INTRODUCTION
LGB is the acronym used for lesbian, gay and bisexual. The American Psychological Association (2008) defines lesbians as women who are attracted to women, gay men as men who are attracted to men, and bisexual as men or women who attracted to both genders. A number of LGB individuals in Thailand who disclose their sexual orientation, or ‘come out,’ are rapidly increasing almost everywhere in Thailand, especially in the big cities. Thailand is often referred to as ‘Gay Paradise’ (Jackson, Reference Jackson1999; UNDP & USAID, 2014) as it is seen to be friendly for the LGB community. LGB people are also seen regularly in the media. Archavanitkul (Reference Archavanitkul2011) found 1,345 articles about LGB from online news in the Thai Health Promotion Foundation’s database from 1997 to 2007. Although the exact number of the LGB population could not be easily found, the number of LGB in general was estimated to vary from 1.5 to 5.0% of the total population (BBC, 2010; Statistics Canada, 2014; Ward, Dahlhamer, Galinsky, & Joestl, Reference Ward, Dahlhamer, Galinsky and Joestl2014; Spitzer, Reference Spitzer2015). Therefore, the estimated number of LGB in Thailand would be around 0.97–3.24 million given the total population of 64.87 million in Thailand (Institute for Population and Social Research, 2014).
According to Center for Disease Control, National Health Statistics Report, 1.6% of US population were gay or lesbian and 0.7% were bisexual in 2014 (Ward et al., Reference Ward, Dahlhamer, Galinsky and Joestl2014). In 2014, 1.7% of populations were gay or lesbians and 1.3% of Canadians were bisexual (Statistics Canada, 2014). In UK, the Office for National Statistics’ survey showed that 1.0% of British were gay or lesbian, and 0.5% of population were bisexual (BBC, 2010). Approximately 5% of Japanese were Lesbian, Gay, Bisexual, and Transgendered (LGBT) (Spitzer, Reference Spitzer2015).
In the United States, LGB employees are found to be between 4 and 17% of the American workforce (Gonsiorek & Weinrich, Reference Gonsiorek and Weinrich1991). Large US corporations have discovered that encouraging employee diversity can boost their bottom lines. For instance, Kleiner (Reference Kleiner2001) found that providing a supportive working environment to LGB employees resulted in attracting a large number of LGB consumers to the business. Currently, 91% of Fortune 500 companies adopted nondiscrimination policies which specifically include sexual orientation (Human Rights Campaign, n.d.).
In this world of businesses moving toward diversified social contexts and with the number of disclosed LGB employees, some research studies have shown that sexual orientation discrimination in workplaces still exists. Previous research consistently finds that many LGB workers live in fear of discrimination in the workplace (e.g., Hall, Reference Hall1986; Boatwright, Gilbert, Forest, & Ketzenberger, Reference Boatwright, Gilbert, Forest and Ketzenberger1996; Irwin, Reference Irwin1998). This issue has been recognized and addressed in many countries in order to gain equality for LGB individuals. A number of empirical studies reveal the benefits of supportive work environment for LGB employees and their employers. For example, LGB employees who find it easier to ‘come out’ at work have higher job performance, and higher job and life satisfaction than LGB employees who are not comfortable disclosing their sexual orientation at work (e.g., Griffith & Hebl, Reference Griffith and Hebl2002; Ragins, Singh, & Cornwell, Reference Ragins, Singh and Cornwell2007; Huffman, Watrous-Rodriguez, & King, Reference Huffman, Watrous-Rodriguez and King2008). Badgett, Durso, Kastanis, and Mallory (Reference Badgett, Durso, Kastanis and Mallory2013) have analyzed 36 research studies of LGB-supportive workplace and concluded that many of them showed positive relationships between LGB-supportive employment policies and business outcomes. Studies on sustainable organizations and diversity emphasize the importance of integrating sexual orientation diversity into organizations’ overall policies (Trau & Härtel, Reference Trau and Härtel2007). Hence, organizations must work on providing equality for LGB employees in their workforce diversity efforts.
There is an increasing body of research examining how work-related support can improve LGB employees’ work and life experiences and company outcomes. However, literature in this area is largely derived from the LGB workforce in Western or developed countries (e.g., Trau & Härtel, Reference Trau and Härtel2007; Huffman, Watrous-Rodriguez, & King, Reference Huffman, Watrous-Rodriguez and King2008). Although Thailand is considered open to LGB and has been referred to as ‘Gay Paradise,’ it is contended to be ‘tolerant but unaccepting’ (Jackson, Reference Jackson1999). In 2013, the president of LGBT rights group in Thailand mentioned to Phuket news that
‘There is a quote that Thai society unofficially accepts, but officially rejects, gay and lesbian people’ ‘I think it’s true that Thai people can only accept gay and lesbian people in superficial ways, such as the way they act or dress. But when it comes to the meaningful circumstances, Thai people tend to be biased against them.’
