Self-employment among same-sex and opposite-sex couples in Canada.
Waite, Sean ; Denier, Nicole
A BURGEONING LITERATURE CONCERNS the labor market experiences of
nonheterosexuals. Much of this research has focused on quantifying and
exploring sources of wage gaps between sexual minorities and their
heterosexual counterparts (Antecol, Jong, and Steinberger 2008;
Arabsheibani, Marin, and Wadsworth 2004, 2005; Berg and Lien 2002; Black
et al. 2003; Blandford 2003; Carpenter 2008; Denier and Waite 2014;
Drydakis 2012; Waite 2015; Waite and Denier 2015, etc.). Although the
current literature establishes the role of sexual orientation in shaping
labor market outcomes, we still know very little about how sexual
orientation influences the type of work individuals pursue. Given
observed disadvantage in the paid labor market, sexual minorities may
pursue alternative employment strategies, such as self-employment.
Further, many determinants of self-employment, such as educational and
occupational attainment, and family composition vary by sexual
orientation. These differences may provide unique opportunities and
incentives for sexual minority self-employment. In this paper, we ask
whether coupled gay men and lesbians differ from coupled heterosexuals
in their propensity for self-employment, and scrutinize the processes
underlying any differences. This is among the first papers to use
population-based surveys to explore the relationship between sexual
orientation and self-employment.
With few exceptions, research finds that gay men earn less than
heterosexual men in wage employment (Antecol et al. 2008; Arabsheibani
et al. 2004, 2005; Berg and Lien 2002; Black et al. 2003; Blandford
2003; Carpenter 2008; Waite 2015; Waite and Denier 2015, etc.). (1)
Audit studies also find that gay men are less likely to receive
interview callbacks when sexual orientation is signaled on a resume
(Adam 1981; Ahmed, Andersson, and Hammarstedt 2013; Tilcsik 2011;
Weichselbaumer 2003). Attitudinal research, self-report studies, and
court cases provide additional qualitative evidence of discrimination
against gay men in wage employment (Kranz and Cusick 2005; Smith 2008).
In this context, self-employment may serve as a protective strategy for
those who experience taste-based discrimination by employers or
customers in either the hiring process or during employment (Becker
[1957] 1971). This argument has been used to explain immigrants'
higher levels of self-employment, that is, that labor market barriers
and discrimination push immigrants into self-employment (Beaujot, Maxim,
and Thao 1994; Borjas 1990; Li 2000, 2001; Nakhaie 2015). Gay men are
also heavily concentrated in large urban areas (Black et al. 2002;
Denier and Waite 2014) and have historically resided in gay villages
within these urban areas (Compton and Baumle 2012; Ghaziani 2011, 2014,
2015). Although greater social acceptance has diffused gay men outside
these "gayborhoods," this urban concentration may provide
niche markets of likeminded gay men to sell their goods and services.
This argument has also been advanced in the ethnic enclave literature,
specifically that residential concentration of ethnic minorities
provides a comparable advantage to immigrant entrepreneurs who have
knowledge of language, culture, and unique tastes and preferences of the
immigrant community (Ley 2006; Sanders and Nee 1996; Wilson and Portes
1980).
Lesbians, on the other hand, earn more than heterosexual women
(Antecol et al. 2008; Baumle 2009; Carpenter 2008; Mueller 2014; Waite
2015; Waite and Denier 2015, etc.). After controlling for observed human
capital and demographic characteristics, these studies typically
interpret this wage advantage as statistical discrimination (Arrow 1973;
Phelps 1972). In other words, employers make assumptions about the
productivity of lesbians based on what they know about all lesbians
relative to women on average. Lesbians may be perceived as closer to the
unencumbered male worker because they are less likely to be married or
have children (Baumle 2009; Carpenter 2008; Denier and Waite 2014;
Mueller 2014; Waite 2015; Waite and Denier 2015). At the same time,
marriage and parenthood increase women's propensity for
self-employment because it provides flexibility for women who must
manage the dual responsibilities of work and family (Boden 1996; Budig
2006; Carr 1996; Connelly 1992).
Given that lesbians earn more than heterosexual women, have lower
rates of marriage, and fewer children, there may be fewer incentives to
move into self-employment. However, lesbians still earn less than
heterosexual men (Denier and Waite 2014; Waite 2015; Waite and Denier
2015) and the absence of a male wage earner results in lower household
incomes than heterosexual households. This may create more incentives
for lesbians to invest in their professional development, careers, and
earn more. In this case, self-employment may appear as an attractive
option for lesbians coping with lower household incomes.
Sexual minorities also differ from heterosexuals in their
accumulation of human capital, work activities, and household
specialization, all of which may affect their decision to be
self-employed. For example, gay men and lesbians are more highly
educated than heterosexuals (Antecol et al. 2008; Carpenter 2008; Waite
2015; Waite and Denier 2015), are more likely to work in gender-atypical
occupations (Antecol et al. 2008; Ueno, Roach, and Pena-Talamantes
2013b; Waite and Denier 2015), and have a more egalitarian division of
labor within the household (Goldberg, Smith, and Perry-Jenkins 2012;
Kurdek 2007). Differences in these characteristics may affect sexual
minority self-employment rates.
Of particular importance is how occupational sorting by sexual
orientation could affect sexual minorities' propensity for
self-employment. The prevalence of self-employment varies by occupation.
In highly paid professional occupations, such as law and medicine,
self-employment may be the norm. Self-employment is also more prevalent
in construction and other trade occupations. Incentives to move into
self-employment may also vary by occupations. Waite and Denier (2015)
found that sexual minority wage gaps vary significantly across 32
occupations in Canada. High levels of education allow gay men to sort
into highly paid occupations but within these highly paid occupations
gay men have the largest wage disadvantage relative to heterosexual men.
It reasons that the incentives to move into self-employment may be
greater in contexts where gay men are most disadvantaged. Conversely,
the lesbian wage advantage relative to heterosexual women also differs
across occupations, but with less variability (Waite and Denier 2015).
We may expect lower rates of self-employment in the occupations where
lesbians are most advantaged and higher rates of self-employment in
occupations where lesbians are least advantaged. To date, there has been
no research on whether employment barriers, urban coresidence, family
formation, or human capital accumulation result in different rates of
self-employment for sexual minorities, relative to heterosexuals. There
has also been no research on whether rates of self-employment vary by
occupation. This study fills these gaps in the literature.
The biggest limitation when studying the labor market experiences
of nonheterosexuals is the dearth of data including information on
sexual orientation, type of employment, earnings, and labor market
experiences. In an attempt to overcome these limitations, it has become
common practice for researchers to use census data and infer sexual
orientation from the relationship status between adult persons in the
household (Antecol et al. 2008; Baumle 2009; Denier and Waite 2015;
Klawitter 2011; Waite 2015; Waite and Denier 2015, etc.). Although
information on singles is lost from this approach, census data provides
a very large sample of cohabiting same- and opposite-sex respondents, as
well as relatively high-quality employment variables. Given that
self-employment is a relatively small fraction of all employment, a very
large sample of gay men and lesbians is needed to produce meaningful
estimates. The census is one of the only data sources where it is
possible to study self-employment and sexual orientation. However, the
loss of singles means that these results are only generalizable to same
and opposite sex couples.
This study bridges three distinct literatures--research on sexual
minority wage gaps, female self-employment, and the ethnic economy--and
presents a theoretical framework for how self-employment may vary by
sexual orientation. We pool cross-sectional data from the 2001 and 2006
Canadian Censuses and the 2011 National Household Survey (NHS) to
determine whether coupled sexual minorities differ from the coupled
heterosexual population in their propensity for self-employment. We
begin by exploring whether rates of self-employment differ by sexual
orientation in the aggregate and within occupational groups. Next, we
use multivariate logistic regression to compare the likelihood of
self-employment by sexual orientation. To establish that the motivation
to move into self-employment may vary by occupation, we estimate sexual
minority wage gaps across eight broad occupation groups and compare the
likelihood of self-employment within these occupations. We then explore
the determinants of self-employment and whether the propensity for
self-employment varies by occupation. We find significant within
occupation variation in the likelihood of self-employment for gay men
and lesbians relative to their heterosexual counterparts. We find that
gay men are less likely and lesbians more likely than heterosexuals to
be self-employed; however, there is significant variation across
occupations. Gay men are more likely to be self-employed in arts and
culture, sales and service, and natural and applied sciences, but less
likely in health-related occupations. Lesbians are much more likely to
be self-employed in health-related occupations, natural and applied
sciences, and arts and culture. Having children and being married are
significant predictors of self-employment for heterosexual women but not
lesbians.
LITERATURE REVIEW
This is among the first studies to explore differences in
self-employment by sexual orientation using nationally representative
data. To frame our analysis, we first build on the sexual minority wage
gaps literature in the United States and in Canada. We also draw on the
literature on immigrant economic incorporation, which emphasizes how
barriers to paid employment may push immigrants to pursue
self-employment and further highlights how ethnic enclaves provide
distinct opportunities for immigrant self-employment. We discuss how
similar mechanisms--barriers in the paid labor market and geographic
concentration--may impact sexual minorities' propensity for
self-employment. Last, we review the gender and self-employment
literature to determine whether sexual minorities share similar
motivations toward self-employment as heterosexuals.
Badgett (1995) was the first to apply econometric modeling
techniques to the study of sexual minority labor market outcomes in the
United States. Using the 1989 to 1991 U.S. General Social Survey, she
found that gay and bisexual men earned between 11 and 27 percent less
than comparable heterosexual men. Lesbian and bisexual women had
earnings that were statistically indistinguishable from heterosexual
women. Subsequent studies have generally found that gay men earn less
and lesbians earn more than their heterosexual counterparts in the
United States (Antecol et al. 2008; Berg and Lien 2002; Black, Makar,
Sanders and Taylor 2003; Blandford 2003; Carpenter 2008; Klawitter 2011,
etc.) and Canada (Carpenter 2008; Denier and Waite 2014; Harris 2012;
Lafrance, Warman, and Woolley 2009; Waite 2015; Waite and Denier 2015).
