draft

Vol.:(0123456789)1 3

Journal of Business Ethics (2020) 167:775–791 https://doi.org/10.1007/s10551-019-04158-z

O R I G I N A L PA P E R

Do LGBT Workplace Diversity Policies Create Value for Firms?

Mohammed Hossain1,2 · Muhammad Atif3 · Ammad Ahmed4 · Lokman Mia1

Received: 27 June 2018 / Accepted: 3 April 2019 / Published online: 25 April 2019 © Springer Nature B.V. 2019

Abstract We show that the U.S. anti-discriminatory laws prohibiting discrimination in the workplace based on sexual orientation and gender identity (i.e. lesbian, gay, bisexual, and transgender (LGBT) identities) spur innovation, which ultimately leads to higher firm performance. We use the Human Rights Campaign’s Corporate Equality Index (CEI) of 398 (1592 firm- year observations) U.S. firms between 2011 and 2014, and find a significantly positive relationship between CEI and firm innovation. We also find that an interacting effect of CEI and firm innovation leads to higher firm performance. We use our understanding of Rawls’ Theory of Justice and stakeholder theory to show that firms with workplace diversity policies are likely to be more innovative and perform better than those without such policies. Our results are robust to endogeneity, reverse causality and simultaneity issues. Our results will trigger debate in similar markets around the globe on the economic benefits of LGBT workplace diversity policies for firms.

Keywords Workplace diversity · LGBT · Innovation · Firm performance

Introduction

Support for LGBT (lesbian, gay, bisexual, and transgender— hereafter LGBT) rights has increased substantially over the last two decades in Australia, UK, USA, and other European countries (Lloren and Parini 2017; Pichler et al. 2018). Con- sequently, corporate equality initiatives, and more specifi- cally, employee equality initiatives are becoming an integral part of firms’ diversity management. These initiatives sig- nal an open and tolerant workplace environment in which employees are not discriminated against on the basis of their

sexual orientation or gender identity.1 Liddle et al. (2004) emphasize that workplace environment plays a key role in employee recruitment, productivity, stress and commitment. It is estimated that approximately eight million people, or 3.5% of the U.S population, identify as LGBT (Gates 2011, 2012) and 30 states have no laws protecting the employment rights of LGBT individuals (Webster et al. 2018). Research has shown that individuals who identify as LGBT face dis- crimination, hostility and negative attitudes (homophobia and transphobia) in the workplace, which negatively affects their performance on the job in terms of higher absenteeism and lower productivity (Bonaventura and Biondo 2016).

To improve the workplace environment, in 2017, the U.N. High Commissioner for human rights released new standards of conduct to eliminate discrimination against LGBT employees in the workplace.2 These anti-discrimi- natory policies have both societal and economic benefits. For example, formal acceptance of LGBT employees in the workplace makes them feel less anxious, less threatened, and more comfortable (Liddle et al. 2004). In terms of economic

* Mohammed Hossain [email protected]; [email protected]

Muhammad Atif [email protected]

Ammad Ahmed [email protected]

Lokman Mia [email protected]

1 Department of Accounting, Finance and Economics, Griffith Business School, Nathan, QLD 4111, Australia

2 Faculty of Business, Sohar University, Sohar, Oman 3 Essex Business School, University of Essex, Colchester, UK 4 College of Business, Zayed University, Abu Dhabi,

United Arab Emirates

1 The extant literature from various fields, including history, soci- ology, and psychology, concurs that discrimination against LGBT groups exists because of both sexual orientation and gender identity (e.g. Badgett 1995; Drydakis 2009; King and Cortina 2010; Ozeren 2014; Bonaventura and Biondo 2016). 2 These standards include respect, elimination and prevention of dis- crimination, support, and taking a stand for LGBT individuals.

776 M. Hossain et al.

1 3

benefits, in Australia, it has been estimated that acceptance of secluded workers in different workplaces could lead to as much as $285 million in savings per year nationally, an increase of 11% in staff retention and an increase of 30% in productivity (Johnson and Cooper 2015). At the organiza- tional level, acceptance of LGBT groups increases the pool of talent from which organizations may draw strategic ben- efits, and such inclusion leads to an increase in diversity in different positions and professional teams within the organi- zation (Barbulescu and Bidwell 2013). However, although promoters of LGBT-supportive policies argue that these enhance the talent pool and improve firm-level diversity, very little attention has been paid to whether these policies create value for firms.

To recognize the economic effects of LGBT workplace policies, this paper focuses on the effect of LGBT workplace policies3 in value creation for firms. A growing number of studies have focused on the social imperative of workplace policies, i.e. discrimination and LGBT workplace policies (Ragins and Cornwell 2001; Priola et al. 2014), diversity and LGBT workplace policies (Ozturk and Tatli 2016), stigma in the workplace and LGBT workplace policies (Ragins 2008), and politics and LGBT workplace policies (Gupta et al. 2017; Rhodes 2017). However, only a limited num- ber of studies have focused on the economic imperative of workplace policies, i.e. firm performance and LGBT work- place policies (Shan et al. 2017; Pichler et al. 2017, 2018). In this paper, we seek to investigate the impact of LGBT workplace policies on firms’ innovation, and ultimately on firm performance.

We use Rawls’ Theory of Justice (1971) to explain ethical corporate behaviour, social responsibility, societal fairness and equality (Chapman 1975). This theory points to fairness as a social good and suggests that institutions (in this case, firms) have a responsibility and an opportunity to demon- strate how to treat others (in this case, employees) fairly. The theory further specifies that fairness should not be con- tingent upon socio-demographical characteristics. Instead, fairness at the workplace should be rooted within a work- ers’ meritocracy. Prior literature that has confirmed Rawls’ (1971) concept of fairness and its value to organizational outcomes is limited to some specific socio-demographical factors, i.e. race and religion (Beckley 1986; Cohen 2010). To date, no research on Rawls’ (1971) contribution is extended specifically to the fairness of LGBT employees in

the workplace. We apply the Theory of Justice to the chal- lenges faced by the LGBT community, and show how trans- parent corporate communication that reflects fairness, i.e. LGBT-supportive workplace policies, may have implications for organizational outcomes. Moreover, we use stakeholder theory, which builds on the premise that all firms’ stake- holders should be treated fairly and equally (Fieseler et al. 2010), to demonstrate that firms, through transparent cor- porate communication, create value for stakeholders, which ultimately adds value to the firm.

To empirically answer our research question, we use data from the Human Rights Campaign (HRC), which provide information on firms’ overall Corporate Equality Index (CEI) score, sexual orientation non-discrimination policies, gender-identity non-discrimination policies, domestic part- ner benefits, and transgender policies from 2011 to 2014. We find a significant positive relationship between the CEI score and individual policies with firms’ levels of innova- tion, which ultimately positively affect firm performance.

Our study may face critique on potential endogeneity bias due to the causal effect of CEI on innovation. For instance, a manager who is sensitive to the benevolent effects of cor- porate social equality may hire more LGBT employees by implementing LGBT-supportive workplace policies that will ultimately improve firm innovation. On the other hand, firms keen on innovation may also be responsive to external and societal expectation in terms of implementing LGBT work- place policies. We employ two techniques to minimize these endogeneity concerns: propensity score matching (PSM) and dynamic panel estimation [Generalized method of moments (GMM)]. Our results are robust to these sensitivity tech- niques and to alternate proxies of firm innovation.

Our contribution to the existing literature is twofold. First, we contribute to the literature on Rawls’ Theory of Justice (1971) by extending its construct of fairness to LGBT employees and policies in the workplace that create value for organizations. Second, we provide empirical evi- dence to support our arguments on how LGBT workplace policies positively affect firm outcomes. More specifically, we contribute to the existing literature by investigating the actual driver of value-addition, i.e. firm innovation. Previ- ous studies that found a positive association between LGBT workplace policies and firm performance failed to provide a channel through which these policies may have an impact on firm performance. In this study, we confirm the argument that LGBT workplace policies improve firm performance through innovation. Our study is also timely and supports the upsurge in calls for LGBT rights in workplaces around the globe.

The remainder of the paper is structured as follows. “Workplace diversity management, accounting and inno- vation” section discusses the intra-relationship between workplace diversity management policies, accounting and

3 LGBT workplace policy is measured through the Corporate Equal- ity Index (CEI), which is published annually by Human Rights Cam- paign (HRC), the largest organization for LGBT rights in the U.S. This index rates the firms from 0 to 100 points with 100 as the high- est score based on different sub-policies (e.g. sexual orientation non- discrimination policies, domestic partner benefits policies, workplace training and LGBT-supportive policy guidelines).

777Do LGBT Workplace Diversity Policies Create Value for Firms?

1 3

innovation. “Theory and hypothesis development” sec- tion discusses Rawls’ theory of Justice and the stakehoder theory and develops the main hypothesis for this research. “Research design” section comprises data collection mode of characteristics of the firms for the study, selection of con- trol variables and firm-specific characteristics and summary of descriptive statistics. “Results and discussion” section presents our empirical results based on workplace diversity policies and innovation followed by discussion on robust- ness checks and a detailed analysis. “Conclusion” section presents our conclusions, including the implications of our findings and the limitations of our research with suggestions for future research.

