Skip over navigation

Journal Issue: Marriage and Child Wellbeing Volume 15 Number 2 Fall 2005

The Hefty Penalty on Marriage Facing Many Households with Children
Authors: Adam Carasso C. Eugene Steuerle

How Marriage Penalties and Subsidies Arise

Lawmakers rarely intend to create marriage penalties; even subsidies are often accidental. Two conditions are necessary to cause marriage penalties and subsidies, and neither is sufficient by itself.24 The first condition is tax rates that vary based on income. The second is joint filing by married couples for benefits or taxes. Both characterize the U.S. tax code.

The effect of the first condition, variable tax rates, is often exacerbated by government transfer programs that are also based on joint filing. During the past several decades, policymakers have pursued the dual objectives of progressivity-—giving greater tax and welfare benefits to those with lower incomes-—and cost containment. As a result, programs like the earned income tax credit or food stamps restrict benefits to lower-income citizens by reducing or “phasing out” the benefits at steep rates as households earn more income (see box on page 164 for an example).

Combining the direct tax rates in the tax code and the benefit reduction rates in the transfer system can result in extremely high tax rates, as an example will illustrate.25 Suppose a single tax filer earns $18,000, placing her in the 10 percent income tax bracket, which means that she faces a marginal tax rate of 10 percent on each additional dollar earned above $18,000.26 (The average tax rate applying to all her income might be well below 10 percent because most of her income below $18,000 may not be taxable at all.)27 Suppose, further, that she has two children and is also receiving the EITC, which decreases by 21.06 cents for every dollar earned above $14,040. Her effective marginal tax rate includes this loss of benefits and amounts to the sum of the 10 percent marginal income tax rate and the 21.06 percent EITC phase-out rate, for a total rate of 31.06 percent.28 (For this example, we are ignoring many other taxes and benefit reductions, such as Social Security tax or food stamps.) Thus, for the income range over which a given benefit phases out, the effective marginal tax rate bumps up by the phase-out rate until the benefit has fallen to zero. When our single tax filer's income (or if she marries, her and her husband's combined income) exceeds about $35,000, her EITC benefit is gone and the 21.06 percent phase-out rate no longer applies, so her effective marginal rate then drops by 21.06 percentage points.

Although one may not typically think of it in this light, the loss of means-tested transfer benefits as earnings increase affects a household in much the same way as higher direct tax rates do-—both are losses of income. Indeed, economists commonly apply the term “tax rates” to transfer programs to identify how much benefit is lost (effectively taxed away) as a family's income rises. This, by the way, is not a comment on the fairness of benefit phase-outs. Some observers believe that there is no entitlement to such benefits, and therefore that benefit reductions are different on equity grounds from direct taxes, which take away what one has earned rather than what one has received as a transfer. But in terms of incentives and size of penalties the issue remains, regardless of the fairness of benefit phase-outs.

Benefits from some programs, like Medicaid and the State Children's Health Insurance Program (SCHIP), do not phase out gradually but instead fall swiftly or end altogether as soon as a household's income exceeds some dollar threshold. In these cases, receiving one more dollar of earnings can strip a household of several thousand dollars of benefits.

The effective marginal tax rate—-the rate created by steep benefit phase-out rates combined with graduated income tax rates—- moves up and down a lot as income increases, as evidenced by the example above, but it is usually highest for low- to moderate-income families. This reality runs counter to the notion that marginal rates rise progressively with income, as one would be led to believe by looking only at the statutory rate schedule in the income tax.

Note that these variable tax rates do not by themselves penalize marriage. A second, simultaneous condition is necessary to create marriage penalties and bonuses-—joint filing by married couples for taxes or benefits. Policymakers often look to the household unit, or joint tax return income, rather than to each individual's income separately to measure the need for transfer benefits or the ability to pay taxes. Their aim is to treat households with equal incomes equally, but in a system with variable rates, individuals with equal incomes will then not be treated equally. If graduated or variable tax rates were accompanied by individual filing, there would be no marriage penalties. Marriage would have no effect on any benefit received or tax paid by the individual. Alternatively, if everything were taxed at a flat rate (including zero, as in the case of a universal grant that did not phase out) there would also be no marriage penalties.