Sexual Orientation and Occupational Attainment

Work and Occupations40(1) 3 –36

© The Author(s) 2013Reprints and permission:

sagepub.com/journalsPermissions.navDOI: 10.1177/0730888412460532

wox.sagepub.com

460532WOX40110.1177/0730888412460532Work and OccupationsUeno et al.© The Author(s) 2013

Reprints and permission:sagepub.com/journalsPermissions.nav

1Florida State University, Tallahassee, FL, USA

Corresponding Author:Koji Ueno, Department of Sociology, Florida State University, 526 Bellamy Building, Tallahassee, FL 32306, USA Email: [email protected].

Sexual Orientation and Occupational Attainment

Koji Ueno1, Abráham E. Peña-Talamantes1, and Teresa A. Roach1

Abstract

Research shows that sexual minorities and heterosexuals differ in the level of success in educational and earnings attainment, but differences in occu-pational attainment have been unclear. To extend the literature, this article examines sexual orientation differences in young adulthood occupational status by analyzing the National Longitudinal Study of Adolescent Health Waves 1 to 4. Drawing from the life course literature, a theoretical frame-work is developed to explain how sexual orientation shifts opportunities and constraints in the occupational attainment process. The analysis shows that the association between sexual orientation and occupational status depends on gender, the type of sexual experience used to measure sexual orientation, and the timing of first sexual experience. Women who report their first same-sex attraction or sexual contact in young adulthood have lower occupational status than those without such experience, and men who report their first same-sex dating relationships in young adulthood tend to have higher occupational status than those without such relationships. These findings extend the existing knowledge on sexual development and status attainment, which has mainly focused on heterosexual development.

Keywords

occupational prestige, labor market outcomes, sexual orientation, worker characteristics, homophobia, work attitudes, youth

4 Work and Occupations 40(1)

Status attainment is a process in which people obtain positions in the status hierarchy, and the level of success is often measured by educational qualifi-cations, occupational status, and earnings (Blau & Duncan, 1967; Sewell & Hauser, 1980). Sexual minorities and heterosexuals differ in the level of suc-cess in status attainment, but the direction of the difference depends on attain-ment outcomes and gender. In educational qualifications, sexual minorities surpass heterosexuals among both women and men (Laumann, Gagnon, Michael, & Michaels, 1994; Turner, Villarroel, Chromy, Eggleston, & Rogers, 2005). In earnings, sexual minority women exceed heterosexual women, but sexual minority men lag behind heterosexual men (Berg & Lien, 2002; Black, Makar, Sanders, & Taylor, 2003).1 Little is known about sexual orientation differences in occupational status, however. Occupational status refers to the standard of living based on the person’s occupation, and is commonly mea-sured by educational qualifications of other people who hold the same occu-pation in the population (i.e., occupational education) and their earnings (i.e., occupational earnings; Hauser & Warren, 1997; Nakao & Treas, 1994; Nam & Boyd, 2004).

Although the literature provides little information about occupational sta-tus differences, sexual minorities’ and heterosexuals’ uneven representations across major occupational categories have been extensively studied. According to the General Social Survey, for example, sexual minority women are overrepresented in craft, operative, and service categories and underrep-resented in managerial, administrative support, and sales categories, com-pared to heterosexual women (Badgett & King, 1997; Blandford, 2003). Sexual minority men are overrepresented in professional, technical, adminis-trative support, and sales categories and underrepresented in managerial, craft, and operative categories. Studies based on census data show similar results (Antecol, Jong, & Steinberger, 2008). Scholars have attributed these differences to several factors including varying levels of discrimination against sexual minorities across occupations (Badgett & King, 1997), their occupational choice based on perceived “gay friendliness” (Berger, 1982; Harry & DeVall, 1978; Hewitt, 1995), and their concentrations in gender-atypical occupations (Blandford, 2003; Whitam, 1983).

Sexual minorities’ and heterosexuals’ uneven representations across major occupational categories hint that the two groups may differ in occupa-tional status. However, the information at the major category level is not necessarily useful for addressing differences at the occupational title level for two reasons. First, some patterns in major categories lead to conflicting expectations for occupational status differences. For example, sexual minority women are underrepresented in major categories that include many

Ueno et al. 5

higher-status occupations (i.e., professional, managerial, and technical cate-gories) compared to heterosexual women, but they are also underrepresented in major categories that include many lower-status occupations (i.e., admin-istrative support and sales categories). Second, within each major occupa-tional category, sexual minorities are concentrated in certain occupations (e.g., gay men in art-related occupations within the professional category; Blandford, 2003). Occupational clustering within major categories is not unique to sexual orientation, but it has been observed for other sociodemo-graphic dimensions such as gender (e.g., Cohen, Huffman, & Knauer, 2009; Ku, 2011). Because occupational status varies considerably across occupa-tions within each major category, these categories are too crude to address sexual orientation differences in occupational status.

A Census 2000 study by Baumle, Compton, and Poston (2009) was the first study to report national estimates of sexual orientation differences in occupational status at the occupational title level. Using resident partner’s gender as a sexual orientation indicator, the study showed that women cohab-itating with other women had higher occupational status than women married to or cohabitating with men. Men cohabitating with other men had lower occupational status than heterosexually married men but higher status than those cohabitating with women. This study had two major limitations. First, the study focused on estimating the extent of the group differences and did not provide much information about what accounted for the differences. Second, the study simultaneously analyzed people in a wide age range and therefore could not address whether the differences already existed in early career stages.

The present study examines sexual orientation differences in occupational status using the National Longitudinal Study of Adolescent Health (Add Health) Waves 1 to 4. Drawing from the life course literature, we conceptual-ize young adult occupation as a status attainment outcome and investigate how sexual orientation creates opportunities and constraints in the occupa-tional attainment process. Important disparities in occupational status emerge across social groups in this early career stage (Miech, Eaton, & Liang, 2003; Warren, Hauser, & Sheridan, 2002), but little is known about whether and how disparities emerge across sexuality groups. Unlike gender and race, sex-ual orientation is a dynamic individual attribute—people become aware of their sexual orientation as they grow up, and the timing of first same-sex experience varies across individuals (Savin-Williams & Ream, 2007). To address this dynamic aspect of sexual orientation, the present study investi-gates whether the timing of first same-sex experience is associated with occu-pational status and what accounts for the association. Although the analysis


is limited to supply-side factors such as worker qualifications and motiva-tions due to a lack of measures for demand-side factors in Add Health, we seek to move the literature forward by increasing knowledge about how sex-ual orientation differences emerge in the occupational attainment process.

Sexual Development and Status AttainmentConceptualizing first vaginal intercourse as the major transition in hetero-sexual development, previous research has demonstrated that people who go through the transition at early ages tend to be less successful in status attain-ment than other people (Frisco, 2008; Spriggs & Halpern, 2008; Steward, Farkas, & Bingenheimer, 2009). Scholars have proposed two explanations by drawing from the life course literature. First, life events that occur at a nonnormative age tend to create negative consequences due to a lack of suf-ficient interpersonal and cognitive skills to deal with the events (Newcomb & Bentler, 1988; Petersen & Taylor, 1980). Specifically, those who have first sexual intercourse at a young age may not be mature enough to handle the responsibilities and emotional intensity in sexual relations (Meier, 2007). Women suffer greater consequences of early timing than men partly because the society attaches more negative meaning to women’s early sexual devel-opment (Spriggs & Halpern, 2008; Tolman, 2006). Events that take place before the normative age could also undermine status attainment by creating unintended consequences (Elder, Johnson, & Crosnoe, 2003). Pregnancy and sexually transmitted disease infections are examples of such unintended consequences of first intercourse at young ages (Frisco, 2008).

Second, the association between early timing of first heterosexual inter-course and unsuccessful status attainment may result from a shared antecedent—a lack of social control in adolescence. Hirschi (1969) demonstrated that detachment from the conventional society (e.g., family, school) reduces restraints to delinquency by lowering the perceived costs of those behaviors. Research shows that a lack of social control also contributes to early transi-tion into first heterosexual intercourse and risky heterosexual behaviors in adolescence (Buhi & Goodson, 2007; Longmore, Eng, Giordano, & Manning, 2009). At the same time, a lack of social control in adolescence has important life course implications—it limits the success in status attainment by distract-ing youth from school work and reducing the ability to obtain educational qualifications (Joyner & Udry, 2000; Sewell & Hauser, 1980; Siennick & Staff, 2008).

The extent to which these processes apply to same-sex sexual develop-ment is unclear. Sexual relations may require maturity and may be subject to

Ueno et al. 7

social control regardless of the partner’s gender. For this reason, early timing of same-sex experience may be associated with unsuccessful status attain-ment. Unlike heterosexual development, however, age norms are not clearly defined for same-sex sexual development, which may blur the impact of the timing of first same-sex experience. Furthermore, stigma linked to same-sex orientation undermines educational preparation and development of psycho-logical resources in early life stages (Pearson, Muller, & Wilkinson, 2007; Ueno, Roach, & Peña-Talamantes, in press). Therefore, the timing of first same-sex experience may shift the timing of initial exposure to sexuality stigma as well as the accumulated level of exposure. The timing of first same-sex experience may also impact when sexual minorities start considering the possibility of facing work discrimination, which may impact educational and occupational plans.

