The Effects of Green- and Brownwashing Strategies on the
Continuity of a Company: Predicting Bankruptcy
June 24, 2019 University of Groningen Faculty of Economics and Business MSc BA Strategic Innovation Management
ABSTRACT
Through the combined efforts of agency- and impression management (IM) theory, this quantitative empirical study explores the novel relationship between green- and brownwashing and the probability of bankruptcy for firms with financial difficulties. The agency relationship between shareholder and managers further characterizes the research context. Thereby, this research responds to the scarcity of related literature, and ambiguity on its effects regarding critical firm survival. A cross-sectional multi-industry sample of bankrupt, and matched non-bankrupt US listed companies between 2002-2018 is studied via logistic regression analyses. Whereas a positive, yet insignificant effect of greenwashing is concluded, the likelihood of bankruptcy is significantly negatively affected by brownwashing. Hence, companies that engage in brownwashing as opposed to greenwashing are less likely confronted with bankruptcy. This outcome therefore proves brownwashing to be a viable managerial IM tactic, and potential bonding mechanism to deceptively align to shareholders’ interests. Furthermore, the saliency of shareholders is accentuated, and improved strategic decision-making on CSR communication by executives, shareholders and alternative stakeholders is enabled.
I.
INTRODUCTION
Firms face birth and prosperity as well as stagnation and default. Hence, they represent a lifetime in which not only profitability and cost-reduction define success. In fact, the continuity of a company is what actually drives businesses. Although the latter is an everlasting subject to study, strongly enlarged environmental and societal pressure incentivized companies to become more responsible and transparent throughout the last decade. As a consequence, both the corporate and academic environment noticed the expanded contemporary urge to improve corporate social responsibility (CSR) and its transparency towards various engaged stakeholders. To ensure the aforementioned continuity, the relevance of strategic CSR information disclosure decisions thereby prospered.
To establish solid arguments, impression management (IM) theory is applied and further elaborated with agency theory to investigate the effects of green- and brownwashing on the probability of bankruptcy. In fact, Eisenhardt (1989) encouraged future research to prioritize the examination of organizational behaviour related subjects from an IM-agency theory perspective to further explore and develop both theories. The recombination of multiple theories is actually advocated by prior studies (e.g. Eisenhardt, 1989; Watson et al., 2002). This study extends both theories by means of its investigation on the consequences of managerial IM tactics, as green- and brownwashing, for potential future bankruptcy under the conditions of financial difficulties. Additionally, the role of asymmetric information and conflicting interests within the critical agency relationship between managers and shareholders as a trigger to engage in green- or brownwashing is particularly important too. The seemingly aligned or misaligned interests between managers, shareholders and non-shareholder stakeholders due to green- and brownwashing, and their consequential degree of support stands out. In fact, this study emphasizes the prioritization of shareholders for both managers and corporations. An additional insightful theoretical extension is the proven viability of brownwashing, in contrast to greenwashing, as a managerial IM tactic and bonding mechanism to prevent bankruptcy for firms facing financial difficulties via deceptively aligning to shareholders’ interests. Subsequently, the contemporary literature on both green- and brownwashing is further developed by virtue of the investigated novel relationship with the likelihood of bankruptcy; equivalently for bankruptcy prediction literature by means of green- and brownwashing. In addressing this disruptive topic the following research questions will be explored: RQ1: Do green- and brownwashing strategies significantly affect the likelihood of bankruptcy for firms facing financial difficulties?
RQ2: Which companies are more likely to go bankrupt when they are under financial difficulties: greenwashing or brownwashing companies?
Especially managers of firms with financial difficulties can learn about the impact of green- and brownwashing on the probability of bankruptcy, as well as the salience of shareholders, and adapt their CSR disclosure strategies accordingly. Furthermore, this study allows for improved decision-making by various stakeholders as customers and investors. For instance, knowledge on the likelihood of facing bankruptcy is essential in the evaluation of whether to, and how much to, invest in a business. Additional awareness is raised on the underlying drivers and consequences of green- and brownwashing by top executives, as well as their ability to shape shareholder perceptions on CSR.
II.
LITERATURE REVIEW AND THEORY DEVELOPMENT
Given that firms with financial difficulties predicate intensified shareholder scrutiny- and pressure above all (Kim & Lyon, 2015), this study explores the principal-agent relationship between respectively shareholders and corporate managers. This research context characterized by the described agency relationship within financially constrained firms allows for the valid comparison and analysis of green- and brownwashing. The accentuated saliency of shareholders for corporations facing financial constraints is emphasized by the stakeholder salience theory prioritization based on power, legitimacy and urgency (Mitchell et al., 1997). The agency relationship accordingly asserts shareholders (i.e. principal) empower corporate management (i.e. agent) with the transfer of decision-making authority to execute appointed tasks (Eisenhardt, 1989; Jensen & Meckling, 1976). The conflicting interests that arise from this agency relationship for financially troubled businesses, combined with the embedded outcome uncertainty and asymmetric information determine incentives for both green- and brownwashing and reflect the relevance of this specified research domain for bankruptcy. Note that previous related studies commonly considered ultimate company failure as bankruptcy (Bellovary et al., 2007), defined as “the inability of a firm to pay its financial obligations as they mature” (Beaver, 1966, p. 71). Yet this might be entitled to financial distress instead (Gilson, 1989), emphasizing the interchangeability of definitions for unsuccessful corporations (Altman, 1993). Hence to formally clarify bankruptcy, “firms are considered bankrupt when a petition is filed under either Chapter 11 or Chapter 7 of the U.S. Bankruptcy Code” (Gilson, 1989, p. 243), and an official declaration of bankruptcy is received by the Federal District Court (Altman, 1993); which still holds. Whereas Chapter 11 entails the reorganization and replacement of debt and assets, Chapter 7 argues corporate liquidation.
Green- and brownwashing
Whilst a study by Lyon and Montgomery (2015, p. 226) emphasized the popularity and impreciseness of the term, they proposed the following most common definition of greenwashing: “communication that misleads people into adopting overly positive beliefs about an organization’s environmental performance, practices or products”. As long as the essence of misleading communication remains, greenwashing is equivalently applicable to overreported CSR performance. Nevertheless, firms can engage in alternative organizational information disclosure strategies too. Although prior research distinguished between exaggerating performance (i.e. greenwashing), full disclosure and non-disclosure strategies, Kim and Lyon (2015) recently extended it with ‘brownwashing’, defined as downplaying the disclosure of environmental achievements, as well as for social and governance related outcomes. Although the aforementioned benefits of CSR might motivate greenwashing, additional drivers and consequences of both green- and brownwashing are identified by foregoing literature.
According to Kim and Lyon (2015), growing businesses are for instance more likely to engage in greenwashing due to increased expectations, and their need for high reputation and regulatory support. Supplementary, Delmas and Burbano (2011) concluded that a lack of clarity about punishment regulations on greenwashing, limited transparency on environmental firm performance information, firm characteristics as a large size and strong brand recognition, as well as external market pressure could all drive greenwashing. With regard to the rationale of brownwashing, investor pressure could cause firms to conceal CSR related costs due to their high-cost perception (Ullmann, 1985). In other words, Kim and Lyon (2015, p. 708) concluded that “shareholders may respond negatively to a firm’s environmentally friendly practices” as this tends to be expensive. Furthermore, as reputation is a trivial strategic resource, “the threat of environmental activism may drive firms to carefully consider how to communicate their positive environmental activities” (Carlos & Lewis, 2018, p. 158). Whilst exposure to reputational threats discourages greenwashing, and possibly even restrains honest CSR communication, ‘strategic silence’ and brownwashing becomes more enticing by virtue of hypocrisy avoidance (Carlos & Lewis, 2018).
