Home | Business News | Browse by Publication | I | International Advances in Economic Research

Kids cause specialization: evidence for Becker's household division of labor hypothesis.

Article Excerpt
Abstract We examine the division of labor within households and marital matching patterns in the USA using both the Current Population Survey (CPS) and the National Longitudinal Survey of Youth (NLSY). We use Becker's theory of marriage markets by estimating household production functions and using the estimates to test for positive or negative assortive matching. We also construct match matrices, which are used to judge how well our model fits Becker's theory. We find positive assortative matching on all traits in young marriages and couples without children, and negative assortment along some traits in marriages with children. This suggests that children induce specialization whereas couples without children exploit household public goods.

Keywords Marital assignments * Division of labor * Market efficiency

JEL C51 C61 C78 * J12

Introduction

This paper examines the division of labor within households and martial matching patterns in the United States using data from both the Current Population Survey (CPS) and the National Longitudinal Survey of Youth, 1979 (NLSY79). Becker (1973, 1974, 1991) presents a model of efficient marriage markets in which marital matching is based on household production. Marriage markets are efficient if marital assignments maximize total "household output." The model predicts positive assortative matching along individual traits that are complements in household production. Importantly, Becker also predicts a division of labor within the household, resulting in negative assortative matching along individual traits that are substitutes in the household production function. While many studies have found evidence of positive assortative matching, there is little or no evidence of negative assortative matching along any traits. In contrast, we do find evidence of negative assortative matching along some traits for couples in the CPS. To our knowledge, this paper is the first to present empirical findings supportive of Becker's division of labor hypothesis.

We examine marital assignments on four dimensions: age, education, income and hours worked. To measure the desirability of each individual as a mate we use these traits to construct marriage indices. We then use these indices to construct "match matrices" to judge how well the selected traits explain matching patterns in the data. The estimated match matrices are consistent with Becker's theory of efficient marriage markets. (1) Finally, we use the match matrices to examine whether marriages in our samples exhibit hypergamy (women marrying up) or hypogamy (men marrying up). Despite popular and some academic belief in the existence of hypergamy, we find no evidence of hypergamy or hypogamy. (2)

Following Becker's model, we estimate household production functions for married couples and use the parameter estimates to test for positive or negative assortative matching. Importantly, we estimate production functions on different samples of married couples. One is a sample of new marriages, with relatively young individuals, constructed from information in the NLSY79. The other is the married couples from the March 1998 CPS. The individuals in the CPS sample are, on average, older as compared to those in the NLSY79 and, because the CPS is a random sample, represent a combination of new and on-going marriages. The results from these two samples are starkly different. Specifically, while the estimates using the NLSY79 sample are consistent with the existing research that finds positive assortative matching on all traits, the CPS estimates provide some evidence of negative assortative matching. To explore the causes of these differences, we divide our CPS sample into sub-samples of couples with and without children. The couples without children sub-sample is comparable to the NLSY79 sample in that it is made up of relatively younger people who are most likely in relatively new marriages. The estimates of the household production function for the childless sample are, in fact, consistent with the NLSY79 estimates in that there is positive assortative matching on all traits. The estimates using the sub-sample of couples with children, on the other hand, are in line with the estimates when using the full CPS sample in that we find evidence of negative assortative matching along some traits. We interpret these findings as evidence that children induce specialization, as predicted by Becker, whereas couples without children are better able to exploit household public goods (Lam 1988).

The rest of the paper is organized as follows. The next section provides a discussion of the existing research and places this study in context. "A Model of Marital Matching" presents an empirical model of marital sorting and explains implementation of the model. The data are described in "Data". "Results" presents and discusses the empirical results. In the concluding section implications of the findings are discussed.

Background

In Becker's model, utility maximizing individuals seek marriage partners. Since marriage is a voluntary arrangement, the utility from being married must exceed the utility from remaining single for each person. Utility is a function of the goods produced using the household production function. (3)...