In an interesting opinion piece for the Wall Street Journal, Robert Barro and Charles Redlick give empirical evidence supporting the claim that stimulus spending doesn’t work. By “doesn’t work” they mean that the government spending multiplier is less than 1. This means that stimulus spending increases national income by less than the amount of new spending. Why? Barro and Redlick are vague on the reasons. They invoke “crowding out” and, presumably, the “permanent income hypothesis”:
Empirically, our research shows that most of the fall was in private investment, with personal consumer expenditure changing little.
It would be good to know more about why private investment fell. Is Robert Higgs’s “regime uncertainty” hypothesis about the Great Depression the explanation? Another possible explanation is the traditional Hayekian one: A substantial part of the problem is the prior misdirection of resources that aggregate government spending does not address or may even make worse.
Is there one explanation for all the periods for which Barro and Redlick had data? And then there is the result that stimulus begins to work when unemployment reaches 12%. Huh? Why?
The problem with this kind of neoclassical answer to the Keynesians is that it doesn’t include a plausible theory to rationalize the admittedly thin empirical results. The Keynesians have a plausible story to tell.
Empirical work is valuable but empirical work with scanty, implausible or incomplete theories will not convince other economists, policymakers and the general public.
One of the most important reasons that both Hayek and Keynes still have a hold on the thinking of mant highly intelligent people is that they told good stories. Don’t underestimate the power of “the story.”