In popular and much debated works, Richard Florida (2002, 2005) has hypothesized that for cities and states to prosper, they must have what he has called a creative class. He has argued that in order to attract and support that creative class, locales must be tolerant, which includes welcoming a vibrant lesbian, gay, bisexual, and transgender (LGBT) community. In Florida’s argument, a strong LGBT community signals to leaders and innovators that diversity is welcomed and that an open-minded and progressive way of life is supported and promoted.

One signal to lesbian, gay, and bisexual (LGB)Footnote 1 individuals that a community is welcoming is the enactment of a nondiscrimination policy that includes sexual orientation. A number of arguments have been made in support of nondiscrimination policies that include sexual orientation (and gender expression or identity; Rubenstein 2001). Among these arguments are basic concerns about ensuring fair and equal treatment of all citizens. Proponents also claim that such policies send an important message to all residents that discrimination on the basis of sexual orientation is considered unacceptable in their jurisdiction.

LGB individuals are members of a stigmatized minority group (based on the nonnormative performance of gender roles through the identification of their same-sex attractions (Herek 1999, 2000). As a result, they are at risk to experience minority stress. Minority stress for LGB individuals includes the stress associated with experiences of discrimination, anticipated rejection or prejudice, decisions about disclosure, internalized homophobia, and the coping strategies that they employ to deal with these stresses (Meyer 2003; Rostosky et al. 2007). As members of a stigmatized group exposed to minority stress and its negative effects (Meyer), feeling welcomed and protected could have positive benefits for LGB individuals’ well-being (c.f. Beals and Peplau 2005).

Although research has been conducted on workplace nondiscrimination policies and minority stress factors, studies of the presence of a state nondiscrimination policy including sexual orientation and the perceptions of LGB residents of a jurisdiction are scarce. Using data from a large national convenience sample of LGB individuals, we tested hypotheses that residence in states with inclusive policies would be associated with perceptions of a more positive and less negative environment, experiences of more social support, and lower levels of minority stress in LGB participants.

Nondiscrimination Policies

The debate about the degree to which civil rights laws reflect or reform environments in the locales of their jurisdiction has a long history among scholars, politicians, and citizens. For example, Allport (1954), in The Nature of Prejudice, argued that laws were a positive factor in persuading the public to be less prejudiced. State and local nondiscrimination policies have their origins in a long history of US civil rights laws, including, for example, the Civil Rights Act of 1866, the 14th Amendment to the US Constitution (1868), the Civil Rights Act of 1964, and the Fair Housing Amendments Act (1988). Originally intended to protect the rights of African Americans, these laws now commonly include national or ethnic origin, sex, religion, age, marital status, disabilities, and veteran status. Nondiscrimination policies affect residents within jurisdictions in a circumscribed set of circumstances (e.g., protecting them from discrimination in employment or housing).

The inclusion of sexual orientation in the nondiscrimination policies of states and cities is relatively new in the US political landscape. In 1972, East Lansing, Michigan was the first city to include sexual orientation in their local nondiscrimination policy. In 1982, Wisconsin became the first state to include sexual orientation in their statewide nondiscrimination policy. As of January, 2009, 20 states and the District of Columbia have policies prohibiting discrimination on the basis of sexual orientation in public or private employment: California*, Colorado*, Connecticut*, Hawaii, Iowa*, Illinois*, Maine*, Maryland, Massachusetts, Minnesota*, Nevada, New Hampshire, New Jersey*, New Mexico*, New York, Oregon*, Rhode Island*, Vermont*, Washington*, and Wisconsin (state names followed by an asterisk also prohibit discrimination on the basis of gender identity). In addition, more than 150 cities prohibit discrimination based on sexual orientation in employment; many of these policies also prohibit discrimination in housing and other public services. Additionally, whereas statewide and local nondiscrimination policies that include sexual orientation are typically passed by legislative bodies (i.e., state legislatures and city councils), limited protections may also be issued by executive order (e.g., orders of governors applied only to the workplace for state employees).

