Just how many households receive government benefits?

Over at the Wall Street Journal, Sara Murray crunches the Census data and finds that 48.5 percent of U.S. households, nearly half, now receive “some type of government benefit.” That category, she notes, includes mean-tested programs like food stamps, subsidized housing, cash welfare, and Medicaid, as well as retirement programs like Medicare and Social Security.

But it’s worth noting that the chart on the right, while interesting, characterizes “government benefits” in a very specific and narrow way. It doesn’t include the vast buffet of tax breaks and deductions for things like employer-provided health care or mortgage-interest payments. As Len Burman and Marvin Phaup have shown, these sorts of tax expenditures totaled $1.2 trillion last year — dwarfing food stamps, welfare, and subsidized housing combined — and reach a much broader range of Americans (many of them in higher income brackets).

And, as Burman argued when I chatted with him last month, tax expenditures are functionally equivalent to direct spending programs. Giving a person a $1,000 tax break for employer-provided health care isn’t terribly different from handing that person a $1,000 check to buy more health insurance. But politicians often prefer to conduct social policy through tax expenditures because it can be sold as a “tax cut” rather than as a spending program.

Still, the net effect, as Suzanne Mettler has argued, is that Americans have a somewhat distorted view of the size of the U.S. welfare state. Check out her chart here showing that most people who benefit from what she calls “the submerged state” insist that they have never used a government program. And this set-up makes it easy to believe that only the bottom 48.5 percent (and rising) relies on Uncle Sam. But, in principle, there’s no reason why government benefits should be split into different tiers.