CSR Disclosure and the Zero-Leverage Phenomenon: Evidence from Pakistan Listed Firms

Abstract

The effect of corporate social responsibility (CSR) disclosure on zero-leverage policies is examined for listed firms at the Pakistan Stock Exchange (PSX) from 2010 to 2021. Binary logistic regression models show a statistically significant positive relationship between CSR disclosure and zero leverage. Increased CSR disclosure raises the propensity of firms to have zero leverage. Moreover, the negative effect of CSR disclosure on debt ratios further confirms these findings. Results show that highly disclosed CSR firms face less information asymmetry and prefer equity financing over bank debt. Regulators should develop incentive programs to increase their CSR disclosure and strengthen stakeholders’ relationships.

1. Introduction

Firms’ financing is seen as the fuel for effective governance of business. However, given the heterogeneity of firms, they utilize financial resources in diverse ways to meet their respective needs. Capital structure reflects a firm’s financing decisions regarding the use of debt or equity. One of the biggest challenges faced by firms is achieving an optimal balance between financing sources. This decision is a determinant of the upcoming financing resources, the cost of capital, and the firm’s overall risk and valuation, among others. This underscores the significance of defining optimal capital structures within the corporate finance domain. However, the literature on the subject offers no definitive guidance on the optimal capital structure (Drobetz & Fix, 2003).

Following the global recession of 2007–2008, regulatory and managerial governance efforts to achieve sustainable growth gained significant momentum. Firms adopted a more resilient and robust funding strategy, resulting in a puzzling debate over financial protectionism, financial conservatism, debt avoidance, and no debt or low-leveraging strategies (Cui, 2019; Gendron & Smith-Lacroix, 2015). Among researchers and practitioners, the zero-leverage phenomenon is gaining recognition as a fundamental element of capital structure decisions (Strebulaev & Yang, 2013; Zaman et al., 2019). A relevant and increasing number of firms following zero-leverage policies have been identified in the United States, the United Kingdom, India, China, and Pakistan, among many others (e.g., Bessler et al., 2013; Zhang & Gregoriou, 2019; Huang et al., 2017; Ehsan et al., 2023).

Zero leverage has been described by El Ghoul et al. (2018) as a rising trend observable in both developed and developing nations. Since zero leverage has been increasing continuously for many years, the phenomenon has been associated with high market-to-book values, substantial cash reserves, robust profits, tax payments, or even dividend policy (Nguyen et al., 2021). This indicates that these firms are more resistant to unexpected negative events and can maintain sustainability. Furthermore, zero-leverage firms have been found to exhibit a greater degree of social responsibility, as evidenced by increased tax contributions to their economies.

The link between zero leverage and social responsibility ties into the larger discussion of Corporate Social Responsibility (CSR). Recently, decision-makers and researchers have increased their focus on CSR (Wang et al., 2016). The rising importance of CSR for firms comes from greater awareness and attention from communities, educational institutions, and regulatory bodies (Alam & Tariq, 2022). As a result, firms increasingly disclose information on their CSR practices to meet the expectations of different investors. This practice helps build positive feelings in the community and the market. The focus on CSR disclosure underscores its crucial role in sustainable growth and value (Jamali, 2008).

Although the link between a firm’s financing structure and CSR disclosure has been considerably examined, studies about the impact of CSR on zero leverage remain scarce. Studies linking capital structure and CSR disclosure are particularly performed in developed countries; however, key structural differences between developed and developing nations could affect this relationship. Developed countries are usually considered to have high performance in labor, capital and goods and service markets (Khanna & Palepu, 2000), financial stability, and better access to funding through capital markets or debt. This makes it easier for firms to adopt policies like zero debt, since changes in policy can happen more readily. In contrast, developing countries’ markets face several frictions due to agency and information issues, such as lack of information disclosure, weak governance, and underdeveloped financial intermediaries, among others. This leads to increased uncertainty and risk, consequently increasing financing costs. At the same time, capital markets also become less attractive to firms and investors.

The ability to disclose CSR also depends on a country’s level of development. Ali et al. (2017) outline differences in CSR disclosure between countries with different levels of development. In developed countries, various stakeholders, such as shareholders, investors, media, and regulators, play important roles in disclosure. However, in developing nations, international forces have a greater influence on CSR disclosure due to lower local public pressure (Ali et al., 2017).

In Pakistan, firms stand to gain significantly from an increase in their CSR, particularly with respect to enhancing their reputation and brand value, expanding their consumer base, and cultivating stronger relationships with stakeholders. However, Pakistan poses a challenging context for studying the zero-leverage phenomenon. The underdeveloped capital market results in firms relying heavily on bank debt, which hinders the adoption of zero-leverage strategies. Furthermore, the disclosure of CSR is still in its early stages, raising the question of whether a CSR commitment could lead firms to pursue sustainable financing policies. While there are guidelines, such as those from the Securities and Exchange Commission of Pakistan (SECP) in 2013, they are not properly implemented. The lack of a legal framework means that disclosure often depends on voluntary reporting or pressure from stakeholders. This raises the issue of whether a commitment to CSR could encourage firms to adopt sustainable financing policies. Consequently, an investigation into this phenomenon within the context of Pakistan’s economic landscape could offer valuable insights about the evolving role of CSR in firms’ financial decisions.