Politicians perceive that nondiscrimination policies that include sexual orientation improve the lives of LGB residents. Button et al. (1995; 1997) conducted a survey of 126 local officials from cities with nondiscrimination ordinances that included sexual orientation. When asked, “What do you think have been the two most important positive effects (if any) of the passage of this ordinance?,” the three most common answers were, “recognition that discrimination based on sexual orientation is legally wrong,” “reduced discrimination based on sexual orientation in public employment and in other covered institutions,” and “gay men and lesbians feel safer and more comfortable in being open.”

The number of complaints filed under state nondiscrimination policies has been estimated at an average of three per 10,000 lesbians and gay men (Badgett 2007; Rubenstein 2001). Many jurisdictions with inclusive nondiscrimination policies have reported no complaints filed and the majority have reported a nominal number of filed complaints (see Button et al. 1995). The lack of longitudinal data makes it impossible to assess whether these low numbers reflect a change in practice after the policy was passed, whether there was always a low incidence rate, or whether there is a continuing reluctance to report incidents.

Despite the lack of filed complaints, self-reports by lesbians and gay men suggest that discrimination does continue. A recent review (Badgett 2007) of samples of lesbians and gay men found that 16–18% reported employment discrimination. Other studies (e.g., Croteau and Lark 1995) have reported that between 25% and 66% of LGB workers experience or anticipate discrimination in the workplace. Past findings have indicated, however, that having inclusive nondiscrimination policies in the workplace promoted disclosure at work, increased job satisfaction, and positive work outcomes (e.g., Badgett 2001; Tejeda 2006; Waldo 1999), including the alleviation of minority stress factors (Rostosky and Riggle 2002). Specifically, inclusive policies or programs in the workplace, such as an inclusive nondiscrimination policy, LGB employees’ association, Pride Week activities, or domestic partner health insurance, facilitated identity disclosure and positive relationships with coworkers (Burgess 1997). Employees covered by an inclusive nondiscrimination policy were more likely to disclose their sexual orientation at work. They also experienced lower stress levels and lower levels of internalized homophobia compared with employees in nonprotected workplaces (Rostosky and Riggle 2002).

In sum, inclusive nondiscrimination policies in the workplace provide a signal that gay and lesbian employees are accepted and that discrimination is not tolerated. The presence of such policies, in turn, has been associated with lower levels of minority stress in the workplace (also see Waldo 1999). It remains an empirical question as to whether inclusive state nondiscrimination policies affect the perceived environment and minority stress of LGB residents.

In this study, we hypothesized that living in a state with an inclusive nondiscrimination policy would be associated with LGB residents’ perceptions of a more positive and less negative environment (measured by perceptions of positive and negative messages), higher levels of social support, and lower reported levels of minority stress factors (i.e., higher levels of disclosure and lower levels of internalized homophobia).

Method

Participant Recruitment

Participants were recruited via e-mail listserv announcements and website postings aimed at the LGB community. The announcements recruited individuals ages 18 and older who identified as gay, lesbian, bisexual, same-gender loving, or same-sex partnered, and who lived in the United States. Participants were directed to a website for more information and to link to the survey. The announcement was originally sent via e-mail to more than 250 electronic mailing lists and websites. These included professional organizations (e.g. business guilds, an LGB engineers group), Pride centers and resource centers, student alliances, advocacy groups, and religious affiliation groups, as well as more general social groups.

A total of 2,511 participants completed the survey. Demographic questions were asked regarding gender identity, sexual identity, racial-ethnic identity, age, education, income, employment status, and relationship status. For the sample, the mean age was 39.1 years (SD 12.7); 58.5% were female and 41.5% male; 88.0% were White or Caucasian, 2.3% African American or Black, 2.7% Hispanic or Latino, and 7.0% biracial or multiracial or other; 27.7% had a high school diploma or some college, 28.1% had a college degree, 28.6% had a master’s degree, and 15.6% had a doctorate or professional degree; the median personal income was $40,000; 26% reported having children; and, 54.2% were in a committed relationship, with an additional 14.4% being in a civil marriage, civil union, or domestic partnership. Compared with the general population, the sample included more women and fewer people of color, was more highly educated and more likely to be in college, and had a higher personal annual income. The sample characteristics were consistent with demographics found in other web-based surveys of LGB samples and web-based survey samples of other populations (see Badgett 2001; Riggle et al. 2005).