Sexual orientation is a multifaceted concept measured by attraction, rela-tionships, contact, and identity, and these measures correlate only moderately (Laumann et al., 1994). As explained below, the extent of disparities in occu-pational status may depend on which dimension of sexuality is measured. In this article, we use the phrase “same-sex experience” to inclusively refer to sexual events and engagement with same-sex individuals assessed by these measures, and we use the phrase “sexual minorities” to refer to people who report the type of same-sex experience in question.

To address individual variations in same-sex sexual development, we use a classification scheme consisting of three categories, depending on the pat-tern of same-sex experience reported in adolescence and young adulthood:

1. No same-sex experience.2. Early development (same-sex experience reported in both adoles-

cence and young adulthood).3. Late development (same-sex experience reported only in young

adulthood). 2

In the discussions below, we formulate expectations about whether these three groups differ in occupational status by drawing from the life course literature.

Early Development of Same-Sex ExperiencePeople who report their first same-sex experience in adolescence may have lower occupational status in young adulthood than other people because their early exposure to sexual stigma may limit their success in occupational


attainment. Heteronormativity in the family and school contexts creates many challenges for sexual minority youth including conflicts with parents, harass-ment and isolation from peers, and academic and mental health problems (Barrett, Pollack, & Tilden, 2002; Pearson et al., 2007). Disadvantages in early life stages tend to persist and in many cases accumulate over the life course because they take away resources and opportunities necessary to move out of the negative trajectories (Elman & O’Rand, 2004; Pearlin, Schieman, Fazio, & Meersman, 2005). Following this argument, the chal-lenges resulting from sexuality stigma in adolescence may lower occupa-tional status by limiting the ability to obtain sufficient educational qualifications. Research shows that sexual minority boys suffer more serious stigma than sexual minority girls (Horn, 2007; Pascoe, 2007), suggesting that they might receive a greater penalty in occupational status in young adulthood.

For the people in this group, work discrimination in young adulthood may further limit the ability to obtain and sustain high-status occupations and thus increase their gap from the No Experience group. A recent audit experiment study illustrates that demand-side factors strongly influence the risk of sexual orientation discrimination—employers tend to be more discriminatory in male-typical occupations and in places with more conservative political cli-mate and no antidiscrimination laws (Tilcsik, 2011). Although the literature on sexual orientation discrimination remains sparse, many studies have been conducted to examine racial and gender discrimination. These studies identi-fied several contributing factors on the demand side, including recruitment and termination practices (Byron, 2010; Kmec, 2005), job-sorting practices (Williams & Connell, 2010), and spatial segregation across work groups (Vallas, 2003). Discrimination not only reduces one’s career chances in the hiring, evaluation, promotion, and firing processes, it also affects worker sat-isfaction and mental health (Sloan, 2011), which may shift their ability and motivation for career development. Although the implications of these demand-side characteristics for sexual orientation discrimination have been unexamined, a study suggests that workplace characteristics such as size and gender composition strongly affect sexual minority women’s work experience (Schneider, 1986). Among the several measures of sexual orientation, dating relationships are generally more stable and visible to others, and engagement in same-sex relationships may have particularly strong impact on the chance of being labeled as sexual minorities and facing discrimination at work. For this reason, dating relationships may show the Early Development group’s disadvantage in occupational status more clearly than other measures.

Sexual minorities’ limited access to marriage and reduced chance of parenthood may also contribute to their occupational status difference from

Ueno et al. 9

heterosexuals because workers’ marital and parental statuses have significant impact on how they are evaluated by employers. Earnings studies show that parenthood provides a premium for men and a penalty for women (Budig & England, 2001; Glauber, 2007). For men, marriage provides an additional premium (Cohen, 2002). Although these contrasting consequences of family formation for men and women may partly reflect changes in their work com-mitment (Greenman, 2010), employer bias plays an important role. For example, employers perceive fathers as more committed and mothers as less committed even when they have qualifications similar to other men and women (Correll, Benard, & Paik, 2007). Given sexual minorities’ limited access to marriage and their lower chance of parenthood (Baumle et al., 2009; Black et al., 2003), these employer biases may reduce occupational status of men in the Early and Late Development groups and increase occupational status of women in these groups, compared to the No Experience group. Furthermore, a recent earnings study shows that unlike heterosexual women, who receive a motherhood penalty, sexual minority women receive an advan-tage, perhaps because motherhood affirms their femininity while still being perceived as committed to work (Baumle, 2009). A similar process may operate for occupational status.

Late Development of Same-Sex ExperienceThe literature leads to conflicting expectations regarding occupational impli-cations of same-sex experience that emerges in adulthood. As mentioned above, delayed sexual development may indicate high levels of social control in adolescence, which facilitates the status attainment process. If this process applies to same-sex sexual development, the Late Development group should show higher occupational status than the other two groups. This group difference may be observed more strongly in analyses based on behav-ioral measures of sexual orientation (dating relationships and sexual contact) as opposed to attraction because these behaviors should more strongly reflect the level of social control and shift the level of distraction from educational preparation.

Contrary to these arguments, research on sexuality stigma suggests that awareness of same-sex sexuality emerging in young adulthood may be asso-ciated with lower occupational status because it marks the entry into a stig-matized role as a sexual minority during the critical period for occupational attainment. The transition may have negative consequences such as identity confusion, self-denial, and fear of rejection by family and friends (Cass, 1979; Troiden, 1989). As a result, people reporting emerging same-sex experience


tend to show greater deteriorations in mental health than other people (Ueno, 2010). These challenges may in turn undermine the ability to obtain high-status occupations by draining psychological resources and reducing the chance to earn educational qualifications. Consistent with this argument, both qualitative and quantitative research suggests that awareness of same-sex sexuality delays career development (Boatwright, Gilbert, Forrest, & Ketzenberger, 1996; Schmidt & Nilsson, 2006).

The negative impact of same-sex experience that emerges in young adult-hood may depend on gender, although it is unclear whether women or men should show a greater penalty. Women exhibit more negative mental health responses to emerging same-sex experience (Ueno, 2010), perhaps due to their early socialization that emphasizes emotional investment into roman-tic relationships (Gilligan, 1996) and sensitivity to interpersonal events (Rudolph, 2002), which exacerbate their emotional and network responses to the emerging awareness of same-sex orientation. At the same time, men are subject to more serious sexuality stigma (Herek, 2002; Horn, 2007), which leads to an opposite expectation that men should face a greater penalty in occupational status. Furthermore, the negative impact of first same-sex expe-rience in young adulthood may depend on types of same-sex experience. Specifically, the negative impact may be more clearly observed for attraction than dating relationships and sexual contact because many sexual minorities first become aware of their same-sex attraction before engaging in sexual behaviors (Cass, 1979; Troiden, 1989). Therefore, attraction may have a stronger ability to detect psychological and occupational consequences of the initial awareness of same-sex orientation.

SummaryThe purpose of this study is to examine sexual orientation differences in occupational status. Drawing from the literatures on life course and hetero-sexual development, we propose that timing of first same-sex experience has important implications for occupational status. People who become aware of same-sex sexuality in adolescence may have lower occupational status because interpersonal, academic, and mental health problems linked to sexu-ality stigma in adolescence reduce their ability to earn educational qualifica-tions. Work discrimination in adulthood may further lower their occupational status. The literature leads to conflicting expectations for occupational status of people who report late timing of their first same-sex experience. Emerging awareness of same-sex orientation may operate as a negative turning point and undermine the ability to obtain high-status occupations, but the late timing may instead indicate a high level of social control in adolescence, which

Ueno et al. 11

generally facilitates the status attainment process. Previous research suggests that the status disparities by sexual orientation may depend on gender, although it is unclear whether women or men should show greater dispari-ties. Furthermore, the magnitude of disparities may vary across different measures of sexual orientation because each measure may be uniquely linked to the underlying mechanisms of occupational status disparities.

MethodData and Sample

Add Health is an ongoing survey study that assesses health and attainment outcomes and their social antecedents (Harris et al., 2009). In 1994, 80 high schools and 52 middle schools in the United States were selected for the school-wide survey (In-School Survey). Among the initial respondents, 20,745 students participated in in-depth interviews at their home in 1995 (Wave 1 in-home interviews). Follow-up interviews were conducted in 1996 (Wave 2), between 2001 and 2002 (Wave 3), and between 2008 and 2009 (Wave 4). Most respondents were between 13 and 18 years of age in Wave 1, between 14 and 19 years in Wave 2, between 20 and 25 years in Wave 3, and between 26 and 31 years in Wave 4. In Wave 1, 18,924 respondents constituted the main component of the sample designed to produce national estimates. Of these people, 12,288 respondents also participated in at least Waves 3 and 4 interviews. From these respondents, we excluded 1.7% of people who reported in Wave 4 that they had not worked since 2001. We also excluded those who had only worked for military jobs (n=25) because their occupational status could not be determined in the scale used for this study. The final operational sample included 6,560 women and 5,513 men.

MeasuresSexual orientation. The analysis used three measures of sexual orientation—

attraction, dating relationships, and sexual contact. Table 1 describes how each measure was used to identify three patterns between adolescence and young adulthood: (1) no same-sex experience, (2) early development (same-sex experience reported in both adolescence and young adulthood), and (3) late development (same-sex experience reported only in young adult-hood). As mentioned earlier, a small number of people reporting same-sex experience only in adolescence were combined to the no same-sex experi-ence group (see Endnote 2). The three measures of sexual orientation were

12

Tab

le 1

. Sam

e-Se

x Ex

peri

ence

Mea

sure

s

Cod

ing

for

Mea

sure

sN

o ex

peri

ence

aEa

rly

deve

lopm

ent

(firs

t ex

peri

ence

in a

dole

scen

ce)

Late

dev

elop

men

t (fi

rst

expe

rien

ce in

yo

ung

adul

thoo

d)

Att

ract

ionb

W1:

“H

ave

you

ever

had

a r

oman

tic a

ttra

ctio

n to

a fe

mal

e/m

ale?

