The winnowing hypothesis posits that transitions from dating to cohabiting to marital unions are marked by increasing selectivity in the mate. Keywords religion, cohabitation, dating, sanctification, and homogamy Homogamy among dating, cohabiting, and married couples. Sociological Quarterly, We argue that assortative mating between cohabiting and married couples tend to .. (from dating, cohabiting and married couples) leads to higher homogamy.
In this regard, we aim to observe if patterns of assortative mating among cohabiting and married couples tend to be similar once their social contexts are similar. By social context we refer to the physical and social setting in which cohabitation or marriage happens or develops. Historically, these contexts were socially and spatially segregated, meaning that cohabitation was mostly found among lower social strata and in specific regions.
Nevertheless, beginning in the s, countries and regions with low or moderate levels of cohabitation witnessed rises in cohabitation. In addition, cohabitation has been rising over time among men and women in all educational groups, and sometimes more among the high educated Esteve et.
Further, recent research shows that the considerable drop in proportions married prior to age 30 is mainly due to the substitution of marriage by cohabitation Esteve et. Why we expect a homogamy gap? Existing hypotheses Education is an important structuring dimension of modern marriage markets.
Pathways to Educational Homogamy in Marital and Cohabiting Unions
Individuals tend to marry or partner within their own same educational groups, and this pattern is more clearly observed at both ends of the educational hierarchy. Societies that have experienced strong educational expansion processes are more likely to present higher levels of homogamy. Recent research has identified three main hypotheses: The looser bonds hypothesis offers the view that cohabitation is a distinct institutional form of union with its own norms, goals, and behaviors.
Cohabitation differs from marriage because it is associated with a weaker sense of commitment and greater personal autonomy. Under this premise, a suitable match is less relevant in short-term relationships, such as cohabitation, than in long-term relationships, such as marriage.
Schwartz provides additional insight on the specific mechanisms by which cohabiting unions are less likely to be educationally homogamous than married couples. Although this conclusion was consistent with the winnowing hypothesis, Schwartz's results did not support the assumption that the partner selection practices of cohabiters and married people differed.
Rather, no statistical differences in partner choices were found at the time that cohabiting and marital unions were formed. Kiernan suggests that the differences between marriage and cohabitation may change as the degree of institutionalization of cohabitation shifts in society. Measures of institutionalization range from raw percentages of cohabiting unions to more nuanced indicators based on the social acceptance of this type of union.
The educational homogamy gap between married and cohabiting couples in Latin America
In societies with intermediate levels of institutionalization, many couples regard non-marital cohabitation as a trial period before marriage.
In such a context, cohabiting unions are more likely to be less homogamous than married couples with regard to education. Finally, in those contexts in which informal cohabitation is widely accepted, educational homogamy patterns for both types of unions will tend to converge. Hamplova and Le Bourdais found no support for this hypothesis when comparing differences between the Quebec province where cohabitation is widespread and the other Canada provinces where it is less widespread.
A later study conducted by Hamplova found partial support for the institutionalization hypothesis when examining differences across European countries Hamplova, The homogamy gap was lower in those countries in which cohabitation was more widespread, but the gap was not observed in the expected direction: Hypotheses for Latin America If a homogamy gap exists in Latin America, will the strength of educational homogamy be higher among cohabiting couples than among married couples?
We hypothesize that cohabiting unions will be less homogamous than marriages. Blackwell and Lichter's winnowing hypothesissuggests that an increased selectivity in the choices of partners based on their levels of commitment from dating, cohabiting and married couples leads to higher homogamy levels among married couples. However, we argue that cohabiting couples will be less likely to be homogamous in the Latin American context because the structuring role of education is less significant in cohabiting partner markets than in marriage markets.
In this hypothesis, the argument developed by Schwartz to explain the lack of significant differences in homogamy between married and cohabiting couples in the U. In Latin America, the context of opportunities in the cohabiting partner market is less structured by education than in the marriage market.
Given that cohabiting couples were historically more likely to be found in the lower social classes, among less educated people and in indigenous populations, education has theoretically had less influence on partner choices. Conversely, marriages are distributed across the educational spectrum and, as result, are more selective with regard to education.
