David R. Henderson  

Henderson and Williams on Free to Exchange

PRINT
The Political Firm... Henderson on The New Tax Law...

My long-time friend Ben Powell, a professor of economics at Texas Tech University and the director of the Free Market Institute there, is quite a good interviewer: asking good questions and coming up with good lines apparently on the fly.

Here's a link to a recently aired "Free to Exchange" episode. The first 13 minutes are with Walter Williams and the last 11 minutes are with me.

Times below are approximate. I've highlighted only the parts from the interview with me, although I highly recommend watching the interview with Walter too.

13:00: Brief history of draft.
14:00: Why draft doesn't lower cost of having a military.
15:00: Great question in optimal tax theory.
16:00: Elvis Presley.
16:40: Bill Gates.
17:00: Not having draft makes military stronger.
18:15: Great line from Ben: "They could have got a lot more privates by letting him [Presley] not be one."
18:20: Economists' effect on the debate.
19:30: Milton Friedman's exchange with General William Westmoreland.
21:00: Ben asks: If national defense is so important, why would how we obtain manpower be an exception to the way we organize other economic activities?
21:50: Happy birthday, 18 year old.


Comments and Sharing


CATEGORIES: Labor Market , Regulation




COMMENTS (18 to date)
Trevor Adcock writes:

Actually the draft is Pareto efficient ex ante.

https://works.bepress.com/ted_bergstrom/14/

Not that there are not other arguments against it, but there is a strong economic argument for it.

Matthias Goergens writes:

Trevor, the first sentence of the abstract of that paper already puts strong caveats:

This paper shows that if workers have identical wealths, abilities, and preferences then a draft lottery is Pareto superior to a voluntary army.
David R Henderson writes:

@Matthias Goergens,
Thanks for saving my reading the paper. Those assumptions are incredibly absurd. Those are not just strong caveats: they show that the model doesn’t have any connection with reality.

Trevor Adcock writes:

Matthias I know the draft isn’t Pareto efficient in real life. But that is just because literally nothing is. Economists still argue for free trade and environmental regulations using similar arguments. This is despite such arguments being vulnerable to the exact same critique.

David R Henderson writes:

@Trevor Adcock,
I know the draft isn’t Pareto efficient in real life.
Exactly. And we economists who have been involved in analyzing the draft are talking about real life.
Economists still argue for free trade and environmental regulations using similar arguments.
Not that similar. If preferences and abilities were identical, there would be NO trade.

Mark writes:

Trevor,

"Not that there are not other arguments against it, but there is a strong economic argument for it."
I couldn't disagree more. If one accepts the axioms of classical (or Austrian) economics, economic reasoning strongly vindicates a voluntary force. If people can rationally and self-interestedly assess the current threat to their well-being (probably more importantly to enlistees, that of their family and friends), then military enlistment should correlate with threat level; meanwhile, the opportunity cost of serving in the military will lead those most productive to enlist less; some would complain that this would mean poor people disproportionately enlisting, but this isn't innately a bad thing. It's good that more factory workers than doctors join the military.

And this isn't just a typical example of the difference between practice and theory isn't really justified here. Assuming "identical wealths, abilities, and preferences" is beyond what meaningful economic theorizing would justify, imo. It's like saying, let's develop a trade theory... assuming all countries only produce the exact same product as each other. Then using said theory to question the validity of competitive advantage.

Thomas Nagle writes:

I fear that the Walter Williams interview could have been much better. First, starting out with the claim that poor black youth are not employed because there are now more generous welfare programs than in the 1940s is at best disingenuous. There is very little “welfare” available for unemployed single poor people. You cannot get “unemployment benefits” unless you have been laid off from employment. Most unemployed black youth have never had a job to be laid off from.
Most of what is spent on welfare is limite to Families with dependent children. Since welfare reform in the 1990s, people (mostly single mothers) receiving cash aid are required to work or be getting training for work. That requirement resulted in in a vast increase in labor participation among poor women (as well as a vast reduction in fatherless children per woman). Black women are not the under employed; their employment rate is higher than that of white women. The underemployment is among single black men who are getting little if anything from the government that can be converted into cash. It is unlikely that whatever little non-cash assistance they might get (some food stamps) is the cause of their low participation rate in income-generating employment.
The reasons for high levels of black male unemployment, higher than in the late 1940s, are many but certainly one has to do with our criminal justice system, where it costs money to get justice. Young people, whether black or white, experiment with drugs—particularly marijuana. If you are white, however, you are much less likely to be searched and, if you do get arrested, Daddy hires a lawyer and you get off without any police record. If you are poor and black, your stash is more likely to be discovered and you will meet your public defender one minute before you go before a judge. Few get off even when innocent. Once a poor black man gets through this nightmare, he has a “criminal” record that in many states prevents him from getting a license to work in many low-skill jobs. A “criminal” record is also a hindrance to getting many non-licensed jobs that require trust, like doorman, security guard, or cashier. Having done nothing worse than violate a bunch of conservatives’ efforts at social engineering, he is now classified along with people who have committed robberies or other real crimes.
I could go on about the government’s destruction of blighted, but relatively safe communities where people chose to live together and care for each other, replacing them with violent public housing where the government assigns neighbors who have little in common except their poverty. Or about the effects of zoning to prevent multi-family housing from being built in the places with good schools. Or about the high cost of poorly performing inner-city schools. But my point is that, as a black man, Williams had a chance to influence young blacks and liberals to reconsider the impact of government on the lives of poor people. There is a good case to be made. Unfortunately, Professor Williams’ poor choice of examples seemed better designed to reinforce conservative prejudices than to change minds about the pernicious roll of big government in the lives of poor people.

Trevor Adcock writes:

"Not that similar. If preferences and abilities were identical, there would be NO trade."

Really? Because a man cannot cut his own hair. Nor can he baby sit his own children while he goes to dine out. I can easily concoct a model where people trade despite identical preferences and abilities.

I was simply pointing your attention to the obvious economic justification for the draft. Which it would seem improper for an economist to ignore. Your interviewer say that the draft in an absurd tax system because of the high marginal tax rate; however, this model explains why it is still a good tax system despite these high marginal tax rates. He was implicitly making assumptions about job choice being continuous. This pushes back against the idea of it being costly to society. Clearly we can see it can be much less costly to society to use a draft if recruiting a large amount of manpower becomes necessary.

The point of the paper isn't to demonstrate the effect in real life its to show a model for how this effect, caused by the discreetness of job choice, leads to the optimality of lotteries in general for jobs like this.

Sure don't reinstate the draft now I would never want that, but this model shows that fighting World War 2 with an all volunteer army would have been absurdly expensive to the point where I don't think it would have been possible.

I would think an Economist would be able to understand what point a model is trying to make. Don't confuse the map with the territory. There is some information in that model even if it doesn't match up 1to1 with reality. That is why it can even usefully convey the information in the first place.

Jon Murphy writes:

@Trevor-

If people have identical preferences and abilities, there's no comparative (or absolute) advantage. Without comparative/absolute advantage, there's no reason to trade. Without any reason to trade, people consume their endowments.

Trade occurs precisely because people are different and there are gains to be had from trade. If everyone is identical, then there are no gains from trade.

Jon Murphy writes:

@Trevor

I know the draft isn’t Pareto efficient in real life. But that is just because literally nothing is. Economists still argue for free trade and environmental regulations using similar arguments. This is despite such arguments being vulnerable to the exact same critique.

Forgive me, I would have included this with my above comment to you, but I thought it sufficiently important to warrant its own post.

Economists make simplifying assumptions all the time, it is true. One has to, otherwise analysis becomes cumbersome. In these analytical models, the assumptions serve to clarify. They are not necessary. However, when expanding said analysis to policy recommendations, that is moving from the theoretical world to the real world, that's where assumptions matter.

When attempting to base a real-world policy decision off of a model outcome, we are no longer using the model analytically but descriptively (ie, we are now claiming the model is descriptive of reality). When this claim is made, the implication is that the assumptions of the model reflect reality, either perfectly or sufficiently so. At this point, questioning the legitimacy of the assumptions is critical. Ronald Coase often made this point in his discussions on "blackboard economics." It's one thing to draw a hypothetical supply/demand graph and implement various interventions to make the market more efficient (eg optimal taxation, optimal tariff, Pigouvian taxes, etc). It's a whole other beast to make that a reality (and all this is even ignoring public choice concerns).