Evidence shows that Pakistani firms with a zero-leverage policy often encounter financial restrictions. Nonetheless, the literature indicates that some firms adopt zero leverage as a strategy when they are quite profitable (e.g., Ehsan et al., 2023). Recently Shahzad et al. (2025) show that CEO characteristics, board composition, and a firm’s ownership structure are determinants for explaining corporate deleveraging policies in Pakistan. This highlights the varying circumstances surrounding the zero-leverage trend. Recent studies have found a positive connection between CSR and zero leverage. Zaman et al. (2022) note that this link is stronger in firms that pay dividends.

To the best of our knowledge, no study has been conducted on the potential impact of CSR on zero leverage in Pakistan. Therefore, this study addresses the following research question: (1) Does corporate social responsibility impact zero-leverage policies in Pakistan? To answer this question, we use a sample of 141 non-financial firms listed on the Pakistan Stock Exchange (PSX) for 2010–2021, applying linear and logistic regressions to examine the relationship between CSR disclosure and leverage or zero leverage.

This research directly links these two broad areas of corporate finance and CSR. In particular, we show that greater levels of CSR disclosure increase the propensity for firms to become debt-free, potentially because they can obtain more favorable conditions from private investors and, in this way, replace debt with equity financing. Therefore, CSR disclosure is a mechanism to decrease information asymmetries and give investors more confidence to participate in capital markets and invest in listed firms. This research makes executives, financial managers, and financing entities aware of the relevance of CSR to capital structure decisions. Therefore, this research will likely benefit managers, policymakers, regulators, and academics.

The remainder of this paper is structured as follows: Section 2 reviews the literature; Section 3 presents the institutional setting in Pakistan and the status of CSR practices in the country; Section 4 summarizes the methods used; Section 5 is dedicated to the presentation and discussion of results; and Section 6 presents the conclusion of the paper by highlighting the main findings and their implications.

2. Literature Review

CSR disclosure has become a central element of corporate governance, influencing interactions between firms and their stakeholders. Ali et al. (2022) review the literature, documenting the factors that influence CSR disclosure. Additionally, Christensen et al. (2021) offer a broad analysis of the main effects of CSR on capital markets, stakeholders, and enforcement of CSR practices. Overall, CSR disclosure plays a key role in corporate governance and decisions about capital structure through several determinants. Variables like firm size (e.g., Zamir et al., 2021), industry size (e.g., Sekhon & Kathuria, 2019), firm age (e.g., Maama, 2021), board gender diversity (Khan et al., 2019), and financial performance (e.g., Haniffa & Cooke, 2005), among others, seem to drive CSR disclosure. According to Ali et al. (2022), most studies reviewed suggest that these variables tend to positively impact CSR disclosure. However, some variables, such as leverage, still have an unclear effect.

Corporate governance and capital structure decisions heavily depend on CSR disclosure. From the viewpoint of agency theory, CSR serves as a tool to align the interests of management and shareholders. This alignment helps firms improve their image and influences their choices for funding and investments (C. Deegan & Blomquist, 2006). Empirical evidence shows that CSR disclosure matters for capital structure and lending activities (Carroll, 1979; Spector, 2008). Consistent with this research, Pijourlet (2013) finds that firms with higher CSR issue more equity and have lower leverage compared to those with lower CSR. Benlemlih (2017) states that CSR disclosure negatively affects long-term loans of American firms by reducing the maturity of these loans. S. Sheikh (2019) finds similar results, arguing that CSR reporting has a negative effect on firms’ leverage in highly competitive product markets. Ye and Zhang (2011) support this view, demonstrating that firms with greater CSR concerns have notably lower debt costs. Moreover, CSR disclosure can greatly enhance the quality of firms’ earnings (Fan et al., 2023).

Recent studies from emerging markets show that mandatory CSR disclosure can raise firms’ financing constraints and lower their willingness to take on risky investments, leading to a more cautious leverage policy (e.g., Guo et al., 2024). Ezzi et al. (2020) explain that this negative impact arises from firms prioritizing their commitments to shareholders, which results in a lower capital structure and increased liquidity. Recent evidence shows that mandatory CSR disclosure affects firms’ financing behavior. In light of this, we propose the following hypothesis:

H1. 

CSR disclosure has a negative effect on firm leverage.

Although CSR may have a negative effect on firm leverage, no prior study has explicitly examined its relationship with zero leverage. Zero-leverage firms exhibit strong profitability, fewer financial constraints, and are economically stable (Dang, 2013). Based on stakeholder theory (Freeman, 1984), zero-leverage firms are seen as more socially responsible and legitimate (C. M. Deegan, 2019). Similarly, firms with higher CSR are seen by society as more reliable and sustainable firms, particularly concerned with the welfare of society and the environment, which gives them better long-term prospects (Weber et al., 2010; Jung et al., 2018), while attracting equity financing due to reduced information asymmetry (Kim & Park, 2023), and increased investor willingness to invest (Herbohn et al., 2019). Taking into account these arguments, the subsequent hypothesis is formulated:

H2. 

