Abstract

This article analyzes the changes in family structure, fertility behavior, and the division of labor within the household from the Silent generation (cohort born in 1940-49) to the Millennial generation (cohort born in 1980-89). Using data from the Panel Study of Income Dynamics, this article documents the main trends and life-cycle profiles for each generation. The main findings are that (i) the wage-age profile has been shifting down over generations, especially for Millennial men; (ii) the returns to a four-year college degree or higher for men have increased for all generations; (iii) Millennials enjoy a higher level of leisure than previous generations; (iv) the housework hours for women have clearly declined over generations, while the housework hours for Millennial men are higher than those of the previous generations of men; (iv) less-educated individuals have retreated from marriage, especially Millennials, while more-educated individuals have delayed marriage; (v) divorce rates have risen, with Millennials most likely to divorce, but the longer a couple is married, the likelihood of divorce has decreased over generations; and (vi) the Millennials' completed fertility rate is likely to be the lowest among all generations. 

George-Levi Gayle is a professor of economics at Washington University in St. Louis and a research fellow at the Federal Reserve Bank of St. Louis. Mariana Odio-Zuniga and Prasanthi Ramakrishnan are PhD candidates at Washington University in St. Louis. The authors thank Limor Golan and Robert Pollak for helpful discussions.

INTRODUCTION

Over the past half-century, some of the most striking socioeconomic changes in developed countries have been the radical changes in the family structure, fertility behavior, and the division of labor within the household. These changes have consequences for labor market productivity and the viability of Social Security and other programs. For example, the decline in fertility below the replacement rate has led to a major concern for pay-as-you-go social security in Organisation for Economic Co-operation and Development countries, including the United States. Moreover, the significant increase in the percentage of women in the workforce has had a considerable effect on the type of benefits employers offer, specifically regarding a family-friendly workplace, parental leave, and other work-family balance policies.

Several articles analyze the changes in the family structure, fertility behavior, and the division of labor within the household for older generations (see Eckstein, Keane, and Lifshitz, 2019; Kong, Ravikumar and Vandenbroucke, 2018; and Ramey and Francis, 2009); however, to the best of our knowledge, no study has analyzed these changes in the latest generation—Millennials—and how their family, fertility, and labor market behavior compares with that of the previous generations. The scarcity of studies analyzing the behavior of Millennials is mainly because of a lack of data. Using data from the Panel Study of Income Dynamics (PSID) from 1968 to 2015, this article presents the first read on the behavior of Millennials as they complete their education, form their families, and transition into adulthood.

This article focuses on three key aspects: work, leisure, and family. For each of these, trends over time as well as life-cycle profiles over generations are presented. In addition to this, an Oaxaca-Blinder decomposition of wages into explained and unexplained components is estimated to understand the changes in the gender wage gap. Given the trends in education and the recent convergence in the wage gap, the returns to the labor market are also estimated and the changes are analyzed over generations by race and gender. Along with the changes in education and hours worked, there has been a major change in housework hours as well. Using a linear regression framework, as a first cut, the differential effects of education, race, gender, marital status, and the presence of children on housework hours are estimated. Lastly, with declining marriage rates, the key question of what predicts partner choice is analyzed using a multinomial logit model.

In the article, we assign the generations the following names and birth cohorts:

• Silent generation: 1940-49
• Baby Boomers-1 (Boomers-1): 1950-59
• Baby Boomers-2 (Boomers-2): 1960-69
• Generation X (GenX): 1970-79
• Millennials: 1980-89

Focusing on labor markets and education decisions, we find that the wage-age profile has been shifting down over generations, with Millennial men having the lowest real wages over the life cycle (up to age 33) so far. However, women's wages have not decreased but have stagnated, with Millennial women earning lower wages than GenX for the part of the life cycle that can be analyzed (up to age 33). This finding indicates that most of the documented rise in wage inequality has come from men and not women. For both Black men and White men, returns to college graduates (those with a four-year degree or higher) have increased for all generations, with the most significant increase for Millennials, specifically Black men.

Despite the increasing returns to college for men, Black women have always graduated from college at higher rates than Black men; and this gap has only increased over the birth cohorts. On the other hand, up to Boomers-1, White men graduated from college at higher rates than White women; this gap reversed in Boomers-2 and has continued to increase over generations (see Goldin, Katz, and Kuziemko, 2006; Murnane, 2013; Blau and Kahn, 2017; and Eckstein, Keane, and Lifshitz, 2019), with Millennials having the biggest gap between White men and White women. While the college graduation gap between White men and White women is more than the college graduation gap between Black men and Black women, this outcome reverses if individuals with some college and college graduates are combined.

