Tax System Discourages Married Women From Working

The New York Times, "Economic Scene" , November 01, 2000

The single women of "Sex and the City" read the New York Times wedding announcements and mock a bride who was "until recently" employed as an account executive. Now that she's found an investment banker to marry, they scoff, she doesn't have to pretend to be interested in her career. These independent women are contemptuous of such behavior. They're looking for love, not a meal ticket. But maybe the new Mrs. Investment Banker doesn't deserve their scorn. Maybe she's just a rational "economic woman," responding to incentives.

Marriage has kicked her into the highest bracket. With federal, state and Social Security taxes, she's now losing about half of every dollar she earns -- starting with the very first one. You've got to really love your work to do it for half price.

The Census Bureau recently said that both spouses now work in 51 percent of married couples with wives of childbearing age -- the first time that number has surpassed 50 percent since the bureau started tracking it. More wives would work if our progressive income tax structure didn't sock second earners with high marginal rates.

"There is a relationship between taxes and labor-force participation," said Nada Eissa, an economist at the University of California at Berkeley and the National Bureau for Economic Research. "We think it's fairly strong. In fact, we think it's strongest for married women," who are usually the family's secondary earners and are more likely to consider not working outside the home.

Far more than men or single women, married women act like supply siders. Cut their marginal tax rates, and they get jobs. Raise their taxes, and they stay home.

By disproportionately punishing married women's work, the tax system distorts women's personal choices. And by discouraging valuable work, it lowers our overall standard of living.

This is a tax story you won't hear from politicians. The "marriage penalty" debate tends to frame the issue as one of family formation: how does the tax code affect a couple's decision to marry? But marriage isn't primarily an economic decision.

The more pertinent question is how does the tax code affect a married woman's decision to work? There are obvious political reasons not to ask that. Democrats don't want to admit that soak-the-rich taxation wallops working wives, lest they split feminists and redistributionists. And Republicans don't want to admit that cutting taxes will lead more married women to get jobs, lest they split economic libertarians and social conservatives. So everyone stays mum.

But the empirical evidence is pretty clear. Tax rates are a feminist issue. Professor Eissa has studied how the tax bite affects behavior at both ends of the income distribution. (Her papers are available here.) In a 1995 paper for the National Bureau of Economic Research, she looked at how married women responded to the Tax Reform Act of 1986. That law created a natural experiment. It chopped the top marginal tax rate to 28 percent, from 50 percent, but had only a minor impact on middle-income taxpayers. That gave her an opportunity to compare the responses of different groups of wives before and after the change.

Before it, the women in the 99th percentile of family income (the top 1 percent) on average paid about 52 cents in taxes on the first dollar they earned. In some states, a wife who earned less than $30,000 a year could have paid in taxes as much as 70 cents of the first dollar she earned.

The 1986 law flattened federal rates and, as a result, slashed the average marginal rate faced by such wives to 38 percent. The percentage of these married women who worked jumped from 46 percent to 55 percent -- a 19 percent increase -- and those who had jobs increased their hours 13 percent.

To make sure this increase was caused by the tax change rather than other trends, Professor Eissa compared it with the behavior of women in the 75th percentile, who got a much smaller tax cut. Their labor force participation and hours worked also rose, but by significantly less -- a 7 percent jump in those working and a 9 percent increase in hours worked.

Since 1986, rates have become steeper, as Washington pursued increased revenue and lost interest in supply-side incentives. "There's been less discussion of the behavioral effects of taxation," Professor Eissa said.

Policy makers have, however, tried to reform the incentives faced by poor women. And here, too, Professor Eissa finds they've managed to prejudice married women against work. The 1993 tax bill expanded the Earned-Income Tax Credit, a subsidy that goes to the working poor and is administered through the tax-filing system. The credit is intended partly to counteract the discouraging effects of Social Security payroll taxes, which hit wage earners at their first dollar.

Like the rest of the tax system, the credit's sliding scale applies to households rather than individuals. One spouse's income is enough to qualify for the subsidy, and the credit is phased out as total family income increases. So if a husband earned $11,650 in 1997, his family of four got $3,656 from the earned-income credit, assuming his wife did not work. For every dollar she earned, the credit was reduced 21 cents -- in effect, a marginal tax rate of 21 percent on top of Social Security and state taxes.

In a 1998 paper, Professor Eissa and her colleague Hillary Williamson Hoynes found that the expanded tax credit did increase married men's labor force participation. But married women were 5 percent less likely to work if their family incomes were high enough that anything they earned would cancel out some of the credit. And if they did work, the tax credit led them to reduce their hours 20 percent.

By paying one spouse to work and reducing the payments as household income rises, Professors Eissa and Hoynes wrote, the tax credit "is effectively subsidizing married mothers to stay at home." It's possible that's what policy makers intended. But it's more likely that they didn't fully consider how married women would react to what amounted to high marginal tax rates.

That's a pervasive problem throughout the tax system. Once a woman is married, it seems, the tax laws don't really want her to work.