The idea that opposites attract sounds nice, but plenty of evidence suggests otherwise. A 2010 survey by the American College of Surgeons indicates many doctors marry doctors, as dual-doctor households made up one-third of all working couples surveyed. According to the book The Two-Body Problem, released by Johns Hopkins University Press in 2004, 35% of male faculty and 40% of female faculty are paired with other faculty members. Another study finds that in recent decades in the United States, college-educated men have become more likely to marry college-educated women, and men without a college degree have tended to marry women without a college degree.
This preference for marrying people with similar earnings ability exacerbates income inequality, says Assistant Professor Alexander P. Frankel. It also affects how a government should tax couples. Frankel finds that when people tend to marry within their own income group, the best tax system is one where a couple’s incomes are taxed separately.
Economists have long studied how the tax system could be used to fight inequality. To some, an optimal tax policy redistributes income from the rich to the poor by taxing high earners more. At the same time, it makes sure not to blunt incentives to work by imposing too high a tax rate.
Previous research has found that when high- and low-earners are just as likely to pair up with each other as they are within their own group, the best tax policy would be one where a spouse’s marginal tax rate (the amount of tax paid on an additional dollar of income) would fall if a husband or wife’s income were to rise. This would induce the spouse to work at least as much as before.
When most couples consist of a main breadwinner and a second earner, the benefit of such a tax scheme is potentially large. The downside of this theoretical tax policy is that if a husband or wife’s income were to fall, the spouse’s tax rate would rise. Or if both partners’ incomes were low, their marginal tax rates would be relatively high, which could persuade them to work less.
When people with the same earnings capacity marry each other, which Frankel can adjust for in his model, the case for linking a couple’s incomes and tax rates weakens, disappearing altogether when a large enough share of the population marries people of similar incomes.
When low-earning people tend to marry each other, a policy that links their incomes and tax rates becomes expensive because it results in a high level of government subsidies. When one low earner loses some income, the spouse faces a higher tax rate and a disincentive to work. That combination makes the partners poorer and may make them eligible for a higher level of government subsidies, creating an additional cost for taxpayers.
In Frankel’s view, as couples with two low- or middle-income earners become more common, maintaining tax rates and the incentive to work, rather than providing subsidies, would be a more cost-effective way of combating income inequality. It becomes more important for the government to provide the right incentives to low-income families. And as for couples made up of two high earners, it makes sense to tax each high earner separately because if one gets a raise, the spouse already has a strong incentive to work in the form of a high income.
Frankel argues that his policy would improve upon the current situation in the US, where most couples file taxes jointly largely because of shared tax credits and deductibles. In the US, when one spouse’s income goes up, the other’s marginal tax rate increases as well. This discourages work, which isn’t optimal by any measure.
Alexander P. Frankel, “Taxation of Couples under Assortative Mating,” American Economic Journal: Economic Policy, forthcoming.