Workplace Diversity Management, Accounting and Innovation

With respect to the workplace, diversity refers to the co- existence of employees from various socio-cultural back- grounds. The equal opportunity philosophy is aimed at ensuring that organizations make the most out of the unique- ness of a diverse workforce, which might assist the organi- zation to be more efficient and effective, rather than losing talent. Broadly, diversity management is the systematic and planned commitment by an organization to recruit, retain, reward and promote a heterogeneous mix of employees (Grobler et al. 2006). Nowadays, for many leading busi- nesses, it is a strategic imperative to create a culture of inclusion and diversity that extends to LGBT people: they know that it correlates to greater individual performance and ultimately, stronger business performance. For example, 85% of Fortune 500 businesses have explicit policies against discrimination based on sexual orientation, and 49% include gender identity (Kelly 2016). Prior studies have found that various forms of diversity are associated with greater inno- vation, improved strategic decision making, and better out- comes when innovation and complex problem-solving are required (Jackson and Joshi 2004; Francoeur et al. 2008; Omankhanlen and Ogaga-Oghene 2011).

Hopwood (1987a, b) states that ‘Accounting is not a static phenomenon’ (p. 207). Over time, accounting has been impli- cated in the creation of very different patterns of organizational segmentation (Hopwood 1987a, b). If accounting is a machine (Burchell et al. 1980), it is a mechanical procedure that offers propositions about problems to be concerned with in the future (Mouritsen and Kreiner 2016). Therefore, account- ing is relevant in many different situations. When decision making is considered as a rational procedure, accounting is understood as an answering machine calculating the economic consequences of various decision alternatives (Mouritsen and Kreiner 2016). If decision making is understood in less rational terms, accounting may play a much more complex role, as a

learning, ammunition and rationalization machine (Stambaugh and Carpenter 1992; Palincsar 1998; Mouritsen and Kreiner 2016). On the other hand, the accounting information also cre- ates conditions for the possibility of the emergence of a new interpretation of the organization’s activities, new criteria for action and managerial structures (Dent 1990; Ezzamel and Bourn 1990). Indeed, it is a general belief that greater transpar- ency is a prerequisite for developing more useful accounting information, as well as improved organizational accountability (Roberts 2009). In this case, our assumption is that corporate accounting information is largely embedded within the corpo- rate strategies that are taken by the board of directors in rela- tion to corporate performance and human resource policy that required workplace contextual support for LBGT employees (Webster et al. 2018; Pichler et al. 2018).

It is established that in the case of workplace discrimina- tion, employers limit their available talent pool by discrimi- nating against qualified applicants because of their sexual orientation and/or gender identity (Tilscik 2011). Literature concurs that LGBT employees experience less discrimina- tion when their employers have non-discrimination policies that include sexual orientation and gender identity, and are more amenable to the efforts of strategic outcomes and val- ues of the organization (Schneider et al. 2013). Moreover, companies that are more diverse and inclusive are better able to compete, and have higher levels of innovation and creativity. In a global survey of companies with a turnover of more than $500 million, 85% agreed that workforce diver- sity encourages different perspectives, which drive innova- tion (Forbes Insights 2011). As diversity in the workplace is related to increased innovation, improved strategic decision making, and greater problem-solving skills within business teams (Jackson and Joshi 2004; Francoeur et al. 2008), we can conclude that there is a relationship between workplace diversity management in relation to LGBT employees, accounting and innovation. LGBT-supportive workplace policies can bring about two specific benefits that can have a positive impact on the corporate bottom line (Sears and Mallory 2011): retention of talented employees, and new ideas and innovation generated by drawing upon a diverse workforce with a wide range of characteristics.

Theory and Hypothesis Development

Rawls’ Theory of Justice (1971)

Rawls’ (1971) Theory of Justice has received enormous attention from scholars in a wide range of disciplines (Chapman 1975; Bond and Park 1991). This theory offers a rational accommodation of freedom and equality (Chap- man 1975). It also provides a foundation, based on the idea of fairness, which links the demands of justice to a more

778 M. Hossain et al.

1 3

general mode of reasoning (Sen 1995). The successful inte- gration of the ideas of fairness, rationality, reasonableness, objectivity, and reflective equilibrium show Rawls’ theory of justice to be remarkably effective.

The tolerance of gender inequality is closely related to notions of legitimacy and correctness (Sen 1995). In theory, the State should guarantee the freedom and liberty of all its citizens, human rights, rule of law, participation, fair- ness and justice (Freeman 2006; Petersmann 2008). Firms should provide guarantees on what is essentially described as “equality of opportunity”, i.e. that there should be no dis- crimination (legal or de facto) against any group of people (or minority) based on the values and identity they uphold (De Hart 1994; Gavrilovic 2016).

Rawls consistently points out that the organizational insti- tution has the resources and the opportunity to treat others fairly, and to reward employees not according to a specific socio-demographical factor such as race, religion or gen- der, but according to their merit, based on work competency within the institution (Rawls 1971). Rawls also adds that because the institution has a unique opportunity to advance fairness and meritocratic values in society at large, a greater ethical imperative is placed upon the institution to do so (Rawls 1971). The employees who work in a company also have a platform to advance the fairness principle in the workplace.

Stakeholder Theory

Researchers have long been interested in the process of social change through activist pressure on corporations (Briscoe et al. 2014) in order to create an environment of equality, sometimes with specific reference to sexual orienta- tion and gender-identity policies in the workplace (Pichler et al. 2018). Other researchers investigate corporate disclo- sures and corporate social responsibility (CSR) practices, which are vehicles of communication between corporations and stakeholders. There are many theoretical perspectives on corporate disclosures, many of which are closely tied to stakeholder theory (Donaldson and Preston 1995; Pichler et al. 2018).

The key proposition of stakeholder theory is that firms have a variety of stakeholders, who are affected by or affect firms’ outcomes (Freeman 1984). It is important to include and represent the stakeholders and their interests within the firm because it is not only management who contribute to the success of an organization, but also stakeholders, such as customers, suppliers, and employees, who also make impor- tant contributions to the organization (Baker and Anderson 2010). A limitation of stakeholder theory, as noted by Fiese- ler et al. (2010), is that it does not include ethical guidelines for communication and treating all stakeholders equally. The existing literature provides a rich discussion on how

stakeholders are valued equally (or not) and how their inter- ests are addressed (or not) within the organization (Turnbull et al. 2011; Van Dijk et al. 2012).

The commitment to diversity, equality and inclusive- ness towards LGBT groups is an important aspect of CSR (Snider et al. 2003; Colgan 2011). Therefore, our assumption is that LGBT-supportive policies in a firm are increasingly important as part of workplace diversity management, which should be communicated in a transparent way to all stake- holders, for example through the CSR report. Stakeholders seek to shape equitable employment practices through nego- tiations inside organizations (Bidwell et al. 2013). Human Rights Campaign (HRC), the largest national LGBT civil rights organization in the USA, collected data based on an annual survey to rate U.S. firms on how they treated LGBT employees, with the compliance of pre-determined ques- tions scoring a maximum of 100.4 This suggests that a pres- sure group like HRC, which is an example of a stakeholder, through its compiled Corporate Equality Index, can create value for other stakeholders, and ultimately for firms. From a stakeholder perspective, if a firm implements LGBT-sup- portive workplace policies, it provides a signal to potential employees and the market that the firm is socially respon- sible in terms of anti-discrimination policies and support for diversity (Theodorakopoulos and Budhwar 2015; Pichler et al. 2018).

Hypothesis Development

Innovation is an important determinant of firm-level com- petitiveness (Porter and Stern 2001) and is of interest to many stakeholders (Fang et al. 2014). The literature on the business case for diversity builds on the assertion that diver- sity brings innovation, creativity and problem-solving skills (Østergaard et al. 2011). Research also indicates that LGBT- supportive workplace policies are increasingly important to employees regardless of their own sexual orientation and gender identity (Badgett et al. 2007; Cordes 2012). Although prior studies have linked LGBT-supportive workplace poli- cies to a variety of social imperatives and firm outcomes, i.e. firm performance and stock returns, our study investigates the actual existence of value-addition (Hypothesis). In short, we believe that firms, through effective workplace diversity management (i.e. implementation of LGBT workplace poli- cies) can improve their competitiveness in the market. Con- sistent with prior research on the importance of innovation to stakeholders and the positive effects of LGBT-supportive workplace policies on firm outcomes, we hypothesize:

4 A detailed discussion has been included in the research design.

779Do LGBT Workplace Diversity Policies Create Value for Firms?

1 3

H LGBT workplace policies are positively associated with firm innovation.

Research Design

Our data on corporate workplace policies are collected manually from the Human Rights Campaign (HRC) annual reports (HRC 2015) that provide information on firms’ over- all Corporate Equality Index (CEI) score, sexual orientation non-discrimination policies, gender-identity non-discrimi- nation policies, domestic partner benefits, and transgen- der policies from 2011 to 2014. The HRC survey includes firms from Standard and Poor’s 500, Forbes’ list of the 200 largest privately held firms, and the Fortune 500 largest publicly traded firms.5 Prior research (e.g. Johnston and Malina 2008; Wang and Schwarz 2010; Shan et al. 2017) has used the HRC CEI scores to study the impact of LGBT- supportive corporate policies on firms’ outcomes (such as firm performance). We collect firm innovation, accounting and governance characteristics from Bloomberg that reports data on firm patents, trademarks, copyrights, research and development, board size, and leverage among others. Con- sistent with the previous studies (e.g. Chen et al. 2017), we match both datasets and require sample firm-years to have corporate workplace policies, governance and accounting data in order to be included as part of the sample. Our final sample consists of 398 firms or 1592 firm-year observations.

