Crosby (1982) described the "paradox of the contented female worker": women's relatively high work satisfaction coexists with relatively lower rewards when compared to men. Explanations of this paradox often point to women's use of other women for comparisons about work. Given the level of gender segregation in most organizations, it is logical that women would choose other women for comparison. Because women are rewarded less than men, using other women for comparison does not result in a perception that rewards are unjust. Extant literature specifically evokes women's use of female comparisons as an explanation of both perceptions of distributive justice and work satisfaction. Empirical testing of women's use of female referents and the consequences for work satisfaction is needed. Using a sample of 288 females from a large human services organization, this study attempts to confirm the propositions set forth in earlier studies. Approximately two-thirds of the women in this study use females exclusively when evaluating their jobs. Referents are predominantly external to the respondent's workplace. There is a surprising lack of relationship between referent sex and global distributive justice perceptions and pay distributive justice perceptions. Using a same-sex referent is directly related, however, to job satisfaction. These findings suggest that the theoretical assumptions made about women's reference process need further investigation.
Although research suggests that although women consistently receive fewer rewards at work than men (Berch 1982; Featherman and Hauser 1976; Phelan 1994), they are just as satisfied with their jobs as men are (Buchanan 2005; Crosby 1982; Graham and Welbourne 1999; Hodson 1989; Keaveny and Inderrieden 2000; McDuff2001; Mueller and Wallace 1996; Steel and Lovrich 1987). Crosby (1982:7) labeled this phenomenon "the paradox of the contented female worker."
Distributive justice is generally defined as "the fairness of the outcomes or rewards that an individual or group receives" (Younts and Mueller 2001:125). Extant literature consistently mentions the referent process as directly related to women's perceptions of fairness in rewards, and, in turn, their satisfaction with their jobs (Kim 2005; Mueller and Wallace 1996; Phelan 1994; Sweeney and McFarlin 1997). These arguments hypothesize that women are content with their rewards because they weigh their own rewards against the rewards of other women. The assumptions of the former processes make logical sense and are theoretically supported. However, there has been very limited examination of women's specific comparison choices, resulting in limited research specifying the consequences of referent choices on satisfaction and distributive justice perceptions.
This research attempts to empirically examine the theoretical hypotheses and assumptions alleged by former studies by focusing specifically on the female referent process. First, I undertake an exploration of women's referent choices to determine the referents women choose in terms of sex, relationship to respondent, whether they work with the respondent, and the referent's status relative to the respondent.
Experiments use predetermined comparisons by design and investigate selfreported pay differences (Major and Forcey 1985; Major and Konar 1984). Another approach to indicating sex of the comparison other has been to imply the process by examining women's attitudes in sex-typed industries or at particular levels of gender segregation (Kirn 2000; Moore 1991). If sex-typed industries are related to satisfaction, the causal connection to sex of referent is assumed by proxy. This approach does not determine with whom specifically women are comparing themselves. Given the prevalence of referent sex in the literature relating to gender and equity, it is surprising how little this process has been specifically investigated.
Next, I examine the consequence of women's choice of same-sex referents. …