By implication, enthusiasm for fiscal activism in macroeconomic management is always a mistake, while the current alarmism about the fiscal cliff is unjustified. However, Friedman never wrote a rigorous paper on the relationship between changes in the budget deficit and demand. Part of the trouble may have been that dissection of the Keynesian argument required detailed statistical work which, late in life, Friedman was reluctant to undertake.

Technically, the right concept in a test of Keynesianism is the change in the budget deficit that is attributable to policy. In practice there is the complication that the deficit is buffeted around by the effects of the business cycle on tax revenues and expenditure. These cyclical effects need to be stripped out of the calculation, to arrive at the cyclically adjusted or "structural" budget balance. Fortunately, in the last few years the International Monetary Fund (IMF) has devoted resources to estimating structural budget balance numbers for all the world's major economies, including the US. Do these numbers endorse the government activists or the fiscal sceptics? Who is right, the Keynesians or Friedman?  

Let us be more precise about the hypothesis. The Keynesians would be proved right if decreases in the structural budget balance (that is, increases in the budget deficit or reductions in the surplus) were associated with above-trend growth. Economists have a highfalutin' way of checking whether growth is above or beneath trend, when they refer to "the output gap". The Keynesian hypothesis, in short, is that the structural budget balance and the output gap ought to change in opposite directions. When growth is above trend, output exceeds its trend level or is less beneath its trend level, and the change in the output gap is positive. (I apologise for the jargon, which to a degree is unavoidable.) Numbers for the output gap are also prepared by the IMF.