002)		(0.002)		(0.002)		(0.002)
GDP.GROWTH	0.009	***	0.009	***	0.009	***	0.008	***	0.008	***
	(0.003)		(0.003)		(0.003)		(0.003)		(0.003)
GAP	−0.046	***
	(0.004)
GAP.LOW.10			0.133	***
			(0.017)
GAP.LOW.25					0.105	***
					(0.013)
GAP.HIGH.25							−0.098	***
							(0.015)
GAP.HIGH.10									−0.099	***
									(0.020)
no. of observations	9087		9112		9112		9112		9112
Wald (joint)	529.52	***	424.86	***	430.33	***	404.81	***	383.80	***
R2	0.77		0.76		0.76		0.76		0.76
AR(1) test	24.27	***	24.07	***	23.75	***	23.65	***	23.62	***
AR(2) test	9.90	***	9.98	***	9.99	***	10.11	***	10.08	***
Panel B. Dependent variable: natural logarithm of EBITDA per employee (PROD.EBITDA)
Model No	1		2		3		4		5
Constant	6.874	***	6.799	***	6.723	***	6.895	***	6.888	***
	(0.335)		(0.335)		(0.336)		(0.336)		(0.337)
SIZE	0.192	***	0.188	***	0.191	***	0.185	***	0.184	***
	(0.013)		(0.013)		(0.013)		(0.013)		(0.013)
DEBT.RATIO	−0.165	***	−0.184	***	−0.190	***	−0.190	***	−0.192	***
	(0.064)		(0.064)		(0.064)		(0.064)		(0.064)
CAPEX	2.263	***	2.280	***	2.304	***	2.298	***	2.306	***
	(0.186)		(0.186)		(0.186)		(0.186)		(0.186)
DIV.PAYOUT	−0.036	***	−0.032	***	−0.036	***	−0.034	***	−0.034	***
	(0.013)		(0.013)		(0.013)		(0.013)		(0.013)
ASSET.TANG	−0.264	***	−0.267	***	−0.270	***	−0.271	***	−0.270	***
	(0.077)		(0.076)		(0.076)		(0.077)		(0.077)
CASH.RESERVES	0.820	***	0.828	***	0.845	***	0.833	***	0.840	***
	(0.092)		(0.092)		(0.092)		(0.092)		(0.092)
GINI	−0.008	***	−0.008	***	−0.008	***	−0.008	***	−0.008	***
	(0.003)		(0.003)		(0.003)		(0.003)		(0.003)
GDP.GROWTH	0.004		0.004		0.005		0.004		0.004
	(0.005)		(0.005)		(0.005)		(0.005)		(0.005)
GAP	−0.031	***
	(0.005)
GAP.LOW.10			0.151	***
			(0.026)
GAP.LOW.25					0.111	***
					(0.019)
GAP.HIGH.25							−0.052	**
							(0.022)
GAP.HIGH.10									−0.026
									(0.030)
no. of observations	8587		8611		8611		8611		8611
Wald (joint)	443.32	***	449.57	***	448.03	***	419.12	***	414.17	***
R2	0.72		0.72		0.71		0.71		0.71
AR(1) test	32.47	***	32.25	***	32.35	***	32.19	***	32.13	***
AR(2) test	7.61	***	7.53	***	7.59	***	7.40	***	7.40	***

Notes: the table presents the GLS (generalized least squares) random-effect static panel model estimates. All models include the time and industry dummies (not reported). The heteroscedasticity robust standard errors are provided in parentheses. ***, ** indicate significance at the 1% and 5% levels, respectively.

contains the results of regression analysis inquiring into the impact of wage gap on employee productivity. Panel A of the table presents the results of the tests of specification (1) with the regressand being the natural logarithm of sales per employee. In Panel B, the explained variable is natural logarithm of EBITDA per employee. In both cases, wage gap is shown to exercise a statistically significant negative impact on employee productivity. The respective regression coefficients at the variable GAP are statistically significant at 1% significance level. In both panels, one can observe an explicit non-linearity in productivity–wage gap relationship. The decile of firms with the lowest degree of pay inequity (GAP.LOW.10) is demonstrated to exhibit relatively higher productivity compared to the remainder of the research sample. In contrast, the quartile of firms reporting the highest degree of wage disparities (GAP.HIGH.25) are evidenced to have a relatively lower employee productivity measured by both logs of sales per employee and EBITDA per employee. The negative impact of wage gap on sales per employee ratio further strengthens in the last decile of the wage gap distribution suggesting that employee discontent is a growing function of wage disparities. Overall, our findings suggest that within non-financial companies, salary inequity may bear detrimental effect on employee productivity.

Further expanding our analysis, we look into the impact of wage gap on employee turnover. Our results presented in

Table 4. Specification (1)–The impact of wage gap on employee turnover.