Empirical Model and Variables

To examine the impact of CEI on firm innovation, we esti- mate the following baseline model:

We measure our dependent variable innovation as the number of patents, trademarks, and copyright (PTC) grants in a year. We choose our measurement of innovation based on several factors. First, this measurement shows the true economic value of innovation that has been created and recognized, through grants of patents, trademarks and

(1)

Innovation i,t = � + �1(Corporate_Equality)i,t

+ �2(Board_Characteristics)i,t

+ �3(Firm_Characteristics)i,t

+ �4

∑

(Industry Effects) i

+ �5

∑

(Year Effects) t + �

i,t

copyrights in a year (Hall et al. 2005). Second, this meas- urement provides precise assessment of the outcome of a firm’s efforts and investment in innovation. Third, our meas- urement is based on innovation outcome rather than input (e.g. research and development expenditure). However, we also employ the input-based measurement of innovation, i.e. research and development expenditures (Ln_R&D) fol- lowing Miller and del Carmen Triana (2009). Other alter- native measures of innovation include patents, trademarks and copyrights per employee, per sales (PTC/Emp, PTC/ Sales), and research and development per sales (R&D/Sales), respectively.

The variable of interest in this study is workplace diver- sity policies. The HRC annual reports on the Corporate Equality Index rates a firm on a scale, ranging from 0 to 100, with 100 being the highest equality.6 The measure includes not only the workplace diversity policies that the firm has in place, but also the training taking place, involvement with the LGBT community, and responsible citizenship of the firm. The rating criteria include points assigned to a firm according to whether its employment policies include sexual orientation, gender identity and diversity training, support- ive gender transition guidelines, domestic partner insurance, and transgender wellness benefits. For example, according to 2014 criteria, a policy such as non-discrimination on sexual orientation earns 15 points, non-discrimination on gender identity earns 15 points, and partner health insurance earns 15 points. The point breakdown for diversity policies is publicly available. The HRC Corporate Equality Index has been commonly used by prior studies to investigate different firm-level outcomes (e.g. Wang and Schwarz 2010; Cook and Glass 2016; Shan et al. 2017). In addition to CEI (the umbrella measure), we also employ a set of dummy variables to measure the individual policies. First, we employ dummy variable (SONDP) that equals 1 if a firm has a sexual orien- tation non-discrimination policy in place and 0 otherwise. Second, we use dummy variable (GINDP) that equals 1 if a firm has a gender-identity non-discrimination policy in place, and 0 otherwise. Third, we assign dummy variable (DPB) that equals 1 if a firm has a domestic partner benefits policy in place, and 0 otherwise. Finally, we employ dummy variable (TG) that equals 1 if a firm has a transgender ben- efits policy in place, and 0 otherwise.

5 In 2002, the HRC (largest national LGBT civil rights organization in the USA) began conducting an annual survey to rate US firms on how they treat their LGBT employees, investors, and consumers. The HRC publishes an annual report on the Corporate Equality Index (CEI).

6 The HRC asks the largest organizations to submit a survey for this index. However, their compliance is voluntary, and organizations can also submit responses, if not asked by HRC. The HRC CEI measure includes a comprehensive set of sexual equality and gender-identity policies and the measure is used by extant literature. However, we acknowledge that it may not be a perfect measure due to the likeli- hood of certain perceptions and limitations of this organization (HRC).

780 M. Hossain et al.

1 3

We use two types of control variables: corporate gov- ernance and firm characteristics. Our selection of control variables is based on prior studies (e.g. Cook and Glass 2016; Chen et al. 2017). Chen et al. (2017) show that cor- porate board characteristics are also important determinants of corporate policies. Therefore, we include a variety of board-specific variables to capture the quality of corpo- rate governance, such as board size (Bsize) (measured as the total number of directors on the board); CEO duality (Duality) serves as proxy for CEO power (a dummy vari- able that equals 1 if the CEO is a chairman of the board, and 0 otherwise); board independence (Bind) is considered an effective monitoring tool (Fama and Jensen 1983) to implement societal shifts towards the LGBT community, as board independence is more likely to promote workplace diversity policies (measured as the number of independent directors divided by the board size); and regular board meet- ings (Ln_bmeeting) which improve the board’s monitoring ability (Rutherford and Buchholtz 2007) and tend to approve diversity-supportive initiatives (measured as the log of the total number of board meetings held in a year).

The firm characteristics include firm-specific variables, such as size of the firm (Firm size), which is measured as the natural log of total assets; ROA, return on assets, which is a measure of financial health; and Leverage, which is measured as total debt (short- and long-term) to total assets. Tobin’s q, a proxy for growth opportunities, is the ratio of the book value of assets minus the book value of equity plus the market value of equity to the book value of assets. Inside ownership (Insideown), a proxy for internal ownership, is measured as shares held by insiders to total outstanding shares. Capex, a proxy for capital expenditure, is measured by total capital expenditure divided by total assets (Table 1).

To test our empirical model, we use ordinary least square (OLS) as the baseline method and include industry (based on two-digit codes of GICS industry sectors) and year effects. The standard errors are corrected for through the clustering of residuals at the firm level to control for heteroscedasticity and within-firm correlation in the residuals (Petersen 2009).7 We also specify 1-year-lagged independent variables by replacing the contemporaneous variables in the regressions to mitigate the endogeneity concerns (Harford et al. 2008).

The underlying rationale is that diversity policies and board characteristics require time to influence firm innovation.

Descriptive Statistics

Table 2 presents the summary statistics. The average of inno- vation measure (PTC) is 420.245 (see Panel A in Table 2). Panel B shows workplace diversity policy measures. CEI has a 58.222 average value; about 89% of firm observations offer a sexual orientation non-discrimination policy (SONDP); 61% of firm observations have a gender-identity non-dis- crimination policy (GINDP); domestic partner benefits are offered by 63% of firm observations (DPB); and only 32% of the firm observations have transgender benefits policies (TG). The CEI score indicates whether a firm fully supports policies or only does so symbolically. In our sample, 39% of all firm observations score 100 points, suggesting that these firms engage in the best practices with their LGBT employees and provide support to the LGBT and non-LGBT workforce in creating a respectful and conducive workplace environment for all. Panel C shows that on average, the board size (Bsize) is 11.051; CEO duality (Duality) has the mean value 0.547; board independence (Bind) is 82.322%; and the average number of board meetings (Ln_bmeeting) is 7.962. Panel D shows that size of the firm (Firm size) has an average value of 4.078; ROA shows 6.177 mean value; and Leverage has an average value of 0.256. Tobin’s q has an average of 1.844; Insideown, a proxy for internal ownership, shows a mean value of 2.084%; and capex has an average of − 0.043. Table 3 shows the number and percentage of firms offering workplace diversity policies. For instance, about 96% of firms (386 firms) have SONDP; 76% of firms (306 firms) have GINDP; DPB and TG are offered by 70% (281 firms) and 49% (198 firms) of firms, respectively. In terms of the combination of policies, 76% of firms (303 firms) have both SONDP and GINDP; 59% of firms (236 firms) have both GINDP and DPB, and only 43% of firms (173 firms) have DPB and TG policies.

Table 4 shows the correlations among variables used in our regression model to check the multicollinearity prob- lem. All the variables measuring workplace diversity are positively correlated with the innovation variable, providing support to our hypothesis. In our sample, the highest correla- tion is among CEI and dummy variables (SONDP, GINDP, DPB, and TG), highlighted in bold. As a general principle, a correlation higher than 0.70 may indicate a multicollinearity issue (Alam et al. 2019; Liu et al. 2014). However, we used highly correlated variables in separate regressions, instead of simultaneously in a model, and therefore, high correla- tion among these variables is not an issue for our study. The remaining variables report no correlation coefficient value higher than 0.70. In addition, to test the potential effect of multicollinearity among these variables, we calculate the

7 Generally, the fixed effects (FE) technique is suggested for panel data estimation in the presence of unobserved firm fixed effects (e.g. Pathan 2009). However, this method (FE) may not be suitable for this study because it requires substantial variation in the variables to generate consistent and unbiased estimates. In our study, the vari- able of interest, CEI lacks substantial variations over time. Therefore, FE may provide imprecise estimation (Wooldridge 2002, p. 286). In addition, FE is not supported in a small number of firms over a lim- ited time period, which is the case of our study (n = 398 and t = 4) (Baltagi et al. 2005, p. 13).

781Do LGBT Workplace Diversity Policies Create Value for Firms?

1 3

variance inflation factor (VIF).8 All the variables have a VIF of less than 1.35, and the overall mean value is 1.28.9 This suggests that multicollinearity is not an issue in the model.