In fact, since “today, deliberate disclosure of false information is rare, due to shareholders, governments, nongovernmental organizations (NGO’s) and social activists being able to provide close monitoring” (Testa et al., 2017), green- or brownwashing might be recognized. Accordingly, various stakeholders could reasonably punish firms in terms of support or trustworthiness for example. Given the indicated potential negative consequences of green- and brownwashing by prior studies, bankruptcy can be a realistic outcome. Nonetheless, managers of firms with financial difficulties may still purposefully engage in green- or brownwashing as the forthcoming theory development section will justify. This study thereby not only explores the effects of green- and brownwashing on bankruptcy, but extends related literature on both its drivers and consequences.
As a widely dispersed array of applied conditional boundaries predicate the differing scope of the aforementioned motivators and effects of green- and brownwashing discussed in previous studies, diverse theories ground different underlying arguments too. For instance, economic theory would argue that credible proof is required to persuade sceptical stakeholders and let them abandon their worse assumptions and ingrained scepticism (Lyon & Montgomery, 2015; Milgrom, 1981;). The inability to deliver credible evidence of magnified CSR performances via greenwashing could result in discovery, causing descended performance and an increased probability of bankruptcy. Au contraire, despite information disclosure theory argues “investors view voluntary disclosures as credible information”, reducing information risk and acquisitions costs as well (Healy & Palepu, 2001, p. 426), the augmented urge of CSR and transparency deteriorates the habitual disclosure credibility of stakeholders. Hence, green- and brownwashing could harm firms, particularly when confronted with financial problems due to further declined disclosure credibility. Notwithstanding, the combined efforts of agency- and impression management (IM) theory will actually ground the upcoming arguments in this study. In contrast to the previously mentioned theories, the novel recombination of agency- and IM theory allows the full, instead of partial, and solid development of a strong underlying logic for all hypotheses in the given research context. Additionally, it acknowledges incentives to engage in both green- and brownwashing strategies as well.
Theory
this symbolic signalling of CSR policies tends to form images and perceptions by individual constituents on for instance CSR or financial conditions; which collectively establish the corporate reputation. Thus, managers and organizations that engage in green- or brownwashing as IM tactics are incentivized by signalling a preferred appearance, and thereby manipulate and shape stakeholder perceptions of CSR effort and performance to capture potential benefits (Highhouse et al., 2009). Agency theory accordingly characterizes the relationship in which green- and brownwashing are involved. Given the agency problem of conflicting interests between shareholders and managers of firms facing financial difficulties, and the inability of shareholders to precisely monitor and verify potential (in)appropriate behaviour of managers due to the presence of asymmetric information (i.e. the agent has more information than the principal), IM via green- and brownwashing tend to be viable (Eisenhardt, 1989; Jensen & Meckling, 1976; Westphal & Graebner, 2010). Both can therefore be used to deceive shareholders’ impressions on CSR performance and effort.
greenwashing is further triggered as the protection of managers against shareholders’ displeasure via the support of non-shareholder stakeholders particularly stands out in a controlled environment with strong internal governance mechanisms (Surroca & Tribó, 2008). Furthermore, since rationally firms facing financial troubles are more likely to go bankrupt, there is not a lot at stake. This discourages engagement in hypocrisy avoiding behaviour due to reputational risks (Carlos & Lewis, 2018), and stimulates the thought that it might be worthy to engage in greenwashing and face the risk of being identified as an untrustworthy greenwashing company (Hawn & Ioannou, 2016). So, risk-seeking behaviour (i.e. greenwashing) might occur due to the presence of upward potential and a seemingly ‘fixed’ bottom-line of bankruptcy as well. Whether an organization can reap the benefits of greenwashing remains uncertain, but certainly depends on the potential discovery by stakeholders.
However, incentives to implement brownwashing as an IM tactic exist too. Since shareholders pursue to maximize profits, significant effort regarding potentially costly CSR is disfavoured, particularly when organizations encounter financial difficulties (Kim & Lyon, 2015). Given these circumstances, shareholder pressure could necessitate firms to disguise CSR related costs, and hence engage in brownwashing (Ullmann, 1985; Kim & Lyon, 2015). With the use of brownwashing, corporate management aims to signal aligned interests towards shareholders regarding a cost-effective focus rather than expensive CSR. The self-interest for managers is embedded in their risk-averse behaviour (Eisenhardt, 1989). To ensure their current job, corporate managers strive to comply to the interests of shareholders by engaging in brownwashing. Hence, the increased risk of losing a job exhibits an implicitly ingrained governance mechanism to seemingly align interests when an organization is confronted with financial troubles from a shareholder perspective, whilst in reality brownwashing solely configures this deceptive perception. Other than greenwashing, managers that engage in brownwashing thereby assume individualistic interests as job security and self-esteem are better accomplished by signalling adherence to shareholders’ interests instead of non-shareholder stakeholders’ interests. Hypotheses
Consequently, corporate transparency and particularly shareholder scrutiny enlarges (Healy & Palepu, 2001; Kim & Lyon, 2015; Watson et al., 2002). In addition to the contemporary relevance of CSR and information disclosure, this incremented shareholder scrutiny increases the likelihood greenwashing practices by self-interested managers are discovered. Shareholder recognition of greenwashing thereby lowers trustworthiness and support of shareholders in their relationship with corporate management, whilst this is fundamental to ensure firm survival. Furthermore, greater agency costs arise under these circumstances as the discovery of greenwashing requests alternative governance measures to facilitate an improved ability to cope with the asymmetric information and conflicting interests between managers and shareholders. For firms with financial difficulties, expanded agency costs could become detrimental. Therefore, in addition to declined shareholder support for unnoticeable greenwashing, identified greenwashing decreases shareholder trustworthiness and further increases agency costs. Thus, in either recognition or non-recognition of greenwashing, it is expected that:
H1: The probability of bankruptcy increases as a consequence of greenwashing under the circumstances of financial difficulties.
Concerning brownwashing, initial intensive shareholder pressure and scrutiny causes managers to conceal CSR related costs to signal cost-effectiveness and aligned interests. Hence, by deceptively downgrading CSR efforts and performance, managers signal the prioritization of shareholders’ interests. As a consequence, managers are expected to obtain stronger shareholder support and involvement. Since interests seemingly align due to brownwashing, shareholders’ need to closely monitor managers becomes redundant too. This both reduces the likelihood brownwashing is recognized, and the agency costs involved. Given the subordinated need of shareholders to monitor managers, brownwashing could actually function as a managerial bonding mechanism to dishonestly ensure non-damaging behaviour to shareholders (Jensen & Meckling, 1976). Due to the expected decline of agency costs, and strengthened shareholder support and involvement, the ability to prevent bankruptcy tends to improve. Therefore, it is hypothesized that:
H2: The probability of bankruptcy decreases as a consequence of brownwashing under the circumstances of financial difficulties.
Given the discussed hypotheses, this research thereby expects the following:
H3: Organizations that engage in greenwashing are more likely to go bankrupt compared to brownwashing under the conditions of financial difficulties.
- -
+
Figure 1: Conceptual model
CSR Information Disclosure Strategies
Greenwashing Brownwashing Control Variables Net-income to assets Cash-flow to debt EBITDA to liabilities
Working capital to assets
Market capitalization to debt
Probability of Bankruptcy
III.
METHODOLOGY
This quantitative empirical study utilizes yearly secondary cross-sectional data from a multi-industry sample of solely US listed firms in the 2002-2018 time frame. Since only businesses facing financial difficulties are investigated, the specificity of the data prevents potentially biased results; yet, at the cost of generalizability of the outcomes. On the other hand, given the multi-industry sample, outcomes may deviate across industries. Note that within the rather broad time frame, company specific time periods of five years are obtained and investigated, including the final year of bankruptcy. Furthermore, despite the distance from the source, secondary data is objective and hence reasonably reliable as well. Still this study is limited to a rather small sample size caused by the absence of simultaneously available CSR and bankruptcy data. This may generate biased outcomes. Although it is important to be aware of this deficiency, the analyses still fulfil the explorative research intent to establish and discover the relationship between green- and brownwashing and the probability of bankruptcy.