CSR disclosure increases the propensity of firms to follow zero-leverage policy.

3. Pakistan Institutional Settings and CSR

Pakistan is classified as a lower-middle-income economy with an underdeveloped capital market, which is generally characterized by small size and low liquidity (Beck et al., 2010). As a result, the information available to investors is small and, combined with lower liquidity, results in lower attractiveness and protection for small shareholders. As a consequence, the attractiveness of the capital market is also reduced for firms, making it more difficult and costly to raise debt or equity on capital markets (Djankov et al., 2007). Due to weak institutional quality and the limited appeal of Pakistan’s capital markets, firms rely predominantly on debt financing (N. A. Sheikh & Wang, 2011), making it even more difficult to study the zero-leverage phenomenon in this context.

Pakistan is a country whose society is deeply affected by terrorism, poor education and health infrastructure, and underdeveloped industry, and is characterized by political and economic instability (Ehsan et al., 2018). This reality is reflected in the weak regulation and control of firms’ activities, which leads to several problems related to working conditions, such as human rights violations, wage rates, and child labor. The overall level of social and environmental disclosure in Pakistan is low and inconsistent (Alam & Tariq, 2022). Therefore, as argued by Ehsan et al. (2018), there is an urgent need to examine CSR-related topics in Pakistan. CSR disclosure was introduced in Pakistan in 2009 by the SECP, which required listed firms to disclose their CSR activities in annual reports. In 2013, the SECP introduced the CSR Voluntary Guidelines, which make it mandatory to provide an articulated CSR policy and vision to promote ethics in business and increase the flow of information for shareholder decisions (Ehsan et al., 2018). An important and unexplored area of research is the potential link between CSR and capital structure. In particular, it is relevant to investigate the potential impact of CSR on sustainable and precautionary financing policies as a zero-leverage phenomenon.

4. Methodology

4.1. Data and Sample

The sample includes listed firms on the Pakistan Stock Exchange. We have excluded financial firms from the analysis as they are observed under different regulatory frameworks. The data were collected from the annual reports of 141 firms during the period 2010 to 2021. Similar to Ehsan et al. (2018), we measure CSR using the disclosure method previously applied by these authors in the Pakistani context. This method looks at all the information that a firm discloses in its yearly report about the activities it has done during the year, which may affect its CSR classification. Therefore, this method considers information about initiatives, programmes, and the number of resources used to influence the public at large and the firm’s stakeholders (Chan et al., 2014; Ehsan et al., 2018).

4.2. Variables

Table 1. Description of variables.

Variables	Symbols	Definitions
Dependent Variables
Total Debt	TD	(Long-term debt + short-term debt)/total Assets
Zero Total Debt	ZTD	A dummy variable that takes a value of 1 when short and long-term debt is zero in a given period; otherwise, it is 0.
Independent Variables
CSR Disclosure	CSRD	CSR score assigned to a firm/maximum CSR score
Profitability	ROA	Earnings of the firm after tax deduction to the total assets of the firm
Tobin’s Q	TOB Q	Share issued by share price divided by the total asset
Size	SIZE	Natural logarithm of total assets
Tangibility	TANG	Total fixed assets to total assets ratio
Liquidity	LIQUI	Current assets to current liabilities ratio
Cash	CASH	Cash and cash equivalents divided by total assets
Dividend	DIVI	Dividend paid during the year to total asset
CEO Duality	CDUAL	1 if the same person takes up the chairman and the chief executive position and 0 otherwise
Board Diversity	BDIVER	1 if any female director is on board and 0 otherwise
Board Size	BSIZE	Natural logarithm of the total number of directors on a firm’s board
Alternative Dependent Variables
Long-Term Debt	LTD	Long-term debt to total assets ratio
Zero Long-Term Debt	ZLTD	1 if the long-term debt is zero and 0 otherwise

