The complex contribution of sociodemographics to ...

2 downloads 0 Views 766KB Size Report
like to thank Lisa Diamond, Timothy Smith, and Donald Strassberg for their helpful ...... Loving, T. J., Heffner, K. L., Kiecolt‐ Glaser, J. K., Glaser, R., & Malarkey, ...
Title page with All Author Information

© 2017, American Psychological Association. This paper is not the copy of record and may not exactly replicate the final, authoritative version of the article. Please do not copy or cite without authors permission. The final article will be available, upon publication, via its DOI: 10.1037/fam0000234

The complex contribution of sociodemographics to decision-making power in gay male couples

Nicholas S. Perry1, David M. Huebner1, Brian R. W. Baucom1, & Colleen C. Hoff2

1

Department of Psychology, University of Utah, Salt Lake City, UT Center for Research and Education on Gender and Sexuality, San Francisco State University, San Francisco, CA 2

Corresponding author: Nicholas Perry Department of Psychology University of Utah, Room 502 380 South 1530 East Salt Lake City, UT, 84112 E-mail: [email protected] Phone: 781-526-5179 Fax: (801) 581-5841 Acknowledgements: Portions of these analyses were previously presented at the 2016 annual meeting of the Association for Behavioral and Cognitive Therapies. Support for this study came from NIMH grant R01MH65141 awarded to Colleen Hoff. Support for analyses came from a Gustafson Fellowship awarded by the University of Utah to Nicholas Perry. The authors would like to thank Lisa Diamond, Timothy Smith, and Donald Strassberg for their helpful comments. Lastly, we thank the participants for their time and for sharing their relationships.

RUNNING HEAD: Power and Gay Male Couples 2 Abstract Relationship power is an important dyadic construct in close relationships that is associated with relationship health and partner’s individual health. Understanding what predicts power in heterosexual couples has proven difficult, and even less is known about gay couples. Resource models of power posit that demographic characteristics associated with social status (e.g., age, income) confer power within the relationship, which in turn shapes relationship outcomes. We tested this model in a sample of gay male couples (N=566 couples), and extended it by examining race and HIV status. Multilevel modeling was used to test associations between demographic bases of power and decision-making power. We also examined relative associations among demographic bases and decision-making power with relationship satisfaction, given the literature on power imbalances and overall relationship functioning. Results showed that individual income was positively associated with decision-making power, as was participant’s HIV status, with HIV-positive men reporting greater power. Age differences within the relationship interacted with relationship length to predict decision-making power, but not satisfaction. HIV-concordant positive couples were less satisfied than concordant negative couples. Higher power partners were less satisfied than lower power partners. Demographic factors contributing to decision-making power among same-sex male couples appear to share some similarities with heterosexual couples (e.g., income is associated with power), as well as have unique features (e.g., HIV status influences power). However, these same demographics did not reliably predict relationship satisfaction in the manner that existing power theories suggest. Findings indicate important considerations for theories of power among same-sex male couples. Key Words: power; same-sex couples; gay couples; relationship satisfaction

Power and Gay Male Couples

1

Introduction In relationship science, power has been construed as an inherently dyadic process between two partners (Huston, 1983), characterized by the ability of one partner to influence the other toward a desired outcome. Intimate relationship power has been highlighted as a central dynamic within couples (Cromwell, 1975; Dunbar & Burgoon, 2005; Gray-Little & Burks, 1983; Huston, 1983) and has been linked to a couple’s relationship health, as individual partners’ mental and physical health (Brock & Lawrence, 2011; Loving et al., 2004). Despite its importance, previous studies have focused primarily on power within samples of heterosexual couples and far less is known about what characteristics influence power for same-sex couples. Indeed, these factors may differ substantially from those that predict for heterosexual couples. It is imperative that the state of the science keeps pace with the rapidly expanding legal and social recognition that same-sex couples are experiencing within the United States, and researchers must work actively to repair these gaps in the empirical base. Some evidence suggests that gay couples are more similar to than different from heterosexual couples in a number of important relationship domains (e.g., satisfaction, commitment, support, conflict) (e.g., Graham & Barnow, 2013; Julien, Bouthillier, & Begin, 2003; see Kurdek, 2005 for a review; Roisman, Clausell, Holland, Fortuna, & Elieff, 2008). However, many of these studies have looked for differences in relationship functioning by comparing samples of heterosexual couples to gay and lesbian couples. Although this approach might have some value in identifying global differences across types of couples based on sexual orientation, it also obscures potentially meaningful variability within distinct groups, such as differences among gay couples based on other demographic characteristics beyond sexual orientation. Further, testing theoretical models within diverse groups of couples allows for