(Mosbergen, Reference Mosbergen2015)
According to the report from International Labor Organization from 2005 to 2014 (Suriyasarn, Reference Suriyasarn2014), the study revealed that there is no real acceptance of LGBT persons in Thai society although there are more public space for them. The study reported that ‘many gays and lesbians still feel the need to hide their true sexuality at home, at school and in the workplace’ (Suriyasarn, Reference Suriyasarn2014). A poll by the National Institute of Development Administration (2013) on ‘How Thai people think about third gender,’ found that 88.9% from over a thousand Thai people surveyed accepted friends or colleagues who are ‘third gender’ if they do not ‘make any trouble.’ The National Human Rights Commission of Thailand reported that there are several barriers to employment for LGB in Thailand, starting from job application, job interview, psychological test, legal identify verification. In Thailand, the public sectors are the least LGB friendly for employment. Especially civil service, male and female employees were expected to behave more traditionally. In academia, teachers and professors are presumed to be role model for students. The study reported that masculine gay men and heterosexual men have the same chance to employment. There may be some barriers to employment for gay not masculine, therefore, gay men may hide their sexuality and come out later at the workplace (Suriyasarn, Reference Suriyasarn2014, Reference Suriyasarn2015). Many LGB do not disclose themselves to avoid rejection at workplace in Thailand. The unique aspect of workplace in Thailand is highly collectivistic culture which focuses on harmony in relationships with supervisor, coworkers among the group (Engelen, Flatten, Thalmann, & Brettel, Reference Engelen, Flatten, Thalmann and Brettel2014). In addition to differences in supportive working environments for LGB employees, there could also be differences in the social security systems and enforcement of nondiscrimination legislation between developed and developing countries. In September 2015, Gender Equality Act, which specifically prohibited discrimination against one ‘of a different appearance from his/her own sex by birth’ was enacted in Thailand. This act was the first national legislation in Southeast Asia to offer legal protections against discrimination on gender expression (Mosbergen, Reference Mosbergen2015). Therefore, the conventional assumption of the relationship between sources of support at work for LGB working in Western or developed countries might not apply to the context in Thailand.
In seeking to address this shortcoming and promote equality and truly supportive environments for LGB workforce in Thailand, we extend previous research by proposing a framework describing the influences of three types of support at work (supervisor, organizational, and coworker) on LGB employees’ job and life satisfaction. In addition, we provide empirical evidence and insight into how organizational support and job satisfaction become more significant on the development of life satisfaction for LGB workforce in the distinctive social context of a developing country like Thailand, as compared with Western or developed countries. Online surveys were administered to 144 Thai employees who identified themselves as LGB and were working in any organizations located in Thailand. The findings offer insights for human resource managers to gain a better understanding of the supportive work environment required for LGB employees. In addition, the empirical results provide guidance for organizations that aim to have ‘good practice’ in equality toward sexual minority groups in Thailand.
LITERATURE REVIEW
Disclosure and concealment
Disclosure or the phrase ‘come out’ is used to refer to the process by which one accepts one’s own sexuality, gender identity, or status as an intersexed person, and then shares it with others. Disclosure is not an easy process and requires a tremendous amount of courage and support from family, friends, and colleagues. According to Ragins (Reference Ragins2004), disclosure is an important part of developing self identity, allowing LGBs to develop their own authentic and stable sense of self, and to positively cultivate the identity of sexual minority (Wells & Kline, Reference Wells and Kline1987), and decrease the negative psychological effects of identity management (Cain, Reference Cain1991).
The feeling of being able to disclose sexual orientation at work can lead to better work performance. It is critical that LGBs feel safe about not hiding their sexual orientation in their work environment. This can help LGBs to be the most productive and involved within the organization (Van Den Bergh, Reference Van Den Bergh1999). In addition, being ‘out’ at work can combat isolation, bolster efficacy, engender social support for occupational goals, and allow the individual to contribute more fully to the organization’s success (Wells & Button, Reference Wells and Button2004). Day and Schoenrade’s (Reference Day and Schoenrade1997) study mentioned that being ‘out’ at work correlates with improved job satisfaction and work–life balance. LGB individuals often decide to what degree they disclose details of their sexual orientation and to whom they disclose this information (Legate, Ryan, & Weinstein, Reference Legate, Ryan and Weinstein2012). Specifically, LGB individuals are likely to reveal their sexual orientation if they perceive the support of their colleagues (Ragins & Cornwell, Reference Ragins and Cornwell2001).
On the other hand, concealment of sexual orientation is a strategy that LGBs often use to avoid stigmatization or negative reactions. Yet, concealment may relate to higher stress over time (Miller & Major, Reference Miller and Major2000). Those who choose not to disclose their sexual orientation may be more likely to be dissatisfied in the workplace (Griffith & Hebl, Reference Griffith and Hebl2002), have more conflicts at work, experience fewer job promotions, and possess negative job attitudes (Ragins, Singh, & Cornwell, Reference Ragins, Singh and Cornwell2007). LGB employees might face obstacles befriending their colleagues who are also LGBs especially if they are reluctant to disclose their sexuality in the workplace (Woods & Lucas, Reference Woods and Lucas1993; Burnett, Reference Burnett2010). Critcher and Ferguson’s (Reference Critcher and Ferguson2011) study finds that LGBs who are asked to conceal their sexual orientation perform significantly worse on cognitive and physical tasks than LGB employees who are not asked to conceal their sexual orientation.
Supportive factors of LGB employees’ Job and life satisfaction
Although sexual orientation is perceived to be outside the domain of work, it is actually a part of people’s working lives and cannot be separated (Gedro, Reference Gedro2009). Since people nowadays spend longer hours at work than at home, having a supportive work environment plays an even greater role in the lives of employees. This is especially true for LGB’s working in an environment where they do not feel discriminated against because of their sexual orientation. According to Badgett et al. (Reference Badgett, Durso, Kastanis and Mallory2013), most LGB research reveals a positive relationship between disclosure and improved health and well-being of LGB employees. Yet, a large number of studies in this area raise concerns about the significance of context specific effects given that most studies are derived from the LGB workforce in Western communities or developed countries (e.g., Trau & Härtel, Reference Trau and Härtel2007; Huffman, Watrous-Rodriguez, & King, Reference Huffman, Watrous-Rodriguez and King2008). Therefore, examining the effects of various types of supportive environments on LGB’s job and life satisfaction in the distinct social context of Thailand is urgently needed.
