From Social Ties to Social Responsibility: How Social Capital Shapes CSR Practices Around the World

Abstract

This paper uses data from 61 countries to document the relationship between country-level social capital and corporate social responsibility (CSR) during the period between 2007 and 2023. We find that firms headquartered in countries with higher levels of social capital exhibit higher CSR engagement than firms headquartered in countries with lower levels of social capital. This association remains robust across a wide range of estimation approaches and sensitivity checks. Additional analyses explore the mechanisms through which social capital affects CSR by disentangling its underlying dimensions. Finally, we show that the economic value of CSR (as measured through its association with Tobin’s Q, sales growth, stock return volatility, and financial distress risk) varies with the level of social capital, with CSR playing a more prominent value-enhancing role in low-social-capital settings.

1. Introduction

Corporate social responsibility (CSR) has become a central element of firms’ strategic decision-making and an increasingly important dimension of corporate governance.1 Over the past two decades, firms have substantially expanded their investments in CSR-related activities. This increase is motivated by reputational considerations, stakeholder engagement, and long-term value creation.2 For example, firms in the United States and the United Kingdom alone spent more than $15 billion on CSR initiatives in 2014. Prior research suggests that firms use CSR to enhance corporate reputation (Crane et al., 2019), strengthen relationships with key stakeholders, and ultimately improve financial performance (Margolis & Walsh, 2003; Orlitzky et al., 2003; Scherer, 2013; Mishra & Suar, 2010; Cheng et al., 2014). In parallel, investors increasingly incorporate CSR considerations into capital allocation decisions, with socially responsible investment now representing a core component of European fund management (EUROSIF, 2018).3

Given the growing economic relevance of CSR, a large number of studies have examined the determinants of firms’ engagement in socially responsible activities. Existing studies highlight the role of firm characteristics, ownership structure, board composition, competitive pressures, regulatory frameworks, and managerial traits in shaping CSR outcomes (Scherer, 2013; Cheng et al., 2014; Orlitzky et al., 2003). However, despite the recognition that corporate behavior is embedded in broader social and cultural contexts, our understanding of how informal social institutions influence CSR remains limited. In particular, relatively little is known about whether and how social capital at the country level shapes firms’ incentives to engage in CSR across heterogeneous institutional environments.

Social capital, commonly understood as the shared norms, trust, and social networks that facilitate cooperation within a society, has long been recognized as an important determinant of economic behavior (Coleman, 1990; Putnam, 1995). A small but growing literature links social capital to CSR, showing that firms located in regions with higher social capital tend to exhibit stronger social responsibility. Notably, Jha and Cox (2015) and Hoi et al. (2018) document a positive association between social capital and CSR within a single institutional setting. While these studies provide important insights, their within-country focus limits the ability to assess how social capital operates in environments characterized by substantial variation in legal systems, investor protection, labor market institutions, and macroeconomic stability. This study advances the literature by adopting a cross-country perspective and conceptualizing social capital as an informal governance mechanism that shapes corporate behavior alongside, and potentially in substitution for, formal institutions. Importantly, our contribution is not only to extend prior evidence across countries, but to clarify what CSR represents under different informal institutional conditions. We argue that social capital strengthens CSR through normative expectations and reputational enforcement, making CSR closer to a baseline, norm-consistent behavior where informal discipline is strong. Conversely, when social capital is weak, CSR is more likely to function as a credibility-building mechanism that mitigates trust deficits and legitimizes firms in the eyes of stakeholders. This framing yields a theory-based implication: social capital should not only predict CSR levels but also condition CSR’s marginal value relevance across contexts.

By examining firms operating across diverse legal, regulatory, and economic contexts, we address a key limitation of prior research and provide a richer understanding of the conditions under which social capital influences CSR. In doing so, we move beyond documenting an association and instead investigate how and when social capital matters for socially responsible corporate conduct. Specifically, this paper addresses three interrelated research questions. First, does social capital prevailing at the country level systematically influence firms’ engagement in CSR across heterogeneous institutional environments? Second, through which underlying mechanisms, such as civic participation, interpersonal trust, and social networks, does social capital affect CSR? Third, does the economic value of CSR depend on the level of social capital, such that CSR plays different roles in environments characterized by strong versus weak informal social norms?

The theoretical foundation of this study builds on the idea that social capital promotes adherence to shared norms and constrains opportunistic behavior by increasing the social costs of deviation (Coleman, 1990; Elster, 1989; Spagnolo, 1999). In societies with high social capital, firms may face stronger informal pressures to act in socially responsible ways, independent of formal regulatory enforcement. Conversely, in low-social-capital environments, CSR may serve as a strategic tool to compensate for weak social trust and limited informal discipline. This perspective suggests that social capital not only influences the level of CSR but may also condition the extent to which CSR contributes to firm value.

To examine these issues, we analyze a large panel of non-financial firms from 61 countries over the period 2007–2023. By integrating country-level measures of social capital with firm-level CSR and performance data, this study contributes to the CSR and the institutional economics literature in three ways. First, it theorizes and tests social capital as an informal governance institution that shapes CSR through normative pressure and reputational enforcement across heterogeneous environments. Second, it clarifies the mechanisms by showing that multiple dimensions of social capital (e.g., participation, trust, and networks) are consistently associated with CSR, consistent with both norm-based and monitoring-based channels. Third, it demonstrates that CSR’s economic meaning is context-dependent: the marginal value relevance of CSR is weaker where social capital already disciplines corporate behavior and stronger where informal discipline is limited.

The remainder of the paper is structured as follows: Section 2 develops the hypothesis. Section 3 summarizes the data and presents the methodology. Section 4 presents an assessment of our arguments. Section 5 presents additional tests. The paper concludes with Section 6.

2. Literature Review and Hypothesis Development

This section develops a theoretical framework that explains how social capital operates as an informal institutional mechanism shaping CSR. Drawing on institutional economics and social capital theory, we conceptualize social capital as a set of shared norms, trust, and social networks that influence managerial incentives, stakeholder expectations, and reputational constraints, thereby affecting firms’ CSR behavior. Formal institutions such as laws, regulations, and enforcement mechanisms play a central role in governing corporate conduct. However, prior research in institutional economics emphasizes that economic behavior is also shaped by informal institutions, including norms and social conventions, which guide behavior even in the absence of formal sanctions (North, 1990). Social capital represents a key informal institution that structures interactions among individuals and organizations by fostering cooperation and reducing opportunism (Putnam, 1995; Fukuyama, 1997). In this framework, CSR is not viewed solely as a discretionary managerial choice or a strategic response to regulation, but as an outcome that reflects the social environment in which firms are embedded. Firms operating in societies with higher levels of social capital face stronger normative expectations and social scrutiny, which shape both managerial decision-making and stakeholder responses to corporate behavior.

2.1. Theoretical Framework

Social capital is a multifaceted construct that has been examined extensively across economics, sociology, and finance. At the micro level, social capital refers to the networks and relationships individuals use to access financial and human capital (Burt et al., 2000; Burt, 1992). At the macro level, social capital is defined as the norms, trust, and shared values that facilitate cooperation and coordination for mutual benefit within societies (Putnam, 1995; Portes & Sensenbrenner, 1993). These definitions collectively highlight social capital as a form of intangible resource embedded in social structures that shape individual and organizational behavior. Prior literature identifies two core dimensions of social capital. The first dimension relates to shared norms and social trust, which define informal rules governing interactions among economic agents and reduce uncertainty in social and economic exchanges (Guiso et al., 2004; Fukuyama, 1997). The second dimension concerns the presence of dense social networks, which facilitate information flow, monitoring, and enforcement of socially acceptable behavior (Coleman, 1990; Portes & Sensenbrenner, 1993; Payne et al., 2011; Hasan et al., 2015). Together, these dimensions suggest that social capital operates as an informal institutional mechanism that complements or substitutes for formal governance structures.

2.1.1. Social Capital, Shared Norms, and Normative Pressure

Shared norms constitute a central dimension of social capital and define socially acceptable behavior within a society. Norms reduce uncertainty by clarifying expectations about appropriate conduct and by shaping individuals’ beliefs about how others are likely to behave (Bicchieri, 2006; Sunstein, 1996). In societies characterized by high social capital, individuals tend to exhibit higher levels of trust and cooperation, which reflects adherence to shared norms and facilitates reciprocal interactions (Putnam, 1995; Fukuyama, 1997).

Within this framework, CSR can be viewed as a form of norm-conforming corporate behavior. Firms headquartered in high-social-capital environments face strong normative expectations to respect stakeholder interests, including employee welfare, human rights, and community well-being. Managers operating in such contexts are therefore more likely to internalize these societal expectations when making corporate decisions. Prior literature also suggests that ethical considerations play an important role in decision-making within cohesive societies, where deviations from socially accepted behavior may entail reputational costs (Akerlof, 2007; Graafland & Van de Ven, 2006). Managers are not isolated agents; rather, they are embedded within social environments that influence their preferences and values. Empirical studies indicate that firms located in high-social-capital regions tend to attract managers whose values align with those of the surrounding society, reinforcing norm-consistent behavior (Hilary & Hui, 2009; Vroom, 1966).

As a result, shared norms and trust generate normative pressure that encourages firms to engage in CSR not only as a strategic choice, but as an expected and socially sanctioned form of behavior. In contrast, in low-social-capital environments, where trust is limited and short-term self-interest dominates, such normative pressure is weaker, reducing incentives for socially responsible conduct.

