« Money and Monetary Policy | Main | Are Prices Sticky? »

Monday, November 13, 2006

Economists Thinking Like Economists

In this "season of non-economists opining about what economics is all about," David Altig comes to the defense, yet again, of economists thinking like economists. Having just posted a "Military Production Function" here earlier today, David's defense is timely:

More Things Economists Don't Say, by David Altig, macroblog: From The Weekly Standard (via Instapundit) comes an article by Harvard law professor William Stuntz, "explaining" how thinking like an economist is the wrong way to go about weighing the options in Iraq. I hesitate to wade into these waters... But it seems to be the season of non-economists opining about what economics is all about, and I seem to be in the mood to object.

So object I will. But before I do, let me make this ... clear: Nothing I'm about to say should construed as support, one way or another, for any particular position on whether the war was a good idea or bad idea, ... or whatever. ... I am not interested Professor Stuntz's position on the war per se, but rather his characterization of what it means to think like an economist.

With that disclaimer, we begin:

When you gamble and lose, the natural tendency is to double your bet--and when that doesn't work, mortgage everything you have to try to retrieve your losses. But as every undergraduate economics student knows, that strategy is a disaster. Hence the principle of "sunk cost." The fact that I've lost a pile on some enterprise or investment is no reason to lose an even bigger pile. ...

All of which seems to apply to Iraq, in spades. A seemingly quick and easy military victory has turned sour. The costs, in blood and treasure, have escalated. Victory looks uncertain and distant. It seems the time has come, if not to cut and run, then surely to cut our losses. If ever the principle of sunk cost applied to warfare, it would seem to apply here.

But that instinct is wrong. Warfare is not like investment banking. At precisely the moment an economist might say to stop throwing good money after bad, a wise military strategist might say to double the bet.

Why?

Why do insurgent gangs, who have vastly smaller resources and manpower than the American soldiers they fight, continue to try to kill those soldiers? The answer is, because they believe they only have to kill a few more, and the soldiers will leave. They need not inflict a military defeat ...--all they need to do is survive until American voters decide to throw in the towel, which might happen at any moment.

All of which fits quite nicely into standard economist-like thinking. What's on the table are the ideas of time inconsistency, commitment, and credibility, concepts that are these days the centerpieces of economic policy discussions. Those ideas -- the germ of which was the basis for a recent Nobel prize -- go something like this: Let's say that it is a good goal of monetary policy to deliver low and stable inflation, but that at any point in time the central bank can pump up the economy by delivering more inflation than people expect. Though a central bank may have the best of intentions in announcing that it plans to deliver low inflation indefinitely, things will almost certainly look different when the economic going gets tough and a little monetary stimulus could help to pick things up a bit. Everyone knows, of course, that these incentives exist -- that is, that the announced policy is time inconsistent -- and will refuse to believe that low inflation will [be] delivered -- and refuse to be fooled. Actual and expected inflation will therefore be higher than they would otherwise be, to no benefit.

There are two ways out of this dilemma One is to adopt some sort of commitment device that will keep the central bankers on the straight and narrow -- a law, for example, that the monetary authorities will be removed from their positions if they renege on their low inflation promises. Such devices are often hard to implement, so an alternative is for the policymaker to earn his or her credibility by consistently following through on plans. This may be painful if that policymaker doesn't start with a lot of credibility, as it may require that monetary policy be especially tight in those times when the central bank is most expected to back off of its promise -- when the going gets tough.

Get the point? The idea of "sunk costs" mentioned in the Stuntz article does in fact mean something like "look forward, don't look back." But in policymaking, in economics as well as warfare, looking forward always means taking into account the fact that your credibility is at stake, and that the long-run costs may be very much higher if that credibility is lost.

Stuntz continues:

There is another reason economic logic does not readily apply to the fighting of wars. When running a business, one aims to invest just as much as is necessary to make the sale or manufacture the product--no less, and no more. Profit equals revenue minus cost, so minimizing cost lies at the core of wise business management.

Warfare could not be more different. Send just enough soldiers and guns and tanks to do the job, and you may soon find you have sent too few. The enemy concludes that if it can raise the marginal cost of the conflict just a bit, if casualties are a little higher or the expense a tad greater than you imagined, you'll quit the field.

Ah, but profit-maximizing does not mean minimizing cost. There is a way to express profit-maximization in terms of cost minimization, but the condition is to minimize costs subject to a particular level of output. If you send too few workers and too little capital to do the job, then what you are trying to do is not profit-maximizing -- it is infeasible. Economics says no less.

In the world of business, decisions are made at the margin: a little more invested here, a bit less there; everywhere, strive to cut waste, to spend no more than is absolutely necessary. In warfare, waste and excess are productive: They send the message that victory is inevitable, that whatever resources are needed to obtain it will be given to the task.

It seems to me that productive "waste and excess" is not waste and excess. Clearly Professor Stuntz is not suggesting that more soldiers and tanks be sent than necessary to ensure victory (at lowest long-run cost in blood and treasure). So surely Professor Stuntz is not suggesting anything with which an economist would disagree -- at least not one obliged to accept his cost and benefit analysis.

In fact, the opinion expressed in the Weekly Standard article is just the well-known position that a favorable outcome in Iraq requires the application of greater levels of force than presently exist. That is matter of great debate among military and political strategists. For my part, I would just as soon we economists be left out of it.

I'll just add a bit more on "waste and excess," and on the statement:

When running a business, one aims to invest just as much as is necessary to make the sale or manufacture the product--no less, and no more. ... Warfare could not be more different. Send just enough soldiers and guns and tanks to do the job, and you may soon find you have sent too few.

In addition to what David noted, adding a buffer (this is the "waste and excess") in the face of uncertainty is not inconsistent with economic principles when losses are asymmetric. Suppose you are going on a trip and expect costs to be $500, but they could be anywhere from $250 to $750. If there were no other sources of funds once the trip was underway, and coming up short would be life-threatening, would you take "just enough," i.e. would you take $500 in such a case? Of course not.

    Posted by on Monday, November 13, 2006 at 04:20 PM in Economics, Iraq and Afghanistan | Permalink  TrackBack (0)  Comments (7)

    TrackBack

    TrackBack URL for this entry:
    https://www.typepad.com/services/trackback/6a00d83451b33869e200d834fd95a069e2

    Listed below are links to weblogs that reference Economists Thinking Like Economists:


    Comments

    Feed You can follow this conversation by subscribing to the comment feed for this post.