Felix Salmon

Last week’s Munk debate featured one of those strange-bedfellow moments, when Paul Krugman agreed with Art Laffer that the tax rate on capital gains should be the same as the tax rate on income. (In fact, Laffer went one step further than that, saying that even unrealized capital gains should be taxed at the same rate.) Normalizing the capital-gains tax rate so that it’s the same as the income-tax rate is an easy way to bring a lot of money into the public fisc — some $161 billion per year, according to the CBO. So why aren’t we doing it?

Evan Soltas does his best to answer that question with his “Defense of the Capital-Gains Loophole”. Here’s the meat of his argument:

Most tax breaks create distortions. The tax break for capital gains does the opposite: It reduces a distortion. Investment is really deferred consumption. Taxing consumption tomorrow at a higher rate than consumption today — which is what a tax on investment income does — encourages people to shift consumption forward in time, and that’s inefficient.

This doesn’t make a lot of sense. Firstly, investment really isn’t deferred consumption. The amount of money invested, in the world, is going up over the long term, not down — which means that once you look past the natural tidal movements of money in and out of various investment vehicles, it’s reasonable to say that money, once it gets invested, stays invested. Pretty much forever. The amount of money being saved, plus the amount that the investments have grown, is nearly always going to be greater, in aggregate, than the amount of money being withdrawn for the purposes of consumption. That’s the inefficient thing: money that could be cycling through the economy at high velocity is instead tied up in investment vehicles, and might not be spent for decades, if at all.

Soltas comes up with an example to show that if I invest my money today and then spend it in ten years’ time, then I’m going to end up being taxed at 50% — a higher rate than the 40% income tax. This example is a subset of the annoying dual-taxation meme. But in any case it ignores the much bigger amount of dual taxation which goes on with regard to spent money.

If I earn money and spend it today, my spending is going to become someone else’s income. If that person then pays tax on that income and spends the remainder, we’ll get yet another round of income tax out of it. And so on and so forth. It’s a constant high-velocity money-go-round, which is driving tax revenues all the way. By contrast, if my money is tied up in savings for a decade, it’s not generating any tax revenues at all. As a result, saved money generates much lower tax revenues than spent money. At the very least, then, it should be taxed at the same rate as spent money.

That said, savings do have an important role to play in the economy. Do we want to endanger that? Here’s Soltas again:

In theory, this is a strong disincentive for saving and investment, leading to less accumulation of capital and lower incomes over time. The empirical evidence is admittedly less impressive. Still, this reasoning explains why economists leant towards a preferential rate of capital-gains tax in a recent survey.

My theory is that economists lean towards a preferential rate of capital-gains tax for two reasons: they like theory over reality, and they tend to be rich people with capital gains income. The fact is that there’s really no empirical evidence to suggest that raising the capital gains tax to the income-tax rate would actually reduce savings; neither is there any good evidence to suggest that if savings were reduced, then incomes would trend lower over time. In order for the capital-gains loophole to be justified, we would need to be reasonably certain on both counts. We’re not even close to certain on either: my feeling, indeed, is that both are downright false.

Soltas does have a good point that capital-gains taxes become particularly onerous when inflation rises — an asset with zero real growth can still be subject to large capital gains if it’s held over an inflationary period. As a result, I’m open to persuasion on the idea that capital gains should be adjusted for inflation before being taxed. But the bigger picture is clear: as Soltas himself explains, “the capital-gains tax ignores investments in human capital and thereby creates a disincentive for that particular form of investment”. Unless and until Soltas can come up with what he calls “an equivalent subsidy for human capital”, we should treat all income equally for tax purposes — whether it comes from income or whether it comes from capital gains.

Every investigative journalist occasionally dreams of what she might be able to do with monster resources and subpoena power. The answer looks something like Carl Levin, whose latest report on Apple’s tax strategies is Pulitzer-worthy stuff. When Apple CEO Tim Cook testifies in front of Levin today, it’s going to be one of the most uncomfortable grillings of his life. Steve Jobs could be intense — but Carl Levin, in full flow, is truly formidable.

The first discrepancy I’d love to see Levin clear up is a simple factual one: how much income tax does Apple pay? The various tax years and fiscal years are rather confusing, but in its testimony, Apple says that its income tax payments to Treasury were “nearly $6 billion” in FY2012, for “a US federal cash effective tax rate of approximately 30.5%”. (Those numbers imply taxable income of about $19.6 billion.)

The Senate report, by contrast, looks at the 2011 calendar year, and reproduces Apple figures showing $3.884 billion in current federal taxes, plus holding on to $2.998 billion in deferred federal taxes, for a total of $6.882 billion; that means an effective tax rate of 20.1%. (Again, working backwards, the implied total taxable income increases here to $34.2 billion.)