”W

2: “

Sinc

e (t

he d

ate

of t

he la

st in

terv

iew

), ha

ve y

ou e

ver

had

a ro

man

tic a

ttra

ctio

n to

a fe

mal

e/m

ale?

”W

4: “

Are

you

rom

antic

ally

att

ract

ed t

o fe

mal

es/m

ales

?”

No

sam

e-se

x at

trac

tion

repo

rted

in W

1,

W2,

or W

4.

Sam

e-se

x at

trac

tion

repo

rted

in

W1

or W

2 an

d ag

ain

in W

4.Sa

me-

sex

attr

actio

n re

port

ed in

W4

only.

Dat

ing

rela

tions

hips

W1,

W2:

Lis

ted

up t

o th

ree

rom

antic

rel

atio

nshi

ps a

nd t

hree

se

xual

rel

atio

nshi

ps in

last

18

mon

ths.

W3:

Lis

ted

rom

antic

and

sex

ual r

elat

ions

hips

sin

ce t

he s

umm

er

of 1

995.

W4:

Lis

ted

part

ners

tha

t re

spon

dent

s ha

d m

arri

ed, l

ived

with

, ha

d a

preg

nanc

y w

ith/im

preg

nate

d, o

r w

ere

curr

ently

invo

lved

w

ith.

No

sam

e-se

x da

ting

rela

tions

hips

re

port

ed in

W1

thro

ugh

W4.

Sam

e-se

x re

latio

nshi

ps

repo

rted

in W

1 or

W2,

or

invo

lvem

ent

in s

ame-

sex

rela

tions

hips

bef

ore

the

18th

bi

rthd

ay b

ased

on

W3

and

W4

data

. In

addi

tion,

rep

ortin

g in

volv

emen

t on

or

afte

r th

e 18

th b

irth

day.

Invo

lvem

ent

in s

ame-

sex

rela

tions

hips

on

ly o

n or

aft

er

the

18th

bir

thda

y ba

sed

on W

3 or

W

4 da

ta.

In e

ach

wav

e, p

artn

er g

ende

r, st

artin

g da

te, a

nd e

ndin

g da

te (

if en

ded)

of e

ach

rela

tions

hip

wer

e re

cord

ed.

Sexu

al c

onta

cts

W4:

“C

onsi

deri

ng a

ll ty

pes

of s

exua

l act

ivity

, with

how

man

y m

ale/

fem

ale

part

ners

hav

e yo

u ev

er h

ad s

ex?”

“. .

. with

how

m

any

mal

e/fe

mal

e pa

rtne

rs d

id y

ou h

ave

sex

befo

re y

ou w

ere

18 y

ears

old

, eve

n if

only

one

tim

e?”

No

sam

e-se

x co

ntac

ts r

epor

ted

in W

4.

The

num

ber

of s

ame-

sex

cont

acts

bef

ore

age

18 w

as

grea

ter

than

0, a

nd t

his

num

ber

was

exc

eede

d by

the

lif

etim

e nu

mbe

r of

sam

e-se

x co

ntac

ts.

No

sam

e-se

x co

ntac

ts b

efor

e ag

e 18

and

one

or

mor

e sa

me-

sex

cont

acts

in li

fetim

e.

Not

e. W

1 =

Wav

e 1;

W2

= W

ave

2; W

3 =

Wav

e 3;

W4

= W

ave

4. S

exua

l exp

erie

nces

bef

ore

the

18th

bir

thda

y w

ere

cons

ider

ed a

s ad

oles

cent

ex

peri

ence

s, a

nd t

hose

on

or a

fter

the

18t

h bi

rthd

ay w

ere

cons

ider

ed a

s yo

ung

adul

thoo

d ex

peri

ence

s. M

ost

resp

onde

nts

wer

e ag

e be

twee

n 13

and

18

in

Wav

e 1,

bet

wee

n 14

and

18

in W

ave

2, b

etw

een

19 a

nd 2

5 in

Wav

e 3,

and

bet

wee

n 26

and

31

in W

ave

4.a In

add

ition

to

thos

e w

ho d

id n

ot r

epor

t an

y sa

me-

sex

expe

rien

ce, t

his

grou

p in

clud

ed a

sm

all n

umbe

r of

peo

ple

repo

rted

sam

e-se

x ex

peri

ence

onl

y in

ad

oles

cenc

e (s

ee m

ore

disc

ussi

ons

in E

ndno

te 2

in t

ext)

.b In

Wav

e 3,

res

pond

ents

rep

orte

d on

life

time

attr

actio

n in

stea

d of

att

ract

ion

sinc

e th

e la

st w

ave.

The

refo

re, t

he t

imin

g of

sam

e-se

x at

trac

tion

repo

rted

in

Wav

e 3

coul

d no

t be

det

erm

ined

. For

thi

s re

ason

, the

Wav

e 3

attr

actio

n m

easu

re w

as n

ot u

sed

for

this

stu

dy.

Ueno et al. 13

analyzed separately because they were expected to show different associa-tions with occupational status.

Occupational characteristics. In Wave 4 (year 2008/2009), respondents were asked about occupations in which they were currently working or had worked at least 10 hr a week for at least 9 weeks between 2001 and the year before the interview. Interviewers helped the respondents identify their occupations in the 2000 Standard Occupational Classification (SOC; US Department of Labor, Bureau of Labor Statistics, 2009). To construct the dependent variable for the present investigation, we converted SOC codes into occupational sta-tus scores using the Nam–Power–Boyd Occupational Status Scale (Nam & Boyd, 2004). The scale is constructed from educational and earnings levels of occupation incumbents, and the score can be interpreted as the percentage of people in the labor force who have the combined level of median education and earnings lower than the occupation.3 Part-time employment was included in the analysis because it is more common in low-status occupations (Magnusson, 2009) and varies by sexual orientation (Baumle et al., 2009). The variable was dichotomously coded to distinguish between full-time employment (0 = 35+ hr) and part-time employment (1 = less than 35 hr). Percentage of women in the occupation was also entered in the analysis because sexual minorities are more likely to hold occupations in gender-atyp-ical occupations (Blandford, 2003) and because gender composition is tied to occupational reward (Cohen et al., 2009; England, 1979; Glauber, 2012). Employment status and gender composition helped examine whether sexual orientation differences in occupational status persisted after considering dif-ferences in these employment and occupational characteristics. Descriptive statistics of the key variables are presented in Table 2.

Predictors in adolescence. The analysis included the following variables from adolescent waves to examine whether they helped explain the associa-tion between sexual orientation and occupational status. Depressive symp-toms measured emotional manifestations of stress possibly resulting from sexuality stigma. The measure used a modified version of the Center for Epidemiologic Studies Depression Scale (CES-D; Radloff, 1977), which summed 19 four-point items (e.g., “you were sad”; α = 0.85). Grade point average measured academic performance in high school, and it was based on official transcripts for respondents who participated in the Adolescent Health and Academic Achievement study as part of Add Health Wave 3 (Riegle-Crumb, Muller, Frank, & Schiller, 2005). For other respondents, GPA was computed by taking the average across self-reported grades in English, math, sciences, and history or social science in Waves 1 and 2. Educational expec-tations summed two five-point items in Wave 1 (“How much do you want to go to college?” and “How likely is it that you will go to college?”; r = 70).


Educational expectations are linked to occupational expectations and attain-ment outcomes (Sewell & Hauser, 1980), although these links seem to have weakened in recent years (Reynolds, Stewart, MacDonald, & Sischo, 2006).

Predictors in young adulthood. As in adolescence, depressive symptoms were measured to assess emotional manifestations of stress. The measure in young adulthood was based on a nine-item version of CES-D scale included in Wave 3 (α = 0.79). Highest educational qualification in Wave 4 was used as a measure of educational attainment. The ordinal variable included the fol-lowing seven categories: (1) less than high school, (2) GED, (3) high school diploma, (4) post-secondary vocational or technical certificate or license, (5) associate degree, (6) bachelor’s degree, and (7) graduate or professional degree. To examine whether family formation accounted for the sexuality group differences in occupational status, the analysis also entered marital status in Wave 4 (0 = unmarried; 1 = married) and parental status in Wave 4 (0 = nonparent; 1 = parent).

Sociodemographic background. The analysis controlled for the following sociodemographic variables due to their correlations with same-sex experi-ence (Laumann et al., 2004; Ueno, 2010) and with status attainment (Kalleberg, 2011; Saltiel, 1988; Warren et al., 2002). Age in Wave 4 was measured in years. Race consisted of four dummy variables including non-Hispanic White, non-Hispanic Black, Hispanic, and others. Parental education was a set of four dummy variables, including less than high school, high school graduate, some college, and college graduate, based on mother’s or father’s educational level, whichever was higher.

