Corporate Social Responsibility and Firm Performance in GCC Countries: A Panel Smooth Transition Regression Model

Abstract

There is evidence for mixed effects of corporate social responsibility (CSR) on corporate financial performance. In particular, evidence is reported to be positive, negative, and insignificant. These controversies are generally explained by two opposing schools of thought, which are the social impact hypothesis and the shift of focus hypothesis. This paper attempts to contribute to the ongoing debate by investigating whether the relationship between CSR and firm financial results is nonlinear. Therefore, this research relies on a panel smooth transition regression (PSTR) model in order to calculate the value transition threshold of CSR in 70 Gulf Cooperation Council (GCC) firms from 2015 to 2020, using the CSR composite index and various CSR dimensions, which include environmental, social, and governance transition dimensions. Empirical findings indicate that investment in CSR does not help to boost corporate value until it exceeds the value transition threshold. However, when the marginal benefit exceeds the cost, CSR investment becomes a positive contributor to corporate performance. Furthermore, results indicate that the nonlinear relationship persists when using the individual CSR dimensions, i.e., governmental, social, and environmental CSR measurements. Finally, an interesting finding shows that the social CSR dimension is associated with the highest threshold level. Hence, firms should invest more in the social aspects of CSR in order to see their profitability increase.

1. Introduction

During the past decades, central issues such as climate change, environmental pollution, philanthropic activities, and community relations, among other reasons, motivated firms to commit to corporate social responsibility (CSR) practices. Engagement with CSR became an integral aspect for firms to develop a long-term sustainable relationship with their business environment [1,2]. Furthermore, it is argued that firms should have environmental responsibility towards their waste emissions and resource use; social responsibility towards their communities and human rights; and economic responsibility to their stakeholders and shareholders [3,4,5].

The significance of CSR and sustainability reporting is further evident in the strides taken by international regulative bodies to harmonize and develop comprehensive global guidelines for sustainability disclosure standards. In this regard, the International Financial Reporting Standards Foundation recently established the International Sustainability Standards Board (ISSB) in 2021 as a new international standard-setting body that aims to develop comprehensive sustainability disclosure standards, thereby helping investors and interested parties make informed decisions about a firms’ sustainability reporting (International Financial Reporting Standards, 2021). In addition, the Securities and Exchange Commission in the U.S. established in 2021 a Climate and Environmental, Social and Governance (ESG) Task Force in order to investigate ESG misconduct and better govern the market (U.S. Securities and Exchange Commission, 2021).

The aforementioned global bodies and initiatives underline the rising interest in CSR and sustainability internationally. In addition, the number of firms committing to CSR and sustainability reporting is increasing yearly, as indicated by the Governance & Accountability Institute Report (Governance & Accountability Institute Report, 2021), which emphasized the global change occurring in the business environment and the growing demand for sustainable information. In the same line, much research has highlighted that engaging with CSR practices can promote stronger work engagement and enhance employee job satisfaction, organizational commitment, employee productivity, and retention [6,7]. Besides, CSR can increase customer loyalty and satisfaction [8] and improve production quality, stakeholder support, firm reputation, and sustainable business performance [9]. Consequently, this fueled the CSR literature with research focusing on the relationship between CSR and firm performance [2,10,11,12,13]. However, other authors reject benefits of CSR for firm performance and rather underline its drawbacks. For instance, CSR is reported to be associated with high costs incurred from social and environmental activities, which weaken the competitive advantage and firm profitability [14,15,16]. Moreover, engaging in CSR practices directed towards environmental activities, employee relations, and society welfare would result in shifting the focus towards practices that do not enhance shareholder value but increase costs [17].

From a theoretical standpoint, the classical economists’ perspective advocates that the goal of firms is to enhance economic value and profit maximization for shareholders [18]. In line with this theory, the shift of focus hypothesis stresses that engaging in CSR would draw the focus towards other aspects other than profit maximization, which in turn establishes a negative relationship between CSR and firm performance [17,19,20]. Furthermore, the trade-off hypothesis asserts a negative relationship exists between CSR and firm performance due to the increased costs that result from investing in CSR [19,21].

Conversely, the stakeholder theory emphasizes that over time firms become more successful when they sustain good relationships with their stakeholders [22] as each stakeholder group can impact, directly [23] or indirectly [24], the economic status of a firm and its future [25]. Hence, the relationship between CSR and firm performance would be positive, which aligns with the social impact hypothesis. The social impact hypothesis proclaims that meeting the needs of stakeholders reflects positively on firm performance [26,27]. Consequently, the theoretical relationship between CSR and firm performance is still inconclusive, as the literature section further indicates.

Similarly, the empirical literature on CSR and firm performance linkages is characterized by a lack of consensus, and studies in developing and emerging markets have yielded mixed results. For instance, [28] conducted a longitudinal survey examining the influence of CSR activities on firm performance in small and medium-sized firms in Pakistan. The findings show that environmental and social CSR are positively correlated with firm performance and, in particular, corporate reputation and employee commitment. Further, [29] examined 365 companies in sensitive industries that are listed in the BRICS countries to examine their environmental, social, and corporate governance performance compared to companies from other industries. The authors concluded that firm profitability is only related with environmental performance. In addition, [13] examined the correlation between CSR and firm performance in 30 listed Bangladesh banks during the years 2006 to 2018 using panel vector autoregression. The authors found that CSR expenditure does not impact financial performance. Likewise, within a developed countries context, a number of studies have reported that a positive relationship exists between CSR and firm performance [1,30], whereas other findings signified either a negative [11,16] or a neutral relationship [31,32]. Another stream of empirical studies examined the impact of COVID-19 on the relationship between CSR and firm performance and conveyed similar findings ranging between positive [21,33], negative [14,15], and neutral relationships [34,35].

The theoretical and the empirical literature reported mixed results regarding CSR effects on corporate performance. Consequently, this research attempts to contribute to the ongoing debate by investigating whether the relationship between CSR and financial performance is nonlinear in the context of firms operating in the Gulf Cooperation Council (GCC) countries. This study emphasizes that the relationship between CSR and business value is nonlinear and is neither completely positive nor completely negative. We particularly assume that the effect of CSR on firm performance depends on the level of CSR investment. To do this, this study will rely on a panel smooth transition regression (PSTR) model to examine the nonlinear linkages.