Explanations for sexual minority wage gaps are typically borrowed from
the gender pay gap literature. Researchers have argued that wage gaps
may, at least partially, be explained by differences in human capital,
labor force engagement, occupation and industrial sorting, and/or
differences in family composition (Allegretto and Arthur 2001; Antecol
et al. 2008; Waite and Denier 2015). However, residual wage gaps persist
after controlling for these characteristics. These residual wage gaps
for gay men are often interpreted as taste-based discrimination--that
is, customers and/or employers prefer not to interact or do business
with gay men (Antecol et al. 2008; Badgett 1995; Becker [1957] 1971;
Clain and Leppel 2001; Waite and Denier 2015). The wage advantage of
lesbians relative to heterosexual women is partially explained by
differences in hours and weeks worked. Lesbians' lower rates of
childbearing and the absence of a motherhood penalty also explain part
of their wage advantage (Baumle 2009; Waite and Denier 2015). Finally,
the persistence of a wage advantage for lesbians, relative to
heterosexual women, even after controlling for these characteristics may
be explained by positive statistical discrimination (Arrow 1973; Phelps
1972). That is, employers may perceive lesbians as less encumbered by
family and childcare responsibilities and thus a less risky investment
in terms of training and promotion.
Early research on gay men's and lesbians' labor market
experiences speculated that self-employment may be a protective strategy
for sexual minorities who face discrimination by employers, coworkers,
and customers (D'Augelli and Patterson 1995; Friskopp and
Silverstein 1995; Humphreys 1972; Russo 1982; Weinberg and Williams
1974; Woods and Lucas 1993). A similar argument has also been made to
explain sexual minorities' higher levels of education, that is,
they perceive occupations requiring higher levels of education as more
tolerant (Hewitt 1995; Ueno, Pena-Tolamantes, and Roach 2013a; Waite and
Denier 2015). It has also been used to explain why sexual minorities may
prefer working in the public sector, where wage determination practices
are less discretionary (Ahmed et al. 2013; Laurent and Mihoubi 2012;
Lewis 2010; Waite and Denier 2015). Historically, one of the only ways
gay men could be open about their sexual orientation in the workplace
was if they were self-employed. Self-employment may have been a
protective strategy for men who decided to live openly gay lives or who
feared discrimination and harassment if their sexual orientation were
exposed. Research on lesbians focused more on sex-typed discrimination
experienced by all women. It has been argued that never-married lesbians
are more likely to behave like men in the labor market, that is, invest
large amounts of time in education and their professional careers
because they will not have a man to provide for them in the future (Berg
and Lien 2002; D'Augelli and Patterson 1995; Fertitta 1984; Turner
1987). This may also increase the relative attractiveness of
self-employment for career-oriented lesbians. Unfortunately, much of
this early research relied on convenience samples rather than
population-based surveys.
Immigrants often work in self-employment to a larger degree than
the native-born population. The blocked mobility hypothesis suggests
that immigrants use self-employment as a protective shield against
discrimination and other labor market barriers. This argument has been
advanced to explain immigrants' higher rates of self-employment in
both the United States (Borjas 1990) and Canada (Beaujot et al. 1994).
In Canada, immigrants experience barriers in the labor market, such as
the devaluation of foreign experience, skills, training, and education
(Basran and Li 1998; Fong and Cao 2009; Gans 2009; Nakhaie 2006, 2007;
Phythian, Walters, and Anisef 2009). They may also experience
taste-based discrimination because employers and customers prefer to do
business with native-born Whites, rather than foreign-born visible
minorities (Pendakur and Pendakur 1998, 2002, 2011; Skuterud 2010).
These have been important factors pushing immigrants into
self-employment; however, earnings for self-employed immigrants in
Canada are generally lower than from other forms of income (Beaujot et
al. 1994; Li 2000; Nakhaie 2015). For example, Nakhaie (2015) used the
2006 Canadian Census and found most ethno-racial immigrant groups had
self-employment incomes that were lower than the incomes of comparable
immigrants working for wages and salaries.
An alternative hypothesis suggests that the concentration of
immigrants in ethnic enclaves within large cities presents unique
incentives and opportunities for self-employment and class mobility.
Research on ethnic enclave economies has been conducted not only in the
United States (Aldrich et al. 1985; Sanders and Nee 1996), but also in
Canada (Fong and Ooka 2002; Fong and Wilkes 2003; Ley 2006). Unlike the
blocked mobility hypothesis, the ethnic enclave argument suggests that
self-employment can be an avenue for upward mobility. Ethnic minorities
have knowledge about unique tastes, preferences, languages, and cultural
practices that may provide incentives pulling immigrants toward
self-employment. This knowledge creates distinct opportunities and an
advantage for moving into self-employment.
The literatures on immigrants' blocked mobility and ethnic
enclave living provide valuable insights into how gay men and lesbians
may choose their type of work, specifically self-employment. Gay men,
particularly older gay men who started their careers when there was
little social acceptance for alternative sexual lifestyles, may perceive
self-employment as a way to ameliorate wage disadvantage or overcome
discrimination. Although lesbians earn more than heterosexual women they
still earn much less than heterosexual men. The incentives for
self-employment and wage maximization may be greater for lesbians
because unlike heterosexual women, they will not have a higher earning
male partner contributing to household income. Another way that
minorities have attempted to overcome intolerance and potentially
discriminatory labor market conditions has been through urban and
enclave living. Although most literature has focused on ethnic enclaves,
more recently there is a growing interest in sexual minority enclave
living (Comptom and Baumle 2012; Ghaziani 2011, 2014, 2015). Sexual
minorities in Canada and the United States are concentrated in large,
high amenity urban areas (Black, Gates, Sanders, and Taylor 2002; Denier
and Waite 2014). Historically, gay men, and to some extent lesbians,
have also concentrated in gay villages within these areas (Comptom and
Baumle 2012; Ghaziani 2011, 2014), often referred to as gay villages or
"gayborhoods" (Ghaziani 2015). Although most of the literature
to date has focused on the gayborhoods in the United States,
Canada's three largest cities have their own gay
enclaves--Montreal's Gay Village, Toronto's Church and
Wellesley, and Vancouver's Davie Village. These enclaves provide
markets where minorities can do business with individuals with the
similar tastes, preferences, life experiences, and perhaps subcultural
affinity. Recently there has been some debate as to whether gayborhoods
have lost their relevance (Ghaziani 2011, 2014, 2015). We argue that the
advantages of geographic concentration are not solely derived from
residing in a gayborhood. When gay men and lesbians choose to reside
outside these gay villages, they often still have access to a niche
market of likeminded others within the city to sell their products and
services.
The incentive to move into self-employment may not be identical
across occupations. If wage gaps between sexual minorities and
heterosexuals vary by occupation, one may expect opportunity costs for
moving into self-employment would vary by occupation. Using U.S. Census
data, Elmsie and Tebaldi (2007) found that gay men's hourly wage
disadvantage, relative to married heterosexual men, was concentrated in
management, building and grounds cleaning, construction and extraction,
and processing occupations. On the other hand, they found no wage
discrimination for gay men in occupations with greater customer
interaction, which they argue suggests that discrimination may be
concentrated on the employer side. Using 2006 Canadian Census data,
Waite and Denier (2015) found significant variation in gay men's
wage disadvantage across 32 occupational groups and 20 industrial
categories. The largest wage disadvantage for gay men was in some of the
highest paid management and professional occupations. Promotions and
reward structures in these occupations tend to be more heavily
performance based, placing discretion in the hands of bosses and/or
peers who may be biased when evaluating the "worth" of an
employee or coworker. Similarly, Roth (2006) argues that the fraternity
culture in business and finance occupations can create barriers for
women's advancement by influencing how women are judged in
performance evaluations. Research has also linked performance pay to
greater wage inequality in general (Lemieux, MacLeod, and Parent 2009).
Waite and Denier (2015) also argue that the degree to which hegemonic
masculinity may be institutionalized within particular occupations may
place gay men at a greater disadvantage. The wage advantage for
lesbians, relative to heterosexual women, did vary across occupations
and industries, but to a far lesser degree than for gay men. Lesbians
earned less than heterosexual men in all occupations.
Sexual minorities also differ from heterosexuals in term of their
human capital, family formation, occupation, and industry of employment
choices. These differences may bring with them different opportunities
or incentives for self-employment. Single and coupled gay men and
lesbians are more highly educated than their heterosexual counterparts
(Antecol et al. 2008; Carpenter 2008; Muller and Arum 2004; Waite 2015;
Waite and Denier 2015). Education tends to provide individuals with
managerial skills and abilities that increase their propensity for
self-employment (Borjas 1986; Boijas and Bronars 1989; Evans and
Leighton 1989; Fuchs 1982; Rees and Shah 1986). Other studies from the
United Kingdom and Israel have found a curvilinear relationship between
education and self-employment, with self-employment being more prevalent
at both the low and high ends of the education spectrum (Luber et al.
2000; Meager, Kaiser, and Dietrich 1992; Shavit and Yucchtman-Yaar
2001). Sexual minorities are also more likely to work in the public or
nonprofit sector and in gender-atypical occupations or industries (Lewis
2010; Ueno et al. 2013b; Waite and Denier 2015). Opportunities and
incentives for self-employment will vary by occupation and across
different professions. Self-employment is a more common form of
employment arrangement in many professional occupations, such as
medicine and law.
