From LRC, here’s one myth:
It is usually added by flat-tax proponents, that eliminating such exemptions would enable the federal government to cut the current tax rate substantially.
But this view assumes, for one thing, that present deductions from the income tax are immoral subsidies or "loopholes" that should be closed for the benefit of all. A deduction or exemption is only a "loophole" if you assume that the government owns 100% of everyone's income and that allowing some of that income to remain untaxed constitutes an irritating "loophole." Allowing someone to keep some of his own income is neither a loophole nor a subsidy. Lowering the overall tax by abolishing deductions for medical care, for interest payments, or for uninsured losses, is simply lowering the taxes of one set of people (those that have little interest to pay, or medical expenses, or uninsured losses) at the expense of raising them for those who have incurred such expenses.
This analysis is correct on its terms. However, I do think that a contrary argument can be made for closing loopholes/exemptions along with rates.
I think that Rothbard fails to consider the cost of loopholes properly. Yes, it’s true that loopholes allow you to keep more of your money. However, finding loopholes or, worse still, paying a lawyer or an accountant to find loopholes for you, is a waste of time and possibly money. Thus, it could be more cost-effective to reduce tax rates while closing loopholes, since doing so would reduce the cost of tax compliance while keeping tax costs roughly the same. Thus, the same amount of your money would go to the government, but less would go to your lawyer and/or accountant, and less of your time would be wasted when paying taxes. I think, then, that overall a simpler tax code can be better than a complex tax code if the cost of taxes remains the same and the cost of compliance declines.
This is the so-called "Laffer curve," set forth by California economist Arthur Laffer. It was advanced as a means of allowing politicians to square the circle; to come out for tax cuts, keeping spending at the current level, and balance the budget all at the same time. In that way, the public would enjoy its tax cut, be happy at the balanced budget, and still receive the same level of subsidies from the government.
It is true that if tax rates are 99%, and they are cut to 95%, tax revenue will go up. But there is no reason to assume such simple connections at any other time. In fact, this relationship works much better for a local excise tax than for a national income tax. A few years ago, the government of the District of Columbia decided to procure some revenue by sharply raising the District's gasoline tax. But, then, drivers could simply nip over the border to Virginia or Maryland and fill up at a much cheaper price. D.C. gasoline tax revenues fell, and much to the chagrin and confusion of D.C. bureaucrats, they had to repeal the tax.
But this is not likely to happen with the income tax. People are not going to stop working or leave the country because of a relatively small tax hike, or do the reverse because of a tax cut.
Again, Rothbard’s analysis is true on its own terms. And once again it is short-sighted. While revenue gains are not going to be seen from tax rate decreases, assuming that the tax rate is too low to yield optimal revenue, the rest of the time nominal tax decreases should lead to revenue increases, ceteris paribus. Drastic tax cuts do generally lead to revenue increases, with few exceptions. As I’ve pointed out before on this blog, tax revenue as a percent of GDP hits a wall at a rate of roughly 20%. Tax revenue simply does not ever exceed 20% of GDP, regardless of tax rate. Revenue may be significantly below that (and historically federal tax rates have been remarkably low at the time), but revenue does not generally exceed that. Anyhow, the Laffer Curve is generally correct, and lowering nominal tax rates can lead to increases in revenue, particularly when the initial rate is quite high.