To our knowledge, no previous research on the nonlinear impact of CSR on corporate value has been conducted in the context of companies operating in GCC countries. In addition, previous studies have rarely used the PSTR model to investigate the nonlinear relationship between the studied constructs. As far as we know, the studies by [27] and [36] are the only ones that used the PSTR model to assess if the CSR index reflects a nonlinear two-regime relationship between CSR and firm value. However, these studies have been conducted in the context of developed countries. Indeed, [27] studied nonlinearities in the context of listed Taiwanese firms and found that CSR negatively impacts firm value when companies are in the early stages of investing in CSR. Nonetheless, once CSR investment surpasses the transition point, the operational performance and reputation of the corporation improves. Furthermore, [36] examined nonlinearities between environmental CSR and firm performance among European firms from 2005 to 2017. Their findings indicate various relationships spanning from nonlinear positive and nonlinear negative to inverted U-shaped relationships. The originality of our research lies in two aspects. First, this is the first paper that examines the nonlinear effect between CSR and corporate performance in the unique context of GCC countries. Secondly, this study investigates the distinct effects of the many dimensions of CSR on corporate performance, namely, the environmental, social, and governance dimensions, which has not been studied previously.

This study is based on a sample of publicly listed GCC firms. The fact that these organizations operate in emerging markets prompted us to investigate them. Indeed, emerging countries differ from industrialized countries in terms of social, political, managerial, and cultural factors, such as poor governance and state control [10,13]. Moreover, research within the context of developing countries is still underdeveloped and there is a need for more empirical evidence [13].

In a variety of ways, this research contributes to the literature on CSR and business profitability. First and foremost, this is one of the most extensive studies on the relationship between CSR and business performance. Second, this study builds on earlier research by looking at this relationship in the unique context of the GCC countries. Finally, it gives academics a starting point for additional research into sustainability issues in developing markets.

This paper is structured as follows. Section 2 reviews the literature on CSR and corporate performance by highlighting contradictory theoretical and empirical studies. The empirical methodology is described in Section 3. Section 4 discusses the empirical results, whereas Section 5 presents further robustness tests. The last section summarizes the policies, the recommendations, and the limits of the study and some future research directions.

2. Literature Review

2.1. The Concept of Corporate Social Responsibility

CSR is considered to be an important concept in which researchers attempt to understand the relationship between social responsibilities and corporations. There is no general agreement in the literature on the definition for CSR or what constructs constitute CSR [37,38], but it has been highlighted that CSR’s definition should be based on philosophical, historical, contextual, and practical aspects that are aligned with the corporation’s objective to avoid resulting in CSR becoming a liability to the corporation [16]. Moreover, [37] conducted content analysis of different CSR definitions and concluded that the CSR definition encompasses five dimensions, which are the stakeholder dimension, the social dimension, the economic dimension, the voluntariness dimension, and the environmental dimension. The environmental dimension focuses on environmental concerns such as a clean environment. The social dimension is concerned with corporations engaging with social aspects to contribute to better societies. The economic dimension focuses on financial aspects and economic development. The stakeholder dimension emphasizes how corporations engage with their stakeholders such as customers, employees, suppliers, and communities. The voluntariness dimension engages with CSR voluntarily driven by ethical and moral values. Finally, other studies have defined CSR as engaging in environmental, social, and governance activities [2,4,39].

2.2. CSR and Firm Performance

The literature on CSR is abounded with studies examining different CSR characteristics through CSR measurement, questioning the social commitment and responsibilities corporations have towards their communities, and examining the impact of engaging in CSR activities on a corporation’s financial performance [10,13,40,41]. Some stress that corporate practices and policies have social consequences on the community and, hence, firms should be socially responsible [42]. However, other scholars argue that resource expenditure on environmental and social objectives leads to increased costs and harming the firm’s competitive advantage and profitability [17]. Therefore, it is still under debate in the literature if corporations should partake in CSR activities as the impact and benefits are not yet conclusive [43]. Despite that, the pursuit of sustainability and the balance of the social ecosystem emphasize the importance of engaging in CSR [44].

Subsequently, the relationship between CSR and firm performance is one of the most debated relationships in the CSR literature, with different theoretical arguments and empirical results depicting the relationship as either positive, negative, or neutral. Thereby, the inconclusive results in the literature warrant further examination. This study is built on the following theories: the social impact hypothesis, the stakeholder theory, the shift of focus hypothesis, and the trade-off theory. The link between CSR and company value is a point of contention in these theories.

2.2.1. Positive Effects of CSR on Business Performance

From a theoretical perspective, the social impact hypothesis is based on the idea that corporations may enhance their financial success by catering to stakeholders’ requirements. For example, focusing on the needs of the employees results in enhancing their productivity, and improving the image of the company and the public’s confidence, which improves the competitive advantage of the firm. Consequently, the benefits that are accumulated from engaging in CSR surpass the costs and fulfil the sustainable development goal, thereby improving firm value [45,46].

The social impact hypothesis and the stakeholder theory proclaim that CSR is driven by corporations’ obligations towards their society [26]. Thereby, the adopted strategies of these corporations tend to take into account the interests of agents and the benefits of the different stakeholders. Following this, the different CSR activities and philanthropic practices that corporations engage in reflect their values and impact their reputation among the stakeholders [47]. This underlies that the corporations’ commitments towards CSR are driven by their moral, ethical, and social commitments and are not based on legal obligations. CSR aims to establish solid relations between corporations and their stakeholders through investing in social or environmental activities and enhancing information transparency. Therefore, stakeholders tend to focus on CSR of firms.

To elaborate, one stream of the literature argues that engaging in corporate philanthropy impacts market return and corporate profitability positively [48,49,50]. In addition, engaging in CSR practices results in decreasing financial risk [50], enhancing reputation and thereby increasing market returns [1], and gaining a competitive advantage [51]. Further, firms engaging in CSR by increasing their transparency and stakeholder engagement have lower capital constraints and have better finance opportunities [52].