Panel A. Dependent variable: employee turnover (E.TURNOVER); independent variables encode the tails of the distribution of salary gap (GAP)
Model No	1		2		3		4		5
Constant	18.320	***	12.625	*	13.573	**	11.864	*	12.012	*
	(6.744)		(6.687)		(6.690)		(6.669)		(6.682)
SIZE	0.434	*	0.518	**	0.479	*	0.513	**	0.536	**
	(0.246)		(0.247)		(0.247)		(0.246)		(0.247)
DEBT.RATIO	3.809	***	3.914	***	3.963	***	3.785	***	3.877	***
	(1.396)		(1.402)		(1.401)		(1.402)		(1.404)
CAPEX	4.505		4.017		3.755		3.999		3.708
	(3.959)		(3.972)		(3.973)		(3.968)		(3.972)
ASSET.TANG	0.719		0.819		0.843		0.926		0.820
	(1.253)		(1.259)		(1.259)		(1.259)		(1.260)
CASH.RESERVES	4.740	**	4.771	**	4.573	**	4.718	**	4.621	**
	(2.149)		(2.159)		(2.158)		(2.157)		(2.159)
DIV.PAYOUT	−0.459	**	−0.507	**	−0.477	**	−0.489	**	−0.505	**
	(0.214)		(0.215)		(0.216)		(0.215)		(0.215)
EBIT.MARGIN	−8.208	***	−8.438	***	−8.424	***	−8.554	***	−8.525	***
	(1.179)		(1.180)		(1.181)		(1.179)		(1.181)
GINI	−0.012		−0.012		−0.011		−0.014		−0.012
	(0.038)		(0.038)		(0.038)		(0.038)		(0.038)
GDP.GROWTH	0.015		0.023		0.024		0.028		0.030
	(0.091)		(0.091)		(0.091)		(0.091)		(0.091)
GAP	0.260	***
	(0.099)
UNION.DUMMY	0.026		0.122		0.143		0.159		0.139
	(0.401)		(0.402)		(0.402)		(0.401)		(0.402)
GAP.LOW.10			−1.015	**
			(0.452)
GAP.LOW.25					−0.705	**
					(0.358)
GAP.HIGH.25							1.166	***
							(0.409)
GAP.HIGH.10									0.792
									(0.603)
no. of observations	2525		2529		2529		2529		2529
Wald (joint)	80.22	***	84.33	***	83.08	***	87.48	***	80.88	***
R2	0.19		0.18		0.19		0.19		0.18
AR(1) test	12.95	***	12.81	***	12.93	***	12.92	***	12.92	***
AR(2) test	0.62		0.30		0.40		0.36		0.37
Panel B. Dependent variable: employee turnover (E.TURNOVER); independent variables encode internal employment policies.
Model No		6			7			8
Constant		16.959		**	16.890		**	17.424		***
		(6.776)			(6.751)			(6.770)
SIZE		0.504		**	0.507		**	0.471		*
		(0.249)			(0.247)			(0.248)
DEBT.RATIO		3.655		***	3.841		***	3.568		**
		(1.397)			(1.394)			(1.404)
CAPEX		4.525			4.275			4.327
		(3.956)			(3.952)			(3.958)
ASSET.TANG		0.625			0.576			0.713
		(1.254)			(1.252)			(1.253)
CASH.RESERVES		4.660		**	4.723		**	4.819		**
		(2.148)			(2.145)			(2.149)
DIV.PAYOUT		−0.476		**	−0.457		**	−0.455		**
		(0.214)			(0.214)			(0.214)
EBIT.MARGIN		−8.288		***	−8.304		***	−8.306		***
		(1.178)			(1.177)			(1.180)
GINI		−0.011			−0.014			−0.012
		(0.038)			(0.038)			(0.038)
GDP.GROWTH		0.013			0.009			0.015
		(0.090)			(0.090)			(0.090)
GAP		0.253		***	0.262		***	0.261		***
		(0.099)			(0.099)			(0.099)
UNION.DUMMY		0.014			0.007			0.048
		(0.401)			(0.400)			(0.401)
POL.CAREER.DEV		−0.884		**
		(0.437)
POL.DIVERS.OPP					−1.340		***
					(0.432)
INT.PROMOTION.D								−0.508
								(0.320)
no. of observations		2525			2525			2525
Wald (joint)		84.38		***	90.11		***	82.75		***
R^2		0.19			0.20			0.19
AR(1) test		12.82		***	12.88		***	12.85		***
AR(2) test		0.57			0.62			0.60

Notes: the table presents the GLS random-effect static panel model estimates. All models include the time and industry dummies (not reported). The heteroscedasticity robust standard errors are provided in parentheses. ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively.

(explained variable–E.TURNOVER) clearly demonstrate that a higher wage gap contributes to an increased staff turnover. Likewise, we document a strong tail effect of wage gap on the analyzed explained variable (Panel A in Table 4). The decile of firms with the lowest wage disparities appears to enjoy the lowest employee turnover in the sample. As the internal inequities grow, staff turnover increases reaching its peak within the top quartile of wage gap distribution. Ultimately, extreme inequality seems to take a toll both on employee productivity and voluntary employment termination.