Results and Discussion

Main Results: Workplace Diversity Policies and Innovation

We examined the impact of workplace diversity policies on firm innovation. Panel A in Table 5 presents the results of the baseline regressions using the OLS specification. We start our analysis (Column 1) by regressing the innova- tion (measured as the number of patents, trademarks, and copyrights granted) on the Corporate Equality Index (CEI, the umbrella measure) with industry and year effects. In Columns 2–5, we incorporate the measures of individual policies. For instance, Column 2 shows the effect of sexual

Table 1 Variables’ definitions

Notation Variable name Measure

Panel A: innovation  PTC Patents, trademark and copyrights Total patents, trademarks, and copyrights issued in a year  PTC/emp Patents, trademark and copyrights per employee Total patents, trademarks, and copyrights issued in a year

divided by number of employees  PTC/sales Patents, trademark and copyrights per sale Total patents, trademarks, and copyrights issued in a year

divided by total sales turnover  Ln_R&D Research and development Natural log of total research and development expenditure  R&D/sales Research and development per sales Total research and development expenditures divided by total

sales turnover Panel B: corporate equality  CEI Corporate equality index The score ranging from 0 to 100 with 100 being the top score

based on firm treatment of individuals within the LGBT community employees

 SONDP Sexual orientation non-discrimination policy A dummy variable equals 1 if firm has a sexual orientation non-discrimination policy and 0 otherwise

 GINDP Gender-identity non-discrimination policy A dummy variable equals 1 if firm has a gender-identity non- discrimination policy and 0 otherwise

 DPB Domestic partner benefits A dummy variable equals 1 if firm offers domestic partner benefits to its LGBT constituents and 0 otherwise

 TG Transgender insurance A dummy variable equals 1 if firm offers health benefits to transgender employees and 0 otherwise

Panel C: corporate governance  Bsize Board size The total number of directors on the firm’s board  Duality CEO duality A dummy variable equals 1 if the CEO is also the chairman of

the board and 0 otherwise  Bind Board independence The number of independent directors divided by the board

size  Ln_bmeeting Board meetings Log of the number of board meetings held in a year

Panel D: firm characteristics  Firm size Size of firm Natural log of total assets  ROA Return on assets Firm net income divided by total assets  Leverage Leverage The sum of short- and long-term debt divided by total assets  Tobin’s q Growth opportunities Market value of equity divided the book value of equity  Insideown Insider ownership The percentage of share held by insiders in total outstanding

capital  Capex Capital expenditure Total capital expenditures divided by total assets

8 We do not report VIF results in the interest of brevity. 9 Lardaro (1993) suggests that multicollinearity can cause an issue if VIF exceeds 10.

782 M. Hossain et al.

1 3

orientation non-discrimination (SONDP) policy; Column 3 illustrates the impact of gender-identity non-discrimination (GINDP) policy; Column 4 presents the impact of domes- tic partner benefits (DPB) policy; and Column 5 shows the effect of transgender (TG) policy on innovation.

In all the regressions, results suggest that workplace diversity policies have a significantly positive impact on firm innovation at the 10% or better level of significance. In addi- tion, individual policies also positively affect innovation. The economic significance is also important. For example, an increase in CEI score by one (sample) standard deviation (i.e. using Table 2), increases innovation by approximately 0.38% [CEI (37.738) × 4.234/PTC (420.425) = 0.380]. In summary, there is consistent and statistically strong evidence that workplace diversity policies have a significantly positive

impact on innovation across all the regressions. Overall, these findings support our hypothesis.

For robustness, we use OLS and 1-year-lagged specifica- tion, without controlling for firm and governance character- istics. We report results in Panel B in Table 5. Our results suggest that the CEI and individual policies are positively and significantly associated with innovation across all the Columns (1–10).10 These findings further support our hypothesis. Overall, the results confirm the imperative of Rawls’ theory that if the workforce is treated equally and provided with conducive environment, it creates value for the firm.

Robustness Checks

In this section, we perform a number of sensitivity checks to ensure the robustness of our results. We specify (i) alter- native variables to measure innovation in addition to PTC, including patents, trademarks, and copyrights scaled by number of employees in the firm (PTC/Emp), PTC scaled by sales turnover (PTC/Sales), log of total research and devel- opment expenditure (Ln_R&D), and research and develop- ment expenditure scaled by sales turnover (R&D/Sales); (ii)

Table 2 Descriptive statistics

Refer Table 1 for variables definitions

Variable N Mean SD 1st quartile Median 3rd quartile

Panel A: innovation  PTC 1592 420.425 2030.575 0 0 119.500

Panel B: corporate equality  CEI 1592 58.222 37.738 15 65 100  SONDP 1592 0.889 0.314 1 1 1  GINDP 1592 0.606 0.489 0 1 1  DPB 1592 0.628 0.483 0 1 1  TG 1592 0.321 0.467 0 0 1

Panel C: corporate governance  Bsize 1561 11.051 2.019 10 11 12  Duality 1549 0.547 0.498 0 1 1  Bind 1592 82.322 15.488 80 88.889 90.909  Ln_bmeeting 1592 7.962 3.560 6 7 10

Panel D: firm characteristics  Firm size 1579 4.078 0.480 3.782 4.023 4.323  ROA 1570 6.177 6.177 2.296 5.24 9.325  Leverage 1577 0.256 0.219 0.118 0.225 0.360  Tobin’s q 1561 1.844 1.094 1.146 1.501 2.111  Insideown 1592 2.084 4.384 0.312 0.642 1.708  Capex 1577 − 0.043 0.053 − 0.058 − 0.029 − 0.013

Table 3 Descriptive statistics of workplace policies

Refer Table 1 for variables definitions

Variable No. of firms % of firms

SONDP 386 0.960 GINDP 306 0.760 DPB 281 0.700 TG 198 0.490 SONDP and GINDP 303 0.760 GINDP and DPB 236 0.590 DPB and TG 173 0.430

10 We further apply restriction in our sample by including only those firms which have SONDP but not GINDP, DPB, or TG and run regression analysis. Our results remain consistent.

783Do LGBT Workplace Diversity Policies Create Value for Firms?

1 3

Ta bl

e 4

C or

re la

tio n

m at

ri x

Ta bl

e  4

sh ow

s th

e co

rr el

at io

n m

at ri

x. R

ef er

T ab

le  1

fo r v

ar ia

bl es

d efi

ni tio

ns

V ar

ia bl

es 1

2 3

4 5

6 7

8 9

10 11

12 13

14 15

16

1 PT

C 1.

00 0

2 C

E I

0. 12

8 1.

00 0

3 SO

N D

P 0.

05 1

0. 52

7 1.

00 0

4 G

IN D

P 0.

10 5

0. 82 8

0. 41

5 1.

00 0

5 D

PB 0.

10 5

0. 86 4

0. 42

3 0. 64 6

1. 00

0 6

T G

0. 10

9 0. 70 3

0. 23

1 0.

54 0

0. 48

5 1.

00 0

7 B

si ze

0. 09

2 0.

21 4

0. 11

7 0.

16 0

0. 18

1 0.

13 6

1. 00

0 8

D ua

lit y

0. 07

6 0.

06 1

0. 02

4 0.

03 8

0. 04

3 0.

04 1

0. 04

9 1.

00 0

9 B

in d

0. 06

0 0.

14 0

0. 08

3 0.

09 6

0. 16

7 0.

06 9

0. 11

0 0.

17 3

1. 00

0 10

L n_

bm ee

tin g

0. 05

5 0.

10 6

0. 04

4 0.

06 9

0. 06

8 0.

14 3

0. 12

7 −

0 .0

23 0.

17 5

1. 00

0 11

Fi rm

s iz

e 0.

20 5

0. 24

4 0.

15 0

0. 17

3 0.

18 1

0. 24

0 0.

34 4

0. 15

5 0.

09 8

0. 09

1 1.

00 0

12 R

O A

0. 03

6 −

0 .0

05 −

0 .0

21 −

0 .0

18 0.

00 7

0. 01

9 −

0 .1

24 0.

00 3

− 0

.0 68

− 0

.2 63

0. 00

6 1.

00 0

13 L

ev er

ag e

0. 00

2 −

0 .0

91 0.

05 6

− 0

.0 87

− 0

.0 53

− 0

.0 88

0. 02

2 −

0 .0

28 −

0 .0

50 0.

02 6

− 0

.0 66

− 0

.0 04

1. 00

0 14

To bi

n’ s

q 0.

02 8

0. 05

1 −

0 .0

26 0.

04 5

0. 06

7 0.

01 0

− 0

.1 61

− 0

.0 10

− 0

.0 96

− 0

.2 31

− 0

.1 16

0. 48

6 0.

20 4

1. 00

0 15

In si

de ow

n −

0 .0

52 −

0 .0

10 −

0 .0

19 0.

00 2

− 0

.0 21

0. 02

3 −

0 .0

98 −

0 .1

03 −

0 .2

84 −

0 .0

98 −

0 .1

49 −

0 .0

08 −

0 .0

03 0.

06 2

1. 00

0 16

C ap

ex 0.

04 9

0. 11

4 0.

00 4

0. 06

2 0.

09 1

0. 11

4 0.

09 8

− 0

.0 50

0. 04

3 0.

07 0

0. 00

1 −

0 .0

52 −

0 .1

51 −

0 .0

44 −

0 .0

09 1.

