Friday, August 26, 2011

A collective action / externalities rationale for Keynesian stimulus

I'm no longer surprised by the frequency with which good economists - of course, not all of them - fail to understand and apply basic economics reasoning. In the fields I write about professionally, I'd have to say I'm grateful, as it's good for business.

A key reason for these failures is that people get lost in the forest and can only see the trees. They play with models, use math, etcetera, but forget the basic underlying intuitions. Or, they become prisoners of simplifying assumptions that are often (but not always) useful, and forget that these assumptions should only be used conditionally and when appropriate.

Today's example is Robert Barro, who has a Wall Street Journal op-ed today (available outside the paywall here) claiming that Keynesian economics, unlike "regular economics," can't possibly make sense. Thus, with regard to the claim that Food Stamps or unemployment benefits could boost demand and help ease the recession, he concludes:

"There are two ways to view Keynesian stimulus through transfer programs. It's either a divine miracle—where one gets back more than one puts in—or else it's the macroeconomic equivalent of bloodletting."

And elsewhere he says of the at one time uncontroversial Keynesian idea that giving money to the cash-constrained can be stimulative:

"How can it be right? Where was the market failure that allowed the government to improve things just by borrowing money and giving it to people?"

David Glasner responds to Barro as follows:

"But wait a second. What does Barro mean by his query: 'Where was the market failure that allowed the government to improve things just by borrowing money and giving it to people?' Where is the market failure? Hello. Real GDP is at least 10% below its long-run growth trend, the unemployment rate has been hovering between 9 and 10% for over two years, and Professor Barro can’t identify any market failure?"

Glasner then asks whether Barro agrees with the real business cycle theorists who explained that the Great Depression merely reflected millions of workers' rational decision to take a nice long vacation until productivity and therefore wages were higher.

Paul Krugman jumps in as well, and mentions some of the standard points from Keynesian economics for which there is empirical evidence, such as sticky prices and wages.

But I have long thought it reasonably clear (as discussed somewhere in here) that a very familiar tool can do a lot of the work here - collective action problems. Economic models often simplify the world, and in the right setting with good reason, by taking people's preferences as given and assuming away interdependence. But suppose I am deciding whether to spend money on a nice vacation. Even in an entirely rational setting and with a flexibly responding price system, this may depend on how well I expect my business to do over the next few years. Suppose people who are considering whether to patronize my business have exactly the same thought in mind. Their willingness to buy goods and services depends importantly on their confidence that others will be buying their own goods or services. So, when everyone gets scared or anxious, there is a coordination problem. If only everyone could agree to shake hands and continue opening their wallets a bit, the problem would ease, but instead it feeds on itself, as belt-tightening here prompts responsive belt-tightening there.

To throw in another common buzzword, there's an externality here, as each individual's belt-tightening causes others to lean more towards belt-tightening, creating collectively self-fulfilling prophecies about low earnings potential.

Against this background, Food Stamps and unemployment benefits, by getting people to spend more (albeit perhaps more because they were cash-constrained than due to the vacation problem above) can get things moving in the other direction. People whose businesses start doing better change their estimats about how much it makes sense for them to spend, and things may start going the other way.

Back in January 2009, Barro had a WSJ op-ed that was almost as skeptical about stimulus, in which he said:

"[Keynesian theory] implicitly assumes that the government is better than the private market at marshaling idle resources to produce useful stuff. Unemployed labor and capital can be utilized at essentially zero social cost, but the private market is somehow unable to figure any of this out. In other words, there is something wrong with the price system."

Seen through the above lens, however, the price system in a recession or depression may be getting it exactly right so far as revealed preferences are concerned. Resources are idle because there isn't enough demand for the production that is being forgone. And given all that the price system is working just fine.

But the seemingly efficient equilibrium is far inferior in human welfare terms to the alternative one that would result if people could coordinate shifting to a higher-expressed demand, higher-output equilibrium.

This presumably is what Glasner means when he says: "Hello. Real GDP is at least 10% below its long-run growth trend, the unemployment rate has been hovering between 9 and 10% for over two years, and Professor Barro can’t identify any market failure?"

And this is why the view that the Great Depression was just a nice long holiday, as people awaited higher productivity that would increase their willingness to swap leisure for work, has never seemed very intuitively persuasive.

But the use of collective action problems and externalities that I suggest here lies outside conventions that economists are accustomed to allowing in their models (which commonly take revealed preferences as given rather than conditional and interdependent, and assume away externalities unless clearly demonstrable like that from pollution). Also, the use I suggest is admittedly a bit informal and ad hoc, which economists may rightly be on guard against. But it is coherent logically, plausible intuitively, and permits one to make more sense of the world. Barro's apparent inability to see that it might be relevant, and thus his entirely misplaced sarcasm about whether Keynesian economics could possibly make sense outside the realm of "divine miracle," is more disappointing than surprising.


cap441 said...

Great write up on the issue. The theory that collective action magnifies the externalities from individual belt tightening seems to reject the rational economic participant. We are not belt tightening because of our own individual application of economic principles to empirical evidence, rather we are simply copying each other and doing what is fashionable. I have long thought that the same phenomenon that explains whether suits have wide or thin lapels explains most peoples’ economic outlook.

It is essential to understand that this is the problem because it will lead us to our solution. We are currently using rational solutions to an assumed rational problem, like unadvertised payroll tax cuts and quantitative easing’s complex increase of the money supply. Suits are sold by advertising. Maybe we should try a different approach.

Daniel Shaviro said...

Copying isn't necessarily irrational. In belt-tightening you might either observe that people aren't spending, or trust the wisdom of crowds re. what to expect. And on wide versus thin lapels you may want to fit in with the crowd for reasons that make sense.

Agreed that in principle advertising can solve the Keynesian problem. E.g., if Obama gave a speech that actually motivated millions of people to start spending more, it would do the job. Unfortunately (even leaving aside his inadequacies) decades of presidential cheap talk have degraded this tool if indeed it ever had any weight (which I suppose is itself unlikely).

Paul said...

I wouldn't gloss over the cash constrained as an initial part and sustainer of collective action. If wealth is substracted (fall in housing prices) and/or no cash (10% or more loose their jobs), belt tightening would seem to be more rational.