Our findings suggest that staff turnover may be significantly reduced through the implementation of targeted internal policies favoring employee development and equality of opportunity (Panel B of Table 4). Firms, which claimed to have been systematically implementing the respective policies, are shown to enjoy an overall lower level of staff turnover. The latter suggests the presence of a long-term positive effect stemming therefrom on employee incentives.

Finally, we analyze the impact of wage gap on the odds of internal controversies and managerial termination. The results of logit models framed in specification (2) are reported in

Table 5. Specification (2)–The impact of wage gap on the probability of employee action.

	Wages and Work Condition Controversies (WAGE.CONTR)		Executive Compensation Controversies (EX.COMP.CONTR)		Management Departure (MGM.DEPARTURE)
Model No	1		2		3
Constant	−25.485	***	−21.863	***	−10.959	***
	(1.168)		(2.165)		(0.6722)
CAPEX	0.809		−1.455		2.063	**
	(1.515)		(3.432)		(0.9724)
SIZE	0.885	***	0.776	***	0.378	***
	(0.04318)		(0.08002)		(0.02637)
EBIT.MARGIN	−1.123	***	−0.431		−0.602	***
	(0.3588)		(0.7994)		(0.2219)
DEBT.RATIO	−0.508		−1.301		0.112
	(0.3951)		(0.8412)		(0.2493)
ASSET.TANG	1.887	***	1.143	*	−0.077
	(0.2983)		(0.6042)		(0.2051)
CASH.RESERVES	2.557	***	0.200		0.548
	(0.5380)		(1.294)		(0.3401)
DIV.PAYOUT	−0.244	**	−0.476	*	0.102
	(0.1199)		(0.2532)		(0.07828)
GDP.GROWTH	−0.094	***	−0.034		0.023
	(0.02554)		(0.05256)		(0.01866)
GINI	0.023	**	−0.036		−0.0198
	(0.009534)		(0.02303)		(0.006922)
UNION.DUMMY	−0.042		0.172		0.297	***
	(0.1471)		(0.3189)		(0.1126)
GAP	0.073	**	0.112	*	0.012
	(0.03121)		(0.06273)		(0.02218)
Log−likelihood	−1136.70		−374.27		−2281.52
No of observations	9774		9774		9774
Chi2	631.75	***	141.08	***	253.29	***

Note: the table presents the maximum likelihood estimates of a binary logit model. Asymptotic standard errors are reported in parentheses under the coefficients. ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively.

. Overall, wage and working condition controversies have been recorded in 351 firm-years representing 3.4% of the entire sample. Typically, wage controversies occur within relatively larger mostly manufacturing companies with significant liquidity reserves and a relatively higher wage gap (model 1 in Table 5). Similarly, salary inequity is documented to significantly increase the odds of executive compensation controversies (model 2 in Table 5). Interestingly, the level of employee unionization has been shown to have no impact on neither the probability of wage and working condition controversies nor the odds of executive compensation disputes.

When it comes to the probability of management departure defined as voluntary turnover of the members of executive or supervisory boards (model 3 in Table 5), the scale of pay gap is demonstrated to exercise no significant impact thereon. Wage disparities may, therefore, seem to be an inefficient tool for management retention. As the alternative employment opportunities for the top executives and board members are expanding, the relative wage disparity appears to play a minor role in shaping their career choices. In fact, the only variable that has been shown to exercise a significant impact on the odds of management departure (apart from firm size, which is self-explanatory) is the one reflecting the scale of employee unionization. It appears that the higher bargaining power of employees may ultimately result in an increased turnover in the companies’ top management.

5.2. The Impact of Wage Gap on Firms’ Operating Performance

Since our prior evidence suggests a negative association between wage gap and employee productivity, we turn to the operating performance indicators to see whether inequity may erode the corporate bottom line. We rerun model specification (1) with different operating performance indicators substituted for regressands. The results are reported in

Table 6. Specification (1)–The influence of wage gap on financial performance.

