So, what would we have expected total government spending — federal, state, and local — to do over the past three years if there had not been a crisis and a change in government control? A first approximation would have been spending growing along with the trend growth in the economy — that is, real GDP growing with the economy’s potential, and government spending growing at real GDP plus inflation.

Now, over the period 2000-2007 — from business cycle peak to business cycle peak — real GDP grew 2.4 percent a year. So a reasonable estimate for trend growth is 2.4 percent, or 7.3 percent since 2007.

We can use actual inflation: the GDP deflator rose 4.1 percent from 2007II to 2010II.

Put these together, and “normal” growth in government spending would have been 11.7 percent over the past three years.

Actual growth has been higher: 19.5 percent. So government spending is about 7 percent, or about $350 billion, higher than a simple trend projection would have suggested. What accounts for the higher spending?

Well, none of it is government consumption; it’s all in transfer payments. BEA data aren’t quite as helpful here as I’d like, but it’s clear that a large chunk, roughly $100 billion, is unemployment benefits, which have surged along with unemployment, and another large chunk is Medicaid spending, which has surged because the slump has impoverished more people. Some more for other safety net programs, like food stamps. Also, Social Security and Medicare outlays have gone up about $85 billion more than my 11.7 percent norm — medical cost growth, aging baby boomers, and maybe some people taking early retirement because they can’t find jobs.

The thing about Krugman is that he's really wedded to the idea that presenting concrete facts will change the discussion. He doesn't understand that the right wing will simply create new "facts" to justify their accusations!