The report then presents the actual amount of cash paid in taxes, as reported on Apple’s tax return. (This is where that subpoena power comes in particularly handy: I’d love to see Apple’s response to a reporter asking to see Apple’s Form 1120 for the past three years.) According to the Form 1120, which is the corporate equivalent of the 1099 1040 for individuals, Apple paid $2.5 billion in actual cash payments to Treasury in FY2011, up from $1.2 billion in FY2010.

The report doesn’t convert those figures into an effective tax rate, just saying that the number would be “well below the statutory tax rate”. But in in the year ended September 23, 2011, Apple overall reported net income of $25.9 billion, while in the following year its net income was $41.7 billion. Much of that income was overseas, of course. Still, it does seem that Apple’s total actual federal tax payments in both FY2010 and FY2011 were less than 10% of its reported net income.

This is particularly shocking to the US public, which has to pay taxes on its global income. Every other country’s billionaires are extremely good at escaping into a state of tax-free statelessness; America’s aren’t, and we expect that if you’re rich American, you’re going to pay a substantial amount of US taxes.

American multinational corporations, in this sense, lie somewhere in the middle: they don’t need to pay income tax on their global income, and so they can avoid billions of dollars in taxes by moving income to tax-friendly jurisdictions like Ireland, or to subsidiaries such as Apple Operations International and Apple Sales International, which pay taxes in no jurisdiction at all. (Their headquarters are in Ireland, so they are sheltered from US taxes, but since their operations are mostly in the US, they don’t pay Irish taxes, either.)

The only real punishment for avoiding taxes, if you’re a US corporation, is that your offshore profits are stuck offshore, where it can be hard to invest them or return them to shareholders. So when Apple claims in its testimony that it “supports comprehensive reform of the US corporate tax system”, note its two key provisos: that such reform be “revenue neutral”, and that it allow “free movement of capital back to the US”. The first would mean that US corporations wouldn’t actually pay the taxes they’re avoiding right now: total corporate taxes would remain at an all-time low. And the second would mean that the biggest corporate tax loophole of all — the ability to pay no taxes on foreign earnings — would be made substantially bigger.

The Senate report quotes Mark Keightley, making a very important point:

Corporate tax revenues have declined over the last six decades. In the post-World War II era, corporate tax revenue as a percentage of gross domestic product (GDP) peaked in 1952 at 6.1%. Today, the corporate tax generates revenue equal to approximately 1.3% of GDP. The corporate tax has also decreased in importance relative to other revenue sources. At its post-WWII peak in 1952, the corporate tax generated 32.1% of all federal tax revenue. In that same year the individual tax accounted for 42.2% of federal revenue, and the payroll tax accounted for 9.7% of revenue. Today, the corporate tax accounts for 8.9% of federal tax revenue, whereas the individual and payroll taxes generate 41.5% and 40.0%, respectively, of federal revenue.

What we’re seeing here is a corporate class which is vastly more effective at evading taxes than individuals are; I don’t see that trend going away any time soon.

Instead, I have a modest proposal of my own: why not at least require all public US companies to file their federal tax returns with the SEC. They already report the amount of taxes that they pay, but as we’ve seen, the reported numbers, calculated under GAAP, can differ substantially from the actual cash numbers. I’m not saying we’d shame companies overnight into suddenly paying more taxes. But at least we’d be able to see which ones are evading taxes most effectively.

James Stewart has a long attack this weekend on the one idea from the presidential campaign which managed to receive genuine bipartisan support: the cap on deductions. He’s a first-rate reporter and columnist, so it’s worth going into some detail about all the different places he’s wrong.

Stewart starts off his column by summing up his two main arguments against a cap on deductions:

Without addressing larger tax preferences, like a lower rate on capital gains, it does almost nothing to cure the so-called Buffett problem, in which Warren Buffett’s secretary pays a higher effective rate than her billionaire boss. It doesn’t even raise much revenue.

Saying that a cap on deductions doesn’t cure the Buffett problem is a bit like saying that some random bit of Dodd-Frank doesn’t solve too-big-to-fail, or wouldn’t have prevented the 2008 financial crisis. It’s true, but it’s irrelevant. You can’t approach the current fiscal negotiations with the idea that solving the Buffett problem is a necessary precondition for any fiscal-policy tweak: you’d never get anywhere if you did. The task right now is to come to an agreement on a set of policies which will raise revenues and cut expenditures; a cap on deductions does exactly that. And what’s more, while it won’t mean Warren Buffett paying a higher tax rate than his secretary, it will at least reduce the distance between them.

As for the idea that a cap on deductions “doesn’t even raise much revenue” — well, that’s in the eye of the beholder. The dog not barking here is that Stewart never actually comes out and say how much money a cap on deductions would raise. Here are the numbers, from the Tax Policy Center: a cap at $50,000 would raise more than $700 billion over ten years, while a cap at $25,000 would raise some $1.2 trillion. That’s real money. Even if you exempt charitable donations from the cap, you’re still raising almost $500 billion at the $50,000 level, and more than $800 billion with a $25,000 cap.

