ESG Disclosure and Employee Turnover. New Evidence from Listed European Companies

Garsaa, Aziza; Paulet, Elisabeth

doi:https://doi.org/10.7202/1097695ar

1. Introduction

As pointed out by Deloitte[1] in an article recently published in NACD Directorship magazine, “companies that don’t harness the power of ESG disclosure risk losing favour with investors or ceding competitive advantage, and they may also be at a disadvantage when attracting and retaining customers and employees.” The stakes are therefore clear: it is not enough to invest in social responsibility activities to reassure stakeholders and limit employee turnover, as suggested by prior research (Carnahan et al., 2017). Providing them with ESG indicators is, increasingly, also essential. Over the last few years, integrating ESG data into annual reports has become the norm for multinational corporations, with 78% of them doing so in 2017 (KMPG, 2017).

Companies are realizing that the costs of high employee turnover far exceed those of investing in social responsibility and, especially, in responsible human resource management (HRM) (Stamolampros et al., 2019). In addition, it is commonly assumed that a high level of social responsibility makes it easier to attract a high-quality workforce (Alareeni and Hamdan, 2020). Because companies are aware that a good reputation may increase job satisfaction and decrease employee turnover, many seek to retain employees by ensuring that they identify with their organization (Lee et al., 2013).

The drivers and consequences of social responsibility practices have long been the subject of numerous studies (Friedman, 1970; Franco et al., 2020). More recently, researchers have started to take an interest in ESG disclosure because stakeholders are increasingly pressuring companies to be transparent about their social responsibility practices (Font et al., 2012). The studies focus essentially on the drivers of such practices (Yang et al., 2017; Wang et al., 2018) and their effects on a company’s financial performance (Albitar et al., 2020; Sassen et al., 2016). However, there have been few empirical studies of ESG disclosure and how it affects employee behaviour. Most of them conclude that ESG disclosure is positively associated with such employee behaviours as creativity, commitment, engagement and volunteering. (Peterson, 2004), which in turn reduce employee turnover (Vitaliano, 2010).

We thus explored the relationship between ESG disclosure and employee turnover, using a mixed theoretical framework. We assumed that a company’s transparency pushes it to implement more responsible human resource management, which in turn helps satisfy the employees’ psychological needs, enhances their job satisfaction (Hudson et al., 2017) and reduces their turnover rate (Farooq et al., 2014). This relationship is essentially due to pressure from stakeholders, who use ESG indicators to measure and judge a company’s social responsibility (Rumambi and Marentek, 2015).

We analyzed how ESG disclosure affects employee turnover, using panel data fixed-effect quantile regression and a sample of 212 listed European companies during the 2010-2017 period. This study contributes to the existing literature on corporate social performance and employee turnover in two ways. First, we investigated how ESG disclosure affects employee turnover over a range of possible turnover rates. Indeed, fixed-effect quantile regression had the key advantage of enabling us to evaluate how ESG disclosure affects the employee turnover rate at any point of that rate’s conditional distribution. Second, this relationship was studied as a function of corporate sustainability reporting (mandatory or voluntary) and economic sector. We could thus measure how it might vary by institutional context and by sectoral characteristics.

The remainder of our paper is broken down into five sections. We will present the theoretical framework of the ESG disclosure/employee turnover relationship and develop our hypotheses in Section 2, the data and empirical analysis in Section 3, the results in Section 4 and the discussion and conclusion in Section 5.

2. Theoretical Framework and Development of Hypotheses

2.1 ESG Disclosure and Employee Turnover

According to the existing literature on corporate social performance (CSP), on employee turnover, on job satisfaction, on corporate governance, on HRM and on labour psychology, ESG disclosure can reduce employee turnover above all by satisfying employee needs through strategic HRM (Page and Vella-Brodrick, 2012).

Since employees are considered to be central to sustainable HRM (Richards, 2022), they will remain in their company to the extent that their needs are satisfied (Mobley et al., 1979). The existing literature generally identifies two kinds of employee incentives: financial and nonfinancial (Peterson and Luthans, 2006). Most empirical studies conclude that both negatively affect employee turnover (Saleem, 2011). Money can certainly help satisfy financial needs (Mitchell and Mickel, 1999). Financial incentives are therefore effective in attracting, motivating and retaining workers (Peterson and Luthans, 2006; Gillan, 2006). Hence, a company can retain productive, high-quality workers by offering them benefits and high salaries (Hope and Mackin, 2007), which are considered to be strong incentives (Korschun et al., 2014).