00 0

784 M. Hossain et al.

1 3

Table 5 Workplace diversity policies and innovation

Variable PTC

(1) (2) (3) (4) (5)

Panel A  CEI 4.234***

(2.698)  SONDP 167.841*

(1.985)  GINDP 303.389***

(2.658)  DPB 251.708**

(2.145)  TG 273.557**

(2.253)  Bsize 16.916 23.931 18.985 19.842 23.157

(0.596) (0.846) (0.671) (0.700) (0.820)  Duality 227.220** 241.226** 231.548** 236.867** 236.421**

(2.100) (2.227) (2.141) (2.189) (2.185)  Bind 4.580 6.253 5.296 4.614 5.772

(0.776) (1.064) (0.901) (0.777) (0.983)  Ln_bmeeting 17.571 21.161 18.929 19.917 16.489

(1.069) (1.290) (1.154) (1.215) (0.998)  Firm size 559.639*** 632.386*** 586.689*** 596.459*** 576.474***

(4.425) (5.123) (4.719) (4.787) (4.564)  ROA 7.608 5.503 7.966 6.675 5.786

(0.627) (0.453) (0.656) (0.550) (0.477)  Leverage − 209.622 − 283.986 − 206.062 − 240.680 − 236.763

(− 0.814) (− 1.099) (− 0.800) (− 0.936) (− 0.920)  Tobin’s q − 15.777 6.440 − 13.403 − 9.428 − 3.215

(− 0.223) (0.091) (− 0.189) (− 0.133) (− 0.046)  Insideown − 5.980 − 4.217 − 5.437 − 5.125 − 6.031

(− 0.479) (− 0.338) (− 0.436) (− 0.410) (− 0.482)  Capex 1299.188 1165.873 1336.538 1281.109 1167.524

(1.184) (1.061) (1.217) (1.166) (1.065)  Constant − 729.857*** − 158.642*** − 877.784*** − 858.258*** − 751.883***

(− 3.163) (− 3.708) (− 3.365) (− 3.322) (− 3.173)  Industry effects Y Y Y Y Y  Year effects Y Y Y Y Y  N 1528 1528 1528 1528 1528  Adj. R2 0.189 0.195 0.198 0.187 0.188

Variable PTC

OLS (1) Lagged (2)

OLS (3) Lagged (4)

OLS (5) Lagged (6) OLS (7) Lagged (8) OLS (9) Lagged (10)

Panel B  CEI 6.402*** 5.635***

(4.550) (4.113)  SONDP 306.882* 267.675*

(1.902) (1.669)  GINDP 422.472*** 354.685***

(3.952) (3.377)

785Do LGBT Workplace Diversity Policies Create Value for Firms?

1 3

a different estimation technique (i.e. Tobit regression); (iii) excluding firm governance characteristics; and (iv) control- ling for firm age and location, since both of these variables may affect firm equality policies depending on whether it is a new or established firm and the state location.11

Table 6 reports the results of the above-mentioned specifi- cations based on Eq. 1 in Panels A–D. The year and industry effects are included in all the regression specifications. We also include the control variables, as specified in Table 1, in all the regressions. We find that CEI is positively associated with innovation across the four panels.

Endogeneity Bias

Our results in the previous section may face critique for potential endogeneity bias, due to a causal effect of diversity policies on firm outcomes (i.e. innovation). For instance, in order to create a conducive and creative workplace environ- ment, managers who are sensitive to the benevolent effect of corporate diversity may hire more LGBT employees by cre- ating LGBT-supportive policies that may ultimately foster innovation. Moreover, innovative firms may be responsive to external and societal trends in accepting LGBT employees

into their workforce. Therefore, our main independent vari- able (CEI) may suffer from bias, and as a result, may not be systematically associated with our dependent variable (PTC). In this section, we address endogeneity concerns using the following two approaches: propensity score match- ing, and a dynamic panel data estimation technique—GMM.

Propensity Score Matching

We use the propensity score matching estimator method (PSM) (e.g. Rosenbaum and Rubin 1983; Lennox et  al. 2011) to test the change in the dependent variable (innova- tion), as a result of workplace diversity policies. First, we estimate the probability that a firm is involved in the best practices of having workplace diversity (LGBT-supportive) policies. We run a logistic regression for CEI_dummy (that equals 1 when the CEI score is greater than mean value and 0 otherwise—we treat firms with 1 as part of the treat- ment group and those with 0 as the control group) with the same explanatory variables as specified in Table 5 (i.e. Bsize, ROA, Leverage, etc.).12 Table 7 (Panel A) reports the pre-match logistic regression results. The pseudo R2 for the regression is high (0.221).

Table 5 (continued)

Variable PTC

OLS (1) Lagged (2)

OLS (3) Lagged (4)

OLS (5) Lagged (6) OLS (7) Lagged (8) OLS (9) Lagged (10)

 DPB 388.083*** 375.059*** (3.563) (3.530)

 TG 441.072*** 365.387*** (3.955) (3.310)

 Constant 628.803** − 169.758 1166.111** − 48.687 1167.429** − 87.797 968.733* − 167.019 986.305* − 29.615 (2.032) (− 0.341) (2.248) (− 0.090) (2.315) (− 0.171) (1.917) (− 0.331) (1.959) (− 0.060)

 Industry effects

Y Y Y Y Y Y Y Y Y Y

 Year effects

Y Y Y Y Y Y Y Y Y Y

 N 1592 1591 1592 1591 1592 1591 1592 1591 1592 1591  Adj. R2 0.261 0.258 0.210 0.249 0.257 0.241 0.255 0.219 0.271 0.154

This table presents the regression results of model (1) Innovation

i,t = � + �1(Corporate_Equality)i,t + �2(Board_Characteristics)i,t + �3(Firm_Characteristics)i,t + �4 ∑

(Industry Effects) i + �5

∑

(Year

Effects) t + �

i,t where innovation is measured by patents, trademarks and copyrights grants (PTC). Corporate Equality Index (CEI) is a measure of diversity policies published by Human Rights Campaign (HRC). Panel A Columns 1 presents the results of CEI on innovation. Columns 2–5 present the results of corporate equality individual policies effect on innovation. Panel B shows the using OLS and 1-year-lagged variables on innovation without including control variables. Robust t-statistics are shown in parentheses. Standardized beta coefficients are reported ***, **, *Denote statistical significance at the 1, 5 and 10% level, respectively. All variables are defined in Table 1

11 As a further robustness check, we also specify industry-adjusted PTC, and industry-adjusted sales turnover. Our results remain consist- ent.

12 As a robustness measure, we use the median to form CEI_dummy. Our results remain consistent.

786 M. Hossain et al.

1 3

We also use the nearest neighbour approach to ensure that firms in the treatment and control groups are suffi- ciently identical. Notably, each firm in the treatment group is matched to a firm in the control group with the closest propensity score. In the case of multiple matches, we retain the pair for which the difference between the propensity scores of the two firms is the smallest. We further require that the maximum difference between the propensity score of each firm and its matched peer does not exceed 0.1% in absolute value.13

To verify that the firms in the treatment and groups are indistinguishable in terms of observable characteristics, we

conduct two diagnostic tests. The first test consists of re- estimating the logit model for the post-match sample. The results in post-match (Panel A in Table 7) suggest that no coefficient is statistically significant, indicating that there are no distinguishable trends between the two groups. More- over, coefficients in post-match are smaller in magnitude than those in the pre-match column, indicating the decline in the degree of freedom in the restricted sample. Finally, the pseudo R2 declines from 0.221 to 0.027 for the post- match sample. This suggests that propensity score matching removes all observable differences, other than the differ- ence in the lower scores for workplace diversity policies. The second test examines the differences in the mean of each observable characteristic between the treatment and the con- trol firms. Panel B of Table 7 shows that none of the differ- ences in the observable characteristics between the treatment

Table 6 Robustness analysis

This table presents the results of robustness analyses using alternative variables, alternative method, excluding corporate governance variables, and controlling for firm age and location in four panels (A–D). Industry and year effects are included in all the regressions. Robust t-statistics are shown in parentheses. Standardized beta coefficients are reported ***, **, *Denote statistical significance at the 1, 5 and 10% level, respectively. All variables are defined in Table 1

Variable PTC PTC/Emp PTC/Sales Ln_R&D R&D/Sales

Panel A  OLS regression (N = 1592)   CEI 4.234*** 0.005** 0.001*** 0.089*** 0.020***

(2.698) (2.040) (3.714) (4.001) (2.301)   Controls Yes Yes Yes Yes Yes   Industry effects Yes Yes Yes Yes Yes   Year effects Yes Yes Yes Yes Yes

Panel B  Tobit regression (N = 1592)   CEI 9.017*** 0.001* 0.003*** 0.008*** 0.0403**

(3.004) (1.990) (3.640) (3.167) (2.231)   Controls Yes Yes Yes Yes Yes   Industry effects Yes Yes Yes Yes Yes   Year effects Yes Yes Yes Yes Yes

Panel C  Excluding governance characteristics (N = 1592)   CEI 3.722*** 0.002* 0.001*** 0.009*** 0.021***

(2.514) (1.921) (3.871) (4.572) (4.660)   Other controls Yes Yes Yes Yes Yes   Industry effects Yes Yes Yes Yes Yes   Year effects Yes Yes Yes Yes Yes

Panel D  Controlling for firm age, and location (N = 1592)   CEI 3.872** 0.013** 0.010*** 0.019** 0.019**

(2.164) (2.040) (2.543) (2.020) (2.011)   Controls Yes Yes Yes Yes Yes   Industry effects Yes Yes Yes Yes Yes   Year effects Yes Yes Yes Yes Yes

13 We allow firms to be matched to multiple firms by changing the permissible difference in propensity scores to 1.0% and 0.5% in value; however, we find consistent results (untabulated).

787Do LGBT Workplace Diversity Policies Create Value for Firms?

