Corporate social responsibility and bankruptcy

Elizabeth Cooper (Department of Finance, La Salle University, Philadelphia, Pennsylvania, USA)
Hatice Uzun (Department of Finance, Long Island University, Brookville, New York, USA)

Studies in Economics and Finance

ISSN: 1086-7376

Article publication date: 19 June 2019

Issue publication date: 21 June 2019

Abstract

Purpose

This paper aims to examine corporate social responsibility (CSR) and corporate bankruptcy. Specifically, the authors ask the following research questions: Does CSR play a role in determining the likelihood of bankruptcy? Does CSR explain the difference in the probability of that firm eventually reorganizing and emerging from bankruptcy?

Design/methodology/approach

The authors address these questions by testing three CSR theories using a sample of 78 firms that filed for Chapter 11 bankruptcy during the period 2007 to 2014 along with a matched sample of firms that did not.

Findings

Overall, the findings indicate that stronger CSR firms are less likely to become bankrupt relative to weaker CSR firms, all else being equal. This result is in line with the stakeholder theory of CSR. However, results do not support the conjecture that CSR matters when it comes to bankruptcy emergence. While CSR seems to influence whether a company experiences bankruptcy in the first place, having strong CSR does not seem to help a firm once it has filed for Chapter 11.

Research limitations/implications

This paper extends the existing CSR literature but looks at CSR not from the angel of financial “success” but rather from financial “failure”.

Practical implications

The results could potentially help academics and practitioners alike in seeking understanding and reason behind CSR involvement and bankruptcy avoidance and success.

Originality/value

This is the first paper to test whether CSR plays a role in bankruptcy. The authors use a recent sample of firms with CSR scores that experienced a bankruptcy and a matched sample of CSR-scored firms that did not experience bankruptcy.

Keywords

Citation

Cooper, E. and Uzun, H. (2019), "Corporate social responsibility and bankruptcy", Studies in Economics and Finance, Vol. 36 No. 2, pp. 130-153. https://doi.org/10.1108/SEF-01-2018-0013

Publisher

:

Emerald Publishing Limited

Copyright © 2019, Emerald Publishing Limited


1. Introduction

Does corporate social responsibility (CSR) improve a firm’s financial performance? This question is asked many times over in the academic literature. While the CSR–financial performance relationship has been studied from various business perspectives including management, finance and accounting, the results do not overwhelmingly support a distinct relationship between the two. On the one hand, there is research showing a positive relationship between CSR and financial performance. In a meta-analysis of studies looking at the relationship between CSR and performance, Margolis and Walsh (2003) find that almost 50 per cent of the studies examined demonstrate a positive relationship between the two. On the other hand, other studies, such as Brammer et al. (2006), demonstrate a negative relationship between CSR and stock market returns. Additionally, studies such as Nelling and Webb (2009) and Mahoney and Roberts (2007) find generally inconclusive or no relationship between measures of CSR and financial performance.

In this paper, we look at an element of performance that to the best of our knowledge so far has not been analyzed in the CSR framework: bankruptcy. Much of the previous literature in CSR has been devoted to whether CSR determines or plays a role in financial outcomes such as stock returns, accounting performance indicators, risk and capital costs. Bankruptcy, while an outcome of financial performance, is certainly not the outcome of choice for a firm[1]. The importance of bankruptcy avoidance, however, cannot be understated. As an example, Noulas and Genimakis (2014) report that bankruptcy avoidance is considered the primary goal for CFOs when focusing on capital structure decisions. Further, the situation of financial distress and bankruptcy provides an ideal setting in which to test the true worth and practicality of CSR. Here, the first research question that we ask is whether CSR plays a role in determining the likelihood of bankruptcy. Does CSR make a difference in whether a firm faces the ultimate in financial distress and files for Chapter 11 bankruptcy? In addition, for those firms that do file for bankruptcy, does CSR explain any difference in the probability of that firm eventually reorganizing and emerging from bankruptcy?

Our analysis is based on a sample of 78 US firms that filed for Chapter 11 bankruptcy during the period 2007-2014. The results suggest that CSR plays a role in determining whether or not a firm files for bankruptcy. In particular, firms with better CSR performance are less likely to file for bankruptcy compared with firms with worse CSR performance. This is based on our use of a matched sample approach so that we can scientifically look at differences in how CSR affects firm performance. This result supports the stakeholder theory of CSR. We also look at the relationship between CSR and bankruptcy probability under conditions of varying leverage ratios, firm size and Z-scores. Under Chapter 11, firms are able to and encourage to reorganize and, eventually, emerge as a stronger company. In this vein, we also ask whether CSR plays a strategic role in firms that are able to successfully reorganize and emerge as a firm with improved performance and viability. We find limited evidence, however, that CSR affects the likelihood of a firm emerging from bankruptcy. Therefore, while CSR may be an effective tool in avoiding bankruptcy in the first place, it is not clear that CSR helps a firm strategically once it is deemed insolvent.

The theories that are commonly used to analyze CSR in a financial performance context are used here with the background of bankruptcy and financial distress. Interestingly, bankruptcy brings together much of the theoretical underpinnings of CSR research in that it directly involves several stakeholder groups. For instance, creditors endure most of the direct cost of bankruptcy (Eckbo et al., 2016). Along with the creditors from the investment side, suppliers are also impacted by the cessation of a client firm. Of course, shareholders are affected by a bankruptcy even before the firm files. A firm that undergoes bankruptcy also affects employees, the community and customers. Bankruptcy is costly for employees in the form of losing income and other benefits of working in a firm. For example, Verwijmeren and Derwall (2010) examine the relationship between employee well-being and debt-ratios. They show that strong employee well-being is associated with lower debt-to-asset ratios. While general economics play a role in the success or failure of a business, at the heart of the firm are the insiders – managers and executives – who ultimately make decisions that lead the firm one way or another.

This paper extends the existing CSR literature but looks at CSR not from the angel of financial “success” but rather from financial “failure.” We argue that this is an important direction for the CSR literature to take as it has been shown that CSR’s impact on the financial success (such as stock returns or profitability) is less than conclusive. It could be the case that CSR does not substantively influence financial measures of successful performance, but rather it acts as a deterrent or precipitator of a firm’s ultimate demise. If, however, it does not play a role in bankruptcy determination or outcome, this non-relationship may actually support the idea that CSR is inherent to a company’s position on the responsibility of the firm and is not done for ulterior motivations.

The results from our research could potentially help academics and practitioners alike in seeking understanding and reason behind CSR involvement and bankruptcy avoidance and success. The question of whether CSR matters at all is worth asking. Does CSR matter if a firm wants to avoid bankruptcy? If so, why is this the case? If a firm does file for bankruptcy, does the amount of CSR involvement help in turning the company around and ultimately emerging from bankruptcy? Alternatively, CSR may not be useful in preventing bankruptcy or emerging from it, but rather it is a standalone decision based on belief and character that is without regard to financial success (or failure). This paper provides some support for the stakeholder theory of CSR and adds to the literature on CSR in this context. Specifically, in determining financial failure, as opposed to success, CSR seems to be important. However, we caution that our results are based primarily on the largest firms that went through bankruptcy proceedings during this time as our CSR data only cover large firms. Therefore, we cannot say that these results hold for all firms but rather for large firms that go to experience bankruptcy.

2. Literature review and hypothesis development

2.1 Bankruptcy

The number of businesses filing for bankruptcy has peaked in recent years because of the financial crisis in 2008. In particular, according to the American Bankruptcy Institute, there were 60,837 business bankruptcies filed in 2009 compared to 24,735 in 2015 and 35,472 in 2000. Financial, corporate governance and ethical issues that contribute to the risk of bankruptcy have increasingly attracted the attention of researchers, practitioners and policymakers.

One line of existing studies examines the link between accounting and market-based financial data and bankruptcy (Kwak et al., 2005; Reisz and Perlich, 2007; Franzen et al., 2007; Dawkins et al., 2007; Singhal and Zhu, 2013; Lyandres and Zhdanov, 2013; Ho et al., 2013). Another line of existing studies examines the link between corporate governance and risk of the bankruptcy (Fich and Slezak, 2008; Parker et al., 2002; Platt and Platt, 2012; Robinson et al., 2012; Darrat et al., 2016; Eckbo et al., 2016). However, to the best of our knowledge, there has been no study on the link between CSR and bankruptcy. To fill this gap in the literature, we examine the relationship between CSR and the likelihood of bankruptcy. Furthermore, we also examine whether CSR explains any difference in the probability of that firm eventually reorganizing and emerging from bankruptcy.