However, money is not the only driver of employee retention; nonfinancial incentives play a non-negligible role (Sawatsky, 1951). Aguilera et al., (2007) proved that responsible management through implementation of social responsibility strategies can help satisfy the psychological needs of employees and increase their motivation (Hilliard, 2013), engagement (Smith and Macko, 2014) and morale (Park and Levy, 2014), which in turn reduce employee turnover (Peterson, 2004; Porter and Kramer, 2006). For example, using a multivariate analysis of French data during the two lockdowns of 2021, Fang et al. (2019) showed that social support from the immediate manager has a strong direct effect on employee burnout. In the same vein, Unsal-Akbıyık and Zeytinoglu (2018), showed that the close family-like work environments within boutique hotels in Istanbul increase the employees’ intention to stay. Seo and Chung (2019) demonstrated that abusive supervision increases turnover intention of young factory workers in northern China. The effect of HRM quality on employee psychological needs seems to be universal and not only European.

Besides helping companies attract and retain a high-quality workforce (Sassen et al., 2016.), social initiatives and, especially, high levels of CSP are generally considered to indicate superior management skills (Waddock and Graves, 1997). Indeed, companies engaged in employee-oriented CSP tend to enjoy higher job satisfaction and lower turnover among their employees (Turban and Greening, 1997; Park and Levy, 2014, Hudson et al., 2017). Recent decades have seen researchers investigate how employees perceive social responsibility practices, and how that perception affects their behaviour (Boonbumroongsuk and Rungruang, 2022). There have also been investigations into such employee behaviours as creativity (Hur et al., 2018), volunteering (Muthuri et al., 2009), commitment (Turker, 2009), job satisfaction (Sims and Keon, 1997; Valentine and Fleischman, 2008) and retention intention (Lee and Chen, 2018). However, the empirical link between ESG disclosure and employee behaviour is still under-explored (Vitaliano, 2010).

In particular, the corporate governance literature considers a company’s HRM skills to be key to its performance and one of its main sustainable competitive advantages (James and Joseph, 2015). Corporate governance mechanisms are thus part of a company’s resources (Wernerfelt, 1984). Consequently, “proper management is related to an organization in having good corporate governance as it has become one of the most important elements in evaluating firm’s performance and sustainability” (James and Joseph, 2015, p. 118).

Most of the empirical research linking strategic HRM to corporate governance focuses on employee involvement, retention and performance, since both the company and its employees may benefit from improvements to corporate governance (Karami et al., 2008). For example, Vitaliano (2010) found that a company can reduce its annual quit rate by adopting business policies that enhance its social ranking. In a context of mergers and acquisitions, Chun (2009) found that responsible behaviour by a company strongly increases employee loyalty, satisfaction and emotional attachment.

In that context, ESG disclosure could therefore increase employee trust (Dawkins and Lewis, 2003). One could suppose that ESG disclosure exerts additional pressure on socially conscious companies to maintain or enhance their reputation because they wish to remain transparent vis-a-vis their stakeholders (DeTienne and Lewis, 2005; Rumambi and Marentek, 2015).

Employee motivation and, thus, employee turnover are affected by psychological factors that may be classified into two groups: (1) recognition and performance feedback (Luthans, 2000; Bradler et al., 2016.); and (2) need for self-identification, as developed in the literature on social identity theory (Tajfel and Turner, 1986). As members of the organization, employees will self-identify with their company and may develop an emotional commitment (Farooq et al., 2014). Increased transparency may thus help employees identify with their company and thereby satisfy some of their psychological needs, increase their motivation and reduce staff turnover.

An increase in ESG disclosure may thus increase the company’s trustworthiness and strengthen its relationship with its employees by making them more satisfied (Ulmann, 1985; Perrini et al., 2009). By making employees feel more loyal and satisfied, ESG disclosure may enhance their self-perception and reduce the turnover rate (Lee and Chen, 2018). We therefore propose:

Hypothesis 1. The ESG disclosure score is negatively correlated with the employee turnover rate.