Panel A. Dependent variable: EBITDA margin (EBITDA.MARGIN).
Model No	1		2		3		4		5
Constant	−0.615	***	−0.615	***	−0.619	***	−0.618	***	−0.617	***
	(0.065)		(0.065)		(0.065)		(0.065)		(0.065)
SIZE	0.039	***	0.040	***	0.040	***	0.040	***	0.040	***
	(0.002)		(0.002)		(0.002)		(0.002)		(0.002)
DEBT.RATIO	−0.072	***	−0.071	***	−0.071	***	−0.072	***	−0.072	***
	(0.012)		(0.012)		(0.012)		(0.012)		(0.012)
CAPEX	0.299	***	0.300	***	0.299	***	0.301	***	0.300	***
	(0.036)		(0.036)		(0.036)		(0.036)		(0.036)
ASSET.TANG	−0.049	***	−0.048	***	−0.048	***	−0.047	***	−0.048	***
	(0.015)		(0.015)		(0.015)		(0.015)		(0.015)
CASH.RESERVES	−0.042	**	−0.039	**	−0.040	**	−0.036	**	−0.039	**
	(0.017)		(0.017)		(0.017)		(0.017)		(0.017)
GINI	−0.001	**	−0.001	**	−0.001	**	−0.001	**	−0.001	**
	(0.001)		(0.001)		(0.001)		(0.001)		(0.001)
GDP.GROWTH	0.005	***	0.005	***	0.005	***	0.005	***	0.005	***
	(0.001)		(0.001)		(0.001)		(0.001)		(0.001)
GAP	0.003	***
	(0.001)
GAP.LOW.10			−0.006
			(0.005)
GAP.LOW.25					−0.000
					(0.004)
GAP.HIGH.25							0.0181	***
							(0.004)
GAP.HIGH.10									0.008
									(0.006)
no. of observations	9451		9476		9476		9476		9476
Wald (joint)	431.73	***	417.34	***	415.30	***	433.12	***	417.29	***
R^2	0.15		0.15		0.15		0.15		0.15
AR(1) test	26.54	***	26.79	***	26.85	***	26.59	***	26.8	***
AR(2) test	−1.21		−0.80		−0.81		−0.75		−0.83
Panel B. Dependent variable: Net profit margin (NET.PROFIT.M).
Model No	1		2		3		4		5
Constant	−0.561	***	−0.566	***	−0.578	***	−0.567	***	−0.566	***
	(0.071)		(0.071)		(0.072)		(0.071)		(0.071)
SIZE	0.032	***	0.033	***	0.034	***	0.033	***	0.033	***
	(0.003)		(0.003)		(0.003)		(0.003)		(0.003)
DEBT.RATIO	−0.235	***	−0.232	***	−0.231	***	−0.233	***	−0.233	***
	(0.016)		(0.016)		(0.016)		(0.016)		(0.016)
CAPEX	0.451	***	0.452	***	0.452	***	0.454	***	0.453	***
	(0.053)		(0.053)		(0.053)		(0.053)		(0.053)
ASSET.TANG	−0.148	***	−0.147	***	−0.147	***	−0.147	***	−0.147	***
	(0.020)		(0.020)		(0.020)		(0.020)		(0.020)
CASH.RESERVES	−0.107	***	−0.105	***	−0.105	***	−0.102	***	−0.104	***
	(0.023)		(0.023)		(0.023)		(0.023)		(0.023)
GINI	−0.000		−0.000		−0.000		−0.000		−0.000
	(0.001)		(0.001)		(0.001)		(0.001)		(0.001)
GDP.GROWTH	0.009	***	0.009	***	0.009	***	0.009	***	0.009	***
	(0.001)		(0.001)		(0.001)		(0.001)		(0.001)
GAP	0.003	**
	(0.001)
GAP.LOW.10			−0.006
			(0.008)
GAP.LOW.25					0.003
					(0.006)
GAP.HIGH.25							0.013	**
							(0.006)
GAP.HIGH.10									0.012
									(0.008)
no. of observations	9451		9476		9476		9476		9476
Wald (joint)	509.46	***	505.12	***	504.89	***	509.50	***	506.62	***
R2	0.15		0.15		0.15		0.15		0.15
AR(1) test	26.76	***	26.73	***	26.76	***	26.66	***	26.71	***
AR(2) test	−4.72	***	−4.53	***	−4.49	***	−4.52	***	−4.54	***

Notes: the table presents the GLS random-effect static panel model estimates. All models include time and industry dummies (not reported). The heteroscedasticity robust standard errors are provided in parentheses. ***, ** indicate significance at the 1% and 5% levels, respectively.

and

Table 7. Specification (1)–The impact of wage gap on the cost structure.