1 3

and control groups are statistically significant.14 Overall, the diagnostic test suggests that the propensity score matching removes all of the observable differences in explanatory variables, other than CEI.

We report propensity score matching estimates in Panel C (Table 7). The results show that there are significant dif- ferences (significant at the 1% level) in innovation between firms with LGBT-supportive policies and those without (symbolic). These findings suggest that an increase in inno- vation is attributable to the systematic difference in the workplace diversity policies.

Dynamic Panel Data Estimation

Second, we use dynamic panel estimation, which accounts for unobserved heterogeneity, simultaneity, and the dynamic relation between the CEI and past innovation (Wintoki et al. 2012; Abdallah et al. 2015; Atif et al. 2019). The two- step ‘system GMM’ (Arellano and Bover 1995; Blundell and Bond 1998) uses first-differenced variables as instru- ments for the equations in levels.15 The estimations are robust to undetected heterogeneity, causality problems and dynamic endogeneity. The stability of the dynamic system GMM depends on two major conditions. The first condi- tion is the serial independence of the residuals, where the first difference residuals should be serially correlated (AR1) by the means of their structure. However, residuals in the second difference should not be serially correlated (AR2). The second condition is the validity of instruments used in dynamic estimation. The Hansen J-statistic of over-identify- ing restrictions tests the null hypothesis of the instruments’ validity. The insignificance of the Hansen J-statistic indi- cates the validity of instruments in the respective estima- tions. Finally, the number of instruments (i.e. 20) used in the model is less than that in the panel (i.e. 398), which adds to the consistency of the Hansen J-statistic.

The diagnostic test in Table 8 shows that the model is statistically well fitted for first-order autocorrelation (AR1), insignificant for second-order autocorrelation (AR2), and for the Hansen J-statistic of over-identifying restrictions. The interpretation of the parameters on CEI and innovation remains quantitatively the same as in Table 5. For instance, CEI positively affects innovation. Hence, the system GMM supports our results, even after controlling for undetected heterogeneity, simultaneity bias and dynamic endogeneity.

Table 7 Propensity score matching

The table presents the results of endogeneity test, propensity score matching in three panels. Panel A shows the pre-match- and post- match results, and Panel B presents the difference in firm character- istics for the matched sample, and Panel C reports propensity score estimators. Standardized beta coefficients are reported ***, **, *Denote statistical significance at the 1, 5 and 10% levels, respectively. All variables are defined in Table 1

Variable CEI_dummy

Pre-match Post- match

Panel A  Bsize 0.160*** 0.004

(4.493) (0.143)  Duality 0.319** − 0.078

(2.462) (− 0.692)  Bind 0.031*** 0.008

(4.222) (1.229)  Ln_bmeeting 0.064*** 0.029

(3.054) (1.728)  Firm size 1.320*** 0.269

(8.264) (1.125)  ROA − 0.056*** 0.012

(− 3.647) (1.051)  Leverage − 1.202*** − 0.064

(− 3.546) (− 0.209)  Tobin’s q 0.474*** 0.103

(5.131) (1.511)  Insideown 0.022 0.001

(1.397) (0.074)  Capex − 3.377** − 4.499

(− 2.391) (− 1.710)  Constant − 9.680*** − 1.906

(− 9.219) (− 1.725)  Industry effects Y Y  Year effects Y Y  N 1528 828  Pseudo R2 0.221 0.027

Variable Treatment Control Difference t-stat

Panel B: difference in firm characteristics  Bsize 11.464 11.158 0.306 1.991  Duality 0.584 0.608 − 0.024 − 0.611  Bind 85.532 85.192 0.341 0.441  Ln_bmeeting 8.413 8.071 0.341 1.360  Firm size 4.174 4.122 0.052 1.450  ROA 6.186 5.916 0.270 0.510  Leverage 0.234 0.236 − 0.001 − 0.070  Tobin’s q 1.916 2.108 − 0.191 − 1.160  Insideown 1.995 2.087 − 0.091 − 0.240  Capex − 0.038 − 0.043 0.004 1.240

Panel C: propensity score estimator  PTC 653.386 426.312 227.074*** 2.139

14 Mean difference between the treatment and the control group is based on the average treatment effect on the treated (ATT). 15 The system GMM estimations are based on (Roodman 2006), using Stata module ‘xtabond2’. Refer to Roodman (2006) and Pathan (2009) for details on dynamic panel data estimations.

788 M. Hossain et al.

1 3

Workplace Diversity Policies and Firm Performance

In this section, we investigate whether workplace diversity policies affect a firm’s performance. If firms tend towards implementing LGBT-supportive policies, in line with external and societal norms, then one may expect a positive

impact on firm value and performance consistent with the responsible firm hypothesis. We measure firm performance using Tobin’s q and ROS (net income scaled by sales turno- ver) following prior studies (e.g. Liu et al. 2014). We esti- mate the following regression model to examine the impact of CEI and PTC on firm performance.

The dependent variable in this equation (firm perf) is measured by Tobin’s q and ROS. The independent variables are the same as specified in Table 1 and discussed in the above sections, except for the interaction between innova- tion and diversity policies (Corporate_Equality × Innova- tion) which is the main variable of interest. Table 9 reports the results using OLS and 1-year-lagged specifications for Tobin’s q and ROS, respectively. The interaction term shows a positive effect on firm performance. These results suggest a positive effect of workplace diversity policies and innovation on firm performance, which is significant at the 5% or better level of significance. These findings sug- gest that LGBT-supportive policies have a positive impact on firms’ innovation that ultimately increases firm perfor- mance, indicating workplace diversity policies as one of the drivers of firm performance.

Conclusion

Innovation is an important function of both accounting and management because it is linked to business performance and serves to accommodate market uncertainties a firm may face in its competitive environment. In this paper, using HRC data for large US firms during 2011–2014, we investigate the effect of CEI score on firm innovation and find a significantly positive relationship. We also find a significantly positive relationship between individual anti-discriminatory policies and firm innovation. More specifically, we show that workplace diversity policies positively affect firm innovation, which ultimately leads to higher firm performance. Our findings are robust to alter- native econometric specifications, alternative measures of innovation, and to individual policies. In a finer analysis, propensity score matching (PSM), and dynamic panel data estimation (system GMM), we further strengthen our findings that workplace diversity policies have a posi- tive causal effect on innovation, and that these results are not due to omitted variables or causality issues. We have

(2)

Firm perf i,t = � + �1(Corporate_Equality)i,t

+ �2(Innovation)i,t + �3(Corporate_Equality × Innovation)i,t

+ �4(Board_Characteristics)i,t + �5(Firm_Characteristics)i,t

+ �6

∑

(Industry Effects) i + �7

∑

(Year Effects) t + �

i,t

Table 8 Generalized method of moments

The table presents the results of dynamic panel data estimation (sys- tem GMM). Industry and year effects are included in the regression. Standardized beta coefficients are reported ***, **, *Denote statistical significance at the 1, 5 and 10% level, respectively. All variables are defined in Table 1

Variable PTC

CEI 11.144** (2.091)

Bsize − 19.273 (− 1.232)

Duality 165.640 (1.463)

Bind − 17.379 (− 0.245)

Ln_bmeeting − 1.926 (− 0.014)

Firm size 16.526* (1.885)

ROA 4.575 (0.068)

Leverage 141.283* (1.945)

Tobin’s q − 33.533 (− 1.181)

Insideown − 30.094 (− 0.168)

Capex 115.934** (1.979)

Constant 182.985 (0.029)

Industry effects Y Year effects Y N 1528 Model fits  AR I 0.176***

(2.860)  AR II 0.125

(0.550)  Wald F-statistics 16.100***

(0.030)  No. of instruments 20  Hansen J-statistic 2.271

(0.518)

789Do LGBT Workplace Diversity Policies Create Value for Firms?

1 3

used Rawls’ Theory of Justice (1971) to argue that firms have a greater ethical imperative to advance fairness and meritocratic values among employees and society at large. In addition, we have used the imperative of stakeholder theory to argue that LGBT-supportive policies in the firms are an important part of workplace diversity management, which should be communicated to all stakeholders. We

conclude this paper with the Conclusion section by laying stress on the importance of innovation in firm performance and its close association to workplace diversity manage- ment higlighting implications for workplace strategies, limitations and scope for future studies.

Our paper has important implications not only for firms’ workplace strategies, but also for developing diversity poli- cies aimed at increasing innovation that may ultimately lead to better firm performance. The implications of our study are based on macro- (overall CEI score) and micro (individual diversity policies)-perspectives. Our results suggest that adoption of anti-discriminatory policies that aim to provide equal employment opportunities can have real economic benefits in terms of higher innovation and higher firm per- formance. Our findings are timely and important because, in the U.S., there is an ongoing debate around banning sexual orientation discrimination in the workplace across the coun- try. Our findings may also trigger debate in similar markets around the globe to enhance disclosure on firms’ LGBT workplace policies to gain comparable economic benefits.

Like most research, our study is subject to potential limi- tations. In particular, it is limited to a consideration of large firms in the U.S. This focus limits the ability to generalize to small firms and those outside the U.S., or not-for-profit organizations. Future studies may investigate this issue in the context of different types of organizations, and may shift the focus towards other developed or developing countries, since workplace diversity policies are increasingly advocated around the globe.

Compliance with Ethical Standards

Conflict of interest All four authors declare that they have no conflicts of interest.