When a company files for Chapter 11 under the US Bankruptcy Code, it is essentially declaring that although it is not able to keep the business afloat, the company wants to reorganize itself so that eventually it can continue as a going concern in the future. At the time, a firm files for Chapter 11 bankruptcy they also must propose a plan of reorganization. At the heart of the plan is the intent to keep the business alive. Generally, firms need to figure out a way to pay back their creditors over time and work with suppliers and other stakeholder groups that are impacted by the firm and that also impact the firm’s own value-creating activities.

The existence of a reorganization plan does not automatically mean it will successfully emerge from bankruptcy. Research points to certain characteristics that tend to predict whether a firm will successfully emerge from bankruptcy. For instance, Denis and Rodgers (2007) find that firms that significantly reduce assets and liabilities increase the likelihood that they will successfully emerge from bankruptcy. Further, they find that the time spent in Chapter 11 is directly related to the pre-bankruptcy size and operating performance of the firm. Heron et al. (2009) find that in firms that do emerge from bankruptcy; their debt ratios are higher than the average of their respective industries. They argue that Chapter 11 proceedings may not be the most efficient model of reorganization for firms experiencing financial distress.

In a study of firms that emerge from Chapter 11 bankruptcy, Bogan and Sandler (2012) find that having new management in place is the strongest determinant of post-bankruptcy survival. So, although Chapter 11 may not be the most efficient reorganization model, it may be the case the procedure works effectively when under proper management. James (2016) finds that relationships with stakeholders (as measured by executory contracts) influence the probability of whether a firm subsequently emerges as a going concern. In particular, James finds that rejected executory contracts (contracts that the firm has with primary stakeholders) increase the likelihood of a firm emerging from bankruptcy. This result supports a stakeholder management view of bankruptcy whereby firms use Chapter 11 to make strategic changes that ultimately increase firm value. Interestingly, one theory of CSR involves stakeholders, and thus, a clear connection between the two kinds of literature becomes evident.

2.2 Corporate social responsibility

Why companies engage in CSR remains a question but three competing arguments take hold in the literature: Stakeholder theory, agency theory and pure altruism. Originating from Freeman (1984), stakeholder theory suggests that CSR resolves conflicts among stakeholder groups, and therefore, should be compatible with better financial performance. Managing and supporting stakeholder groups effectively should be associated with stronger long-term value. In this sense, CSR is not purely an altruistic ideal but is integral to the success of the firm financially. Stakeholder theory also suggests firms that purposefully and strongly engaging in CSR ultimately take on lower risk. Erhemjamts et al. (2013) support the idea that CSR can be effective as a stakeholder management tool and firms that are strong CSR performers exhibit favorable investment and organizational strategies. Wu and Shen (2013) find that CSR is positively associated with financial performance in the banking industry and argue that this supports the idea that CSR is a strategic choice of the firm. Dayanandan et al. (2018) find that stronger CSR firms are more likely to provide earnings warnings (notably before negative announcements) and tend to have smaller abnormal returns following the announcement relative to weak CSR firms. Further, from stakeholder theory, CSR activities may reduce the cost of capital. Some recent studies focus on the link between CSR and the overall cost of the cost of capital. Cajias et al. (2014), Ghoul et al. (2011), Oikonomou et al. (2014) and Cooper and Uzun (2015), all find that better CSR tends to reduce a firm’s cost of capital. Erragragui (2018) finds that environmental and governance strengths (components of CSR) reduce the cost of debt.

Agency theory derives its roots from Jensen and Meckling (1976), in which the authors state that there is a distinction and information asymmetry between agents (managers) and principles (shareholders). In this context, CSR is ultimately a value-destroying activity as it is seen as a waste of valuable resources in the face of the firm’s true purpose of increasing shareholder wealth. Specifically, agency theory suggests that firms that engage in CSR activities do so for the benefit of insiders as it improves their reputation, but it is at the expense of the firm’s investors, who would rather see that money used for value-creating investments or returned to shareholders through dividends. Ultimately, agency theory predicts a detrimental financial impact on firms engaging in overt CSR activities.

Finally, pure altruism suggests that firms engage in CSR activities without an ulterior motive; they do so because of an inherent belief that a company has a duty beyond financial performance. In line with this argument is the finding that successful firms have the resources available to address social issues and are seen to do so to a greater extent compared to firms that do not have the financial means. Waddock and Graves (1997) point to a cycle of financial success leading companies to engage more heavily in social issues and those social endeavors helping to lead to more financial success. At the heart is the idea that a company may choose CSR simply because it is able to do so and managers, with shareholder support, justify the spending on the firm’s responsibility beyond financial performance and reward. In fact, the term “CSR” suggests just that; a company’s ultimate responsibility lies beyond shareholder success.

While the relationship between CSR and bankruptcy has not specifically been studied before (to the best of our knowledge), several papers have connected the two in different capacities. Hogan et al. (2014) find that firms with a low probability of bankruptcy as measured by Altman’s Z-score tend to participate in more community giving relative to less financially sound firms. They also find that board size and the percentage of women on board are directly related to the extent of community spending as a percentage of EBITDA. Ahn and Park (2018) study the CSR practices in eight long-lived Korean firms and come to the conclusion that long-term survival is at least partially determined by social performance. The authors state that building social capital amongst primary stakeholders through CSR activities helps firms to survive over time while competing firms without these relationships fail.

In this paper, the CSR theories will each be tested in the context of bankruptcy. First, we test whether CSR plays a role in the probability that a firm experiences bankruptcy. Specifically, we look at a sample of firms that have filed for Chapter 11 bankruptcy and a matched sample of firms that have not filed for bankruptcy and see whether CSR differentiates between the two. Our test results should lend support to one of the three preceding theories. While each of the theories is a possibility, based on the bankruptcy literature we expect that stakeholder theory is the predominant theory to emerge. Hence, our H1 is the following:

H1.

There is a negative relationship between CSR and the likelihood of bankruptcy.

This would imply that a firm with strong CSR has a lower probability of bankruptcy, all else being equal. Stakeholder theory suggests that CSR creates a positive atmosphere with the firm’s various stakeholder groups and that this would create financial benefits for the firm. Therefore, if this is the case the firms with stronger CSR programs and actions should see a better financial performance, and therefore, exhibit a lower likelihood of filing for bankruptcy. While previous literature has shown mixed results (albeit most findings suggest a positive relationship between CSR and financial performance), it may be the case that the impact of CSR on typical measures of financial performance is difficult to detect. Therefore, looking at bankruptcy may give a different and clearer perspective on this relationship.

However, should we find that there is a positive relationship between CSR and the likelihood of bankruptcy, the result would lend support to the agency theory of CSR. Further, if no relationship exists between CSR and bankruptcy this would suggest that CSR may be a purely altruistic activity and one that is more related to firm choice and preference rather than as an entrenchment tool (agency theory) or as a means to enhance financial performance (stakeholder theory).

We extend this analysis to see whether certain factors about the firm would make CSR more (or less) important in terms of bankruptcy prediction. We examine the interaction effect of two variables. Specifically, if a firm is highly leveraged, the stakeholder benefits associated with CSR may not be enough to steer a company away from the path of bankruptcy. As highly leveraged firms need to generate and retain cash to pay higher interest on the debt, this might reduce their ability to fund more CSR and CSR reporting (Barnea and Rubin, 2010). Further, at some point, the risks associated with high amounts of debt would outweigh any benefit to the firm associated with strong CSR.

Additionally, the size of the firm may make a difference in terms of how CSR impacts the probability of bankruptcy. For large firms, the impact of CSR should be stronger on bankruptcy determination as these firms may be more visible and would benefit the most from stakeholder considerations. According to Barnea and Rubin (2010), large firms have a greater impact on communities (stakeholders, investors, customers and authorities), and therefore, firm size is likely to influence the amount of CSR disclosure to address the concern of various stakeholders groups.

Altman’s Z-score, a gauge of the likelihood of bankruptcy that incorporates profitability, leverage, liquidity, solvency and activity should also influence the relationship between CSR and the probability of bankruptcy. Firms with low Z-scores, indicating a stronger possibility of bankruptcy, should see less of a relationship between CSR and bankruptcy probability as these firms (particularly those with Z-scores less than 1.8) are likely headed to bankruptcy regardless of their CSR involvement. However, firms with strong Z-scores may see a stronger relationship between CSR and bankruptcy, as common measures of bankruptcy probability are not conclusive regarding the outcome.

Hence, we propose three additional hypotheses:

H1a.

The inverse relationship between CSR and the likelihood of bankruptcy is stronger for firms with low debt ratios relative to firms with high debt ratios.

H1b.

The inverse relationship between CSR and the likelihood of bankruptcy is stronger for large firms relative to small firms.

H1c.

The inverse relationship between CSR and the likelihood of bankruptcy is stronger for firms with strong Z-scores relative to firms with weak Z-scores.