Differences in family formation may also be an important factor in
explaining sexual minority differences in self-employment. This will be
particularly important when comparing lesbians to heterosexual women,
since the majority of family responsibilities fall on heterosexual
women's shoulders. Lesbians are less likely to be married and have
fewer children than heterosexual women (Baumle 2009; Carpenter 2008;
Mueller 2014; Waite 2015; Waite and Denier 2015). Same-sex couples also
have a more egalitarian division of labor within the household (Goldberg
et al. 2012; Kurdek 2007). Unlike heterosexual women, managing the dual
responsibilities of employment and family will be less relevant for
lesbians. This may decrease propensity for self-employment, specifically
in nonprofessional, nonmanagerial self-employment. Self-employment often
provides work flexibility, which is an incentive for women managing work
and family responsibilities. A number of early studies found that
marriage and parenthood increase women's propensity for
self-employment (Boden 1996, 1999; Carr 1996; Connelly 1992). However,
self-employed women are a heterogeneous group and not all enter
self-employment for the flexibility. Budig (2006) finds that there are
two groups of self-employed women. The first consists of those who enter
nonprofessional and nonmanagerial self-employment to manage the dual
responsibilities of work and family. The second group consists of women
in professional and managerial self-employment, for whom family factors
have little influence over their entry into self-employment. The latter
group consists of ambitious, single, childless women who enter
self-employment to advance their careers and their economic
opportunities. It is unclear whether all forms of partnership have the
same effect on women's propensity for self-employment. It is
reasonable to assume that marriage would bring greater familial
obligations than merely cohabitation, but to a lesser extent than the
movement from singlehood to marriage. For this reason, a sample
comprising exclusively couples will likely show a much smaller effect of
marriage on propensity for self-employment than a sample including
couples and singles. This is an unavoidable limitation of this study.
Coupled gay men are also far less likely than heterosexual men to
be married and have children (Mueller 2014; Waite 2015; Waite and Denier
2015). These are important determinants of self-employment for
heterosexual men, since marriage and the arrival of children often
results in heterosexual men increasing their work intensity and
productivity to wage maximize (Hodges and Budig 2010; Killewald 2012;
Lundberg and Rose 2000, 2002). This may also include professional and
career development and self-employment. Since gay men are less likely to
be married and have children this may decrease their overall motivation
for self-employment. When gay men do marry and have children a more
egalitarian division of labor may make specialization in either
workplace or household responsibilities less important and reduce the
effect of marriage and children on self-employment.
A final consideration is changes in antidiscrimination legislation
over the last few decades. The social and legal climate has changed
significantly since the 1960s, which would have impacted sexual
minorities of different ages in significant ways. In Canada, sexual
activity between members of the same sex was illegal up until 1969. In
1977, Quebec became the first province to include sexual orientation as
a protected class in the Quebec Charter of Rights and Freedoms (Smith
2008). Other provinces gradually followed suit. There has also been the
federal legalization of same-sex marriage in July 20, 2005.
Self-employment may have been a more important protective strategy for
older cohorts than for more recent ones. Younger lesbian cohorts may
have fewer incentives to enter self-employment because they are less
likely to have children and be married, which are important for
heterosexual women's transition into self-employment. The exclusion
of singles in this analysis will result in a slightly older sample. For
this group, self-employment may have been a more important protective
strategy to guard against intolerance and discrimination in the labor
market.
The current literatures on self-employment, sexual minorities'
labor market experiences, and immigration and enclave living raise
interesting questions and lead us to ask: (1) Do sexual minorities
differ from their heterosexual counterparts in their propensity for
self-employment? (2) Does propensity for self-employment vary across
occupations? (3) Do the determinants of self-employment differ by sexual
orientation?
DATA AND METHODS
Data
We use the 20 percent samples of the confidential 2001 and 2006
Census of Canada, as well as the 2011 NHS. The NHS replaced the
mandatory long-form Census in 2011. However, unlike the census, the NHS
was not mandatory and resulted in a larger nonresponse rate. The NHS had
a weighed response rate of 77.2 percent, but in highly populated census
metropolitan areas (CMAs) the response rates were higher (Statistics
Canada 2013). Given that most same-sex couples reside in major
metropolitan areas and/or larger CMAs, comparability should be less of a
concern between the NHS and census data.
Identifying Same-Sex Couples
The Canadian government began collecting information on same-sex
families in 2001 by including a "same-sex common law partner of
Person 1" response in the 2001 Census. On July 20, 2005 the
Canadian government legalized same-sex marriage. The 2006 Census
reflected this change by further allowing same-sex couples the option of
selecting "wife or husband of Person 1" or writing in their
relationship status. The quality of the same-sex response was high and
postcollection edits, using first names, were applied to correct for
misreporting of sex. Data quality studies were conducted by Statistics
Canada on the final same-sex responses and revealed "no
problems" with the final data in the 2006 Census (Statistics Canada
2007). Unfortunately, Statistics Canada became aware of a potential
overestimation of same-sex couples using the 2011 short-form Census,
which affected the quality of NHS estimates. (2) Although this
overestimation introduces a potential bias, this remains the only census
data in North America that contains any direct questions on same-sex
relationships.
Sample
The sample is restricted to employed individuals between the ages
of 25 and 64 in either same-sex or opposite-sex relationships. We
exclude self-employed farmers since the land and capital investment in
this type of self-employment differs from most other types of
self-employment. Those under 25 are excluded because they may still be
in school and not fully engaged in the labor market. Also, cohabitation
and marriage is less prevalent among younger individuals. Those over the
age of 64 are excluded since many individuals retire around this age. We
drop immigrants from our sample, since immigrants tend to have higher
rates of self-employment for reasons specific to their labor market
integration process, such as the devaluation of foreign credentials and
experience (Boijas 1986; Bradley 2004; Maxim 1992, etc.). (3) We also
exclude aboriginals from our sample because a significant proportion of
aboriginal population reside on First Nation reserves, which have very
different labor markets from the rest of Canada. The aboriginal
community also has a long history of marginalization and exclusion from
education and conventional employment opportunities. Rates of
self-employment are also significantly lower for aboriginals than
nonaboriginals (Hiebert 2002; Usalcas 2011). Last, we drop those
residing in Nunavut, Yukon, and the North West Territories, as the labor
markets in these scarcely populated areas are very different from the
rest of Canada.
Dependent Variable
Self-employment is operationalized using the class of worker
variable in the census, which refers to the type of work conducted in
the week prior to enumeration. The census distinguished between
self-employed with or without paid help and incorporated or not
incorporated. To maintain the largest sample of self-employed sexual
minorities possible, we aggregate type of self-employment. We ran a
series of sensitivity models to determine whether the propensity for
self-employment or determinants of self-employment differ by
self-employment type. Our findings are generally robust across all types
of self-employment. We mention the few cases where our sensitivity
models produced divergent outcomes.
To add support to our position that the motivation for
self-employment may vary by occupations, we estimate sexual minority
wage gaps across eight broad occupations. We estimate wage gaps using
total wages and salaries, which is a measure of gross employment
earnings, including tips commissions, and bonuses, before deductions for
items such as income taxes. Wages and salaries do not include other
sources of income, such as self-employment income, social assistance
transfers, or investment interest. All earnings have been adjusted for
inflation to 2010 dollars.
Independent Variables
We control for many known determinants of self-employment. Work
experience is controlled using the Mincer Proxy (age--years of
education--five years), (4,5) which we enter as five 10-year categories.
Education is controlled using four categories of educational attainment
(less than high school, high school certificate, college/trade
certificate, bachelor's degree, or above). We also control for
birth cohort, which is entered in five-year categories. Demographic
characteristics controlled for in our models include common-law status,
presence of children, visible minority status, province of residence,
residence in rural areas, and census/survey year. In some models, we
also control for occupation, coded using the National Occupational
Classification for Statistics broad occupation groups. (6) We run
separate models excluding doctors, lawyers, dentists, veterinarians,
optometrists, etc., since these professions tend to have higher rates of
self-employment and because sexual minorities' higher levels of
education allow them to disproportionately sort into them. Last, our
wage gap models also control for sector employment (public vs. private).
Studies have found that sexual minorities are more likely to work in the
public sector (Denier and Waite 2014; Lewis and Ng 2013; Waite and
Denier 2015), where more rigid wage structures, higher rates of
unionization, and a stronger entrenchment of antidiscrimination
legislation limit wage differentials (Ahmed et al. 2013; Laurent and
Mihoubi 2012; Waite and Denier 2015).
Methods of Analysis
We start with a sample description (Table 1) and then compare how
the rates of self-employment vary by age group (25-44 years old and 4564
years old) and eight broad occupations by sexual orientation (Table 2).
Next, we move on to multivariate logistic regression to determine
whether gay men and lesbians differ from their heterosexual counterparts
in their propensity for self-employment, controlling for covariates
(Table 3). We include a dichotomous same-sex partner variable in our
models. The same-sex partner dummy variable can be interpreted as the
difference in self-employment between sexual minorities and their
heterosexual counterparts. These models are run with and without
occupation controls. We also run a separate model where we drop
professionals, such as doctors, veterinarians, optometrists, lawyers,
surgeons, etc. We do so since self-employment may be standard practice
in these professions and not necessarily guided by individual
preference.
To establish that the motivation for gay men and lesbians to move
into self-employment may vary by occupation, we estimate sexual minority
wage gaps using ordinary least squares regression (OLS) for each broad
occupation (Table 4). We use a dichotomous same-sex partner variable,
which can be interpreted as the residual wage gap after controls. For
ease of interpretation and to minimize table space, we only provide the
coefficients that represent wage gaps between groups. These models
control for socio-demographic and human capital variables (see table
note for a list of all controls). To answer whether the propensity for
self-employment differs by occupation, we run a series of interactions
between sexual minority status and occupation (Table 5). Significant
interactions are interpreted in conjunction with the main effects of the
same-sex partner dummy variable to determine the degree to which the
propensity for self-employment varies for sexual minorities across
occupations. We present only the main and interaction effects in this
table. As in previous models, these are also run excluding
professionals.
Last, to compare the determinants of self-employment by sexual
orientation we run separate multivariate logistic regression models for
heterosexual men, gay men, heterosexual women, and lesbians (Table 6).
RESULTS
Descriptive Results
Table 1 provides a descriptive summary of our sample. Our
descriptive table confirms existing literature showing that gay men and
lesbians are slightly younger, are more educated, have less potential
experience, and are less likely to be married or have children compared
to their heterosexual counterparts. The mean age of respondents in our
sample is slightly older than would be expected because we exclude
singles. In terms of occupational distribution by sexual orientation, we
find that gay men work in greater proportion in a number of highly paid
occupations, such as management and business, finance, and
administration, but they are significantly underrepresented in trades,
transport, and manufacturing occupations. Lesbians are more likely than
heterosexual women to work in management occupations, but much less
likely to work in business, finance, and administration occupations.