Previous research studies have tested the conceptualization of support for LGB at work by combining several types of support together (e.g., Griffith & Hebl, Reference Griffith and Hebl2002; King, Reilly, & Hebl, Reference King, Reilly and Hebl2008). Research from Trau and Härtel (Reference Trau and Härtel2007) indicates that the more support and greater diversity of sexual orientations in an organization, the more disclosure and positive work–life responses among LGB employees. In addition, Huffman, Watrous-Rodriguez, and King (Reference Huffman, Watrous-Rodriguez and King2008) found that each type of support – supervisor, organizational, and coworker – plays a different role in supporting LGB workforces at work and that all types of support are positively correlated with the rate of identity disclosing, life satisfaction, and work satisfaction.
Supervisor support
Support from supervisor is perhaps the most important source of support at work (Moyle, Reference Moyle1998). Employees see their supervisors as the representative of the organization who holds power over them, and developing and keeping a good relationship with their supervisors can ensure their longevity in the organization. Supervisor support can have effects on emotional distress (Peeters & le Blanc, Reference Peeters and le Blanc2001), job satisfaction (Lobban, Husted, & Farewell, Reference Lobban, Husted and Farewell1998), and other tangible benefits, for example, bonus and salary increase. Goldberg and Smith (Reference Goldberg and Smith2013) reported that supervisor support is indirectly related to employees’ anxious and depressive symptoms. Similarly, Waldo’s (Reference Waldo1999) study confirms that having an unsupportive supervisor may contribute to stress and negative mood – a concern of discrimination and job loss – for sexual minorities.
Organizational support
For LGB employees, organizations could provide both instrumental and psychosocial forms of support. In order to succeed in adopting diversity of workforces, organizations must recognize the importance of their LGB employees’ perception toward organizational support and how the organizations value their LGB employees’ contributions at work (Kossek & Zonia, Reference Kossek and Zonia1993; Thomas & Ely, Reference Thomas and Ely1996; Williams & O’Reilly, Reference Williams and O’Reilly1998; Mor Barak & Levin, Reference Mor Barak and Levin2002). Examples of organizational support include nondiscrimination policies, diversity training, LGB union, diversity department, and support for LGB activities. Many US and UK firms are advanced in adopting these kinds of support. They offer LGB non-discriminatory policies not because of legal requirements, but because of the organizations’ commitment to fairness and their intention to recruit and retain qualified employees regardless of sexual orientation (Gurchiek, Reference Gurchiek2005). Some employers foster an LGB-friendly workplace culture by providing diversity training or by modeling an inclusive culture among their managers and executives (Riley, Reference Riley2008). The impact of working in a positive or non-discriminatory environment is that employees are happier at work, which enhances job satisfaction and improves employees’ productivity and effectiveness (Colgan, Creegan, McKearney, & Wright, Reference Colgan, Creegan, McKearney and Wright2007).
Coworker support
Supportive coworkers are fellow employees who respect their LGB colleagues, are sensitive to their LGB colleagues’ concerns, and are aware of their own status in regards to sexual orientation and gender identity (Riley, Reference Riley2008). It is normal for coworkers to have some degree of knowledge about their colleagues’ personal lives, and this knowledge can be a critical element in establishing the trust upon which networking and mentoring relationships are built (Kronenberger, Reference Kronenberger1991; Neely Martinez, Reference Neely Martinez1993). Coworker support helps create a safe climate for disclosed LGBs. Supportive coworkers will sometimes also act as advocates for their LGB coworkers and support them when they are faced with discriminatory situations (Melton & Cunningham, Reference Melton and Cunningham2014). This helps to enhance job satisfaction and work commitment (Ragins, Singh, & Cornwell, Reference Ragins, Singh and Cornwell2007).
Three types of support for life satisfaction and the mediating role of job satisfaction for LGB employees in Thailand
Previous research defines job satisfaction as an affective response to one’s job, while life satisfaction is a general affective evaluation of one’s life (Huffman, Watrous-Rodriguez, & King, Reference Huffman, Watrous-Rodriguez and King2008). Since most people spend at least 8 hr a day, 5 days a week at work, the work environment is important in influencing job and life satisfaction. Elsesser and Peplau (Reference Elsesser and Peplau2006) and Hutlin and Szulkin (Reference Hutlin and Szulkin2003) agree that employees who are excluded from friendship networks at work can struggle to advance their careers because access to important organization resources and sources of influence are limited. Similar to everyone else, LGBs benefit from being a part of the environment in which they live; and being able to share their lives with and receive support from family, friends, and acquaintances (American Psychological Association, 2008). At work, social support can lessen negative organizational outcomes, for example absenteeism (Godin & Kittel, Reference Godin and Kittel2004) and employee turnover rates (Acker, Reference Acker2004), thereby increasing positive organizational outcomes, for example, job performance (Shanock & Eisenberger, Reference Shanock and Eisenberger2006), organization commitment (Redman & Snape, Reference Redman and Snape2006), and job satisfaction (Acker, Reference Acker2004; Redman & Snape, Reference Redman and Snape2006). Trau and Hartel (2004) explored gay men’s career development issues and found that support from colleagues and family was pivotal in comfortably disclosing gay sexual identity at work. Consistent with Trau and Hartel (2004), Rumens (Reference Rumens2010) confirms that workplace support leads to constructing and sustaining managerial career identities.
Past research studies find that several types of work-related support influence LGB’s job and life satisfaction. However, the literature is largely derived from the LGB workforce in Western communities. To our knowledge, empirical evidence of Thai LGB workforce is very limited. Since the social security support in Thailand is much weaker than in Western countries, once a Thai employee loses his or her job, he or she would not receive guaranteed support from the social security system and might not survive financially. Most LGB workforces in Thailand are not well protected by the nondiscrimination legislation that requires equal employment rights for lesbians and gay men which have been widely adopted by employers in Western countries (Graham, Reference Graham1986; Susser, Reference Susser1986; Seal, Reference Seal1991). Within this distinctive social context of Thailand, we examine the importance of three types of support at work – supervisor, organizational, and coworker – on Thai LGB’s job and life satisfaction, while highlighting the important role of job satisfaction on life satisfaction for LGB employees in Thailand. We propose the following hypotheses (see also Figure 1):
Hypothesis 1: A positive relationship exists between supervisor support and LGB employees’ job satisfaction.
