The bad habit of defining “tax reform” in terms of fairness or “closing loopholes” sidesteps the most essential task of effective tax policy – namely, to collect taxes in ways that do the least possible damage to incentives for productive effort, investment and entrepreneurship.

The Joint Committee on Taxation list of “tax expenditures” is arbitrary accounting, not economics, and tax expenditures are not necessarily “loopholes.” These estimates do not take taxpayer behavior into account and therefore do not estimate revenues that could be raised by closing the so-called loopholes (e.g., a higher tax on capital gains would shrink asset sales and revenues). Policies that make sense in terms of economic incentives can therefore be portrayed as useless tax subsidies in the purely static accounting of “tax expenditures.”

For example, a recent New York Times article by former vice presidential adviser Jared Bernstein complains that tax deferral for retirement savings is unfair because, “most savings subsidies go to households that would surely save anyway, while almost nothing goes to the households that need help to save.”

These “subsidies” for high-bracket taxpayers mainly consist of deferring rather than avoiding taxes, which only partly offsets the way savings are double-taxed. Even if higher-income households would actually save the same without 401(k) accounts (which contradicts research), they would still end up with much smaller retirement savings. Dividends and capital gains would then be repeatedly taxed, year after year, rather than being continually reinvested within a tax-deferred pension, IRA or 401(k) account.

Estimated “subsidies” from tax deferral are deceptive: Instead of having recent dividends and capital gains taxed at a 15–20 percent rate in recent years, distributions from tax-deferred accounts will later be taxed at rates up to 39.6 percent. It’s a subsidy only if you don’t live much past 70.

Bernstein presents a graph showing the top 20 percent getting a 66 percent share of these “subsidies” for pensions and defined-contribution plans while the middle fifth gets only nine percent and the poorest 20 percent just two percent. What these figures actually demonstrate is that (1) people who work full-time for many years have more income to save than those who don’t, and that (2) people who pay no income tax cannot benefit from any policy that reduces taxable income, even temporarily.

There are five times as many workers in the top 20 percent than there are in the bottom 20 percent. To exclude young singles and old retirees, Gerald Mayer examined the work experience of households headed by someone between the working ages of 22 and 62. Average work hours among the poorest 20 percent still amounted to just 1,415 hours a year in 2010, while those in the middle fifth worked 2,771 hours, and the top 20 percent worked 4060 hours.

If Bernstein’s “subsidies” were properly expressed as shares of income, rather than as shares of foregone tax revenue, the differences nearly vanish. The Congressional Budget Office (the undisclosed source of his estimates) shows tax benefits for retirement savings worth only about twice as much to the top 20 percent (2 percent of net income) as to the middle 20 percent (0.9 percent of income). Retirement savings incentives appear to be worth only 0.4 percent of income to the poorest 20 percent, since they rarely owe taxes, yet annual benefits are a poor guide to lifetime benefits. Those in low income groups while they are young commonly move up to higher tax brackets by the time they start saving for retirement.

The alleged unfairness of lower-income households not getting the same dollar tax break as couples earning more than $115,100 (the top 20 percent) could be alleviated by reducing marginal tax rates on two-earner families. But Bernstein instead suggests “closing loopholes that make it easy for wealthy individuals to exceed contribution limits to tax-preferred accounts (as was found to be the case with Mitt Romney), reducing contribution limits for high-income filers, or simple limiting the value of tax breaks for the wealthiest of filers (e.g. allowing them to deduct such contributions at 28 percent instead of 39.6 percent.” None of these schemes would add a dime to the savings of low or middle-income households, of course, and they wouldn’t work.

It is not legal — and therefore not “easy”– to exceed strict contribution limits for high-income taxpayers, and Mitt Romney certainly did not do so. What Romney did was to roll over qualified retirement plans into an IRA and then earn high compounded returns on very successful investments. Similarly, albeit on a much smaller scale, I rolled-over a lump-sum pension into an IRA in 1990 when I changed jobs, and that IRA is now 12-times larger thanks to compound interest and bold investments. Since I never contributed another dollar after 1990, tougher or lower contribution limits would have been entirely irrelevant.

Bernstein’s final proposal is from the Obama budget — “allowing taxpayers to deduct contributions at 28 percent instead of 38.6 percent.” But that too is irrelevant. Any alleged “loopholes” for retirement savings have nothing to do with itemized deductions for top-bracket taxpayers, who are not allowed to deduct contributions to an IRA. Failure to include employer contributions as taxable income is not an itemized deduction to begin with, nor is the exclusion from adjusted gross income for contributions to a Keogh retirement plan for the self-employed.

In the process of giving “tax reform” a bad name, Jared Bernstein uses a sham fairness argument to justify arbitrary and unworkable anti-affluence policies that are irrelevant to any ill-defined problems.