The link between CSR and company performance has been the subject of several empirical research studies. For instance, [41] conducted a second-order meta-analysis on 25 previous meta-analyses and their results showed a positive bilateral relationship between corporate social performance and corporate financial performance. In addition, [53] in their research found that corporations with high customer awareness had a positive relationship between CSR and firm value, whereas it was insignificant or negative in other corporations. Furthermore, [54] used firm-level panel regressions and calendar-time portfolio stock return regressions and discovered that profitability of corporations that focus on material sustainability issues was high in margins, stock returns, and sales growth. Additionally, [30] adopted a competitive-action perspective as a contingency aspect of CSR that impacts a firm’s financial performance relationship in 113 US software corporations between the years of 2000 and 2005. Their findings showed that a positive relationship existed in corporations with a high level of competitive action in which CSR initiatives improved the corporation’s financial performance. Further, when corporations engaged in irresponsible activities their financial performance was improved when the competitive action level was low. The research emphasized the role of competitive action in understanding CSR’s performance implications on corporations.

Moreover, [12] conducted research on 34 Saudi publicly traded firms during the period between 2015 and 2020 to examine the usefulness of CEO characteristics in moderating the relationship between CSR and firm performance. The author’s findings showed a positive moderation of CEO tenure and education with CSR and firm performance. In particular, CEOs with an MBA or a science degree and long-tenured CEOs were reported to impact positively the CSR and firm performance relationship.

Another strand of the literature emerged following the outbreak of COVID-19, which focused on examining the impact of the pandemic on the CSR and firm performance relationship. In a recent research study, [21] examined the effect of sustainable investments on the G20 companies’ performances during COVID-19 and found that environmental, social, and governance (ESG) yields benefited in a time of crisis. However, this result was dependent on firm-specific aspects, income level, and ESG pillars. For example, a positive correlation was found between ESG and accounting performance measures and an insignificant or negative correlation was found between ESG and market performance measures. Further, [21] emphasized that ESG practices are highly contingent on country-level mandatory or voluntary codes, environmental obligation, and integrated sustainability reporting. From a Chinese context, the authors of [1] indicated that during the COVID-19 pandemic, high environmental, social, and governance portfolios outperformed the low environmental, social, and governance portfolios, as environmental, social, and governance performance decreased the financial risk. Their research reflects that during the pandemic, the environmental, social, and governance performance was positively related with the short-term stock returns. Relatedly, [33] used Thomson Reuters’ ESG Database to examine all U.S. Stocks on their environmental and social policies. They found that high environmental and social rating stocks have low return volatility and high operating profit margins and returns. Similarly, [55] conducted a cross-country analysis of 6700 firms from 61 economies during the first months of 2020 to examine the relationship between CSR activities and stock prices following the COVID-19 outbreak. They concluded that firms strongly engaged in CSR before the pandemic had superior stock performance after the pandemic’s outbreak, thereby suggesting that CSR improves loyalty and stakeholder relations, which in turn results in stakeholders supporting the firm in times of crisis. Furthermore, [56] found that CSR firms were less impacted by the pandemic outbreak compared to the non-CSR firms in the Taiwanese stock market. Consequently, following the above studies, CSR correlates positively with firm performance.

2.2.2. Negative Effects of CSR on Business Performance

In contrast to the social impact hypothesis, the shift of focus hypothesis builds on the notion that engaging in CSR practices directs a firm towards environmental activities, employee relations, or society welfare and results in shifting the focus towards practices that do not enhance shareholder value but increase the costs. Subsequently, CSR negatively impacts firm performance and the competitive advantage [17,19]. Further, others claim that managers engage in CSR to shift the focus from the firm’s shortcomings [57]. Therefore, the shift of focus hypothesis indicates that meeting the needs of the different stakeholders reflects negatively on firm performance [20].

In addition, advocates of the traditionalist perspective or the trade-off hypothesis argue that there is a negative relationship between CSR and firm performance due to the high costs incurred from social and environmental activities, which weaken the competitive advantage and firm profitability. Moreover, CSR can be used as part of a corporate strategy for social legitimacy and control [58].

Part of the empirical studies have provided evidence on negative results as they emphasize that investments in CSR incur additional costs and expenses, which in turn reflects negatively on firm performance [14,15,16]. During the years between 2009 to 2010, [15] examined 329 listed companies in the U.S., Europe, and Asia-Pacific to determine the relationship between CSR and financial performance. They used the Granger causality test and linear regression analysis and found that engaging in CSR does not result in good financial performance and that financial performance negatively effects CSR. Further, [11] investigated firm-level moderators such as firm size and age to facilitate understanding the CSR and firm performance correlation in Western European listed companies. The moderation analysis showed that CSR and firm performance were negatively correlated in small and/or young firms. This could be due to limited experience and resources and lack of reputation in these firms.

Relatedly, [16] used the panel vector autoregression approach to estimate the relationship between CSR and firm performance. The findings of their research suggest that better financial performance results improved CSR commitment, but good CSR did not result in greater financial performance. Further, significant negative influence was reported in regard to CSR and financial performance measures, in particular, a return on invested capital, return on assets, and return on equity. In similar veins, [14] explored 179 Canadian firms using the Granger causality approach and did not find any significant relationship between corporate social performance and firm performance, excluding market returns. In particular, a significant negative effect was reported for the environmental dimension and three measures of financial performance, in particular, market returns, return on equity, and return on assets. Therefore, these previous empirical research results are in accordance with the trade-off hypothesis.

2.2.3. Neutral Effects of CSR on Business Performance

Proponents of a neutral relationship argue that being socially responsible does not impact firm profitability [31,32]. In this regard, [59] conducted an empirical research study to explore the relationship between CSR and firm financial performance through examining the financial indicators and social responsibility policies of companies listed in the Istanbul Stock Exchange between 2005 and 2007. Although the researchers found a relationship between firm size and CSR, there was no significant relationship between CSR and firm financial performance. Moreover, analyzing data from 599 companies in 28 countries, [31] reported that no direct correlation was found between CSR and financial performance; only an indirect one that depends on the mediating impact of the intangible resources of a firm was found.

During COVID-19, [35] found little support that highly rated environmental, social, and governance corporations have enhanced performance in the stock market. In support, [34] examined the relationship between CSR and market returns in 1750 U.S. firms during COVID-19 and concluded that they found no proof that CSR impacted the stock returns, indicating that CSR before the pandemic was not effective in preserving shareholder’s wealth from an unanticipated crisis such as COVID-19.