2.1.2. Social Networks, Monitoring, and Social Enforcement

A second key dimension of social capital is the presence of dense social networks. Social networks facilitate the flow of information and enable the monitoring of individual and organizational behavior (Coleman, 1990; Portes, 1998). In societies with dense networks, information about unethical or uncooperative behavior spreads more easily, increasing the likelihood that deviant actions are detected and sanctioned (Putnam, 2000). From a theoretical perspective, dense social networks strengthen CSR by increasing the reputational costs of socially irresponsible behavior. Firms operating in such environments are subject to informal monitoring by stakeholders, communities, and peer organizations, which constrains opportunistic behavior even in the absence of formal regulatory enforcement.

Social network theory emphasizes that reputational concerns play a critical role in sustaining cooperation. Individuals and organizations that violate social expectations may face informal sanctions, such as loss of trust, social exclusion, or diminished access to valuable resources (Coleman, 1990). Conversely, actors who enforce norms may be rewarded with social approval and enhanced status, reinforcing collective monitoring mechanisms. Applied to corporate behavior, these mechanisms imply that managers in high-social-capital societies face strong incentives to maintain a positive social reputation by engaging in CSR activities. CSR, thus, emerges as a socially enforced action, where firms comply with stakeholder expectations to avoid reputational penalties and preserve access to social and economic resources.

2.1.3. Social Capital, CSR, and Institutional Substitution

Shared norms combined with dense social networks position social capital as a partial substitute for weak formal institutions. In environments where legal enforcement and regulatory oversight are limited, social capital can fill institutional voids by disciplining corporate behavior through informal mechanisms.

This perspective aligns with the broader institutional economics literature, which emphasizes that informal institutions can complement or substitute for formal governance structures in shaping economic outcomes (North, 1990; Banfield, 1967). In the context of CSR, social capital may, therefore, not only influence the level of socially responsible behavior but also affect the strategic and economic role of CSR across different institutional environments. While social capital can mitigate certain governance shortcomings by disciplining behavior through informal mechanisms, it does not replace the need for strong formal institutions, such as investor protection and contract enforcement, which remain fundamental for sustainable corporate governance and economic development.

2.2. Hypothesis Development

As discussed above, social capital shapes corporate behavior by fostering shared norms, trust, and cooperation, and by constraining opportunistic actions through informal social enforcement. In high-social-capital societies, firms face stronger normative expectations to behave responsibly toward stakeholders, while managers internalize these expectations as part of socially acceptable conduct. Deviations from socially responsible behavior are more likely to entail reputational costs, making CSR a socially sanctioned action rather than a discretionary choice. Accordingly, we expect firms headquartered in countries with higher levels of social capital to exhibit higher levels of corporate social responsibility. Hence, we advance the following hypothesis:

Hypothesis 1 (H1). 

Firms headquartered in countries with higher levels of social capital exhibit higher levels of corporate social responsibility.

3. Data and Methodology

3.1. Sample

This paper uses the data of non-financial firms from 61 countries to document the effect of social capital prevailing in the country on CSR. The sample period spans from 2007 to 2023 and comprises 61,521 firm-year observations. The sample includes all firms for which the data on variables used in this paper is available.

3.2. Methodology

To document the effect of social capital prevailing in the country on CSR, we estimate various versions of the following pooled OLS regression equation.

$$\begin{array}{cc}\hfill \mathrm{C}\mathrm{S}\mathrm{R}=\mathsf{\alpha } +& {\mathsf{\beta }}_1\left(\mathrm{S}\mathrm{O}\mathrm{C}\mathrm{I}\mathrm{A}\mathrm{L}\right)+{\mathsf{\beta }}_2\left(\mathrm{S}\mathrm{I}\mathrm{Z}\mathrm{E}\right)+{\mathsf{\beta }}_3\left(\mathrm{L}\mathrm{E}\mathrm{V}\mathrm{E}\mathrm{R}\mathrm{A}\mathrm{G}\mathrm{E}\right)+{\mathsf{\beta }}_4\left(\mathrm{L}\mathrm{O}\mathrm{S}\mathrm{S}\right)+{\mathsf{\beta }}_5\left(\mathrm{G}\mathrm{R}\mathrm{O}\mathrm{W}\mathrm{T}\mathrm{H}\right)+{\mathsf{\beta }}_6\left(\mathrm{D}\mathrm{I}\mathrm{V}\mathrm{I}\mathrm{D}\mathrm{E}\mathrm{N}\mathrm{D}\right)\hfill \\ & +{\mathsf{\beta }}_7\left(\mathrm{C}\mathrm{A}\mathrm{P}\mathrm{E}\mathrm{X}\right)+{\mathsf{\beta }}_8\left(\mathrm{B}\mathrm{E}\mathrm{T}\mathrm{A}\right)+{\mathsf{\beta }}_9\left(\mathrm{A}\mathrm{N}\mathrm{A}\mathrm{L}\mathrm{Y}\mathrm{S}\mathrm{T}\right)+{\mathsf{\beta }}_{10}\left(\mathrm{I}\mathrm{N}\mathrm{V}\mathrm{E}\mathrm{N}\mathrm{T}\mathrm{O}\mathrm{R}\mathrm{Y}\right)+{\mathsf{\beta }}_{11}\left(\mathrm{R}\mathrm{E}\mathrm{T}\mathrm{A}\mathrm{I}\mathrm{N}\mathrm{E}\mathrm{D}\right)\hfill \\ & +{\mathsf{\beta }}_{12}\left(\mathrm{E}\mathrm{N}\mathrm{F}\mathrm{O}\mathrm{R}\mathrm{C}\mathrm{E}\mathrm{M}\mathrm{E}\mathrm{N}\mathrm{T}\right)+{\mathsf{\beta }}_{13}\left(\mathrm{P}\mathrm{R}\mathrm{O}\mathrm{T}\mathrm{E}\mathrm{C}\mathrm{T}\mathrm{I}\mathrm{O}\mathrm{N}\right)+{\mathsf{\beta }}_{14}\left(\mathrm{S}\mathrm{T}\mathrm{A}\mathrm{B}\mathrm{I}\mathrm{L}\mathrm{I}\mathrm{T}\mathrm{Y}\right)+{\mathsf{\beta }}_{15}\left(\mathrm{C}\mathrm{O}\mathrm{M}\mathrm{P}\mathrm{E}\mathrm{T}\mathrm{I}\mathrm{T}\mathrm{I}\mathrm{V}\mathrm{E}\right)\hfill \\ & +{\mathsf{\beta }}_{16}\left(\mathrm{L}\mathrm{A}\mathrm{B}\mathrm{O}\mathrm{R}\right)+{\mathsf{\beta }}_{17}\left(\mathrm{F}\mathrm{I}\mathrm{S}\mathrm{C}\mathrm{A}\mathrm{L}\right)+{\mathsf{\beta }}_{18}\left(\mathrm{G}\mathrm{D}\mathrm{P}\right)+{\displaystyle \sum _{\mathrm{Y}=1}^{\mathrm{N}-1}}{\mathsf{\gamma }}_{\mathrm{Y}}\left(\mathrm{Y}\mathrm{D}\mathrm{U}\mathrm{M}\right)+{\displaystyle \sum _{\mathrm{I}=1}^{\mathrm{N}-1}}{\mathsf{\phi }}_{\mathrm{I}}\left(\mathrm{I}\mathrm{D}\mathrm{U}\mathrm{M}\right)+\mathsf{\epsilon }\hfill \end{array}$$

(1)

In the above regression equation, the dependent variable (CSR) measures the corporate social responsibility. For this paper, we use “Refinitiv’s social pillar score” as a measure of corporate social responsibility.4 This variable indicates the capacity of a firm to generate trust and loyalty in its workforce, customers, and society. This variable takes the values between 0 and 100. with higher values indicating higher levels of corporate social responsibility. The main independent variable (SOCIAL) measures the level of social capital prevailing in a country. It is an aggregate index that is based on five factors. These factors are civic and social participation, social tolerance, interpersonal trust, personal and family relationships, and social networks. This variable is obtained from the Legatum Institute and is used in several other studies, such as Jabbouri et al. (2024), Pasiouras and Samet (2022) and Papadimitri et al. (2021).

Furthermore, Equation (1) includes several firm-specific and country-specific characteristics as control variables. These variables are expected to affect CSR to varying degrees. The firm-specific characteristics include log of market capitalization in dollars (SIZE), total debt to total assets ratio (LEVERAGE), dummy variable representing a loss-making firm (LOSS), one-year growth in total assets (GROWTH), dividend yield (DIVIDEND), capital expenditures to total assets ratio (CAPEX), market risk of a firm (BETA), number of analysts covering a firm (ANALYST), total inventory to total assets ratio (INVENTORY), and retained earnings to total assets ratio (RETAINED). The country-specific characteristics include the extent of contract enforcement (ENFORCEMENT), strength of investor protection mechanisms (PROTECTION), extent of macroeconomic stability (STABILITY), level of competitiveness (COMPETITIVE), extent of labor market flexibility (LABOR), level of fiscal sustainability (FISCAL), and log of gross domestic product in dollars (GDP). In addition, the regression equation also includes industry dummies (IDUM) and year dummies (YDUM) to control for industry-specific and year-specific effects on CSR. See

Table 1. Definition of variables.