Hypothesis 2: A positive relationship exists between organizational support and LGB employees’ job satisfaction.
Hypothesis 3: A positive relationship exists between coworker support and LGB employees’ job satisfaction.
Hypothesis 4: A positive relationship exists between job satisfaction and life satisfaction for LGB employees.
Hypothesis 5: LGB employees’ job satisfaction is a mediator variable between supervisor, organizational, or coworker support and LGB employees’ life satisfaction.
The conceptual model of types of support at work and their effects on Thai lesbian, gay, and bisexual’s (LGB) job and life satisfaction
Methodology
Sample and data collection
The data used for developing the model and testing our hypotheses were collected by an online survey containing questions adapted from other surveys which we describe below. Although the respondents in this study are LGB individuals who have already disclosed themselves, it is still sensitive to conduct a research study in this area using paper-based questionnaires which could create an uncomfortable climate. In order to offer the respondents the most comfortable environment and obtain a good representation of a full range of diversity in the LGB population, the online survey methodology was more appropriate (Trau & Härtel, Reference Trau and Härtel2007). In this study, the online survey was promoted via Facebook, LINE, and other online communities of LGB in Thailand. The questionnaire was posted in the Facebook in a group of activity news for LGBT’s right support in Thailand and Thai Sexual Diversity Network, in which several people shared the questionnaire in their Facebook personal page to help distribute the questionnaire. In addition to Facebook, the questionnaire was also distributed to Line groups in which all members are gays. The respondents could access the online survey link from their comfort zone and answer the questions without identifying themselves in public. A total of 161 questionnaires were obtained through the online survey within 3 days after distribution. Due to a large amount of missing data on key variables, 17 responses were excluded, yielding a final sample of 144 questionnaires (n=144).
To ensure the appropriateness and quality of the respondents, we screened the potential respondents with the following criteria:
(a) Thai employees who identified themselves as LGB and were working in any organizations located in Thailand.
(b) The origin of the organizations was not limited. The respondents might work in either local or multinational organizations.
(c) The respondents must have high school diplomas or higher degrees.
Respondents who fit all the screening criteria were invited to proceed with the survey. All respondents were informed about the confidentiality of their responses. For robustness, the sampling frame was obtained from various professions, age groups, and educational backgrounds. Respondents who agreed to answer the questionnaire were asked to answer seven sets of questions, 36 items in total, including demographic information, level of disclosure, supervisor support, organizational support, coworker support, job satisfaction, and life satisfaction. As this study aimed to investigate the factors that affect job and life satisfaction of LGB employees in Thai context, the survey questions were translated into Thai language for the respondents’ convenience. The procedures recommended by Brislin (Reference Brislin1990) for survey translations were applied.
Respondent profiles
Details of demographic characteristics of LGB participants are presented in Table 1.
Demographic distribution of disclosed lesbian, gay, and bisexual participants
Among the participants (n=144), 74.31% identified their physical gender as male and 25.69% as female, while revealing their sexual orientation as 68.06% gay, 17.36% lesbian, and 14.58% bisexual. For education levels, 59.3% had earned a diploma or bachelor’s degree or equivalent, and 40.97% had a higher education. The percentage of participants working in private sector versus public sector/voluntary sector was 77.78 and 22.22%, respectively. The majority of participants reported that they work in operational levels (63.89%), while the rest work in either middle or senior management levels (36.11%).
In the second part of the questionnaire, participants were asked about their level of disclosure in the organizations. Of all 144 responses, 49.31% reported to have disclosed their sexual orientation or ‘come out’ to everyone at work, 27.08% disclosed their sexual orientation to some people at work, while 23.61% were out to only a few people in their organizations.
Measures
In general, the key variables in this study are operationalized using existing well-validated scales or measures slightly adapted from existing scales reported in previous studies (i.e., Liddle, Luzzo, Hauenstein, & Schuck, Reference Liddle, Luzzo, Hauenstein and Schuck2004; Huffman, Watrous-Rodriguez, & King, Reference Huffman, Watrous-Rodriguez and King2008). Full citation of each measurement item is provided in Appendix 1.
Independent variables
We measured supervisor support using an 8-item scale adapted from Eisenberger, Huntington, Hutchison, and Sowa (Reference Eisenberger, Huntington, Hutchison and Sowa1986) to examine LGB employees’ perception toward their supervisors’ values of their contribution and care of their well-being. After measurement purification based on an exploratory factor analysis, supervisor support scale retained seven items. An example of supervisor support items was ‘My supervisor really cares about my well-being.’ Organizational support was assessed by a revisited version of Lesbian, Gay, Bisexual, and Transgendered Climate Inventory (LGBTCI) (Liddle et al., Reference Liddle, Luzzo, Hauenstein and Schuck2004). The selected seven items capture the perceptions of LGB employees toward the level of support from their organizations. The final cleaned scale includes five items to measure organizational support, for example, ‘At my workplace, LGB employees consider it a comfortable place to work.’ Finally, we adapted Liddle et al.’s (Reference Liddle, Luzzo, Hauenstein and Schuck2004) coworker support measure. This 7-item scale taps the perception of LGB employees toward the level of support from their coworkers. After measurement purification using exploratory factor analysis, we retained five items referring to coworker support. An example of coworker support items was ‘My coworkers are as likely to ask nice, interested questions about a same-sex relationship as they are about a heterosexual relationship.’ The anchor points for item rating in all questions were adjusted from a 4-point to 5-point scale anchored by 1=‘strongly disagree,’ and 5=‘strongly agree’ for consistency. The adjusted scale helped reduce respondents’ confusion when answering the questions.