2.2.4. Mixed Effects of CSR on Business Performance

Other empirical studies have shown a mix of results of positive, negative, and neutral impacts on the relationship between some CSR elements/dimensions and firm performance. For example, [60] examined CSR and firm performance in Korean listed firms during the period from 2008 until 2014. They reported that environmental responsibility performance had a negative correlation with firm performance, governance responsibility performance had a positive correlation, and social responsibility performance did not have any significant correlation with firm performance. Additionally, [61] examined the CSR effect of internal and external stakeholders on the financial profitability of a restaurant. The findings denote that although internal CSR enhanced the operational profitability, it did not have any impact on market value. On the other hand, the external CSR increased the firm’s market value; however, it was negatively correlated with the operational performance. As such, some empirical studies indicate that the relationship between CSR and firm performance is not a linear relationship by which it is either a positive, negative, or neutral relationship per se.

In support of this, [62] incorporated Bloomberg’s Environmental Social Governance Disclosure Score during the period of 2007 until 2011 to examine CSR performance and corporate financial performance in all listed firms in the S&P 500 stock market. They examined both linear and nonlinear relationships and reported that although the linear model reported a negative and significant relationship between SCR and the dependent variable, the nonlinear model indicated a U-shaped correlation that exists between CSR and the corporate financial performance of accounting measures. This corresponds with the findings of [27], who used the panel smooth transition regression model to assess if the CSR index reflects a nonlinear two-regime relationship between CSR and corporate value in firms listed on the Taiwan stock exchange. Their findings highlighted that corporations at the beginning stages of their investments in CSR yield negative results in regard to firm value due to increased costs. Nonetheless, once CSR investment surpasses the transition point, the operational performance and reputation of the corporation improves. Hence, the benefits exceed the costs, thereby enhancing the firm value. Furthermore, their research emphasized that firm value is positively related with firm size, institutional shareholding, and CSR, which can result in profit maximization.

Focusing on the context of the European firms, [36] examined nonlinearities between environmental CSR and firm performance among French, German, Italian, and Spanish firms from 2005 to 2017. Using a PSTR model, their findings indicated various relationships spanning from nonlinear positive and nonlinear negative to inverted U-shaped relationships.

In summary, the above literature review highlights the inconclusive results of the studies examining the relationship between CSR and firm performance in different contexts. Consequently, this research strives to contribute to the ongoing debate by investigating the potential nonlinearity of the relationship between CSR and financial performance. This study underlines, in particular, that the link between CSR and corporate value is nonlinear and depends on the CSR value. Therefore, we propose the following hypotheses:

H1. 

The relationship between CSR and firm performance is nonlinear.

H1a. 

CSR has a negative effect on firm performance when CSR is below a certain threshold.

H1b. 

CSR has a positive effect on firm performance when CSR is above a certain threshold.

H2. 

Firm size has a positive effect on firm performance.

H3. 

Firm leverage has a negative effect on firm performance.

H4. 

Sales growth has a positive effect on firm performance.

H5. 

Assets tangibility has a negative effect on firm performance.

3. Empirical Analysis

3.1. Data

The initial study’s sample covered all GCC companies that disclosed environmental, social, and governance statistics in the Bloomberg database between 2015 and 2020. CSR information was obtained from the Bloomberg database. The Bloomberg database assesses the extent to which companies disclose ESG data and accordingly calculates scores for each firm. It comprises 247 potential criteria in the ESG dimensions. The score is based on the amount of genuine information provided in each of the environmental, social, and governance sectors.

These rankings were collected from publicly available data on CSR in sustainability reports, firms’ annual reports, and websites, as well as information collected directly from the firm. The ESG Bloomberg database was chosen because it has several benefits over other open-ended databases, including the following: According to the Bloomberg sustainability team, ESG data may “give investors with a macro-level evaluation of how businesses are managing their ESG capital,” allowing investors to “integrate ESG factors into their fundamental analysis.” Furthermore, “the ESG database employs the most complete approach to assess corporations’ environmental, social, and governance actions and outcomes.” Finally, the database assigns a score of 0–100 to each ESG category. As a result, the ESG database is thought to include enough information to investigate CSR reporting and actions.

Financial indicators were taken from annual reports of listed companies. They consist of financial performance factors as well as numerous control variables. The final sample included 70 publicly listed firms from 2015 to 2020. Financial businesses with specific reporting standards, along with firms with missing data, were omitted in this research.

3.2. Dependent, Independent, and Control Variables’ Measurements

3.2.1. The Dependent Variable

The firm’s performance was measured using three variables: return on asset (ROA), return on equity (ROE), and Tobin’s Q (Tobin-Q) [12]. These variables allowed for the assessment of various aspects of firm profitability. The first two variables are specific to the firm’s accounting performance, whereas the third is a proxy for its market performance. The ratio of net income to total assets (ROA) displays how a business uses its resources to generate returns. The return on equity (ROE) is computed by dividing net income by total equity throughout the period. It shows the profit as a proportion of the company’s stock value. Lastly, Tobin-Q is calculated by dividing the sum of the book value of total assets and the market value of equity by the book value of total assets.

3.2.2. The Independent Variables

This study examined the influence of CSR on company performance. Consequently, it used the environment, social, and governance performance scores collected from the Bloomberg database. Bloomberg’s environmental data incorporate a wide range of inputs, including emissions, materials, hazardous waste, renewable energy, water contamination, and environmental operating rules. The social disclosure score is calculated by Bloomberg using data primarily related to employee, product, and community impact policies, such as customers’ complaints, human rights, and community spending. Governance details include the structure and function of the board, executive salaries, board committee activity, and business political participation.

This study used four independent variables, the CSR composite index (CSR)) and the three ESG components that are the environmental score (ENV-CSR), social score (SOCIAL-SCR), and governance score (GOV-CSR). The (CSR) composite index is rated from 0 to 100. The index measures the quantity of CSR data provided by companies between 2015 and 2020. Each ESG category is assigned a score ranging from 0 to 100 in the Bloomberg database. Particularly, the sustainability scores vary from 0.1 for firms that reveal the bare minimum of ESG information to 100 for companies who disclose every data variable gathered by Bloomberg.