Sample and data collection
As CSR data is the primary requirement of this research, the university’s Eikon database is first used to retrieve an initial sample of all roughly 2500 US firms available within ASSET4. Secondly, using the Wharton Research Data Services (WRDS), it is verified that around 300 firms are confronted with financial difficulties based on the Standard and Poor’s (S&P) domestic long term issue credit ratings available in Compustat. Whereas all ratings lower than BB maintain speculative characteristics, this study presumes all businesses with credit ratings lower or equal to B+ to be confronted with financial troubles; following the S&P ratings definitions (S&P Global Ratings, 2018). To ensure a reliable matching procedure, bankrupt firms are thereby not only matched to non-bankrupt firms on the basis of industry and time period (Beaver et al., 2005), but the presence of financial difficulties too.
Subsequently, whilst CSR information is gathered from ASSET4, unavailable necessary data on bankrupt firms caused a reduction from 58 to 19 cases of bankruptcy. As a consequence, industry averages based on the 3-digit NAICS and identical years are constructed from all roughly 2500 available US corporations within ASSET4, and function as substitutes to establish a complete modified sample of 116 companies. Moreover, financial- and remaining company specific data is particularly retrieved from Compustat and WRDS Financial Ratio Suite. Further complementary databases used are CRSP and Yahoo Finance for market information, and Orbis. As a final procedure to complete the data, 10-k annual report filings are attained via the U.S. Securities and Exchange Commission (SEC) or corporate websites, and researched accordingly.
Core measures
The dependent variable in this study is a dichotomous measure of one for bankruptcy and zero for non-bankruptcy. To identify bankruptcy, all D till CC credit rated firms are manually inspected for filings under Chapter 7 or 11 of the U.S. Bankruptcy Code at the SEC using the EDGAR company search engine. Further authentication of bankruptcies occurred via the UCLA-LoPucki Bankruptcy Research Database (BRD). Additionally, status measures within ASSET4, Orbis and especially Compustat allowed to classify bankruptcies from the entire initial sample too, which are confirmed via desk research and 8-k report filings at the SEC.
To measure green- and brownwashing, the main explanatory variables of interest, Hawn and Ioannou’s (2016) internally consistent measures for internal- and external CSR actions will be offset against each other. As external CSR concentrates on communication and disclosure (e.g. reporting), internal CSR maintains an in-house orientation on operations and policies instead. Based on the pre-selected underlying variables of Hawn and Ioannou (2016), reliable measurement of internal- and external CSR is ensured. Accordingly, the variables demonstrated in appendix B provide guidance to ultimately determine green- and brownwashing. The raw datapoints of these variables enable the calculation of company-specific scores between [0,1] for internal- and external CSR; averaged by the total number of available data inputs per company. Yet, it is the difference that indicates green- or brownwashing. Given the strong affiliation with the CSR gap instituted by Hawn and Ioannou (2016), green- and brownwashing are measured in the following manner:
Greenwashing: |𝐸𝑥𝑡𝑒𝑟𝑛𝑎𝑙 𝐶𝑆𝑅𝑡− 𝐼𝑛𝑡𝑒𝑟𝑛𝑎𝑙 𝐶𝑆𝑅𝑡−1| → 𝐸𝑥𝑡𝑒𝑟𝑛𝑎𝑙 𝐶𝑆𝑅𝑡 > 𝐼𝑛𝑡𝑒𝑟𝑛𝑎𝑙 𝐶𝑆𝑅𝑡−1
Brownwashing: |𝐸𝑥𝑡𝑒𝑟𝑛𝑎𝑙 𝐶𝑆𝑅𝑡− 𝐼𝑛𝑡𝑒𝑟𝑛𝑎𝑙 𝐶𝑆𝑅𝑡−1| → 𝐸𝑥𝑡𝑒𝑟𝑛𝑎𝑙 𝐶𝑆𝑅𝑡 < 𝐼𝑛𝑡𝑒𝑟𝑛𝑎𝑙 𝐶𝑆𝑅𝑡−1
Thereby, following the aforementioned computation, only an intensity measure for brownwashing is realized. Given its inability to assess and compare green- and brownwashing, the initial measurement is therefore centred. To enable this procedure, demeaned values of internal- and external CSR formed on the respective year-based sample means are used. To validly indicate green- or brownwashing, respectively the third (Q3) and first quartile (Q1) of the distribution of the difference between current year’s external CSR and prior year’s internal CSR serve as benchmarks. Subsequently, the indicator variables for green- and brownwashing are one when the criteria below are satisfied.
Greenwashing: 𝐸𝑥𝑡𝑒𝑟𝑛𝑎𝑙 𝐶𝑆𝑅𝑡− 𝐼𝑛𝑡𝑒𝑟𝑛𝑎𝑙 𝐶𝑆𝑅𝑡−1 > 𝑄3
Brownwashing: 𝐸𝑥𝑡𝑒𝑟𝑛𝑎𝑙 𝐶𝑆𝑅𝑡− 𝐼𝑛𝑡𝑒𝑟𝑛𝑎𝑙 𝐶𝑆𝑅𝑡−1 < 𝑄1 Control variables
Furthermore, a set of relevant control variables are included to phase out alternative explanations for bankruptcy, and increase the validity of results. To predict bankruptcy, Beaver (1966) formerly initiated accounting-based prediction using financial ratios alone. In line with Beaver’s proposed usage of accounting measures to predict bankruptcy, Altman (1968) was the first to construct a discriminant multivariate prediction model. The benchmarking literature of Altman (1968, 1993) additionally introduced the critical Z-score based on five predictive accounting ratios which allows a concrete classification of future (non-) bankruptcy. Despite traditional bankruptcy models solely included accounting ratios, prediction accuracy tend to improve through the inclusion of market variables (Shumway, 2001). Due to the long history of using financial ratios in predicting bankruptcy, Beaver et al. (2005) reassessed their validity and found distinct robustness overtime. In fact, all preceding bankruptcy prediction literature concentrates on at least three aspects, “profitability, cash flow generation, and leverage” (Beaver et al., 2005, p. 95). Regardless of a small reduction in the predictive power of financial ratios, a strong combined predictive ability of both financial and market measures exists, caused by the enhanced prediction power of market measures (Beaver et al., 2005). Given the fact accounting ratios are certainly reflected by market information as well, caution on potential partial substitutive relationships is notified. Still, market-based measures provide supplementary insights that financial figures cannot grant on for instance competitiveness and market confidence.
The net income to total assets ratio (NITA) is a prevailing profitability measure used in firm-level research, also defined as ROA (return-on-assets). In fact, this ratio is most common in bankruptcy prediction studies (Bellovary et al., 2007), and second-best in predicting bankruptcy according to Beaver (1966). Although profitability reflects overall operational results and firm healthiness, it indicates the degree of payment capacity as well (Beaver, 1966; Beaver et al., 2005). Hence, to control for bankruptcy caused by drastic operational results and overall performance, the result-oriented net-income ratio is included.
Secondly, the asset turnover ratio is incorporated to account for the asset productivity, and the ability of corporate managers to deal with competitive conditions (Altman, 1968). It is measured by dividing sales to the total assets. Interestingly, similar to Altman (1968, 1993), the ratio is insignificant on a univariate basis, yet significant in combination with other variables. The expected relationship with bankruptcy is negative as augmented sales and asset efficiency rationally decrease the odds of bankruptcy.
Concerning the cash-flow to total debt ratio (CFTD), previous research accentuated its worthiness for predicting bankruptcy. In fact, it is concluded that “the ability to predict failure is strongest in the cash-flow to debt ratio” (Beaver, 1966, p.85). Note that solely cash-cash-flows from operations define this ratio. Given the fact cash-flows from operations of failed firms are subordinated (Beaver, 1966; Beaver et al., 2005), strongly diminishing cash-flows tend to stimulated potential default (e.g. Gilson, 1989). The ratio simultaneously measures the cash generating ability of firm operations, and the coverage of debt. Thereby, it encloses important alternative explanations for bankruptcy.