2.2 ESG Reporting Regulations

Institutional context is considered to be a key driver of social responsibility initiatives in both the empirical literature (Cavalcanti Sá de Abreu et al., 2012) and the theoretical literature (Crifo and Forget, 2015). The role of institutions in regulating the European market is discussed in the sixth chapter of Amable (2003), which examines the “… opposition between a project of regulated capitalism, which corresponds to a renewal of the Continental model, and a neo-liberal project, which aims at transforming the EU countries into market-based economies” (p. 225). The author concludes that the attempted transformation has been a relative failure, which confirms his theory of the diversity of capitalism. He states that different economic models, each endowed with internal consistency, can coexist and perform each in its own way. In the same vein and more recently, Fainschmidt et al. (2018) distinguished among seven types of institutional systems. Drawing on the comparative capitalism literature (Lane and Wood, 2013), Doering et al., (2015) analyzed how multinational corporations may exploit different institutional contexts to develop diverse strategies to achieve sustainable competitive advantages. They “…argue that the institutional context in which a company is embedded can provide an environment in which companies, and other social actors, perceive and act upon sustainability, for example in relation to environmental regulation, in a number of ways” (Doering et al., 2015, p. 621).

There seems to be a positive correlation between institutional pressure and social responsibility practices (Yang et al., 2017). Furthermore, it is commonly argued that European companies are more engaged in social responsibility activities than non-European ones (Young and Marais, 2012) since “stakeholder dialogue is more established in Europe, where CSR[2] has developed most extensively, than elsewhere” (Tokoro, 2007, p. 143). Indeed, the European Commission made huge efforts during the last decade to encourage companies to be more socially responsible (European Commission, 2019). Most of the time they have no choice, since in many countries corporation acts oblige them to publish sustainability reports that disclose the extent to which labour standards and environmental, social or ethical considerations are respected.

Some empirical studies have looked into the effects of non-financial reporting regulation on company transparency, and most have found that disclosure of both financial and non-financial information improved after the adoption of non-financial reporting regulation (Gulenko, 2018; Hoffmann et al., 2018; Wang et al., 2018). We therefore propose:

Hypothesis 2. Mandatory ESG disclosure increases the ESG disclosure score and conversely decreases employee turnover

2.3 Economic Sector

It is commonly argued that manufacturing contributes more than other industries to environmental pollution and social costs (Handayani et al., 2017). According to Garcia et al. (2017), “companies with manufacturing processes that negatively influence the environment will have greater disclosure compared with companies in other industries” (p. 145). Hence, ESG may have a stronger effect on employee turnover for companies in manufacturing than for those in services. We therefore propose:

Hypothesis 3. The effect of ESG on employee turnover is stronger for companies in manufacturing than for those in services

3. Data and Methodology

3.1 Sample Selection and ESG Disclosure Score

We measured how the ESG disclosure score affects the employee turnover rate by using an unbalanced panel of 212 multinational corporations (1,041 observations) listed in the European capital market and operating in different areas of manufacturing and services from 2010 to 2017. We focused on the European capital market because European regulations have been strengthened since the 2007-2008 financial crisis in order to increase the social responsibility awareness of public interest entities (Velte, 2017). To avoid creating a trend effect due to the last financial crisis, we limited our analysis to the 2010-2017 period. Financial and non-financial data were extracted from the Bloomberg database, using the company “security” identifier. We extracted 898 listed companies that were in the European capital market (all indices combined) and which were operating in several economic sectors. A lot of them had missing values in the selected variables. To be included in our unbalanced panel, a company had to have a minimum of two successive years of available data during the 2010-2017 period. That criterion severely reduced the total number of companies in the final sample. The final sample had a wide variety of companies with headquarters in different countries all over the world, more than 90% of which were in Europe. About 50% of the selected companies were in manufacturing, and more than 60% mandatorily reported their ESG indicators.