Variable	Definition	Source
CSR	Corporate social responsibility index	Thomson Reuters Eikon
SOCIAL	Social capital index	Legatum Institute
SIZE	Natural log of market capitalization (in dollars)	Worldscope
LEVERAGE	${\displaystyle \frac{\mathrm{T}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l} \mathrm{D}\mathrm{e}\mathrm{b}\mathrm{t}}{\mathrm{T}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l} \mathrm{A}\mathrm{s}\mathrm{s}\mathrm{e}\mathrm{t}\mathrm{s}}}$	Worldscope
LOSS	$\left\{\begin{array}{c}1, \mathrm{If} \mathrm{Earnings} \mathrm{Before} \mathrm{Interest} \mathrm{and} \mathrm{Taxes}<0\\ 0, \mathrm{If} \mathrm{Earnings} \mathrm{Before} \mathrm{Interest} \mathrm{and} \mathrm{Taxes}\ge 0\end{array}\right.$	Worldscope
GROWTH	One-year growth in total assets	Worldscope
DIVIDEND	Dividend yield	Worldscope
CAPEX	${\displaystyle \frac{\mathrm{C}\mathrm{a}\mathrm{p}\mathrm{i}\mathrm{t}\mathrm{a}\mathrm{l} \mathrm{E}\mathrm{x}\mathrm{p}\mathrm{e}\mathrm{n}\mathrm{d}\mathrm{u}\mathrm{t}\mathrm{u}\mathrm{r}\mathrm{e}\mathrm{s}}{\mathrm{T}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l} \mathrm{A}\mathrm{s}\mathrm{s}\mathrm{e}\mathrm{t}\mathrm{s}}}$	Worldscope
BETA	Sensitivity of stock returns to market returns (based on the linear regression of the logarithmic adjusted returns and the returns of the corresponding market index—using previous five years of monthly data)	Datastream
ANALYST	Total number of analysts covering a firm	Thomson Reuters Eikon
INVENTORY	${\displaystyle \frac{\mathrm{T}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l} \mathrm{I}\mathrm{n}\mathrm{v}\mathrm{e}\mathrm{s}\mathrm{n}\mathrm{t}\mathrm{o}\mathrm{r}\mathrm{y}}{\mathrm{T}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l} \mathrm{A}\mathrm{s}\mathrm{s}\mathrm{e}\mathrm{t}\mathrm{s}}}$	Worldscope
RETAINED	${\displaystyle \frac{\mathrm{R}\mathrm{e}\mathrm{t}\mathrm{a}\mathrm{i}\mathrm{n}\mathrm{e}\mathrm{d} \mathrm{E}\mathrm{a}\mathrm{r}\mathrm{n}\mathrm{i}\mathrm{n}\mathrm{g}\mathrm{s}}{\mathrm{T}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l} \mathrm{A}\mathrm{s}\mathrm{s}\mathrm{e}\mathrm{t}\mathrm{s}}}$	Worldscope
YDUM	Set of year dummies	Worldscope
IDUM	Set of industry dummies	Worldscope
ENFORCEMENT	Contract enforcement index represents the efficacy and efficiency of a country’s system to enforce the rights of a contract holder. It is a function of quality of judicial administration, time to resolve commercial cases, legal costs and alternate dispute resolution mechanisms.	Legatum Institute
PROTECTION	Investor protection index represents the degree of investor protection. It includes strength of insolvency framework, insolvency recovery rate, auditing and reporting standards, conflict of interest regulations, and extent of shareholder governance.	Legatum Institute
STABILITY	Macroeconomic stability index represents the macroeconomic stability. It is based on the GDP per capita growth rate and the volatility of the inflation rate.	Legatum Institute
COMPETITIVE	Competitiveness index represents the efficiency with which inputs can be converted into outputs and the level of diversification in the economy. It is based on labor productivity, export quality, high-tech manufactured exports, and economic complexity.	Legatum Institute
LABOR	Labor flexibility index represents the flexibility of the workplace in terms of employment contracts. It includes cooperation in labor-employer relations, flexibility of hiring practices, redundancy costs, flexibility of employment contracts, and flexibility of hiring practices.	Legatum Institute
FISCAL	Fiscal sustainability index represents the ability of a government to sustain its current spending, tax, and other policies in the medium-to-long term. It is based on government budget balance, government debt, country’s credit rating, country risk premium, and gross savings.	Legatum Institute
GDP	Log of gross domestic product in dollars	World Bank
PERFORM	Variable representing the performance of a firm. It is measured by one of the following variables: (1) Standard deviation of returns over the past five years, (2) Tobin’s Q, (3) one year growth in sales, and (4) Altman Z-Score.	Worldscope
CIVIC	Civic and social participation index	Legatum Institute
TRUST	Interpersonal trust index	Legatum Institute
RELATIONSHIP	Family and personal relationship index	Legatum Institute
NETWORKS	Social networks index	Legatum Institute
TOLEARANCE	Social tolerance index	Legatum Institute

for definition of these variables.

3.3. Summary Statistics

Table 2. Average values of corporate social responsibility and social capital.

Country	Observations	CSR	SOCIAL
Argentina	197	29.6133	58.6310
Australia	3445	38.6792	76.6983
Austria	263	57.7797	68.6380
Bahrain	27	15.4866	59.7889
Belgium	376	53.7043	62.3279
Brazil	790	53.1682	56.9771
Bulgaria	1	37.8657	54.6000
Canada	3039	39.3765	75.1208
Chile	327	42.1422	57.2437
China	5185	31.3509	51.6309
Colombia	105	60.1401	58.3152
Cyprus	23	22.2482	51.4348
Denmark	427	55.9691	81.0391
Egypt	51	14.1763	37.7549
Finland	508	58.9892	73.9398
France	1386	65.0039	60.6841
Germany	1627	57.7462	68.7760
Greece	187	43.0947	46.7690
Hong Kong	1690	43.5848	59.2579
Hungary	56	54.3745	49.9071
India	1890	49.0460	47.3641
Indonesia	373	48.2010	57.6239
Ireland	429	52.0331	73.1664
Israel	169	37.2396	51.9308
Italy	601	63.2167	60.5358
Japan	5353	40.8462	43.8319
Jordan	1	33.0617	40.9000
Kenya	9	40.1819	57.0444
Kuwait	54	36.8044	55.5593
Luxembourg	182	57.6907	65.8934
Malaysia	1000	45.7592	54.3936
Malta	33	31.6222	68.8697
Mexico	421	48.8985	52.8834
Morocco	10	36.1547	39.5400
The Netherlands	571	66.0362	76.7112
New Zealand	481	37.3514	78.8004
Nigeria	2	55.4082	53.0000
Norway	368	56.2388	79.0533
Oman	29	10.6124	54.1379
Pakistan	15	34.6199	47.4267
Peru	150	39.1349	50.7653
Philippines	240	41.5024	59.9721
Poland	245	37.5898	48.5449
Portugal	104	69.5203	58.4962
Qatar	110	14.1238	55.3791
Romania	11	44.6387	48.7636
Russia	341	37.6743	45.5164
Saudi Arabia	163	24.0137	50.5515
Singapore	638	44.1811	58.7580
South Africa	1113	52.0485	57.2291
South Korea	1163	42.3926	46.7132
Spain	548	71.9677	66.9434
Sri Lanka	11	59.7202	54.8636
Sweden	1469	52.0544	76.1624
Switzerland	1102	50.8703	70.7787
Thailand	652	59.0425	58.3201
Turkey	381	55.8003	42.7412
United Arab Emirates	127	32.0473	56.3118
United Kingdom	3896	50.4442	70.3672
United States of America	17,314	42.9732	72.3772
Vietnam	42	26.1733	62.6810

Note: The table reports the average values of social capital index (SOCIAL) and corporate social responsibility (CSR) along with total observations in each country. The sample period is from 2007 to 2023. See Table 1 for definition of variables.

documents the average values of main variables (CSR and SOCIAL) in each country along with total number of observations. The table shows that there is a considerable variation in CSR across countries. For example, firms headquartered in Oman, Egypt, Qatar, Bahrain, Saudi Arabia, and Cyprus have the lowest scores on social responsibility. The average value of this variable is less than 25 for firms headquartered in these countries. In contrast, firms headquartered in Spain, Portugal, the Netherlands, and France score relatively high on social responsibility. The average value of this variable is more than 65 for firms headquartered in these countries. As expected, the table shows high levels of CSR among firms headquartered in developed countries. Similar divergence is observed in the level of social capital across countries. Morocco, Turkey, Jordan, and Egypt have the lowest scores on social capital, while the Netherlands, New Zealand, Norway, and Denmark have the highest scores on social capital. It also indicates that, on average, more developed countries have higher social capital than less developed countries.5

Table 3. Descriptive statistics.