Measures

Nondiscrimination Policies

Respondents indicated their current state of residence. Respondents who lived in a state with an inclusive nondiscrimination policy at the time of the survey were coded “1” (“0” otherwise). Participants were also asked, “Does the state you reside in have a statewide nondiscrimination statute protecting gay, lesbian and bisexual citizens?” Response categories included “yes,” “no,” and “don’t know.”

Exposure to Messages

Participants in the study were asked to respond to the following direction:

Please indicate how often during the PAST MONTH you encountered positive or negative messages about gays, lesbians, bisexuals or same-sex couples in the following media. Positive messages include presentations of happy or resilient LGB individuals, same-sex couples and their families, support for LGB rights or marriage rights for same-sex couples. Negative messages include negative stereotypes, derogatory terms, and opposition to the rights of LGB people or same-sex couples, including marriage rights. Note: Some presentations (such as news stories) may include both positive and negative messages. These may be counted in each category.

Participants were asked to report separately how often in the past month they had encountered (a) positive messages and (b) negative messages in television news reports, including messages spoken by people interviewed, reporters for a story, or commentators; newspaper or magazine articles, including messages from individuals quoted in an article, editorials, or the overall article message itself; billboards, yard signs, bumper stickers, or other public advertisements (such as flyers); and conversations that the participant engaged in or overheard. Participants were asked to respond to each message type and source along a six-point scale (range 0–5): not at all, once or twice, about once a week, a couple of times a week, daily or almost daily, or more than once a day. A summary score of positive messages and (separately) negative messages was created by averaging the responses across the four types of messages.

Disclosure

To assess disclosure of sexual orientation, participants responded to the following prompt: “The following questions concern the degree to which you have disclosed your sexual orientation or identity to the following groups of people. Choose the number that best represents your answer.” Response categories ranged from 0% (0) to 100% (10) in 10% increments. Participants were asked about their level of disclosure to parents and siblings; extended family (grandparents, aunts, uncles, cousins, etc.); children (if applicable); straight friends; supervisors, colleagues, coworkers, peers, clients, or customers; and cultural community (e.g., the African American or Latino community, the hearing- or sight-impaired community, a church community).

Social Support and Support for Sexual Identity

Perceptions of general social support and support for sexual identity were measured separately. Perceptions of social support were measured using Sarason et al. (1986) six-item index rating satisfaction with the overall level of emotional support one received. A six-point response scale ranged from very dissatisfied to very satisfied. Scores ranged from 6 to 36 (Cronbach’s alpha reliability was 0.93).

Perceptions of social support for sexual identity were assessed with the following two items: “Think about the people who really accept and support your sexual identity or same-sex orientation. Please indicate how satisfied you are with the OVERALL support you have in this area.” and “Think about the people whom you can really count on to encourage you when you are experiencing difficulty as a lesbian, gay, or same-sex oriented person. Please indicate how satisfied you are with the OVERALL support you have in this area.” A six-point response scale ranged from very dissatisfied to very satisfied. Scores ranged from 2 to 12.

Internalized Homophobia

Internalized homophobia was assessed using The Internalized Homophobia Scale (Wright et al. 1999), a self-report inventory comprising nine items. Participants responded to each item on a Likert-type scale ranging from 1 (strongly agree) to 5 (strongly disagree). Sample items included, “I wish that I wasn’t attracted to the same sex” (reverse scored) and “I feel proud that I am gay/lesbian/bisexual.” Item responses were summed to create a composite measure. Scores ranged from 9 to 45, with lower scores indicating lower levels of internalized homophobia. For this sample, Cronbach’s alpha reliability was 0.81.

Results

Inclusive Policies and Perceptions

Of the 2,511 respondents to the survey, 41% lived in the 17 states with inclusive nondiscrimination policies and 59% lived in states without policy protections. These numbers are based on the state of residence reported by the participant and whether the state had a policy on the date of the survey in 2006 (Colorado, Iowa, and Oregon passed state policies after the survey administration date). We examined whether respondents correctly knew the status of policy protections in their state of residence. Of the respondents living in states with policy protection, 70.4% correctly answered that their state had an inclusive nondiscrimination statute, 9.4% incorrectly answered that their state did not have an inclusive nondiscrimination statute, and 20.2% answered that they did not know. Of the respondents living in states without policy protections, 77.4% answered correctly that their state did not have an inclusive nondiscrimination statute, 4.9% answered incorrectly, and 17.7% answered that they did not know.