Panel A. Dependent variable: Sales. general and administrative expenses (SGA.EXP).
Model No	1		2		3		4		5
Constant	0.513	***	0.511	***	0.510	***	0.516	***	0.515	***
	(0.045)		(0.045)		(0.045)		(0.045)		(0.045)
SIZE	−0.016	***	−0.017	***	−0.017	***	−0.017	***	−0.017	***
	(0.002)		(0.002)		(0.002)		(0.002)		(0.002)
DEBT.RATIO	0.006		0.005		0.004		0.005		0.006
	(0.008)		(0.008)		(0.008)		(0.008)		(0.008)
CAPEX	−0.042	*	−0.042	*	−0.041	*	−0.042	*	−0.042	*
	(0.022)		(0.022)		(0.022)		(0.022)		(0.022)
ASSET.TANG	0.001		0.000		0.000		0.000		0.000
	(0.010)		(0.010)		(0.010)		(0.010)		(0.010)
CASH.RESERVES	0.025	**	0.024	**	0.025	**	0.023	**	0.024	**
	(0.011)		(0.011)		(0.011)		(0.011)		(0.011)
DIV.PAYOUT	0.000		0.000		0.000		0.000		0.000
	(0.002)		(0.002)		(0.002)		(0.002)		(0.002)
GDP.GROWTH	−0.001	**	−0.001	**	−0.001	**	−0.001	**	−0.001	**
	(000)		(000)		(000)		(000)		(000)
GAP	−0.004	***
	(0.001)
GAP.LOW.10			0.007	**
			(0.003)
GAP.LOW.25					0.004	*
					(0.002)
GAP.HIGH.25							−0.013	***
							(0.003)
GAP.HIGH.10									−0.013	***
									(0.004)
no. of observations	9961		9987		9987		9987		9987
Wald (joint)	168.56	***	136.03	***	133.22	***	155.57	***	141.69	***
R2	0.18		0.18		0.18		0.18		0.18
AR(1) test	30.38	***	30.6	***	30.65	***	30.31	***	30.48	***
AR(2) test	8.71	***	8.58	***	8.58	***	8.55	***	8.55	***
Panel B. Dependent variable: Cost of goods sold (COGS).
Model No	6		7		8		9		10
Constant	1.643	***	1.641	***	1.639	***	1.647	***	1.646	***
	(0.058)		(0.058)		(0.058)		(0.058)		(0.058)
SIZE	−0.034	***	−0.035	***	−0.035	***	−0.035	***	−0.035	***
	(0.002)		(0.002)		(0.002)		(0.002)		(0.002)
DEBT.RATIO	0.095	***	0.094	***	0.093	***	0.094	***	0.094	***
	(0.012)		(0.012)		(0.012)		(0.012)		(0.012)
CAPEX	−0.535	***	−0.538	***	−0.536	***	−0.537	***	−0.537	***
	(0.034)		(0.034)		(0.034)		(0.034)		(0.034)
ASSET.TANG	0.112	***	0.111	***	0.111	***	0.110	***	0.111	***
	(0.014)		(0.014)		(0.014)		(0.014)		(0.014)
CASH.RESERVES	−0.012		−0.014		−0.013		−0.017		−0.015
	(0.016)		(0.016)		(0.016)		(0.016)		(0.016)
DIV.PAYOUT	−0.003		−0.003		−0.003		−0.003		−0.003
	(0.002)		(0.002)		(0.002)		(0.002)		(0.002)
GDP.GROWTH	−0.003	***	−0.003	***	−0.003	***	−0.003	***	−0.003	***
	(0.001)		(0.001)		(0.001)		(0.001)		(0.001)
GAP	−0.004	***
	(0.001)
GAP.LOW.10			0.009	**
			(0.005)
GAP.LOW.25					0.005
					(0.004)
GAP.HIGH.25							−0.016	***
							(0.004)
GAP.HIGH.10									−0.009	*
									(0.006)
no. of observations	9845		9871		9871		9871		9871
Wald (joint)	592.83	***	576.92	***	575.48	***	589.82	***	576.06	***
R2	0.44		0.44		0.44		0.44		0.44
AR(1) test	35.28	***	35.23	***	35.28	***	35.07	***	35.28	***
AR(2) test	1.93	*	2.02	**	2.01	**	2.07	**	2.01	**

Notes: the table presents the GLS random-effect static panel model estimates. All models include the time and industry dummies (not reported). The heteroscedasticity robust standard errors are provided in parentheses. ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively.

.

In Table 6, the explained variables are EBITDA margin (Panel A) and net profit margin (Panel B). The results in both panels bring us to similar conclusions: wage gap appears to exercise a positive and statistically significant impact on firms’ operating performance. The performance–inequity relationship appears to be impacted by the presence of a tail effect. The profitability ratios are found to be relatively higher in the top quartile of firms from the wage gap distribution.

Delving further into the transmission mechanism intermediating the performance–wage gap relationship, we analyze the potential sources of superior performance of firms reporting the highest wage disparities. Since our earlier findings demonstrated that wage gap may negatively affect employee productivity measured by sales per employee, we turn to the analysis of cost structure hoping to find the source of enhanced efficiency there. Table 7 presents the results of an empirical inquiry into the impact of wage gap on the cost structure of sampled firms. In Panel A, the explained variable is the value of sales/general and administrative expenses (SGA.EXP) scaled by the value of total revenue. In Panel B, the explained variable is cost of goods sold scaled by the value of total revenue (COGS). In many respects, the findings reported in Table 7 represent a mirror reflection of our prior evidence regarding the profit margins. Wage gap is confirmed to have an economically significant beneficial impact on the cost structure. By contributing to cost optimization, it enhances the profit margins. The relationship between costs and wage gap is evidenced to manifest nonlinearity. The bottom 10% of firms with the lowest scale of wage disparities are shown to report the highest cost-to-revenue ratios (both for SGA.EXP and COGS). In contrast, the top quartile of firms from the wage gap distribution report relatively lower cost-to-revenue ratios than the remainder of the sample.