Second, examining the firms that experience bankruptcy, we analyze whether CSR is a contributing factor in whether that firm ultimately survives and emerges from bankruptcy. We also expect to see the stakeholder theory of CSR emerging as the preeminent theory. Specifically, the second hypothesis is the following:

H2.

There is a positive relationship between CSR and the likelihood of emergence from bankruptcy.

Stakeholder theory would suggest that with the support of stakeholders, better performing CSR firms should be able to successfully navigate through bankruptcy proceedings and emerge eventually as a newly formed company. Their ties to stakeholders through CSR engagement should help the firm emerge from bankruptcy at a higher rate than firms with weak CSR and less stakeholder support.

As with H1, alternate theories may be supported. A negative relationship between CSR and the likelihood that the firm successfully emerges from bankruptcy would indicate that managers simply use CSR as a tool to protect themselves. Ultimately, CSR would negatively impact financial performance, and therefore, would be a deterrent to bankruptcy emerge rather than a support. A non-significant relationship could indicate that CSR is a strategic preference of the managers and owners of the firm, and therefore, an altruistic behavior rather than a behavior with a goal to influence management entrenchment or firm performance. It could also indicate that CSR is not something that can help (or hurt) a firm’s chances of recovering from bankruptcy. Other factors may outweigh this particular indicator in determining the likelihood of emergence.

3. Data

We begin by using the LoPucki Bankruptcy Research Database. This database includes firms that have filed for bankruptcy since October 1979 with the condition that the firms be considered “large” (assets over $100m in 1980), publicly traded and had filed a 10-K within the three years prior to filing for Chapter 11. Here, we identify those firms that filed for Chapter 11 bankruptcy from 2007 to 2014. There are 246 firms in the LoPucki Database during this time period. From this database, we also collect information about the bankruptcy itself including financial data about the firm one year prior to bankruptcy and characteristics regarding the bankruptcy itself. The database identifies whether the firm filed bankruptcy along with pre-packaged or pre-negotiated terms that could ease the negotiations of large creditors regarding debt contracts. Further, we collect information on how long the CEO was in that position prior to the bankruptcy filing date and whether the CEO was replaced during the bankruptcy proceedings. Additionally, we identify whether a firm eventually emerged from Chapter 11 and ultimately “survived”.

The sample of firms from LoPucki’s Database is matched to firms in MSCI’s STATS database over the same time period. The STATS database, formerly known as KLD, is a prominent tool in CSR research (Erragragui, 2018). The STATS data include CSR data for selecting US publicly-traded companies since 1991. It varies by year, but as of 2012 environmental, social and governance (ESG) STATS contained data on the largest 3,000 US publicly traded companies by market capitalization. The data consist of annual scores of ESG performance indicators. The indicators identify positive and negative performance attributes in seven areas including community relations, diversity, employee issues, environmental matters, product safety, corporate governance and human rights. Here, we include all categories except human rights[2].

Community relations includes issues relating to charitable giving and community engagement. Diversity contains a wide range of issues and some include workplace diversity and minority contracting. Employee issues pertain to profit sharing, benefits and professional development opportunities as examples. Environmental issues include such areas as clean technology, waste management and carbon emissions. Product safety, where applicable, refers to areas such as chemical safety, financial product safety and privacy and data safety. Corporate governance includes issues relating to governance structures and risk controls.

The firm rates each firm separately on these different dimensions of CSR. The rating is binary in that an indicator is given a score of “0” if a particular indicator is not present in the firm or a score of “1” if it is relevant. Strengths and concerns are rated separately. For instance, Bally Total Fitness, which filed for bankruptcy in 2007 (and later emerged from bankruptcy) was rated a “1” for the concern indicator variable relating to marketing and advertising in 2006 (the year prior to the bankruptcy filing). According to MSCI’s documentation, this means that there was evidence of controversies involving the firm’s marketing and advertising practices, which include factors such as “widespread or egregious instances of false, discriminatory or improper marketing/advertising, marketing targeted at disadvantaged groups, resistance to improved practices and criticism by NGOs and/or other third party observers.” As another example, Blockbuster Inc., which filed for bankruptcy in 2010 (and was later acquired by Dish Network), was giving a rating of “0” for the community impact concern variable in 2009. This means that MSCI saw no evidence that the firm engaged in controversies related to the community in which it does business.

Overall, we are able to match 78 firms between LoPucki and STATS databases during our examination period of 2007-2014. It should be noted that the firms were able to match in the MSCI data tended to be the largest firms from the LoPucki database (as MSCI data contains only the largest firms by market cap per year). In fact, the sample of bankrupt firms that were not included in our final sample (i.e. did not have CSR data from MSCI) were, on average, ten times smaller than the firms used in our final sample. We use the CSR data for the year prior to the bankruptcy filing. To measure CSR, we follow Rekker et al. (2014) and Dunbar et al. (2016) and use Net CSR Score. The Net CSR Score is the sum of the scores of each of the six categories. Specifically, NET_CSR_SCORE = COM_score + DIV_score + EMP_score + ENV_score + PRO_score + CGOV_score, where the individual categorical scores are calculated by subtracting the total number of concerns from the total number of strengths in each category. In addition to using the continuous variable Net CSR Score, we also divided the score into high and low values where HI_CSR_SCORE is equal to 1 if the NET_CSR_SCORE is above the median for the sample and equal to 0 if is below the median.

Tables I and II present frequency distributions of the bankruptcy filings in our sample by year and by industry. The year 2009 accounted for almost 40 per cent of our sample of bankruptcies, with 2013 and 2014 having the fewest. Overall the financial crisis (the period from 2008 to 2010) contained 52 of the 78 bankrupt firms used in the sample. As Table II reports, the industry with the highest frequency of bankruptcies was manufacturing followed by finance and transportation/communication (industry is a variable classified in the LoPucki database). Manufacturing firms account for 31 out of the 78 bankruptcies in the sample.

To study the determinants of bankruptcy, we create a matched sample of firms that did not experience bankruptcy. Specifically, we match each of the 78 firms in our bankruptcy sample to firms in the STATS universe based on three characteristics, namely, size, industry and year. The matched sample must have CSR data available for us to test the hypothesis that CSR matters when it comes to probability of bankruptcy. The closest matched firm to each bankrupt firm in the sample was done by finding firms with the same 2-digit SIC code as the bankrupt firm and also within 15 per cent of the bankrupt firms’ total assets as of the year of the bankruptcy filing. In cases where multiple firms fit these criteria, we chose the firm closest in size to the bankrupt firm as the match. In nine cases, we needed to find firms slightly outside of the 15 per cent total asset differential window. We collected CSR data on the matched firms as of the year prior to the Chapter 11 filing of the bankrupt sample.

We use board characteristics to control for governance factors that may influence a firm’s probability of bankruptcy and bankruptcy emergence. Specifically, we use board size, percentage of insiders on the board, firm age and CEO/chairman duality.

We collect information about the company’s board of directors and governance from MSCI ESG Research. Previous studies on the relationship between corporate bankruptcy and corporate governance factors are mixed. In this study, we focus on most common corporate governance characteristics, namely, board size, board independence, CEO duality and firm age. Gales and Kesner (1994) find that bankrupt companies are more likely to have small boards relative to non-bankrupt firms. However, some studies, such as Jensen (1993), Lipton and Lorsch (1992) and Yermack (1996) suggest that larger boards can be less effective overall than small boards. Darrat et al. (2016) find that a larger board is likely to reduce the probability of bankruptcy in complex firms relative to smaller boards. The presence of independent directors on the board has been shown to be a factor in firm performance. Some studies argue that independent directors are in a better position to monitor the action of the CEO. Further, insider directors may not be able to effectively monitor the action of the CEO because of their position in the firm and the existence of possible inherent contract with the CEO (Weisbach (1988). Hambrick and D’Aveni (1992) find that firms with a large proportion of independent directors are less likely to file for bankruptcy relative to peers with fewer outside directors. Darrat et al. (2016) find that, on average, firms with a large proportion of insiders on the board are more likely to file for bankruptcy relative to firms with more outsiders. More specifically, they find that a greater proportion of inside directors increases the probability of bankruptcy for firms with low research and development (R&D) intensity and reduces the probability of bankruptcy for firms with high R&D intensity.

Daily and Dalton (1994) and Darrat et al. (2016) find that duality (when the CEO also serves as the board chairman) is relatively more common in bankrupt firms than in firms that do not go through bankruptcy.