Additionally, the working capital to total assets ratio (WCTA) eliminates the possibility bankruptcy is triggered by illiquidity. The working capital of a firm is defined as the difference between its current assets and current liabilities. Whereas Beaver (1966) argued future bankrupt companies to own less liquid assets, Altman (1968) found that amongst various liquidity ratios the working capital to total assets ratio is most valuable. Ultimately, decreased liquidity due to descended performance and consequential operating losses cause an inability to fulfil short-term financial obligations, and an increased likelihood of bankruptcy.
Prior excess market returns are measured by the difference between annual company stock- and value-weighted index returns (Beaver et al., 2005; Shumway, 2001). Accordingly, firm stock returns are assessed by the change in market capitalization, whereas for the value-weighted index returns the combination of NASDAQ, AMEX and NYSE returns are utilized since firms in the sample are mainly listed on these indices. Other than financial performance measures, prior excess market returns benchmarks firm performance to the market. Thereby, it measures relative performance, and market confidence, which entails the competitive atmosphere as well. Significant underperformance could cause an inability to compete, with potential negative consequences for the continuity of a firm. Despite the fact including excess market return as a control variable reduces the significance of NITA, ASSETTURNOVER and CFTD, its strong significance with bankruptcy and the unnecessity to obtain parsimonious models approves this decision.
Despite the fact previous research concluded that predictive ability resides within leverage ratios as debt-to-assets and liabilities-debt-to-assets (Beaver, 1966; Beaver et al., 2005; Gilson, 1989), both leverage ratios are excluded due to strong multicollinearity concerns. Nevertheless, although the financial structure of companies is not explicitly measured by the control variables, the model still accounts for the solvency and consequential financial risk via the market capitalization to total debt ratio, and for debt coverage with the cash-flow to total debt ratio. In a similar manner, whereas prior studies controlled for firm size (and market valuation) via the logarithm of market capitalization (Beaver et al., 2005; Shumway, 2001), it is ingrained in other control variables. For instance, firm size characteristics are part of the working capital to total assets ratio according to Altman (1968), and additionally embedded in the remaining market capitalization ratio.
Preliminary empirical analysis
Although the independent variables are assumed to correspond linearly to the odds ratios, their relationship with bankruptcy is not required to be linear for logit models. Thereby, logistic regression models are applied, which are equivalent to generalized linear models with a binomial (i.e. Bernoulli) distribution and logit link function. The baseline control variable model is therefore defined as follows: (i) 𝑃(𝐵𝐴𝑁𝐾𝑅𝑈𝑃𝑇 = 1) = 𝐹(𝛽0+ 𝛽1𝑁𝐼𝑇𝐴 + 𝛽2𝐴𝑆𝑆𝐸𝑇𝑇𝑈𝑅𝑁𝑂𝑉𝐸𝑅 + 𝛽3𝐶𝐹𝑇𝐷 + 𝛽4𝑊𝐶𝑇𝐴 +
𝛽5𝑀𝐴𝑅𝐾𝐸𝑇𝐶𝐴𝑃𝑇𝐷 + 𝛽6𝐸𝑋𝐶𝐸𝑆𝑆𝑀𝐾𝑇_𝑅𝐸𝑇𝑈𝑅𝑁)
Whereas 𝑃 declares the probability, 𝛽0 is the constant, 𝛽𝑖 the regression coefficients, and 𝐹(𝑧) = 𝑒𝑧 (1+𝑒𝑧)
denotes the cumulative logistic distribution. Yet, model modifications are necessary to investigate the effects of greenwashing (H1) and brownwashing (H2) on the probability of bankruptcy, and their comparison (H3). Since potential multicollinearity with green- or brownwashing is ignored in the model above, the final logistic regression models below account for these criteria. As separate inclusion of green- or brownwashing to the baseline model decreases the significance, various alternative specifications are considered. Ultimately, ASSETTURNOVER is replaced with the EBITDA to total liabilities ratio (EBITDATL), measured using the earnings before interest, taxes, depreciation and amortization (EBITDA). Although this ratio tends to measure the equivalent of the cash-flow to total debt (Beaver, 1966; Beaver et al., 2005), and a sales-based efficiency ratio is removed, it yields significant results for all variables included. Additionally, given the matching criteria of this study, the crucial difference between bankrupt and financially constrained non-bankrupt companies might not be made at a sales level (i.e. ASSETTURNOVER), but rather after the subtraction of costs. Hence, more towards profitability and performance instead. To investigate and compare the effects of green- and brownwashing on the likelihood of bankruptcy, whilst both significantly correlate, two comparable separate logistic regression models are constituted.
(ii) 𝑃(𝐵𝐴𝑁𝐾𝑅𝑈𝑃𝑇 = 1) = 𝐹(𝛽0+ 𝛽1𝐺𝑅𝐸𝐸𝑁𝑊𝐴𝑆𝐻𝐼𝑁𝐺 + 𝛽2𝑁𝐼𝑇𝐴 + 𝛽3𝐶𝐹𝑇𝐷 + 𝛽4𝐸𝐵𝐼𝑇𝐷𝐴𝑇𝐿 +
𝛽5𝑊𝐶𝑇𝐴 + 𝛽6𝑀𝐴𝑅𝐾𝐸𝑇𝐶𝐴𝑃𝑇𝐷 + 𝛽7𝐸𝑋𝐶𝐸𝑆𝑆𝑀𝐾𝑇_𝑅𝐸𝑇𝑈𝑅𝑁)
(iii) 𝑃(𝐵𝐴𝑁𝐾𝑅𝑈𝑃𝑇 = 1) = 𝐹(𝛽0+ 𝛽1𝐵𝑅𝑂𝑊𝑁𝑊𝐴𝑆𝐻𝐼𝑁𝐺𝑅 + 𝛽2𝑁𝐼𝑇𝐴 + 𝛽3𝐶𝐹𝑇𝐷 +
𝛽4𝐸𝐵𝐼𝑇𝐷𝐴𝑇𝐿 + 𝛽5𝑊𝐶𝑇𝐴 + 𝛽6𝑀𝐴𝑅𝐾𝐸𝑇𝐶𝐴𝑃𝑇𝐷 + 𝛽7𝐸𝑋𝐶𝐸𝑆𝑆𝑀𝐾𝑇_𝑅𝐸𝑇𝑈𝑅𝑁)
Finally, to control for potential time-dependent external explanatory factors, dummy variables measuring the year of bankruptcy were initially generated. However, due to perfect multicollinearity all are omitted in the regression, and no further actions can be undertaken. Although each bankrupt firm is matched to a non-bankrupt peer from the same year, further clarification on the distribution of bankruptcies overtime remains insightful. Given the distribution presented in figure 2, a significant amount of bankruptcies occurred in 2009. This is reasonably caused by the global financial crisis from 2008, which led to the destruction of companies in its aftermath. In addition, the number of post-crisis bankruptcies slightly exceed the amount of pre-crisis cases in the sample. Since these deviations are neglectable, solely the external year-dependent impact of the crisis in 2009 is a considerable explanatory factor for bankruptcy. Concerning the industry background of all bankrupt companies from the sample, respectively the manufacturing and quarrying, oil, and gas extraction sector are best represented (figure 3a). More specifically, the individual subsectors in which most of the firms operate are the transportation equipment manufacturing, and mining (support) activities (figure 3b). Yet, the large share of firms devoted to “other” portrays the vast variety of industries included.
Figure 3a: Industry sector distribution
IV.
RESULTS
In advance of the regression analyses, table 1 and 2 respectively represent the general descriptive statistics and correlation matrix. Regarding the variables of interest, the dichotomous measure for bankruptcy is expectedly equally distributed, whereas green- and brownwashing retain the forced quartile-based distribution. Interestingly, on average, firms tend to underperform compared to the market due to problems concerning profitability and cash-flow generation. Given the fact solely firms with financial difficulties comprise the sample, a rational justification exists. Furthermore, the correlation matrix in table 2 suggests that similar to the hypotheses, a positive interaction between greenwashing and the probability of bankruptcy exists, whilst brownwashing maintains a strong negative correlation. With regard to the control variables, all declare the expected correlation signs. Moreover, multicollinearity concerns emerge from excess market returns due to its particularly strong correlation with NITA, WCTA and MARKETCAPTD, and further appear around the significant correlation between WCTA and MARKETCAPTD. To verify whether multicollinearity restricts the regression analyses, variance inflation factors are computed. As its values are all acceptable, multicollinearity is assumed to be harmless (appendix C).
