FEBRUARY 2005

• • • • • • • • • • • • • • • • • • • • • • • • • •

For Richer or For Poorer

Working Spouses and Labor Inequality

Research by Kevin M. Murphy

Since the 1960s, have married women increased their participation in the labor force to compensate for the decline in employment and disappointing earnings growth of their husbands? Are married men working less today because their wives are working more?

In 1969, the average prime-age male in the United States was employed a little more than 95 percent of the year and earned roughly $23,000 per year (in 1982 dollars). His wife, in contrast, earned approximately $4,000 per year and was employed in the labor market only about 39 percent of the year.

By 1989, the average prime-age male worked slightly less and earned slightly more than in 1969, working about 93 percent of the year and earning $25,000. The story for the wives of these men is quite different.

By the late 1980s, wives of prime-age males worked 66 percent of the year, and they earned an average of $9,000 per year. Hence, over the 1970s and 1980s, the amount of time worked for married males declined, and their earnings increased slightly, while the time worked and earnings of their spouses roughly doubled.

At first glance, the two trends appear related. In the study "Wage Inequality and Family Labor Supply," University of Chicago Graduate School of Business professor Kevin M. Murphy and Chinhui Juhn of the University of Houston refute this argument, and uncover a very different labor market reality. The study describes earnings and employment changes for married couples in different types of households classified by the husband's hourly wage.

As the authors note, examining differing levels of men's income, rather than wages and earnings in the aggregate, tends to negate the theory that the two trends are strongly related.

"Historically, there has been an enormous increase in married women in the labor market," Murphy says. "Many people have assumed this phenomenon is the result of husbands' earnings becoming stagnant. However, we find that the largest increase in earnings has been among the wives of middle- and high-income men."

During the sample period, employment rates increased among all women, but increased the most among wives of men not from the lowest ranges, but from the middle and top wage ranges. The result was a dramatic shift in women's likelihood to work from 1969 to 1989. In 1969, the higher the husband's wage, the less likely his wife was to hold a job. By 1989, wives of men in the middle of the wage spectrum were the ones working the most.

"The very groups of men who are earning the least compared with the 1960s are men from the lowest income levels, and their wives have shown the least increase in labor participation," says Murphy. "In the case of these low-income men, their wives' increases just compensate for the shortfall in their earnings. By contrast, the earnings of the highest income men from 40 years ago have gone up, and their wives' incomes have just added to that increase."

Data on Married Couples

Murphy and Juhn used data from the Annual Demographic Files on the March Current Population Surveys (CPS) for the survey years 1968 to 1992. They singled out three years of comparable labor market activity-1969, 1979, and 1989-to study long-term changes.

Murphy and Juhn also incorporate data from the 1960 census into their study. Degrees of difference in wage and employment were stable or contracting over the 1960s, while real earnings increased sharply for all wage categories of men over this same period.

Wives of low-wage men had the fastest employment growth in the 1960s, while their husbands' earnings grew 42 percent. However, this group of women slowed their entry into the labor market in the 1970s and the 1980s as their husbands' earnings fell sharply. The biggest slowdowns in earnings growth occurred for low-wage men, while the biggest acceleration in female participation came from the wives of high-wage men.

Murphy and Juhn conclude that increased market opportunities for women-particularly highly skilled women-may have played a greater role in fueling the acceleration of female employment in the 1970s and 1980s than shortfalls in husbands' earnings.

Changes in Earnings and Employment

What role does fertility play in labor market participation?

While it has traditionally been thought that married women with children were less likely to work than those without children, here again Murphy and Juhn's findings seem to stand conventional wisdom on its head. Their results show that having children did little to slow women's gains in the labor market.

From their estimates, the authors conclude that changes in employment across all categories of women rather than changes in fertility patterns play the predominant role in accounting for the aggregate increase in female labor supply as well as the differences across these households.

"The increase in the labor market participation of women with children was as large, or larger than the increase of women without children," notes Murphy. "Change in fertility was not the biggest factor."

Having addressed these issues, the authors next examined the extent to which the fall in male employment could be linked to the rise in wives' earnings and to what extent the growth in female employment rates can be linked to the poor economic performance of their husbands.

In both cases, the answer was the same: the performance of their spouses in the labor market did not greatly impact the losses and gains of married men and women.

"The answers to these questions are pretty symmetrical," observes Murphy. "Far and away, the decline in opportunities for lower-wage men outweighed in importance the growth in their wives' earnings. The reasons women were working more had a lot more to do with greater opportunities for them."

In addition, Murphy points out, "Seventy to 80 percent of the change has to do with labor opportunities-factors directly impacting their participation behavior, not that of their spouses. Only about 20 to 30 percent of the change could be attributable to changes in spouses' earnings."

Labor Inequality

Murphy and Juhn's findings highlight the enormous rise in labor inequality between the highest earning men and the lowest earning men over the years 1969 to 1989.

"One of the dominant features of the '70s and '80s was increasing labor inequality," says Murphy. "You might expect lower income wives to have earned more than middle and higher income wives to make up for that. But we really didn't see that."

In contrast to the 1970s and 1980s, the employment of low-wage men actually increased during the 1960s in spite of the fact that earnings growth for their wives was greater in the 1960s than during the later decades. Thus, female earnings were not the only significant factor determining the pattern of male employment over these three decades. The authors note that declining wages generated by falling demand for low-skilled workers played a major role in the decline of male employment rates in the 1970s and 1980s.

While real wages of men in the top wage range increased by 15 percent over this time period, real wages of men in the bottom wage range declined by 29 percent. When both wages and employment rates are considered, the real annual earnings for men in the bottom range declined about 35 percent.

The 1970s and 1980s were a time when the relative participation rates of married women shifted enormously, with the wives of high-wage men increasing their employment rates significantly more than the wives of low-wage men. Since the slowdown in male earnings growth was greatest for low-wage men and smallest for high-wage men, these relative shifts run completely contrary to what would be expected.

Changes in women's wages and market opportunities seem to be consistent with biased growth in female employment, since wage growth of women in low-wage households slowed considerably while wage growth of women in high-wage households held steady during the 1970s and 1980s, as compared to the 1960s.

Taken together, the findings support the long-held belief that high-wage men have traditionally tended to marry high-wage women.

"That's always been true, but it used to be relatively flat," says Murphy. "In other words, because they worked fewer hours, the spouses of high-wage men used to earn as much as spouses of low-wage men."

"However, there has been a substantial increase in the labor market participation of high-wage women who are married to high-wage men," Murphy continues. "So what used to be true in wages now comes through in earnings. It's evolved into the fact that high-earning men tend to marry high-earning women."

 

 

• • • • • • • • • • • • • • • • • • • • • • • • • •

Kevin M. Murphy is George J. Stigler Professor of Economics at the University of Chicago Graduate School of Business.

>> Request a copy of the research paper featured in this article