When I wrote a post a couple of days ago about the size of Obamacare’s tax increase, I thought I was bending over backward to be fair. I could have estimated the average size of the increase over ten years — which would have included all the early low-tax years — but even though that’s fairly standard for budgetary purposes I thought it would be cherry picking. Likewise, I didn’t use Jerry Tempalski’s estimate of the four-year cost of the bill — a very modest 0.18% of GDP — even though that’s the standard he used for all the other tax increases he analyzed. Technically, this would have been an apples-to-apples comparison and it would have made Obamacare’s taxes look very small, but again, it seemed like cherry picking.
So instead I used the figure that PolitiFact calculated for the end of the initial ten-year period, the largest possible figure you could reasonably use, and compared it to Tempalski’s four-year estimates for the other bills. Even at that, Obamacare still came in as only the tenth largest tax increase since 1950.
But it turns out this isn’t good enough for Sen. Jim DeMint. He insists on a different standard, based on a single sentence in a CBO long-term budget outlook from 2010:
Under the extended-baseline scenario, the impact of the legislation on the revenue share of GDP would rise over time, CBO estimates, boosting revenues by about 1.2 percent of GDP in 2035…..
That would be a pretty sizeable tax increase, but are you wondering why DeMint used a CBO estimate from 2010? That’s because CBO’s 2012 estimate clocks in at only 0.8 percent of GDP, something I imagine DeMint knows perfectly well.
That’s pretty hackish, but the larger problem here, which DeMint also knows, is that he’s inventing a brand new standard for comparing tax legislation out of thin air. Nobody knows how a tax law is going to play out over 25 years.1 Nobody knows what the economy is going to look like a quarter of a century from now. Nobody can estimate the cumulative size of incentive effects between now and 2035. There’s nothing wrong with CBO taking its best shot at a long-term estimate, but for all these reasons and more, a 25-year time horizon has never been a standard for comparing tax bills. DeMint knows this, just like he knows that CBO’s most recent estimate is a lot lower than the outdated 2010 estimate he tried to put over on us.
1That’s especially true in this case, since CBO’s revenue estimate is based largely on the effect of Obamacare’s Cadillac tax on healthcare plans. This tax is extremely sensitive to assumptions about the rise of healthcare costs and to the response of both corporations and consumers to the tax.