Oh boy. We’re about to have one of these very, very frustrating debates in which people say things like the following:
Of course, even if the bill is not very expensive, it is worth passing only if it will make a difference. And economists say it will. [Emphasis added.]
Last year’s big stimulus program certainly did. The Congressional Budget Office estimates that 1.4 million to 3.4 million people now working would be unemployed were it not for the stimulus. Private economists have made similar estimates.
That’s a quote from David Leonhardt, the economics columnist for the New York Times who can be as frustrating as he is insightful. It’s not obvious to me that “economists say it will” is terribly useful, given that other economists say that it won’t make enough of a difference to justify the cost.
My skepticism is reinforced by Leonhardt’s confidence regarding the impact of the stimulus. I agree that the stimulus had an impact on GDP. GDP = private consumption + gross investment + government spending + (exports − imports). Increasing government spending will tend to increase GDP. And it’s certainly true that not laying off large numbers of redundant workers will help maintain private consumption. The tougher question is that impact of government dissaving on long-term economic outcomes, as Leonhardt understands.
There is another question: given that GDP relies on a raw measure of government spending rather than a measure of the value of government spending, it counts value-destroying public-sector activities as a plus. Remember the stories about Chinese workers melting down valuable farm implements to increase steel output to reach arbitrary targets set during the Great Leap Forward? Nothing that egregious is happening in the U.S. public-sector. But it’s worth remembering that we don’t mark government spending to market.
Earlier this week, Ed Glaeser published a terrific post at Economix — never let it be said that the NYT doesn’t allow a great variety of views to be expressed in its pages — that offers a somewhat different take on stimulus spending.
The chart below shows a plot of the change in unemployment between January 2009 and March 2010 on per capita Federal Recovery Act funds received in each state.This relationship is negative and almost statistically significant at the 5 percent level, which lends a bit of support to the view that the recovery spending reduced unemployment. But that negative relationship is driven entirely by three states with very few people — Alaska and the Dakotas. If I weight by population, or eliminate those three states, or even control for state unemployment as of January 2009, there is no longer any significant relationship between spending and change in unemployment. [Emphasis added.] I’m not suggesting that spending did or didn’t reduce unemployment; I am asserting that we can’t tell anything with any degree of certainty.
To add more complexity to the mix, even if we found that recovery funds did significantly reduce unemployment, that wouldn’t necessarily justify their cost. If you hire thousands of people on make-work jobs, then you are wasting their time. That cost needs to be weighed against the benefits of countering the recession.
Might this have been a useful finding for Leonhardt to make note of in his column? It would certainly give readers a different impression of the state of the debate among economists to follow his observations re: the impact of the federal stimulus bill.
But I actually agree with the spirit of Leonhardt’s column, as expressed in his conclusion:
But this new spending needs to be accompanied by something more credible than Augustine-like vows of future parsimony. It should be paired with substantive cuts to continuing policies, like subsidies for oil companies and agribusinesses, outdated weapons systems, NASA’s moon program and at least some Bush tax cuts, among many other things.
We just disagree as to the proportions involved. My sense is that there is far more wasteful spending in the federal budget, not to mention state and local budgets, than Leonhardt suggests with his brief list of proposed cuts. And that is why I’m so concerned about the idea that we should sharply increase public-sector hiring, an idea that, as Annie Lowrey reports, is very popular with liberal legislators.
Washington has not run out of options to tackle the unemployment crisis. The government could expand, directly hiring workers. (As the census demonstrates, when the government hires workers, unemployment goes down.) It could push tax incentives and stimulus, so that Americans spend more and companies decide to hire more workers. It could take up Rep. George Miller’s (D-Calif.) Local Jobs for America Act, providing $75 billion to local governments to keep employees on the payroll. Or it could have considered Sen. Tom Harkin’s (D-Iowa) proposal to grant $23 billion to keep public school teachers in their classrooms, the Keep Our Educators Working Act. Both of those bills would have preserved or created local government jobs, and a spate of other congressional initiatives are aimed at private-sector hiring.
Because fiscal discipline begets organizational discipline, I tend to think that this is a very bad strategy. This will lead public officials to delay the retrenchment of bloated workforces, stymieing productivity growth and creating problems down the road. If we’re going to spend, we should place a heavier emphasis on investment rather than consumption. Merely labeling public-sector salaries a “public investment” doesn’t make it so. Rick Hess has pointed to one strategy that might work — offer more money, but demand concessions on tenure and other protections. (Joel Klein, one of my heroes, believes that Michelle Rhee has accomplished more than initially appeared to be the case in D.C.) He has also called for serious investments in improving the quality of the data we collect about public-sector outcomes.
Rather than throw money helter-skelter at state and local governments, taxpayers have a right to ask whether big investments yield big reforms that will deliver better performance at lower cost over time. A simple way to think about stimulus spending is that it makes sense to purchase valuable goods when they are on sale. It does not make sense to purchase valueless goods at a high price. Propping up broken bureaucracies is the latter, not the former.