How Keynesian Are We?
June 5, 2010
Last year, Congress passed the American Recovery and Reinvestment Act (or ARRA), known colloquially as “the stimulus”. Justifying this massive federal outlay ($787 billion over a couple years) was the ongoing economic meltdown following the mortgage and financial crises and the predicted decline in output and upsurge in unemployment (here’s Krugman’s pessimistic and, in retrospect, relatively accurate take). Obama signed the bill, and the federal government began spending (well, ramped up spending). A year and a half later, economists are debating the stimulus and its effects (e.g. Glaeser at the NYTimes Economix blog). 70 years later, the debate still rages – does Keynesianism work? Can we spend our way out of a recession? New and old estimates of the effects of various programs (the New Deal, Kennedy’s tax cut, etc.) are bandied in a debate that shows no signs of abating (though it may have gotten a bit more humorous).
Here’s my question though: is ARRA enough to label the current situation a Keynesian stimulus? Clearly, the act increased Federal spending from what it would have otherwise been. But the Federal government is not the whole of public spending. In fact, a quick glance at the National Income and Product Accounts Table 1.1.2. Contributions to Percent Change in Real Gross Domestic Product shows that total government expenditures have actually decreased for the last two quarters, with only contributions to growth the two quarters before that (Hat Tip to Mark Thoma). State and local governments, mostly constrained by balance budget amendments, have been slashing spending almost as fast or faster than the federal government could ramp it up. The trajectory that the US economy takes over the next couple years will be a key data point in debates about Keynesian spending and the like. And yet, when you look at the numbers, the whole thing seems a bit… small. The government as a whole is much larger than it was in 1929 or 1932. But the changes from the trend are tiny. For example, also according to the NIPA, total government spending rose 12.8% in 1934, which in turn contributed 2 percentage points of the increase in GDP.
(To clarify, the contributions I am talking about here are direct ones – there is no way to track inside the NIPA any multiplier (positive or negative). Rather, the NIPA simply calculate the growth in total GDP and then derives arithmetically how much each portion (consumption, investment, government, trade) contributed to the increase. Calculating the multiplier is a much-debated problem in macroeconomics, but I would think a subsequent one to establishing just how much government spent in the first place.)
In short, the New Deal was a big increase in spending in a much smaller government. The ARRA was a big increase in federal government spending that (almost) made up for a big decrease in state government spending. So the question is, how Keynesian are we? How Keynesian have we ever been? And perhaps of more interest for the crowd that reads this blog, why does popular perception maintain that government spending has increased dramatically when the official figures show very modest changes? I think it might have something to do with the focus of media on the federal government, and the failure to aggregate up local issues (laying off teachers and cops and the like) into their macroeconomic effects. But I’m not entirely sure. Also, what should the proper baseline be to compare the effectiveness of a stimulus against – a world where government expenditure was flat, or a world where government expenditure did what it would have without the stimulus (i.e. go way down due to cutbacks at the state level)? There are some interesting questions here, I think, in the realm of civic epistemology.