Table III presents the descriptive statistics of the bankrupt firms in our sample. The average NET_CSR_SCORE was −0.910 with a median score of −1.0. This means that across all of the six categories that our CSR measure covers, the sum of the positive minus the negative scores nets to close to −1. Overall, CSR scores for the bankruptcy firm are leaning slightly toward the “concern” areas as opposed to “strength” but the variance in scores is high with a standard deviation of 2.169. By design, the average binary CSR measure, HI_CSR_SCORE, is about half (59 per cent). The average debt ratio for the 78 bankrupt firms is 95 per cent with a slightly lower median at 91.8 per cent. On average, 26 per cent of the firms in the sample came with a prepackaged/pre-negotiated restructuring plan and 57.8 per cent of the bankrupt firms in our sample eventually emerged from bankruptcy. The time period that a CEO was in the job prior to the bankruptcy was 1,461 days (4 years) on average, and the firm replaced the CEO in about half (48.5 per cent) of the firms.

We were able to match the governance data to many of the firms in the sample (but not all). The average company age at the time of bankruptcy was 38 years. On average, 16 per cent of the boards of the bankruptcy firms are comprised of insiders. The average board size is nine members and most firms in the sample had CEOs who also act as the chairman of the board.

4. Methodology

We compare our matched sample of non-bankrupt firms to the sample of bankrupt firms to see whether there are any univariate differences between the two samples. We report the results of the t-tests in Table IV. Interestingly, the only characteristic that is statistically significantly different between the two matched samples is the debt ratio (LEVERAGE). Unsurprisingly, firms that experienced bankruptcy had a might higher debt ratio on average (95 per cent) relative to firms in the matched sample that did not experience bankruptcy (61 per cent). However, Z-score, a measure of the likelihood of bankruptcy, is surprisingly not statistically significant between the two samples. The p-value on the difference in means is close to 18 per cent, and the non-bankruptcy sample has a higher average Z-score than the bankruptcy sample. Total assets (SIZE), return on assets (ROA) and cash-to-assets (CASH/TA), were not statistically significantly different between samples. As we matched the samples by size, the result that total assets are not statistically different is not surprising. We do see that ROA and CASH/TA are smaller on average for the bankruptcy sample relative to the matched sample. Finally, while the average CSR scores (NET_CSR_SCORE) are higher for the non-bankruptcy matched sample, the differences are not statistically significant. The differences are pronounced, however, the average Net CSR Score for the bankruptcy sample is −0.910 compared to −0.538 for the matched sample.

Also, using the matched sample, we find that firms with higher Net CSR Scores are, from a frequency standpoint, less likely to experience bankruptcy relative to firms with lower Net CSR Scores. Specifically, for firms with Net CSR Scores greater than the median (HI_CSR_SCORE = 1) 39 of them were in the non-bankruptcy matched sample and 31 were in the bankruptcy sample (these results are not reported in a table).

In terms of governance characteristics, the two samples are not statistically significantly different. The percentage of inside directors on the board, the size of the board of directors, CEO/chair duality and company age for the two samples are relatively close. One that is somewhat close to being statistically significantly different is CEO/chair duality; the bankruptcy sample has a higher instance of duality relative to the non-bankruptcy sample and the difference is significant at the 15 per cent level.

We next look at the bankruptcy sample and matched firms in a multivariate setting. This will enable us to test H1, which states that, based on stakeholder theory we should see a negative relationship between CSR and the likelihood of bankruptcy. Our logistic regression model is based on Darrat, et al. (2016), who look at how corporate governance characteristics effect the risk of bankruptcy. Specifically, our model is as follows:

Pr(BANKRUPTCY)=α+β1CSR+β2LEVERAGE+β3SIZE+β4ROA+β5CRISIS+β6CASHTA+β7INDUSTRY+ε

CSR represents the CSR score. Here, we use the binary variable HI_CSR_SCORE (1 indicates higher or stronger CSR; 0 indicates weaker CSR firms) and the continuous measure of CSR (NET_CSR_SCORE). Our hypothesis suggests that this coefficient should be negative as a firm with stronger CSR would be less likely to experience bankruptcy relative to a firm with weak CSR. LEVERAGE is equal to the debt ratio of the firm; the more levered the firm, the higher the likelihood of bankruptcy. The log of total assets measures firm SIZE. Darrat et al. (2016) noted a negative relationship between firm size and bankruptcy probability. We measure profitability with ROA. The more profitable the firm, the lower the likelihood that the firm will experience bankruptcy.

We also include a binary variable CRISIS, which takes the value of 1 if the bankruptcy (and subsequently the matched firm pair) is in the year 2008, 0 otherwise. We propose that firms in the year 2008 were prone to experiencing financing distress more than in other years due to the financial crisis, so we control for this by incorporating the crisis dummy variable into the analysis[3]. We control for liquidity using the measurement CASH/TA, which is cash and short-term assets divided by total assets. Presumably, the more liquid the firm, the lower the probability of bankruptcy. However, this variable is not present for some of our sample firms, and therefore, reduced our sample size.

We also control for industries in each model. Specifically, we use the US Department of Labor’s industry classification breakdown and divide our sample into size industry categories: mining, manufacturing, transportation/communication, retail, finance and services. We run both logistic and probit regressions for each of the models (here, we report the results of the logistic regressions). The dependent variable is equal to 1 if the firm experienced bankruptcy, and 0 otherwise.

In subsequent models, we add several board characteristics as control variables. Specifically, we include the following variables: COMPANY_AGE, INSIDER_PCT, BOARD_SIZE and DUALITY. COMPANY_AGE represents the age of the firm in years. INSIDER_PCT represents the percentage of the board made up of insider directors (directors with past or present affiliation with the firm). The number of directors on the board constitutes BOARD_SIZE. DUALITY is equal to 1 if the chairman of the board is also the firm’s CEO (0 otherwise). As the addition of these variables reduces the overall sample, we run the regressions both with and without these controls. All financial characteristics, controls and the CSR variables are as of one year prior to the bankruptcy announcement. Similar to Dunbar et al. (2016), the use of lagged CSR scores mitigates any potential for reverse causality.

We follow up the analysis of predictors of bankruptcy by including several interaction terms in the model. Specifically, we look at how CSR may have a moderating impact on the effect of leverage, size and Z-score in predicting the likelihood of bankruptcy. Therefore, we include three additional indicator variables to the original logistic regression model presented earlier: CSR × LEVERAGE, CSR × SIZE and CSR × Z-SCORE.

Finally, we turn to an analysis of the firms that did in fact file for bankruptcy. Our research question is whether CSR is a determinant for bankrupt firms to eventually emerge from bankruptcy. H2, in this paper, states that we should see a positive relationship between CSR and the probability that the firm will emerge from bankruptcy.

The specific logistic regression model is as follows:

Pr(Emerge)=α+β1CSR+β2LEVERAGE+β3SIZE+β4ROA+β5PREPACK+β6CEOTENURE+β7CEOTTURNOVER+β8CRISIS+β9INDUSTRY+ε

We again measure CSR performance using the continuous measure (NET_CSR_SCORE) and the binary measure (HI_CSR_SCORE). Denis and Rodgers (2007) find that smaller firms spend less time in Chapter 11 and have better operating performance relative to larger companies. We control for SIZE using the log of total assets. In a prepackaged bankruptcy, the firm files a reorganization plan at the same time as filing for bankruptcy. These types of restructurings tend to emerge from bankruptcy faster and with lower costs relative to bankruptcies filed without a prepackaged/pre-negotiated restructuring plan. We control for this using the variable PREPACK, which is equal to 1 if the firm files an organization plan, at the same time, as filing for bankruptcy. The length of time that a CEO is in office may also influence whether or not the firm emerges from bankruptcy. Longer-tenured CEOs might be able to better navigate through the bankruptcy and restructuring process since older CEO are generally more experienced, which would increase the probability that they would successfully navigate the firm through the bankruptcy process. On the other hand, it might be the case that longer-tenured CEOs, who were at the helm when the firm experienced bankruptcy in the first place, would not be best-suited help the company emerges from bankruptcy. We use the variable CEO_TENURE to account for the number of days the CEO has been in office. Similarly, firms that replace their CEOs during the bankruptcy process may be more likely to emerge from bankruptcy if the replacement CEO is successful at turning the company around. Betker (1995) shows that 91 per cent turnover of CEOs in office two years prior to filing by the time the firm emerges from bankruptcy. Bogan and Sandler (2012) find that management turnover is a determinant of bankruptcy survival. Ayotte and Morrison (2007) find that 70 per cent of CEOs are replaced within two years of a bankruptcy filing. We use the variable CEO_TURNOVER to account for companies that replaced the CEOs after filing for Chapter 11.

In additional iterations of the model, we also incorporate governance variables as controls for bankruptcy emergence. In particular, we include COMPANY_AGE, INSIDER_PCT, BOARD_SIZE and DUALITY. Each of these variables is collected one-year prior to the bankruptcy announcement.