5.3. The Factors Curbing the Wage Gap

In this subsection, we discuss the organizational factors contributing to or constraining the scale of wage gap in the sampled companies. The results of empirical tests of specification (3) are presented in

Table 8. Specification (3)–The organizational determinants of wage gap.

Panel A. Dependent variable: wage gap (GAP); independent variables encode internal employment policies.
Model No	1		2		3		4		5
Constant	−1.113	*	−1.267	**	−1.456	**	−1.441	**	−1.272	**
	(0.580)		(0.579)		(0.582)		(0.582)		(0.579)
SIZE	0.294	***	0.300	***	0.309	***	0.307	***	0.301	***
	(0.022)		(0.022)		(0.022)		(0.022)		(0.022)
DEBT.RATIO	0.712	***	0.702	***	0.698	***	0.700	***	0.702	***
	(0.116)		(0.116)		(0.116)		(0.116)		(0.116)
CAPEX	−0.771	**	−0.795	**	−0.806	**	−0.803	**	−0.795	**
	(0.358)		(0.358)		(0.358)		(0.358)		(0.358)
EBIT.MARGIN	0.468	***	0.465	***	0.470	***	0.471	***	0.470	***
	(0.100)		(0.099)		(0.099)		(0.099)		(0.099)
CASH.RESERVES	−0.524	***	−0.528	***	−0.541	***	−0.534	***	−0.531	***
	(0.164)		(0.164)		(0.164)		(0.164)		(0.164)
ASSET.TANG	0.388	***	0.389	***	0.382	***	0.384	***	0.384	***
	(0.142)		(0.142)		(0.142)		(0.142)		(0.142)
DIV.PAYOUT	−0.060	**	−0.060	**	−0.062	**	−0.062	**	−0.062	**
	(0.025)		(0.025)		(0.025)		(0.025)		(0.025)
GINI	0.009	**	0.009	**	0.009	**	0.009	**	0.009	**
	(0.005)		(0.005)		(0.005)		(0.005)		(0.005)
GDP.GROWTH	−0.000		−0.001		−0.001		−0.001		−0.001
	(0.010)		(0.010)		(0.010)		(0.010)		(0.010)
UNION.DUMMY	−0.328	***	−0.329	***	−0.331	***	−0.329	***	−0.329	***
	(0.048)		(0.048)		(0.048)		(0.048)		(0.048)
INT.PROMOTION.D			−0.126	***	−0.101	***	−0.102	***	−0.099	***
			(0.035)		(0.036)		(0.036)		(0.036)
POL.CAREER.DEV					−0.111	***	−0.118	***	−0.098	***
					(0.038)		(0.038)		(0.038)
POL.DIVERS.OPP							0.045
							(0.039)
MGM.TRAINING.D									−0.149	***
									(0.037)
no. of observations	9451		9451		9451		9451		9451
Wald (joint)	411.26	***	427.62	***	436.91	***	438.44	***	439.34	***
R2	0.25		0.26		0.26		0.26		0.26
AR(1) test	28.18	***	28.16	***	28.10	***	28.10	***	28.49	***
AR(2) test	7.71	***	7.69	***	7.65	***	7.64	***	7.85	***
Panel B. Dependent variable: wage gap (GAP); independent variables encode corporate governance settings and management training policies.
Model No		6				7
Constant		−1.089		*		−1.088		*
		(0.578)				(0.578)
SIZE		0.293		***		0.293		***
		(0.022)				(0.022)
DEBT.RATIO		0.710		***		0.710		***
		(0.116)				(0.116)
CAPEX		−0.779		**		−0.784		**
		(0.358)				(0.358)
EBIT.MARGIN		0.471		***		0.474		***
		(0.100)				(0.100)
CASH.RESERVES		−0.529		***		−0.529		***
		(0.164)				(0.164)
ASSET.TANG		0.389		***		0.388		***
		(0.142)				(0.142)
DIV.PAYOUT		−0.059		**		−0.059		**
		(0.025)				(0.025)
GINI		0.009		**		0.009		**
		(0.005)				(0.005)
GDP.GROWTH		−0.000				−0.000
		(0.010)				(0.010)
UNION.DUMMY		−0.331		***		−0.330		***
		(0.048)				(0.048)
COMP.C.INDEP		−0.092		*		−0.087		*
		(0.052)				(0.053)
COMP.C.NONEX						−0.101
						(0.132)
no. of observations		9451				9451
Wald (joint)		417.7		***		418.3		***
R2		0.26				0.26
AR(1) test		28.35		***		28.33		***
AR(2) test		7.86		***		7.86		***

Notes: the table presents the GLS random-effect static panel model estimates. All models include the time and industry dummies (not reported). The heteroscedasticity robust standard errors are provided in parentheses. ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively.