TABLE 1: Descriptive Statistics
Variables Obs. Mean Std. Dev. Min. Max.
(1) BANKRUPT 116 0.5 0.5021692 0 1 (2) GREENWASHING 116 0.25 0.4348913 0 1 (3) BROWNWASHING 116 0.25 0.4348913 0 1 (4) NITA 116 -0.1018183 0.2107239 -1.23616 0.3577811 (5) CFTD 112 -0.0003569 0.9141044 -9.5 0.6734937 (6) EBITDATL 116 -0.0616135 0.4372754 -1.455616 3.621625 (7) WCTA 115 0.0319386 0.3225987 -1.783954 0.9382803 (8) MARKETCAPTD 108 1.970759 5.13445 -.0015493 43.86623 (9) EXCESSMKT_RETURN 110 -0.3451136 0.6075171 -1.381154 3.419019 Note: NITA=net-income to assets, CFTD= cash-flow to debt, EBITDATL= EBITDA to liabilities, WCTA= working capital to
TABLE 2: Correlation Matrix Variables (1) (2) (3) (4) (5) (6) (7) (8) (9) (1) BANKRUPT 1.000 (2) GREENWASHING 0.060 1.000 (3) BROWNWASHING -0.139 -0.333*** 1.000 (4) NITA -0.295*** -0.023 0.139 1.000 (5) CFTD -0.129 -0.155 0.045 0.148 1.000 (6) EBITDATL -0.031 -0.067 -0.012 0.198** 0.024 1.000 (7) WCTA -0.299*** 0.048 0.027 0.139 0.030 -0.022 1.000 (8) MARKETCAPTD -0.092 -0.066 0.111 0.144 0.047 0.015 0.307*** 1.000 (9) EXCESSMKT_RETURN -0.351*** 0.024 0.092 0.299*** -0.009 0.160* 0.305*** 0.282*** 1.000 Notes: *** P<0.01, ** P<0.05, * P<0.1 ;
In table 3, the logistic regression outputs of model (ii) are demonstrated. Interestingly, the coefficients of EBITDATL and MARKETCAPTD reversed into positive values, in contrast to the correlation statistics and the expected direction. Although this might be due to the narrow sample size, an alternative explanation is provided by the matching criteria of financial difficulties. As consequently differences between bankrupt and non-bankrupt companies are naturally small, the expected sign of both control variables with bankruptcy may reasonably reverse. The regression analyses however concentrate on the effects of green- and brownwashing on the probability of bankruptcy.
Although the expected positive effect of greenwashing on bankruptcy is reflected and confirmed by the positive regression coefficient, greenwashing does not significantly affect the probability of bankruptcy according to the p-value. As a consequence, H1 is not supported since greenwashing does not significantly increase the likelihood of bankruptcy. Table 4 presents the regression outputs of model (iii), which confirms the negative relationship between brownwashing and bankruptcy. As a matter of fact, the negative coefficient of brownwashing retains a p-value of 0.08, which is statistically significant at a 10% level. Thereby, as brownwashing significantly reduces the likelihood of bankruptcy, support for H2 is provided. Given the proven, yet insignificant positive effect of greenwashing, and the significant negative effect of brownwashing on the probability of bankruptcy, support is found for the hypothesized claim that engaging in greenwashing compared to brownwashing increases the likelihood of bankruptcy as well (H3). According to the pseudo R-squared, Akaike- and Bayesian criteria, and the significant chi-squared test statistics, the inclusion of brownwashing instead of greenwashing slightly increases the model fit as well.
TABLE 3: Logistic Regression Model (ii)
BANKRUPT Coef. Std. Err. Z-statistic P > | z | [95% Conf. Interval]
GREENWASHING 0.759 0.588 1.29 0.197 -0.393 1.911 NITA -7.502 2.909 -2.58 0.010** -13.203 -1.801 CFTD -6.974 3.001 -2.32 0.020** -12.856 -1.092 EBITDATL 4.356 2.017 2.16 0.031** 0.403 8.310 WCTA -4.534 1.882 -2.41 0.016** -8.222 -0.845 MARKETCAPTD 0.146 0.055 2.67 0.008*** 0.039 0.253 EXCESSMKT_RETURN -1.299 0.674 -1.93 0.054* -2.620 0.022 CONSTANT -1.388 0.548 -2.53 0.011** -2.462 -0.313
Mean dependent var 0.453 SD dependent var 0.5
Pseudo R-squared 0.338 Number of obs. 106
Chi-square 49.382 Prob > chi2 0.000
Akaike crit. (AIC) 112.620 Bayesian crit. (BIC) 133.928 Notes: *** P<0.01, ** P<0.05, * P<0.1 ;
TABLE 4: Logistic Regression Model (iii)
BANKRUPT Coef. Std. Err. Z-statistic P > | z | [95% Conf. Interval]
BROWNWASHING -1.162 0.665 -1.75 0.080* -2.466 0.141 NITA -6.993 2.897 -2.41 0.016** -12.671 -1.314 CFTD -6.978 2.956 -2.36 0.018** -12.773 -1.183 EBITDATL 4.095 1.944 2.11 0.035** 0.286 7.905 WCTA -4.341 1.903 -2.28 0.023** -8.070 -0.612 MARKETCAPTD 0.144 0.056 2.57 0.010** 0.034 0.254 EXCESSMKT_RETURN -1.228 0.665 -1.85 0.065* -2.530 0.075 CONSTANT -0.861 0.498 -1.73 0.084* -1.836 0.115
Mean dependent var 0.453 SD dependent var 0.5
Pseudo R-squared 0.350 Number of obs. 106
Chi-square 51.043 Prob > chi2 0.000
Akaike crit. (AIC) 110.959 Bayesian crit. (BIC) 132.267 Notes: *** P<0.01, ** P<0.05, * P<0.1 ;
NITA=net-income to assets, CFTD= cash-flow to debt, EBITDATL= EBITDA to liabilities, WCTA= working capital to assets, MARKETCAPTD= market capitalization to debt, EXCESSMKT_RETURN= excess market return
To shed further light, the corresponding odds ratios of both logistic regression models are declared in table 5 and 6. In contrast to probabilities, odds ratios represent a constant effect based on the
comparison between the probability of bankruptcy versus the probability of non-bankruptcy, ranging from zero to infinity. An odds ratio exceeding one thereby demonstrates a positive relationship, and vice versa for values lower than one. Regarding the odds ratios of green- and brownwashing, 2.136 times greater odds of bankruptcy would have occurred if greenwashing was a significant predictor of bankruptcy, whilst the odds of bankruptcy are around three times lower (1/0.313) when a firm engages in brownwashing. For the remaining continuous control variables, the odds of bankruptcy rise due to unit increases in EBITDATL and MARKETCAP, and descend through unit increases in NITA, CFTD, WCTA and EXCESSMKT_RETURN.
Despite this study does not intent to find parsimonious models that strongly predict bankruptcy, a condensed and plain side-analysis using classification tables is executed as a matter of interest. The tables in appendix D show that the overall prediction accuracy of the baseline model (80.19%) remains equal when greenwashing is included, but increments for the inclusion of brownwashing (81.13%). Yet, the correct classifications of non-bankruptcies exceed that of bankruptcies as roughly 72% of bankruptcy events are properly predicted in all models; contrary to roughly 88% of non-bankruptcies. Ultimately, differences amongst the models are neglectable.