Analytical ProcedureThe association between same-sex experience and occupational status was examined in ordinary least square (OLS) regression models. Each model included dummy variables for the Early and Late Development groups, and their coefficients indicated whether these groups differed from the No Experience group (reference) in occupational status. The difference between the Early and Late Development groups was examined in F tests of coeffi-cient equality. The base model in each analysis controlled for sociodemo-graphic background. In the subsequent models, we sequentially added predictors in adolescence and those in young adulthood and observed which variables increased or decreased the dummy coefficients for the Early and Late Development groups. As mentioned earlier, the analyses were limited to people who had worked since 2001. To reduce bias resulting from the selection into the labor force, we applied Heckman (1979) correction to the OLS models.4 Our investigation focused on the association between

Ueno et al. 15

Table 2. Sample Characteristics by Gender

Women Men

Range Mean or % SD Mean or % SD

Occupational characteristicsOccupational status 1,100 54.06 26.77 53.25 25.24Working part time 0,1 19.71% 10.38% % Women in the occupation

1,98 62.94 24.00 31.47 23.88

Sexual orientationSame-sex attraction No SS attraction 90.56% 95.81% Early development 1.73% 1.51% Late development 7.71% 2.68% Same-sex dating relations No SS SS dating relations

95.46% 96.67%

Early development 1.04% 1.01% Late development 3.50% 2.32% Same-sex contact No SS contact 88.15% 93.98% Early development 2.08% 1.98% Late development 9.77% 3.30%

Sociodemographic backgroundAge at W4 24,34a 28.15 1.88 28.35 1.81Race Non-Hispanic White 0,1 69.10% 68.61% Non-Hispanic Black 0,1 15.21% 14.11% Hispanic 0,1 11.11% 11.73% Others 0,1 4.58% 5.56% Parent education Less than high school 0,1 11.23% 11.17% High school graduate 0,1 39.82% 38.62% Some college 0,1 14.45% 14.11% College graduate 0,1 34.50% 36.10%

Predictors in adolescenceDepressive symptoms 0,54 11.78 8.37 10.03 6.44Educational expectations 2,10 8.79 1.96 8.37 2.05GPA 0,4 2.73 .83 2.45 .81

Predictors in young adulthoodDepressive symptoms 0,25 4.91 4.49 4.01 3.49Educational attainment 1,7 4.32 1.76 3.95 1.63Married at W4 0,1 46.80% 38.40% Parent by W4 0,1 53.17% 31.83%

Note. n = 6,560 women and 5,513 men; W4 = Wave 4.aMost respondents were between age 26 and 31 in Wave 4.


same-sex experience and occupational status within each gender, and there-fore all analyses were conducted separately for women and men. The analy-ses conducted in Stata 11.2 imputed missing data from other key variables, corrected standard errors for the nested data structure (people attending the same schools in adolescence), and weighted estimates for the oversampling of certain sociodemographic groups (Chantala, 2006).

ResultsThe first part of this section will present the results from base models and discuss how the association between same-sex experience and occupational status varied by gender and by the three sexuality measures. The second part will demonstrate what factors accounted for the association by presenting intermediate and full models.

Base ModelsTable 3 presents results from the base models. Each model used a different sexuality measure (attraction, dating relationships, or sexual contact) to define the three sexuality groups (No Experience, Early Development, and Late Development). Among women, the Early Development group did not differ significantly from the No Experience group in the analyses of attrac-tion and dating relationships, as shown in the insignificant dummy coeffi-cients for the Early Development group (−2.95 and −0.25, respectively). In the analysis of sexual contact, however, this group’s status score was lower than the No Experience group by 11.16 points (p < .01). Given the great degree of variability in occupational status scores (SD = 26.77), the group difference was modest in size. The Late Development group also showed lower occupational status than the No Experience group in the analysis of attraction (b = −5.91, p < .001) and sexual contact (b = −6.64, p < .001). The result was consistent with the argument that emerging same-sex experience marks an entry into a stigmatized role and operates as a negative turning point in the status attainment process. The analysis of dating relationships showed no difference between the Late Development group and the No Experience group, however.

Men showed fewer group differences than women. The Early Development group did not differ from the No Experience group regardless of the sexuality measures used. The Late Development group also did not differ from the No Experience group in the analyses of attraction and sexual contact, but the group showed a modest occupational status premium in the analysis of dating

Ueno et al. 17

relationships (b = 9.83, p < .001). The last finding was consistent with the argument that delayed timing of first same-sex experience indicates a high level of social control in adolescence, which promotes success in status attainment. For both women and men, the two groups reporting same-sex attraction (the Early and Late Development groups) did not significantly dif-fer from each other in occupational status regardless of the sexuality measure used, as shown in F tests at the bottom of the table.

In sum, the difference between the Early Development group and the No Experience group was significant only in the analysis of women’s sexual contact with the former group showing lower occupational status. Women and men showed contrasting results for the Late Development group: Women in this group received a status penalty compared to the No Experience group in the analysis of attraction and sexual contact, whereas men in this group showed a status premium in the analysis of dating relationships.5

Intermediate and Full ModelsTo identify factors that accounted for the sexual orientation differences in occupational status, predictors were sequentially added to the base models. Due to space limitations, we will present results only for women’s attraction and men’s dating relationships, which showed large group differences in the base models. Results from other intermediate and full models are available from the authors upon request.

Table 4 presents results for women’s attraction. Model 1 is the base model shown earlier. In this model, the Late Development group had lower occupa-tional status than the No Experience group by 5.91 points (p < .001), whereas the Early Development group did not differ from the No Experience group. When adolescent predictors were entered in Model 2, GPAs and educational expectations showed positive associations as expected. However, depressive symptoms did not show a significant association in this model because the variable was related to occupational status only indirectly through GPAs and educational expectations as revealed in a follow-up analysis not shown here. In Model 2, the coefficient for the Late Development group shrank to −4.10 (p < .05) because women in this group had lower GPAs and lower educa-tional expectations. We will return to this point in the discussion section.

When adulthood predictors were added in Model 3, depressive symptoms showed a negative association with occupational status, and educational attainment showed a positive association as expected. In this model, the coef-ficient for the Late Development group shrank substantially to 1.88 and lost significance because this group reported higher levels of depressive

18

Tab

le 3

. OLS

Reg

ress

ion

Mod

els

Pred

ictin

g O

ccup

atio

nal S

tatu

s by

Sam

e-Se

x Ex

peri

ence

(Ba

se M

odel

s)

Wom

enM

en

A

ttra

ctio

nD

atin

g r e

latio

nSe

xual

con

tact

Att

ract

ion

Dat

ing

rela

tion

Sexu

al c

onta

ct

Sam

e-se

x ex

peri

ence

(re

f = n

o SS

E)

Earl

y de

velo

pmen

t−

2.95

−0.

25−

11.1

6**

3.88

3.04

.67

La

te d

e vel

opm

ent

−5.

91**

*−

4.01

−6.

64**

*3.

669.

83**

−3.

40A

ge a

t W4

0.61

*0.

67**

0.63

*.8

6**

.86*

*.8

6**

Rac

e (r

ef =

Non

-His

pani

c W

hite

)

Non

-His

pani

c Bl

ack

−1.

74−

1.54

−1.

76−

7.25

***

−7.

13**

*−

7.20

***

H

ispa

nic

5.28

***

5.30

***

5.29

***

3.30

*3.

37**

3.41

**

Oth

ers

4.50

*4.

65*

4.40

*7.

08**

7.06

**7.

13**

Pare

nt e

duca

tion

(ref

= h

igh

scho

ol)

Le

ss t

han

high

sch

ool

−4.

56**

-4.7

6**

−4.

74**

−4.

48**

-4.5

9**

−4.

53**

So

me

colle

ge5.

74**

*5.

70**

*5.

80**

*7.

51**

*7.

47**

*7.

51**

*

Col

lege

gra

duat

e12

.47*

**12

.48*

**12

.49*

**12

.92*

**12

.87*

**12

.96*

**In

vers

e m

ills

ratio

−64

.83*

**−

64.9

6***

−65

.48*

**−

56.1

6***

−55

.91*

**−

55.6

0***

Con

stan

t34

.42*

**32

.47*

**34

.45*

**25

.57*

*25

.35*

*25

.55*

*R2

0.10

0.10

0.11

0.10

0.10

0.10

F(1,

128)

for

b(SS

E ea

rly)

= b

(SSE

late

)0.

630.

591.

820.

001.

830.

62

Not

e. n

= 6

,560

wom

en a

nd 5

,513

men

; W4

= W

ave

4.*p

< .0

5. *

*p <

.01.

***

p <

.001

.

Ueno et al. 19

symptoms and lower levels of educational attainment. When family status variables were added in Model 4, marital status showed a positive associa-tion, and parental status showed a negative association. The result for paren-tal status was consistent with the argument that employers discriminate against mothers (Correll et al., 2007; England, 2005) and with the argument that mothers focus on home production rather than market production (Becker, 1991). However, the differences across the sexuality groups did not change much in this model, contrary to previous studies that attributed sexual orientation differences in attainment outcomes to family formation (e.g., Black et al., 2003). Model 5 added part-time employment status and percent-age of women in the occupation because previous studies showed that these variables are correlated with sexual orientation (Baumle et al., 2009; Blandford, 2003) and occupational reward (England, 1979; Magnusson, 2009). These variables showed negative associations with occupational sta-tus as expected. More importantly, the coefficient for the Late Development group did not change much and remained significant, indicating that the group difference in occupational status existed net of these occupational and employment characteristics. In short, the Late Development group’s disad-vantage in occupational status relative to the No Experience group was partly explained by the group’s higher levels of depressive symptoms and lower levels of educational attainment in young adulthood, consistent with the argument that the heteronormative society imposes hardships on people who become aware of their same-sex orientation, which operates as a nega-tive turning point in their status attainment process.