An h score was computed based on an arcsine transformation of the percentages comparing the policy knowledge of those who lived in an inclusive policy state with that of those who lived in a state without inclusive policy protections. Significantly more survey respondents who lived in a state without policy protections correctly knew that they were not protected than respondents from inclusive states correctly knowing that they were protected (h = 0.089, p < 0.05). Correspondingly, respondents in states with policy protections were significantly more likely to incorrectly report that they were not protected than respondents in states without policy protections incorrectly reporting that they were protected (h = 0.16, p < 0.01). Additionally, respondents from states with protections were more likely to report that they did not know whether they were protected (“don’t know”) than respondents from states without protections (h = 0.102, p < 0.05).

Messages

As Table 1 shows, LGB residents of states with inclusive nondiscrimination policies reported exposure to significantly more positive messages and fewer negative messages from all measured sources than LGB residents of states without an inclusive nondiscrimination policy.

Table 1 Means and group comparisons of perceptions of exposure to messages, disclosure, social support and internalized homophobia by policy protection status (N = 2,511)
Full size table

Disclosure

LGB residents of states with inclusive nondiscrimination policies were more likely to have disclosed their sexual orientation than LGB residents of states without inclusive policies (see Table 1). This difference extended across all types of disclosure, except disclosure to children (where applicable). The highest overall rates of disclosure were to children, straight friends and parents or siblings. The lowest rates of disclosure were to extended family (who may live in other jurisdictions) and supervisors, peers, coworkers, and clients.

Although most state nondiscrimination policies include workplace protections, only 37% of LGB respondents residing in a state with an inclusive nondiscrimination policy reported being 100% out to supervisors, peers, coworkers, or clients. However, significantly fewer respondents (32%) reported being totally out in states without an inclusive policy (h = 0.105, p < 0.05).

Social Support and Internalized Homophobia

The presence of an inclusive nondiscrimination policy at the state level was associated with significantly higher levels of perceived social support for participants’ sexual orientation, as well as perceived social support in general. Likewise, living in a state with an inclusive nondiscrimination policy was associated with significantly lower levels of internalized homophobia compared with participants who lived in states no inclusive policy (see Table 1).

To control for possible demographic effects on internalized homophobia and social support outcomes, stepwise regression analyses were run with age, gender, and education, entered as demographic controls in step 1, and policy status entered in step 2 on perceived social support (general), perceived social support for sexual identity, and internalized homophobia. Results of the final model of the regression analyses revealed significant associations of age with the dependent variables; older participants reported more social support in general, more social support for their sexual identity, and less internalized homophobia than younger participants. Female participants perceived more social support in general than male participants. Level of education was not significantly related to the dependent variables.

Beyond these effects, policy status was a significant increment in the model (step 2) and was significantly associated with the variance in perceived social support in general, perceived social support for sexual identity, and internalized homophobia (see Tables 2, 3, and 4). That is, living in a state with an inclusive nondiscrimination policy was associated with higher levels of perceived social support in general and for sexual identity, and less internalized homophobia.

Table 2 Final step results of regression of demographic variables and inclusive policies on perceptions of social support
Full size table
Table 3 Final step results of regression of demographics and inclusive policies on perceptions of social support for sexual identity
Full size table
Table 4 Final step results of regression of demographics and inclusive policies on reported level of internalized homophobia
Full size table

Discussion

The results reported support the hypotheses that inclusive state nondiscrimination laws are associated with more positive, supportive environments, and less minority stress for LGB residents. Residents of the states with inclusive policies reported hearing more positive and fewer negative messages, were more likely to disclose their sexual orientation, experienced more social support in general and for their sexual identity, and reported less internalized homophobia.

Interestingly, residents of states with inclusive policy protections were less likely to accurately know about that protection, more likely to wrongly think they were not protected, or less certain about the protections (i.e., indicated they “don’t know”) than those who resided in states that did not have an inclusive policy. Alternatively, residents of states without policy protections were more likely to accurately know that they were not protected. This difference in knowledge could be the result of a general assumption, absent any other conclusive knowledge, that LGB residents are not protected by nondiscrimination policies. Or, residents of states with policies may have known that they were included in some type of policy protection, but may have been unclear about the nature or extent of that policy. In any case, LGB people and the general population may need more information and education about nondiscrimination policies in their jurisdictions. LGB individuals need to know what legal recourse is available to them if they experience discrimination in protected areas.