To measure company transparency, we used the Bloomberg’s Disclosure Score Index, which is the proxy most often used in the existing empirical literature (e.g., Giannarakis, 2014). The index is calculated from the ESG data that a company presents in its public reports. Each data point is weighted in terms of importance and tailored to different economic sectors. The ESG score has various definitions (Billio et al., 2021), metrics (e.g., the S&P ESG Index and the STOXX Global ESG Index) and components (three disaggregation levels: economic, social and environmental). The ESG score may thus have different values because the data can be calculated, weighted and collected in different ways. Our methodological choice is justified by three main reasons. First, it would be complicated for us to construct our own ESG indicator by collecting the data directly from the companies. Second, since stakeholders do not perceive a company’s CSR initiatives solely in terms of one component of the ESG score (Nitkin and Brooks, 1998), we used an aggregate ESG score instead of a disaggregated one to measure company transparency. Third, had we used multiple sources instead of a single one (i.e., the Bloomberg ESG index), we would have had to discard much more data because of a lower likelihood that the data for one company would be comparable to the data for the others.

3.2 Econometric Strategy

To capture the effect of ESG disclosure on employee turnover, we used a model that includes the company’s ESG disclosure score as an explanatory variable. In addition to year dummies that enabled us to capture some macroeconomic and cyclical effects, three control variables were introduced, i.e., the company’s size (lnEmp), its efficiency (Efficiency) and its return on assets (ROA). These variables are consistent with the existing empirical literature. First, it is commonly assumed that ROA is a major predictor of employee turnover. A positive relationship between a company’s financial performance and its employee job satisfaction has been widely confirmed (e.g., Stamolampros et al., 2019). Second, the company’s efficiency, as measured by the ratio of total sales to number of employees, is considered to be a proxy for the company’s level of productivity and, hence, for employee motivation and job satisfaction (Imran et al., 2015). Finally, the company’s size is commonly considered to be an indicator of the company’s organizational complexity (Hausknecht et al., 2009).

Hence, our model can be simplified as follows:

Where is the company, the year, and the composite error term, with being the individual fixed effects and the idiosyncratic error term. Detailed definitions and descriptive statistics for the selected variables are given in the Appendix (cf. Table 1).

Figure 1

**Kernel Density Estimation of Employee Turnover (2010–2017)**

The characteristics of our sample and our selected variables determined our econometric strategy. First, we ran a fixed-effects model because the null hypothesis of the Hausman test (Hausman, 1978) was rejected at a 2% level of significance. Second, the distribution of the dependent variables (cf. Figure 1) prevented us from using standard regression estimators that are not robust to extreme values and heavy-tailed distributions (Galvao, 2011). The distribution of the employee turnover rate is centred at 10%, with most of the companies having a rate close to the median (11.2%). Some have a rate as high as 50% or more, and some a rate as low as 0.2%. In such cases, several authors (e.g., Coad and Holzl, 2012) have advocated using the fixed-effects quantile regression or “two-step estimator” (denoted 2-STEP) developed by Canay (2011). It has essentially two advantages. First, it controls for fixed[3] effects that cannot be measured or observed (e.g., intangible capital, like employees’ skills, institutional or sectoral context). Second, it is robust to extreme values and heavy-tailed distributions, like employee turnover (cf. Figure 1).

This approach also has a practical advantage. It is well known that some specific economic sectors suffer more than others from high employee turnover (e.g., retail, wholesale, and services). Hence, the 2-STEP estimator can better explain the relationship between ESG disclosure and employee turnover because it identifies the estimated correlation at all points of the conditional distribution of the turnover rate.

Since the selected sample includes only companies that reported ESG values, our results may suffer from selection bias due to the correlation between the time-variant error term () and the ESG disclosure score. To control for the potential bias, and to check the robustness of the results, we ran fixed-effects instrumental variable estimations, also called two-step GMM-HAC (Arellano, 1987).

4. Main Findings

To check the robustness of our results, we ran several estimations using three estimation techniques. To take into account the conditional distribution of the employee turnover rate, and to ensure a high degree of robustness, we interpreted only the results from the fixed-effects quantile regression (2-STEP)[4].