Table 5 presents the results for men’s dating relationships. Model 1, the base model shown earlier, demonstrated that the Late Development group exceeded the No Experience group in occupational status by 9.83 points (p < .001) whereas the Early Development group did not differ from the No Experience group. When adolescent predictors were added in Model 2, the coefficient for Late Development group shrank to 7.24 (p < .05) because the group reported higher GPAs and educational expectations than the No Experience group. When adulthood predictors were added in Model 3, the coefficient for the Late Development group was further reduced to 5.39 and no longer significant due to the group’s higher levels of educational attainment.

Model 4 introduced family status variables and showed a positive coeffi-cient for marriage, consistent with previous earnings research documenting a marriage premium for men (Cohen, 2002; Correll et al., 2007). As expected, the coefficient for the Late Development increased slightly to 6.47 (p < .05) in this model because these men were less likely to get married.


Table 4. OLS Regression Models Predicting Occupational Status by Same-Sex Attraction Among Women

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

Same-sex attraction (ref = no SSA) Early development −2.95 −1.03 −0.88 −0.68 −1.41 Late development −5.91*** −4.10* −1.88 −1.63 −2.29Adolescent predictors

Depressive symptoms −0.09 0.01 0.01 0.01Educational expectations 1.19*** 0.58* 0.57* 0.50*GPA 8.97*** 3.68*** 3.41*** 3.15***

Young adult predictorsDepressive symptoms −0.31** −0.30** −0.21*Educational attainment 5.44*** 5.31*** 5.31***Married at W4 3.14*** 3.51***Parent by W4 −2.07* −0.27Working part time −11.69***Percent women in the occupation

−0.25***

Sociodemographic backgroundAge at W4 0.61* 0.91*** 0.55** 0.50* 0.31Race (ref = Non-Hispanic White) Non-Hispanic Black −1.74 1.49 0.72 1.65 1.04 Hispanic 5.28*** 7.31*** 6.84*** 6.91*** 6.49*** Others 4.50* 2.80 2.28 2.48 2.08Parent education (ref = high school) Less than high school −4.56** −2.76 −1.79 −1.90 −2.44 Some college 5.74*** 3.28* 1.53 1.43 1.23 College graduate 12.47*** 6.94*** 3.32*** 3.18*** 3.07***

Inverse mills ratio −64.83*** −48.67*** −36.68*** −36.98*** −32.76***Constant 34.42*** −7.12 0.50 2.82 26.54***R2 0.10 0.19 0.27 0.27 0.37F(1,128) for b(SSA early) = b(SSA late)

0.63 0.69 0.08 0.08 0.08

Note. n = 6,560; W4 = Wave 4.*p < .05. **p < .01.***p < .001.

This suppression effect suggested that the group’s advantage in occupational status would have been even greater if they had been as likely as the No Experience group to get married. When employment status and percentage of women in the occupation were added in Model 5, the coefficient for the Late Development group slightly increased again to 6.64 (p < .05). The change reflected the group’s higher tendency to hold female-typical occupations (Blandford, 2003) and their greater likelihood to work part time (Baumle

Ueno et al. 21

et al., 2009), which suppressed their occupational status advantage to some extent. In sum, the analysis of men’s dating relationships showed that the Late Development group’s higher status relative to the No Experience group was partly explained by their higher GPAs, educational expectations, and educational attainment, consistent with the argument that delayed sexual development indicates a high level of social control in adolescence, which contributes to success in status attainment by keeping people focused on school work.

DiscussionExtending previous research that documented sexual minorities’ and heterosexuals’ uneven representations in major occupational categories (e.g., Badgett & King, 1997; Blandford, 2003), this study examined sexual orientation differences in occupational status focusing on three issues. First, the study took the dynamic aspect of sexual orientation into account and demonstrated that attainment implications of same-sex experience depend on whether it emerges in adolescence or young adulthood. Second, taking advantages of an extensive list of behavioral and attitudinal mea-sures in Add Health, the study examined the ways in which sexual orienta-tion affects the occupational attainment process, and the results confirmed the major arguments developed from the life course framework. Third, the study employed three measures of sexual orientation and demonstrated that the magnitude of the association between sexual orientation and occupa-tional status depends on these measures, highlighting that each measure captures a unique aspect of sexual orientation and relates to the occupa-tional attainment process in unique ways.

People who report first same-sex experience in adolescence (the Early Development group) may face challenges in adolescence such as emotional and academic problems, but these people, on average, have about the same occupational status in young adulthood as those who do not report same-sex experience (the No Experience group). The results are consistent with the view that emphasizes sexual minorities’ resilience against sexuality stigma: They develop effective coping skills over time (Weinberg & Williams, 1974) and strive to counterbalance discrimination in the labor market (Hewitt, 1995). Similar compensating behaviors in the labor market have been reported for other socially disadvantaged groups (e.g., Kmec & Gorman, 2010). These results by no means suggest that sexual minorities’ challenges in adolescence are unimportant. These challenges are negatively associated with occupational status, so the occupational status of these sexual minorities


Table 5. OLS Regression Models Predicting Occupational Status by Same-Sex Dating Relationships Among Men

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

Same-sex dating (ref = no SSD) Early development 3.04 1.77 1.88 2.86 5.92 Late development 9.83** 7.24* 5.39 6.47* 6.64*Adolescent predictors

Depressive symptoms −0.11 −0.07 −0.07 −0.07Educational

expectations1.62*** 0.88*** 0.87*** 0.83***

GPA 9.04*** 4.05*** 3.89*** 4.26***Young adult predictors

Depressive symptoms −0.25* −0.21 −0.13Educational

attainment5.59*** 5.61*** 5.83***

Married at W4 4.31*** 3.11***Parent by W4 −0.12 −0.52Working part time −12.22***Percent women in

the occupation−0.08***

Sociodemographic backgroundAge at W4 0.86** 1.29*** 0.95*** 0.73** 0.64**Race (ref = Non-Hispanic

White) Non-Hispanic

Black−7.13*** −4.61** −4.01** −3.42** −2.70*

Hispanic 3.37** 3.80** 3.52** 3.49** 2.97* Others 7.06** 4.79* 3.95* 4.45* 5.18**Parent education

(ref = high school) Less than high

school−4.59** −2.06 0.10 −0.11 0.12

Some college 7.47*** 4.72*** 3.09** 3.20** 3.36** College graduate 12.87*** 6.89*** 2.84** 3.00** 3.64***

Inverse mills ratio −55.91*** −43.93*** −35.42*** −31.57** −25.00**Constant 25.35** −19.56* −11.93 −7.38 −2.58R2 0.10 0.22 0.29 0.30 0.33F(1,128) for b(SSD

early)=b(SSD late)1.83 1.23 0.54 0.57 0.03

Note. n = 5,513; W4 = Wave 4.*p < .05. **p < .01. ***p < .001.

Ueno et al. 23

would be even higher if they do not experience sexuality stigma in adoles-cence. The results therefore provide support for the continuation and further development of social policy to create a safe social environment, which would allow sexual minority youth to achieve their full potential. One excep-tion to this pattern is women who report early timing of same-sex contact; these women tend to have lower occupational status than the No Experience group. Sexual contact involves physical interaction and could take place out-side committed relationships. Therefore, the measure might have emphasized the self-exploratory aspect of sexuality. In this respect, the present results are consistent with a previous finding in the heterosexual literature that sexual exploration has negative consequences for status attainment, especially for women (Spriggs & Halpern, 2008).

People who report their first same-sex experience in young adulthood (the Late Development group) differ from the No Experience group in occupa-tional status, but the direction of the difference depends on gender. For women, those who report late development of same-sex attraction have lower occupational status, and the disadvantage is partly explained by their higher levels of depressive symptoms and lower levels of educational attainment in young adulthood. These findings are consistent with the argument that emerg-ing awareness of same-sex sexuality marks an entry into a socially stigma-tized role of a sexual minority and creates negative changes in everyday life such as identity confusion and relational strains (Ueno, 2010). The negative changes in turn undermine educational preparation and career planning (Boatwright et al., 1996; Schmidt & Nilsson, 2006). Alternatively, the result may suggest that low educational attainment in adolescence signals psycho-logical traits or social circumstances that delay the timing of first same-sex experience and constrain the occupational attainment process at the same time. This interpretation is consistent with another finding from this study. These women report lower GPAs and educational expectations in adoles-cence, before they report their first same-sex experience. More research is necessary to tease out the processes that account for these women’s occupa-tional status penalty.

This disadvantage linked to emerging same-sex experience is observed only for women, consistent with previous findings in the heterosexual litera-ture that sexual transitions affect women more negatively (Joyner & Udry, 2000; Meier, 2007). The association is more clearly observed when attraction is used as the sexual orientation measure, as opposed to dating relationships. This result may reflect the fact that many people first become aware of attrac-tion before engaging in dating relationships (Cass, 1979; Troiden, 1989), thereby amplifying the initial psychological shock and confusion resulting


from attraction. Although sexual contact also shows an occupational status penalty for women in the Late Development group, depressive symptoms account for only a small portion of their disadvantage (detailed results not presented in this article), suggesting that the link between emerging same-sex contact and low occupational status develops through different processes.