presents the variables used in the study. Following the previous literature, zero leverage is a binary variable assuming the value of 1 if total debt is null (ZTD) and 0 otherwise (Morais et al., 2024; Bessler et al., 2013). We also use traditional measures of leverage, such as the total debt (TD) ratio, to examine the effect of CSR on firms’ debt levels, rather than just on the decision to have or not to have debt. For robustness purposes, zero long-term debt (ZLTD) and long-term debt ratio (LTD) are used. The main explanatory variable, CSR Disclosure (CSRD), is measured through the CSR disclosure method of Ehsan et al. (2018), followed by a content analysis to measure firms’ environmental and social performance and to explore their commitment to the social strength of society (Ehsan et al., 2018). The content analysis followed an index approach, which allows the presence or absence of a CSR activity to be verified through a binary coding method (1 if the firm performed the activity or applied the item related to CSR during the year and 0 in case of absence). Finally, an aggregate score is assigned to each firm based on the presence or absence of the selected items (Ehsan et al., 2018). Similar to Chan et al. (2014), Ehsan et al. (2018) and Malik and Kanwal (2018), the CSR items were examined through the firm’s annual reports, and a disclosure frequency approach was used to determine whether information on specific CSR activities is reported in firms’ annual reports. Following a procedure similar to Ehsan et al. (2018), who construct a CSR disclosure index for Pakistan firms, we apply a checklist of 40 items from 5 main themes/categories of CSR: (1) community welfare; (2) contribution to education and health sector; (3) environmental and energy importance; (4) product services and customer; and (5) workforce. The items considered for each CSR could be seen in Ehsan et al. (2018). The rationale is to assign a value of 1 or 0 for the presence or absence of the item in the annual report. After examining the presence or absence of the 40 items, an aggregate score is assigned to each firm (sum of all items assigned a value of 1) divided by the maximum score a firm can receive, 40 (1 for all items examined). The final aggregated score is, therefore, determined by the following equation:

$$CSRD=\sum d_i^{n40}/n_j$$

(1)

where $d_i=1$ if information on an item is disclosed and 0 otherwise. $n_j$ is the total number of items for $j_{th}$ firms, $n_j$ = 40. The greater the score, the greater the firm’s commitment to CSR.

The models incorporate firm-specific control variables, such as Size, Tangibility, Profitability, Liquidity, Cash, Dividend, and Tobin’s Q, a set of variables commonly identified by the extant literature as being significant in explaining a firm’s leverage (e.g., Dang, 2013; Strebulaev & Yang, 2013). Considering the role of corporate governance mechanisms on capital structure decisions, particularly on zero-leverage policies, some variables such as CEO Duality, Board Diversity, and Board Size were used to control for these effects. Unobserved specific effects are controlled by including, in all models, dummies for industry and years.

4.3. Model and Technique

This research is conducted to examine the effect of CSR on a firm’s debt ratio and zero leverage. In order to explore the overall impact of CSR on the firms’ debt ratio, we use standard linear regression models such as OLS. We estimate the effect of the independent variables on TD using the following equation:

$$\begin{gathered} {TD}_{it}={\beta }_0+{\beta }_1{CSRD}_{it}+{\beta }_2{BSIZE}_{it}+{\beta }_3{CDUAL}_{it}+{\beta }_4{BDIVER}_{it}+{\beta }_5{ROA}_{it}+{\beta }_6{TANG}_{it}+{\beta }_7{LIQUI}_{it} \\ +{\beta }_8{SIZE}_{it}+{\beta }_9{CASH}_{it}+{\beta }_{10}{DIVI}_{it}+{\beta }_{11}{TOBQ}_{it}+e_{it} \end{gathered}$$

(2)

A different econometric tool is used to estimate the specific impact of CSR on zero-leverage policies. In particular, we have to use a non-linear method. Given that the variables proxying for zero-leverage policies are binary (1 if the firm has zero debt and 0 otherwise), linear methods, such as standard linear regression models (e.g., OLS), cannot be used because they are designed for continuous dependent variables. Therefore, logit regression is the appropriate econometric technique for binary dependent variables such as ZTD (Yasmin & Rashid, 2019). Logit regression is a statistical method for calculating the likelihood of a categorical event, which relies on independent variables. By using independent variables that predict the dependent variable, logit regression is a predictive technique commonly used with categorical dependent variables (Ebrahimi et al., 2020). We estimate the effect of the independent variables on ZTD using the following equation:

$$\begin{gathered} {ZTD}_{it}={\beta }_0+{\beta }_1{CSRD}_{it}+{\beta }_2{BSIZE}_{it}+{\beta }_3{CDUAL}_{it}+{\beta }_4{BDIVER}_{it}+{\beta }_5{ROA}_{it}+{\beta }_6{TANG}_{it}+{\beta }_7{LIQUI}_{it} \\ +{\beta }_8{SIZE}_{it}+{\beta }_9{CASH}_{it}+{\beta }_{10}{DIVI}_{it}+{\beta }_{11}{TOBQ}_{it}+e_{it} \end{gathered}$$

(3)

5. Results

5.1. Univariate Results

Descriptive statistics are presented in

Table 2. Descriptive statistics.

Variables	Obs.	Mean	Std. Dev.	Min	Max
TD	1656	0.234	0.197	0.000	0.870
CSRD	1656	0.557	0.188	0.150	0.975
ROA	1656	0.060	0.094	−0.214	0.338
TANG	1656	0.707	0.351	0.007	1.810
SIZE	1656	3.939	0.726	2.098	5.617
LIQUI	1656	1.642	1.261	0.064	7.024
TOB Q	1656	1.977	2.609	0.123	12.406
CASH	1656	0.066	0.092	0.000	0.364
DIVI	1656	0.044	0.088	0.000	0.432
BSIZE	1656	2.103	0.197	1.099	3.045

. On average, CSRD assumes the value 0.557, with a standard deviation of 0.188, ranging between 0.15 and 0.975. On average, TD is 0.234, with a standard deviation of 0.197, ranging between 0 and 0.87.