5. Results

We report results of the logistic regressions in Table V. The probability modeled in each analysis is bankruptcy = 1 (so that a positive coefficient would mean a direct relationship with the likelihood of bankruptcy). The coefficient on HI_CSR_SCORE is negative and statistically significant at the 10 per cent level in Model 1. This indicates that weaker CSR scores (below the median) are associated with a higher probability of bankruptcy compared to firms with strong CSR scores. This result is consistent with H1, which states that CSR strength and likelihood of bankruptcy are inversely related. Further, we find that the coefficient on LEVERAGE is positive and statistically significant at the 1 per cent level, adding to the evidence that higher debt ratios are associated with a higher probability of bankruptcy relative to firms with low debt ratios. Interestingly, we find that the coefficient on ROA is positively and statistically significantly related to the probability of bankruptcy but the size of the firm is not related. In terms of industries, the probability of bankruptcy seems to be highest for mining and manufacturing firms, compared to the non-modeled industry (services). In each model, we use year fixed effects[4]. Thus, we find initial support for the stakeholder theory of CSR.

In Model 2, we again use the binary measure of CSR, HI_CSR_SCORE, and find that it is negative and significant at the 5 per cent level. This result is again consistent with the first hypothesis. In this model, we include the variable CASH/TA, which is a measure of liquidity. This reduces the sample from 156 institutions to 139. The control variables in Model 2 are also consistent with that of Models 1 and 3. In Model 3 of Table V, the continuous measure of CSR, NET_CSR_SCORE, is not statistically significantly related to the likelihood of bankruptcy. However, we do note that the coefficient is negative. We cannot use this as evidence to support the first hypothesis, however. Otherwise, the control variables are consistent with Model 1 in that the debt ratio, ROA, mining and manufacturing are all positively associated with the likelihood of bankruptcy.

In the next iteration of regressions, we include the board characteristics. So, Models 4-6 are in all ways similar to Models 1-3 respectively, yet with the addition of COMPANY_AGE, INSIDER_PCT, BOARD_SIZE and DUALITY. The sample size is reduced in each model. We find, consistent with Models 1 and 2, that the binary variable HI_NET_CSR_SCORE is negative and statistically significantly related to the likelihood of bankruptcy. A firm with a strong CSR score is less likely to become bankrupt, all else being equal. The continuous variable NET_CSR_SCORE found in Model 6 is not statistically significantly related to bankruptcy probability (but the coefficient is negative). In Models 4- 6, the variable DUALITY is positive and statistically significant, indicating that firms with the same individual serving as CEO and Chairman are more likely than non-duality firms to become bankrupt are. This is consistent with the bankruptcy literature. For instance, Daily and Dalton (1994) and Darrat et al. (2016) show that there is a positive relationship between CEO power concentration and the probability of bankruptcy. The other governance characteristics are not statistically significantly related to bankruptcy probability.

In all six specifications, we find that the chi-squared test statistics are statistically significant indicating good model fit. It is worth noting that for each model, we also perform a probit analysis. These results (not reported) are consistent with the logistic regression results presented in Table V. In summary of the results in Table V, we find evidence to support H1 – stronger CSR deters bankruptcy – in particular, when CSR is measured using a distinction between “high” and “low.” It may be the case that because CSR scores do not vary substantially by the firm, looking at them as “good” and “bad” is more informative than the continuous measure of CSR performance.

Next, we examine the interaction effects of variables on the likelihood of bankruptcy. In addition to the main effects of CRS variables on the likelihood of bankruptcy, the interaction effects among the two variables could provide better predictions of bankruptcy. Therefore, we test H1a-H1c by including three separate interaction terms. Specifically, we test whether the coefficients of CSR and leverage, CSR and size and CSR and Z-score are significant predictors of the probability of bankruptcy.

The results of the interaction effects are presented in Table VI. In all six models, we see that the relationship between CSR and the probability of bankruptcy is negative and statistically significant. In other words, we continue to see that firms with strong CSR have a lower likelihood of bankruptcy relative to firms with weaker CSR. Further, for Models 1-4, we see that leverage and the probability of bankruptcy is positive and statistically significant at the 1 per cent level. Firms with a higher proportion of debt relative to equity have a higher chance of bankruptcy, all else being equal. However, we do not see that the interaction between CSR and leverage is significant in the model (the coefficient is positive but not statistically significant). Therefore, we cannot conclude that the inverse relationship between CSR and bankruptcy changes with differences in debt ratios (H1a). This is surprising as we expected that at high levels of debt, the relationship between CSR and bankruptcy probability would weaken but it appears that CSR impacts bankruptcy at all debt ratios.

In Models 3 and 4 of Table VI, we see that again, CSR is inversely related to the likelihood of bankruptcy and leverage is positively related to bankruptcy. Size of the firm is not statistically significantly related to bankruptcy probability in and of itself. In Model 4, we also include CASH/TA as an additional control variable in the model. H1b said that we expected that large firms would have a stronger relationship between CSR and bankruptcy probability as larger firms are more apt to disclose more information to stakeholders and are likely monitored more closely relative to small firms. The analysis here shows that the hypothesis is not supported as the interaction between size and CSR score is not statistically significant in Models 3 and 4.

In Models 5 and 6 of Table VI, we look at Z-Score as an indicator of bankruptcy and also interact this variable with CSR to see what impact it has on the relationship between CSR and likelihood of bankruptcy. As Z-score includes a measure of leverage, we take this out of the model predicting bankruptcy. We find that Z-score is negatively related to the likelihood of bankruptcy, holding all else constant. Higher Z-scores are indicative of lower levels of bankruptcy, as expected. Further, we find that the interaction between Z-score and CSR is positive and statistically significant at the 1 per cent level. This result provides support for H1c, which stated that the inverse relationship between CSR and the probability of bankruptcy is stronger for firms with stronger (better) Z-scores than for firms with low (worse) Z-scores. This does appear to be the case both with and without a measure of firm liquidity. All models in Table VI have statistically significant chi-squared test statistics indicating models of a good fit.

In summary of the results in Table V and Table VI, overall we find evidence that firms with strong CSR have a lower likelihood of bankruptcy relative to firms with weaker CSR. When we look at by industry, the probability of bankruptcy seems to be highest for mining and manufacturing firms, compared to other industries. Further, we find that the inverse relationship between CSR and the probability of bankruptcy is stronger for firms with stronger Z-scores than for firms with low Z-scores. This is consistent with the literature on the likelihood of bankruptcy. In terms of internal monitoring, we find that firms with the same individual serving as CEO and chairperson of the board are more likely than non-duality firms to become bankrupt. Duality gives to CEO more power and authority and this leads to a higher probability of bankruptcy. This finding is consistent with Daily and Dalton (1994) and Darrat et al. (2016).

Finally, we report the findings for our analysis of the determinants of bankruptcy emergence in Table VII. The Model 1 reports results without industry controls and the Model 2 controls for industry. The logistic regressions modeling the probability of emerging from bankruptcy fail to show any consistent and significant relationship between CSR and emergence for both models. Specifically, the coefficient HI_CSR_SCORE is negative in both models (−0.699 in Model 1 and −0.676 in Model 2), but the results are not statistically significant. This is contrary to our hypothesis and conjecture where we posit that CSR would have a positive influence on bankruptcy emergence. Compared with our results in Tables V and VI, which showed a negative relationship between CSR and the probability of bankruptcy, these results make it reasonable to consider that CSR does not impact successful emergence from bankruptcy. Kahl (2002) argues that only the worst firms fail to navigate through bankruptcy, so it appears that CSR is not representative of a determinant factor of bankruptcy emergence. However, as expected, a bankruptcy filing accompanied by a prepackaged or pre-negotiated restructuring arrangement is significantly more likely to emerge from bankruptcy relative to a bankruptcy that was not prepackaged/pre-negotiated. The coefficient on PREPACK is positive and significant at the 5 per cent level in all four models of Table VII. None of the other control variables in Models 1 and 2 are statistically significant, including debt ratio, total assets and CEO turnover and CEO tenure.

In Models 3 and 4, we incorporate the governance characteristics as control variables. Again, we do not see a statistically significant relationship between CSR and the likelihood of emergence from bankruptcy. We do see that PREPACK is positive and statistically significantly related to bankruptcy outcome, consistent with Models 1 and 2. We also see that firms with dual CEO/chairman roles are less likely to emerge from bankruptcy than firms with without “duality.” The variable DUALITY is negative and statistically significant in Models 3 and 4. It should be noted that the addition of the governance variables substantially reduces the sample size of the models (43 observations from 69 observations); however, the chi-squared test statistics are statistically significant across all four models in Table VII.