Dependent variables
The questionnaires were developed based upon earlier studies (Cammann, Fichman, Jenkins, & Klesh, Reference Cammann, Fichman, Jenkins and Klesh1983; Diener, Emmons, Larsen, & Griffin, Reference Diener, Emmons, Larsen and Griffin1985; Huffman, Watrous-Rodriguez, & King, Reference Huffman, Watrous-Rodriguez and King2008), two dependent variables – job satisfaction and life satisfaction – were measured with 3-item and 5-item scales, respectively. An example of job satisfaction items was ‘In general, I like working here,’ while that of life satisfaction items was ‘The conditions of my life are excellent.’ All items used a 5-point scale anchored by 1=‘strongly disagree,’ and 5=‘strongly agree.’ The mean of the scores provided the composite score for each variable.
For the missing data, we use mean imputation technique due to the low level of missing data in this sample (<5%), as suggested by Hair, Black, Babin, Anderson, and Tatham (Reference Hair, Black, Babin, Anderson and Tatham2006) and Rubin, Witkiewitz, St. Andre, and Reilly (Reference Rubin, Witkiewitz, St. Andre and Reilly2007).
Data analysis and results
The hypotheses were tested simultaneously using a partial least squares, structural equation modeling (PLS-SEM) approach which was a variance-based structural equation modeling technique commonly used in management research (Bontis & Serenko, Reference Bontis and Serenko2007; Bontis, Booker, & Serenko, Reference Bontis, Booker and Serenko2007). Recently, PLS-SEM application has been increasingly used in organizational and marketing studies with the recognition that PLS-SEM’s distinctive methodological features make it a possible alternative to the more popular covariance-based structural equation modeling approaches (Sosik, Kahai, & Piovoso, Reference Sosik, Kahai and Piovoso2009; Hair, Sarstedt, Ringle, & Mena, Reference Hair, Sarstedt, Ringle and Mena2012). In particular, a number of recent reviews of PLS-SEM have generated evidence supporting its advantages over covariance-based structural equation modeling (Hair, Hult, Ringle, & Sarstedt, Reference Hair, Hult, Ringle and Sarstedt2013; Mitchell, Obeidat, & Bray, Reference Mitchell, Obeidat and Bray2013). PLS-SEM is appropriate for this data analysis, as it has some significant advantages over other modeling techniques: (1) PLS-SEM has no identification issues regarding small sample sizes and allows us to achieve high levels of statistical power (Reinartz, Haenlein, & Henseler, Reference Reinartz, Haenlein and Henseler2009), (2) it estimates more accurate path relationships when there is correlation between independent variables since PLS-SEM attempts to minimize the error terms and maximize R 2 values of the target endogenous constructs (Sosik, Kahai, & Piovoso, Reference Sosik, Kahai and Piovoso2009; Hair et al., Reference Hair, Hult, Ringle and Sarstedt2013), and (3) it is a non-parametric method that relaxes the assumption of multivariate normality needed for maximum likelihood-based structural equation modeling estimations, and is suitable for applications where strong assumptions cannot be fully met (Hair et al., Reference Hair, Sarstedt, Ringle and Mena2012; Camp, Oltra, Aldás-Manzano, Buenaventura-Vera, & Torres-Carballo, 2016).
The data were analyzed in two stages employing the SmartPLS 2.0 M3 release (Ringle, Wende, & Will, Reference Ringle, Wende and Will2005). First, the psychometric properties of the measurement instrument including individual item reliability/content validity, internal consistency reliability, convergent validity, and discriminant validity were assessed using approaches that Fornell and Larcker (Reference Frone1981) developed for a PLS context. Second, the hypothesized causal relationships were tested using the structural model. The objective of this stage is to confirm the extent to which the causal relationships proposed in this study are consistent with the collected data. The sample size obtained in this study is large enough to carry out a statistical study based on a PLS-SEM approach (Hair et al., Reference Hair, Hult, Ringle and Sarstedt2013). According to the complexity level of the model to be tested, the minimum sample size required was calculated using a 10 times rule as suggested by Barclay, Higgins, and Thompson (Reference Barclay, Higgins and Thompson1995) that the sample size should be 10 times the maximum number of structural paths directed at a particular construct in the inner path model, and this was made up of a minimum required sample size of 40. This figure is within the bounds of the sample size obtained in this study (n=144), indicating that the sample size of 144 is adequate for a robust path modeling.
Analysis of the measurement model
Following the approach that Fornell and Larcker (Reference Frone1981) developed for a PLS context as previously described, we assessed individual item reliability, construct reliability, convergent validity, and discriminant validity. Similar to other structural equation modeling techniques, PLS-SEM generates data that enable the assessment of the measurement and structural components of research models (Mitchell, Boyle, Parker, Giles, Chiang, & Joyce, Reference Mitchell, Boyle, Parker, Giles, Chiang and Joyce2015). For the measurement model, PLS-SEM analysis provides factor loadings (individual item reliability) that can be interpreted similarly to principal components factor analysis (Sosik, Kahai, & Piovoso, Reference Sosik, Kahai and Piovoso2009). Details of factor loading are shown in Table 2. All the tests have shown satisfactory results. Factor loadings were all above the 0.707 guideline (Barclay, Higgins, & Thompson, Reference Barclay, Higgins and Thompson1995), indicating that all constructs possess adequate individual item reliability. Also, the values for both the Cronbach’s α coefficient and composite reliability were all above 0.8, which meet Nunnally and Bernstein’s (Reference Nunnally and Bernstein1994) stricter guideline for testing construct reliability in basic research. Average variance extracted (AVE) in constructs were greater than 0.5 minimum threshold value suggested by Bagozzi and Yi (Reference Bagozzi and Yi1988), indicating convergent validity. To assess discriminant validity, we compared the square root of AVE or diagonal values in Table 3 with the correlations among constructs or off-diagonal values in the same table. As shown in Table 3, the square root of AVE is higher than the construct correlation, indicating that each construct relates more strongly to its own measures than to the others and thus confirming discriminant validity. Table 3 provides descriptive statistics of each construct, Cronbach’s α, composite reliability, AVE, and a correlation matrix with the square root of AVE on the diagonal. Based on this assessment, it can be concluded that all the tests carried out have shown satisfactory results and that these psychometric properties are sufficiently strong to enable an interpretation of structural model parameters.
