The Circular Logic of Stimulus Spending
There's a circular logic to calls for more and bigger economic stimulus: Every time stimulus fails to kick the economy back into gear, the keepers of the Keynesian consensus inevitably say that the problem is that the last round of spending was too small. And so the calls begin for more.
Starting in the 1990s, Japan followed those recommendations for ever more stimulus. But as Harvard economist Robert Barro points out in The Wall Street Journal, the country never saw the results it hoped for:
For the U.S., my view is that the large fiscal deficits had a moderately positive effect on GDP growth in 2009, but this effect faded quickly and most likely became negative for 2011 and 2012. Yet many Keynesian economists look at the weak U.S. recovery and conclude that the problem was that the government lacked sufficient commitment to fiscal expansion; it should have been even larger and pursued over an extended period.
This viewpoint is dangerously unstable. Every time heightened fiscal deficits fail to produce desirable outcomes, the policy advice is to choose still larger deficits. If, as I believe to be true, fiscal deficits have only a short-run expansionary impact on growth and then become negative, the results from following this policy advice are persistently low economic growth and an exploding ratio of public debt to GDP.
The last conclusion is not just academic, because it fits with the behavior of Japan over the past two decades. Once a comparatively low public-debt nation, Japan apparently bought the Keynesian message many years ago. The consequence for today is a ratio of government debt to GDP around 210%—the largest in the world.
This vast fiscal expansion didn't avoid two decades of sluggish GDP growth, which averaged less than 1% per year from 1991 to 2011. No doubt, a committed Keynesian would say that Japanese growth would have been even lower without the extraordinary fiscal stimulus—but a little evidence would be nice.
Read Reason's summer 2009 story on Japan's post-bubble policies here.