Variables	25th Percentile	Mean	Median	75th Percentile	Standard Deviation	Observations
CSR	25.6613	44.8894	43.2667	63.4580	23.8960	61,521
SOCIAL	55.7000	64.0828	70.4000	72.8000	11.6895	61,521
SIZE	13.7064	14.7421	14.8421	15.8295	1.6862	61,521
LEVERAGE	11.0800	25.8035	24.6700	37.6000	18.4973	61,521
LOSS	0.0000	0.1620	0.0000	0.0000	0.3684	61,521
GROWTH	−0.9700	12.3153	5.7500	15.4700	34.7379	61,521
DIVIDEND	0.0000	2.0744	1.5000	3.1800	2.3118	61,521
CAPEX	0.0152	0.0467	0.0333	0.0618	0.0475	61,521
BETA	0.7017	1.0864	1.0300	1.4000	0.5856	61,521
ANALYST	4.0000	10.6601	9.0000	16.0000	8.7522	61,521
INVENTORY	0.0065	0.0946	0.0605	0.1460	0.1098	61,521
RETAINED	0.0341	0.0633	0.2105	0.3904	0.8389	61,521
ENFORCEMENT	83.2000	83.4828	88.9000	91.2000	11.6184	61,521
PROTECTION	71.6000	72.6973	77.0000	78.7000	10.5371	61,521
STABILITY	61.9000	65.0056	64.7000	66.8000	7.0017	61,521
COMPETITIVE	76.3000	81.8431	87.3000	89.2000	10.3361	61,521
LABOR	63.6000	67.0779	69.9000	73.0000	10.1187	61,521
FISCAL	51.1000	59.4074	56.2000	68.0000	11.7827	61,521
GDP	27.8208	28.8045	28.7762	30.4744	1.5319	61,521

Note: The table documents the descriptive statistics for variables used in this paper. The sample period is from 2007 to 2023. See Table 1 for definition of variables.

documents the descriptive statistics for control variables used in this study. The table shows that most of the firms have positive earnings before interest and taxes. Only around 16 percent of firms generate losses. The growth rate of an average firm is 12.31 percent and the average analyst coverage is around 10.66. The dividend yield is around 2 percent and the average firm looks very similar to the market with the beta of around 1. The table also shows that almost 25 percent of firm’s assets are financed by debt and around 9 percent of assets are held in inventory. We argue that average characteristics of the firms included in this analysis are relatively strong. It may be due to the fact that well-established firms are more likely to disclose information on CSR in each country.

Table 4. Difference between the average values of control variables for firms headquartered in high-social-capital countries (75th percentile of social capital) and firms headquartered in low social capital countries (25th percentile of social capital).

Variables	Low-Social-Capital Countries	High-Social-Capital Countries	Difference
SIZE	15.1319	14.5194	0.6124 *** (34.0658)
LEVERAGE	24.1947	25.2657	−1.0710 *** (−5.2677)
LOSS	0.0743	0.1970	−0.1227 *** (−31.9648)
GROWTH	10.5653	13.7963	−3.2310 *** (−8.5668)
DIVIDEND	2.0723	2.1550	−0.0826 *** (−3.2353)
CAPEX	0.0497	0.0531	−0.0033 *** (−5.8627)
BETA	1.0214	1.1234	−0.1019 *** (−16.3815)
ANALYST	11.1483	10.2807	0.8675 *** (8.7238)
INVENTORY	0.1112	0.0817	0.0295 *** (24.6764)
RETAINED	0.2888	−0.0560	0.3448 *** (42.5656)
ENFORCEMENT	76.3541	90.1441	−13.7889 *** (−120.00)
PROTECTION	65.3885	78.6496	−13.2610 *** (−130.00)
STABILITY	67.4960	64.0000	3.4959 *** (38.8262)
COMPETITIVE	77.8624	80.6002	−2.7378 *** (−21.8228)
LABOR	61.4165	70.4497	−9.0332 *** (−71.3147)
FISCAL	58.7055	63.4501	−4.7446 *** (−34.6142)
GDP	28.5837	28.5015	0.0822 *** (5.2331)

Note: This table reports the difference between the average values of control variables for firms headquartered in high-social-capital countries (more than 75th percentile of social capital) and firms headquartered in low-social-capital countries (below 25th percentile of social capital). The t-values are presented in parentheses. The sample period is from 2007 to 2023. The symbols *** correspond to p < 0.1, p < 0.05, p < 0.01, respectively. See Table 1 for definition of variables.

documents that there are statistically significant differences in the average values of control variables between firms headquartered in countries with high social capital and firms headquartered in countries with low social capital.6 It is an indication that characteristics of firms headquartered in countries with high social capital and characteristics of firms headquartered in countries with low social capital are significantly different from each other. Therefore, it is important to control for these characteristics. Some of the interesting observations that can be made from this table are regarding the profitability and retained earnings of firms. The table shows that around 7.4 percent of firms headquartered in countries with low social capital generate losses, while this statistic is more than 19.7 percent for firms headquartered in countries with high social capital. In case of retained earnings, the table reports that retained earnings as a percent of total assets is more than 28 percent for firms headquartered in countries with low social capital and negative for firms headquartered in countries with high social capital. The country-specific characteristics also exhibit significant differences between high- and low-social-capital countries. For example, differences in fiscal stability, investor protection, and labor market flexibility are significant between the two groups.

Table 5. Correlation matrix.

No.	Variables	1	2	3	4	5	6	7	8	9
1	SIZE	1.00
2	LEVERAGE	0.07	1.00
3	LOSS	−0.34	−0.03	1.00
4	GROWTH	0.03	−0.04	−0.01	1.00
5	DIVIDEND	0.11	0.12	−0.26	−0.09	1.00
6	CAPEX	0.05	0.05	−0.05	0.04	0.02	1.00
7	BETA	−0.09	0.03	0.17	−0.01	−0.18	−0.02	1.00
8	ANALYST	0.66	0.02	−0.14	−0.02	0.02	0.07	0.01	1.00
9	INVENTORY	−0.05	−0.09	−0.12	−0.05	−0.01	−0.10	0.04	−0.02	1.00
10	RETAINED	0.34	−0.04	−0.45	−0.02	0.18	0.07	−0.12	0.17	0.14
11	ENFORCEMENT	−0.03	−0.01	0.09	−0.03	−0.03	−0.04	0.01	0.04	−0.08
12	PROTECTION	−0.08	0.01	0.10	−0.01	−0.01	−0.02	0.03	0.02	−0.10
13	STABILITY	0.06	−0.03	0.00	0.03	−0.07	0.02	0.06	0.04	0.05
14	COMPETITIVE	0.12	0.03	0.05	−0.04	−0.13	−0.11	0.06	0.13	−0.02
15	LABOR	0.00	0.00	0.11	−0.01	−0.02	−0.07	0.04	−0.05	−0.08
16	FISCAL	0.00	−0.04	−0.04	0.00	0.13	0.05	−0.06	0.04	0.03
17	GDP	0.12	0.01	0.09	0.04	−0.24	−0.03	0.17	0.07	−0.01
Variables		10	11	12	13	14	15	16	17
RETAINED		1.00
ENFORCEMENT		−0.08	1.00
PROTECTION		−0.10	0.77	1.00
STABILITY		−0.01	−0.36	−0.35	1.00
COMPETITIVE		−0.03	0.63	0.57	−0.09	1.00
LABOR		−0.10	0.65	0.54	−0.11	0.64	1.00
FISCAL		0.07	−0.09	−0.14	0.38	−0.16	0.11	1.00
GDP		−0.13	0.09	0.05	0.18	0.42	0.17	−0.46	1.00

Note: The table documents the correlation between variables used in this paper. The sample period is from 2007 to 2023. See Table 1 for definition of variables.

documents correlation between variables used in this analysis. The table indicates low correlation between variables. Therefore, it is possible to include all of them together in a single regression equation.

4. Results and Discussion

4.1. Baseline Analysis

Table 6. Social capital and corporate social responsibility.

Variables	Model (1)	Model (2)	Model (3)	Model (4)
SOCIAL	0.1555 ***	0.2579 ***	0.2009 ***	0.2862 ***
	(18.6595)	(33.6757)	(17.4435)	(27.8735)
SIZE		3.9113 ***		5.0481 ***
		(49.2253)		(62.6724)
LEVERAGE		0.0819 ***		0.0653 ***
		(17.2696)		(14.4151)
LOSS		1.1611 ***		1.5384 ***
		(4.3782)		(5.9981)
GROWTH		−0.049 ***		−0.0463 ***
		(−20.7306)		(−20.3227)
DIVIDEND		1.143 ***		0.8503 ***
		(27.3941)		(21.0804)
CAPEX		−23.5851 ***		−14.8186 ***
		(−12.9252)		(−8.3494)
BETA		−1.2524 ***		0.2916 **
		(−8.3069)		(1.9796)
ANALYST		0.6927 ***		0.6082 ***
		(48.5584)		(43.7598)
INVENTORY		2.2966 ***		5.1185 ***
		(2.7908)		(6.5575)
RETAINED		0.5202 ***		−0.4078 ***
		(4.9995)		(−3.8948)
ENFORCEMENT			0.0411 **	0.0299 **
			(2.5657)	(2.152)
PROTECTION			−0.0345 **	0.0873 ***
			(−2.0213)	(5.947)
STABILITY			−0.3531 ***	−0.3483 ***
			(−16.6296)	(−18.2483)
COMPETITIVE			0.5669 ***	0.2698 ***
			(36.5477)	(19.4922)
LABOR			−0.491 ***	−0.3528 ***
			(−31.9266)	(−25.4148)
FISCAL			−0.0613 ***	−0.1587 ***
			(−4.0947)	(−11.7968)
GDP			−3.1825 ***	−3.9168 ***
			(−30.7068)	(−42.7099)
Industry Dummies	Yes	Yes	Yes	Yes
Year Dummies	Yes	Yes	Yes	Yes
Observations	61,521	61,521	61,521	61,521
R-Square	0.0193	0.2611	0.0838	0.3349

Note: This table documents the relationship between social capital and corporate social responsibility. The t-values are presented in parentheses. The outcome variable is CSR (corporate social responsibility) and the key independent variable is SOCIAL (social capital). The sample period is from 2007 to 2023. The pooled ordinary least squares regression with heteroscedasticity-robust standard errors is used. The symbols **, *** correspond to p < 0.1, p < 0.05, p < 0.01, respectively. See Table 1 for definition of variables.

documents the relationship between social capital and corporate social responsibility. The main variable of interest in this table is SOCIAL. As expected, we report significantly positive coefficient estimate of SOCIAL for all models. It indicates that firms headquartered in countries with high social capital score higher on CSR than firms headquartered in countries with low social capital. Consistent with Hypothesis 1, this result suggests that social capital plays a systematic role in shaping firms’ engagement in socially responsible activities. These findings are consistent with Jha and Cox (2015) who also document high CSR among firms headquartered in regions with high social capital in the United States. We argue that CSR is considered as an act that is consistent with the prescribed values associated with social norms. Therefore, it is relatively hard for firms headquartered in countries with high social capital to deviate from these norms by not engaging in socially responsible practices. Consequently, we observe higher levels of CSR in these countries relative to countries with low social capital. These findings are consistent with the shared-norms and social-network mechanisms discussed in Section 2.1 and Section 2.2, whereby high social capital creates informal societal expectations and reputational pressures that constrain managerial discretion and incentivize firms to internalize socially responsible behavior (Vroom, 1966; Akerlof, 2007; Hilary & Hui, 2009).

