The Congressional Budget Office projects a federal deficit of $693 billion for the current fiscal year. Take away the tax expenditures in the above table (which add up to $971.8 billion) and in theory you get a budget surplus of $278.8 billion.

Of course, we wouldn’t want to take away all of those tax expenditures, would we? The earned income tax credit and the Social Security exclusion, for example, are targeted at people with pretty low incomes. And while the direct benefits of reduced tax rates on capital gains and dividend income flow disproportionately to the very wealthy, there are economists who argue that these lower rates stimulate investment and thus economic growth.1

It’s a lot harder to find reasoned defenses of the upper-middle-class tax breaks. In fact, the empirical evidence just keeps piling up against them.

A study of retirement saving published in the Quarterly Journal of Economics in 2013 -- using data from Denmark because it was abundant and because a 1999 reduction in retirement tax benefits for the wealthy enabled handy before-and-after comparisons -- found that “each $1 of government expenditure on subsidies increases total saving by only 1 cent.” Wealthy people who would save for retirement in any case respond to subsidies by shifting assets into tax-sheltered accounts; the less wealthy don’t respond much at all.2

Meanwhile, the National Bureau of Economic Research has just released a working paper that finds the mortgage interest deduction sorely wanting. It also is based on data from Denmark, where the deduction was partly removed in the late 1980s, and concludes that:

First, the mortgage deduction has a precisely estimated zero effect on homeownership. This holds even in the very long run. Second, the mortgage deduction has a sizeable impact on housing demand at the intensive margin, inducing homeowners to buy larger and more expensive houses. Third, the largest effect of the mortgage deduction is on household financial decisions, inducing them to increase indebtedness.

Higher education in Denmark is free for citizens of the European Union, so we won’t be getting a similar study analyzing the impact of college-savings subsidies. But the conclusion would surely be similar: The benefits flow mostly to affluent people who would be saving anyway, the impact on overall saving is modest, and the impact on college affordability is somewhere near and possibly below zero.

Why do these subsidies continue nonetheless? Mainly, it seems, because they’ve been granted to a sizable, influential population who, it is feared, will fight any effort to take them away. There are other interested parties, too -- the real estate industry and mortgage lenders in the case of the mortgage interest deduction, asset managers in the case of retirement and college-savings subsidies, and of course colleges and universities in the case of the latter. But mainly it’s the millions of upper-middle-class Americans who, like me and my family, are beneficiaries of tax subsidies for home mortgages, retirement accounts and/or college savings. (The health-care and state-and-local tax deductions figure significantly in our finances, too, but I’m not going to get into those here because I think it would overcomplicate things.)