In summary of Table VII, we do not support the conjecture that CSR matters when it comes to bankruptcy emergence. So, while CSR seems to influence whether a company experiences bankruptcy in the first place, having strong CSR does not seem to help (or hurt) a firm once it has filed for Chapter 11. This may provide some support for the altruistic theory of CSR, but more than likely, it is evidence that other variables are more important to a firm once it files for bankruptcy. Stakeholder relationships may not provide much support for a firm that does not have a restructuring plan in place, for instance. Therefore, although we do not support stakeholder theory in this context, these results do not suggest that the CSR stakeholder theory is not valid.

6. Conclusion

Motivated by recent increases in filing for bankruptcy (according to the American Bankruptcy Institute) and increases in awareness of CSR on financial performance, we address the question of whether CSR matters when a firm faces with bankruptcy. Theory suggests that CSR can help a firm by engaging stakeholders and can use that support for financial gain (stakeholder theory). Alternatively, CSR could pose a financial drag on the firm where capital is used for things outside of the firm’s business projects; thus, making the firm worse off than it would be had they not engaged in CSR (agency theory). Else, CSR may be a factor separate from other business decisions, making no difference at all in the outcomes of the firm’s financial performance, but rather is part of the internal structure of the firm – a true altruistic behavior for firms that can afford it.

Here, we use a matched sample approach to study the difference between CSR of firms that do become bankrupt compared to those that do not. Our analysis is comprehensive as we matched a sample of 78 bankrupt firms through the period 2007-2014, which includes the financial crisis, based on year, size and industry. Because of CSR data limitations, our results can only be applied to large firms (as small firms are generally not included in the MSCI database).

Our results provide several contributions to literature. First, our results show that stronger CSR firms are less likely to become bankrupt relative to weaker CSR firms, all else being equal. This result is in line with the stakeholder theory of CSR, as it may be the case that firms with better CSR involvement are able to use that stakeholder support for financial advantage and gain that firms with weaker CSR do not have the stakeholder support available. Second, leverage (debt-ratio) and size variations do not seem to impact the CSR-bankruptcy relationship. Third, Z-score variations seem to affect the CSR-bankruptcy relationship. The inverse relationship between CSR and the probability of bankruptcy is stronger for firms with stronger (better) Z-scores than for firms with low (worse) Z-scores. Fourth, firms with a dual CEO/Chairman tend to have a higher probability of bankruptcy relative to firms with separate individuals serving as CEO and Chairman of the Board. This is the case proper internal mechanism limit managerial entrenchment. Finally, when we look at whether CSR determines bankruptcy outcome (emergence), we find that it does not play a vital role. Once a firm is bankrupt, it is more about the bankruptcy itself; in particular, bankruptcies that are pre-packaged and pre-negotiated that determines whether the firm eventually emerges. This result neither supports stakeholder theory nor does it support agency or pure altruism. We argue that in bankruptcy, variables beyond bankruptcy specifics do not make a difference, ultimately, in terms of emerging from Chapter 11.

Some limitations of these results include the fact that because of data availability, our results can only be applied to large, publicly-traded firms. Future studies could use samples that include small, private firms and firms outside of the USA. We use an aggregated CSR score that is applied equally to all industries; it might be the case that different industries are exposed to different CSR factors, however. Future research can address industry-specific indicators of CSR and their impact on bankruptcy.

The results of this study have important implications for researchers, practitioners and policymakers. From an academic standpoint, this paper fills in a research gap in a busy stream of CSR and bankruptcy literature. It is the first paper to study the link between CSR and bankruptcy. This paper provides some support for the stakeholder theory of CSR and adds to the literature on CSR in this context. Specifically, in determining financial failure, as opposed to success, CSR seems to be important. From a practitioner’s standpoint, this paper provides evidence to show that CSR does matter to a firm in the context of financial performance. For managers deciding whether to invest in CSR activities beyond altruistic reasons can find this study as many others that point to a positive relationship between CSR and financial performance. From policymakers’ standpoint, this paper provides evidence of a benefit to companies for being socially responsible and can enhance any rules or regulations that motivate firms to move in that direction.

Distribution of bankruptcy firms by filling year

Filling year N (%) Cumulative (%)
2007 6 7.69 7.69
2008 11 14.10 21.79
2009 31 39.74 61.54
2010 10 12.82 74.36
2011 8 10.26 84.62
2012 7 8.97 93.54
2013 2 2.56 96.15
2014 3 3.85
Total 78   100

Distribution of bankruptcy firms by industry

Industry N (%) Cumulative (%)
Finance 21 26.96 26.96
Manufacturing 31 39.74 66.67
Mining 7 8.97 75.64
Retail 4 5.13 80.77
Services 5 6.41 87.18
Trans/comm/elec 10 12.82
Total 78   100
Notes:

Table I and II report frequency distributions of the bankruptcy filings in our sample by year and by industry during the period from 2007-2014

Summary statistics (bankruptcy firms)

Variables N Mean Median SD
CSR variables
NET_CSR_SCORE 78 −0.910 −1.000 2.169
HI_NET_CSR_SCORE 78 0.589 1.000 0.495
Firm characteristics
LEVERAGE 78 0.950 0.918 0.499
Z-SCORE 47 0.4969 0.4899 1.816
SIZE 78 20,614.31 2,859.22 86,385.60
Bankruptcy characteristics
PREPACK 78 0.256 0.000 0.439
EMERGED 76 0.578 1.000 0.497
CEO Measures
CEO_TENURE 77 1,460.670 568.00 1975.02
CEO_TURNOVER 70 0.485 0.000 0.503
Board characteristics
COMPANY_AGE 59 38.170 24 37.650
INSIDER_PCT 74 0.162 0.125 0.0829
BOARD_SIZE 74 9.068 9 2.235
DUALITY 61 0.574 1 0.4986
Notes:

Table III reports the summary statistics for the major variables during the sample period of 2007-2014. See the Appendix for the definitions of all the variables

t-tests between the non-bankruptcy matched sample and bankruptcy sample

Variables Non-bankruptcy sample Bankruptcy sample t-value p-value
N Mean SD Minimum Maximum N Mean SD Minimum Maximum
CSR variables
NET_CSR_SCORE 78 −0.538 2.062 −8.000 5.000 78 −0.931 2.169 −7.000 5.000 1.10 0.274
HI_NET_CSR_SCORE 78 0.500 0.503 0 1.00 78 0.397 0.492 0 1.00 1.29 0.202
Firm financial characteristics
LEVERAGE 78 0.613 0.260 0.084 1.206 78 0.950 0.499 0.275 4.529 −5.29*** 0.000***
SIZE 78 11,898.5 31,796.7 228.0 218,328 78 20,614.31 86,385.6 264.4 691,063 −0.84 0.405
ROA 78 −1.120 12.743 −46.754 21.284 78 −0.143 0.187 −0.930 0.108 −0.68 0.500
CASH/TA 78 0.093 0.101 0.000 0.565 78 0.072 0.092 0.000 0.427 1.18 0.238
Z-SCORE 55 2.258 8.6947 −54.1493 19.990 47 0.4969 1.816 −6.088 4.012 1.36 0.176
Board characteristics
COMPANY_AGE 64 38.234 40.748 0 177 59 38.170 37.650 0 157 0.01 0.9927
INSIDER_PCT 72 0.1649 0.0697 0 0.4286 74 0.1619 0.0829 0.0625 0.4286 0.23 0.8169
BOARD_SIZE 72 9.7083 3.8178 5 34 74 9.0676 2.235 3 16 1.24 0.2164
DUALITY 65 0.4462 0.501 0 1 61 0.5738 0.4986 0 1 −1.43 0.1546
Notes:

This table reports the summary statistics for the variables used in the empirical analysis during the sample period of 2007-2014. The sample contains 78 bankruptcy firms and 78 matched non-bankruptcy firms. *** indicate significance based on a difference in means t-test at the 1% levels, respectively. See the Appendix for the definitions of all the variables

Logistic regressions: determinants of bankruptcy

Variable Model 1 Model 2 Model 3 Model 4 Model 5 Model 6
INTERCEPT −5.142 (1.848)*** −5.289 (1.982)*** −4.582 (1.788)** −7.373 (3.018)** −7.352 (3.296)** −6.401 (2.894)**
HI_NET_CSR_SCORE −0.898 (0.478)* −1.058 (0.513)** −1.475 (3.018)** −1.990 (0.895)**
NET_CSR_SCORE −0.091 (0.112) −0.186 (0.167)
LEVERAGE 6.453 (1.215)*** 6.034 (1.283)*** 6.139 (1.163)*** 8.983 (2.037)*** 10.170 (2.487)*** 8.247 (1.901)***
SIZE −0.080 (0.184) −0.073 (0.195) −0.149 (0.177) 0.143 (0.287) 0.231 (0.327) 0.085 (0.279)
ROA 0.053 (0.027)* 0.050 (0.028)* 0.053 (0.028)* 0.076 (0.040)* 0.078 (0.043)* 0.079 (0.041)*
MINING 2.263 (1.204)* 2.392 (1.325)* 2.007 (1.163)* 1.508 (1.748) 0.844 (2.010) 1.111 (1.715)
MANUFACTURING 2.229 (1.091)** 2.471 (1.217)** 1.917 (1.033)* 0.883 (1.668) 0.858 (1.913) 0.522 (1.619)
RETAIL 2.313 (1.430) 2.660 (1.529)* 1.810 (1.375) 0.381 (2.336) 1.168 (2.601) −0.397 (2.161)
(0.152) (0.188) (0.151)
DUALITY 1.127 (0.6424)* 1.750 (0.779)** 1.070 (0.636)*
Year fixed effects Yes Yes Yes Yes Yes Yes
No. of Observations 156 139 156 105 97 105
Chi-square test statistic 67.337*** 54.970*** 64.306*** 65.704*** 64.889*** 62.987***
Notes:

This is the result of logistic regressions during the sample period of 2007-2014 where the DV = 1 where the firm experienced bankruptcy. Insignificant variables not presented in the table include: CASH/TA, Finance and Trans/Comm industry variables, COMPANY_AGE, INSIDER_PCT, BOARD_SIZE. The standard errors are reported in the parentheses; *, ** and *** indicate significance at the 10, 5 and 1% levels, respectively. See the Appendix for the definitions of all variables

Logistic regressions with interactions

Variable Model 1 Model 2 Model 3 Model 4 Model 5 Model 6
INTERCEPT −4.959 (1.833)*** −5.051 (1.970)** −3.829 (1.993)* −3.558 (2.148)* −1.087 (1.920) −0.537 (1.975)
HI_NET_CSR_SCORE −1.579 (0.861)* −1.653 (0.889)* −0.971 (0.487)** −1.167 (0.530)** −2.101 (0.690)*** −2.213 (0.738)***
LEVERAGE 6.138 (1.254)*** 5.797 (1.308)*** 6.647 (1.251)*** 6.302 (1.327)***
Z-SCORE −0.746 (0.237)*** −0.758 (0.238)***
CSRxZ-SCORE 0.671 (0.222)*** 0.694 (0.225)***
ROA 0.052 (0.028)* 0.050 (0.028)* 0.058 (0.028)** 0.057 (0.029)** 0.105 (0.046)** 0.095 (0.044)**
CASH/TA 0.047 (2.586) −0.529 (2.616) −2.669 (2.817)
MINING 2.288 (1.194)* 2.381 (1.302)* 2.298 (1.219)* 2.629 (1.382)* 1.111 (1.309) 1.052 (1.313)
MANUFACTURING 2.258 (1.078)** 2.463 (1.190)** 2.251 (1.114)** 2.708 (1.283)** 2.234 (1.237)* 2.328 (1.242)*
RETAIL 2.292 (1.428) 2.609 (1.515)* 2.339 (1.467) 2.917 (1.603)* 2.304 (1.500) 2.343 (1.511)
Year fixed effects YES YES YES YES YES YES
No. of observations 156 139 156 139 102 98
Chi-square test statistic 68.285*** 55.660*** 70.449*** 59.504*** 40.493*** 41.460***
Notes:

These models use logistic regressions to examine the relationship between CSR and the probability of bankruptcy during the sample period of 2007-2014 where the DV = 1 where the firm experienced bankruptcy. Insignificant variables not presented in the table include: Finance and Trans/Comm industry variables, CSR × LEVERAGE, SIZE, CSR × SIZE. The standard errors are reported in the parentheses. *, ** and *** indicate significance at the 10, 5 and 1% levels, respectively. See the Appendix for the definitions of all variables

Logistic regression: determinants of emerging from bankruptcy

Variable Model 1 Model 2 Model 3 Model 4
HI_NET_CSR_SCORE −0.900 (0.658) −1.196 (0.798) −2.344 (1.525) −6.340 (4.271)
SIZE 0.039 (0.201) 0.567 (0.293)* 0.405 (0.367) 2.898 (1.308)**
PREPACK 1.873 (0.882)** 2.277 (1.033)** 3.036 (1.549)** 7.586 (3.530)**
CEO_TENURE −0.00004 (0.0001) 0.00001 (0.0001) 0.00006 (0.0002) 0.0005 (0.0005)
(0.396) (0.683)
DUALITY −3.702 (1.636)** −7.570 (3.171)**
Year fixed effects Yes Yes Yes Yes
No. of observations 69 69 43 43
Chi-square test statistic 16.685** 25.744** 18.124* 29.769**
Notes:

These models use logistic regressions to examine the relationship between CSR and the probability that the firm will emerge from bankruptcy during the sample period of 2007-2014 where the DV = 1 if the firm emerged from bankruptcy. Insignificant variables not presented isn the table include: all industry indicator variables, LEVERAGE, CEO_TURNOVER, COMPANY_AGE, INSIDER_PCT, BOARD_SIZE. The standard errors are reported in the parentheses. * and ** indicate significance at the 10 and 5% levels, respectively. See the Appendix for the definitions of all variables

Variable definitions

Variable Definitions
BANKRUPTCY Takes the value of 1 if the firm experienced bankruptcy, and 0 otherwise
NET_CSR_SCORE The score for each category (COM, ENV, DIV, EMP and PRO) is calculated as the number of strengths minus the number of concerns in each category by year. Overall net score is calculated as COM_score + DIV_score + EMP_score + ENV_score + HUM_score + PRO_score
HI_CSR_SCORE This is equal to 1 if the Net CSR Score is above the median for the sample and equal to 0 if is below the median
SIZE The log of the total assets
LEVERAGE Book value of debt over book value of assets
ROA Operating income before depreciation, scaled by book value of assets
CASH/TA Cash and short-term assets divided by total assets
CRISIS Takes the value of 1 if the bankruptcy is in year 2008, and 0 otherwise
PREPACK Takes value of 1 if the firm files a reorganization plan at the same time as filing for bankruptcy, and 0 otherwise
EMERGED Takes value of 1 if the firm emerged from bankruptcy, 0 otherwise
CEO_TENURE The length of time that a CEO is in the office (in days)
CEO_TURNOVER Takes value of 1 if the firms replace the CEO during the bankruptcy process, 0 otherwise
COMPANY_AGE The age in years of the company
INSIDE_PCT The percentage of the board comprised of insiders (people who are current or past employees of the company or related to current or past employees)
BOARD_SIZE The number of directors on the board
DUALITY Takes the value of 1 if the CEO is also the Chairman of the Board and 0 if these positions are separate
Z-SCORE A gauge of the likelihood of bankruptcy that incorporates profitability, leverage, liquidity, solvency and activity

Notes

1.

In some cases, however, bankruptcy may be a good option for a firm; particularly, when a business owner has personal assets at risk when a firm is in financial distress. Bankruptcy helps to protect personal assets from creditors. However, in general, bankruptcy is time-consuming and costly for all involved (including creditors).

2.

Human Rights issues include things such as involvement in exceptional human rights initiatives or, on the concern side, supporting controversial regimes and involvement in egregious human rights violations such as killings or physical abuse. It is not a highly-populated category so we do not include it in the analysis here.

3.

Because the financial crisis may not be seen as being limited to 2008, we also run the analyses using a dummy variable = 1 if the bankruptcy is in year 2008 or 2009, and 0 if otherwise. The results of this robustness check are consistent with the results presented in here using the aforementioned definition of the crisis dummy.

4.

For brevity’s sake we did not include several insignificant variables in the tables shown here.

Appendix

Table AI

References

Ahn, S. and Park, D.-J. (2018), “Corporate social responsibility and corporate longevity: the mediating role of social capital and moral legitimacy in Korea”, Journal of Business Ethics, Vol. 150 No. 1, pp. 117-134.

Ayotte, K.M. and Morrison, E.R. (2007), “Creditor control and conflict in chapter 11”, Working paper, Columbia Law School.

Barnea, A. and Rubin, A. (2010), “Corporate social responsibility as a conflict between shareholders”, Journal of Business Ethics, Vol. 97 No. 1, pp. 71-86.

Betker, B.L. (1995), “Managements’ incentives, equity’s bargaining power, and deviation from absolute priority in chapter 11 bankruptcies”, The Journal of Business, Vol. 68 No. 2, pp. 161-183.

Bogan, V.L. and Sandler, C.M. (2012), “Are firms on the right page with chapter 11? An analysis of firm choices that contribute to post-bankruptcy survival”, Applied Economics Letters, Vol. 19 No. 7, pp. 609-613.

Brammer, S.J., Brooks, C. and Pavelin, S. (2006), “Corporate social performance and stock returns: UK evidence from disaggregate measures”, Financial Management, Vol. 35 No. 3, pp. 97-116.