Ethical Approval This article does not contain any studies with human participants or animals performed by any of the authors.

References

Abdallah, W., Goergen, M., & O’Sullivan, N. (2015). Endogeneity: How failure to correct for it can cause wrong inferences and some remedies. British Journal of Management, 26(4), 791–804.

Alam, M. S., Atif, M., Chien-Chi, C., & Soytaş, U. (2019). Does cor- porate R&D investment affect firm environmental performance? Evidence from G-6 countries. Energy Economics, 78, 401–411.

Arellano, M., & Bover, O. (1995). Another look at the instrumen- tal variable estimation of error-components models. Journal of Econometrics, 68(1), 29–51.

Atif, M., Huang, A., & Liu, B. (2019). Effect of say on pay on CEO compensation and spill-over effect on corporate cash holdings: Evidence from Australia. Pacific-Basin Finance Journal.

Table 9 Workplace diversity policies, innovation and firm perfor- mance

This table presents the results of innovation, CEI (interaction term) and firm performance measured by Tobin’s q and ROS as dependant variables using OLS and 1-year-lagged specifications, respectively. Industry and year effects are included in all the regressions. Robust t-statistics are shown in parentheses. Standardized beta coefficients are reported ***, **, *Denote statistical significances at the 1, 5 and 10% levels, respectively. All control variables are defined in Table 1

Variable Tobin’s q Lagged Tobin’s q

ROS Lagged ROS

CEI 0.001* 0.003*** 0.002** 0.013** (1.990) (5.110) (2.190) (2.005)

PTC 0.001** 0.001** 1.320** − 1.200 (2.110) (2.041) (2.190) (0.090)

CEI × PTC 1.320** 1.220*** 1.580** 3.001** (2.181) (2.480) (2.120) (2.105)

Bsize − 0.088*** − 0.044*** 0.006 0.001 (− 7.010) (− 3.540) (0.530) (0.081)

Duality 0.018 0.058 − 0.008 0.004 (0.038) (1.220) (− 0.180) (0.670)

Bind − 0.002 − 0.002 − 0.001 − 0.001* (− 0.001) (− 0.900) (− 0.720) (− 0.600)

Ln_bmeeting − 0.047*** − 0.020*** 0.001*** 0.001* (− 6.790) (− 2.730) (2.410) (1.192)

Firm size − 0.913*** − 0.303*** − 0.087*** − 0.086*** (− 13.921) (− 5.530) (− 12.250) (− 9.660)

ROA 0.201*** 0.120*** 0.011*** 0.006*** (9.109) (14.130) (20.73) (9.791)

Leverage 0.833*** 1.240*** 0.074*** 0.066*** (7.651) (11.160) (6.570) (4.661)

Tobin’s q − 0.017*** − 0.001 (− 5.380) (− 0.013)

Insideown − 0.004 0.008 0.001** 0.001 (− 0.080) (1.640) (2.130) (1.540)

Capex − 0.628 0.883* − 0.036 − 0.517 (− 1.131) (1.960) (− 0.760) (− 0.849)

Constant 5.001*** 2.712*** 1.054** 0.385*** (10.970) (9.071) (2.042) (6.930)

Industry effects

Y Y Y Y

Year effects Y Y Y Y N 1528 1327 1528 1327 Adj. R2 0.380 0.423 0.440 0.344

790 M. Hossain et al.

1 3

Badgett, M. L. (1995). The wage effects of sexual orientation discrimi- nation. Industrial and Labor Relations Review, 48(4), 726–739.

Badgett, M. V., Lau, H., Sears, B., & Ho, D. (2007). Bias in the work- place: Consistent evidence of sexual orientation and gender iden- tity discrimination.

Baker, H. K., & Anderson, R. (Eds.). (2010). Corporate governance: A synthesis of theory, research, and practice (Vol. 8). New York: Wiley.

Baltagi, B. H., Bratberg, E., & Holmås, T. H. (2005). A panel data study of physicians’ labor supply: the case of Norway. Health Economics, 14(10), 1035–1045.

Barbulescu, R., & Bidwell, M. (2013). Do women choose different jobs from men? Mechanisms of application segregation in the market for managerial workers. Organization Science, 24(3), 737–756.

Beckley, H. R. (1986). A Christian affirmation of Rawls’s idea of jus- tice as fairness-part II. Journal of Religious Ethics, 14(2), 229.

Bidwell, M., Briscoe, F., Fernandez-Mateo, I., & Sterling, A. (2013). The employment relationship and inequality: How and why changes in employment practices are reshaping rewards in organi- zations. Academy of Management Annals, 7(1), 61–121.

Blundell, R., & Bond, S. (1998). Initial conditions and moment restric- tions in dynamic panel data models. Journal of Econometrics, 87(1), 115–143.

Bonaventura, L., & Biondo, A. E. (2016). Disclosure of sexual orienta- tion in the USA and its consequences in the workplace. Interna- tional Journal of Social Economics, 43(11), 1115–1123.

Bond, D., & Park, J. C. (1991). An empirical test of Rawls’s theory of justice: a second approach, in Korea and the United States. Simulation & Gaming, 22(4), 443–462.

Briscoe, F., Chin, M. K., & Hambrick, D. C. (2014). CEO ideology as an element of the corporate opportunity structure for social activists. Academy of Management Journal, 57(6), 1786–1809.

Burchell, S., Clubb, C., Hopwood, A., Hughes, J., & Nahapiet, J. (1980). The roles of accounting in organizations and society. Accounting, Organizations and Society, 5(1), 5–27.

Chapman, J. W. (1975). Rawls’ theory of justice. American Political Science Review, 69(2), 588–593.

Chen, J., Leung, W. S., & Goergen, M. (2017). The impact of board gender composition on dividend payouts. Journal of Corporate Finance, 43, 86–105.

Cohen, M. A. (2010). The narrow application of Rawls in business ethics: A political conception of both stakeholder theory and the morality of markets. Journal of Business Ethics, 97(4), 563–579.

Colgan, F. (2011). Equality, diversity and corporate responsibility: Sexual orientation and diversity management in the UK private sector. Equality, Diversity & Inclusion, 30(8), 719–734.

Cook, A., & Glass, C. (2016). Do women advance equity? The effect of gender leadership composition on LGBT-friendly policies in American firms. Human Relations, 69(7), 1431–1456.

Cordes, C. L. (2012). The business case for offering domestic partner benefits. Compensation & Benefits Review, 44(2), 110–116.

De Hart, J. S. (1994). Equality challenged: Equal rights and sexual difference. Journal of Policy History, 6(1), 40–72.

Dent, J. F. (1990). Strategy, organization and control: Some possi- bilities for accounting research. Accounting, Organizations and Society, 15(1–2), 3–25.

Donaldson, T., & Preston, L. E. (1995). The stakeholder theory of the corporation: Concepts, evidence, and implications. Academy of Management Review, 20(1), 65–91.

Drydakis, N. (2009). Sexual orientation discrimination in the labour market. Labour Economics, 16(4), 364–372.

Ezzamel, M., & Bourn, M. (1990). The roles of accounting information systems in an organization experiencing financial crisis. Account- ing, Organizations and Society, 15(5), 399–424.

Fama, E. F., & Jensen, M. C. (1983). Separation of ownership and control. The Journal of Law and Economics, 26(2), 301–325.

Fang, V. W., Tian, X., & Tice, S. (2014). Does stock liquidity enhance or impede firm innovation? The Journal of Finance, 69(5), 2085–2125.

Fieseler, C., Fleck, M., & Meckel, M. (2010). Corporate social respon- sibility in the blogosphere. Journal of Business Ethics, 91(4), 599–614.

Francoeur, C., Labelle, R., & Sinclair-Desgagné, B. (2008). Gender diversity in corporate governance and top management. Journal of Business Ethics, 81(1), 83–95.

Forbes Insights. (2011). Global Diversity and inclusion Fostering Innovation Through a Diverse Workforce. Retrieved April 2, 2011 from https ://i.forbe simg.com/forbe sinsi ghts/Study PDFs/ Innov ation _Throu gh_Diver sity.pdf.

Freeman, R. E. (1984). Strategic management: A stakeholder approach. Boston, MA: Pitman.

Freeman, S. (2006). The law of peoples, social cooperation, human rights, and distributive justice. Social Philosophy and Policy, 23(1), 29–68.

Gates, G. J. (2011). How many people are lesbian, gay, bisexual and transgender? Los Angeles, CA: Williams Institute, University of California, Los Angeles School of Law. Retrieved March 26, 2018 from http://willi amsin stitu te.law.ucla.edu/wp-conte nt/uploa ds/Gates -How-Many-Peopl e-LGBT-Apr-2011.pdf.

Gates, G. J. (2012). LGBT identity: A demographer’s perspective. Loy- ola of Los Angeles Law Review, 45, 693–714.

Gavrilovic, V. (2016). Toleration and pluralism: an analysis of the con- temporary debate on toleration and its practical implications on LGBT social movements and advocancy groups.

Grobler, B. R., Moloi, K. C., Loock, C. F., Bisschoff, T. C., & Mestry, R. J. (2006). Creating a school environment for the effective man- agement of cultural diversity. Educational Management Adminis- tration & Leadership, 34(4), 449–472.

Gupta, A., Briscoe, F., & Hambrick, D. C. (2017). Red, blue, and pur- ple firms: Organizational political ideology and corporate social responsibility. Strategic Management Journal, 38(5), 1018–1040.