Factor loadings for the measurement model
Notes. The following items – CoSup4, OrgSup6, OrgSup7, and SupSup3 – were dropped from the scale after measurement purification.
The bold numbers indicate loadings within factors. While the unbold numbers indicate the cross-loadings between factors.
CoSup=coworker support; JobSat=job satisfaction; LifeSat=life satisfaction; OrgSup=organizational support; SupSup=supervisor support.
Descriptive Statistics and correlation matrix
Notes. The diagonal elements in bold numbers are square root of the average variance extracted (AVE). Off-diagonal elements are correlations among constructs.
a Mean: the average score for all of the items in this measure.
CA=Cronbach’s α; CR=composite reliability; CoSup=coworker support; JobSat=job satisfaction; LifeSat=life satisfaction; OrgSup=organizational support; SupSup=supervisor support.
Finally, since the data regarding all dependent and independent variables derive from the same self-reported survey, this could lead to what is known as ‘common method bias’ (Podsakoff, MacKenzie, Lee, & Podsakoff, Reference Podsakoff, MacKenzie, Lee and Podsakoff2003). Even though common method bias is not considered as a threat to the validity of the findings in the context of PLS (Chin, Thatcher, & Wright, Reference Chin, Thatcher and Wright2012), we decided to diagnose the extent to which the common method variance effects could be a problem in this study by conducting Lindell and Whitney’s (Reference Lindell and Whitney2001) marker variable assessment test. Results from this test suggest that for all significant effects of the antecedents and their consequences on the dependent variables, the corresponding bivariate correlations coefficients remain statistically significant at p<.05 when partialling out an unrelated ‘marker variable’ (Lindell & Brandt, Reference Lindell and Brandt2000; Lindell & Whitney, Reference Lindell and Whitney2001). Therefore, it can be concluded that common method effects are not likely to contaminate the results observed in this study.
Analysis of the structural model
To assess the research hypotheses, test the mediating relationships, and determine the degree of statistical significance of the coefficients of both the measurement and the structural models, we used a bootstrap re-sampling procedure with 500 randomly generated sub-samples. Structural evaluation was conducted to assess nomological validity by examining the size and significance of path coefficients and R 2 values of the dependent variables (Smith & Barclay, Reference Smith and Barclay1997; Hair et al., Reference Hair, Hult, Ringle and Sarstedt2013). Results obtained from structural model analysis are shown in Table 4. Overall, the results show that the predictors offer good explanation for focal constructs: R 2 for job satisfaction=0.49; and R 2 for life satisfaction=0.47.
Structural results of types of support at work and their effects on lesbian, gay, and bisexual’s Job and life satisfaction model
Notes. * p<.05; *** p<.001 (one-tailed test).
CoSup=coworker support; JobSat=job satisfaction; LifeSat=life satisfaction; OrgSup=organizational support; SupSup=supervisor support.
Results
The positive and significant effects of supervisor support (β=0.59, p<.001) and coworker support (β=0.13, p<.05) on job satisfaction support Hypothesis 1 and Hypothesis 3, respectively. However, the effect of organizational support on job satisfaction is not significant, thus Hypothesis 2 is not supported. In addition, we found a positive and significant effect of job satisfaction on life satisfaction (β=0.60, p<.001), which support Hypothesis 4. The details are shown in Table 4, Model 1.
Following Baron and Kenny’s (Reference Baron and Kenny1986) and Tippins and Sohi’s (Reference Tippins and Sohi2003) approach, we tested for the presence of a mediating effect of LGB employees’ job satisfaction by comparing the direct effect between variables and the competing links where the mediated variable occurs. As shown in Table 4, Model 1 and Model 2, the inclusion of job satisfaction leads to a significant decrease in the effect size of supervisor support (from 0.39 to 0.04), which is not significant, suggesting a full mediation (total effect β=0.39; direct effect β=0.04, n.s.; indirect effect β=0.35, p<.001). Also, the mediation model explains more variance in life satisfaction than the direct effect model (0.47 vs. 0.29). In addition, the results show that the direct effect from coworker support to job satisfaction is not significant, suggesting a full mediation (total effect β=0.04; direct effect β=−0.04, n.s.; indirect effect β=0.08, p<.05). There is a significant direct relationship between organizational support and life satisfaction (β=0.19, p<.05); nevertheless, the effect of organizational support on job satisfaction is not significant as previously mentioned. Therefore, Hypothesis 5 is partially supported.