These data cannot resolve the debate of which comes first—the more positive environment or the law. Most likely, the law and environment are mutually supportive and interact to produce these outcomes. States with inclusive nondiscrimination policies may be the result of an environment that was already tolerant and supportive of an active LGB community. In this case, the policy may further reinforce and extend that tolerant, supportive environment. This perceived support may make residents even more likely to disclose their sexual orientation and may attract new LGB residents. For example, Egan et al. (2008) have found evidence of residential patterns suggesting that LGB individuals may be more likely to live in locales with inclusive policies. Also, higher levels of education and an individualistic state culture have been associated with having inclusive nondiscrimination policies (Dorris 1999; Wald et al. 1996). On the other hand, jurisdictions without inclusive policies may be perceived as condoning prejudicial attitudes and discriminatory actions, making LGB individuals less likely to feel welcomed or supported in these areas.

Likely, a reinforcing loop exists in policy creation, with residents disclosing their sexual orientation and petitioning for protection, followed by policy inclusion, followed by more residents coming out and more individuals moving to that locale so that they can live openly in a protected jurisdiction. Furthermore, living in a supportive environment may reduce internalized homophobia, or individuals with lower levels of internalized homophobia may consciously relocate to more supportive environments. Perhaps living in an environment with more positive and fewer negative messages provides more supportive messages to internalize, which may also enhance LGB individuals’ perceptions of general and specific forms of social support. Modeling these bidirectional relationships awaits future empirical investigation with longitudinal datasets.

The data also suggest that there is a continuing stigmatization of LGB identity. The low rates of disclosure in some environments likely reflect the anticipation of prejudice or discrimination, even in states with inclusive nondiscrimination policies. For example, the lower rates of disclosure in the workplace may reflect a lack of specific nondiscrimination policies in participants’ workplaces, or a fear of loss of clients (by the individuals or their employers) because of stigmatization.

This study does not include a measure of the presence of local inclusive nondiscrimination policies; therefore, participants may have lived in cities with inclusive policies even if their state of residence did not. More than 150 cities, including many cities in states without inclusive policies, have inclusive nondiscrimination policies. These protections may provide more proximal indications of a welcoming and supportive environment. Also, people travel from one jurisdiction to another to work or for services, and thus are exposed to multiple environments of which some, but not all, may be in protected areas. All of these factors may contribute to the relatively low effect sizes of these findings and suggest the need for future research on proximal and distal sources of effects.

This study does not address actual experiences of discrimination. Other findings (e.g., Badgett 2001) have suggested that increased disclosure leads to a greater risk of discrimination. Therefore, for measures of rates of discrimination, the inclusive effects of the policy perhaps would be offset by the risks associated with disclosure. In addition, the types of discrimination covered by state policies might be limited. Thus, LGB residents are not likely to be protected from all forms of discrimination nor from everyday experiences of prejudice. Therefore, even in a state with an inclusive nondiscrimination policy, LGB individuals may not feel safe or welcome in all communities or areas. Future study is needed to determine whether these policies impact rates of discrimination and how those who would discriminate are affected (and possibly deterred) by the presence of the inclusive policy.

The reported results must be interpreted with caution due to the lack of longitudinal data and limitations on the generalizability posed by a volunteer sample of participants (rather than a probability sample). Also, the sample did not include a large enough subsample of African American, Latino/a, or other non-White participants to be able to test for the effects of policy protections on different racial and ethnic groups, who may be subject to multiple forms of discrimination and prejudice.

Discrimination in its many forms takes a toll on minority citizens in the form of minority stress (see Meyer 2003). Minority stress includes perceiving discrimination and prejudice, as well as anticipating incidents of discrimination and prejudice. If inclusive nondiscrimination policies provide recourse for perceived experiences and also lessen the expectations of discrimination and prejudice, then those policies have done double duty in reducing minority stress and creating a supportive environment for LGB residents.