Hall, B. H., Jaffe, A., & Trajtenberg, M. (2005). Market value and pat- ent citations. RAND Journal of Economics, 16–38.

Harford, J., Mansi, S. A., & Maxwell, W. F. (2008). Corporate gov- ernance and firm cash holdings in the US. Journal of Financial Economics, 87(3), 535–555.

Hopwood, A. G. (1987a). The archeology of accounting systems. Accounting, Organizations and Society, 12(3), 207–234.

Hopwood, A. G. (1987b). Accounting and gender: An introduction. Accounting, Organizations and Society, 12(1), 65–69.

Human Rights Campaign. (2015). LGBTQ Equality at the Fortune 500. Retrieved from https ://www.hrc.org/resou rces/lgbt-equal ity- at-the-fortu ne-500.

Jackson, S. E., & Joshi, A. (2004). Diversity in social context: A multi- attribute, multilevel analysis of team diversity and sales perfor- mance. Journal of Organizational Behavior, 25(6), 675–702.

Johnson, I., & Cooper, D. (2015). LGBT diversity: Show me the busi- ness case. Retrieved from https ://www.outno wcons ultin g.com/ marke t-repor ts/lgbt-diver sity-show-me-the-busin ess-case-repor t.aspx.

Johnston, D., & Malina, M. A. (2008). Managing sexual orientation diversity: The impact on firm value. Group & Organization Man- agement, 33(5), 602–625.

Kelly, G. M. (2016). Transitioning into modern society: Why the law is not keeping up with gender identity. JL & Com., 35, 101.

King, E. B., & Cortina, J. M. (2010). The social and economic imper- ative of lesbian, gay, bisexual, and transgendered supportive

791Do LGBT Workplace Diversity Policies Create Value for Firms?

1 3

organizational policies. Industrial and Organizational Psychol- ogy, 3(1), 69–78.

Lardaro, L. (1993). Applied econometrics. New York: Harper Collins. Lennox, C. S., Francis, J. R., & Wang, Z. (2011). Selection models

in accounting research. The Accounting Review, 87(2), 589–616. Liddle, B. J., Luzzo, D. A., Hauenstein, A. L., & Schuck, K. (2004).

Construction and validation of the lesbian, gay, bisexual, and transgendered climate inventory. Journal of Career Assessment, 12(1), 33–50.

Liu, Y., Wei, Z., & Xie, F. (2014). Do women directors improve firm performance in China? Journal of Corporate Finance, 28, 169–184.

Lloren, A., & Parini, L. (2017). How LGBT-supportive workplace policies shape the experience of lesbian, gay men, and bisexual employees. Sexuality Research and Social Policy, 14(3), 289–299.

Miller, T., & del Carmen Triana, M. (2009). Demographic diversity in the boardroom: Mediators of the board diversity–firm per- formance relationship. Journal of Management Studies, 46(5), 755–786.

Mouritsen, J., & Kreiner, K. (2016). Accounting, decisions and prom- ises. Accounting, Organizations and Society, 49, 21–31.

Omankhanlen, A. E., & Ogaga-Oghene, J. (2011). The impact of workforce diversity on organizational effectiveness: A study of a Nigerian bank. Annals of the University of Petroşani, Economics, 11(3), 93–110.

Østergaard, C. R., Timmermans, B., & Kristinsson, K. (2011). Does a different view create something new? The effect of employee diversity on innovation. Research Policy, 40(3), 500–509.

Ozeren, E. (2014). Sexual orientation discrimination in the workplace: A systematic review of literature. Procedia-Social and Behavioral Sciences, 109, 1203–1215.

Ozturk, M. B., & Tatli, A. (2016). Gender identity inclusion in the work- place: Broadening diversity management research and practice through the case of transgender employees in the UK. The Interna- tional Journal of Human Resource Management, 27(8), 781–802.

Palincsar, A. S. (1998). Social constructivist perspectives on teaching and learning. Annual Review of Psychology, 49(1), 345–375.

Pathan, S. (2009). Strong boards, CEO power and bank risk-taking. Journal of Banking & Finance, 33(7), 1340–1350.

Petersen, M. A. (2009). Estimating standard errors in finance panel data sets: Comparing approaches. Review of Financial Studies, 22(1), 435–480.

Petersmann, E. U. (2008). Human rights, international economic law and ‘constitutional justice’. European Journal of International Law, 19(4), 769–798.

Pichler, S., Blazovich, J. L., Cook, K. A., Huston, J. M., & Strawser, W. R. (2018). Do LGBT-supportive corporate policies enhance firm performance? Human Resource Management, 57(1), 263–278.

Pichler, S., Ruggs, E., & Trau, R. (2017). Worker outcomes of LGBT- supportive policies: A cross-level model. Equality, Diversity and Inclusion: An International Journal, 36(1), 17–32.

Porter, M. E., & Stern, S. (2001). Innovation: Location matters. MIT Sloan Management Review, 42(4), 28.

Priola, V., Lasio, D., De Simone, S., & Serri, F. (2014). The sound of silence. Lesbian, gay, bisexual and transgender discrimination in ‘inclusive organizations’. British Journal of Management, 25(3), 488–502.

Ragins, B. R. (2008). Disclosure disconnects: Antecedents and con- sequences of disclosing invisible stigmas across life domains. Academy of Management Review, 33(1), 194–215.

Ragins, B. R., & Cornwell, J. M. (2001). Pink triangles: Antecedents and consequences of perceived workplace discrimination against gay and lesbian employees. Journal of Applied Psychology, 86(6), 1244.

Rawls, J. (1971). A theory of justice. Cambridge, MA: Harvard Uni- versity Press.

Rhodes, C. (2017). Ethical praxis and the business case for LGBT diversity: Political insights from judith butler and emmanuel levi- nas. Gender, Work & Organization, 24(5), 533–546.

Roberts, J. (2009). No one is perfect: The limits of transparency and an ethic for ‘intelligent’ accountability. Accounting, Organizations and Society, 34(8), 957–970.

Roodman, D. (2006). How to do Xtabond2: An introduction to differ- ence and system GMM in Stata (Center for Global Development Working Paper No. 103). Retrieved from https ://ssrn.com/abstr act=98294 3.

Rosenbaum, P. R., & Rubin, D. B. (1983). The central role of the pro- pensity score in observational studies for causal effects. Biom- etrika, 70, 41–55.

Rutherford, M. A., & Buchholtz, A. K. (2007). Investigating the rela- tionship between board characteristics and board information. Corporate Governance: An International Review, 15(4), 576–584.

Schneider, B., Ehrhart, M. G., & Macey, W. H. (2013). Organizational climate and culture. Annual Review of Psychology, 64, 361–388.

Sears, B., & Mallory, C. (2011). Economic motives for adopting LGBT-related workplace policies. Retrived from: http://willi amsin stitu te.law.ucla.edu/wp-conte nt/uploa ds/Mallo ry-Sears -Corp-State ments -Oct20 11.pdf.

Sen, A. (1995). Gender inequality and theories of justice. In M. C. Nuss- baum & J. Glover (Eds.), Women, culture and development: A study of human capabilities (pp. 259–273). Oxford: Clarendon Press.

Shan, L., Fu, S., & Zheng, L. (2017). Corporate sexual equality and firm performance. Strategic Management Journal, 38(9), 1812–1826.

Snider, J., Hill, R. P., & Martin, D. (2003). Corporate social responsi- bility in the 21st century: A view from the world’s most successful firms. Journal of Business Ethics, 48(2), 175–187.

Stambaugh, C. T., & Carpenter, F. W. (1992). The roles of accounting and accountants in executive information systems. Accounting Horizons, 6(3), 52.

Theodorakopoulos, N., & Budhwar, P. (2015). Guest editors’ introduc- tion: Diversity and inclusion in different work settings: Emerg- ing patterns, challenges, and research agenda. Human Resource Management, 54(2), 177–197.

Tilscik, A. (2011). Pride and prejudice: Employment discrimination against openly gay men in the United States. Journal of Sociol- ogy, 117, 586–626.

Turnbull, H., Greenwood, R., Tworoger, L., & Golden, C. (2011). The inclusion skills measurement profile: Validating an assessment for identification of skill deficiencies in diversity and inclusion. Journal of Organizational Culture, Communications and Conflict, 15(1), 11.

Van Dijk, H., van Engen, M., & Paauwe, J. (2012). Reframing the busi- ness case for diversity: A values and virtues perspective. Journal of Business Ethics, 111(1), 73–84.

Wang, P., & Schwarz, J. L. (2010). Stock price reactions to GLBT non-discrimination policies. Human Resource Management, 49(2), 195–216.

Webster, J. R., Adams, G. A., Maranto, C. L., Sawyer, K., & Thor- oughgood, C. (2018). Workplace contextual supports for LGBT employees: A review, meta-analysis, and agenda for future research. Human Resource Management, 57(1), 193–210.

Wintoki, M. B., Linck, J. S., & Netter, J. M. (2012). Endogeneity and the dynamics of internal corporate governance. Journal of Finan- cial Economics, 105(3), 581–606.

Wooldridge, J. M. (2002). Inverse probability weighted M-estimators for sample selection, attrition, and stratification. Portuguese Eco- nomic Journal, 1(2), 117–139.

Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Reproduced with permission of copyright owner. Further reproduction prohibited without permission.