Although PLS-SEM does not generate indicators of model fit, the model R 2 statistic indicates the extent to which hypothesized pathways combine to predict the dependent variable. In this study, the R 2 were 0.49 for job satisfaction and 0.47 for life satisfaction which can be interpreted as indicating good model fit (Chin, Reference Chin1998). As recommended by several organizational studies using PLS-SEM analysis (e.g., Mitchell et al., Reference Mitchell, Boyle, Parker, Giles, Chiang and Joyce2015; Camps, Oltra, Aldás-Manzano, Buenaventura-Vera, & Torres-Carballo, Reference Camps, Oltra, Aldás-Manzano, Buenaventura-Vera and Torres-Carballo2016), we assessed the model’s predictive performance by using PLS-SEM to generate the Stone–Geisser criterion (Q 2) with an omission distance of 7 (Geisser, Reference Geisser1974; Stone, Reference Stone1974). Analysis resulted in a Stone–Geisser criterion Q 2 value of 0.39 for job satisfaction and 0.28 for life satisfaction, which is substantially above the threshold value of 0, and which indicates the model’s predictive relevance since values of 0.02, 0.15, and 0.35 indicate that an exogenous construct has a small, medium, or large predictive relevance for a selected endogenous construct (Hair et al., Reference Hair, Hult, Ringle and Sarstedt2013). The results suggest that all three types of support have a significant impact on job and life satisfaction, and also support the utility of the pathways that we have investigated.
In sum, the results suggest the following conclusions. First, supervisor support and coworker support are positively related to LGB employees’ job satisfaction. Second, supervisor support and coworker support have indirect effects on life satisfaction (through LGB employees’ job satisfaction). Third, organizational support only has a direct effect on life satisfaction. Together these results provide evidence that there is a significant mediating effect of job satisfaction on the relationships between supervisor/coworker support and LGB employees’ life satisfaction, and that the mediation model represents a significant improvement over the direct effect model since the mediation model explains a good amount of the variance of the LGB employees’ life satisfaction. The finding highlights the direct and significant effect of organizational support on LGB employees’ life satisfaction in the distinctive context of Thailand.
DISCUSSION AND CONCLUSION
Our goal in this study is to extend the LGB literature by untangling the effect of supportive working environments toward LGB employees, especially those in Thailand. The empirical results shed light on the importance of supervisor support and coworker support on job satisfaction. Huffman, Watrous-Rodriguez, and King (Reference Huffman, Watrous-Rodriguez and King2008) stated that supervisor support had a stronger relationship toward job satisfaction than coworker and organizational support and that only supervisor support was significantly related to job satisfaction. However, our findings reveal that both supervisor and coworker support are significantly related to job satisfaction (β=0.59, p<.001, and β=0.13, p<.05, respectively).
The empirical findings of this study provide interesting aspects of LGB employees’ perception toward their workplace environment in a different geographical and cultural context. In Thailand, which has a predominantly collectivisticFootnote 1 culture, that is, focusing on group orientation, teamwork, and mutual dependence, LGB employees’ job satisfaction is influenced by both supervisors’ and coworkers’ attitudes toward their sexual identity. We believe that strong support from supervisors and coworkers in the workplace creates the climate of a safe-place for LGB workers as their identities are affirmed and accepted within their groups or organizations. In such climates, their supervisors and coworkers often act as advocates and protect LGB from discrimination (Melton & Cunningham, Reference Melton and Cunningham2014). On the contrary, past research conducted with LGB employees in Western communities posited that supervisors were seen as directly tied to the job and that their supervisor-subordinate relationship was recognized as an essential part of their successful task outcomes; therefore, supervisor support would affect job attitude to a greater extent than other types of support (Frone, Reference Fornell and Larcker2000; Huffman, Watrous-Rodriguez, & King, Reference Huffman, Watrous-Rodriguez and King2008). This is consistent with characteristics of members of countries which have predominantly individualisticFootnote 2 cultures, that is, more independent of the groups or organizations that they are part of, focus on personal goals, act according to attitudes rather than social norms, and emphasize task outcomes (Hofstede, Reference Hofstede1991; Markus & Kitayama, Reference Markus and Kitayama1991; Triandis, Reference Triandis1995).
Huffman, Watrous-Rodriguez, and King (Reference Huffman, Watrous-Rodriguez and King2008) claimed that only coworker support was significantly related to life satisfaction. Interestingly, our results show that organizational support offers a direct and significant relationship with life satisfaction of Thai LGB employees, while both supervisor and coworker support have indirect effects on life satisfaction through job satisfaction. This is an important finding, as it might imply that the potential for any organizations to increase their LGB workforce’s life satisfaction will depend substantially on their ability to increase the employees’ job satisfaction and organizational support. A plausible explanation for this finding could come from differences between Thai and Western LGB employees’ lifestyle outside of work. When considering the factors that might affect quality of life, work is one of the most important concerns because having a job means having income for living. Moreover, it is commonly known that the social security support is stronger in Western countries than in developing countries like Thailand. Therefore, once a person in developing countries loses his or her job, he or she would certainly get support from the effective social security system and can most likely survive financially for a while. In addition, several Western countries have adopted nondiscrimination legislation that requires equal employment rights for lesbian and gay men (Graham, Reference Graham1986; Susser, Reference Susser1986; Seal, Reference Seal1991); thus the LGB workforces in Western countries are well protected both during their work and after they lose their job. However, the situation is quite different in Thailand. If a person loses his or her job, it could affect his or her financial security. Thai laws that prevent or protect against sexual orientation discrimination at work are not as widely enforced as in Western countries. The only regulation that was partly related to the supportive scheme for LGB employees at work is Corporate Social Responsibility TLS8001-2010. This regulation is voluntarily adopted by companies. This means that there is no guaranteed protection at work for LGB employees in Thailand. Further, the compensation that Thai people receive from the social security scheme, if they join the program, is barely enough to live on with the rising cost of living in Thailand. Hence, one way to ensure their security at work is staying in the job as long as possible. This distinctive social context in Thailand highlights a significant direct effect of organizational support on life satisfaction, including the importance of the mediating role of job satisfaction on Thai LGB employees’ life satisfaction.
