Last week, we examined Mitt Romney's contention that 47 percent of the country was "dependent on government." A look at Census data suggests that, yes, about 49 percent of Americans lived in a household last year that received at least one direct benefit from the federal government, be it Social Security or food stamps or unemployment insurance.
But political scientists Suzanne Mettler and John Sides argue that this definition of "government benefit" is far too narrow. If you include all federal benefits that go to specific households, from Social Security to even tax expenditures like the mortgage-interest deduction, then survey data from 2008 reveals that 96 percent of Americans have received assistance from the federal government at some point in their life:
What the data reveal is striking: nearly all Americans — 96 percent — have relied on the federal government to assist them. Young adults, who are not yet eligible for many policies, account for most of the remaining 4 percent.
On average, people reported that they had used five social policies at some point in their lives. An individual typically had received two direct social benefits in the form of checks, goods or services paid for by government, like Social Security or unemployment insurance. Most had also benefited from three policies in which government’s role was “submerged,” meaning that it was channeled through the tax code or private organizations, like the home mortgage-interest deduction and the tax-free status of the employer contribution to employees’ health insurance.
The contentious bit here is how one views tax expenditures. There are two ways to look at the vast buffet of credits and deductions that Congress offers through the tax code. One is that these are essentially tax cuts. People making mortgage interest payments get to keep a bigger chunk of their paychecks than they otherwise would. A firm that offers its employees health insurance pays less in taxes than it would if it spent that money on extra wages.
The other way to look at these credits and deductions is that they’re essentially government spending programs in disguise. After all, if these deductions didn’t exist, then either the deficit would be smaller or everyone else could pay fewer taxes. The mortgage-interest deduction is essentially a subsidy from renters to homeowners. A tax credit that subsidizes the construction of affordable housing is no different than an explicit grant to do the same thing. And so on.
Suzanne Mettler's research on "the submerged state" takes the latter view—tax expenditures are no less government programs than Social Security or unemployment insurance. True, they're handled by the IRS rather than other government agencies, but they have the exact same effect on the budget.
In a 2010 paper, Mettler created this table showing that many Americans don't even realize they are receiving these benefits. Below is the percentage of beneficiaries of various programs who say they have never benefited from a government program.* For instance, 43 percent of those who have received unemployment insurance say they have never used a government program:
Mettler and Sides also note that "submerged policies" like the mortgage-interest deduction or student loans are more likely to benefit higher-income families: "Low-income people have used more of the direct policies than have the affluent: the average household with income under $10,000 per year used four of them, compared to only one by the households at $150,000 and above. But the proportions were reversed in the case of the submerged policies: wealthy families had typically used three of them, and the poor just one."
Of course, you could also broaden the definition of "government benefit" much, much further to include everything from highways to farm subsidies (which provide us with cheaper food) to food-safety regulations to government-granted patents (which, in theory, promote innovation and benefit patent-holders). Both Michael Grunwald and Matthew Yglesias have taken that view. But even setting that aside, it's striking to note that the number of people who receives government benefits varies quite drastically depending on whether or not we include tax expenditures.
* Correction: I originally described this chart incorrectly. It shows the portion of beneficiaries of various programs who say they have never used a government program. It does not show the total number of Americans who have used each program. Apologies for the error.