As cohabitation spreads into higher social strata and escapes its traditional boundaries, opportunities in partner markets will become similar.
In other words, given the low expansion of education in the contexts where the institutionalization of marriage was not present, it is to be expected that education played a less important role in partner selection among cohabiting couples.
Alternatively, other dimensions where important such as ethnicity, race or class origins. How will the homogamy gap vary with regard to various levels of cohabitation? A straightforward application of the institutionalization hypothesis in Latin American countries should find a smaller homogamy gap in areas in which cohabitation is more widespread. When this proportion is larger, the homogamy gap should be smaller. This relationship should be observed both over time and across countries.
Nevertheless, a nuanced view of this hypothesis should also consider the differences between countries regarding the importance and historical roots of unmarried cohabitation in the region as described in Section 2.
Rather, unmarried cohabitation has long been socially accepted but confined to certain subgroups of the population and to particular regions within Latin America Castro, ; De Vos, Thus, we hypothesize that the cross-country correlation between the homogamy gap and the spread of cohabitation will be weaker than the over-time correlation.
This hypothesis is based on our observation in Section 2 that the initial level of cohabiting unions as measured by census round microdata does not correspond to different stages of development. As a follow-up hypothesis, we predict that the association between the homogamy gap in assortative mating and the spread of cohabitation will be stronger in those countries where cohabitation was less widespread in the s.
This association will be stronger because cohabitation in these countries should be less mixed between traditional and modern forms than in countries with stronger traditions of cohabitation.Being Married Vs. Dating
In countries such as Argentina or Chile, we should expect a close correlation between the homogamy gap and the spread of cohabitation that is in line with the institutionalization hypothesis. All in all, we expect that the patterns of assortative mating will be increasingly similar as the contexts in which cohabitation and marriage occurs also become increasingly similar.
The empirical evidence also varies widely. Using data from the late s and early s, one study found that cohabitors were more likely to be educationally homogamous than married couples Schoen and Weinick ; another study found the reverse pattern Blackwell and Lichterand still others found no difference Jepsen and Jepsen ; Qian or that the results vary by educational level Blackwell and Lichter A limitation of previous research is that it has often compared the educational resemblance of cohabitors and married couples by using cross-sectional data Blackwell and Lichter ; Jepsen and Jepsen ; Spanierwhich do not allow researchers to identify the mechanisms through which differences in couple resemblance arise.
There is also a growing body of literature on homogamy as couples transition into and out of cohabitation and marriage Blackwell and Lichter ; Goldstein and Harknett ; Sassler and McNallybut none of these studies has systematically examined how these transitions work together to affect overall differences in homogamy by union type. In this article, I use log-linear models and data from the National Longitudinal Survey of Youth to examine how educational homogamy varies as couples move into and out of cohabitation and marriage.
In so doing, this article makes three main contributions. First, by using a stock-and flow framework, I show how many of the disparate findings of past research are, in fact, coherent pieces of a larger process of assortative entry and exits from unions. In this respect, the article adds to literature cautioning researchers against using cross-sectional data to draw conclusions about the mechanisms that generate differences between cohabitors and married couples Kenney and McLanahan Second, the results help adjudicate between competing hypotheses about the nature of marriage and cohabitation with respect to sorting on education, and offer new insights into these processes.
The educational homogamy gap between married and cohabiting couples in Latin America
Finally, I use the findings as a basis for tentative speculation about the impact of the rise of cohabitation on trends in educational assortative mating in marriage. Studies of trends in the educational resemblance of spouses have generally found an increase in spousal resemblance since at least the s e.
Figure 1 shows the different flows into and out of cohabitation and marriage that may affect resemblance in the stock of prevailing cohabiting and marital unions boxes A and B. The stock of unions represents all cohabiting and marital unions that exist in the population at a given time; researchers that have relied on cross-sectional data have examined stocks of unions. Differences in educational homogamy in the stocks of unions can be generated in a variety of ways. First, the odds of homogamy among newly formed cohabiting and marital unions may differ transitions 1, 3, and 4.
Second, the odds of homogamy among cohabitors and married couples who exit their unions may also differ transitions 2, 3, and 5.