Pretty much the most interesting blog on the Internet.— Prof. Steven Landsburg

Once you get past the title, and the subtitle, and the equations, and the foreign quotes, and the computer code, and the various hapax legomena, a solid 50% English content!—The Proprietor

Thursday, May 30, 2013

The Tax Expenditure Which Dares Not Speak Its Name

One topic of frequent exhortation on taxes among progressive policy wonkeries, such as the Tax Policy Center, are tax expenditures, or internal revenue code provisions which reduce tax revenues below what they would otherwise be.

According to the Tax Policy Center, the largest tax expenditures and related tax revenue loss in 2008 were:

Even nominally less partisan institutions, such as the Congressional Budget Office have come up with more recent (and even larger) figures for tax expenditures.

The argument which generally follows this exposition is that these tax expenditures are the equivalent of government spending, but that the beneficiaries are disproportionately the middle and upper classes (as lower income tax payers generally do not directly pay any federal income taxes in any case). This in turn sparks outrage at "welfare for the rich" and calls for either abolition of these tax expenditures or, if that should prove infeasible, higher taxes for the well-off and more government programs for the less well-off.

A principled objection to this argument is obvious: A tax expenditure is the government's forgone revenue (i.e., allowing income earners to keep more of what is their own), not a government handout; that to treat foregone revenue the same as a handout is implicitly to assume that the government is entitled to 100% of all income and any share that the earner gets to keep is effectively an expression of the government's generosity.

I will not pursue this objection here. Let it suffice to say that as emotionally appealing it is, it is intellectually not entirely satisfactory. Any government expenditure program could readily be reconstructed as a tax credit to the beneficiary (or, if necessary, a third party) without materially affecting its substance. (See, for example, the current vogue of state tax credit for non-governmental schools.) Thus a distinction between advances of liberty (such as a lowering of tax rates) and retreats (such as increases in most government programs) must turn on factors other than where they show up in the government's ledgers.1

Rather, I propose here to take the logic that tax expenditure is indistinguishable from other government expenditure for granted and point to a glaring omission in the lists of tax expenditures pushed by progressives: The preferential income tax rate for low income (or non-high income) individuals.

On substantial incomes, the IRS exacts an income tax rate of 39.6% (in 2013, not considering effectively higher marginal rates due to phase outs of various provisions, FICA taxes, etc.). However, due to loopholes inserted in the tax codes by the various politicians who have tampered with it since its inception, many lower income individuals receive preferential tax rates for parts or all of their income of 35%, 33%, 28%, 25%, 15%, 10% or even 0%. As a consequence of this tax expenditure to preferred tax payers, the fisc loses an enormous amount of revenue.

How much? Strangely enough, while the tables of tax expenditures pushed by progressive advocates cited above do list the tax expenditures from such rate preferences as those for capital gains or dividends, there is no listing for the tax expenditures relating to the rate preference for moderate and low income individuals.

So this writer resorted to self-help. Using the IRS published data for the most recent available year (2010, Table 1), one can obtain the individual incomes subject to each marginal rate and resulting revenue. It is straightforward to calculate how much higher the IRS revenues would have been had all these incomes been subject to the then-highest income tax rate (35%). The total tax expenditure on the preferential rate for low and moderate income earners is over $237 billion (or 1.8% of 2010 GDP).2

This renders this forgotten tax expenditure, in absolute terms, either the largest tax expenditure (comparing to the 2008 TPC data) or the second largest tax expenditure (after the health care exclusion, comparing 2013 CBO data). As a share of GDP, it is the largest tax expenditure by either comparison year.

One may honestly reject the idea of tax expenditure in its entirety. Or one may honestly account for tax expenditures like other parts of the federal budget. But one cannot, as the TPC and CBO do, treat the concept of tax expenditure as valid and then blithely ignore the largest tax expenditure because one approves of its distributional impacts. To do so is hackery disguised as objective analysis.

Update: Great minds think alike or, less likely, the WSJ Best of the Web reads this blog and followed it with this lead item the following day.

1 My thoughts on this issue are tentative but it seems that the crucial factor to consider is how distortionary the governmental policy is (i.e., how far it causes individuals' actions to diverge from the absence of governmental policy). However that is an issue for another post.

2 I excluded the relatively small amount of Form 8514 income for which no tax rates were provided. Also, the IRS tables list no marginal income for the 0% rate, which is surely an understatement. Both of these errors tend to underestimate the calculated size of the tax expenditure.

I also made a static calculation, assuming no behavioral effect of the changed tax rates. This is the same methodology the CBO used to estimate the size of its tax expenditures.