Stewart is at least honest about the main reason he opposes this cap:

It would hit people like me: taxpayers in higher brackets who rely on earned income as opposed to investment income or an inheritance, who give to charity and live in a high-tax state. Assuming a $35,000 limit on itemized deductions, my federal tax last year would have risen to 27 percent of my adjusted gross income, from 22 percent.

Stewart talks about his own personal tax rate a lot in his column; he must think it’s of great interest to the rest of us. Interestingly, he always talks about his tax rate as a percentage of his adjusted gross income, which is surely a lot higher than his tax rate as a percentage of the total amount of money he makes every year. (As a self-employed professional, Stewart can take a large number of expenses, including housing expenses, and deduct them from his income before calculating any tax at all.)

Stewart’s point is absolutely correct, as far as it goes. The three major deductions are state and local taxes; mortgage interest payments; and charitable contributions. So people who spend a lot of money on those three things every year — people like Stewart — are going to be precisely the people who are most hit by a cap on deductions.

At the same time, however, Stewart is rich, and everybody knows that the rich are going to have to pay more in taxes, one way or another. Indeed, Stewart says he’s OK with that: he claims that he “wouldn’t mind paying more” in taxes, just so long as the top 400 taxpayers in the country all paid more in taxes as well.

But here’s the thing — they would! According to Stewart’s own calculations, the taxable income of the top 400 taxpayers would rise by $32 million, on average, while their overall tax rate could go up to 25% from 20%. Seems like a big hike to me. But because that 25% is lower than Stewart’s own 27%, he’s decided that we’d be better off not capping deductions at all.

This is profoundly myopic. I can see how on a philosophical level it makes sense to ask the top 400 taxpayers in the country to pay a higher tax rate than James Stewart. But the top 400 taxpayers are, by definition, a highly exceptional bunch, who spend millions of dollars a year on tax-avoidance strategies. It might or might not be possible to construct a tax regime which makes the top 0.0001% pay a higher tax rate than James Stewart, but I really don’t think that failure to do is reason to do nothing at all.

Maybe realizing that he’s on to a losing argument here, Stewart shifts course at this point, describing the deduction cap as “a stake aimed at the heart of the charitable deduction”. And once again, the dog doesn’t bark: he quotes lots of people who work in the non-profit sector, saying that this move would reduce the amount of money that people give to charity. But not once does he hazard a guess at the amount by which charitable giving might decline; indeed, he quotes Patrick Rooney, of the Center on Philanthropy at Indiana University, as saying that he hasn’t studied that question.

Rooney has studied similar questions, however. For instance, his institute looked at the effect of capping the deduction at 28%, even for taxpayers with a higher marginal tax rate. That tweak would reduce charitable giving by some $2 billion per year, they found — but it would raise ten times that amount in new tax revenues.

And when the CBO recently looked at various different ways of changing the charitable tax deduction, they came to much the same conclusion:

In each case that CBO examined, the reduction in the subsidy (and thus the increase in revenues) would exceed the reduction in charitable contributions, whether measured in dollars or as a percentage change.

If there’s one constant when it comes to the charitable deduction, it’s this: its opponents love to get quantitative, while its defenders generally refuse to talk numbers at all. For instance, check out Bob Shiller’s column this weekend: despite the fact that he’s a fine economist, he never once talks costs and benefits, instead relying on general principles such as the one saying that “income that is freely given away should not even be considered as taxable income”. And then compare Dick Thaler, or any of the many other critics of the charitable deduction: they ground their arguments in reality, rather than in the clouds.

So when Stewart starts saying that capping deductions will hurt the poor, on the grounds that the poor go to hospitals and museums, and those hospitals and museums are reliant on charitable donations — well, take it all with a pinch of salt. And move on to Stewart’s next argument, which revolves around the deductibility of state and local taxes:

According to the Census Bureau, state taxes per capita in 2011 ranged from $3,491 in New York to $1,674 in South Dakota. For many higher-income taxpayers in high-tax states, state and local taxes alone would exceed the cap limit, completely depriving them of the mortgage and charitable deductions.

This is an interesting use of the word “many”. If the cap was put at $25,000, that would be more than seven times the average state taxes in the state with the highest taxes in the union. If the cap was at $35,000, it would be more than ten times New York’s average state taxes. So yes, if you pay ten times the average amount of taxes in your state, and if you live in New York, then you might use up all of your cap with state taxes alone. You’ll excuse me if my heart doesn’t bleed.

Stewart concludes by reiterating that he would rather see other people pay more in taxes, rather than himself — especially people who rely less on income and more on capital gains. I’m inclined to agree with him, on a policy level: I too would like to see unearned income taxed at the same rate as earned income. But the fact is that all the big deductions — charitable, mortgage-interest, even state and local taxes — are bad public policy. We should cap them, at a high level if necessary, and then bring down the cap over time, until it reaches zero. That, in turn, will help income tax rates to converge on capital-gains tax rates, again over time. Few things in fiscal policy happen overnight. But capping deductions is a step in the right direction. And Stewart should embrace that, rather than fighting it.