For dichotomous variables, descriptive statistics are shown in

Table 3. Descriptive statistics of dichotomous variables.

Variables	Frequency Case = 1	Percent	Frequency Case = 0	Percent
ZTD	216	13.04	1440	86.96
CDUAL	205	12.38	1451	87.62
BDIVER	558	33.72	1097	66.28

. Among the total observations, 216 cases (13.04%) have a value of 1 for the variable ZTD, while 1440 cases (86.96%) have a value of 0. This means that around 13% of the observations correspond to zero-leverage firms, i.e., with no debt both on short and long term. Regarding the TD variable, an average total debt-to-assets ratio of 23% suggests that Pakistani firms present low debt ratios, which may indicate their preference for self-financing and/or equity issuance.

Table 4. Pairwise correlation.

Variable	CSRD	TD	ROA	TANG	SIZE	LIQUI	TOB Q	CASH	DIVI	BSIZE	CDUAL	BDIVER
CSRD	1.000
TD	−0.206 ***	1.000
ROA	0.255 ***	−0.460 ***	1.000
TANG	−0.125 ***	0.285 ***	−0.232 ***	1.000
SIZE	0.567 ***	−0.043 *	0.167 ***	−0.110 ***	1.000
LIQUI	0.004	−0.496 ***	0.394 ***	−0.246 ***	−0.055 **	1.000
TOB Q	0.102 ***	−0.105 ***	0.365 ***	−0.134 ***	−0.025	0.014	1.000
CASH	0.144 ***	−0.354 ***	0.261 ***	−0.266 ***	0.003	0.259 ***	0.191 ***	1.000
DIVI	0.047 *	−0.266 ***	0.365 ***	−0.178 ***	−0.116 ***	0.157 ***	0.350 ***	0.274 ***	1.000
BSIZE	0.208 ***	−0.047 *	0.161 ***	−0.036	0.322 ***	0.011	0.173 ***	0.001	0.111 ***	1.000
CDUAL	−0.141 ***	0.074 ***	−0.074 ***	0.025	−0.074 ***	−0.053 ***	−0.073 ***	−0.069 ***	−0.017	−0.088 ***	1.000
BDIVER	−0.059 **	0.054 **	−0.049 **	−0.027	−0.034	0.020	0.016	0.051 **	0.040	−0.075 ***	0.117 ***	1.000

*** p < 0.01; ** p < 0.05; and * p < 0.1.

shows the pairwise correlations between variables. The correlations between the independent variables are below 0.6 in all cases. This result suggests that there is no multicollinearity in this analysis. Table 4 also highlights a negative correlation between CSRD and TD (correlation coefficient of −0.206); the firm’s debt ratio decreases as CSR disclosure increases.

5.2. Multivariate Analysis

Table 5. Linear regression—effect of CSR on total debt.

DEP VAR. TD	Coef.	SE	t-Value	p-Value	Sig
CSRD	−0.152	0.026	−5.940	0.000	***
BSIZE	0.009	0.021	0.430	0.667
CDUAL	0.007	0.012	0.570	0.570
BDIVER	0.024	0.008	2.880	0.004	***
ROA	−0.538	0.052	−10.440	0.000	***
TANG	0.057	0.012	4.880	0.000	***
LIQUI	−0.043	0.003	−13.430	0.000	***
SIZE	0.018	0.007	2.690	0.007	***
CASH	−0.344	0.046	−7.530	0.000	***
DIVI	−0.181	0.050	−3.610	0.000	***
TOB Q	0.006	0.002	3.520	0.000	***
Constant	0.300	0.045	6.690	0.000	***
R-squared	0.386			Number of obs.	1656.000
F-test	93.802			Prob > F	0.000
Akaike crit. (AIC)	−1462.036			Bayesian crit. (BIC)	−1397.097

*** p < 0.01.

shows the results of linear regression using Total Debt (TD) as a dependent variable. These results allow us to analyze the effect of CSR on a firm’s debt ratio (Hypothesis 1). The table reports the coefficients, robust standard errors adjusted for heteroscedasticity, Wald tests for individual significance, and the respective significance levels evaluated in the last two columns of the table.

Table 5 shows that CSRD has a negative and statistically significant coefficient; thus, CSR has a negative effect on the firm’s debt ratio. An increase in CSR disclosure reduces the firm’s debt levels. In particular, an increase of one standard deviation in the CSRD variable, ceteris paribus, corresponds to a decrease of 15.2 percentage points (pp) in the firm’s debt ratio.

Table 6. Logistic regression—effect of CSR on zero leverage (Zero Total Debt).