3.2.3. The Control Variables

We followed the literature and included a set of control variables. We used four control variables: firm size (Size), which is calculated as the natural logarithm of total assets at year-end; firm leverage (Leverage), which is calculated as the ratio of total debts to total assets; sales growth (Sales-Growth), which is measured as current year sales minus previous sales divided by current year sales; and asset tangibility (Tangible), which is determined as the ratio of fixed assets to total assets.

3.3. The Econometric Methodology

We employed a panel smooth transition regression model to investigate the potential nonlinearity of the relationship between CSR and financial performance in the context of GCC companies. This model was created by [63,64] and has several characteristics that make it particularly suitable for the needs of the analysis. First, it allows the panel’s observations to be classified into a few homogenous groups, each with its own coefficient. In this study, homogeneous groups depended on the level of CSR. Secondly, it allows for the computation of various elasticities which are allowed to vary gradually when transitioning from one regime to another. As a result, the PSTR model considers the variability of the relationship between firm performance and CSR.

Furthermore, the PSTR model thrives at smooth transitions rather than sharp ones. This model varies from [65]’s PTR model in that it allows the elasticity of financial performance in relation to CSR to vary over time and between firms, depending on the level of CSR. Finally, the PSTR model was chosen since we do not know which type of transition exists in the model.

We employed the below model:

$$Y_{it}={\alpha }_i+{\beta }_{0}CS{R}_{it}+{\beta }_{1}CS{R}_{it}\ast g\left(q_{it}; \gamma ;c\right)+{\beta }_2{Z}_{it}+e_{it}$$

(1)

where $Y_{it}$ is the financial performance variable and $CS{R}_{it}$ is the vector of the various CSR measures of the company i at time t. These measures were described in the previous section. ${\alpha }_i$ represents an individual fixed effect, and $Z_{it}$ is a k-dimensional vector of control variables. Following [63], the transition function in the logistic form $g\left(q_{it}; \gamma ;c\right)$ is a continuous function of the transition variable q_it, bounded between 0 and 1, and defined as:

$$g\left(q_{it}; \gamma ;c\right)={\left[1+\exp (-\gamma {{\displaystyle \prod }}_{j=1}^m\left(q_{it}-c_j\right))\right]}^{-1}$$

(2)

with $\gamma$ > 0 and ϲ₁ ≤ ϲ₂ ≤…. ≤ ϲ_m. γ is the slope of the transition function and c = (ϲ₁, …, ϲ_m)’ is an m-dimensional vector of threshold parameters. For m = 1 and m = 2, there are one or two levels of CSR around which the effect on firm performance is nonlinear; [63] argue that considering m = 1 or m = 2 is sufficient because these values allow for frequently occurring types of variation in the parameters. This nonlinear effect is represented by a parameter continuum between the extreme regimes.

Equation (1) allows for the investigation of whether nonlinearity is related to various levels of CSR. Indeed, depending on the value of the CSR index, various CSR levels may have a distinct influence on financial performance. Whereas the elasticity in a linear model is constant and equal to 0 in Equation (1), the elasticities in the PSTR model vary across firms and over time depending on the value of the transition function.

Finally, [63] stressed the need of running a homogeneity or specification test prior to estimating Equation (1). This test determines whether or not utilizing a PSTR model is appropriate and, thus, allows a selection between using a linear model and a nonlinear model to estimate Equation (1).

4. Results and Discussions

The estimation process using the PSTR model consists of two steps. The first step is to perform the homogeneity test, and the second step is to estimate the Equation (1). The results of the two stages in addition to the results of the descriptive statistics and the correlation matrix are presented separately in the sections that follow.

4.1. Results of the Homogeneity Test

Table 1. LM and F tests of homogeneity (p-values).

	CSR	ENV-CSR	SOCIAL-CSR	GOV-CSR
	m = 1	m = 1	m = 1	m = 1
LM test	0.0030	0.0013	0.0021	0.0019
F test	0.0023	0.0022	0.0021	0.0022

Notes: LM and F are the Lagrange multiplier and F tests for linearity; H₀: linear model, H₁: PSTR model with m = 1.

shows the results of the homogeneity tests. The p-values of the Lagrange multiplier and Fisher-type tests for the null hypothesis of linearity versus the alternative of nonlinearity specifications are shown in Table 1.

The findings indicate that the null hypothesis of homogeneity is rejected at the 1% significance level. They further demonstrate that for m = 1, the rejection of linearity is significant. Here, m = 1 reveals that there is a single CSR threshold beyond which the impact of CSR on business performance becomes nonlinear, resulting in two extreme regimes. We followed [66] and assume that “there is just one threshold (m = 1), hence we did not check for the sample’s remaining nonlinearity. Furthermore, there are no aspects in the theoretical or empirical literature that may support the fact that m could become equivalent to 2 in our study”. Consequently, we proceeded to estimate our model using the PSTR model since the findings reveal that the effect of CSR on firm performance depends on the amount of sustainability performance.

4.2. Descriptive Statistics

Table 2. Descriptive statistics.

Variable	Mean	Std. Dev.	Min	Max
Tobin-Q	3.286	0.923	0.422	9.721
ROA	4.723	2.038	−1.121	35.699
ROE	3.043	1.021	−2.432	28.502
CSR	45.33	6.23	4.65	60.32
ENV-CSR	44.94	5.76	3.84	50.43
SOCIAL-CSR	46.16	4.54	3.12	54.04
GOV-CSR	39.48	4.04	5.93	55.93
Leverage	32.945	11.546	3.564	60.237
Sales	13.854	3.122	−6.456	25.433
Tangible	0.622	0.054	0.0132	0.988
Size	15.656	1.032	10.202	21.456

displays descriptive statistics of the variables used in the study. It displays the mean, standard deviation, and minimum and maximum values for each variable. Regarding the variables that measure corporate social responsibility, it appears from the results in Table 2 that the average CSR value for GCC companies is 45.33, indicating a relatively low level of CSR reporting (45 out of a possible 100). The CSR standard deviation is 6.23, demonstrating major disparities in CSR performance among the companies studied. These results are also confirmed when examining each individual component of the composite index, i.e., ENV-CSR, SOCIAL-CSR and GOV-CSR.

4.3. Correlation Matrix

The correlation matrix is shown in

Table 3. Correlation matrix.

