Arpit recently pointed me to a very worthwhile post by Tino Sanandaji which raises questions about Paul Krugman’s recent effort to demonstrate that collapsing tax revenue has been the main driver of increasing deficits:

Paul Krugman asks why deficits have grown. He gives the impression that the explosion in the deficit is due to tax revenue being too low, rather than spending being too high.

But as Tino explains, Krugman presents his graph in nominal dollars, not in terms of percentage of GDP. After starting the chart in 2000 rather than 2005 and making this commonsense adjustment to using percentage of GDP, the picture looks very different:

Notice that the “natural” increase in spending isn’t there anymore. Between 2000 and the period of the crisis in late 2007, spending is essentially flat, while revenue decreases (thanks to the IT-bubble bursting and the Bush tax cuts). Between late 2007 and now however, spending rises sharply as a share of GDP.

It is perfectly understandable that the federal government would engage in countercyclical spending increases. Some of these increases were in so-called “automatic stabilizers.” But Congress could have decided to keep spending in line with shrinking GDP. I wouldn’t endorse that course of action.

And that raises the question of why Krugman felt the need to use nominal dollars to mask this effect? The fact that spending increases account for two-thirds of the increase in the deficit in percentage of GDP terms isn’t in itself dispositive. One gets the sense that point-scoring is taking precedence over honest, forthright analysis. Then again, Tino could be missing something important, but he is, to his credit, quite transparent about his numbers.