DEP VAR. ZTD	Coef.	dydx	SE	t-Value	p-Value	Sig
CSRD	1.459	0.104	0.635	2.300	0.022	**
BSIZE	0.201	0.014	0.482	0.420	0.677
CDUAL	0.567	0.040	0.290	1.950	0.051	*
BDIVER	−1.596	−0.113	0.241	−6.610	0.000	***
ROA	1.675	0.119	1.209	1.390	0.166
TANG	−0.698	−0.050	0.295	−2.370	0.018	**
LIQUI	0.580	0.041	0.066	8.730	0.000	***
SIZE	−0.141	−0.010	0.177	−0.800	0.425
CASH	10.108	0.717	0.879	11.500	0.000	***
DIVI	3.361	0.238	0.956	3.510	0.000	***
TOB Q	0.019	0.001	0.038	0.500	0.614
Constant	−4.474		1.040	−4.300	0.000	***
Pseudo r-squared	0.372			Number of obs.		1656.000
Chi-square	476.951			Prob > chi2		0.000
Akaike crit. (AIC)	829.217			Bayesian crit. (BIC)		894.156

*** p < 0.01; ** p < 0.05; and * p < 0.1.

shows the logistic regression for Zero-Leverage Total Debt (ZTD). The results allow us to examine the impact of CSR on zero-leverage policies (Hypothesis 2). This table reports coefficients, robust standard errors adjusted for heteroscedasticity, Wald tests for individual significance, and the respective significance levels. We also present the average partial effect (dydx) to interpret the economic impact of the independent variables on the dependent variable.

Results show that the CSRD variable has a positive and significant coefficient. Therefore, CSR increases the propensity for firms to have zero leverage. A one-standard-deviation increase in CSRD is associated with a 10.4 percentage point increase in ZTD, ceteris paribus.

Regarding corporate governance variables, a positive change in BDIVER by one standard deviation corresponds to a 11.3 percentage point decrease in ZTD. A result that is also confirmed by the positive coefficient of the variable in Table 5, indicating that BDIVER contributes to an increase in the firm’s debt ratios. BSIZE and CDUAL have little explanatory power on zero-leverage policies. Among the firm-specific variables, ROA has an insignificant effect on zero leverage. However, it reduces the firm’s debt ratios (Table 5). A positive change in LIQUI and CASH by one standard deviation corresponds to an increase in ZTD of 4.11 and 71.1 percentage points, respectively. One standard deviation of TANG has a negative effect of 5pp on ZTD, while DIVI has a positive effect of 23.8pp on ZTD. Finally, SIZE and TOB Q have non-statistically significant coefficients.

5.3. Robustness Section

The robustness of the results discussed above is evaluated in this section through the presentation of some tests. In particular, given that several authors claim that capital structure decisions relate only to long-term debt, we replace the dependent variables TD and ZTD with the variables long-term debt (LTD) and zero-leverage long-term debt (ZLTD). Therefore,

Table 7. Linear regression—effect of CSR on long-term debt.

DEP VAR. LTD	Coef.	SE	t-Value	p-Value	Sig
CSRD	−0.126	0.018	−7.170	0.000	***
BSIZE	0.012	0.014	0.860	0.390
CDUAL	−0.005	0.008	−0.610	0.539
BDIVER	0.013	0.006	2.240	0.025	**
ROA	−0.230	0.036	−6.460	0.000	***
TANG	0.092	0.008	11.460	0.000	***
LIQUI	−0.008	0.002	−3.600	0.000	***
SIZE	0.050	0.005	10.450	0.000	***
CASH	−0.150	0.032	−4.740	0.000	***
DIVI	−0.072	0.035	−2.070	0.039	**
TOB Q	0.001	0.001	0.610	0.539
Constant	−0.088	0.031	−2.840	0.005	***
R-squared	0.270			Number of obs.	1656.000
F-test	55.240			Prob > F	0.000
Akaike crit. (AIC)	−2690.163			Bayesian crit. (BIC)	−2625.224

*** p < 0.01; and ** p < 0.05.

shows the re-estimated model of Table 5 using LTD as the dependent variable, and

Table 8. Logistic regression—effect of CSR on zero-leverage long-term debt.

DEP VAR. ZLTD	Coef.	dydx	SE	t-Value	p-Value	Sig
CSRD	2.584	0.363	0.445	5.810	0.000	***
BSIZE	0.688	0.097	0.358	1.920	0.055	*
CDUAL	0.803	0.113	0.196	4.090	0.000	***
BDIVER	−1.152	−0.162	0.156	−7.380	0.000	***
ROA	1.149	0.162	0.868	1.320	0.186
TANG	−1.651	−0.232	0.214	−7.720	0.000	***
LIQUI	0.235	0.033	0.054	4.330	0.000	***
SIZE	−1.110	−0.156	0.123	−9.020	0.000	***
CASH	5.239	0.737	0.732	7.150	0.000	***
DIVI	3.045	0.428	0.785	3.880	0.000	***
TOB Q	0.091	0.013	0.026	3.490	0.000	***
Constant	0.506		0.748	0.680	0.498	***
Pseudo r-squared	0.258			Number of obs.		1656.000
Chi-square	502.783			Prob > chi2		0.000
Akaike crit. (AIC)	1471.680			Bayesian crit. (BIC)		1536.619

*** p < 0.01; and * p < 0.1.

shows the re-estimated model of Table 6 using ZLTD as the dependent variable.