	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)	VIF
Tobin-Q	1											2.22
ROA	0.322	1										3.23
ROE	0.295	0.756	1									3.66
CSR	0.435	0.587	0.589	1								2.98
ENV-CSR	0.345	0.487	0.423	0.821	1							2.27
SOCIAL-CSR	0.312	0.490	0.519	0.909	0.434	1						2.86
GOV-CSR	0.376	0.380	0.470	0.923	0.576	0.480	1					2.92
Lev	0.655	−0.564	−0.387	0.034	0.022	−0.036	0.098	1				3.66
Sales	0.213	0.410	0.603	0.136	0.029	0.017	0.078	−0.041	1			2.34
Tang	0.123	0.134	0.232	0.164	0.202	−0.017	0.128	0.234	0.138	1		1.71
Size	0.563	0.239	0.134	0.058	0.066	0.221	0.099	0.350	0.229	0.341	1	2.12

. Results show a positive link between the (CSR) composite index variable and the financial performance indicators. Furthermore, several CSR dimensions which are (ENV-CSR), (SOCIAL-CSR), and (GOV-CSR) are positively correlated with the various performance measures. In addition, every control variable has the expected sign and is statistically connected to the dependent variable. The many explanatory variables are weakly connected, which means they might be included in the same model. The variance inflation factor VIF values were much lower than the cutoff level of 5, indicating that multicollinearity is not a problem.

4.4. Results and Discusions of the PSTR Model

Table 4. CSR and firm financial performance: PSTR results using the aggregate CSR index.

	(1)	(2)	(3)
	ROA	ROE	Tobin-Q
LagROA	0.638 ***
	(0.000)
CSR	−0.031 *	−0.027 **	−0.028 *
	(0.054)	(0.019)	(0.053)
CSR× g (.)	0.082 **	0.072 **	0.066 ***
	(0.047)	(0.029)	(0.003)
Leverage	−0.302 ***	−0.247 ***	−0.032
	(0.000)	(0.000)	(0.234)
Sales-growth	0.0854 ***	0.0909	0.070 *
	(0.002)	(0.152)	(0.095)
Tangible	−0.012 **	−0.0233 *	0.023
	(0.020)	(0.076)	(0.234)
Size	0.030 *	0.045 *	0.076 **
	(0.066)	(0.071)	(0.043)
LagROE		0.502 ***
		(0.000)
LagTobin-Q			0.492 ***
			(0.001)
Transition parameters
CSR (c1,1)	44.34 ***	43.67 ***	44.89 *
	(0.004)	(0.000)	(0.062)
CSR (γ)	34.34 (0.873)	33.23 (0.673)	32.34 (0.435)
AIC criterion	−0.625	−0.625	−0.625
N. of Obs.	420	420	415

Notes: CSR represents the composite index. c_1,1 and γ represent the estimated location parameter and estimated slope parameter in Equation (2); the estimated location parameters can be interpreted directly as the level of CSR. *** p < 0.01, ** p < 0.05, * p < 0.1.

and

Table 5. CSR and firm financial performance: PSTR results using the individual CSR components.

	(1)	(2)	(3)
	ROA	ROE	Tobin-Q
LagROA	0.543 ***
	(0.000)
ENV-CSR	−0.039 *	−0.043 ***	−0.044 **
	(0.058)	(0.003)	(0.012)
ENV-CSR× g (.)	0.064 ***	0.071 ***	0.054 ***
	(0.001)	(0.000)	(0.001)
SOCIAL- CSR	−0.024 *	−0.038 **	−0.033 ***
	(0.063)	(0.020)	(0.002)
SOCIAL-CSR× g (.)	0.044 **	0.052 **	0.053 ***
	(0.047)	(0.020)	(0.003)
GOV-CSR	−0.021 *	−0.028 ***	−0.015 **
	(0.070)	(0.002)	(0.033)
GOV-CSR× g (.)	0.064 **	0.059 **	0.063 ***
	(0.037)	(0.030)	(0.002)
Leverage	−0.212 *	−0.327	−0.123 *
	(0.064)	(0.123)	(0.073)
Sales-growth	0.044 ***	0.096	0.080 *
	(0.001)	(0.142)	(0.085)
Tangible	−0.112 **	−0.113 *	0.033
	(0.030)	(0.066)	(0.164)
Size	0.040 *	0.055 *	0.066 **
	(0.076)	(0.074)	(0.023)
LagROE		0.402 ***
		(0.000)
LagTobin-Q			0.392 ***
			(0.001)
Transition parameters
ENV-CSR (c1,1)	41.32 ***	41.53 ***	42.04 *
	(0.003)	(0.000)	(0.072)
ENV-CSR (γ)	30.24 (0.773)	31.22 (0.543)	33.34 (0.604)
SOCIAL-CSR (c1,1)	46.12 **	46.02 ***	47.86 **
	(0.023)	(0.000)	(0.032)
SOCIAL-CSR (γ)	32.97 (0.673)	33.09 (0.533)	32.65 (0.622)
GOV-CSR (c1,1)	39.11 *	38.01 **	39.61 *
	(0.082)	(0.077)	(0.072)
GOV-CSR (γ)	31.66 (0.674)	31.11 (0.543)	32.43 (0.572)
AIC criterion	−0.329	−0.329	−0.329
N. of Obs.	420	420	415

Notes: ENV-CSR, SOCIAL-CSR, and GOV-CSR represent the various components of the composite index. c_1,1 and γ represent the estimated location parameter and estimated slope parameter in Equation (2); the estimated location parameters can be interpreted directly as the level CSR. *** p < 0.01, ** p < 0.05, * p < 0.1.

present the estimation results using the PSTR model. The various measures of financial performance are respectively used in columns 1, 2, and 3. The (CSR) composite index is included in the model in Table 4, whereas Table 5 reports the results using the multiple CSR dimensions, which are (ENV-CSR), (SOCIAL-CSR), and (GOV-CSR). Starting with the results in Table 4, findings indicate that the direct impact of the CSR composite index on firm performance, as measured by β₀, is significantly negative through all the regressions. This finding is consistent with some empirical evidence, which implies that CSR has a detrimental influence on financial performance [14,15,16]. It particularly corroborates the shift of focus hypothesis, according to which social responsibility for labor relations, social welfare, and environmental protection necessitates a shift in emphasis that does not maximize shareholder profit [17,19]. The greater investment of resources raises costs and has a negative impact on financial performance [19].