Lesbians are also more likely to work in natural and applied sciences,
social services, education, and government occupations, and trades,
transport, and manufacturing.
Table 2 provides the proportion of our sample self-employed by
sexual orientation and by occupation. Our findings show very little
difference in the overall rates of self-employment between gay men and
heterosexual men, although gay men are slightly less likely to be
self-employed. Women overall have lower rates of self-employment: 10
percent of coupled heterosexual women in our sample are self-employed
compared to 15 percent of coupled heterosexual men. The difference
between lesbians and heterosexual women is very small. Those who are
older have higher rates of self-employment, and this holds true
regardless of sexual orientation. Although there is little difference in
self-employment by age group for gay men and heterosexual men, the
difference between lesbians and heterosexual women grows in the older
age group.
The proportion of our sample working in self-employment varies
considerably by occupation, with the highest rates of self-employment
found in arts, culture, and recreation and sport occupations, regardless
of gender or sexual orientation. Differences in rates of self-employment
between sexual minorities and their heterosexual counterparts diverge in
stark contract across occupations. For example, there are far more
self-employed heterosexual men in management, business, finance, and
administrative occupations, as well as health-related occupations.
Conversely, there is a slightly greater proportion of self-employed gay
men in arts, culture, recreation, and sport occupations, sales and
service, and natural and applied sciences.
Turning to women, there are more self-employed heterosexual women
in business, finance, and administration; social services, education,
and government; and sales and services occupations, whereas
self-employed lesbians have a greater proportion in applied sciences,
health-related occupations, and arts and culture occupations.
Multivariate Results
Our first research question asks whether sexual minorities differ
from their heterosexual counterparts in their propensity for
self-employment. The descriptive results provided in Table 2 show very
little difference in overall self-employment rates by sexual
orientation; however, there were noticeable differences in
self-employment by sexual orientation across occupations. In Table 3, we
further explore differences in self-employment using multivariate
logistic regression and control for various determinants of
self-employment. We run models with and without occupation controls
(models 1 and 2) and dropping professionals (model 3). We find no
difference between gay men and heterosexual men in their propensity for
self-employment in model 1; however, after introducing occupation
controls, gay men are roughly 10.2 percent less likely to be working in
self-employment. This means that occupational differences suppress an
otherwise statistically significant difference in self-employment, which
is likely because gay men tend to work in occupations that have high
rates of self-employment. The direction of this relationship holds after
removing professionals from the sample, but the same-sex coefficient is
not statistically significant. (7) There is also significant variation
in the likelihood of self-employment by occupation. Compared to social
science, education, and government occupations, our reference category,
men in most other occupations are more likely to be self-employed. The
one exception is sales and services occupations, where men are less
likely to be self-employed. Self-employment is most likely in
health-related occupations and arts, culture and recreation, and sport
occupations.
For women, all models show that lesbians have a much greater
propensity for self-employment than heterosexual women. In our fully
specified model, lesbians are roughly 16.9 percent more likely to be
self-employed than heterosexual women. This greater likelihood of being
self-employed remains after dropping professionals. (8) Unlike men, the
likelihood for self-employment was lower for many occupations. For
example, women are less likely to be self-employed in business, finance,
and administration; natural and applied sciences; and health-related
occupations, compared to employment in social science, education, and
government occupations. They are more likely to be self-employed in
arts, culture, recreation, and sport; sales and service; and
manufacturing, compared to the reference.
If sexual minority wage gaps vary considerably across occupations,
there is reason to suspect that motivations for moving into
self-employment will be higher in occupations that have greater wage
disadvantage. We estimate sexual minority wage gaps by occupation using
OLS regression and controlling for the standard wage determination
variables (Table 4). We find that overall gay men earn roughly 6 percent
less and lesbians 12 percent less than heterosexual men but lesbians
earn 7 percent more than heterosexual women (calculated using
([e.sup.[beta]] - l) x 100). We also find considerable variation in wage
gaps by broad occupation. Gay men earn significantly less than
heterosexual men in management; business; finance, and administration;
sales and service; and manufacturing, trades, and primary industry
occupations. The largest wage gaps were for gay men working in sales and
service occupations and manufacturing, trades, and primary industry.
Lesbians earn less than heterosexual men across all occupations. On the
other hand, comparing lesbians' earnings to heterosexual women
reveals a wage advantage in business, finance, and administration,
social science, education and government, sales and service, and the
biggest wage advantage for lesbians was in manufacturing, trades, and
primary industry occupations.
After establishing that sexual minority wage gaps vary across our
broad occupations, we tackle our second research question, which asks
whether the difference in the propensity for self-employment varies
across occupations. To answer this question, we introduce a series of
interaction terms between same-sex partner and occupation in Table 5.
These models are run with and without professionals (model 1 and 2).
Statistically significant interactions can be interpreted as gay men or
lesbians being more or less likely than their heterosexual counterparts
to be self-employed in a particular occupation. Our results indicate
that men are less likely to be self-employed in business, finance, and
administration occupations and health-related occupations, compared to
heterosexual men. However, gay men are more likely to be self-employed
in natural and applied sciences; arts, culture, and recreation; and
sales and service, compared to heterosexual men.
Lesbians are much more likely to be self-employed in natural and
applied sciences, health-related occupations, and arts, culture,
recreation, and sport occupations. Of particular note is the
significantly higher rate of self-employment for lesbians in
health-related occupations. This is likely because higher levels of
education move lesbians into higher level health occupations where
self-employment is more prevalent (e.g., medical doctors), whereas,
heterosexual women are more likely working in lower skilled health
occupations where self-employment is less prevalent (e.g., nurses and
assistants). Taken together, there is considerable variability in the
propensity for self-employment by sexual orientation. This would suggest
that there may be unique incentives, opportunities, or rewards for gay
men and lesbians who choose to work in self-employment in particular
occupations.
Next, we explore whether the determinants of self-employment differ
by sexual orientation. Table 6 addresses this research question by
running separate models by sexual orientation. We also ran these models
after dropping professionals but there was very little change in the
determinants of self-employment and these models are available upon
request. For men, more years of work experience increases the likelihood
of self-employment. This relationship is stronger for gay men than
heterosexual men, but significance is lost for gay men with 40+ years of
experience. For heterosexual women the pattern is not linear. The
likelihood of self-employment is highest for women with some experience
(10-19 years) and many years of experience (40+ years). For lesbians,
there is no statistically significant relationship between experience
and the likelihood of self-employment; however, the direction of the
coefficients suggests that lesbians with more experience may be less
likely to be self-employed. The reason for this is unclear.
Before turning to demographic determinants, we remind our reader
that our sample is limited to same- and opposite-sex couples.
Heterosexual men and women who are living common-law, rather than
married, are less likely to be self-employed. (9) Coupled heterosexuals
with children are more likely to be self-employed than those without
children. Marriage and children do not have an effect on gay men or
lesbian's likelihood for self-employment. Interestingly, although
the coefficients did not reach statistical significance, cohabiting
lesbians were more likely than married lesbians to be self-employed. Our
analysis also shows that partnered heterosexual women with children are
more likely to be self-employed than those without children. This may
suggest that balancing family responsibilities and work is less of a
motivation for lesbians to enter self-employment than heterosexual
women. Perhaps because coupled lesbians tend to have a more egalitarian
division of labor in the household (Goldberg et al. 2012; Kurdek 2007),
making the flexibility of self-employment less of an incentive. Caution
should be exercised interpreting these results for gay men given the
relatively small sample sizes for married same-sex couples with children
(see Table 1). Having a bachelor's degree or above increases the
likelihood of self-employment for all groups; however, the effect of
education appears to be stronger for gay men and lesbians. Independent
of potential experience, birth cohort was a significant determinant of
self-employment, with younger cohorts significantly less likely than
older cohorts to be self-employed.
DISCUSSION AND CONCLUSION
This is among the first studies to explore whether sexual
minorities differ from their heterosexual counterparts in their
propensity for self-employment. We borrow from the econometric
literature on sexual minority wage gaps, determinants of female
self-employment, and the immigrant self-employment and ethnic enclave
literatures to frame our analysis. At the aggregate, we find that gay
men are less likely and lesbians more likely than their heterosexual
counterparts to be working in self-employment. This overshadows
considerable differences across occupations in rates of self-employment.
Although gay men are less likely to be self-employed in business,
finance, and administrative occupations and health-related occupations,
they are more likely to be self-employed in natural and applied
sciences, arts, culture, recreation and sport occupations, and sales and
service occupations. With the exception of sales and service
occupations, these are not the occupations where gay men experience the
greatest wage disadvantage, which seems to be inconsistent with the
blocked mobility hypothesis. An alternative explanation may be that gay
men are pulled into self-employment in particular occupations because
their sexual orientation offers unique incentives or opportunities.
Adding further support to this argument is the findings that
self-employment is higher in sales and service and the arts. These are
occupations where knowledge of unique tastes and preferences of the gay
and lesbian communities may be particularly advantageous.
Lesbians are more likely to be self-employed overall but again this
conceals considerable variation by occupation. Our results show that
lesbians are more likely to be self-employed in health-related
occupations, natural and applied sciences, and arts and culture. Again,
this does not seem to be consistent with a blocked mobility hypothesis,
since these are not occupations where lesbians are most disadvantaged,
relative to heterosexual men. These are, however, occupations where
lesbians have earnings that are comparable to heterosexual women. The
finding that lesbians have higher rates of self-employment than
heterosexual women are further evidence that lesbians, at least coupled
lesbians, behave more like men in the labor market.
Marriage and having children are strong predictors of
self-employment among partnered heterosexual men and women; however,
they do not bear on the propensity of sexual minorities. Previous
studies have found that same-sex couples with children tend to have a
more egalitarian division of labor within the household; this may
decrease the attractiveness of self-employment as a strategy for
managing the dual responsibilities of work and domestic duties.
