Separating Hyperplanes

The Bush Tax Cuts

Matthew Martin 12/16/2012 12:00:00 PM

By now everyone has heard about the "fiscal cliff," perhaps under its other names "fiscal slope" or "fiscal bluff." The biggest component of the slope is the expiration of the Bush tax cuts. Here's what happened. Before Bush, tax rates were higher across the board. With the exception of the top bracket, the rates when Bush took office were the same as the ones when Reagan left office--Clinton had raised just the top marginal rate in order to balance the budget. Bush passed a series of marginal rate cuts that concentrated on the top income brackets, but included cuts for all brackets as a way to make it politically palatable. But here's the problem: Bush couldn't find the money to pay for the cuts. Even with screwy math that obviously understated the impact on the deficit, administration officials were unable to make it appear as though the tax cuts could be paid for. So they opted for an accounting trick. They made the tax cuts expire at the end of the decade so that when the CBO did its "current law" projections, it would look like the tax cuts wouldn't push the budget into deficit. Of course, this was dishonest from the beginning--republicans had every intent of making the cuts permanent, and if they cared at all about the empirical evidence, knew they would push the budget from Clinton's surplus into deficit (which also violated the budget rules in place at the time).

Those tax cuts expired at the end of 2010. The democrats, who controlled congress and the presidency, but did not have enough to overcome filibuster in the senate, wanted to renew only the portion of the tax cuts for those making less than $250,000. Doing so raises a significant amount of revenue--to the tune of $849 billion over ten years according to Treasury estimates. Moreover, these tax rates would be quite similar to the plan that Obama campaigned on in 2008, which called for tax breaks for those making less than $250,000 (enacted as part of the stimulus act) paid for by tax increases for those making more than $250,000 (which have yet to be passed). The democrats faced two problems. First, they never actually had 60 seats in the senate, meaning that the GOP always had the ability to torpedo any democratic bills. Second, there were a group of roughly eight democratic senators that were "professional bipartisans" who would not vote for most democratic bills unless at least one republican voted for them. That means that the democrats could only wrangle 51 votes for the tax bill. Rather than see all the tax cuts expire, they compromised by renewing the all the cuts for just two years, in exchange for the GOP lending them enough votes to bypass the filibuster.

We are now two years later, and the tax cuts, all of them, will expire on December 31.

There is a certain misfortune that these tax cut debates had to come at a time when the economy is weak. The universally accepted premise inside the beltway and across the popular press is that recessions are a particularly bad time to raise taxes. The thing is, there is remarkably little scientific evidence for thinking this. Theory suggests that higher marginal income tax rates will reduce potential GDP by causing workers to choose to work fewer hours. Some studies have found that this is basically true, but it is particularly hard to obtain proper identification for econometric analysis of tax cuts--that means that we can't be sure the effects we observe are the result of tax cuts. But the problem is that at present, all the theory and all the data suggest that we are already operating well below potential GDP, meaning that there is no reason to suppose that raising the marginal tax rates right now will have any supply-side effects at all: for there to be a supply-side effect, there would have to be people who would be able to find jobs at the current tax rate who would be unwilling to work at the higher tax rate. But what we observe is quite the contrary, there are tons people who do not currently have jobs that would be willing to work at either the current or the higher tax rates. That means that even if people drop out of the labor force as a result of the higher rates, we won't see changes in employment, and therefore no change in GDP in response to the higher marginal tax rates.

But that is only if we consider just the supply-side effects of marginal tax rates. There is also the demand-side effect of the fact that more government revenue means households have less spending money. And for an economy already constrained by demand, this probably means that higher revenues will reduce employment and GDP.

This suggests an easy solution. We should completely ignore the state of the economy when setting the marginal tax rates, and if we are concerned about the impact of taking away people's spending money, then simply refund that revenue in the form of a bigger standard deduction, which would expire as soon as the economy is recovered. This is, actually, what the optimal tax research suggests we should do. Few people seem to realize that the science actually tells us that optimal marginal tax rates are largely independent of the size of government or the state of the economy. That's because marginal tax rates are primarily about redistribution, not financing government spending or stabilizing the economy. In fact, a basic New Keynesian model suggests that lowering marginal tax rates on labor income can be dangerous in a liquidity trap, since it puts downward pressure on wages, causing demand to fall further.

This all suggests that we really ought to be asking whether to renew any of the Bush tax cuts. Letting all of them expire is the single best thing we can do for the long term budget picture, and any negative impact this might have on the recovery (which, as we've noted, would be due purely to demand side effects) can be offset by a temporary increase in the standard deduction.