TABLE 5: Odds Ratios - Logistic Regression Model (ii)
BANKRUPT Odds Ratio Std. Err. Z-statistic P > | z | [95% Conf. Interval]
GREENWASHING 2.136 1.255 1.29 0.197 0.675 6.758 NITA 0.001 0.002 -2.58 0.010** 0.000 0.165 CFTD 0.001 0.003 -2.32 0.020** 0.000 0.336 EBITDATL 77.958 157.254 2.16 0.031** 1.496 4063.246 WCTA 0.011 0.020 -2.41 0.016** 0.000 0.429 MARKETCAPTD 1.157 0.063 2.67 0.008*** 1.040 1.287 EXCESSMKT_RETURN 0.273 0.184 -1.93 0.054* 0.073 1.023 CONSTANT 0.250 0.137 -2.53 0.011** 0.085 0.731 Notes: *** P<0.01, ** P<0.05, * P<0.1 ;
NITA=net-income to assets, CFTD= cash-flow to debt, EBITDATL= EBITDA to liabilities, WCTA= working capital to assets, MARKETCAPTD= market capitalization to debt, EXCESSMKT_RETURN= excess market return
TABLE 6: Odds Ratios - Logistic Regression Model (iii)
BANKRUPT Odds Ratio Std. Err. Z-statistic P > | z | [95% Conf. Interval]
BROWNWASHING 0.313 0.208 -1.75 0.080* 0.085 1.151 NITA 0.001 0.003 -2.41 0.016** 0.000 0.269 CFTD 0.001 0.003 -2.36 0.018** 0.000 0.306 EBITDATL 60.054 116.721 2.11 0.035** 1.331 2709.825 WCTA 0.013 0.025 -2.28 0.023** 0.000 0.542 MARKETCAPTD 1.155 0.065 2.57 0.010** 1.035 1.289 EXCESSMKT_RETURN 0.293 0.195 -1.85 0.065* 0.080 1.078 CONSTANT 0.423 0.210 -1.73 0.084* 0.159 1.122 Notes: *** P<0.01, ** P<0.05, * P<0.1 ;
TABLE 7: Average Predictive Margins
Margin Delta-method Std. Err.
Z-statistic P > | z | [95% Conf. Interval] GREENWASHING 0 0.4231714 0.0428737 9.87 0.000*** 0.3391404 0.5072024 1 0.5379878 0.0746787 7.20 0.000*** 0.3916201 0.6843554 BROWNWASHING 0 0.4908647 0.0441198 11.13 0.000*** 0.4043915 0.5773379 1 0.3233055 0.0757497 4.27 0.000*** 0.1748389 0.4717721 Note: *** P<0.01, ** P<0.05, * P<0.1
TABLE 8: Adjusted Predictive Margins
Margin Delta-method Std. Err.
Z-statistic P > | z | [95% Conf. Interval] GREENWASHING 0 0.566238 0.1013071 5.59 0.000*** 0.3676797 0.7647964 1 0.7360314 0.1192094 6.17 0.000*** 0.5023853 0.9696775 BROWNWASHING 0 0.6696943 0.0937697 7.14 0.000*** 0.485909 0.8534796 1 0.3880322 0.1514823 2.56 0.000** 0.0911325 0.684932 Note: *** P<0.01, ** P<0.05, * P<0.1
TABLE 9: Logistic Regression Model – GREENWASHINGt3
BANKRUPT Coef. Std. Err. Z-statistic P > | z | [95% Conf. Interval]
GREENWASHINGt3 1.342 0.783 1.71 0.087* -0.193 2.878 NITA -7.889 3.251 -2.43 0.015** -14.261 -1.517 CFTD -8.904 3.563 -2.50 0.012** -15.888 -1.920 EBITDATL 4.980 2.206 2.26 0.024** 0.658 9.303 WCTA -6.362 2.439 -2.61 0.009*** -11.141 -1.582 MARKETCAPTD 0.179 0.063 2.85 0.004*** 0.056 0.302 EXCESSMKT_RETURN -1.367 0.805 -1.70 0.090* -2.945 0.211 CONSTANT -1.250 0.629 -1.99 0.047** -2.483 -0.017
Mean dependent var 0.483 SD dependent var 0.503
Pseudo R-squared 0.375 Number of obs. 87
Chi-square 45.237 Prob > chi2 0.000
Akaike crit. (AIC) 91.267 Bayesian crit. (BIC) 110.994 Notes: *** P<0.01, ** P<0.05, * P<0.1 ;
GREENWASHINGt3= three-year lagged greenwashing, NITA=net-income to assets, CFTD= cash-flow to debt, EBITDATL= EBITDA to liabilities, WCTA= working capital to assets, MARKETCAPTD= market capitalization to debt,
EXCESSMKT_RETURN= excess market return Robustness
To verify the robustness of the previous outcomes, alternative model specifications are examined. First of all, probit regressions are executed to guarantee deviating results did not occur due to the logistic model choice. The regression outputs in appendix E actually demonstrate increased significance levels of both green- and brownwashing for the probit modifications of model (ii) and (iii). Despite an ongoing lack of support for H1, brownwashing almost became significant at a 5% level given its p-value of 0.058. Thereby, consistent support for H2 and H3 exists.
Concerning modifications of the market-based variables, the annual percentage change of market capitalization (ΔMARKETCAP) is considered as a substitute for EXCESSMKT_RETURN, which was found to correlate with various control variables. According to Beaver et al. (2005), ΔMARKETCAP similarly measures market-based returns. In the initial baseline model (i), this replacement results in a p-value of 0.091 for brownwashing whilst ΔMARKETCAP becomes insignificant. Yet, to elaborate on the baseline model without EBITDATL, substituting EXCESSMKT_RETURN for ΔMARKETCAP leads to exclusively significant effects, including brownwashing (table 10b). All in all, the preceding results from the main analysis remain stable.
TABLE 10a: Logistic Regression Model – Excluding EBITDATL
BANKRUPT Coef. Std. Err. Z-statistic P > | z | [95% Conf. Interval]
BROWNWASHING -1.121 0.640 -1.75 0.080* -2.375 0.133 NITA -1.975 1.425 -1.39 0.166 -4.769 0.819 CFTD -4.768 2.729 -1.75 0.081* -10.118 0.582 WCTA -4.608 1.789 -2.58 0.010** -8.114 -1.101 MARKETCAPTD 0.125 0.056 2.24 0.025** 0.016 0.234 EXCESSMKT_RETURN -1.412 0.657 -2.15 0.032** -2.699 -0.124 CONSTANT -0.460 0.451 -1.02 0.308 -1.345 0.425
Mean dependent var 0.453 SD dependent var 0.5
Pseudo R-squared 0.312 Number of obs. 106
Chi-square 45.566 Prob > chi2 0.000
Akaike crit. (AIC) 114.436 Bayesian crit. (BIC) 133.080 Notes: *** P<0.01, ** P<0.05, * P<0.1 ;
NITA=net-income to assets, CFTD= cash-flow to debt, WCTA= working capital to assets, MARKETCAPTD= market capitalization to debt, EXCESSMKT_RETURN= excess market return
TABLE 10b: Logistic Regression Model – Substituting EXCESSMKT_RETURN
BANKRUPT Coef. Std. Err. Z-statistic P > | z | [95% Conf. Interval]
BROWNWASHING -1.045 0.632 -1.65 0.098* -2.283 0.194 NITA -2.437 1.418 -1.72 0.086* -5.216 0.342 CFTD -5.460 2.691 -2.03 0.042** -10.734 -0.186 WCTA -4.517 1.728 -2.61 0.009*** -7.904 -1.129 MARKETCAPTD 0.125 0.055 2.28 0.023** 0.017 0.233 ΔMARKETCAP -0.890 0.536 -1.66 0.097* -1.940 0.160 CONSTANT -0.162 0.395 -0.41 0.682 -0.936 0.612
Mean dependent var 0.453 SD dependent var 0.5
Pseudo R-squared 0.297 Number of obs. 106
Chi-square 43.297 Prob > chi2 0.000
Akaike crit. (AIC) 116.705 Bayesian crit. (BIC) 135.350 Notes: *** P<0.01, ** P<0.05, * P<0.1 ;
V.