These results demonstrate congruence with those previously reported. The CSRD variable retains its sign and significant coefficients. Table 7 shows that CSR disclosure significantly decreases long-term debt ratios. In particular, a positive change in CSRD by one standard deviation reduces long-term debt by about 12.6 percentage points, ceteris paribus. Table 8 adds that the CSRD variable also has a positive effect on zero long-term debt. In particular, the propensity of firms to become debt-free increases by 36.3 percentage points with a one-standard-deviation increase in CSRD, ceteris paribus. Overall, the study finds that firms with a higher level of CSR are more likely to be debt-free, with this finding being robust to the use of various measures that proxy the capital structure of the firm.

Finally, the robustness of the results to the use of alternative econometric models is tested. Considering the panel nature of our data, we use firm-fixed effects estimators, which account for individual-specific factors that may also explain firms’ capital structure, including firm- and industry-specific characteristics (Morais et al., 2025). Avoidance of biased estimation caused by unobserved heterogeneity is therefore enabled by these methods.

Table 9. Linear regression—effect of CSR on total debt using firm-fixed effects estimators.

DEP VAR. TD	Coef.	SE	t-Value	p-Value	Sig
CSRD	−0.090	0.041	−2.21	0.027	**
BSIZE	−0.015	0.023	−0.63	0.531
CDUAL	0.012	0.011	1.06	0.288
BDIVER	0.004	0.008	0.54	0.587
ROA	−0.528	0.045	−11.78	0.000	***
TANG	−0.019	0.015	−1.26	0.208
LIQUI	−0.034	0.004	−8.84	0.000	***
SIZE	0.032	0.018	1.77	0.077	*
CASH	0.036	0.041	0.88	0.377
DIVI	−0.069	0.050	−1.38	0.167
TOB Q	0.006	0.002	3.35	0.001	***
Constant	0.276	0.084	3.29	0.001	***
R-squared	0.316			Number of obs.	1656.000
F-test	30.130			Prob > F	0.000
Akaike crit. (AIC)	−2916.681			Bayesian crit. (BIC)	−2851.742

*** p < 0.01; ** p < 0.05; and * p < 0.1.

and

Table 10. Logistic regression—effect of CSR on zero leverage using logit specifications with firm-fixed effects estimators.

DEP VAR. ZTD	Coef.	SE	z-Value	p-Value	Sig
CSRD	6.715	3.049	2.200	0.028	**
BSIZE	−0.019	1.063	−0.020	0.986
CDUAL	−0.660	0.804	−0.820	0.412
BDIVER	−2.617	0.591	−4.430	0.000	***
ROA	9.005	2.851	3.160	0.002	***
TANG	1.087	0.939	1.160	0.247
LIQUI	0.596	0.187	3.190	0.001	***
SIZE	2.976	1.577	1.890	0.059	*
CASH	5.282	1.897	2.790	0.005	***
DIVI	−0.401	2.091	−0.190	0.848
TOB Q	−0.229	0.120	−1.920	0.055	*
Pseudo r-squared	0.258		Number of obs.		480.000
Chi-square	125.660		Prob > chi2		0.000
Akaike crit. (AIC)	267.674		Bayesian crit. (BIC)		317.759

*** p < 0.01; ** p < 0.05; and * p < 0.1.

re-estimate the models of Table 5 and Table 6, respectively, using fixed-effects estimators. While applying fixed-effects estimators to continuous dependent variables (total debt—Table 9) is not expected to affect the number of observations included in the estimations, applying fixed-effects models to binary dependent variables (Zero Total Debt—Table 10) is not always feasible and can substantially reduce the number of observations retained in the estimation. In the context of binary dependent variables, logit models with fixed-effects estimators entail a substantial loss of observations because observations without variation within groups are automatically excluded. In particular, this approach leads to a severe reduction in sample size (in our case, from 1656 firm-year observations to 480), as firms that exhibit during the whole period a zero-leverage policy or maintain a constant leverage status are dropped from the analysis.

Table 9 and Table 10 show that our main results are robust to the use of fixed-effects estimators. In particular, Table 9 confirms the negative effect of CSR disclosure on a firm’s debt ratio. Even considering the specificities of fixed-effects estimators in the framework of binary dependent variables, Table 10 confirms that CSR disclosure increases the propensity for firms to have zero leverage.

6. Discussion

This paper focuses on the impacts of CSR disclosure on the zero-leverage phenomenon. The results validated Hypothesis 1, which posited that CSR disclosure has a negative effect on firm leverage. Firms that report their CSR have a higher propensity to behave morally, which may lead them to collaborate with other stakeholders and the public at large on social issues and to make more prudent financial decisions. Firms that effectively incorporate CSR disclosure may benefit financially (improving a firm’s efficiency and investing prospects). One possible explanation for this negative effect of CSR disclosure on debt ratios is that firms with more CSR disclosure have a better reputation in society, which may increase investors’ willingness to invest in these firms, favoring equity financing and reducing the firm’s dependence on bank debt. These results are consistent with those presented by Pijourlet (2013), Harjoto (2017), and S. Sheikh (2019), which suggest that CSR disclosure has a negative impact on leverage, due to a lower cost of equity, improved reputation, or even reduced information asymmetry.