4.1 Company Transparency: Key to Less Employee Turnover?

Figure 2 shows the estimated coefficients of the explanatory variables as a function of the conditional distribution of the employee turnover rate. The results for the full sample confirm our first hypothesis (H1). ESG disclosure decreases employee turnover; therefore, the most transparent companies enjoy low turnover. When we used lagged explanatory variables by two periods, the results confirmed the sign and significance of the estimated coefficient associated with the ESG disclosure score. Employee turnover seemed to be higher for large companies, perhaps because of organizational problems due to coordination difficulties and reduced motivation (Hausknecht et al., 2009). Contrary to expectation, employee turnover seems unaffected by the company’s efficiency. As expected and already found by earlier empirical studies (e.g., Yanadoria and Katob, 2009), there is a negative correlation between employee turnover and financial performance, as measured by the company’s return on assets (ROA). Furthermore, the correlation seems to be stronger at higher turnover rates, ranging from -5.547 at the 10^th percentile to -7,376 at the 90^th percentile.

Figure 2

**Correlation between Employee Turnover and Company Characteristics per Quantile of Turnover Rates (full sample)**

These findings support the claim that better financial performance enables a company to retain its employees and slow down their turnover because it can better meet their financial needs (Peterson and Luthans, 2006; Gillan, 2006).

Companies draw on all their resources (financial and non-financial) to retain their most critical employees. That strategy pays off. By improving HRM skills, a company will enjoy a greater chance of success (Karami et al., 2008). Whatever the nature of the business environment and their individual characteristics, companies understand they must guarantee their staff some degree of stability and develop adequate strategies to that end. If, for example, a company is large and thus suffers from organizational problems and a high turnover rate, it should try to compensate by offering its employees interesting financial incentives to remain.

4.2. ESG Disclosure and Employee Turnover: Does Non-Financial Reporting Regulation Matter?

To test our second hypothesis, we divided our sample into two groups: companies for which ESG disclosure is mandatory and those for which it is voluntary.

Figure 3

**Correlation between Employee Turnover and Company Characteristics per Quantile of Turnover Rates (where corporate sustainability reporting is mandatory)**

Figures 3 and 4 present the correlation coefficients over the full range of turnover rates for, respectively, companies with mandatory ESG disclosure, and those with voluntary ESG disclosure. Disclosure score shows a significant negative correlation with employee turnover only in jurisdictions where corporate sustainability reports are mandatory. Our findings corroborate previous ones and confirm our second hypothesis: pressure from the institutional context (Helmig et al., 2016) increases the effect of company transparency on employee turnover.

For companies that must submit corporate sustainability reports, the correlation between ESG disclosure and employee turnover ranges from -0.0472 at the 10^th percentile to -0.0924 at the 90^th percentile. Mandatory ESG disclosure seems to reduce employee turnover much more among companies suffering from a high turnover rate. Such companies pay much more attention to disclosure in order to establish a trusting relationship with their stakeholders, especially current and potential employees (Perrini et al., 2009). Similar reasons explain why employee turnover decreases with increasing company size in jurisdictions where disclosure is mandatory. Such companies probably try harder to implement sophisticated human resource practices and/or offer different kinds of benefits in order to increase employee satisfaction and to limit the turnover rate (Hausknecht et al., 2009). That is not the case with companies that are not obliged to publish corporate sustainability reports. For such companies, employee turnover increases with increasing company size (Jackson and Schuler, 1995; Guthrie, 2000).

Finally, for both groups of companies, employee turnover is negatively correlated with ROA. The negative correlation is more than four times stronger for companies that report voluntarily (cf. Figure 4). Clearly, such companies improve employee satisfaction more through financial incentives than through non-financial benefits (Yanadoria and Katob, 2009).

Companies adapt to their environment, in particular their institutional context. Even if we consider ESG reporting systems as a single aspect of regulation, we see that companies operate within a wide variety of institutional configurations and may use, for instance, different strategies to reduce employee turnover (Amable 2003). In our sample, employee turnover is approximately the same on average for the two groups of companies. The outcomes are similar despite differences in institutional context, a finding in line with the concept of sustainable varieties of capitalism developed by Doering et al. (2015).

4.3. ESG Disclosure and Employee Turnover: Does Economic Sector Matter?

To test our third hypothesis, we subdivided the full sample into two groups: companies operating in manufacturing and those in services. The service sub-sectors were accommodation and food service activities (section I, NACE Rev.2 classification), information and communication (section J, NACE Rev.2 classification), professional, scientific and technical activities (section M, NACE Rev.2 classification) and administrative and support service activities (section N, NACE Rev.2 classification).