Conceptually, this result supports the view that CSR partly reflects informal governance: where social capital is high, societal norms and reputational mechanisms narrow managerial discretion and shift CSR toward an expected standard of conduct. This helps clarify why CSR engagement can arise even when formal enforcement is not the sole driver of responsible behavior.

4.2. Sample Reconstruction

Table 2 shows that almost 50 percent of observations belong to the United States, Japan, China, and the United Kingdom. For instance, out of 61,521 observations, 17,314 (28.14 percent) belong to the United States, 5353 (8.70 percent) belong to Japan, 5185 (8.42 percent) belong to China, and 3896 (6.33 percent) belong to the United Kingdom. Such a high density of observations from four countries has the potential to drive the main findings. These concerns are tackled by excluding these four countries one-by-one in separate regressions and also by excluding all of them at the same time.

Table 7. Social capital and corporate social responsibility: Sample reconstruction.

Variables	Excluding USA	Excluding Japan	Excluding China	Excluding United Kingdom	Excluding All
SOCIAL	0.3408 ***	0.3334 ***	0.2945 ***	0.2827 ***	0.3856 ***
	(29.134)	(18.9922)	(28.1066)	(27.3738)	(20.3832)
SIZE	4.5974 ***	4.919 ***	5.542 ***	4.9621 ***	5.0495 ***
	(45.6835)	(59.6676)	(68.6426)	(58.8778)	(44.3171)
LEVERAGE	0.0991 ***	0.0632 ***	0.0641 ***	0.0576 ***	0.0897 ***
	(16.809)	(13.5331)	(13.668)	(12.2795)	(12.5548)
LOSS	1.9569 ***	1.3643 ***	1.2815 ***	1.4141 ***	1.0907 ***
	(5.834)	(5.1883)	(4.9307)	(5.2683)	(2.7622)
GROWTH	−0.0516 ***	−0.046 ***	−0.0463 ***	−0.0452 ***	−0.0519 ***
	(−15.6094)	(−20.1481)	(−19.9337)	(−19.3788)	(−14.0183)
DIVIDEND	0.7877 ***	0.7532 ***	0.9009 ***	0.861 ***	0.707 ***
	(16.9323)	(18.4161)	(21.592)	(20.3893)	(13.3722)
CAPEX	−8.9344 ***	−17.6568 ***	−14.1296 ***	−14.8792 ***	−11.9311 ***
	(−4.0972)	(−9.7722)	(−7.8037)	(−8.144)	(−4.8737)
BETA	0.2852	−0.1589	0.3742 **	0.2463	−0.4376 *
	(1.3942)	(−1.0628)	(2.4972)	(1.6068)	(−1.8229)
ANALYST	0.7333 ***	0.5901 ***	0.5683 ***	0.6054 ***	0.6315 ***
	(44.4614)	(41.4048)	(37.94)	(42.5444)	(31.5472)
INVENTORY	6.5421 ***	2.4423 ***	6.7987 ***	5.63 ***	6.8901 ***
	(6.913)	(3.0665)	(8.0124)	(6.8614)	(5.6776)
RETAINED	−0.1465	−0.3166 ***	−0.5815 ***	−0.3863 ***	0.1279
	(−0.7913)	(−3.0079)	(−5.4896)	(−3.6083)	(0.6259)
ENFORCEMENT	−0.0073	0.0077	0.0085	0.0371 ***	−0.0904 ***
	(−0.5100)	(0.4895)	(0.5997)	(2.6559)	(−5.4001)
PROTECTION	0.1028 ***	0.0658 ***	0.0274 *	0.0696 ***	−0.0099
	(6.9187)	(4.4201)	(1.7426)	(4.6753)	(−0.6004)
STABILITY	−0.3881 ***	−0.3488 ***	−0.2520 ***	−0.3506 ***	−0.2265 ***
	(−19.1976)	(−18.2155)	(−11.3896)	(−18.2812)	(−9.4998)
COMPETITIVE	0.2896 ***	0.3065 ***	0.2793 ***	0.2672 ***	0.3544 ***
	(20.016)	(21.0731)	(20.0726)	(19.1556)	(22.5003)
LABOR	−0.3356 ***	−0.396 ***	−0.3510 ***	−0.3453 ***	−0.2700 ***
	(−23.4362)	(−24.3504)	(−24.846)	(−24.6536)	(−14.8649)
FISCAL	−0.1803 ***	−0.1519 ***	−0.1328 ***	−0.1576 ***	−0.1943 ***
	(−13.0812)	(−10.5238)	(−9.573)	(−11.6395)	(−12.2019)
REGULATION	−3.0978 ***	−3.9225 ***	−3.5741 ***	−3.8901 ***	−1.0869 ***
	(−27.0071)	(−42.3123)	(−36.6948)	(−42.1765)	(−7.3167)
Industry Dummies	Yes	Yes	Yes	Yes	Yes
Year Dummies	Yes	Yes	Yes	Yes	Yes
Observations	44,207	56,168	56,336	57,625	29,773
R-Square	0.3379	0.3377	0.339	0.3287	0.3394

Note: This table documents the relationship between social capital and corporate social responsibility after removing some of the most represented countries from analysis. The t-values are presented in parentheses. The outcome variable is CSR (corporate social responsibility) and the key independent variable is SOCIAL (social capital). The sample period is from 2007 to 2023. The pooled ordinary least squares regression with heteroscedasticity-robust standard errors is used. The symbols *, **, *** correspond to p < 0.1, p < 0.05, p < 0.01, respectively. See Table 1 for definition of variables.

reports the results of this analysis. As was the case before, we show that firms headquartered in countries with high social capital score high on CSR. Our findings show that coefficient estimate of SOCIAL remains positive and significant in all estimations. Importantly, the persistence of the results after excluding dominant countries suggests that the documented relationship is not driven by country-specific institutional peculiarities but instead reflects a general mechanism through which social capital operates as an informal institution across diverse national contexts. This reinforces our theoretical argument that shared norms and dense social networks exert systematic influence on corporate behavior beyond formal regulatory environments.

This pattern is consistent with our theoretical argument that the effect is tied to slow-moving informal institutions rather than idiosyncratic features of a small set of countries, strengthening the claim that social capital operates as a general governance mechanism influencing CSR.

4.3. Alternate Estimation Strategies

In this section, we assess the robustness of Equation (1) using a range of alternative estimation strategies. The results from these approaches are reported in

Table 8. Social capital and corporate social responsibility: Alternate estimation procedures.

Variables	Model (1)	Model (2)	Model (3)	Model (4)	Model (5)	Model (6)	Model (7)	Model (8)
SOCIAL	0.1795 ***	0.1575 ***	0.0212 ***	0.1667 ***	2.9690 ***	0.2815 ***	0.2941 ***	0.2068 ***
	(5.0386)	(6.4831)	(4.9636)	(6.2665)	(7.5918)	(4.4784)	(6.0641)	(10.3525)
CSR (Lagged)	No	No	Yes	No	No	No	No	No
Firm Controls	Yes	Yes	Yes	Yes	Yes	Yes	No	Yes
Country Controls	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Within Effects (Firm)	No	No	No	No	No	No	Yes	No
Between Effects (Firm)	No	No	No	No	No	No	Yes	No
Industry Dummies	No	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Year Dummies	Yes	Yes	Yes	Yes	Yes	No	Yes	Yes
Observations	61,521	61,521	49,294	61,521	35,172	61,521	61,521	56,266
R-Squared	0.4014		0.8865		0.3682	0.3681	0.3471	0.3448
Partial R-Square								0.2706
Eigen Value								20,859.1

Note: This table documents the relationship between social capital and corporate social responsibility. The t-values and z-values are reported in parentheses. The outcome variable is CSR (corporate social responsibility) and the key independent variable is SOCIAL (social capital). Model (1) is based on panel regression with fixed effects (robust standard errors). Model (2) is based on panel regression with random effects (robust standard errors). Model (3) is based on dynamic pooled OLS regression with robust standard errors. Model (4) is based on population-averaged panel-data regression. Model (5) is based on propensity score matching. Model (6) is based on Fama–MacBeth regression. Model (7) is based on hybrid regression. Model (8) is based on instrument variable regression. The sample period is from 2007 to 2023. The symbols *** correspond to p < 0.1, p < 0.05, p < 0.01, respectively. See Table 1 for definition of variables.