The second line of Table 4 confirms the homogeneity test results: The effects of CSR on financial performance appear to be highly nonlinear. In fact, the coefficient β₁, which is associated with the model’s nonlinear component, is always positive and significant at the 1% level. As a result, the influence of CSR on financial performance is conditional on the (CSR) level. As the (CSR) index moves from low to high values, this finding means that changes in financial performance with regard to CSR vary from β₀ to β₁. The transition between these severe regimes occurs around the endogenous location parameter c_1,1. Our finding corroborates previous studies that have demonstrated a nonlinear relationship between CSR and business performance. In particular, it is aligned with the findings of [27,36], who used the PSTR model to assess if the CSR index reflects a nonlinear two-regime relationship between CSR and firm value. Indeed, [27] found that CSR negatively impacted Taiwanese firm values when companies were in the early stages of investing in CSR. However, once CSR investment surpassed the transition point, the operational performance and reputation of the corporation improved. Furthermore, [36] examined nonlinearities between environmental CSR and firm performance among European firms. Their findings indicate various relationships spanning from nonlinear positive and nonlinear negative to inverted U-shaped relationships.

Furthermore, findings in Table 4 show that the predicted threshold for CSR is around 44. As a result, in the first regime, when the CSR level is less than 44, more investment in corporate social responsibility tends to hinder the firm performance. However, when the level of CSR exceeds the threshold of 44, investing in more CSR promotes the firm’s financial profitability. Therefore, companies below the threshold value are those which are not investing enough in social responsibility. In this case, CSR investments do not contribute to increasing business value unless they surpass the 44-point value transition threshold. These results may explain the heterogeneous findings in the empirical literature [12,14,15,16,27,30,41,60,62] by demonstrating that the CSR effect on financial performance could be positive or negative depending on the CSR investment level. Our findings can be interpreted as follows: Initial investments in CSR may incur high marginal costs; however, these costs are reduced by recurrent investments. Companies with little expertise or a poor reputation in CSR may also discover that the marginal advantage is small at first. However, their intangible assets, such as reputation [48], may grow as their CSR investment expands and receives more public notice, increasing marginal benefit. When the marginal benefit exceeds the cost (as indicated in Equation (1), CSR > C), CSR investment becomes a positive contributor to corporate value. In this paper, we claim that the shift of focus hypothesis is established at the outset of CSR investing. This is because if a corporation does not place a high value on CSR, the costs it incurs become external costs that are passed on to the public, lowering the firm’s operating costs. As a result, when a company first engages in CSR, it has to raise its operational expenses and risk losing its viability. That is, businesses must pay the opportunity cost of allocating resources to CSR without reaping significant advantages [27]. This endeavor may still be viewed as a chore by the general public and investors. Some of them may even consider it insufficient, lowering the worth of their firm as a result. External expenses are currently a significant portion of a firm’s internal costs. As a result, a company’s profits earned through stakeholder discussions can only be reduced by emphasizing CSR. This has various advantages, including reducing the problem of rising operating costs due to external costs, removing any impediments to a business’ activities, and improving operational efficiency, reputation, performance, and corporate value.

In order to further evaluate the findings in Table 4, the slope of the transition function (γ) should be interpreted. The transition’s speed is indicated by the slope parameter. The sharper the transition from one extreme regime to another, the greater the value of γ. Table 4 shows that the slope appears to be high. Consider a company with a CSR level that is barely below the 44 threshold. According to the transition function, investing more in CSR results in a relatively fast financial performance. Our result is different from [27]’s results where the authors found a threshold value of the CSR index equal to 13,082. Hence, our threshold value is higher than the [27] cutoff value. This may be explained by the fact that the threshold of 13,082 is obtained in the context of a developed market, i.e., Taiwanese firms. Our threshold is related to developing markets where companies might invest more in CSR in order to see its benefits on financial results.

Moving on to Table 5, these results confirm the preceding ones. Indeed, coefficients associated to the variables ENV-CSR, SOCIAL-CSR and GOV-CSR are negative and significant across all regressions. These results confirm the fact that the CSR dimensions’ direct impact on firm performance, as assessed by β0, is negative. Furthermore, the homogeneity test findings are confirmed in Table 5: (ENV-CSR), (SOCIAL-CSR), and (GOV-CSR) appear to have a significantly nonlinear impact on financial profitability. Particularly, at the 1 percent level the β1 coefficient, which is connected with the model’s nonlinear component, is always positive and significant. As a result, CSR’s impact on financial performance is dependent on its various components’ levels. These results are consistent with earlier findings [60,62] in the sense that they show that the effect of CSR on business performance may be different according to the CSR dimension. For instance, our result confirms the findings of [60], where the authors reported that the environmental responsibility performance had a negative correlation with firm performance; the governance responsibility performance had a positive correlation whereas the social responsibility performance did not have any significant correlation with firm performance.

Table 5 also yields interesting findings when looking at the various thresholds for the variables (ENV-CSR), (SOCIAL-CSR), and (GOV-CSR). First, it appears that the threshold value varies between the three variables. Indeed, the cutoff values for the variables (ENV-CSR), (SOCIAL-CSR), and (GOV-CSR) are respectively 41, 46, and 39. Consequently, financial performance is improved only when investment in environmental CSR exceeds the level of 41. The level of investment in social CSR should be higher than the threshold of 46 in order to increase firm profitability, whereas the company should invest higher than the threshold of 39 in governance CSR in order to promote its financial performance. Furthermore, another notable finding from Table 5 is that the variable SOCIAL-CSR is associated to the highest threshold level. Hence, the firm should invest more on the social aspects of CSR in order to see its profitability increased. This can be explained by the fact that when an issue happens, customers are more inclined to forgive a firm that works to connect with the community and create a favorable reputation on a regular basis. This is like being protected by a “CSR shield”. As a result, when a business’ social CSR investment reaches a particular level, the general public begins to connect with it; at that moment, CSR begins to successfully improve corporate value. This result provides additional support for the social impact hypothesis which proclaims that meeting the needs of stakeholders reflects positively on firm performance [26].