There has been a significant increase in college graduation rates for Boomers-2 and GenX for both Blacks and Whites, a pattern that has continued to accelerate for White Millennials. This is in sharp contrast to the graduation rates for the Silent and Boomers-1 generations, which were generally stable across the two. Another striking feature of the change in the education distribution over the generations is the significant reduction in high school dropout rates; this has been most pronounced for Blacks. The percentage of Black high school dropouts fell from 27.5 percent for the Silent generation to 7.1 percent for the Millennial generation.

A few studies in the existing literature analyze leisure over generations and the life cycle. Three notable exceptions are Aguiar and Hurst (2007); Ramey and Francis (2009); and Aguiar, Hurst, and Karabarbounis (2012). This article finds that over generations, while the amount of leisure enjoyed by women has been increasing, the amount of leisure enjoyed by men shows no clear pattern. However, for both men and women, Millennials enjoy a higher level of leisure than previous generations. This rise in leisure for women is primarily coming from married women. Splitting the sample by education, this article finds that women who are college graduates enjoy less leisure than those who are high school graduates, and the same is true for men. The finding that the amount of leisure enjoyed by men is stable over generations masks different dynamics in the components of leisure. There has been a significant reduction in hours worked by men, with Millennials working the least. At the same time, there has been an increase in the housework hours for men, with Millennial men devoting more hours to housework than men of any previous generation. The opposite is true for women: Their reduction in hours devoted to housework over generations has been more than offset by the increase in hours devoted to market work. More importantly, most of this movement has happened within married couples.

The findings of this article are in contrast to Aguiar and Hurst's (2007), who find that leisure is increasing for everyone. Several factors may account for these conflicting results. First, different data sources are used in this article. Second, this article compares leisure over generations and the life cycle, while they do not. Third, the measure of leisure used here may not coincide with the measure of leisure they use, as they are using self-reported data from time diaries, whereas the measure in this article is the residual of hours worked in the market and in household production. Because of these differences, the measure of leisure used in this article is more comparable with the measure used in Ramey and Francis (2009), and results here confirm the life-cycle analysis reported in that article.

The media has often conjectured that the marriage, cohabitation, divorce, and fertility behaviors of Millennials are radically different from previous generations. A detailed analysis of these behaviors is provided, which has been lacking so far. Marriage (and, thus, fertility) has seen many drastic changes over generations. Lower-educated individuals have retreated from marriage, and this finding is the most pronounced among Millennials. However, while Millennial college graduates are delaying marriage significantly, they do catch up with previous generations by ages 30 to 33. Thus, it is not clear whether Millennials would be marrying at an overall rate lower than previous generations, as the composition of education has also changed, with a much higher number of college graduates among Millennials. No clear trends in assortative matching are found over generations. Although this is in contrast to the findings of Greenwood and Guner (2008) and Santos and Weiss (2016), it is in line with those of Gihleb and Lang (2016). However, Millennials are cohabiting at a much higher rate than previous generations. For GenX and Millennials, cohabitation rates are much higher early on in life relative to previous generations; however, the rates drop off significantly after that, indicating cohabitation itself is transitory. This was not true for the Silent and Boomer generations. With respect to divorce, the Silent generation has a fairly flat profile over the number of years since marriage. The subsequent generations present rising divorce rates and more pronounced duration dependence; that is, as the length of the marriage increases, the likelihood of divorce decreases, with Millennials having the highest probability of divorce after five years of marriage. The major changes across generations in divorce rates are primarily coming from Blacks and couples where at least one spouse has completed education of less than or equal to a high school diploma. Couples where both spouses have completed some college or more have significantly lower divorce rates, as also documented by Lundberg, Pollak, and Stearns (2016).

Finally, concerning fertility trends, the significant decline in completed fertility is confirmed in this article, with Boomers-2 and GenX fertility rates falling below the replacement rate. Similar trends exist across race. There has been a steady decline across generations at each age parity of the proportion of births to married women, although this decline has accelerated among Millennials. For example, the proportion of births to married women ages 31 to 35 for the Silent generation is 88 percent, which fell to 83 percent for GenX and even further to 75 percent for Millennials. The age-specific fertility rate has declined for every generation up to GenX. Fertility rates for Millennial women ages 18 to 30 are below every previous generation's; however, for Millennial women ages 31 to 35, the fertility rate is higher than for GenX women of the same ages. This finding suggests that the completed fertility rate for Millennials may not necessarily be below that for GenX; although Millennials are delaying fertility, their fertility rates at later ages might be higher than those of the immediate previous generation.

Read the full article.