Power and Gay Male Couples

2

refinement or adaptations to existing theories that expand our understanding of universal relationship processes and increases the generalizability of relationship theory. Given this dearth in the literature, current theories of power are inherently limited in the scope of couples they can be applied to by researchers or clinicians. Understanding relationship power Despite its theoretical importance, models of power have been inconsistently defined (Gray-Little & Burks, 1983). One factor complicating the study of power is the idea that power shapes other relationship processes through its presence, but is difficult to directly observe (Komter, 1989). In a seminal work on relationship power, Cromwell (Cromwell & Olson, 1975) outlined a tripartite model organized by power bases, power processes, and power outcomes. Power bases were considered attributes of individual partners (i.e., resources or characteristics) that provided one with capital within their relationship. Common examples have included income and education (Gray-Little & Burks, 1983). Power processes were described as the process by which couples express power, such as communication. Power outcomes were the result of decision-making. Functionally, power outcomes were defined as how a conflict is resolved or which partner “wins” a power struggle. Within these resource-based models, power bases would provide individual partners with influence (i.e., power process) to use in negotiating outcomes for the couple (Cromwell & Olson, 1975; Thibault & Kelly, 1959). However, the majority of studies using these multiple components of power have looked for convergent evidence in their influence on an outcome, rather than directly testing the theoretical relationships among them. For example, Babcock and colleagues (1993) examined the correlations among multiple measures of power (i.e., bases – education, income; processes – demand/withdraw behavior; outcome – decision-making). They

Power and Gay Male Couples

3

then examined convergent validity across these measures in their prediction of intimate partner violence. Because most of the literature has either used this approach of looking for convergent evidence across components of power or examined only one power construct singly, the suggestion in resource-based theories that individual characteristics predict influence within a couple remains largely untested empirically, in either heterosexual or same-sex couples. Relationship power and relationship health Various studies have consistently found associations between intimate relationship power and the quality of romantic relationships, as well as the health of individual partners. This literature suggests that power is a prominent relationship process that influences other, larger outcomes within the relationship, such as partner’s general relationship satisfaction. For example, lower power partners can experience a lack of control within the relationship or less freedom to pursue individual interests, reducing their overall happiness in their relationship. An extensive review by Gray-Little and Burks (1983) among heterosexual couples suggested that egalitarian couples were generally the most satisfied with their relationships, although findings varied somewhat depending on the operationalization of power. However, there is little empirical evidence documenting the importance of relationship power for general relationship satisfaction among same-sex couples. Predictors of power among gay couples A small body of evidence suggests that individual resources are important in determining who holds power within same-sex male couples (see Peplau & Fingerhut, 2007 for a review). As Peplau and Fingerhut note in their review, although studies found that the vast majority of samesex male couples felt the ideal balance of power was an equal one, certainly not all couples are able to achieve this optimal balance. For example, research has found that older men and men

Power and Gay Male Couples

4

with greater income in same-sex male couples have been shown to have more power in their relationships (Harry 1984; Harry & DeVall, 1978). Blumstein and Schwartz (1983) also reported income as a determinant of power for gay male couples. However, the social and political climate has changed dramatically since these studies were originally conducted, and a more current study of these factors’ association with power is needed. Further, these studies lacked the statistical tools that are currently available (e.g., multilevel modeling) to appropriately examine how differences between partners within a couple impact power. Another line of research on HIV risk among young gay and bisexual men has suggested that young men, particularly Black men, are less likely to use condoms with older partners (Arrington-Sanders et al., 2013; Bingham et al., 2003; Joseph et al., 2011; Newcomb & Mustanski, 2013). Some have suggested that these age differences confer power over condom use to the older partner (Mustanski & Newcomb, 2014). However, direct associations between age differences and power have not been explored. Further, participant age is often confounded by relationship length (and vice-versa) and exploring their interaction might help better specify when during the relationship these age differences are meaningful. We also know far less about the role of other demographic indicators (e.g., HIV status, race) in structuring power for same-sex male couples, despite some qualitative and quantitative evidence that they are likely relevant (e.g., Remien, Carballo-Dieguez, & Wagner, 1995; Diaz, Ayala, & Bien, 2004). For example, Remien and colleagues conducted a qualitative study of gay men in HIV-serodiscordant relationships (i.e., where one partner is HIV-positive and one partner is HIV-negative). They found that differences in HIV status affected a number of the emotional dynamics of their relationships, including concerns about sexual health and perceptions of attractiveness. In addition, HIV is a heavily stigmatized chronic illness and, in combination with