Figure 5

**Correlation between Employee Turnover and Company Characteristics per Quantile of Turnover Rates (Manufacturing)**

The results from companies in manufacturing and services are shown in Figures 5 and 6 respectively. The sign and statistical significance of the estimated coefficients for manufacturing companies confirm and corroborate the results from the full sample (cf. Section 5.1). The results for manufacturing are similar to those for the full sample, essentially because more than 50% of the companies in our full sample are in manufacturing.

Figure 6

**Correlation between Employee Turnover and Company Characteristics per Quantile of Turnover Rates (Services)**

Except for the ESG disclosure score, the results for service companies are the complete opposite of those for manufacturing companies. Employee turnover decreased with increasing company size, probably because larger companies can invest more in human resource management (Hausknecht et al., 2009). That explanation is all the more plausible because customer relationships are very important to service companies, which must manage and organize their employees to deliver quality service (Wildes and Parks, 2005; Mukherjee et al., 2009). The importance of human resource management is supported by the negative sign of correlations for the Efficiency variable, which range from -15.65 at the 10^th percentile to -11.18 at the 90^th percentile.

Employee turnover in services seems to be strongly and positively affected by financial performance, being highest for the highest performing companies. This finding may seem paradoxical because it is commonly assumed that employee turnover is a significant cost for businesses (Kwon and Rupp, 2013; Stamolampros et al., 2019) and that financial performance reduces staff turnover (Bakotić, 2016.). On the other hand, the quest for financial performance may actually worsen working conditions and thus increase employee turnover, since it is commonly argued that the cost of improvements to working conditions reduces a company’s profitability (Maqbool and Zameer, 2018).

5. Discussion and Conclusion

In this paper, we show that the relationship between a company’s transparency and its employee turnover rate strongly depends on the institutional context, especially disclosure regulation. The more a company is scrutinized, the more it will try to maintain and/or improve its reputation of social responsibility and thus reassure and satisfy its stakeholders.

We specifically show that the turnover rate is higher for service companies than for manufacturing companies (OECD, 2001). This is essentially due to the poor working conditions in services (e.g., rude customers, part-time work, job insecurity) (Han et al., 2016). In service companies, employee turnover correlates negatively with company size and efficiency, and positively (and in a higher proportion) with profitability. It seems that higher profitability often comes at the expense of working conditions and, thus, job satisfaction. Service companies often fail to measure the cost of employee turnover, preferring to prioritize financial performance over human resource management (Hinkin and Tracey, 2000).

Our findings have several managerial implications. First, the employee turnover rate can be decreased not only by investing in social responsibility but also by informing stakeholders about such efforts. Second, because they are organized differently, service companies should develop more sophisticated human management practices and offer additional employee benefits while seeking to boost financial performance, which by itself seems to increase employee turnover. Finally, governments could reduce employee turnover by encouraging ESG disclosure through more effective ESG reporting regulation and a better-defined legal framework.

To conclude, our study suffered from four limitations. First, we did not distinguish between voluntary and involuntary turnover, as the data did not allow us to take this difference into account. The distinction is furthermore difficult to make, as the decision to leave a job could have both individual and organizational causes (Campion, 1991; Lee and Jung, 2016). Second, we did not use a direct proxy for financial and non-financial employee incentives. We instead approximated them by using indicators of company-level characteristics. Third, company transparency was approximated by the Bloomberg ESG disclosure index. Because the E, S, and G components are not defined by common/unique standards, attributes and characteristics, there may be a wide variety of results. Future researchers may resolve this problem by considering different measures and definitions of the ESG score. Finally, even though company fixed effects were already controlled, the 2-STEP estimator did not allow us to introduce a dummy variable, such as the categorization of national context inspired by Bruno Amable’s arguments on the diversity of modern capitalism (Amable, 2003). Such a variable could add an interesting theoretical contribution.

ESG Disclosure and Employee Turnover. New Evidence from Listed European Companies

Summary

Résumé

1. Introduction