Academic journal article Journal of Social Structure

A Longitudinal Analysis of Gendered Association Patterns: Homophily and Social Distance in the General Social Survey

Academic journal article Journal of Social Structure

A Longitudinal Analysis of Gendered Association Patterns: Homophily and Social Distance in the General Social Survey

Article excerpt

Introduction

How have changing structural conditions impacted the composition of male and female ego networks? Existing scholarship argues that our social networks emerge largely from the structural constraints, opportunities, and settings that we experience in the course of our lives (e.g., Blau 1977; Feld 1981; McPherson, Smith-Lovin and Cook 2001). As a result, those who have different structural opportunities and constraints (e.g., different educational or occupational trajectories) should have networks of different compositions. These effects should extend beyond the individual level, aggregating to encompass entire demographic categories and producing characteristic patterns of association. A variety of studies have confirmed this insight, finding, for example, both that male and female networks differ in composition (Ibarra 1992; Marsden 1988), and that some of this difference can be attributed to structural inequalities (Moore 1990).

Recent decades have seen dramatic increases in the proportion of women who have achieved advanced educational credentials (Buchmann et al. 2006; Jacobs 1996) as well as success in the workplace (e.g., Carlson 1992; Cohen 2004). When coupled with burgeoning communications technologies that may help compensate for the traditional differences in male/female voluntary association participation (McPherson and Smith-Lovin 1982, 1986; Popielarz 1999), there is significant reason to believe that male and female networks should be growing more similar. This paper uses nationally representative data on ego networks to analyze differences in the association patterns of males and females, as well as to determine how association patterns have changed over a twenty-year period.

I begin with a discussion of association, evaluated both in terms of homophily, the well-known tendency for like to associate with like, and social distance, or the likelihood of association with various dissimilar categories. The study of association patterns in general is important because they reveal the nature of the underlying social structure (Blau 1977; Blau and Schwartz 1984) and changes in such patterns therefore reflect changes in the nature of that underlying structure. Additionally, association represents the day-to-day social environment that respondents navigate; to understand a person's experience of the social world, we must analyze their associations and not merely their structural location. Next, I introduce the data and correction factors employed in this analysis and describe the log-multiplicative models used to compare association patterns across years and sexes. Then, I describe my data and analytic techniques. Finally, I discuss the results and draw conclusions.

Structure, Homophily, and Social Distance

Homophily, or the tendency to associate with those like oneself (Lazarsfeld and Merton 1954), is one of the most robust of all social science findings (e.g., McPherson et al. 2001; Smith et al. 2014). This tendency does not arise primarily from choice but rather from structural availability. We select our associates from among those with whom we come into contact and because those others often live, work, and recreate in particular places for the same reasons as ourselves (e.g., their wealth or education level), our selections are made from among a pool of others who are already like us. Much as a diner is forced to select a preference from a limited menu, we must select our friends and spouses from among the limited, and comparatively homogeneous, set of people with whom we come into contact. While homophily is a nearly universal finding, the strength of homophily is often unequal between dimensions (e.g., age or education) as well as between levels of the same dimension (e.g., high school degree versus college degree) (Marsden 1988). As such, both the cross-dimension and within-dimension differences in the strength of homophily provide useful information about the underlying structural forces that guide association. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.