The occupational status disadvantage observed for women in the Late Development group is contrary to Baumle et al.’s (2009) census study, which showed higher occupational status for women cohabitating with same-sex partners relative to those heterosexually married or cohabitating. Three fac-tors may account for the discrepancy. First, the present study showed that the penalty for sexual minority women is limited to the Late Development group, and this specificity might have been masked in the census study, which did not address the timing of first same-sex experience. Second, unlike the cen-sus study, the present study focused on young adults, and sexual minority women in this life stage might not have accumulated their occupational advantages. Third, the census study limited the analysis to those who were married or cohabitating to determine sexual orientation from the partner’s gender. Sexual orientation difference in this subpopulation may not be gen-eralized to a broader population that includes both people with and without partners (i.e., the target population of the present study).

Among men, those who report first same-sex dating relationships in young adulthood tend to have higher occupational status than those who do not report such relationships. The advantage is partly explained by their higher educational performance, expectations, and attainment. The results are consistent with the argument that delayed sexual development indicates a high level of social control in adolescence, which promotes success in status attainment by keeping people focused on school work (Sewell & Hauser, 1980; Siennick & Staff, 2008). These men’s advantage is more pro-nounced in the analysis of dating relationships than in the analysis of attrac-tion, consistent with the argument that dating relationships more strongly reflect levels of social control, which are generally intended to regulate behaviors rather than affects. Although sexual contact also involves behav-iors, it does not show this advantage linked to late development. As men-tioned earlier, the measure of sexual contact emphasizes the self-exploratory aspect of sexuality, and the increasing level of self-exploration in young adulthood, as indicated in these people’s first same-sex contact, perhaps counterbalances the attainment advantage. Consistent with this interpreta-tion, men reporting first same-sex contact in young adulthood show signifi-cantly greater increases in the numbers of romantic and sexual partners between adolescence and young adulthood compared to other men (detailed

Ueno et al. 25

results not shown here). The interpretations of these men’s advantage remain tentative because the present study cannot rule out other interpretations. For example, men with high occupational status may feel more confident about their status and therefore less concerned about negative consequences of acknowledging their same-sex attraction or engaging in same-sex behaviors (Weinberg & Williams, 1974). These men may also live in a more tolerant social environment (workplaces, neighborhoods), which reduces constraints on same-sex behaviors in adulthood.

The occupational status advantage for men in the Late Development group partly contradicts Baumle et al.’s (2009) census study, which reported that men cohabitating with same-sex partners have lower occupational status than heterosexually married men, although they exceed heterosexually cohabitat-ing men. In the present study, the Late Development group’s advantage is observed before and after considering marital status, indicating that the group has higher status than both married and unmarried men. The discrepancy may have resulted from the life stage focus on young adulthood in this study. Contrary to previous earnings studies that included men of a wide age range (Budig & England, 2001; Glauber, 2007), the present study does not show an occupational status advantage for fathers, perhaps because fathers do not receive or accumulate the advantage until later life stages. Similarly, although the analysis shows a marriage premium for men, husbands may not receive the full benefit until later life stages as suggested in earnings research (Ahituv & Lerman, 2007). In the absence of substantial marriage and parenthood ben-efits, the limited access to marriage and reduced chance of parenthood may not create serious constraints on young sexual minority men’s occupational attainment.

Although the study cannot directly assess the role of discrimination due to a lack of measures, the results contradict the argument that the occupational status differences across sexuality groups emerges from work discrimination. If work discrimination plays a major part, sexual minority workers’ disad-vantage should be more pronounced among men, whose same-sexuality is more negatively viewed in society (Herek, 2002; Horn, 2007). Furthermore, the disadvantage should be observed for both the Early and Late Development groups because there is no reason to expect that timing of first same-sex experience shifts the chance of facing work discrimination. Finally, the dis-advantage should be more pronounced in the analysis based on dating rela-tionships than attraction, assuming that people who engage in dating relationships are more likely than those who are only aware of same-sex attraction to disclose their sexual orientation to their current or potential employers. As mentioned above, the analysis instead shows that the


occupational disadvantage is observed only for women in the Late Development group, and that the disadvantage is more clearly observed for attraction than dating relationships.6

Before dismissing the discrimination argument, a couple of issues need to be considered. First, discrimination may negatively impact sexual minorities’ occupational status over the course of their occupational careers, and the present results based on a young adult sample may indicate that sexual minorities in this life stage may not have experienced discrimination or its occupational consequences yet. Second, the specificity of occupational status penalty to the Late Development group does not necessarily undermine the discrimination argument if one considers both exposure to discrimination and resilience against discrimination. The lack of occupational status disadvan-tage for the Early Development group may indicate their resilience, instead of their lack of exposure to discrimination, whereas the disadvantage for the Late Development group may indicate that they have not developed effective coping skills against work discrimination. This interpretation highlights the need to measure both exposure and coping attempts in future research to fully understand the role of discrimination.

The present results have important implications for earnings research because occupational status is strongly linked to earnings (Magnusson, 2009). A follow-up analysis using personal earnings as the dependent vari-able showed significant group differences only when dating relationships were used as the sexual orientation measure. Among men, the Late Development group had higher earnings than other men after controlling for sociodemo-graphic attributes, and their higher occupational status partly accounted for the difference. Among women, the Early Development group had higher earnings than other women, but occupational status did not contribute to the explanation of the group difference. The results for women are consistent with previous research (e.g., Black et al., 2003; Blandford, 2003), and those for men are contrary to some previous studies that reported sexual minority penalty (e.g., Badgett, 1995; Black et al., 2003; Blandford, 2003) but partly consistent with recent Census 2000 studies, which reported that sexual minor-ity men earn more money than unmarried, heterosexual men (Antecol et al., 2008; Baumle et al., 2009). The follow-up analysis thus highlighted the importance of considering timing of first same-sex experience and types of same-sex experience in earnings research as well as occupational status research. Earnings disparities emerge through complex processes, and further analysis is necessary to uncover processes that give rise to these earnings disparities.

Ueno et al. 27

The study has four important limitations. First, taking advantage of data strengths in Add Health, the study sought to identify supply-side factors that link same-sex sexual development to occupational status. These factors are by no means exhaustive, and some unmeasured differences across sexuality groups may account for differences in occupational status. For example, work orientation (Johnson & Monserud, 2010) and work effort (Kmec & Gorman, 2010) may vary across the sexuality groups, but the study was unable to assess the role of these factors due to a lack of direct measures. Second, the data did not provide much information about demand-side fac-tors such as recruitment and termination methods (Byron, 2010; Kmec, 2005) and job sorting by employers (Williams & Connell, 2010). More research is needed to examine the extent to which these factors account for sexual orien-tation differences in occupational status. Third, the present study focused on occupational status and could not address sexual orientation differences in workplace characteristics. Workplaces vary substantially by organizational structure, worker sociodemographic composition, and level of segregation across worker groups (Robinson, Taylor, Tomaskovic-Devey, Zimmer, & Irvin, 2005), which have important implications for disparities in earnings (Avent-Holt & Tomaskovic-Devey, 2010), work commitment and motivations (Dickerson, Schur, Kruse, & Blasi, 2010), and development of informal ties among workers (Vallas, 2003). Similarly, occupational status does not pro-vide much information about work experiences in specific jobs. Add Health included several variables that scholars consider as elements of job quality (Kalleberg, 2011; Padavic & Reskin, 2002)—autonomy, repetition, supervi-sory responsibilities, job satisfaction, and perceived contribution to the long-term goal. A supplemental analysis revealed that sexuality groups did not show any systematic differences in these variables, but more research is needed to investigate how sexual minorities and heterosexuals may differ in other dimensions of job quality and work experience. Fourth, factors that account for sexual orientation differences in occupational status, job quality, and work experience, especially those on the demand side, may vary system-atically by cities, states, and regions because they are strongly influenced by the local labor market, labor laws and regulations, and social climate. Incorporating contextual information would therefore provide important insights into occupational differences between sexual minorities and hetero-sexuals in future research. For the same reason, more studies in non-U.S. countries are needed to expand the existing literature. As one of very few examples in this area, a recent Swedish study reports very different patterns of sexual orientation differences in occupational status, highlighting the importance of contextual factors (Ahmed, Andersson, & Hammarste, 2011).


Conclusions

The purpose of this study was to examine whether sexual orientation contributes to occupational status disparities. The results highlighted that sexual orientation operates very differently from other dimensions of occupational disparities such as gender and race because of its dynamic aspect. Important changes occur in sexual development during adolescence and young adulthood, which are also critical life stages for status attainment. Like heterosexual development, devel-opment of same-sex sexuality varies across individuals in the timing and creates opportunities and constraints in the occupational attainment process. People experience changes in occupational status over the course of their careers, and the trajectories differ across social groups (Warren et al., 2002). Similarly, sexu-ality continues to be somewhat fluid in middle and later stages of adulthood (Jensen, 1999; Pope & Schultz, 1991). These changes underscore the need to extend the present investigation by examining the dynamic association between sexual orientation and occupational status throughout adulthood.

Acknowledgment

The authors thank Irene Padavic and Karen Campbell for their thoughtful comments on earlier drafts.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by the National Science Foundation (1060988). This research uses data from Add Health, a program project directed by Kathleen Mullan Harris and designed by J. Richard Udry, Peter S. Bearman, and Kathleen Mullan Harris at the University of North Carolina at Chapel Hill, and funded by grant P01-HD31921 from the Eunice Kennedy Shriver National Institute of Child Health and Human Development, with cooperative funding from 23 other federal agencies and foundations. Special acknowledgment is due Ronald R. Rindfuss and Barbara Entwisle for assistance in the original design. Information on how to obtain the Add Health data files is available on the Add Health website (http://www.cpc.unc.edu/addhealth). No direct support was received from grant P01-HD31921 for this analysis.