Recent evidence from the United States shows similar patterns of
self-employment for gay men and lesbians in the aggregate. Leppel (2016)
uses the 2012 American Community Survey and finds that heterosexual men
have the highest rates of self-employment, followed by gay men,
lesbians, and finally heterosexual women.
With the absence of population-based surveys that include
high-quality employment variables and questions on sexual orientation,
the couple approach has become standard practice when studying the labor
market experiences of sexual minorities. For this reason, these results
should not be generalized to those who are single. Due to small samples
of sexual minorities, we aggregated different types of self-employment
in this study. Sensitivity models show only small differences in our
findings depending on the type of self-employment. It may be worthwhile
for future studies to further untangle differences in type of
self-employment and the effect this has on self-employment wages for
sexual minorities. We have been cautious in our interpretation of the
results, especially the fatherhood effect for gay men because in some
cases small sample sizes may have produced large standard errors and
insignificant estimates. This is an unavoidable limitation. We exclude
the aboriginal and foreign-born populations since these groups have
either very low or much higher rates of self-employment. Untangling the
complex interaction between aboriginal status, nativity and sexual
orientation is outside the scope of this analysis; however, future
research may shed light on how foreign-born or aboriginal sexual
minorities negotiate labor market experiences.
References
Adam, B. 1981. "Stigma and Employability: Discrimination by
Sex and Sexual Orientation in the Ontario Legal Profession."
Canadian Review of Sociology and Anthropology 18(2):216-21.
Ahmed, A., L. Andersson and M. Hammarstedt. 2013. "Are Gay Men
and Lesbians Discriminated against in the Hiring Process?" Southern
Economic Journal 79(3):565-85.
Aldrich, H., J. Cater, T. Jones, D. McEvoy and P. Velleman. 1985.
"Ethnic Residential Concentration and the Protected Market
Hypothesis." Social Forces 63(4):996-1009.
Allegretto, S.A. and M.M. Arthur. 2001. "An Empirical Analysis
of Homosexual/Heterosexual Male Earnings Differentials: Unmarried and
Unequal?" Industrial and Labor Relations Review 54(3):631-46.
Antecol, H., A. Jong and M. Steinberger. 2008. "The Sexual
Orientation Wage Gap: The Role of Occupational Sorting and Human
Capital." Industrial and Labor Relations Review 61(4):518-43.
Arabsheibani, G.R., A. Marin and J. Wadsworth. 2004. "In the
Pink: Homosexual-Heterosexual Wage Differentials in the UK."
International Journal of Manpower 25(3/4):343-54.
Arabsheibani, G.R., A. Marin and J. Wadsworth. 2005. "Gay Pay
in the UK." Economica 72:333-47.
Arrow, K. 1973. "The Theory of Discrimination." Pp. 3-33
in Discrimination in the Labor Markets, edited by O. Ashenfelter and A.
Rees. Princeton, NJ: Princeton University Press.
Badgett, M.V.L. 1995. "The Wage Effects of Sexual Orientation
Discrimination." Industrial and Labor Relations Review
48(4):726-39.
Basran, G. and Z. Li. 1998. "Devaluation of Foreign
Credentials as Perceived by Visible Minority Professional
Immigrants." Canadian Ethnic Studies 30:7-23.
Baumle, A.K. 2009. "The Cost of Parenthood: Unravelling the
Effects of Sexual Orientation and Gender on Income." Social Science
Quarterly 90(4):983-1002.
Beaujot, R., P.S. Maxim and J.Z. Thao. 1994. "Self-Employment
among Immigrants: A Test of the Blocked Mobility Hypothesis."
Canadian Studies in Population 21(2):81-96.
Becker, G. [1957] 1971. The Economics of Discrimination. 2d ed.
Chicago, IL: University of Chicago Press.
Berg, N. and D. Lien. 2002. "Measuring the Effect of Sexual
Orientation on Income: Evidence of Discrimination." Contemporary
Economic Policy 20(4):394-414.
Black, D., G. Gates, S. Sanders and L. Taylor. 2002. "Why Do
Gay Men Live in San Francisco?" Journal of Urban Economics
51(l):54-76.
Black, D., G. Gates, S. Sanders and L. Taylor. 2003. "The
Earnings Effects of Sexual Orientation." Industrial and Labor
Relations Review 56(3):449-69.
Blandford, J.M. 2003. "The Nexus of Sexual Orientation and
Gender in the Determination of Earnings." Industrial and Labor
Relations Review 56(4):622-42.
Boden, R. 1996. "Gender and Self-Employment Selection: An
Empirical Assessment." Journal of Socio-Economics 25:671-82.
Boden, R. 1999. "Flexible Working Hours, Family
Responsibilities, and Female Self-Employment: Gender Differences in
Self-Employment Selection." American Journal of Economics and
Sociology 58(l):71-83.
Boijas, G.J. 1986. "The Self-Employment Experiences of
Immigrants." Journal of Human Resources 21:485-506.
Borjas, G.J. 1990. Friends or Strangers, the Impact of Immigrants
on the U.S. Economy. New York: Basic Books.
Boijas, G.J. and S.G. Bronars. 1989. "Consumers Discrimination
and Self-Employment." Journal of Political Economy 97:581-605.
Bradley, D.E. 2004. "A Second Look at Self-Employment and the
Earnings of Immigrants." International Migration Review
38(2):457-583.
Budig, M. 2006. "Intersections on the Road to Self-Employment:
Gender, Family and Occupational Class." Social Forces
84(4):2223-39.
Carpenter, C. 2005. "Self-Reported Sexual Orientation and
Earnings: Evidence from California." Industrial and Labor Relations
Review 58(2):258-73.
Carpenter, C. 2008. "Sexual Orientation, Work, and Income in
Canada." Canadian Journal of Economics 41(4): 1239-61.
Carr, D. 1996. "Two Paths to Self-Employment? Women's and
Men's Self-Employment in the United States, 1980." Work and
Occupations 23:26-53.
Clain, S.H. and K. Leppel. 2001. "An Investigation into Sexual
Orientation Discrimination as an Explanation for Wage Differences."
Applied Economics 33(1):37-47.
Compton, D.R. and A.K. Baumle. 2012. "Beyond the Castro: The
Role of Demographics in the Selection of Gay and Lesbian Enclaves."
Journal of Homosexuality 59(10):1327-55.
Connelly, R. 1992 "Self-Employment and Providing Child
Gore." Demography 29(1):17-29.
D'Augelli, A.R. and C.J. Patterson. 1995. Lesbian, Gay and
Bisexual Identities over the Lifespan: Psychological Perspectives. New
York: Oxford University Press.
Denier, N. and S. Waite. 2014. "Gay Pay in Canadian Cities:
Local Labour Market Effects on Sexual Minority Earnings Gaps."
Presented at the Canadian Population Society Annual Meeting, May 30, St.
Catharines, Ontario.
Drydakis, N. 2012. "Sexual Orientation and Labour Relations:
New Evidence from Athens, Greece." Applied Economics 44:2653-65.
Elmslie, B. and E. Tebaldi. 2007. "Sexual Orientation and
Labor Market Discrimination." Journal of Labour Research 28:436-53.
Evans, D. and L. Leighton. 1989. "Some Empirical Aspects of
Entrepreneurship." American Economic Review 25:519-35.
Fertitta, S. 1984. "Never Married Women in the Middle Years. A
Comparison of Lesbians and Heterosexual." Wright University, Los
Angeles, CA. Unpublished doctoral dissertation.
Fong, E. and X. Cao. 2009. "Effects of Foreign Education on
Immigrant Earnings." Canadian Studies in Population 36:87-110.
Fong, E. and E. Ooka. 2002. "The Social Consequences of
Participating in the Ethnic Economy." International Migration
Review 36(1): 125-46.
Fong, E. and R. Wilkes. 2003. "Racial and Ethnic Residential
Patterns in Canada." Sociological Forum 18(4):577-602.
Friskopp, A. and S. Silverstein. 1995. Straight Jobs /Gay Lives.
New York: Scribner.
Fuchs, V. 1982. "Self-Employment and Labor Force Participation
of Older Males." Journal of Human Resources 17(3):339-57.
Gans, H.J. 2009. "First Generation Decline: Downward Mobility
among Refugees and Immigrants." Ethnic and Racial Studies 32(9):
1658-70.
Ghaziani, A. 2011. "Post-Gay Collective Identity
Construction." Social Problems 58(1):99-125.
Ghaziani, A. 2014. "Measuring Urban Sexual Cultures."
Theory and Society 43:371-93.
Ghaziani, A. 2015. There Goes the Gayborhood? Princeton, NJ:
Princeton University Press.
Goldberg, A., J. Smith and M. Perry-Jenkins. 2012. "The
Division of Labor in Lesbian, Gay, and Heterosexual New Adoptive
Parents." Journal of Marriage and Family 74(4):812-28.
Harris, B. 2012. "Less Pay for a Less Stressful Day. Sexual
Orientation and Household Composition in the Labor Market." Simon
Fraser Department of Economics Working Paper.
Hewitt, C. 1995. "The Socioeconomic Position of Gay Men: A
Review of the Evidence." American Journal of Economics and
Sociology 54(4):431-79.
Hiebert, D. 2002. "Economic Associations of Immigrants
Self-Employment in Canada." International Journal of
Entrepreneurial Behaviour & Research 8(l/2):93-112.
Hodges, M.J. and M.J. Budig. 2010. "Who Gets the Daddy Bonus?
Organization Hegemonic Masculinity and the Impact of Fatherhood on
Earnings." Gender & Society 24:717-45.
Hou, F. and S. Coulombe. 2010. "Earnings Gaps for
Canadian-Born Visible Minorities in the Public and Private
Sectors." Canadian Public Policy 36(1):29-43.
Humphreys, L. 1972. Out of the Closets: The Sociology of Homosexual
Liberation. Englewood Cliffs, NJ: Prentice-Hall, Inc.
Killewald, A. 2012. "A Reconsideration of the Fatherhood
Premium: Marriage, Coresidence, Biology and Fathers' Wages."
American Sociological Review 78(1):96-116.
