Current proposals from Democratic presidential hopefuls—from heavier wealth and income taxation to free college, Medicare for All, and universal government savings accounts—show that, as in the past, their upcoming campaigns will promise to reduce poverty and inequality. Also as in the past, however, their “solutions” are highly questionable. But before we even reach such issues, there is another major problem that must be reckoned with first: the income measures used to justify their proposals are unreliable.
Unreliable and Inconsistent
One of those issues is the failure to appropriately adjust for age (and a typical lifetime cycle of earnings, from low in early years, when people have fewer skills, to higher in peak earning years and back to lower in retirement, which means most income redistribution is actually from you in middle age to you when both young and old). Others include failing to appropriately adjust for the size of households and the number of workers in families and treating ever-changing groups of people in an income category as if it represented the same people over time, opening the door to false class-based arguments.
While those are very serious—and generally overlooked—issues, those I wish to discuss now illustrate the data’s unreliability more clearly in that programs to provide more resources to lower-income Americans or reduce inequality often make the poor appear poorer in official data. And it is hard to measure worse than by counting help as harm. The source of this problem is that official income and poverty data ignore in-kind transfers and taxes.
Hundreds of billions of dollars of means-tested (varying with income) in-kind benefits, which predominantly go to lower-income households, go uncounted annually. For instance, the food stamp (now SNAP) program’s benefits are omitted from income. However, as affected households earn more money, those benefits are phased out at a rate of 24 percent (after some administrative adjustments), creating an effective 24 percent income tax rate on their earnings. And that reduction in their take-home incentives leads them to earn less than they otherwise would, which is reflected in the data, making them appear poorer than before despite providing them additional resources.
Subsidies Create Disincentives
Similar effects affect housing programs. While providing valuable housing services to recipients, because they are in-kind rather than in cash, they are also omitted from official measures of income. But public housing and Section 8 subsidy programs require recipients to pay 30 cents more for the same unit for each dollar they earn, which acts like a 30 percent income tax. Just as for SNAP, that reduces earned incomes, making recipients look poorer as a result.
Such disincentives are made worse by the fact that many poverty program recipients are in multiple programs, creating multiple effective tax rates as their incomes rise. For example, someone in both SNAP and the Section 8 program would face an effective 54 percent cumulative tax rate as a result. And participating in still more programs would boost that rate further.
Medicaid creates another effective income tax, but rather than reducing benefits as more income is earned, there is an eligibility cliff once a certain income level is exceeded. Obamacare extended eligibility for children up to 133 percent of the federal poverty level (but most states cover children to higher income levels), and states were given the option to similarly extend eligibility to adults. That means that a recipient who earns more than 133 percent of the federal poverty level loses access to Medicaid for their family. Since Medicaid spent $6,641 per person covered in 2012, that amounts to a very large income tax, which is dramatically worsened when there are more children in a household.
The omission of taxes in official income data reinforces the misrepresentation. The most obvious bias introduced is to make higher-income families, who disproportionately bear tax burdens, look better off than they really are. But there is a less-noticed effect. Because taxes are ignored, that means that tax credits are also ignored in the official data. It means that tens of billions of dollars annually in Earned Income Tax Credits to lower-income families aren’t counted. However, the vast majority of recipients are in the 21 percent phaseout range of those credits, which acts like an additional 21 percent income tax rate, and consequent reductions in earnings over that range are counted.
Even increasing taxes on the rich can make the poor appear poorer (beyond the fact that higher tax rates provide those with a great deal of wealth less incentive to use it to benefit others). Increasing taxes on higher-income earners compresses the after-tax wage differentials between high- and low-income career paths, reducing the payoff to making the necessary additional investments and sacrifices. That reduces the supply of higher-income workers over time, raising their pre-tax earnings. The opposite happens for low-income workers. Because data counts only the changed market earnings, measured poverty would increase and measured incomes grow more unequal.
Dramatic mis-measures of income have long sustained efforts to justify political redistribution. And they provide convenient cover for a never-ending story supporting expanded redistribution because so many policies that provide resources for the poor actually make them appear poorer in official data. But the fact is they render such data incapable of justifying such policies. So it looks like one of 2020’s big political questions will be whether such faulty premises can fool enough voters to create even more faulty policies.
Author: Gary M. Galles* (FEE)
*Gary M. Galles is a professor of economics at Pepperdine University. His recent books include Faulty Premises, Faulty Policies (2014) and Apostle of Peace (2013). He is a member of the FEE Faculty Network.