DISCUSSION AND CONCLUSION
This study is the first to investigate the novel relationship between green- and brownwashing, and bankruptcy under the circumstances of financial difficulties. With regard to the initial research questions, the foregoing findings explain the ability of green- and brownwashing to affect the probability of bankruptcy, and their comparison. Thereby, its explorative research intent to clarify the outlook and comparison of the relation between respectively green- and brownwashing and the likelihood of bankruptcy is accomplished.
Although the hypothesized and proven positive effect of greenwashing on the probability of bankruptcy is concluded to be insignificant (H1), empirical evidence at a 10% significance level is found for the expected negative effect of brownwashing (H2). This implies that the probability of bankruptcy does not increase as a consequence of greenwashing practices, whereas brownwashing tends to be capable of reducing the likelihood firms will go bankrupt. Therefore, firms with financial difficulties that engage in greenwashing are more likely to go bankrupt as opposed to companies with a brownwashing strategy, in consonance with the final hypothesis (H3).
Implications and theoretical contribution
Concerning the insignificant positive effect of greenwashing on the probability of bankruptcy, a potential explanation might arise from the underestimated salience of non-shareholder stakeholders compared to shareholders of firms with financial constraints. For example, employees (Bhattacharya et al., 2008), financial institutions (Cheng et al., 2014) and other stakeholders (e.g. Eccles et al., 2014) could be positively attracted towards enlarged CSR friendly practices. Whereas increased non-shareholder stakeholder support incentivized greenwashing, it might actually negatively moderate the positive effect of greenwashing on the likelihood of bankruptcy too. This potentially justifies the insignificance of the expected relationship. Nonetheless, the conflicting interests and presence of asymmetric information between managers and shareholders (Eisenhardt, 1989; Jensen & Meckling, 1976), as well as the eminence of descended shareholder support and involvement that underlie the positive relation between greenwashing and bankruptcy, remain decisive given the approval of the expected positive effect. In other words, the prioritization of shareholders over non-shareholder stakeholders to ensure firm survival is confirmed by means of the validated positive effect of greenwashing on the probability of bankruptcy.
Yet, with regard to the proven significant negative effect of brownwashing on the likelihood of bankruptcy, misleadingly aligned interests between managers and shareholders tend to expectedly decrease the probability of bankruptcy via the consequential declined agency costs and enlarged shareholder support and involvement. Given the demand for corporate transparency and presence of intense shareholder pressure and scrutiny under the conditions of financial difficulties (Healy & Palepu, 2001; Kim & Lyon, 2015; Watson et al., 2002), brownwashing may actually be a suitable bonding mechanism (Jensen & Meckling, 1976). This is embedded in its proven ability to reduce the likelihood of bankruptcy, which implies brownwashing can undermine agency problems of conflicting interests and asymmetric information between managers and shareholders by means of impression management. The result additionally indicates that financially constrained firms are minimally willing to further invest in governance instruments as this increases agency costs, and brownwashing might then have been discovered. Since this would have caused different outcomes, the findings indicate brownwashing is executed undetectably. Hence, in contrast to the implication that brownwashing is more likely recognized due to the contemporary incremented ability of stakeholders to closely monitor firms (Testa et al., 2018), the results represent the opposite as recognition would have resulted in a larger probability of bankruptcy.Testa et al. (2018) additionally argued brownwashing to diminish profitability, whilst Hawn and Ioannou (2016) indirectly suggested a similar negative effect on market value. Despite its misrepresentation is not proven, the curtailed likelihood of bankruptcy caused by brownwashing suggests contrasting results. Furthermore, the notion that “shareholders may respond negatively to a firm’s environmentally friendly practices” may well be true, but remains uncertain (Kim & Lyon, 2015, p. 708). Although it is not proven that greenwashing increases the likelihood of bankruptcy via shareholders’ reactions, the negative impact of brownwashing at least shows shareholder satisfaction of avoiding active CSR engagement by managers.
Given that brownwashing entails the false, but trusted alignment of shareholders’ interests, which scales down the likelihood of bankruptcy, shareholders safeguard continuity and are therefore prioritized. This outcome thereby contributes to the stakeholder salience theory and prioritization related research (Mitchell et al., 1997), and studies in which the role of stakeholders (e.g. shareholders) for bankruptcy is investigated.
Accordingly, in practice, managers could consider brownwashing strategies to potentially prevent bankruptcy given the troubling financial circumstances. The results thereby add an additional dimension to strategic decision-making by managers regarding the communication of CSR practices. In particular, managers should be aware of the significant importance of shareholders for firms with financial difficulties. As illustrated by the proven, yet insignificant positive effect of greenwashing on the probability of bankruptcy, and the validated significant negative effect of brownwashing, firms get rewarded by shareholders when managers seemingly conduct business in accordance with shareholder interests (i.e. brownwashing) contrary to misaligned interests (i.e. greenwashing). The outcomes therefore proclaims managers to fulfil shareholders’ needs to ensure firm survival and their career. Furthermore, stakeholders should take into account that financially constrained firms engaging in greenwashing, as opposed to brownwashing, are more likely to face bankruptcy. For instance customers, policy makers and especially capital supplying stakeholders as banks and investors could benefit from expanded knowledge on the odds of bankruptcy for firms with financial difficulties due to the importance of risk evaluations in investment decisions. In addition, this study increases shareholder awareness on the potential existence of green- and brownwashing IM tactics to misleadingly shape perceptions on managerial CSR efforts and performance.
Limitations and recommendations for future research
Nevertheless, this study is strongly restricted with regard to the data collection. As the availability of simultaneously available CSR and bankruptcy data is limited, a small sample size exists. The contemporary development of CSR databases in which currently solely large listed firms are included underlies the aforementioned. Large companies namely tend to experience bankruptcy occasionally, restricting the assembly of convenient data. Due to the small sample size, results may not only be biased, but additionally less reliable.
Nonetheless, the ultimate conversion towards separate indicator variables for green- and brownwashing still provided insightful indications of their effects on bankruptcy. Moreover, the research design does not account for the fact outcomes potentially deviate across industries, and constrains the inclusion of controls for the year of bankruptcy via dummy variables in the logistic regression analyses. To conveniently incorporate time in the analysis, the application of hazard models is suggested (Bauer & Agarwal, 2014; Beaver et al., 2005; Shumway, 2001). Subsequently, although the novelty embedded in this study partially complicates the comparison with alternative studies, it exhibits the existence of sufficient interesting and relevant fields of future research.
Given that the amount of available CSR data expands and advances, the suitability of this research topic increases in the forthcoming years. The recombination of data from multiple sources is advocated for future researchers that intent to study the equivalent. Whereas company-specific scores for CSR disclosure could be gathered from ASSET4, operational performance measures might be gathered from different databases. Company-based scores can consequently be computed and compared to industry benchmarks to indicate the intensity of green- and brownwashing. In that manner, substitutive measures for green- and brownwashing can be established and inspected too.
As solely the direction of the effects green- and brownwashing have on bankruptcy are studied, it could be interesting for future research to investigate the actual shape of these relationships as well. Since greater intensities of green- and brownwashing rationally increase the chance of discovery, exponential effects might be reasonably present. In fact, larger degrees of brownwashing potentially adjusts the relationship with the probability of bankruptcy towards a U-shaped relation. Hawn and Ioannou (2016) actually concluded that a significant misalignment between internal- and external CSR actions had the most detrimental effect on market value. Supplementary, given that the fit between real and communicated CSR, and in particular its consistency overtime, tend to determine whether organizations actually benefit from CSR (Du et al., 2010; Wagner et al., 2009), researchers are advised to further examine the importance of consistently using green- and brownwashing too.
VI.
REFERENCES
Altman, E.I. (1968). Financial ratios, discriminant analysis and the prediction of corporate bankruptcy. The Journal of Finance, 23(4), 589-609.