Power and Gay Male Couples

5

these emotional dynamics, may shape HIV-positive men to be more submissive or acquiescent to their partners. With respect to race and ethnicity, qualitative evidence among Latino gay men has indicated that social oppression (e.g., discrimination, low SES) might drive their engagement in condomless sex (Diaz, 1998). Racial discrimination and homophobia among young Black men have also been associated with increased condomless sex, as a function of their engagement in challenging sexual situations where they might have less power (e.g., in public, under the influence of drugs or alcohol) (Huebner et al., 2014). These same mechanisms of social oppression might also work to reduce the agency or power ethnic minority gay men have in their romantic relationships more generally. Finally, the existing studies on power among gay male couples have generally failed to take into account the dyadic context of the couple. For example, age differences within the couple might vary across a relationship’s developmental trajectory, or income differences between partners might have very different social and practical significance when a couple, on average, makes very little money versus being middle class or wealthy. Intuitively, this dyadic context could be an essential consideration when exploring associations among demographic differences between partners and power. Current study and hypotheses Given these limitations in the existing literature, the field would benefit from a clearer understanding of what shapes power and how it impacts relationship functioning among samesex couples. The current paper addresses these gaps by applying one of the most frequently used models of power in heterosexual couples (Cromwell’s resource-based model) to decision-making power in gay male couples.

Power and Gay Male Couples

6

Consistent with previous resource-based models of power, we hypothesized that older partners, higher earning partners, more highly educated partners, HIV negative men, and White men would hold greater social resources and, thus, be afforded greater decision-making power within their relationships. Second, we hypothesized that differences between partners in age, education, income, race and HIV status would be associated with lower relationship satisfaction. We anticipated that partners with higher demographic status (i.e., older men, more highly educated men, higher earning men, HIV-negative men, and White men) would reporter greater relationship satisfaction. We also explored whether self-reported power differences within couples were associated with relationship satisfaction. However, given the limited literature on this association at the within-couple level, we did not specifically hypothesize a direction for this effect a priori. Finally, we explored cross-level interactions (partner variable x couple variable) of key predictors to contextualize these within-couple effects across different types of relationships (e.g., do partner differences in age matter more for longer-term vs. newer couples?). Method Procedure Data for the current study come from a larger study of HIV risk among sexual minority men in primary relationships. Couples were recruited in the San Francisco Bay Area between 2005 and 2007. Research staff used both active (e.g., community outreach at gay-identified social venues and health centers) and passive (e.g., advertisements in gay newspapers and websites) recruitment strategies. Eligibility criteria included each partner being over 18 years old, having been in a primary relationship together for at least 3 months, being fluent in English, and being a resident of the San Francisco Bay Area. ‘Primary partner’ was defined for eligible participants as a man

Power and Gay Male Couples

7

one is “committed to above anyone else and with whom he has had sex.” Each partner also needed to have knowledge of his own and his partner’s self-reported HIV status. Eligible couples were then scheduled to complete self-report batteries at the local research offices in San Francisco. Both partners provided written informed consent and then completed self-report questionnaires via audio computer-assisted interview (ACASI) independently. Each partner received $40 for completing the self-report battery. Questionnaires took approximately 70 minutes to complete. An Institutional Review Board where the data were collected approved all study procedures. Participants One thousand one hundred and thirty-two men (566 couples) completed study procedures. The sample was racially and economically diverse. 47% of couples identified as interracial (90% White/non-White dyads), 45% of couples both partners identified as White, 5% as African-American, 2% as Latino, 1% as Asian-American/Pacific Islander, and less than 1% Native-American. For individual partners, 45% reported earning less than $30,000 per year, 30% earned $30,000-59,999, 16% earned $60,000-99,999, and 9% earned $100,000 or more. 4.5% of participants had less than a high school education, 12.5% had a high school diploma or GED, 32% had completed some college education, 28% had a college degree, and 23% had a graduate degree. The average age of individual participants was 41.7 years (SD = 11.4). The average length of relationship was 6.9 years (SD = 8.5; median = 4 years). 77% of partners reported they were living together at the time of the study. With regards to HIV-status, efforts were made to specifically recruit dyads that represented the spectrum of dyadic HIV status (concordant HIV-negative, concordant HIV-

