Skip to main content

Commentary: Policymakers Often Overstate Marginal Tax Rates — And Understate Trade-Offs In Reducing Them

Some Washington policymakers are increasingly focused on whether government benefits for low- and moderate-income people create disincentives to work — in particular, when these benefits phase down as the earnings of beneficiaries rise.

That phase-down rate is often called the “marginal tax rate” because it resembles a tax — benefits fall as earnings rise.  The relationship between marginal tax rates and disincentives to work is an important issue, one worthy of serious debate.  Some policymakers, however, often overstate the size of marginal tax rates and their impacts on work, and understate the trade-offs in trying to lower these rates.

Marginal Tax Rates Are Often Misunderstood

Economists use the term “marginal tax rate” to describe the extent to which a modest increase in income generates higher taxes and/or lower government benefits — “marginal” refers to the effect of the next dollar of income.  Analysts generally examine these benefit phase-outs together with income and payroll taxes.  If, for instance, a worker faces a marginal tax rate of 30 percent, that means if she earns another $100, she will lose $30 of it through lower benefits and/or higher taxes.  High marginal tax rates cause concern in large part because they could prompt workers to work less or choose not to look for higher paying jobs.

The report on the safety net that House Budget Committee Chairman Paul Ryan issued in March (“The War on Poverty 50 Years Later”) offers an example of how policymakers can oversimplify or mischaracterize the marginal tax rate issue.  It argues that marginal tax rates for low-income people are “extraordinarily high” — and often create a “poverty trap” in which families have little or no incentive to increase their earnings.

The report, however, substantially overstates the marginal tax rates that most low-income people face (as discussed in the next section).  It also glosses over the leading academic research, which suggests that even with the marginal tax rate issue, the effects of the safety net on work are modest and the overall effect of safety net programs is to keep millions of people out of poverty and make poverty less severe for millions of others (as discussed in a later section).

Rates Vary Widely, Most Often Are Low or Moderate

Low-income families don’t all face high marginal tax rates; the marginal rates they face vary significantly based on such factors as their income, the number of people in the family, and the benefit programs in which they participate.  In general, the poorest families — those who are out of work or have earnings that leave them well below the poverty line — face the lowest marginal tax rates, while families that have incomes somewhat above the poverty line and receive more than one benefit that is phasing down face higher marginal rates.
Johns Hopkins University’s Robert A. Moffitt, a leading expert in this field, examined marginal tax rates for low-income families in the Supplemental Nutrition Assistance Program (SNAP, formerly called food stamps) and found that high rates are “a significant problem only for a small portion of the SNAP caseload.”  Researchers Stephen Holt and Jennifer Romich examined tax and administrative records for about 2.5 million Wisconsin residents over the 2000-2002 period and found that only 11 percent of poor single parents with children — and fewer than 10 percent of those with incomes between the poverty line and 250 percent of the poverty line — faced marginal tax rates of more than 50 percent.  (By contrast, the Ryan report highlights rates that range from 55 percent to “nearly 100 percent.”)

Though a frequently overlooked fact, many families with little or no earnings actually face the opposite situation — their marginal rate is “negative,” meaning the family is rewarded for increasing its earnings by being provided somewhat more in benefits, so that its overall income rises by morethan a dollar for each additional dollar of earnings from work.  That’s because the Earned Income Tax Credit (EITC) and Child Tax Credit (CTC) rise as earnings rise above very low levels (and because SNAP benefits do not start to phase down with the first dollar of earnings).

That, in turn, creates a strong incentive for parents to work rather than not work.  Indeed, regarding families with children that receive both SNAP and other income supports — and that, in theory, might face high marginal tax rates — Moffitt found, “the vast majority have earnings so low that they face negative cumulative marginal tax rates.”  Families that receive child care or housing benefits, Moffitt explains, would face higher marginal tax rates but most families do not receive all of those benefits.

Impact of High Marginal Rates on Work Should Not Be Overstated

To be sure, some low-income families, typically those with incomes modestly above the poverty line, can face marginal tax rates near or above 50 percent, particularly if they receive benefits such as housing assistance or child care assistance that only a small fraction of low-income households get.  But research suggests that, in practice, families do little to reduce their work hours or wages in response to the marginal tax rates that benefit phase-downs create.  Interviews with low-income families — by Jennifer Romich and her colleagues and by Laura Tach and Sarah Halpern-Meekin — suggest that may be in part because families do not fully understand how benefits (particularly tax credits) adjust as earnings rise and because low-wage workers may have limited control over the hours they work or their ability to find higher-paying jobs.

After exhaustively reviewing the research on how various income-tested programs affect people’s choices about work, scholars Yonatan Ben-Shalom, Robert A. Moffitt, and John Karl Scholz concluded that most benefit programs have, at most, a modest negative effect on hours worked.  Overall, they found, the programs’ work disincentives are small enough to have “almost no effect” on diminishing the safety net’s success in reducing poverty.  Even after accounting for the impact of marginal tax rates on employment, they concluded, the safety net was responsible for reducing the poverty rate from nearly 29 percent to 13.5 percent in 2004 (the year they examined) — equivalent to lifting well over 40 million people out of poverty.

No Easy Route to Lower Rates

Still, policymakers across the ideological spectrum share concerns about marginal tax rates and agree that, all else being equal, lower marginal tax rates are preferable to higher ones.  Unfortunately, all else is not equal, and lowering marginal tax rates entails significant and very challenging policy trade-offs.

For any given income-tested program, policymakers essentially have three ways to reduce its marginal tax rate: (1) allow benefits to phase down more slowly and extend higher up the income scale, which adds significantly to program costs; (2) significantly scale back the assistance that poorer families receive, thereby increasing the extent and depth of poverty, so benefits can phase down more gradually (and possibly extend to families higher up the income scale) without increasing costs; or (3) eliminate assistance for needy individuals and families altogether.

Thus, marginal tax rates are the product not of bad policy design but rather of competing policy goals:  providing needed assistance to financially struggling individuals and families and limiting costs by not providing help to those with more adequate income.  Any serious discussion of the marginal tax rate issue must grapple with the fundamental tension between limiting assistance, controlling costs, and reducing marginal tax rates.

No such serious discussion is likely to result, however, from exaggeration of the marginal tax rates that most low-income families face, overstatement of the impact those marginal rates have on actual work behavior by low-income households, or glossing over the tough policy trade-offs that policymakers must face when seeking to reduce marginal rates.