. Model (1) and Model (2) employ panel regressions with fixed effects and random effects, respectively. Model (3) relies on a dynamic pooled OLS specification with robust standard errors, incorporating lagged values of the dependent variable to capture persistence in firm behavior over time. The inclusion of lagged outcomes also helps mitigate concerns related to autocorrelation and omitted variable bias. Model (4) applies a population-averaged panel-data regression framework, which estimates effects at the population level rather than for individual firms, while appropriately accounting for within-panel correlation and unobserved heterogeneity in longitudinal data. Model (6) implements the Fama–MacBeth two-step procedure, which estimates cross-sectional regressions in each period and averages the coefficients over time. This approach is particularly well-suited to addressing cross-sectional dependence. Model (7) employs a hybrid (within–between) specification to account for potential heterogeneity across firms and countries. This model decomposes time-varying regressors into within-unit deviations and between-unit means, allowing the inclusion of slow-moving variables such as social capital while controlling for firm-level unobserved heterogeneity. In this framework, within effects capture how changes over time influence CSR, whereas between effects reflect structural differences across firms or countries. To further address selection and endogeneity concerns, Model (5) adopts a propensity score matching approach, pairing firms with similar observable characteristics to better isolate the effect of social capital on CSR. By reducing selection bias and addressing observable heterogeneity, this method enhances the comparability between treated and control firms and strengthens causal inference. Finally, Model (8) uses an instrumental variables approach, with ethnic fractionalization serving as the instrument. Post-estimation diagnostics indicate that the instrument is strong, as evidenced by eigenvalue statistics exceeding conventional thresholds and a partial R-squared greater than 0.27. The consistency of the results across diverse estimation strategies strengthens the interpretation that the observed effects are not statistical artifacts but reflect a stable behavioral mechanism. In line with institutional and social capital theories, the results suggest that CSR decisions are persistently shaped by slow-moving social norms and trust structures rather than short-term firm-specific shocks.

5. Additional Tests

5.1. Various Aspects of Social Capital and CSR

Social capital encompasses distinct dimensions, such as the following: (a) civic and social participation, (b) social tolerance, (c) interpersonal trust, (d) personal and family relationships, and (e) social networks. These dimensions influence CSR through complementary channels. For example, personal relationships and interpersonal trust generate normative pressure that encourages firms to conform to societal expectations, while dense social networks facilitate monitoring, information diffusion, and reputational enforcement. If CSR is shaped by both normative expectations and social enforcement mechanisms, then multiple dimensions of social capital should independently contribute to higher CSR engagement. In this section, we identify the channels via which social capital exerts its impact on CSR. We believe that this analysis is important because it can help us recognize the mechanisms through which social capital works in improving social responsibility. It is possible that only some of these components affect CSR. In order to test this conjecture, we replace SOCIAL with its components in Equation (1) and re-estimate it. The findings are reported in

Table 9. Different aspects of social capital and corporate social responsibility.

Variables	Model (1)	Model (2)	Model (3)	Model (4)	Model (5)
CIVIC	0.1485 ***
	(18.6058)
TRUST		0.0613 ***
		(7.0682)
RELATIONSHIP			0.3809 ***
			(37.1807)
NETWORKS				0.2114 ***
				(27.9376)
TOLERANCE					0.1761 ***
					(24.4216)
SIZE	5.0098 ***	4.858 ***	5.0388 ***	5.0572 ***	4.9157 ***
	(61.5203)	(60.8001)	(62.843)	(63.0768)	(61.6568)
LEVERAGE	0.0710 ***	0.0739 ***	0.0588 ***	0.0650 ***	0.0667 ***
	(15.5985)	(16.2264)	(13.0377)	(14.3439)	(14.7239)
LOSS	1.6918 ***	1.6028 ***	1.4272 ***	1.5572 ***	1.4584 ***
	(6.5791)	(6.2368)	(5.5747)	(6.0826)	(5.6843)
GROWTH	−0.0445 ***	−0.0436 ***	−0.0466 ***	−0.0457 ***	−0.046 ***
	(−19.6332)	(−19.2496)	(−20.5155)	(−20.0764)	(−20.1712)
DIVIDEND	0.8904 ***	0.8604 ***	0.8067 ***	0.8292 ***	0.8614 ***
	(21.9874)	(21.2552)	(20.1579)	(20.5437)	(21.3584)
CAPEX	−15.1794 ***	−15.5172 ***	−14.9284 ***	−14.3009 ***	−14.6924 ***
	(−8.5599)	(−8.7608)	(−8.4664)	(−8.0572)	(−8.2768)
BETA	0.3248 **	0.4810 ***	0.3514 **	0.2780 *	0.3746 **
	(2.1961)	(3.2552)	(2.3884)	(1.8911)	(2.5423)
ANALYST	0.6079 ***	0.6357 ***	0.615 ***	0.6013 ***	0.6224 ***
	(43.3493)	(45.6179)	(44.5671)	(43.2081)	(44.9121)
INVENTORY	4.7247 ***	3.9745 ***	5.4251 ***	4.9544 ***	5.0326 ***
	(6.0347)	(5.0902)	(6.9696)	(6.353)	(6.4229)
RETAINED	−0.5375 ***	−0.5618 ***	−0.3073 ***	−0.4194 ***	−0.3902 ***
	(−5.1285)	(−5.3712)	(−2.9205)	(−4.0112)	(−3.7223)
ENFORCEMENT	0.0589 ***	0.0699 ***	0.1200 ***	0.1274 ***	−0.0253 *
	(4.2899)	(4.9274)	(8.874)	(9.285)	(−1.7034)
PROTECTION	0.0703 ***	0.2193 ***	0.1719 ***	0.1602 ***	0.109 ***
	(4.3448)	(15.8766)	(12.8007)	(11.5397)	(7.4304)
STABILITY	−0.382 ***	−0.4718 ***	−0.3239 ***	−0.384 ***	−0.3494 ***
	(−19.8984)	(−25.5755)	(−17.2272)	(−20.3299)	(−18.1717)
COMPETITIVE	0.2279 ***	0.192 ***	0.2221 ***	0.2271 ***	0.2604 ***
	(16.4127)	(13.7025)	(16.6865)	(16.7537)	(18.549)
LABOR	−0.3104 ***	−0.3501 ***	−0.5287 ***	−0.3393 ***	−0.3066 ***
	(−21.5932)	(−24.9928)	(−37.5868)	(−24.4221)	(−21.783)
FISCAL	−0.0739 ***	−0.085 ***	−0.069 ***	−0.1377 ***	−0.1186 ***
	(−5.8841)	(−5.7207)	(−5.7346)	(−10.4931)	(−9.2327)
GDP	−3.2866 ***	−3.36 ***	−3.1847 ***	−3.6159 ***	−3.7664 ***
	(−38.3775)	(−33.5142)	(−38.5943)	(−41.2649)	(−41.4385)
Industry Dummies	Yes	Yes	Yes	Yes	Yes
Year Dummies	Yes	Yes	Yes	Yes	Yes
Observations	61,521	61,521	61,521	61,521	61,521
R-Square	0.3292	0.3252	0.3418	0.3347	0.3325

Note: This table documents the relationship between different aspects of social capital and corporate social responsibility. The t-values are presented in parentheses. The outcome variable is CSR (corporate social responsibility) and the key independent variables are different aspects of social capital (CIVIC, TRUST, RELATIONSHIP, NETWORKS, and TOLERANCE). The sample period is from 2007 to 2023. The pooled ordinary least squares regression with heteroscedasticity-robust standard errors is used. The symbols *, **, *** correspond to p < 0.1, p < 0.05, p < 0.01, respectively. See Table 1 for definition of variables.

. The findings suggest that all aspects of social capital positively affect CSR. The positive and statistically significant effects observed across the different dimensions of social capital provide further insight into the mechanisms discussed in Section 2. Consistent with the shared-norms channel (Holland, 1976; Tom, 1971; Putnam, 1995; Fukuyama, 1997; Arthur & Joshua, 2005; Graafland & Van de Ven, 2006), components such as civic and social participation and social tolerance reflect societal expectations that promote cooperative and ethically responsible behavior. At the same time, dimensions related to interpersonal trust and social networks capture the monitoring and reputational enforcement mechanisms emphasized by Coleman (1990), Portes (1998), and Putnam (2000), whereby dense networks facilitate the detection and sanctioning of deviant behavior. Together, these findings support our theoretical argument that social capital influences CSR through both normative pressure and social enforcement, rather than through a single isolated channel.

5.2. Social Capital and Various Aspects of CSR

This section documents the impact of social capital on various aspects of CSR. Our variable representing CSR comprises four components. The first component (workforce) measures a firm’s effectiveness in terms of providing job satisfaction, a healthy and safe workplace, maintaining diversity and equal opportunities, and providing development opportunities for its workforce. The second component (human rights) measures a firm’s effectiveness in terms of respecting fundamental human rights conventions. The third component (community) measures a firm’s commitment to being a good citizen, protecting public health and respecting business ethics. The last component (product responsibility) reflects a firm’s capacity to produce quality goods and services, integrating the customer’s health and safety, integrity and data privacy. We re-estimate Equation (1) by replacing CSR with its various components one-by-one. The findings are reported in

Table 10. Social capital and different aspects of corporate social responsibility.

