Limits on income tax deductions could raise a lot of money, as I've said -- perhaps more than $1 trillion over 10 years -- in a way that also makes the tax code much more progressive. Now come Jason Furman and Gene Sperling, two of President Barack Obama's top economic advisers, to say they would raise much less. But that's because they impose unwise restrictions on the new cap.

Furman and Sperling say their analysis "takes into account the reality of their impact on middle-class families" and then takes steps to hold families making less than $250,000 entirely harmless from tax increases. As they note, imposing this restriction eats away nearly half of the new revenue -- which is a reason that holding families under $250,000 harmless is bad policy.

Furman and Sperling warn that if we don't take that step, limiting deductions to $25,000 "would raise taxes by an average of $2,400 on 17 million households with incomes below $250,000 ($200,000 for singles)." But that number is a lot less alarming than it sounds.

First of all, it means that more than 90 million households earning less than $250,000 would face no tax increase at all. Those facing a tax increase, as the Tax Policy Center shows, would be overwhelmingly likely to have incomes of more than $75,000. And the lower a household's income, the lower its typical tax increase.

It's also worth considering who these people with very high deductions relative to their incomes -- the ones whose taxes would go up -- are. They are big winners under the current tax code, with very low effective tax rates relative to their incomes, meaning that they have unusually high ability to pay. Some of them are also people with highly variable incomes (small business owners, for example) experiencing a low-income year, which explains how they managed to accrue so many deductions relative to their incomes.

There will always be sympathetic cases: You might have high deductions and a low income because you were recently laid off from a well-paying job. But you can always identify a taxpayer treated unfairly by any given tax policy. In aggregate, a deduction cap imposes a light burden on people with moderate incomes, a moderate one on the affluent, and the largest burden on the top 1 percent.

The key "fairness" question about an itemized deduction cap is this: is it fair to ask people earning between $75,000 and $100,000 to pay an additional $376 a year in taxes (0.4 percent of income), on average, in the context of a tax increase that collects an added $84,889 (2.4 percent of income) on average from people making more than $1,000,000?

The White House's answer to this, apparently, is "no": The "middle class" (meaning the approximately 97 percent of Americans with household incomes of less than $250,000) cannot be made to pay a dime more. But this is a weird answer.

First of all, Obama has already broken his promise not to raise taxes on the "middle class." He did so when he raised cigarette taxes by 61 cents a pack less than 100 days into his first term. Various small taxes in Obamacare will also hit the middle class. Apparently, under some circumstances, some government programs are valuable enough to ask the middle class to pay for them.

A blanket ban on tax increases for families with incomes of less than $250,000 also makes Democrats' larger policy goals impossible. Whatever tax increase the president gets on the wealthy as part of the fiscal cliff negotiations isn't going to be enough to finance the entitlement state in the long term -- in a few years, we're going to need another fiscal adjustment that includes another tax increase. If "protecting the middle class" entails never raising broad-based taxes on them, then Democrats won't be able to afford to preserve Medicare, Medicaid, Social Security and Obamacare in anything similar to their current form.

But even if they admit this for the long term, Democrats point out that top rate increases are an alternative that can raise some revenue today without imposing any additional costs on the middle class, or even on affluent families earning the $150,000 or so that politicians, for whatever reason, now call "middle class." That's true, but unlike expanding the tax base, this policy increases the marginal tax rates faced by taxpayers, providing a disincentive to work and harming GDP growth.

These effects are not as dire as Republicans sometimes claim they are, and they are not so extensive as to mean that tax rate increases would lose revenue. But neither are they nonexistent, and they are a reason (glossed over by Josh Marshall, among others) to prefer that a given dollar of new revenue come from base-broadening -- which raises effective tax rates (and therefore revenue) without affecting marginal tax rates (and therefore incentives to work and produce).

Finally, I'll note another objection raised by Furman and Sperling. They note that a limitation on charitable deductions "would be expected to reduce giving by these households by at least $10 billion per year under the Congressional Budget Office’s assumptions." As such, they argue that the cap would have to be adjusted to exclude charitable deductions, leading to a revenue loss of $200 billion over 10 years.

I think Furman and Sperling are probably right about this -- the charitable deduction is sacrosanct, which is why it's excluded from the AMT. But even if charitable deductions are excluded from the cap, adopting a proposal that Peter Orszag recently put forward -- converting the charitable tax deduction to a matching credit -- would produce some additional revenues while retaining an incentive for Americans to give to charity. This is a better response to the charitable deduction problem than seeking revenue through a rate increase.

(Josh Barro is lead writer for the Ticker. E-mail him and follow him on Twitter.)

Read more breaking commentary from Bloomberg View at the Ticker.