Managerial implications
There is a lack of empirical research in LGB workforce in Thailand as most literature in this area is largely derived from the LGB workforce in Western countries. This study sheds light on the importance of supportive environments at work for LGB employees’ job and life satisfaction. According to the empirical results, it is confirmed that both supervisor and coworker support are strongly related to LGB employees’ life satisfaction through job satisfaction, while also demonstrating that there is a strong and direct relationship between organizational support and life satisfaction.
From the light of our findings, both supervisor and coworker support to LGB employees can be encouraged by policy-wide approach from top management to embrace diversity on sexual orientation, gender, race, or ethnicity to create inclusive environment. Human resource initiatives can follow by incorporating nondiscrimination and equal opportunity code of conducts for all employees encouraging not to discriminate on the basis of race, religious, gender, age, national origin, disability, marital status, and sexual orientation. This would include equal opportunity for all individuals for employment, benefits, and promotion in their career. Leading companies with high performance results, for example IBM, Google, and Apple, have established and practiced diversity and inclusive policy for many years. Recent study stated that ‘Companies that embrace diversity and inclusion in all aspects of their business statistically outperform their peers’ and showed that diversity in workplace was highly correlated with strong business performance (Bersin, Reference Bersin2015). In addition to implementing a code of conduct in the workplace, human resource initiatives can be extended to a rigorous and systematic evaluation of the 360° process to assess performance and provide anonymous feedback from the people who work around supervisors. This includes feedbacks from the employee’s supervisor, peers, and subordinates. By creating mechanism to welcome all feedback from others and use them part of performance evaluation, supervisors would be more careful and provide more support to their employees.
Furthermore, top management could encourage coworker and supervisor support by increasing level of soft skills. Compassion cultivation education is essential for all employees to have more compassion, empathy, and mindfulness toward others in workplace. Compassion cultivation education would lead supervisors and coworkers to spread compassion in the workplace, feel more connected with people, try to put themselves in others’ shoes, be mindful with action to colleagues, and ultimately provide an overall sense of human well-being. With mindfulness to all actions taken to colleagues, diverse and inclusive workplace can be created and supported. However, it takes a lot more than supervisor and coworker support to make a workplace LGB friendly. Even with LGB-supportive activities and environments, many organizations still face challenges in implementing these activities and creating these environments effectively (Young, Reference Young1992; Creegan, Colgan, Charlesworth, & Robinson, Reference Creegan, Colgan, Charlesworth and Robinson2003). Some employees often misinterpret LGB-supportive projects at work as providing higher privilege in the organization to LGB employees when these projects are meant to create the equality and fair treatment. Therefore, managers who decide to adopt LGB support activities at work should carefully consider how to effectively implement these activities within their organizations. The support activities should aim to (1) make LGB employees more visible, (2) drive diversity and equality efforts, (3) prevent the discrimination, and (4) support LGB employees who suffer from sexual orientation discrimination. Another study in diversity also supports that, by shifting individuals’ focus away from judging people to recognizing the individual characteristics which are essential to the target job, openness to non-job-related individual differences as well as an objective and open-minded view of job requirements can be encouraged (Hartel, Hartel, & Trumble, Reference Hartel, Hartel and Trumble2013). To change the culture and climate within an organization to create an LGB-friendly workplace, employers requires a set of activities such as diversity training, LGB-supportive activities, LGB role model, or LGB unions that go beyond what is required by laws and regulations. These activities have been widely adopted and also increasingly welcomed in several developed countries as managers increasingly recognize that workforce diversity can be a source of competitive advantage. Therefore, in order to stay in a competitive business environment, companies in Thailand should consider adopting these policies and activities to attract and maintain LGB talents. For example, to change the state of ‘tolerant but unaccepted’ in Thailand, programs for managing diversity and/or diversity training on LGB should be implemented.
Limitations and future research directions
This study has some limitations which highlight several avenues for future research. First, the data were based on only ‘disclosed’ LGB employees in Thailand. These individuals may differ from ‘closeted’ LGB employees (those who are not ready to disclose their sexual orientation at work). Although we cannot generalize our findings to ‘closeted’ LGB employees, future research might be able to find ways to include them as representative samples of LGB employees.
Second, the study focused only on LGB employees in Thailand because of limitation on data collection. Before data collection, expert interviews suggested that transgendered (T) or queer (Q) were mostly found in fashion, media, and entertainment businesses rather than corporate offices. The targeted workplaces in this study are mostly regular offices which T or Q cannot obviously be seen, thus it was very difficult to get the targeted T or Q respondents. In Thailand workplace, LGB may physically be seen as their true gender. Future research might include T or Q into sample of research.
Third, the model does not purport to represent all possible antecedents and consequences of a supportive environment toward LGB employees’ job and life satisfaction. As the objective of implementing all LGB-supportive activities and creating supportive environments are from the organizations’ attempts to prevent, reduce, and correct issues that are related to sexual orientation discrimination, it is recommended to further study whether discrimination can be prevented or reduced by engaging supervisors or coworkers in particular processes or activities.
Finally, this study does not collect information regarding organizations of our samples. For example, LGB employees in different types of organizations (e.g., different size, having or not having diversity programs, etc.) might have different perceptions toward their working environment, and different types of organizations might also approach their LGB employees differently. Further study may offer insights into how different types of organizations affect LGB employees’ job and life satisfaction.
Acknowledgements
The authors gratefully acknowledge the contributions and valuable comments of anonymous reviewers and our friends to this article. This manuscript is an original work that has not been submitted to nor published anywhere else. All authors have read and approved the paper.
Supplementary material
To view supplementary material for this article, please visit https://doi.org/10.1017/jmo.2017.32