Cajias, M., Fuerst, F. and Bienert, S. (2014), “Can investing in corporate social responsibility lower a company’s cost of Capital?”, Studies in Economics and Finance, Vol. 31 No. 2, pp. 202-222.

Cooper, E.W. and Uzun, H. (2015), “Corporate social responsibility and the cost of debt”, Journal of Accounting and Finance, Vol. 15 No. 8, pp. 11-29.

Daily, C.M. and Dalton, D.R. (1994), “Corporate governance and the bankrupt firm: an empirical assessment”, Strategic Management Journal, Vol. 15 No. 8, pp. 643-654.

Darrat, A.F., Gray, S., Park, J.C. and Wu, Y. (2016), “Corporate governance and bankruptcy risk”, Journal of Accounting, Auditing, and Finance, Vol. 31 No. 2, pp. 163-202.

Dawkins, M.C., Bhattacharya, N. and Smith, B.L. (2007), “Systematic share price fluctuations after bankruptcy filings and the investors who drive them”, Journal of Financial and Quantitative Analysis, Vol. 42 No. 2, pp. 399-419.

Denis, D.K. and Rodgers, K.J. (2007), “Chapter 11: duration, outcome, and post-reorganization performance”, Journal of Financial and Quantitative Analysis, Vol. 42 No. 1, pp. 101-118.

Dayanandan, A., Donker, H. and Nofsinger, J. (2018), “Corporate goodness and profit warnings”, Review of Quantitative Finance and Accounting, Vol. 51 No. 2, pp. 553-573.

Dunbar, C., Li, F. and Shi, Y. (2016), “Corporate social responsibility and CEO risk-taking incentives”, Working Paper.

Eckbo, E.B., Thourburn, K.S. and Wang, W. (2016), “How costly is corporate bankruptcy for the CEO?”, Journal of Financial Economics, Vol. 121 No. 1, pp. 210-229.

Erhemjamts, O., Li, Q. and Venkateswaran, A. (2013), “Corporate social responsibility and its impact on firm’s investment policy, organizational structure, and performance”, Journal of Business Ethics, Vol. 118 No. 2, pp. 395-412.

Erragragui, E. (2018), “Do creditors price firms’ environmental, social and governance risks?”, Research in International Business and Finance, Vol. 45, pp. 197-207.

Fich, E.M. and Slezak, S.L. (2008), “Can corporate governance save distressed firms from bankruptcy? An empirical analysis”, Review of Quantitative Finance and Accounting, Vol. 30 No. 2, pp. 225-251.

Franzen, L.A., Rodgers, K.J. and Simin, T.T. (2007), “Measuring distress risk: the effect of R and D intensity”, The Journal of Finance, Vol. 62 No. 6, pp. 2931-2967.

Freeman, R.E. (1984), Strategic Management: A Stakeholder Approach, Pitman, Boston, MA.

Gales, L.M. and Kesner, I.F. (1994), “An analysis of board director size and composition in bankrupt organizations”, Journal of Business Research, Vol. 30 No. 3, pp. 271-282.

Ghoul, S.E., Guedhami, O., Kwok, C.C.Y. and Mishra, D. (2011), “Does corporate social responsibility affect the cost of Capital?”, Journal of Banking and Finance, Vol. 35 No. 9, pp. 2388-2406.

Hambrick, D. and D’Aveni, R. (1992), “Top team deterioration as part of the downward spiral of large corporate bankruptcies”, Management Science, Vol. 38 No. 10, pp. 1445-1466.

Heron, R.A., Lie, E. and Rodgers, K.J. (2009), “Financial restructuring in fresh-start chapter 11 reorganizations”, Financial Management, Vol. 38 No. 4, pp. 727-745.

Ho, C.Y., McCarthy, P., Yang, Y. and Ye, X. (2013), “Bankruptcy in the pulp and paper industry: market’s reaction and prediction”, Empirical Economics, Vol. 45 No. 3, pp. 205-1232.

Hogan, K., Olson, G. and Sharma, R. (2014), “The role of corporate philanthropy on ratings of corporate responsibility and shareholder return”, Journal of Leadership, Accountability, and Ethics, Vol. 11 No. 3, pp. 108-125.

James, S.D. (2016), “Strategic bankruptcy: a stakeholder management perspective”, Journal of Business Research, Vol. 69 No. 2, pp. 492-499.

Jensen, M. (1993), “The modern industrial revolution, exit and the failure of internal control systems”, The Journal of Finance, Vol. 48 No. 3, pp. 831-880.

Jensen, M. and Meckling, W. (1976), “Theory of the firm: managerial behavior, agency costs, and ownership structure”, Journal of Financial Economics, Vol. 3 No. 4, pp. 305 -360.

Kahl, M. (2002), “Economic distress, financial distress, and dynamic liquidation”, The Journal of Finance, Vol. 57 No. 1, pp. 135-168.

Kwak, W., Shi, Y., Cheh, J.J. and Lee, H. (2005), “Multiple criteria linear programming data mining approach: an application for bankruptcy prediction”, Lecture Notes Computer Science, Vol. 3327, pp. 164-173.

Lipton, M. and Lorsch, J.W. (1992), “A modest proposal for improved corporate governance”, Business Law, Vol. 48 No. 1, pp. 59-77.

Lyandres, E. and Zhdanov, A. (2013), “Investment opportunities and bankruptcy prediction”, Journal of Financial Markets, Vol. 16 No. 3, pp. 439-476.

Mahoney, L. and Roberts, R.W. (2007), “Corporate social performance, financial performance and institutional ownership in canadian firms”, Accounting Forum, Vol. 31 No. 3, pp. 233-253.

Margolis, J.D. and Walsh, J.P. (2003), “Misery loves companies: rethinking social initiatives by business”, Administrative Science Quarterly, Vol. 48 No. 2, pp. 268-305.

Nelling, E. and Webb, E. (2009), “Corporate social responsibility and financial performance: the ‘virtuous cycle’ revisited”, Review of Quantitative Finance and Accounting, Vol. 32 No. 2, pp. 197-209.

Noulas, A.G. and Genimakis, G. (2014), “How do CFOs make Capital structure decisions? A survey of Greek listed companies”, Studies in Economics and Finance, Vol. 31 No. 1, pp. 72-87.

Oikonomou, I., Brooks, C. and Pavelin, S. (2014), “The effects of corporate social performance on the cost of corporate debt and credit ratings”, Financial Review, Vol. 49 No. 1, pp. 49-75.

Parker, S., Peters, G.F. and Turetsky, H.F. (2002), “Corporate governance and corporate failure: a survival analysis”, Corporate Governance: The International Journal of Business in Society, Vol. 2 No. 2, pp. 4-12.

Platt, H. and Platt, M. (2012), “Corporate board attributes and bankruptcy”, Journal of Business Research, Vol. 65 No. 8, pp. 1139-1143.

Reisz, A.S. and Perlich, C. (2007), “A market-based framework for bankruptcy prediction”, Journal of Financial Stability, Vol. 3 No. 2, pp. 85-131.

Rekker, S.A., Benson, K.L. and Faff, R. (2014), “Corporate social responsibility and CEO compensation revisited: do disaggregation, market stress, gender matter?”, Journal of Economics and Business, Vol. 72, pp. 84-103.

Robinson, D., Robinson, M. and Sisneros, C. (2012), “Bankruptcy outcomes: does the board matters?”, Advances in Accounting, Incorporating Advances in International Accounting, Vol. 28, pp. 270-278.

Singhal, R. and Zhu, Y. (2013), “Bankruptcy risk, costs and corporate diversification”, Journal of Banking and Finance, Vol. 37 No. 5, pp. 1475-1489.

Verwijmeren, P. and Derwall, J. (2010), “Employee well-being, firm leverage and bankruptcy risk”, Journal of Banking and Finance, Vol. 34 No. 5, pp. 956-964.

Waddock, S.A. and Graves, S.B. (1997), “The corporate social performance-financial performance link”, Strategic Management Journal, Vol. 18 No. 4, pp. 303-319.

Weisbach, M. (1988), “Outside directors and CEO turnover”, Journal of Financial Economics, Vol. 20, pp. 431-460.

Wu, M.W. and Shen, C.H. (2013), “Corporate social responsibility in the banking industry: motives and financial performance”, Journal of Banking and Finance, Vol. 37 No. 9, pp. 3529-3547.

Yermack, D. (1996), “Higher market valuation of companies with a small board of directors”, Journal of Financial Economics, Vol. 40 No. 2, pp. 185-211.

Acknowledgements

The authors gratefully acknowledge and thank the editor and two anonymous referees for their helpful comments and suggestions.

Corresponding author

Elizabeth Cooper can be contacted at: cooper@lasalle.edu