The debate about former Governor Mitt Romney’s (R-MA) tax plan has reached a fevered pitch. In fact, The Washington Post published two articles (here and here) about it. The second repeated the confusion surrounding a poorly understood policy known as “step-up.”

Recently, I weighed in to correct the false statement that Governor Romney’s plan “mathematically necessitates” a tax increase on the middle class. That falsehood has been perpetuated by a misleading report from the Tax Policy Center (TPC). The report is misleading in part because the authors of the study made a material error in their analysis in regard to step-up.

Step-up arises because of the interaction of two taxes—the tax on capital gains and the death tax—and the need to avoid yet another instance of double taxation. When someone dies, an estate is created. Typically, the estate contains appreciated assets on which capital gains tax would have been owed had they been sold prior to the decedent’s passing.

The death tax levies tax on the total value of the estate less an exemption amount. In contrast, the capital gains tax is levied on the increase in value of the asset, which is the difference between the current value and the purchase price, or “basis.”

Taxing the capital gain when an inheritor sells an asset using the basis of the original owner would be taxing the same wealth twice. Thus, after death tax is paid, the basis of the asset is “stepped up” to the current value so the recipient of the asset will owe capital gains tax only on appreciation occurring after receipt.

On the other hand, if the death tax were repealed, then step-up basis would be inappropriate as long as the capital gains tax remains. The appropriate policy is for the new owner to retain the basis of the original owner and thus receive the asset along with its accrued capital gain. This is called “carry-over,” as the tax basis carries over from one owner to the next.

With this as background, consider now the second Washington Post article:

Dubay’s estimates for exactly how much revenue it raises are flawed. His numbers compare the current “step basis” [sic] policy not to carryover basis, but to a policy in which all assets are taxed at capital gains rate when someone dies. That’s a much more dramatic change than carryover basis, and likely to raise much more revenue. So Dubay’s revenue estimates are considerably too high. [Emphasis added.]

This is factually incorrect. My estimate explicitly involves replacing the step-up in basis with carry-over basis. The Post does not make the important point that step-up is valid with the death tax in place and that carry-over is the proper system for the capital gains tax after the death tax is repealed. (The Romney plan repeals the death tax.)

The Washington Post says I overestimate how much repealing step-up basis could raise. According to the Office of Management and Budget (OMB), in 2013 alone, step-up reduces capital gains tax revenue by $24 billion.

That $24 billion is what OMB determined the IRS would collect if the step-up basis is repealed with the death tax still in place. This estimate accounts for the revenue that would be raised under the capital gains tax if the step-up basis is repealed and inheritors of assets pay the capital gains tax with a carry-over basis. The OMB estimate of the step-up remains the same whether the death tax is in place or not.

In my paper, I calculated that repealing step-up would raise $19 billion in one year, because some inherited assets would go to taxpayers with incomes less than $200,000. As such, I determined that the distribution of step-up would roughly follow that of long-term capital gains. (About 78 percent would therefore go to taxpayers with incomes over $200,000, according to the IRS.)

The capital gains rate under current law will be 20 percent in 2013, and Romney’s plan would keep the rate at 15 percent. That differential in rate could account for some overestimate on my part, except that the TPC report estimates how the Romney plan would apply in 2015. The growth over two years would likely make up for the difference.