Power and Gay Male Couples

8

positive, and “serodiscordant,” where one partner is HIV-positive and the other is HIV-negative). Three hundred and ten couples identified as concordant HIV-negative, 124 couples identified as concordant HIV-positive, and 132 couples identified as HIV serodiscordant. Measures Demographics Single items assessed participants’ self-reported age, income, education, racial/ethnic identity, and HIV-status. Age was self-reported as a continuous variable. Individual’s education was reported as highest level of education completed. Individual’s annual income was reported in equal income brackets, as listed above. HIV-status was self-reported as HIV-positive or negative. Race was self-reported using the categories listed above. Because of small cell sizes in some of the racial categories (e.g., American Indian/Alaskan Native, n=15 men) and the theoretical reasoning that non-White men are generally socially disadvantaged compared to White men, participants were categorized as 0 (“Non-White) or 1 (“White”) for primary analyses. Relationship power A psychometrically sound scale developed to measure sexual relationship power in heterosexual women (Pulerwitz, Amaro, Jong, Gortmaker, & Rudd, 2002) was adapted for the larger study. Several items relevant to MSM were added (e.g., regarding condomless anal sex) and some of the original scale items were removed to reduce participant burden (e.g., those that loaded less strongly onto their respective factor in the original factor analysis, Pulerwitz, Gortmaker, & DeJong, 2000). Items used a 5-point Likert scale to assess power to control an

Power and Gay Male Couples

9

outcome relative to one’s partner across a variety of contexts. For example, a sample item was “I make most of the important decisions that affect us.” In preliminary factor analyses of the adapted 12-item scale, three subscales emerged: ‘Lack of power about barebacking’, ‘Power in condom negotiation’, and ‘Power in decisionmaking’ (R01MH065141, PI: Hoff, unpublished data). Given the theoretical basis for the current study, only the ‘Power in decision-making’ subscale was used. This resulted in a final 7-item scale, showing acceptable internal consistency (Cronbach’s α = .80). Relationship satisfaction The 10–item dyadic satisfaction subscale of the Dyadic Adjustment Scale (DAS; Spanier, 1975) was used to measure relationship satisfaction. This subscale of the DAS has demonstrated reliability in distinguishing distressed heterosexual couples from non-distressed couples (Mark, Heyman, & Weiss, 1991) and has shown similar construct validity in samples of gay couples (Kurdek, 1992). In the current sample, the subscale demonstrated acceptable internal consistency (Cronbach’s α =.81). Analysis plan Multilevel modeling was used to account for the interdependence in the data introduced by having measures from both members of the couple. All continuous within-couple predictors were couple-mean centered (around the average for the couple) and all continuous couple-level predictors were grand-mean centered (around the average for all couples). Dichotomous variables were entered at the within-couple level dummy coded for the individual partner (e.g., 0 for HIV-negative, 1 for HIV-positive; 0 for non-White, 1 for White) and entered at the betweencouple level dummy coded if the couple was the same or different on the within-couple variable

Power and Gay Male Couples 10 (e.g., 0 for HIV-concordant, 1 for HIV-serodiscordant; 0 for White/White or non-White/nonWhite (i.e., minority), 1 for non-White/White).1 We also explored cross-level interactions between partner-level and couple-level variables of interest. Specifically for HIV-status and race, this allowed us to understand differences across couples’ concordance on HIV serostatus and racial composition. For variables treated continuously (i.e., income, age, education), these interactions tested whether the magnitude of the effect of a demographic difference between partners varied as a function of the couple’s average on that variable. For example, we were able to test whether an income difference between partners was associated with power more or less strongly at different levels of a couple’s average income. Finally, we examined the cross-level interaction between age relative to one’s partner and relationship length to better understand the developmental features of the impact of age differences on relationship functioning. An example equation using income as the predictor of decision-making power is included below. Level 1: Decision-making Powerij = β0j + β1j* Partner’s Difference from Couple’s Average Income + rij Level 2: β0j = γ00 + γ01* Couple Average Income1j + u0j Here, our hypothesized effects are embedded within the above equation, with i denoting individuals and j denoting couples. γ00 is the intercept of decision-making power, γ10 represents the effect of a partner’s difference from the couple’s average income on the outcome (individual power), and γ01 represents the effect of couples’ average income on power, relative to other

There are multiple possible ways couples’ racial composition could be coded within the current data. We elected to code couples based on whether their race was the majority race within the United States (White) or not, as we felt this best aligned with theories suggesting that power within couples operates in part as a function of individual’s larger social status. 1