. Similar to [11], we find that salary disparities are higher within larger and more mature firms. The novel part of the analysis consists in the empirical inquiry into the impact of internal corporate policies on the scale of inequity.

To start with, we document a negative associative link between the degree of employee unionization and the scale of pay gap. Inequality is found to be significantly lower in firms where the level of unionization exceeds 50%. Previously, we also reported a strong impact of union coverage on the turnover of top management. It might well be the case, that strong unions preclude excessive wage disparities, thereby positively contributing to strengthening employee incentives and workers’ bargaining position.

Secondly, our findings suggest that implementation of internal policies facilitating and favoring internal employee promotion (INT.PROMOTION.D) and career development (POL.CAREER.DEV) is negatively associated with the degree of wage gap. The implementation of such policies is particularly vital in the case of scalable industries, where human capital and know-how play a crucial role in forging the competitive advantage of firms. The need to constantly invest in employee development and to retain the most capable and competitive staff may limit the potential scale of pay disparities. Policies favoring internal promotion and managerial training (MGM.TRAINING) reduce the companies’ demand for managerial talent from outside. Since external candidates are more likely to negotiate a risk premium to their remuneration attributable to a lack of acquaintance with their new employers’ operational context, internal recruitment may reduce the upside pressure on remuneration.

Finally, we focus on the possible impact of the composition of the board’s remuneration committee on the scale of wage gap. In particular, we verify whether the dominance of independent (COMP.C.INDEP) and non-executive members (COMP.C.NONEX) on the board’s remuneration committee may reduce the scale of wage disparities by limiting the appetites of CEOs. Our results suggest that the composition of the remuneration committee plays only a minor role in shaping the scale of salary disparities. The variable COMP.C.INDEP is significant only at 10% level and should thus be interpreted with caution. COMP.C.NONEX is insignificant. These results point to the possible existence of an imbalance in bargaining power between the CEO and the compensation committee. In many cases, supervisory boards may simply be captured by the top executives with their appointment being largely impacted by the CEO. At the same time, the board may apprehend the possible negative consequences of a failed negotiation with the CEO or a subsequent CEO departure [31]. Finally, the members of compensation committee may have a direct interest in laying a solid foundation for a productive cooperation with the CEO; therefore, they may have no incentive to bargain or initiate a remuneration dispute. After all, unsatisfactory relations with the CEO may bear much more unpleasant consequences than the popular outcry or employee discontent.

6. Discussion

The results reported in the study shed some new light on the role of salary gap in shaping the operational performance of non-financial companies. In particular, we demonstrate that both beneficial and detrimental effects are accompanying elevated within-firm salary disparities. Among the negative repercussions, decreased employee productivity and higher staff turnover appear to be particularly disturbing. Additionally, we note a higher likelihood of employee disobedience actions and collective controversies accompanying a higher salary gap. Among the benefits accruing mostly to shareholders, we document lower costs and higher operational margins. Overall, we demonstrate that within-firm salary gap is associated with inferior employee well-being, but with improved company-wide performance. Notably, the observed effects appear to be particularly strong in the tail of the distribution of firms in terms of the magnitude of salary gap. Firms with the highest wage disparities are the most likely to suffer the negative consequences stemming therefrom. The latter suggests that the possible impact of salary gap on firms’ performance and employee productivity is rising non-linearly.

The present paper set a goal of answering three principal research questions. Below we summarize the responses to those questions relying on the empirical evidence obtained through econometric tests.

The first research question inquires into the possible link between the scale of intra-firm wage gap and employee incentives. Overall, our analysis unambiguously suggests that wage gap exercises an economically significant negative influence on employee incentives. The inequity aversion appears to outweigh the incentivizing impact of internal competitive tournaments thereby engendering empirically observable losses in employee productivity and potentially encouraging shirking and retaliatory employee action taking the form of internal disputes. Firms reporting extreme wage disparity appear to be the most exposed to this detrimental impact.

The second research question we attempted to answer concerns the possible link between the scale of wage gap and firms’ operational performance. In particular, we try to establish whether the higher salary disparities may be an acceptable price to pay for effective stewardship by better managers, who mandate higher compensations. Our findings suggest that despite possibly reducing employee productivity, wage gap generally exercises a beneficial impact on operating performance of sampled companies. The discovered relationship may point to the incentivizing role of CEO remuneration schemes in shaping managerial decision making. Since the level of CEO pay is frequently tracking corporate performance, it may well be the case that top management is guided by the motive of cost optimization in pursuit of superior operating performance. Overall, it appears that the positive impact of wage gap on corporate performance outweighs the negative consequences it bears for employee incentives. The transmission mechanism between wage gap and corporate performance is circular to a certain extent. On one hand, either because of capture or the need to retain scarce managerial talent, the board devises generous remuneration packages to align the interests of the CEO with those of shareholders. The lavish remuneration packages, which frequently reward managers for success factors they do not control, widen and perpetuate within-firm wage disparities. On the other hand, guided by the motive of pay maximization, CEOs invest significant efforts to enhance operating performance by, for example, initiating cost optimization programs. The improvements in operating performance translate into higher remuneration, which in turn widens the wage gap further. We would like to emphasize that the mechanism described above features only discretionary decisions taken by the board and senior management, hence, we do not factor in the possibility of implicit reverse-causal interaction between the analyzed variables. While arguing for the presence of impact of managerial decision making on the scale of salary gap, we cannot accurately quantify the effect that salary gap may exercise on operational and tactical decision making by executives. Thus, we explore the associative link between the scale of salary gap and firms’ operational performance without being able to measure the role of each of the enumerated transmission mechanisms in establishing this link.