Klawitter, M. 2011. "Multilevel Analysis of the Effects of
Antidiscrimination Policies of Earnings by Sexual Orientation."
Journal of Policy Analysis and Management 30(2):334-58.
Kranz, R. and T. Cusick. 2005. Gay Rights. New York, NY: Facts on
File, Inc.
Kurdek, L. 2007. "The Allocation of Household Labor by
Partners in Gay and Lesbian Couples." Journal of Family Issues
28:132-48.
Lafrance, A., C. Warman and F. Woolley. 2009. "Sexual Identity
and the Marriage Premium." Queen's Economics Department
Working Paper No. 1219.
Laurent, T. and F. Mihoubi. 2012. "Sexual Orientation and Wage
Discrimination in France: The Hidden Side of the Rainbow." Journal
of Labor Research 33:487-527.
Lemieux, T., W. MacLeod and D. Parent. 2009. "Performance Pay
and Wage Inequality." Quarterly Journal of Economics 124(1): 1-49.
Leppel, K. 2016. "The Incidence of Self-Employment by Sexual
Orientation." Small Business Economics 46(3):347-63.
Lewis, G. 2010. "Modeling Nonprofit Employment: Why Do So Many
Lesbians and Gay Men Work for Nonprofit Organizations?"
Administration & Society 42(6):720-48.
Lewis, G.B. and E.S. Ng. 2013. "Sexual Orientation, Work
Values, Pay, and Preference for Public and Nonprofit Employment:
Evidence from Canadian Postsecondary Students." Canadian Public
Administration 56(4):542-64.
Ley, D. 2006. "Explaining Variation in Business Performance
among Immigrant Entrepreneurs in Canada." Journal of Ethnic and
Migration Studies 32(5):743-63.
Li, P.S. 2000. "Economic Returns of Immigrants'
Self-Employment." Canadian Journal of Sociology 25(1):1-16.
Li, P.S. 2001. "Immigrants' Propensity to
Self-Employment: Evidence from Canada." International Migration
Review 35(4):1106-28.
Luber, S., H. Lohmann, W. Mueller and P. Barbieri. 2000. "Male
Self-Employment in Four European Countries." International Journal
of Sociology 30(3):5-44.
Lundberg, S. and E. Rose. 2000. "Parenthood and the Earnings
of Married Men and Women." Labour Economics 7:689-710.
Lundberg, S. and E. Rose. 2002. "The Effects of Sons and
Daughters on Men's Labor Supply and Wages." Review of
Economics and Statistics 84(2):251-68.
Maxim, P. 1992. "Immigrants, Visible Minorities, and
Self-Employment." Demography 29:181-91.
Mueller, R. 2014. 'Wage Differentials of Males and Females in
Same-Sex and Different-Sex Couples in Canada, 2006--2010." Canadian
Studies in Population 41(3-4):105-16.
Muller, W. and R. Arum. 2004. "Self-Employment Dynamics in
Advanced Economies." Pp. 1-35 in Reemergence of Self-Employment: A
Comparative Study of Self-Employment Dynamics and Social Inequality,
edited by R. Arum and W. Muller. Princeton, NJ: Princeton University
Press.
Nakhaie, M.R. 2006. "A Comparison of the Earnings of the
Canadian Native-Born and Immigrants, 2001." Canadian Ethnic Studies
38(2):19-46.
Nakhaie, M.R. 2007. "Ethnocultual Origins, Social Capital, and
Earnings." International Migration & Integration 8:307-25.
Nakhaie, M.R. 2015. "Economic Benefits of Self-Employment for
Canadian Immigrants." Canadian Review of Sociology 52(4):377-401.
Pendakur, K. and R. Pendakur. 1998. "The Colour of Money:
Earnings Differentials among Ethnic Groups in Canada." Canadian
Journal of Economics 31(3):518-48.
Pendakur, K. and R. Pendakur. 2002. "Colour My World: Have
Earnings Gaps for Canadian-Born Ethnic Minorities Changed over
Time?" Canadian Public Policy 28(4):489-512.
Pendakur, K. and R. Pendakur. 2011. "Colour by Numbers:
Minority Earnings in Canada 1995-2005." Journal of International
Migration and Integration 12(3):305-29.
Phelps, E. 1972. "The Statistical Theory of Racism and
Sexism." American Economic Review 62:659-61.
Phythian, K., D. Walters and P. Anisef. 2009. "Entry Class and
the Early Employment Experience of Immigrants in Canada." Canadian
Studies in Population 36:363-82.
Rees, H. and A. Shah. 1986. "An Empirical Analysis of
Self-Employment in the U.K." Journal of Applied Econometrics
1:95-108.
Roth, L.M. 2006. Selling Women Short: Gender and Money on Wall
Street. Princeton, NJ: Princeton University Press.
Russo, A.J. 1982. "Power and Influence in the Homosexual
Community: A Study of Three California Cities." Dissertation
Abstracts International, 43, 561B. (University Microfilms No.
DA8215211).
Sanders, J.M. and V. Nee. 1996. "Immigrant Self-Employment:
The Family as Social Capital and the Value of Human Capital."
American Sociological Review 61(2):231-49.
Shavit, Y. and E. Yuchtman-Yaar. 2001. "Ethnicity, Education
and Other Determinants of Self-Employment in Israel." International
Journal of Sociology 31(1):59-91.
Skuterud, M. 2010. "The Visible Minority Earnings Gap across
Generation of Canadians." Canadian Journal of Economics
43(3):860-81.
Smith, M. 2008. Political Institutions and Lesbian and Gay Rights
in the United States and Canada. New York: Routledge.
Statistics Canada. 2007. "Families and Households, 2006
Census--Families Reference Guide, 2006 Census." Statistics Canada.
Catalogue number 97-553-G.
Statistics Canada. 2013. "Immigration and Enthnocultural
Diversity in Canada." Statistics Cariada Analytic Document.
Catalogue no. 99-010-X2011001.
Tilcsik, A. 2011. "Pride and Prejudice: Employment
Discrimination against Openly Gay Men in the United States."
American Journal of Sociology 117(2):586-626.
Turner, B. 1987. "Developmental Perspective on Issues for
Lesbians at Midlife." Paper Presented at the Meeting of the
American Psychological Association, August, Atlanta, GA.
Ueno, K., A.E. Pena-Talamantes and T. Roach. 2013a. "Sexual
Orientation and Occupational Attainment." Work and Occupations
40:3-37.
Ueno, K., T. Roach and A.E. Pena-Talamantes. 2013b. "Sexual
Orientation and Gender Typicality of the Occupation in Young
Adulthood." Social Forces 92(1):81-108.
Usalcas, J. 2011. "Aboriginal People and the Labour Market:
Estimates from the Labour Force Survey, 2008-2010." Statistics
Canada. Catalogue number 71-588-X, no. 3.
Waite, S. 2015. "Does it Get Better? A Quasi-Cohort Analysis
of Sexual Minority Wage Gaps." Social Science Research 54:113-30.
Waite, S. and N. Denier. 2015. "Gay Pay for Straight Work:
Mechanisms Generating Disadvantage." Gender & Society
29(4):561-88.
Weichselbaumer, D. 2003. "Sexual Orientation Discrimination in
Hiring." Labor Economics 10:629-42.
Weinberg, M. and C. Williams. 1974. Male Homosexuals: The Problems
and Adaptations. New York: Oxford University Press.
Wilson, K. and A. Portes. 1980. "Immigrant Enclaves: An
Analysis of the Labor Market Experiences of Cubans in Miami."
American Journal of Sociology 86:295-319.
Woods, J.D. and J.H. Lucas. 1993. The Corporate Closet: The
Professional Lives of Gay Men in America. Toronto: Maxwell Macmillan
Canada.
Sean Waite and Nicole Denier
McGill University
We thank Michael R. Smith for his comments on earlier versions of
this manuscript. We also thank Anthony Lombardi and Marie-Eve Gagnon for
assisting with the translation of our abstract. Sean Waite acknowledges
doctoral support from the Social Sciences and Humanities Research
Council of Canada. The analysis presented in this paper was conducted at
the Quebec Interuniversity Centre for Social Statistics, which is part
of the Canadian Research Data Centre Network. The services provided by
QICSS are made possible by the support of the SSHRC, CIHR, CFI,
Statistics Canada, FRQSC, and Quebec universities. The views expressed
in this paper are those of the authors, and not necessarily those of the
CRDCN or its partners.
Sean Waite, Department of Sociology, McGill University, Room 718,
Leacock Building, 855 Sherbrooke Street West, Montreal, Quebec, Canada
H3A 2T7. E-mail: sean.waite@mail.mcgill.ca
(1.) There are a few deviations from these findings, which appear
to be driven by the uniqueness of the data sets used (see Carpenter
2005; Mueller 2014). Using the California Health Interview Survey,
Carpenter (2005) found no wage disadvantage for gay men. He suggests
this could be the result of more liberal attitudes, the large and
historical importance of gay communities, as well as strong
antidiscrimination legislation in California. Mueller (2014) pooled
Canadian General Social Survey data from 2006 to 2010 and found that gay
men had earnings that were indistinguishable from heterosexual men and
lesbians earned 16 percent more than comparable heterosexual women. The
deviation from previous studies may be the result of small sample sizes;
estimates were drawn from 90 gay men and 118 lesbians.
(2.) In 2011, Statistics Canada counted roughly 65,000 same-sex
couples, of which roughly 20,000 were married. The overestimation at the
national level ranged between 0 and 3,800 individuals.
(3.) Recent estimates show that 20.6 percent of the Canadian
population are foreign born (Statistics Canada 2013).
(4.) A "years of schooling" variable was not included in
the 2006 Census; therefore, a variable must be constructed using the
number of years of schooling for specific diploma obtainment (see Hou
and Coulombe 2010).
(5.) One limitation of the Mincer proxy is its inability to account
for periods of part-time employment and/or absence from the labor
market. This may be a source of bias for heterosexual women who will
have more absences from the labor market for childbearing and rearing
than all men and lesbians. We expect little bias in the estimates for
gay men relative to heterosexual men since gay men have few children and
most children rearing responsibilities fall to women in heterosexual
couples.