Variables	Human Rights	Community	Workforce	Product Responsibility
SOCIAL	0.3462 ***	0.6543 ***	0.1299 ***	−0.1221 ***
	(25.3881)	(49.8603)	(10.2799)	(−8.4654)
SIZE	5.5913 ***	5.2648 ***	5.2179 ***	3.6145 ***
	(50.755)	(53.2407)	(55.324)	(33.0959)
LEVERAGE	0.0926 ***	0.0814 ***	0.0493 ***	0.0688 ***
	(14.3684)	(14.1942)	(8.5346)	(10.646)
LOSS	0.7509 **	2.4574 ***	1.0774 ***	0.5239
	(2.0464)	(7.5466)	(3.3435)	(1.4198)
GROWTH	−0.0606 ***	−0.0514 ***	−0.0338 ***	−0.0476 ***
	(−18.8743)	(−17.9191)	(−11.6997)	(−15.2501)
DIVIDEND	0.7582 ***	0.4935 ***	1.133 ***	0.9313 ***
	(13.4403)	(9.4684)	(22.6929)	(16.4199)
CAPEX	−13.9717 ***	−11.6463 ***	−6.7761 ***	−21.3371 ***
	(−5.7022)	(−5.0834)	(−3.0161)	(−8.6054)
BETA	0.2718	2.4214 ***	−0.9292 ***	−0.6047 ***
	(1.3212)	(12.9925)	(−5.0238)	(−2.8664)
ANALYST	0.6258 ***	0.5579 ***	0.7272 ***	0.4856 ***
	(32.0713)	(32.9366)	(43.1702)	(24.871)
INVENTORY	20.9682 ***	2.2755 **	3.4735 ***	9.3846 ***
	(18.8541)	(2.2775)	(3.5209)	(7.9834)
RETAINED	1.1309 ***	−0.4719 ***	−0.0206	0.0853
	(8.497)	(−3.6822)	(−0.1489)	(0.5634)
ENFORCEMENT	0.0661 ***	−0.2164 ***	0.1645 ***	0.1328 ***
	(3.3995)	(−12.1194)	(9.784)	(6.7089)
PROTECTION	−0.019	0.4782 ***	−0.0114	−0.1346 ***
	(−0.9322)	(25.692)	(−0.628)	(−6.5423)
STABILITY	−0.5193 ***	−0.3917 ***	−0.1514 ***	−0.3486 ***
	(−20.0664)	(−16.6441)	(−6.353)	(−13.3903)
COMPETITIVE	0.5086 ***	−0.0816 ***	0.1795 ***	0.4221 ***
	(26.0234)	(−4.5833)	(10.1555)	(21.6991)
LABOR	−0.4548 ***	−0.3874 ***	−0.3538 ***	−0.1512 ***
	(−24.5855)	(−21.8274)	(−19.5603)	(−7.6499)
FISCAL	−0.1213 ***	−0.2512 ***	−0.0667 ***	−0.2098 ***
	(−6.7728)	(−15.3509)	(−3.9721)	(−11.6081)
GDP	−5.8764 ***	−0.8403 ***	−5.189 ***	−3.6539 ***
	(−45.1255)	(−7.3822)	(−45.5355)	(−28.5797)
Industry Dummies	Yes	Yes	Yes	Yes
Year Dummies	Yes	Yes	Yes	Yes
Observations	61,521	61,521	61,521	61,521
R-Square	0.2908	0.2622	0.2776	0.1574

Note: This table documents the relationship between social capital and different aspects of corporate social responsibility. The t-values based on the heteroscedasticity-robust standard errors are presented in parentheses. The outcome variable is CSR (different aspects of corporate social responsibility) and the key independent variable is SOCIAL (social capital). The sample period is from 2010 to 2021. The pooled ordinary least squares regression is used. The symbols **, *** correspond to p < 0.1, p < 0.05, p < 0.01, respectively. See Table 1 for definition of variables.

. The results demonstrate that social capital positively affects all aspects of CSR, except product responsibility. The heterogeneous effects across CSR dimensions are consistent with the theoretical framework developed in Section 2. CSR dimensions such as workforce, human rights, and community are closely tied to societal norms, ethical expectations, and stakeholder scrutiny, making them particularly sensitive to the shared-norms and social-enforcement mechanisms associated with social capital (Putnam, 1995; Fukuyama, 1997; Coleman, 1990). In contrast, product responsibility is more strongly shaped by formal regulation, industry standards, and technological requirements, which limit the role of informal institutions such as social capital in influencing firm behavior. This distinction reinforces the view that social capital primarily affects CSR activities that are socially observable and normatively enforced rather than those governed predominantly by formal institutional constraints.

5.3. Social Capital and Value of CSR

The theoretical arguments developed in Section 2 suggest that the value relevance of CSR depends critically on the surrounding social environment. In countries characterized by high social capital, strong shared norms and dense social networks constrain managerial behavior and make socially responsible conduct a baseline expectation rather than a strategic choice (Putnam, 1995; Fukuyama, 1997; Coleman, 1990). In contrast, in low-social-capital environments, where informal enforcement mechanisms are weaker and stakeholder trust is lower, CSR can serve as a credible signal of firm quality and commitment to collective interests. In order to test this conjecture, we estimate the following pooled OLS regression. In the following equation, PERFORM can be represented by Tobin’s Q, stock price volatility, sales growth, and financial distress risk. See Table 1 for definition of these variables.

$$\begin{array}{cc}\mathrm{P}\mathrm{E}\mathrm{R}\mathrm{F}\mathrm{O}\mathrm{R}\mathrm{M}& =\mathsf{\alpha }+{\mathsf{\beta }}_1\left(\mathrm{S}\mathrm{O}\mathrm{C}\mathrm{I}\mathrm{A}\mathrm{L}\right)+{\mathsf{\beta }}_2\left(\mathrm{C}\mathrm{S}\mathrm{R}\right)+{\mathsf{\beta }}_3\left(\mathrm{S}\mathrm{O}\mathrm{C}\mathrm{I}\mathrm{A}\mathrm{L}\mathrm{*}\mathrm{C}\mathrm{S}\mathrm{R}\right)+{\mathsf{\beta }}_4\left(\mathrm{S}\mathrm{I}\mathrm{Z}\mathrm{E}\right)+{\mathsf{\beta }}_5\left(\mathrm{L}\mathrm{E}\mathrm{V}\mathrm{E}\mathrm{R}\mathrm{A}\mathrm{G}\mathrm{E}\right)+{\mathsf{\beta }}_6\left(\mathrm{L}\mathrm{O}\mathrm{S}\mathrm{S}\right)\hfill \\ & +{\mathsf{\beta }}_7\left(\mathrm{G}\mathrm{R}\mathrm{O}\mathrm{W}\mathrm{T}\mathrm{H}\right)+{\mathsf{\beta }}_8\left(\mathrm{D}\mathrm{I}\mathrm{V}\mathrm{I}\mathrm{D}\mathrm{E}\mathrm{N}\mathrm{D}\right)+{\mathsf{\beta }}_9\left(\mathrm{C}\mathrm{A}\mathrm{P}\mathrm{E}\mathrm{X}\right)+{\mathsf{\beta }}_{10}\left(\mathrm{B}\mathrm{E}\mathrm{T}\mathrm{A}\right)+{\mathsf{\beta }}_{11}\left(\mathrm{A}\mathrm{N}\mathrm{A}\mathrm{L}\mathrm{Y}\mathrm{S}\mathrm{T}\right)\hfill \\ & +{\mathsf{\beta }}_{12}\left(\mathrm{I}\mathrm{N}\mathrm{V}\mathrm{E}\mathrm{N}\mathrm{T}\mathrm{O}\mathrm{R}\mathrm{Y}\right)+{\mathsf{\beta }}_{13}\left(\mathrm{R}\mathrm{E}\mathrm{T}\mathrm{A}\mathrm{I}\mathrm{N}\mathrm{E}\mathrm{D}\right)+{\mathsf{\beta }}_{14}\left(\mathrm{E}\mathrm{N}\mathrm{F}\mathrm{O}\mathrm{R}\mathrm{C}\mathrm{E}\mathrm{M}\mathrm{E}\mathrm{N}\mathrm{T}\right)+{\mathsf{\beta }}_{15}\left(\mathrm{P}\mathrm{R}\mathrm{O}\mathrm{T}\mathrm{E}\mathrm{C}\mathrm{T}\mathrm{I}\mathrm{O}\mathrm{N}\right)\hfill \\ & +{\mathsf{\beta }}_{16}\left(\mathrm{S}\mathrm{T}\mathrm{A}\mathrm{B}\mathrm{I}\mathrm{L}\mathrm{I}\mathrm{T}\mathrm{Y}\right)+{\mathsf{\beta }}_{17}\left(\mathrm{C}\mathrm{O}\mathrm{M}\mathrm{P}\mathrm{E}\mathrm{T}\mathrm{I}\mathrm{T}\mathrm{I}\mathrm{V}\mathrm{E}\right)+{\mathsf{\beta }}_{18}\left(\mathrm{L}\mathrm{A}\mathrm{B}\mathrm{O}\mathrm{R}\right)+{\mathsf{\beta }}_{19}\left(\mathrm{F}\mathrm{I}\mathrm{S}\mathrm{C}\mathrm{A}\mathrm{L}\right)+{\mathsf{\beta }}_{20}\left(\mathrm{G}\mathrm{D}\mathrm{P}\right)\hfill \\ & +{\displaystyle \sum _{\mathrm{Y}=1}^{\mathrm{N}-1}}{\mathsf{\gamma }}_{\mathrm{Y}}\left(\mathrm{Y}\mathrm{D}\mathrm{U}\mathrm{M}\right)+{\displaystyle \sum _{\mathrm{I}=1}^{\mathrm{N}-1}}{\mathsf{\phi }}_{\mathrm{I}}\left(\mathrm{I}\mathrm{D}\mathrm{U}\mathrm{M}\right)+\mathsf{\epsilon }\hfill \end{array}$$

(2)

The findings are reported in

Table 11. Social capital and the value of corporate social responsibility.