Ueno et al. 29

Notes

1. This pattern in earnings disparities may be changing. Unlike previous studies, recent studies based on Census 2000 showed that men cohabitating with other men make more money than unmarried men cohabitating with women although they make less money than heterosexually married men (Antecol et al., 2008; Baumle et al., 2009).

2. In addition to these three groups, some people report same-sex experience only in adolescence. For example, in the analysis using dating relationships as a sexuality measure, 3% of women and 2% of men in the present sample reported this pat-tern. These people tended to show lower occupational status than those without same-sex experience. One interpretation is that temporary same-sex experience indicates self-exploration in the early stage of sexual development, which under-mines the status attainment process by causing distraction and lowering motiva-tions to achieve conventional goals (Brook & Newcomb, 1995; Hagan, 1997). Our main focus is to contrast people who report same-sex experience in young adulthood and those who do not and to examine whether timing of first same-sex experience has any implications for occupational status. For this reason, those who reported adolescence-only same-sex experience were combined to the No Expe-rience group in the analysis. An alternative analysis that excluded these people produced the same substantive conclusions.

3. Although these scales that combine education and earnings continue to be widely used, scholars have argued that each component should be separately analyzed because the combined scales tend to underestimate gender differences, intergen-erational status stability, and health consequences (Hauser & Warren, 1997; Miech & Hauser, 2001; Warren et al., 2002). Regarding sexual orientation differences, however, a supplemental analysis based on separate measures of occupational edu-cation and occupational earnings showed results very similar to the main analysis based on the combined measure (detailed results available upon request). In this article, we elected to present the results based on the combined measure to sim-plify the presentation and facilitate comparisons to other studies of occupational status disparities based on the frequently used composite scale.

4. A probit model was run to predict participation in the labor force. The model included the following predictors: physical limitation, age, race, parental educa-tion, depressive symptoms in Wave 1 and Wave 3, high school GPA, educational qualification in Wave 4, and marital and parental statuses in Wave 4. From this model, inverse Mills ratio was computed and entered in each OLS model as a predictor (Heckman, 1979).

5. Considering the possibility that sexual orientation differences may be more pro-nounced in some segments of the occupational status hierarchy, a supplemental


analysis was conducted using a trichotomized version of the dependent variable (top 25%, middle 50%, and bottom 25% in the occupational status score). The analysis showed significant sexuality group differences between top and middle occupations and between bottom and middle occupations, indicating that sexual orientation differences found in the primary analysis reflected differences over a wide spectrum of the status hierarchy.

6. The current policy discussions regarding discrimination against sexual minority workers tend to focus on those who identify as gay, lesbian, and bisexual. The analysis did not use self-identity as a sexual orientation measure because Add Health included sexual identity measures only in young adulthood waves (Waves 3 and 4), which precluded the determination of timing of sexual development—the main focus of this article. Nonetheless, a supplemental analysis showed that among women, those who identified as bisexual in young adulthood had lower occupational status than those identified as heterosexual, whereas homosexual women did not differ from heterosexual women. Among men, homosexual people had higher occupational status than heterosexual people, and bisexual people had lower status than these two groups.

References

Ahituv, A., & Lerman, R. I. (2007). How do marital status, work effort, and wage rates interact? Demography, 44, 623-647.

Ahmed, A. M., Andersson, L., & Hammarstedt, M. (2011). Sexual orientation and occupational rank. Economics Bulletin, 31, 2422-2433.

Antecol, H., Jong, A., & Steinberger, M. (2008). The sexual orientation wage gap: The role of occupational sorting and human capital. Industrial and Labor Rela-tions Review, 61, 518-543.

Avent-Holt, D., & Tomaskovic-Devey, D. (2010). The relational basis of inequality: Generic and contingent wage distribution processes. Work and Occupations, 37, 162-193.

Badgett, M. V. L. (1995). The wage effects of sexual orientation discrimination. Industrial and Labor Relations Review, 48, 726-739.

Badgett, M. V. L., & King, M. C. (1997). Occupational strategies of lesbians and gay men. In A. Gluckman & B. Reed (Eds.), HomoEconomics: Capitalism, commu-nity and lesbian and gay life (pp. 73-86). New York, NY: Routledge.

Baumle, A. K. (2009). The cost of parenthood: Unraveling the effects of sexual orien-tation and gender on income. Social Science Quarterly, 90, 983-1002.

Baumle, A. K., Compton, D. R., & Poston, D. L. (2009). Same-sex partners: The demography of sexual orientation. Albany, NY: SUNY Press.

Ueno et al. 31

Barrett, D. C., Pollack, L. M., & Tilden, M. L. (2002). Teenage sexual orientation, adult openness, and status attainment in gay males. Sociological Perspectives, 45, 163-182.

Berg, N., & Lien, D. (2002). Measuring the effect of sexual orientation on income: Evidence of discrimination? Contemporary Economic Policy, 20, 394-414.

Berger, R. M. (1982). Gay and gray: The older homosexual man. Urbana, IL: Uni-versity of Illinois Press.

Black, D. A., Makar, H. R., Sanders, S. G., & Taylor, L. J. (2003). The earnings effects of sexual orientation. Industrial and Labor Relations Review, 56, 449-469.

Blau, P. M., & Duncan, O. D. (1967). The American occupational structure. New York, NY: Wiley.

Blandford, J. M. (2003). The nexus of sexual orientation and gender in the determina-tion of earnings. Industrial and Labor Relations Review, 56, 622-642.

Boatwright, K. J., Gilbert, M. S., Forrest, L., & Ketzenberger, K. (1996). Impact of identity development upon career trajectory: Listening to the voices of lesbian women. Journal of Vocational Behavior, 48, 210-228.

Brook, J. S., & Newcomb, M. D. (1995). Childhood aggression and unconventional-ity: Impact on later academic achievement, drug use, and workforce involvement. The Journal of Genetic Psychology, 156, 393-410.

Budig, M. J., & England, P. (2001). The wage penalty for motherhood. American Sociological Review, 66, 204-225.

Buhi, E. R., & Goodson, P. (2007). Predictors of adolescent sexual behavior and inten-tion: A theory-guided systematic review. Journal of Adolescent Health, 40, 4-21.

Byron, R. A. (2010). Discrimination, complexity, and the public/private sector ques-tion. Work and Occupations, 37, 435-475.

Cass, V. C. (1979). Homosexual identity formation: A theoretical model. Journal of Homosexuality, 4, 219-235.

Chantala, K. (2006). Guidelines for analyzing add health data. Retrieved from http://www.cpc.unc.edu/projects/addhealth/files/wt_guidelines.pdf

Cohen, P. N. (2002). Cohabitation and the declining marriage premium for men. Work and Occupations, 29, 346-363.

Cohen, P. N., Huffman, M. L., & Knauer, S. (2009). Stalled progress? Gender segre-gation and wage inequality among managers, 1980-2000. Work and Occupations, 36, 318-342.

Correll, S. J., Benard, S., & Paik, I. (2007). Getting a job: Is there a motherhood pen-alty? American Journal of Sociology, 112, 1297-1338.

Dickerson, N., Schur, L., Kruse, D., & Blasi, J. (2010). Worksite segregation and performance-related attitudes. Work and Occupations, 37, 45-72.


Elder, G. H., Jr., Johnson, M. K., & Crosnoe, R. (2003). The emergence and develop-ment of life course theory. In: J. T. Mortimer & M. J. Shanahan Handbook of the life course (pp. 3-19). New York, NY: Kluwer Academic.

Elman, C., & O’Rand, A. M. (2004). The race is to the swift: Socioeconomic origins, adult education, and wage attainment. American Journal of Sociology, 110, 123-160.

England, P. (1979). Women and occupational prestige: A case of vacuous sex equal-ity. Signs, 5, 252-265.

England, P. (2005). Gender inequality in labor markets: The role of motherhood and segregation. Social Politics, 12, 264-288.

Frisco, M. L. (2008). Adolescents’ sexual behavior and academic attainment. Sociol-ogy of Education, 81, 284-311.

Gilligan, C. (1982). In a different voice: Psychological theory and women’s develop-ment. Cambridge, MA: Harvard University Press.

Glauber, R. (2007). Marriage and the motherhood wage penalty among African Americans, Hispanics, and Whites. Journal of Marriage and Family, 69, 951-961.

Glauber, R. (2012). Women’s work and working conditions: Are mothers compen-sated for lost wages? Work and Occupations, 39, 115-138.

Greenman, E. (2011). Asian American-White differences in the effect of motherhood on career outcomes. Work and Occupations, 38, 37-67.

Hagan, J. (1997). Defiance and despair: Subcultural and structural linkages between delinquency and despair in the life course. Social Forces, 76, 119-134.

Harris, K. M., Halpern, C. T., Whitsel, E., Hussey, J., Tabor, J., Entzel, P., & Udry, J. R. (2009). National Longitudinal Study of Adolescent Health: Research design. Retrieved from http://www.cpc.unc.edu/projects/addhealth/design

Harry, J., & DeVall, W. B. (1978). The social organization of gay males. New York, NY: Praeger.

Hauser, R. M., & Warren, J. R. (1997). Socioeconomic indexes for occupations: A review, update, and critique. Sociological Methodology, 27, 177-298.

Heckman, J. J. (1979). Sample selection bias as a specification error. Econometrica, 47, 153-161.

