Another Stimulus?

Even before the full effects of the last stimulus bill have been fully evaluated, economists like Paul Krugman argue that America should be instituting a second stimulus plan. They believe that last year’s American Recovery and Reinvestment Act was too small and that a wait-and-see strategy is risky. The true effects of the stimulus measures are still being estimated, and the debate will continue for many months (or perhaps years) about whether or not the money spent was worth it. In the midst of this evaluation, calls for another stimulus plan seem premature.

Would such a plan make sense? It’s hard to know without being able to conclude that the last one was worth it. The massive amount of spending that took place is daunting no matter the outcome, so surely it makes sense before acting again to evaluate the effects of the first effort. Unfortunately, estimating these things is a foggy science at best, because there’s no control environment to compare the economy to. It’s impossible to say what would have happened if we hadn’t taken the stimulus measures.

Last year, Administration estimates of stimulus effectiveness were large, but were based on assumptions of spending “multipliers”—the economy-wide benefit from an additional dollar of government spending. These multiplier estimates were averaged from a range of estimates used by the Federal Reserve and a private forecasting firm, and were as high as 1.6.

The CBO has since released a rosy estimate of the stimulus plan’s effects on the economy. But, crucially, these estimates also rely on high multiplier estimates, ranging from 1-2.5 depending on the particular policy.

Estimating these multipliers from real-world data is a challenge. In the past, physicians applied leeches to their patients, and observed that many of them eventually recovered. Similarly, economists can observe government spending followed by economic recovery. In neither case is it obvious that the treatment was the cause of the cure. Robert Barro has derived different—and far lower—estimates of multipliers that attempt to better capture the causal impact of spending.

It’s also not obvious that the impact of government spending is a fixed number, as opposed to a dynamic variable that depends on numerous other factors. The Center for Economic Policy Research has an analysis that suggests that multipliers depend on broader economic conditions.

America is a mixed case when it comes to stimulus effectiveness. On one hand, America is an open economy with a floating exchange rate and a high debt load. The CEPR finds that fiscal stimulus has a relatively weak effect under these conditions. On the other hand, America is also a high-income country.

Putting these different factors together, the CEPR found that America’s cumulative multiplier from 1960-2007 was 1.19—higher than the estimates Barro prefers, but lower than the ones that the CBO and the Administration uses.

But, significantly, the CEPR also found a large time trend in the data. Stimulus spending was dramatically more effective before 1980 than afterwards. The long-run multipliers after 1980 were just .4.

One interpretation of this data is that more flexible monetary handling from the Federal Reserve has hijacked the effectiveness of stimulus programs. If the Fed responds to greater government spending by hiking rates, monetary and fiscal policies work at cross-purposes, and stimulus spending will be much less effective.

Tyler Cowen and Scott Sumner have also raised this concern with respect to the effects of a possible second stimulus. The Fed has repeatedly stated its intention to hike rates and return to a more normal policy. If a further stimulus does improve economic recovery, its effects could be neutralized if the Fed begins rate hikes earlier than it would otherwise.

There is a certain degree of attraction to Krugman’s proposal to match current stimulus with long-run cuts designed to restore future fiscal balance. In that scenario, additional government spending now could be the sweetener that paves the way for tough budget-balancing measures. But in reality, there are few politicians willing to sign up for that bargain. To the extent there is any political will at all to push new spending bills, they take the form of deficit-raising stimulus measures whose effectiveness is not as ironclad as supporters suggest.