Variables	Tobin’s Q	Financial Distress Risk	Stock Return Volatility	Sales Growth
SOCIAL	−0.0009	0.0099 ***	−0.0007 ***	0.0398
	(−0.6365)	(3.5493)	(−6.5964)	(1.4727)
CSR	0.0294 ***	0.0481 ***	−0.0011 ***	0.1628 ***
	(23.6342)	(19.0068)	(−10.3252)	(6.6206)
SOCIAL*CSR	−0.0001 ***	−0.0005 ***	0.0001 ***	−0.0019 ***
	(−6.341)	(−10.3191)	(6.0983)	(−4.3503)
SIZE	0.3667 ***	0.4045 ***	−0.0336 ***	1.1149 ***
	(58.0213)	(31.8834)	(−53.4572)	(8.08)
LEVERAGE	−0.0196 ***	−0.0885 ***	−0.0005 ***	−0.0398 ***
	(−50.9144)	(−98.4881)	(−14.3821)	(−4.4451)
LOSS	0.031	−0.5377 ***	0.0612 ***	−6.0522 ***
	(1.5485)	(−12.3161)	(27.0382)	(−10.0081)
GROWTH	0.0052 ***	0.0076 ***	0.0005 ***	0.3469 ***
	(18.5952)	(12.8891)	(18.1153)	(30.7359)
DIVIDEND	−0.0837 ***	−0.1368 ***	−0.0093 ***	−0.3349 ***
	(−37.1094)	(−26.4752)	(−34.9971)	(−4.9881)
CAPEX	1.7737 ***	3.1091 ***	0.0826 ***	28.0578 ***
	(12.4762)	(9.8572)	(5.7886)	(7.2629)
BETA	−0.1021 ***	−0.1772 ***	0.1132 ***	1.5051 ***
	(−8.0574)	(−6.7047)	(66.6303)	(4.3286)
ANALYST	−0.0055 ***	−0.0091 ***	0.0007 ***	−0.0304
	(−5.0363)	(−4.1465)	(8.6015)	(−1.5866)
INVENTORY	0.1368 **	1.0551 ***	0.003	−0.481
	(2.5092)	(8.8301)	(0.5892)	(−0.4087)
RETAINED	−0.2135 ***	1.339 ***	−0.053 ***	−2.7405 ***
	(−15.8346)	(27.864)	(−32.6949)	(−6.5568)
ENFORCEMENT	−0.0061 ***	−0.0121 ***	0.0008 ***	−0.0083
	(−6.4954)	(−6.1981)	(9.3731)	(−0.4035)
PROTECTION	0.0071 ***	0.0174 ***	−0.002 ***	0.0033
	(7.8652)	(8.8321)	(−21.2173)	(0.1652)
STABILITY	0.0006	0.0027	0.0006 ***	−0.0607 **
	(0.5205)	(1.08)	(5.0266)	(−2.1918)
COMPETITIVE	0.0005	−0.0051 **	−0.0002 **	−0.0382 *
	(0.5144)	(−2.3558)	(−1.9616)	(−1.708)
LABOR	−0.0281 ***	−0.0413 ***	−0.0001	−0.1559 ***
	(−25.2835)	(−17.2896)	(−1.3164)	(−7.4883)
FISCAL	0.0041 ***	0.0076 ***	0.0009 ***	0.0566 ***
	(5.35)	(4.7161)	(11.7468)	(3.1195)
REGULATION	−0.0079	0.0352 ***	0.0088 ***	0.0008
	(−1.3502)	(2.8362)	(15.8881)	(0.0061)
Industry Dummies	Yes	Yes	Yes	Yes
Year Dummies	Yes	Yes	Yes	Yes
Observations	60,037	55,077	54,367	60,707
R-Square	0.2785	0.3789	0.581	0.1625

Note: This table documents the relationship between social capital and the value of corporate social responsibility. The t-values are presented in parentheses. The outcome variable is PERFORM (Tobin’s Q, stock return volatility, sales growth, and financial distress risk) and the key independent variable is the interaction between SOCIAL (social capital) and CSR (corporate social responsibility). The sample period is from 2007 to 2023. The pooled ordinary least squares regression is used. The symbols *, **, *** correspond to p < 0.1, p < 0.05, p < 0.01, respectively. See Table 1 for definition of variables.

. The main variable of interest in this table is SOCIAL*CSR. We show that SOCIAL*CSR is significant and negative when financial distress risk, sales growth or Tobin’s Q is used as a measure of performance. It indicates that the positive impact of CSR on performance is less pronounced in countries with high social capital. The table also shows that the coefficient estimate of this variable is significant and positive when stock return volatility is used as a measure of performance. It also reveals that the ability of CSR to reduce stock return volatility is less pronounced in countries with high social capital. These findings indicate that CSR is more valuable when social capital is low. These findings provide empirical support for the argument that CSR acts as a substitute governance mechanism in environments with weak informal institutions. Consistent with the literature on social capital and cooperation (Putnam, 1995; Banfield, 1967), CSR investments are more informative and value-enhancing in low-social-capital countries, where they help mitigate information asymmetries, reduce stakeholder skepticism, and enhance firm legitimacy. In high-social-capital countries, however, the marginal impact of CSR on firm value is attenuated, as socially responsible behavior is already widely expected and enforced through social norms and networks.

This moderation result has a direct theoretical implication: social capital does not only increase CSR; it changes CSR’s function. In high-social-capital environments, CSR is closer to a baseline expectation supported by informal enforcement, so its marginal informativeness and value relevance are weaker. In low-social-capital environments, CSR is more distinctive and credibility-enhancing, helping firms mitigate trust deficits and legitimacy concerns, thereby strengthening its value relevance. More broadly, this contingency helps reconcile why CSR–performance relations may vary across institutional settings.

6. Conclusions

This study provides robust evidence that country-level social capital plays a significant role in shaping corporate social responsibility (CSR) practices. Firms headquartered in countries characterized by higher levels of social capital consistently exhibit stronger CSR performance than those operating in low-social-capital environments. These results are robust across multiple estimation strategies and specifications. Overall, the findings support the view that CSR practices are deeply embedded in prevailing social norms and shared values, and that firms operating in high-social-capital societies are more likely to internalize and comply with these normative expectations.

Beyond documenting this relationship, our analysis highlights important heterogeneity in the value relevance of CSR across institutional contexts. While CSR appears to reflect baseline societal expectations in high-social-capital countries, it emerges as a more powerful value-enhancing and signaling mechanism in environments characterized by weaker social capital. In such contexts, CSR appears to play a more informative signaling and legitimacy-enhancing role in environments characterized by weaker social capital. This finding underscores the contingent nature of CSR and suggests that its economic and strategic importance depends critically on the surrounding social environment.

The study carries several important implications. For investors, the results indicate that CSR should be interpreted through a contextual lens. Firms headquartered in high-social-capital countries may exhibit higher CSR due to societal norms rather than superior governance, whereas in low-social-capital environments, strong CSR engagement may signal exceptional managerial quality and commitment to stakeholder interests. As such, CSR metrics may be particularly informative for investment decision-making in countries where social capital is weak. For managers, the findings suggest that CSR strategies should be aligned with the institutional environment. In low-social-capital settings, CSR investments can function as an effective reputational and trust-building tool, enhancing firm credibility and stakeholder relationships. In contrast, in high-social-capital environments, CSR may be necessary to meet minimum societal expectations but less likely to generate differentiation or excess returns. From a policy perspective, the results highlight the importance of informal institutions in shaping responsible corporate behavior, suggesting that policies aimed at fostering social trust and civic engagement may indirectly promote CSR and sustainable business practices. Importantly, while social capital can mitigate certain governance shortcomings through informal social enforcement and normative pressure, it does not replace the role of formal institutions such as investor protection and contract enforcement, which remain essential for sustainable corporate governance and economic development.

This study contributes to the CSR and institutional economics literature by showing that social capital, as an informal institution, not only predicts higher CSR engagement but also shapes the interpretation and economic role of CSR across contexts. By documenting that CSR is more value-relevant where social capital is weaker, the findings support a contingency view in which CSR functions as a stronger credibility-building and substitute governance mechanism when informal discipline is limited.

Finally, we acknowledge a major limitation. Our CSR measure is based on Refinitiv’s social pillar score, which primarily covers larger, more established, and more visible firms with sufficient disclosure and data availability. Consequently, the sample may not fully represent the population of firms within each country, and the results should be interpreted as applying to firms with adequate CSR reporting. This data characteristic may also introduce selection concerns. Future research could extend our analysis by controlling for industry-specific CSR norms, using alternative CSR measures or datasets that cover smaller and privately held firms, and examining additional institutional factors that may interact with social capital to shape CSR outcomes.