Belligerati

« A law to keep down tyrants | Main | Selection effect »

February 23, 2009

What I'd want to know to believe in the stimulus is not certainty

Justin Fox says, "I guess what continues to perplex me at least a little is how lacking in the customary rigor of modern academic economics the arguments for stimulus are." (Brad DeLong tutors me on fiscal stimulus). Then Brad Delong says, "Lacking in the customary rigor..." Justin could mean either of two things: 1.Rigorous economics should produce tightly-estimated conclusions based on statistical sieving of mountains of data, like: when Safeway cuts grocery prices by 1%, its sales rise by 1.456%. 2.Rigorous economics should involve lots of theoretical equations with sigmas and rhos and betas in them.

Can you say false dichotomy? He's saying you could only remain opposed to his position if either you demanded a high level of significance or if a lot of algebra underlies your estimates. I'm saying that one could easily currently be skeptical without requiring an high level of statistical precision. For example, you could just want external validity and perhaps 5 or even 10% significance.I'd like more data to be convinced of stimulus efficacy, but I don't necessarily require enough for a 1% confidence interval. When you have as few data points as we have on a stimulus, you can barely do any statistics at all. Standard errors are huge and intervals around the mean easily do not contain the true value.

In general, shouldn't the burden of proof be on those who want the policy intervention? If the best evidence we have of high stimulus multiples is a couple of mildly relevant anecdotes, I challenge that the meeting of that standard, and I don't need a high degree of (necessarily model specific) parameter precision to make that challenge.

Given the proceedure with which I see my government spending money, the timing of the spending in the stimulus, the things the money is set to purchase (categories without particularly high unemployment and therefore a lot of unused labor resources) , and the size of the stimulus, I struggle to see relevant external validity in any multiplier number that anyone has discussed. I retain a prior that with the exception of a few public goods and private goods but publicly provided infrastructure projects, that in general the government spending multiplier is 0 or negative.

Could we really believe that at all levels of government expenditure, during a recession like ours, that government stimulus spending not only generates enough private activity to completely repay the taxes on the money spent on that stimulus, but that it further creates economic activity beyond that? From our government of $300 ashtrays, mars rovers that crash, schools that don't teach, bridges to nowhere, where Boston's big dig was 9 billion dollars over budget, and NYC hasn't built a second avenue Subway despite 40 years of trying, I am skeptical.

It is true that we are seeing a rather sudden shift to savings which may be triggering the so called paradox of thrift is a problem. However, we do actually need to save much more as a society. American's have sae too little for too long, and now that the boomers are retiring, we have an enormous amount of benefits to pay them, and we need savings to do so. The benefits of thrift may outweigh the costs, even if Keynes was right and those costs are real. I'm ignorant of how to estimate these costs and benefits, but then, I'm not proposing a trillions be spent to fund my ideas.

Others might say, "If we were considering intervening in a perfectly competitive market then, yes, the burden of proof should be on those proposing the policy. But this is a world of the second best. There are numerous distortions in the aggregate so it's not the case that "doing nothing" is actually doing nothing. This is especially true since the policy environment is dynamic---e.g. the Bush tax cuts will expire, etc." I appreciate this point, but you cannot claim that this stimulus what a natural continuation of government policy as usual. Practically by construction the stimulus was filled with projects that were second rate and not worth funding in general circumstances. Most of the first rate projects they have been built already. I think we have to do better than this. This certainly is a world where we are far from the competitive general equilibrium. But the theory of second best is something that does not hold as a law. Sometimes movements towards a competitive economy create efficiency gains, and sometimes they don't.

That is why I retain my belief that this stimulus is a wasteful policy with negative consequences on national productivity and efficiency.

Posted by OneEyedMan at February 23, 2009 8:46 AM

Comments

Post a comment