The results are consistent with the argument that firms with more CSR disclosure are more likely to have zero-leverage policies. Thus, these results confirm the relationship between firms’ socially conscious actions and their propensity to go debt-free. This confirms that society perceives firms with greater CSR as more reliable and sustainable. Therefore, firms reduce information asymmetries and contribute to increasing investor willingness to invest, which ultimately increases the firm’s propensity to obtain alternative sources of financing to debt on more favorable terms, such as equity financing (Weber et al., 2010; Jung et al., 2018). Overall, we conclude that hypothesis H2 is supported.

Regarding additional variables, the results show that the presence of women on the board reduces the propensity toward zero-leverage policies. Moreover, tangibility has a negative effect on zero leverage, under the argument that asset tangibility acts as collateral to debt, increasing their propensity to use debt financing. Dividends, liquidity, and cash holdings seem to increase the susceptibility of firms to follow zero-leverage policies. The likelihood of firms becoming debt-free is increased by the payment of higher dividends, perhaps because they can obtain more equity financing as they are also expected to pay higher dividends. At the same time, internal sources of liquidity represent alternative sources of financing, which agrees with the pecking order theory. These results are in line with the conclusions of Zaman et al. (2022), since the authors concluded that the link between CSR disclosure and zero leverage is stronger in firms that pay dividends. Overall, these results seem to support the fact that a firm is more prone to adopt zero leverage when it is profitable, in line with the conclusions of Ehsan et al. (2023).

This research adds to the literature by testing long-term debt and zero leverage in the Pakistani context. The results support the idea that in a developing country context, bank debt tends to be short-term only, possibly due to a lack of savings (N. A. Sheikh & Wang, 2011). Also, board gender diversity shows a counterintuitive role that brings novelty to the literature on governance–CSR.

Although this study provides evidence of the relationship between CSR and zero leverage in Pakistan, it should be noted that several lines of future research are emerging. For instance, the way in which firms or sector types respond to CSR may vary, so it may be necessary to consider possible sample heterogeneity. Despite recent evidence of varying CSR–leverage responses, heterogeneity across firm types (size and sector) remains under-explored. Extending the research to non-listed firms is also noteworthy when available, given that they face different types of stakeholders, have concentrated ownership, and follow different financing strategies.

7. Conclusions

The zero-leverage phenomenon has gained popularity in the last decade, as it has been identified as a growing phenomenon that enhances the financial performance of a firm. Given this growing relevance, we analyzed a sample of 141 listed Pakistani firms between 2010 and 2021 and found that more than 13% of the observations correspond to zero-leverage firms. Building on this, our study aimed specifically to investigate the effect of CSR on the adoption of zero-leverage policies.

Using linear regression methods to estimate the effect of CSR on firms’ debt ratios and nonlinear regression methods, such as logit models to estimate the effect of CSR on zero-leverage policies, we find that CSR disclosure significantly affects firms’ financial decisions. Indeed, we show that a higher level of CSR disclosure decreases a firm’s debt ratio. Besides that, we found that CSR disclosure increases the likelihood of firms having zero leverage. These conclusions proved to be robust to different capital structure measures. Therefore, our findings further support the argument that firms that are more committed to sustainability and social responsibility are more likely to adopt sustainable financing policies, which are generally identified in the literature as factors that contribute to improving the firm’s financial performance, as they are zero-leverage policies. A possible explanation for these findings is that firms that are more concerned about and contribute to solving the problems of the surrounding society promote a greater flow of information that is interpreted by investors as a safer and more sustainable firm. Ultimately, this increases the propensity of investors to invest in these higher-CSR firms, as they feel that their private funds are better protected when invested in these firms than in firms with lower CSR levels. CSR disclosure can, therefore, help to increase firms’ equity financing and reduce their reliance on bank debt.

Beyond theoretical contributions, by offering distinctive perspectives on the integration of CSR and financial strategy, this paper also has practical implications. For executives and managers of firms in emerging markets, who are mostly dependent on bank debt due to their underdeveloped capital markets, we show that they can potentially gain easier and more favorable access to capital markets and equity financing by increasing their CSR disclosure, enabling these firms to reduce their debt levels. For investors and shareholders, we show that by encouraging greater CSR disclosure, firms become less opaque, reducing information asymmetries and potentially making them safer firms to invest in. For regulators and policymakers, this study draws attention to the critical role that CSR disclosure plays in creating important links between business and society. We show that this can boost investor participation in capital markets, providing firms with alternative financing. Hence, it is important that regulators promote and develop incentive programs for firms to increase their CSR to build relationships and trust with their stakeholders. Studies conducted in different countries and areas can help us determine how cultural, economic, and institutional factors affect the link between CSR and zero leverage. At the same time, we can learn a lot about the complex relationship between CSR and financial decisions by looking at how CSR and its effects on zero leverage change in different situations.