Finally, regardless of the specification employed, the bulk of the control variables have the predicted sign. Indeed, the majority of the control variables have the anticipated sign and are significant, as shown in Table 4 and Table 5. (Size) and (Sales-Growth) have positive and significant coefficients, whereas the variables (Leverage) and (Tangible) are negative and significant. These findings are aligned with the previous studies. Debts are, in reality, used to gauge a company’s risk. More debt, in particular, may raise the likelihood of default and lower firm performance. Because enterprises may use tangible assets as collateral to acquire funding, the variable (Tangible) has a negative and significant coefficient. As a result, asset tangibility and business profitability are negatively connected. Furthermore, the Sales-Growth coefficient is positive and significant. Therefore, strong sales growth is predicted to result in higher profit margins and increased company value. Finally, (Size) has a substantial positive coefficient indicating that huge firms are better able to negotiate attractive financing interest rates, which leads to enhanced efficiency.

5. Robustness Tests

Findings obtained thus far show that the influence of CSR on financial performance in GCC companies is nonlinear. We present additional evidence of the robustness of these results in this section. The results of the different robustness tests are shown in Appendix A and Appendix B in the appendices. The robustness of our results are checked using the ROA financial performance measure.

First, the robustness of the results was checked using other estimation methods. Particularly, we used the fixed effects model and the generalized method of moments to estimate the above relationship [67,68]. The GMM estimator has the benefit of producing instruments automatically. Thus, lag values were used to instrument endogenous variables. To do this, we estimated a model that includes the CSR interaction term to account for nonlinearity. The specification includes a quadratic interaction term to account for the nonlinear effects of the CSR level as a threshold variable (CSR²). With the addition of an interaction term, it is possible to see if at greater levels of this variable, CSR’s marginal effect changes.

Second, we also examined how sensitive our results were to the addition of more control variables. We particularly included two additional control variables: employees (Employee), defined as the number of employees in the firm. It indicates the important features of the manufacturing process. The second variable is the firm age (Age), defined as the number of years since the company was founded. Indeed, because the sample chosen for the study covers organizations at various phases of growth and is diversified across industries, firm age is an essential control variable. Older firms may have a stronger presence in the CSR and are consequently deemed more CSR focused.

As shown in Appendix A and Appendix B, our main results remain unchanged even after using other estimation methods and integrating additional control variables.

6. Conclusions

A growing emphasis on CSR activities has accompanied global economic progress. Most prior research argued that CSR and corporate value have a positive, a negative, or a neutral linear connection. Some argue that engaging in CSR enhances the business image and attracts additional resources, resulting in improved operational success. From this vantage point, CSR has a beneficial influence on the value of a company. Other research, on the other hand, claims that investing in CSR increases expenses and reduces operational performance, lowering competitiveness and, as a result, lowering company value. Similarly, the empirical literature highlights the inconclusive results of studies that have examined the relationship between CSR and firm performance in different contexts. Consequently, this research strove to contribute to the ongoing debate by investigating the potential nonlinearity of the relationship between CSR and financial performance. Our paper relied on the PSTR model in order to test if the relationship between CSR and firm performance is nonlinear. Moreover, it examined whether the relationship between both constructs depends on the various dimensions of CSR (environmental, governance, and social aspects).

The study used a sample of 70 publicly traded companies in GCC countries from 2015 to 2020. Results indicate that the relationship between CSR and corporate performance is indeed nonlinear. Particularly, findings suggest that in the first regime, when the CSR level is less than 44, more investment in corporate social responsibility tends to hinder the firm performance. However, when the level of CSR exceeds the threshold of 44, investing in more CSR promotes the firm’s financial profitability. Therefore, companies below the threshold value are those which are not investing enough in corporate social responsibility, and CSR expenditures do not contribute to boosting company value until they reach above the 44-point value transition barrier. These findings indicate that initial investments in CSR might incur high marginal costs; however, these costs are reduced by recurrent investments. Companies with limited competence or a bad reputation in CSR may find that the marginal advantage is initially minimal. Nevertheless, when their CSR investment expands and attracts more public attention, their intangible assets, such as reputation, rise as well, boosting marginal benefit. When the marginal benefit exceeds the cost, CSR investment becomes a positive contributor to corporate value.

Furthermore, results indicate that the nonlinear relationship persists when using the individual CSR dimensions, i.e., governmental, social, and environmental CSR. Most importantly, it is found that the highest threshold level is associated with the social CSR dimension. Hence, the company should invest more in CSR’s social dimension in order to boost its profitability. This indicates that when an issue happens, customers are more inclined to forgive a firm that works to connect with the community and create a favorable reputation on a regular basis. This is like being protected by a “CSR shield”. As a result, when a business’ social CSR investment reaches a particular level, the public starts to connect with it; at that moment, CSR begins to successfully improve corporate value.

Policymakers might benefit from this study’s unique implications, using these results to improve company sustainability practices in GCC countries. This study revealed that the average CSR disclosure score of GCC enterprises is quite low. This might be due to a lack of legislative push in certain countries to implement CSR policies. According to the findings, consumers and investors are placing a higher emphasis on CSR, prompting many firms to participate in the community in order to avoid friction with key stakeholders. Furthermore, our findings suggest that CSR expenditures improve an organization’s corporate image, promote a better understanding of its goods and services, and, most importantly, develop its ties with a wide range of stakeholders. Interestingly, excellent CSR performance may increase investor trust, thus helping with long-term growth. As a result, governments should emphasize the construction of corporate governance and CSR frameworks in order to protect investors’ interests and strengthen the financial market’s health. Government regulatory bodies in GCC nations should impose and encourage ESG disclosure through legislative mandates and recommendations in order to enhance it. Finally, our findings can assist organizations’ management better understand CSR’s link with and relevance to long-term business performance.

Finally, this study might have a significant shortcoming in that it does not account for sectoral disparities in CSR. Firms in the same industry are more inclined to emulate each other’s CSR spending. Inter-industry CSR spending may thus have a different median value than intra-industry CSR. Future studies may look at specific sectors and assess the impact of CSR on corporate performance across industries. Future empirical models may also include economic cycle variations caused by the COVID-19 pandemic to examine the relationship between both constructs.