(6.) We collapse the following three broad occupations into a
single category due to low cell counts for both gay men and lesbians:
(1) trades, transportation, and equipment operations and related
occupations; (2) natural resources, agriculture, and related production
occupations; and (3) occupations in manufacturing and utilities.
(7.) The direction of the relationship held across sensitivity
models. Both models 1 and 2 were statistically significant when limiting
our sample to wage earners and self-employed not incorporated. When
limiting our sample to wage earners and the self-employed without paid
help no same-sex male partner coefficients were statistically
significant.
(8.) This relationship held across all sensitivity models.
(9.) Sensitivity models where we limit our sample to self-employed
without paid help find that heterosexual men living common law are more
likely to be self-employed than married heterosexual men. We find that
same divergent result when we limit our sample to wage earners and
self-employed not incorporated.
Table 1
Sample Description
Heterosexual Gay Heterosexual Lesbian
men men women women
Age (mean) 44.19 42.78 42.97 42.64
Potential experience 24.44 21.63 22.94 21.47
(mean)
Married 75.74 12.40 76.46 15.74
Children 63.83 3.59 62.61 17.42
Visible minority 1.68 2.11 1.75 1.91
Rural 24.03 10.74 24.76 15.54
No education 15.49 6.10 11.25 6.00
certificate or
diploma
High school 22.28 18.90 24.93 18.52
certificate or
equivalent
College, CEGEP *, 41.28 35.66 39.99 36.32
trade, or
apprenticeship
BA or above 20.96 39.35 23.83 39.16
Before 1950 14.90 7.71 11.15 6.72
1950-1954 13.82 11.11 12.87 11.62
1955-1959 16.59 16.34 16.73 16.06
1960-1964 16.83 22.91 17.52 22.84
1965-1969 13.54 17.00 14.27 16.20
1970-1974 11.91 10.53 12.79 11.65
1975-1979 8.07 8.23 9.15 8.64
1980-1984 3.79 5.28 4.72 5.36
1985 or later 0.56 0.90 0.80 0.90
Management occupations 15.84 18.42 9.27 13.10
Business, finance, and 9.61 16.79 30.77 19.13
administration
Natural and applied 10.13 8.47 2.81 5.88
sciences
Health and related 2.28 6.15 11.19 9.04
occupations
Social services, 6.00 12.80 14.36 19.91
education, and
government
Art, culture, 2.13 8.10 3.13 5.65
recreation, and
sport
Sales and service 14.10 20.30 22.05 17.28
Manufacturing, trades, 39.91 8.97 6.41 10.00
and primary industry
N 22,27,745 18,950 20,66,650 17,250
Notes: Sample includes employed individuals between the ages of 25
and 64 who are married or in a common-law relationship. The sample
excludes immigrants, aboriginal people, and those with self-
employment farm income. All estimates weighted using individual
population weights.
* College d'enseignement general et professionnel.
Table 2
Proportion Self-Employed by Sexual Orientation (Select
Variables)
Heterosexual Gay Heterosexual Lesbian
men men women women
Total (percent) 14.78 13.80 9.58 10.35
25-44 11.96 11.63 8.74 8.91
45-64 17.67 16.62 10.59 12.13
Occupation
Management 21.36 15.06 15.03 11.59
occupations
Business, finance, 12.04 6.40 6.48 4.80
administration
Natural and applied 10.63 12.03 7.47 11.17
sciences
Health-related 34.60 20.65 6.73 17.24
occupations
Social service, 11.90 8.30 8.55 6.76
education, and
government
Art, culture, 36.25 37.12 32.18 36.22
recreation,
and sport
Sales and service 9.45 12.29 9.87 7.58
Manufacturing, 13.92 12.38 12.70 10.00
trades, primary
industry
N 22,27,745 18,950 20,66,650 17,250
Notes: Sample includes employed individuals between the ages of 25
and 64 who are married or in a common-law relationship. The sample
excludes immigrants, aboriginal people, and those with self-
employment farm income. All estimates weighted using individual
population weights.
Table 3
Determinants of Self-Employment Odds Ratios from Logistic Regression,
Census of Canada 2001 to 2011
Men
1 2
B SE B SE
Same-sex partner 1.021 .026 0.898 *** .024
Work experience (ref: 0-9 years)
10-19 years 1.372 .018 1.400 *** .019
20-29 years 1.475 *** .026 1.556 *** .027
30-39 years 1.409 *** .030 1.554 *** .033
40+ years 1.438 *** .036 1.646 *** .042
Demographic characteristics
Common law (ref: 0.962 *** .006 0.966 *** .006
married)
Have children (ref: 1.043 *** .005 1.046 *** .006
no children)
Visible minority 1.031 .020 0.998 .019
(ref: White)
Education (ref: HS degree)
Less than HS degree 1.058 *** .008 1.088 *** .009
College/trade degree 0.962 *** .006 0.950 *** .006
BA+ 1.352 *** .010 1.218 *** .010
Birth cohort (ref: before 1950)
1950-1954 0.757 *** .007 0.762 *** .007
1955-1959 0.689 *** .007 0.701 *** .008
1960-1964 0.645 *** .008 0.670 *** .009
1965-1969 0.594 *** .009 0.629 *** .010
1970-1974 0.493 *** .009 0.531 *** .010
1975-1979 0.393 *** .009 0.434 *** .010
1980-1984 0.337 *** .010 0.384 *** .011
1985 or later 0.280 *** .019 0.328 *** .023
Occupations (ref: social science, education, government)
Management 2.119 *** .024
occupations
Business, finance, 1.115 *** .015
and administration
Natural and applied 1.036 ** .013
sciences
Health-related 4.364 *** .064
occupations
Art, culture, 5.204 *** .081
recreation, and
sport
Sales and service 0.869 *** .011
Manufacturing, 1.283 *** .015
trades, primary
industry
Constant 0.141 *** .003 0.094 *** .002
N 22,46,695 22,46,695
Men Women
3 1
B SE B SE
Same-sex partner 0.954 .026 1.266 *** .039
Work experience (ref: 0-9 years)
10-19 years 1.413 *** .020 1.419 *** .021
20-29 years 1.634 *** .030 1.259 *** .026
30-39 years 1.692 *** .038 1.190 *** .031
40+ years 1.813 *** .048 1.431 *** .045
Demographic characteristics
Common law (ref: 0.970 *** .006 0.831 *** .007
married)
Have children (ref: 1.029 *** .006 1.062 *** .007
no children)
Visible minority 0.962 .020 0.904 *** .021
(ref: White)
Education (ref: HS degree)
Less than HS degree 1.086 *** .009 1.044 *** .011
College/trade degree 0.974 *** .006 1.100 *** .008
BA+ 1.082 *** .009 1.216 *** .011
Birth cohort (ref: before 1950)
1950-1954 0.768 *** .007 0.808 *** .010
1955-1959 0.721 *** .008 0.741 *** .011
1960-1964 0.712 *** .010 0.734 *** .013
1965-1969 0.689 *** .011 0.695 *** .014
1970-1974 0.598 *** .012 0.621 *** .015
1975-1979 0.496 *** .012 0.548 *** .015
1980-1984 0.447 *** .014 0.498 *** .018
1985 or later 0.387 *** .027 0.406 *** .025
Occupations (ref: social science, education, government)
Management 4.356 *** .068
occupations
Business, finance, 2.280 *** .039
and administration
Natural and applied 2.132 *** .036
sciences
Health-related 3.962 *** .087
occupations
Art, culture, 10.641 * .202
recreation, and
sport
Sales and service 1.727 *** .029
Manufacturing, 2.506 *** .041
trades, primary
industry
Constant 0.043 *** .001 0.088 *** .002
N 22,11,995 20,83,900
Women
2 3
B SE B SE
Same-sex partner 1.169 *** .037 1.147 *** .038
Work experience (ref: 0-9 years)
10-19 years 1.383 *** .021 1.519 *** .024
20-29 years 1.225 *** .025 1.442 *** .031
30-39 years 1.184 *** .031 1.462 *** .040
40+ years 1.432 *** .046 1.781 *** .059
Demographic characteristics
Common law (ref: 0.796 *** .006 0.792 ** .007
married)
Have children (ref: 1.100 *** .007 1.095 *** .008
no children)
Visible minority 0.916 * * .022 0.853 *** .022
(ref: White)
Education (ref: HS degree)
Less than HS degree 0.969 ** .011 0.964 *** .011
College/trade degree 1.136 *** .009 1.190 *** .010
BA+ 1.117 *** .012 0.979 .011
Birth cohort (ref: before 1950)
1950-1954 0.804 *** .010 0.813 *** .010
1955-1959 0.732 *** .011 0.745 *** .011
1960-1964 0.721 *** .013 0.752 *** .014
1965-1969 0.685 *** .014 0.741 *** .016
1970-1974 0.605 *** .015 0.685 *** .017
1975-1979 0.528 *** .015 0.625 *** .018
1980-1984 0.481 *** .017 0.599 *** .022
1985 or later 0.397 *** .025 0.525 *** .033
Occupations (ref: social science, education, government)
Management 1.855 *** .022 1.918 *** .023
occupations
Business, finance, 0.712 *** .008 0.712 ** .009
and administration
Natural and applied 0.925 *** .019 0.972 .020
sciences
Health-related 0.751 *** .010 0.510 *** .008
occupations
Art, culture, 5.284 *** .070 5.567 *** .076
recreation, and
sport
Sales and service 1.166 *** .014 1.156 .015
Manufacturing, 1.520 *** .023 1.504 *** .023
trades, primary
industry
Constant 0.080 * .002 0.067 *** .002
N 20,83,900 20,63,510
Notes: * p [less than or equal to] .05; ** p [less than or equal to]
.01; *** p [less than or equal to] .001.
Sample includes employed individuals between the ages of 25 and 64
who are married or in a common-law relationship. The sample excludes
immigrants, aboriginal people, and those with self-employment farm
income. All models control province of residence, residence in a
rural area, and survey year. Model 3 excludes professionals. All
estimates weighted using individual population weights.