Having concluded that an elevated salary gap may be detrimental to employee productivity and well-being, we attempt to establish whether internal corporate policies may remedy the negative repercussions stemming from high wage disparities without exercising severe downward pressure on contemporaneous corporate performance. To that end, we conducted an in-depth analysis of factors associated with the degree of salary gap. The third research question we pose concerns the organizational determinants of wage gap. Our results suggest that internal promotion policies, unionization and dedicated managerial training are factors that may allow to alleviate the consequences of high wage disparities. The results on the relationship between unionization and wage gap are broadly consistent with prior empirical studies. Policies promoting and facilitating internal promotions allow employees to build a long-term career within the same company, reduce the level of staff turnover, improve productivity and broaden the pool of talent from which the firm may recruit managers. We demonstrate that unionization exercises a direct effect on the degree of wage disparities by facilitating collective bargaining, giving employees an effective tool of voicing their concerns towards senior management. At the same time, unionization may impact firms’ operational performance. Thirdly, managerial training and increased investments into human capital within the firm may allow to retain staff, improve productivity and alleviate concerns over elevated salary gap while simultaneously expanding the skillset of the firm’s workforce.

Surprisingly, we document only a weak relationship between the composition of the remuneration committee of the supervisory boards, which shapes executive compensations and the degree of the within-firm salary gap. The independence of the committee appears to play no significant role in alleviating wage disparities. Our findings suggest that extreme wage disparities may not be effectively tackled through corporate governance mechanisms. Alternative mechanisms (internal corporate policies and centralized governmental policies) may be more effective at coping with the problems of growing wage disparities, which may ultimately threaten social cohesion and undermine healthy economic incentives based on productivity.

The findings reported in the present study are broadly in line with inequity aversion theory, which postulates that beyond a certain level, wage disparities start exercising a negative effect on firms’ performance through deteriorating employee incentives, lower morale and lower productivity of the workforce. Dedicated internal policies within firms may partially curb the negative repercussions stemming from salary disparities.

It remains unclear whether the documented negative effects of wage gap on performance are observed at the macro-level. If so, targeted governmental policies may be necessary in order to alleviate the negative consequences of extreme salary disparities for economic growth. Comparative international studies inquiring into this problem are warranted.

7. Conclusions

The empirical findings reported in this paper demonstrate that the consequences of wage disparities for corporate performance are multi-faceted and, therefore, mandate a multidimensional analysis. On one hand, we convincingly demonstrate that an excessive wage gap may exercise a negative effect on employee incentives by depressing individual productivity, inflating staff turnover and increasing the probability of internal remuneration and working condition controversies. On the other hand, wage gap which rewards CEOs for enhanced corporate performance may contribute to the optimization of firm operations by reducing costs and enhancing profitability. Ultimately, wage gap seems to benefit shareholders and top management with the possible byproduct being discontent and employee dissatisfaction caused by inequity aversion [14].

We document that the negative consequences of wage inequality on employee incentives are particularly pronounced in firms reporting outsized wage disparities. While inequality does seem to exercise an incentivizing effect up to a certain point, excessive inequality appears to generate unwanted externalities with the losses seemingly accruing to the employees. What is more, even the high wage gap appears to have no impact on top management retention.

We show that specific organizational measures may stifle inequality growth. In particular, employee unionization and corporate policies favoring employee development and promotion may reduce the dynamics of inequality growth. In contrast, the composition of the boards’ compensation committees seems to bear no consequences for the scale of wage inequality. Overall, our results suggest that tackling excessive wage gap should proceed by means of increasing the bargaining power of employees as opposed to devising alternative corporate governance mechanisms.

In our opinion, further studies regarding the repercussion of wage gap should focus on the quantification of externalities generated by extreme inequalities. While the surplus for shareholders and top management may be positive, factoring in the implicit costs borne by the employees due to lower productivity, strikes and voluntary terminations may yield a large loss in efficiency.

Empirical literature frequently suggests that the higher wage gap is the price for better performance. At least, this is a quintessential explanation for a widening pay gap. If one assumes that social welfare is measured by the well-being of the most well-off than it is indeed justified.