One way to make a Keynesian angry is to point out that stagflation -- a simultaneous increase in inflation AND unemployment -- is not possible under the Keynesian scheme. The Phillips Curve supposedly "proves" that the way to get unemployment down is to ramp up inflation.
In a recent blog post, "Stagnation Nation," Krugman laments the woeful GDP numbers and the fact that unemployment is going up, and he blames it on a lack of government spending. However, inflation also is increasing, no matter what Krugman claims.
Anyone who purchases food, fuel, and consumer items can tell that prices are going up, and Krugman can tout the government's CPI all he wants, but that does not change the fact that we are seeing a huge increase in commodity prices at the same time it becomes harder and harder to find work. Yes, yes, he claims that "volatility" is the answer to rising commodity prices, as though volatility explains why they are rising together. What we know is that commodity price changes tend to be more volatile than price changes for finished goods and labor, but that does not explain why commodity prices are going up.
I recall watching the late Richard Gill neatly try to explain stagflation in a scholastic film. His trick was to move the aggregate supply curve to the left and, Voila!, stagflation! But, it only was a trick and explained nothing and certainly did not explain production of goods in the USA during that time. No, the Keynesians have no way to explain stagflation, and so they either ignore it or try to explain it away, claiming higher oil prices are the key. (The cause of inflation, in their view, is...higher prices, a nice example of "begging the question.")
In the end, Krugman claims that governments should spend more, as though a government magically can produce resources out of nothing. And when someone points out that simple fact, he gets angry and resorts to insults or simply pulls another logical fallacy out of his Keynesian hat.