Like it or hate it, policymakers in Washington are still obsessed with the deficit. That's why think tanks keep churning out clever plans to cut spending and raise taxes.

Pumping out oil... and budget surpluses! (AP)

And here's a new paper from the Council on Foreign Relations offering an interesting twist on the theme. Using economic modeling, Michael Levi and Citgroup's Daniel Ahn suggest that a tax on oil consumption could be one of the least harmful ways to trim the budget deficit.

How do they figure? Levi and Ahn first assume that Congress will enact a big deficit-reduction package over the next 10 years that cuts spending by 3 percent of GDP by 2020 and raises corporate and income taxes by 1 percent of GDP by 2020. That may be unlikely in the real world, but it's fairly similar to the much-discussed Simpson-Bowles proposal.

Next, the authors look at what would happen if Congress scrapped some of those tax hikes and spending cuts and instead replaced them with a tax on oil consumption. This would could involve simply raising existing taxes on gasoline, diesel fuel, and jet fuel. They assume the oil tax would be phased in over time and come to about $50 per barrel of crude oil in 2020, or an extra $1.20 per gallon of gasoline.

After running their economic model, Levi and Ahn found that using the oil tax to fend off some of the spending cuts and income tax hikes could be beneficial to the U.S. economy. In other words, a deficit package with an oil tax could be less harmful than a deficit package without one. Here's the key chart:

In Variation 1, the gold line, the oil tax is used to restore part of the government spending cuts in the big deficit-reduction deal. In Variation 2, the blue line, the oil tax is used to restore part of the spending cuts and keep taxes lower. In Variation 3, the red line, the oil tax revenue is used to keep income and corporate tax rates at their current levels.

The end result: The U.S. economy performs better when there's oil tax revenue to fend off spending cuts and tax hikes. GDP rises faster and unemployment falls further.

Why might this be? For one, Levi explained in a phone interview, a portion of the oil tax would fall on foreign countries, since the United States still imports about 40 percent of its crude. What's more, oil in the United States is relatively lightly taxed. "Raising taxes on something that's under-taxed, like oil, rather than something that's already heavily taxed, like income, can yield good results," Levi said.

Of course, this is a rather simplistic scenario, and Levi and Ahn model a few other possibilities in their full paper (pdf). For instance, it's quite possible that an oil tax would curb U.S. fuel consumption, which might in turn lower global oil prices. (Though that's hardly certain; a lot would depend on how OPEC responded.) In that case, the U.S. economy could see a slightly bigger boost.

Meanwhile, there are distributional consequences to consider. An oil tax is likely to be quite regressive — many poorer Americans spend a greater fraction of their income on gasoline. So Levi and Ahn looked at what would happen if half of the oil tax revenue was kicked back to consumers as lump-sum rebates, while the other half was used to reduce taxes and maintain spending levels. Even in that case, the economy performs better than it does under a standard deficit-reduction plan.

In theory, a tax on oil could have other benefits as well — if it reduces domestic fuel consumption, that would make the U.S. economy less vulnerable to large swings in global oil prices. But those benefits aren't factored in here.

Last year, a similar study from MIT looked at the effects of using a broader carbon tax to trim the deficit. That study found that carbon taxes only offered a slight advantage over other budget-cutting measures. But there's an important difference here — unlike the MIT study, Levi and Ahn's paper doesn't assume that the U.S. economy will be running at full employment anytime soon. And in that case, finding ways to blunt the impact of deficit reduction over the next 10 years could have a big effect on the course of the economy.