Herek, G. M. (2002). Gender gaps in public opinion about lesbians and gay men. The Public Opinion Quarterly, 66, 40-66.

Hewitt, C. (1995). The socioeconomic position of gay men: A review of the evidence. American Journal of Economics and Sociology, 54, 461-479.

Hirschi, T. (1969). Causes of delinquency. Berkeley, CA: University of California Press.Horn, S. S. (2007). Adolescents’ acceptance of same-sex peers based on sexual ori-

entation and gender expression. Journal of Youth and Adolescence, 36, 363-371.Jensen, K. L. (1999). Lesbian epiphanies: Women coming out in later life. New York,

NY: Harrington Park Press.

Ueno et al. 33

Johnson, M. K., & Monserud, M. A. 2010. Judgments about work and the features of young adults’ jobs. Work and Occupations, 37, 194-224.

Joyner, K., & Udry, J. R. (2000). You don’t bring me anything but down: Adolescent romance and depression. Journal of Health and Social Behavior, 41, 369-391.

Kalleberg, A. L. (2011). Good jobs, bad jobs: The rise of polarized and precarious employment systems in the United States, 1970s to 2000s. New York, NY: SAGE.

Kmec, J. A. (2005). Setting occupational sex segregation in motion: Demand-side explanations of sex traditional employment. Work and Occupations, 32, 322-354.

Kmec, J. A., & Gorman, E. H. (2010). Gender and discretionary work effort: Evi-dence from the United States and Britain. Work and Occupations, 37, 3-36.

Ku, M. C. (2011). When does gender matter? Gender differences in specialty choice among physicians. Work and Occupations, 38, 221-262.

Laumann, E. O., Gagnon, J. H., Michael, R. T., & Michaels, S. (1994). The social organization of sexuality: Sexual practices in the United States. Chicago, IL: Uni-versity of Chicago Press.

Longmore, M. A., Eng, A. L., Giordano, P. C., & Manning, W. D. (2009). Parenting and adolescents’ sexual initiation. Journal of Marriage and Family, 71, 969-982.

Magnusson, C. (2009). Gender, occupational prestige, and wages: A test of devalua-tion theory. European Sociological Review, 25, 87-101.

Meier, A. M. (2007). Adolescent first sex and subsequent mental health. American Journal of Sociology, 112, 1811-1847.

Miech, R. A., Eaton, W., & Liang, K. Y. (2003). Occupational stratification over the life course: A comparison of occupational trajectories across race and gender dur-ing the 1980s and 1990s. Work and Occupations, 30, 440-473.

Nakao, K., & Treas, J. (1994). Updating occupational prestige and socioeconomic scores: How the new measures measure up. Sociological Methodology, 24, 1-72.

Nam, C. B., & Boyd, M. (2004). Occupational status in 2000: Over a century of census-based measurement. Population Research and Policy Review, 23, 327-358.

Newcomb, M. D., & Bentler, P. M. (1988). Consequences of adolescent drug use: Impact on the lives of young adults. Newbury Park, CA: SAGE.

Padavic, I., & Reskin, B. F. (2002). Women and men at work (2nd ed). Thousand Oaks, CA: Pine Forge Press.

Pascoe, C. J. (2007). Dude, you’re a fag: Masculinity and sexuality in high school. Berkeley: University of California Press.

Pearlin, L. I., Schieman, S., Fazio, E. M., & Meersman, S. C. (2005). Stress, health, and the life course: Some conceptual perspectives. Journal of Health and Social Behavior, 46, 205-219.


Pearson, J., Muller, C., & Wilkinson, L. (2007). Adolescent same-sex attraction and academic outcomes: The role of school attachment and engagement. Social Prob-lems, 54, 523-542.

Petersen, A. C., & Taylor, B. (1980). The biological approach to adolescence. In J. Adel-son (Ed.), Handbook of adolescent psychology, (pp. 115-155). New York: Wiley.

Pope, M., & Schulz, R. (1990). Sexual attitudes and behavior in midlife and aging homosexual males. Journal of Homosexuality, 20, 169-177.

Radloff, L. S. (1977). The CES-D scale: A self-report depression scale for research in the general population. Applied Psychological Measurement, 1, 385-401.

Reynolds, J., Stewart, M., MacDonald, R., & Sischo, L. (2006). Have adolescents become too ambitious? High school seniors’ educational and occupational plans, 1976 to 2000. Social Problems, 53, 186-206.

Riegle-Crumb, C., Muller, C., Frank, K., & Schiller, K. S. (2005). National Longi-tudinal Study of Adolescent Health Wave III Transcript Data. Chapel Hill, NC: Carolina Population Center, University of North Carolina at Chapel Hill.

Robinson, C. L., Taylor, T., Tomaskovic-Devey, D., Zimmer, C., & Irvin, M. W., Jr. (2005). Studying race or ethnic and sex segregation at the establishment level: Methodological issues and substantive opportunities using EEO-1 reports. Work and Occupations, 32, 5-38.

Rudolph, K. D. (2002). Gender differences in emotional responses to interpersonal stress during adolescence. Journal of Adolescent Health, 30, 3-13.

Saltiel, J. (1988). The Wisconsin model of status attainment and the occupational choice process: Applying a continuous-choice model to a discrete-choice situa-tion. Work and Occupations, 15, 334-355.

Savin-Williams, R. C., & Ream, G. L. (2007). Prevalence and stability of sexual ori-entation components during adolescence and young adulthood. Archives of Sexual Behavior, 36, 385-385.

Schmidt, C. K., & Nilsson, J. E. (2006). The effects of simultaneous developmental processes: Factors relating to the career development of lesbian, gay, and bisexual youth. The Career Development Quarterly, 55, 22-37.

Schneider, B. E. (1986). Coming out at work: Bridging the private/public gap. Work and Occupations, 13, 463-487.

Sewell, W. H., & Hauser, R. M. (1980). The Wisconsin longitudinal study of social and psychological factors in aspirations and achievements. Research in Sociology of Education and Socialization, 1, 59-99.

Siennick, S., & Staff, J. (2008). Explaining the educational deficits of delinquent youths. Criminology, 46, 609-635.

Sloan, M. M. (2012). Unfair treatment in the workplace and worker well-being: The role of coworker support in a service work environment. Work and Occupations, 39, 3-34.

Ueno et al. 35

Spriggs, A. L., & Halpern, C. T. (2008). Timing of sexual debut and initiation of postsecondary education by early adulthood. Perspectives on Sexual and Repro-ductive Health, 40, 152-161.

Steward, N. R., Farkas, G., & Bingenheimer, J. B. (2009). Detailed educational path-ways among females after very early sexual intercourse. Perspectives on Sexual and Reproductive Health, 41, 244-252.

Tilcsik, A. (2011). Pride and prejudice: Employment discrimination against openly gay men in the United States. American Journal of Sociology, 117, 586-626.

Tolman, D. L. (2006). In a different position: Conceptualizing female adolescent sex-uality development within compulsory heterosexuality. New Directions for Child and Adolescent Development, 2006, 71-89.

Troiden, R. R. (1989). The formation of homosexual identities. Journal of Homosexu-ality, 17, 43-73.

Turner, C. F., Villarroel, M. A., Chromy, J. R., Eggleston, E., & Rogers, S. M. (2005). Same-gender sex among U.S. adults: Trends across the twentieth century and dur-ing the 1990s. Public Opinion Quarterly, 69, 439-462.

Ueno, K. (2010). Same-sex experience and mental health during the transition between adolescence and young adulthood. Sociological Quarterly, 51, 484-510.

Ueno, K., Roach, T. A., & Peña-Talamantes, A. E. (in press). The dynamic association between sexual orientation and educational attainment. Advances in Life Course Research.

US Department of Labor, Bureau of Labor Statistics. (2009). 2000 Standard Occupa-tional Classification. Retrieved from http://www.bls.gov/soc/socstruc.pdf.

Vallas, S. P. (2003). Rediscovering the color line within work organizations: The “Knitting of Racial Groups” revisited. Work and Occupations, 30, 379-400.

Warren, J. R., Hauser, R. M., & Sheridan, J. T. (2002). Occupational stratification across the life course: Evidence from the Wisconsin Longitudinal Study. American Sociological Review, 67, 432-455.

Weinberg, M. S., & Williams, C. J. (1974). Male homosexuals: Their problems and adaptations. New York, NY: Oxford University Press.

Whitam, F. L. (1983). Culturally invariable properties of male homosexuality: Tenta-tive conclusions from cross-cultural research. Archives of Sexual Behavior, 12, 207-226.

Williams, C. L., & Connell, C. (2010). “Looking good and sounding right”: Aesthetic labor and social inequality in the retail industry. Work and Occupations, 37, 349-377.

Bios

Koji Ueno is an associate professor of sociology at Florida State University, FL, USA. His research draws from life course and social psychological theories and


investigates the effects of sexual orientation on social networks, mental health, and status attainment outcomes.

Abráham E. Peña-Talamantes is a McNair Scholar and doctoral candidate in the Department of Sociology at Florida State University, FL, USA. His research focuses particularly in the experiences, identity work processes, and linguistic and educational inequalities faced by GLBT Latinas/os in higher education.

Teresa A. Roach is a doctoral candidate in the Department of Sociology at Florida State University, FL, USA. Her research is focused in the areas of gender and sexual-ity. Most recently, Teresa is conducting research on gender dynamics within consen-sual nonmonogamous relationships.

Sexual Orientation and Occupational Attainment

Documents