Power and Gay Male Couples 11 couples in the sample. rij represents the variability in the outcome around the average for partners and u0j represents the variability in the effect around the average for all couples. Two sets of models were run: one set to predict individual relationship power from relevant demographics and the second set to predict relationship satisfaction from relevant demographics and decision-making power2. Both sets of models were run first with each independent variable separately. Then, a final multiple predictor model was run that included all predictors significant in the independent models. Results are presented for both the betweencouple and within-couple effects of our predictors of interest. However, we focus our interpretation on within-couple results given the purpose of the current paper was to understand power dynamics within relationships (rather than across couples). All models were run in HLM 7.0 (Bryk, Raudenbush, & Congdon, 1996) and all results are reported using robust standard erors. Simple slopes for significant cross-level interactions were tested using recommended methods for MLMs (http://quantpsy.org/interact/hlm2.htm; Preacher, Curran, & Bauer, 2006). Results Predictors of decision-making power Descriptive statistics and intercorrelations among study variables are reported in Table 1. The minimum score on the scale in the sample was 7 and the maximum score was 32. The bottom 10% of the sample reported a score at or below 10 and the top 10% of the sample reported a score at or above 22, indicating a wide distribution of our outcome variable in our

We wish to note that findings on relationship satisfaction and couples’ HIV status in this sample have been previously reported in Hoff et al., 2009. However, the current study is the first to examine the effects of HIV at the partner-level to distinguish serodiscordant partners and is also the first to document this association controlling for other relevant demographics (e.g., race, SES). 2

Power and Gay Male Couples 12 sample. Results for the independent and final multiple predictor models of decision-making power are reported in Table 2. Results from the independent models indicated that men earning more than their partners, men who were more educated than their partners, HIV-positive men, and non-White men all reported higher levels of decision-making power. Additionally, results indicated that age relative to one’s partner interacted with relationship length to predict decisionmaking power. No other cross-level interactions between couple-level and partner-level variables were significant (all p’s >.05). In the final multiple predictor model, education and race were no longer significant predictors of decision-making power (p’s > .05). We evaluated the tolerance and variance inflation factor (VIF) of predictors to ensure these changes in significance were not due to multicollinearity. Tolerance was greater than .10 and VIF was less than 10 for all predictors, indicating multicollinearity between these variables was not a concern in these models. All other associations remained significant, including the interaction between participant’s relative age and relationship length. Testing the simple slopes revealed that in shorter relationships, older partners reported greater decision-making power (B=.058, SE=.027, p=.03). In longer relationships, younger partners reported greater decision-making power (B= -.078, SE=.030, p=.01). Effect size calculations (standardized betas for continuous predictors) indicated these were small effects, consistent with our other demographic predictors of power (shorter relationships = .061; longer relationships = -.082). Predictors of relationship satisfaction Results of the independent and multiple predictor models predicting relationship satisfaction are reported in Table 3. In the independent models, decision-making power was a significant predictor of relationship satisfaction, such that men reporting greater decision-making

Power and Gay Male Couples 13 power relative to their partner were less satisfied in their relationships. Income, age, and education were not significantly related to relationship satisfaction in our sample (all p’s >.05). The conditional main effects of race and HIV-status were significantly associated with relationship satisfaction, but these effects are qualified by significant cross-level interactions involving couple-level composition of race and HIV-status, respectively. However, the crosslevel interaction between a partner’s relative age and relationship length was not significantly associated with relationship satisfaction. Simple slopes analysis of the race interaction demonstrated that non-White men were more satisfied in non-White/White relationships than in non-White/non-White relationships (B=3.40, SE=1.04, p.05)3. However, the interaction between individual HIV-status and couple’s HIV concordance remained significant. Testing the simple slopes revealed that HIV-negative men in concordant relationships reported greater satisfaction than HIV-positive men in concordant relationships (B= -1.75, SE=.723, p=.016). However, partners within HIV-serodiscordant relationships did not significantly differ on relationship satisfaction 3

We conducted supplemental analyses for both outcomes using racial/ethnic identity dummy coded at Level 1 (Black, Native American, Asian, Latino, Mixed Race, and Other Race, with White men as the reference group), racial/ethnic match of couples at Level 2 (dummy coded to reflect matched ethnicity vs. unmatched) and their interaction. In models including other significant correlates, race/ethnicity was not a significant predictor of decision-making power at either level, nor was the cross-level interaction (p’s>.05). However, Native American men and Asian/Pacific Islander men reported significantly lower relationship satisfaction (B= -3.54, SE=1.80, p=.05; B= -1.57, SE=.657 p