Altman, E.I. (1993). Corporate financial distress and bankruptcy: A complete guide to predicting and avoiding distress and profiting from bankruptcy (2nd ed.). New York: Wiley
Bansal, P. & Clelland, I. (2004). Talking trash: Legitimacy, impression management, and unsystematic risk in the context of the natural environment. Academy of Management Journal, 47(1), 93-103. Bauer, J. & Agarwal, V. (2014). Are hazard models superior to traditional bankruptcy prediction approaches? A comprehensive test. Journal of Banking and Finance, 40(1), 432-442.
Beaver, W.H. (1966). Financial ratios as predictors of failure. Journal of Accounting Research, 4, 71-111.
Beaver, W.H., McNichols M.F. & Rhie, J. (2005). Have financial statements become less informative? Evidence from the ability of financial ratios to predict bankruptcy. Review of Accounting Studies, 10(1), 93-122.
Bellovary, J.L., Giacomino, D.E. & Akers, M.D. (2007). A review of bankruptcy prediction studies: 1930 to present. Journal of Financial Education, 33, 1-42.
Bhattacharya, C.B., Sen, S. & Korschun, D. (2008). Using corporate social responsibility to win the war for talent. MIT Sloan Management Review, 49(2), 37-44.
Bolino, M., Kacmar, K., Turnley, W., Gilstrap, J. (2008). A multi-level review of impression management motives and behaviors. Journal of Management, 34(6), 1080-1109.
Carlos, W.C. & Lewis, B.W. (2018). Strategic silence: Withholding certification status as a hypocrisy avoidance tactic. Administrative Science Quarterly, 63(1), 130-169.
Carroll, A.B. & Shabana, K.M. (2010). The business case for corporate social responsibility: A review of concepts, research and practice. International Journal of Management Reviews, 12(1), 85-105. Cheng, B., Serafeim, G. & Ioannou, I. (2014). Corporate social responsibility and access to finance. Strategic Management Journal, 35(1), 1-23
Delmas, M.A. & Burbano, V.C. (2011). The drivers of greenwashing. California Management Review, 54(1), 64-87.
Dutton, J. & Dukerich, J. (1991). Keeping an eye on the mirror: Image and identity in organizational adaptation. The Academy of Management Journal, 34(3), 517-554.
Eccles, R.G., Ioannou, I. & Serafeim, G. (2014). The impact of corporate sustainability on organizational processes and performance. Management Science, 60(11), 2835-2857.
Eisenhardt, K. (1989). Agency theory: An assessment and review. The Academy of Management Review, 14(1), 57-74.
Executive Office of the President Office of Management and Budget (2017). North American Industry Classification System, United States. Retrieved from: census.gov/naics
Gilson, S. (1989). Management turnover and financial distress. Journal of Financial Economics, 25(2), 241-262.
Goffman, E. (1959). The presentation of self in everyday life. Garden City, NY: Doubleday & Company Hawn, O. & Ioannou, I. (2016). Mind the gap: The interplay between external and internal actions in the case of corporate social responsibility. Strategic Management Journal, 37(13), 2569-2588.
Healy, P.M. & Palepu, K.G. (2001). Information asymmetry, corporate disclosure, and the capital markets: A review of the empirical disclosure literature. Journal of Accounting and Economics, 31(1), 405-440.
Highhouse, S., Brooks, M., & Gregarus, G. (2009). An organizational impression management perspective on the formation of corporate reputations. Journal of Management, 35(6), 1481-1493. Hull, C.E. & Rothenberg, S. (2008). Firm performance: The interactions of corporate social performance with innovation and industry differentiation. Strategic Management Journal, 29(7), 781-789.
Jensen, M.C. & Meckling, W.H. (1976). Theory of the firm: Managerial behavior, agency costs and ownership structure. Journal of Financial Economics, 3(4), 305-360.
Kim, E. & Lyon, T.P. (2015). Greenwashing vs brownwash: Exaggeration and undue modesty in corporate sustainability disclosure. Organization Science, 26(3), 705-723.
Lioui, A. & Sharma, Z. (2012). Environmental corporate social responsibility and financial performance: Disentangling direct and indirect effects. Ecological Economics, 78, 100-111.
Lyon, T.P. & Montgomery, A.W. (2015). The means and end of greenwash. Organization and Environment, 28(2), 223- 249.
McWilliams, A. & Siegel, D. (2000). Corporate social responsibility and firm financial performance. Strategic Management Journal, 21(5), 602-609.
McWilliams, A. & Siegel, D. (2001). Corporate social responsibility: A theory of the firm perspective. Academy of Management Review, 26(1), 117-127.
Milgrom, P.R. (1981). Good news and bad news: Representation theorems and applications. The Bell Journal of Economics, 12(2), 380-391.
Mitchell, R., Agle, B. & Wood, D. (1997). Toward a theory of stakeholder identification and salience: Defining the principle of who and what really counts. The Academy of Management Review, 22(4), 853-886.
S&P Global Ratings (2018, October 31). S&P Global Ratings Definition. Retrieved from: https://www.standardandpoors.com/en_US/web/guest/home
Shumway, T. (2001). Forecasting bankruptcy more accurately: A simple hazard model. The Journal of Business, 74(1), 101-124.
Surroca, J. & Tribó, J.A. (2008). Managerial entrenchment and corporate social performance. Journal of Business Finance & Accounting, 35(5-6), 748-789.
Testa, F., Miroshnychenko, I., Barontini, R. & Frey, M. (2018). Does it pay to be a greenwasher or a brownwasher? Business Strategy and the Environment, 27(7), 1104-1116.
Ullmann, A.A. (1985). Data in search of a theory: A critical examination of the relationships among social performance, social disclosure, and economic performance of US firms. Academy of Management Review, 10(3), 540-557.
Vilanova, M., Lozano, J.M. & Arenas, D. (2009). Exploring the nature of the relationship between CSR and competitiveness. Journal of Business Ethics, 87(1), 57-69.
Waddock, S.A. & Graves, S.B. (1997). The corporate social performance-financial performance link. Strategic Management Journal, 18(4), 303-319.
Wagner, T., Lutz, R.J. & Weitz, B.A. (2009). Corporate hypocrisy: Overcoming the threat of inconsistent corporate social responsibility perceptions. Journal of Marketing, 73(6), 77-91.
Watson, A., Shrives, P. & Marston, C. (2002). Voluntary disclosure of accounting ratios in the UK. The British Accounting Review, 34(4), 289-313.
VII.
APPENDIX
A. Supporting Conceptual Model
B1. Internal CSR Variables
Environmental
1. Does the company have a policy to reduce emissions?
2. Does the company develop products or technologies that are used for water treatment, purification or that improve water use efficiency?
3. Does the company have a policy to improve its water efficiency? 4. Does the company have a policy to improve its energy efficiency? 5. Does the company make use of renewable energy?
6. Does the company use environmental criteria (ISO 14000, energy consumption, etc.) in the selection process of its suppliers or sourcing partners?
Social
7. Does the company have a diversity and equal opportunity policy, and a work-life balance policy? 8. Does the company have a competitive employee benefits policy, or a policy to ensure good employee relations within its supply chain?
9. Does the company have a policy to avoid child labour, and a policy to avoid forced labour? 10. Does the company have a policy to improve employee health & safety within its supply chain? 11. Does the company have a policy to support the skills training or career development of its employees?
Govenance
12. Percentage of non-executive board members on the audit committee as stipulated by the company. 13. Percentage of non-executive board members on the nomination committee as stipulated by the company.
14. Do the company's statutes or bylaws require that stock options are only granted with a majority vote at a shareholder meeting?
15. Does the company have a CSR committee or team? 16. Percentage of women on the board of directors.
17. Percentage of independent board members as reported by the company.
18. Does the company have a policy for ensuring equal treatment of minority shareholders, facilitating shareholder engagement or limiting the use of anti-takeover devices?
19. Does the company have a policy for performance-oriented compensation that attracts and retain the senior executives and board members?