Table 4
Sexual Minority Wage Gaps by Occupation, Coefficients Only
Gay men Lesbians
(vs. hetero. men) (vs. hetero. men)
B SE B SE
Overall -0.066 *** .008 -0.133 *** .007
N 18,61,250 18,60,340
Management -0.063 *** .016 -0.196 *** .021
occupations 2,79,440 2,78,430
Business, -0.032 * .016 -0.142 *** .015
finance, and 1,86,630 1,86,695
administration
Natural and 0.060 .021 -0.123 *** .026
applied 2,04,895 2,04,360
sciences
Health-related 0.044 .029 -0.090 .031
occupations 33,495 33,835
Social science, 0.017 .018 -0.055 *** .014
education, and 1,21,035 1,21,950
government
Art, culture, -0.021 .046 -0.091 *** .051
recreation, and 32,975 32,620
sport
Sales and service -0.136 *** .017 -0.176 *** .017
2,74,240 2,73,775
Manufacturing, -0.130 *** .030 -0.156 *** .025
trades, 7,28,550 7,28,680
primary
industry
Lesbians
(vs. hetero. women)
B SE
Overall 0.066 *** .007
N 18,06,335
Management 0.030 .020
occupations 1,61,940
Business, 0.032 * .015
finance, and 5,76,440
administration
Natural and 0.015 .026
applied 54,690
sciences
Health-related 0.000 .028
occupations 2,10,120
Social science, 0.038 ** .013
education, and 2,70,920
government
Art, culture, -0.060 .051
recreation, and 46,605
sport
Sales and service 0.175 *** .017
3,78,795
Manufacturing, 0.208 *** .030
trades, 1,06,385
primary
industry
Notes: * p [less than or equal to] .05; ** p [less than or equal to]
.01; *** p [less than or equal to] .001.
Sample includes employed individuals between the ages of 25 and 64
who are married or in a common-law relationship. The sample excludes
immigrants, aboriginal people, and those with self-employment farm
income. All models control work experience, part-time work, weeks
worked, public sector employment, marriage, presence of children,
visible minority status, education, birth cohort, province of
residence, residence in a rural area, and survey year. Sample sizes
in italics under estimates. All estimates weighted using individual
population weights.
Table 5
Determinants of Self-Employment, Odds Ratios from Logistic
Regression, Census of Canada 2001 to 2011
Men
1
B SE
Same-sex partner 0.772 ** .067
Occupations (ref: social science, education,
government)
Management occupations 2.117 *** .024
x Same sex 0.965 .101
Business, finance, and administration 1.118 *** .015
x Same sex 0.738 * .090
Natural and applied sciences 1.031 * .013
x Same sex 1.670 *** .210
Health-related occupations 4.393 *** .065
x Same sex 0.728 ** .086
Art, culture, recreation, and sport 5.158 *** .081
x Same sex 1.442 *** .157
Sales and service 0.861 *** .011
x Same sex 2.029 *** .211
Manufacturing, trades, and primary industry 1.280 *** .015
x Same sex 1.226 .157
Constant 0.094 *** .002
N 2,246,695
Men
2
B SE
Same-sex partner 0.967 .105
Occupations (ref: social science, education,
government)
Management occupations 4.365 *** .069
x Same sex 0.778 * .096
Business, finance, and administration 2.293 *** .040
x Same sex 0.588 *** .081
Natural and applied sciences 2.128 *** .036
x Same sex 1.347 * .191
Health-related occupations 3.988 *** .089
x Same sex 0.745 .116
Art, culture, recreation, and sport 10.590 *** .203
x Same sex 1.162 .147
Sales and service 1.717 *** .030
x Same sex 1.607 *** .197
Manufacturing, trades, and primary industry 2.507 *** .041
x Same sex 0.979 .141
Constant 0.043 *** .001
N 2,211,995
Women
1
B SE
Same-sex partner 0.916 .072
Occupations (ref: social science, education,
government)
Management occupations 1.855 *** .022
x Same sex 0.990 .115
Business, finance, and administration 0.711 *** .008
x Same sex 1.015 .119
Natural and applied sciences 0.915 *** .019
x Same sex 1.933 *** .300
Health-related occupations 0.742 *** .010
x Same sex 3.823 *** .422
Art, culture, recreation, and sport 5.257 *** .070
x Same sex 1.556 *** .184
Sales and service 1.165 *** .014
x Same sex 1.037 .116
Manufacturing, trades, and primary industry 1.520 *** .023
x Same sex 1.040 .132
Constant 0.081 *** .002
N 2,083,900
Women
2
B SE
Same-sex partner 0.799 * .072
Occupations (ref: social science, education,
government)
Management occupations 1.916 .023
x Same sex 1.169 .145
Business, finance, and administration 0.710 *** .009
x Same sex 1.190 .150
Natural and applied sciences 0.960 * .020
x Same sex 2.250 *** .366
Health-related occupations 0.503 *** .008
x Same sex 4.898 *** .627
Art, culture, recreation, and sport 5.530 *** .076
x Same sex 1.835 *** .233
Sales and service 1.153 *** .015
x Same sex 1.215 .147
Manufacturing, trades, and primary industry 1.502 *** .023
x Same sex 1.204 .162
Constant 0.067 *** .002
N 2,063,510
Notes: p [less than or equal to] .05; ** p [less than or equal to]
.01; *** p [less than or equal to] .001.
Sample includes employed individuals between the ages of 25 and 64
who are married or in a common-law relationship. The sample excludes
immigrants, aboriginal people, and those with self-employment farm
income. All models control province of residence, residence in a
rural area, work experience, common-law status, presence of children
in the house, visible minority status, education, birth cohort, and
survey year. Model 2 excludes professionals. All estimates weighted
using individual population weights.
Table 6
Determinants of Self-Employment, Odds Ratios from Logistic
Regression, Census of Canada 2001 to 2011
Heterosexual men Gay men
B SE B SE
Work experience
(ref: 0-9 years)
10-19 years 1.400 *** .019 1.552 ** .213
20-29 years 1.556 *** .027 1.721 ** .309
30-39 years 1.554 *** .034 1.713 * .396
40+ years 1.647 *** .042 1.712 .500
Demographic
characteristics
Common-law 0.967 *** .006 0.884 .072
(ref: married)
Have children 1.046 *** .006 1.248 .170
(ref: no
children)
Visible minority 0.995 .019 1.247 .253
(ref: White)
Education (ref: HS
degree)
Less than HS 1.087 *** .009 1.155 .148
degree
College/trade 0.948 *** .006 1.425 *** .115
degree
BA+ 1.217 *** .010 1.516 *** .134
Birth cohort (ref:
before 1950)
1950-1954 0.762 *** .007 0.741 * .087
1955-1959 0.701 *** .008 0.648 *** .083
1960-1964 0.671 *** .009 0.574 *** .088
1965-1969 0.630 *** .010 0.546 *** .096
1970-1974 0.532 *** .010 0.456 *** .100
1975-1979 0.434 *** .010 0.385 *** .097
1980-1984 0.383 *** .011 0.439 ** .136
1985 or later 0.328 *** .023 0.355 * .180
Occupations (ref:
social science,
education,
government)
Management 2.117 *** .024 2.042 *** .214
occupations
Business, 1.118 *** .015 0.839 .106
finance, and
administration
Natural and 1.031 * .013 1.720 *** .218
applied
sciences
Health and 4.394 *** .065 3.155 *** .382
related
occupations
Art, culture, 5.159 *** .081 7.443 *** .808
recreation, and
sport
Sales and service 0.861 *** .011 1.772 *** .194
Manufacturing, 1.280 *** .015 1.645 *** .224
trades, and
primary
industry
Constant 0.094 *** .002 0.070 *** .019
N 22,27,745 18,950
Heterosexual women Lesbian women
B SE B SE
Work experience
(ref: 0-9 years)
10-19 years 1.385 *** .021 1.070 .173
20-29 years 1.230 *** .026 0.745 .157
30-39 years 1.188 *** .031 0.774 .210
40+ years 1.442 *** .047 0.483 * .176
Demographic
characteristics
Common-law 0.792 *** .006 1.143 .105
(ref: married)
Have children 1.099 *** .007 1.140 .093
(ref: no
children)
Visible minority 0.915 *** .022 1.018 .230
(ref: White)
Education (ref: HS
degree)
Less than HS 0.968 ** .011 1.037 .170
degree
College/trade 1.137 *** .009 1.094 .109
degree
BA+ 1.114 *** .012 1.332 * .150
Birth cohort (ref:
before 1950)
1950-1954 0.805 *** .010 0.801 .106
1955-1959 0.732 *** .011 0.773 .120
1960-1964 0.724 *** .013 0.530 *** .096
1965-1969 0.688 *** .014 0.503 ** .108
1970-1974 0.610 *** .015 0.291 *** .073
1975-1979 0.533 *** .015 0.200 *** .061
1980-1984 0.486 *** .018 0.178 *** .066
1985 or later 0.401 *** .025 0.192 * .123
Occupations (ref:
social science,
education,
government)
Management 1.854 *** .022 2.019 *** .233
occupations
Business, 0.710 *** .008 0.866 .110
finance, and
administration
Natural and 0.915 *** .019 2.034 *** .319
applied
sciences
Health and 0.741 *** .010 3.187 *** .370
related
occupations
Art, culture, 5.254 *** .070 9.046 *** .108
recreation, and
sport
Sales and service 1.163 *** .014 1.514 *** .185
Manufacturing, 1.517 *** .023 2.035 *** .287
trades, and
primary
industry
Constant 0.080 *** .002 0.090 *** .029
N 20,66,650 17,250
Notes: * p [less than or equal to] .05; ** p [less than or equal to]
.01; *** p [less than or equal to] .001.
Sample includes employed individuals between the ages of 25 and 64
who are married or in a common-law relationship. The sample excludes
immigrants, aboriginal people, and those with self-employment farm
income. All models control province of residence, residence in a
rural area, and survey year. All estimates weighted using individual
population weights.