So, given the draft is a policy prescription, and given the assumptions necessary in the paper, I think we can reasonably treat Dr. Bergstorm's paper as a neat theoretical paper, but be suspicious of his analysis as anything more than that. I think your initial claim ("Actually the draft is Pareto efficient ex ante...Not that there are not other arguments against it, but there is a strong economic argument for it,") is far stronger than the evidence you cited allows.

Jon Murphy writes:

The comments on Bill Gates at the 16:40 mark got me thinking: if Gates were drafted and subsequently killed in combat, we'd not have Microsoft and the billions of dollars worth of improvements it has made in our lives. Fortunately for us, this did not happen to Gates, but how many times did it happen? What if one of the 600,000 Americans killed in the Civil War had an idea for some kind of medicine or invention a la Gates or Rockefeller? What if one of the people killed in World War I had the mind of von Neumann? How much wealthier would we be today?

I realize my above paragraph rests on quite a few assumptions, but if even partially true, it would suggest the cost of war is far higher than we even calculate now.

Brian writes:

David said "If preferences and abilities were identical, there would be NO trade."

Jon said "If people have identical preferences and abilities, there's no comparative (or absolute) advantage. Without comparative/absolute advantage, there's no reason to trade."

I think I agree with Trevor here. Comparative advantage does not depend solely on wealth, preferences, and abilities.

A guy living in the forest might be just as good a farmer as the field guy, but the trees are in the way. The field guy might be just as good a lumberman as the forest guy, but he doesn't have any trees. Isn't it in both their interests to do what is available to each of them and then trade with each other--lumber for food?

In any case, assuming identical abilities and preferences in the case of the draft hardly seems like a major assumption. One can realistically assume that everyone prefers not to die, wants to defend their country, and dies with equal likelihood when shot with bullets. Sorry to be crass about this, but we're talking about war here. In large-scale war, the military mainly needs lots of warm bodies, some portion of which will be killed. That's what makes it so terrible. But it's also what makes abilities and preferences much less important. Brain surgeons can take a bullet as well as farm boy.

The paper should not be dismissed because you don't like the assumptions. Read it and learn about the logical content of the model. You might learn something you never knew or thought could be true.

Jon Murphy writes:

@Brian

A guy living in the forest might be just as good a farmer as the field guy, but the trees are in the way.

Then they are not identical.

David R Henderson writes:

@Trevor Adcock, Brian, and Jon Murphy,
Now that I’ve thought more about it, I think I overstated. Trevor’s haircut example is what jogged my thinking. He’s making a similar point to that of Adam Smith and, more recently, Michael Munger. You don’t need comparative advantage to get gains from trade. I erred.
Now to the issue at hand: The analysis I gave, and that other economists have given, depended heavily on the assumption, justified by reality, by the way, that when people have different preferences, proclivities, and productivities (alliteration intended), the draft is highly inefficient. So pointing to a model that has people with the same preferences and same abilities is hardly a way to argue against what other economists and I are saying. It’s kind of the like the old joke: “If I had ham, I could make ham and eggs, if I had eggs."

Brian writes:

Jon,

They are not identical in circumstance, no, but identical in "wealth, ability, and preferences," right?

Jon Murphy writes:

@Brian,

I think we're getting a tad off-topic, so let me say this one thing and I will finish up:

They are not identical in circumstance, no, but identical in "wealth, ability, and preferences," right?

I'd argue that they differ in technical ability, which is dependent on one's circumstances, and thus they are not identical.

Brian writes:

Jon,

Perhaps off topic, so to bring it back I'll come back to the paper that used that assumption. The idea behind the paper is that one can either draft people or raise a volunteer army by paying them enough to choose the army as an option. The assumption that everyone has identical wealth, ability, and preferences appears to be used only to say that for a given payment, everyone will choose the army with the same probability. It's a simplifying assumption, yes, but doesn't seem to have any essential impact on the result.

This does not appear to be any different than my example. The forest guy only chooses the lumber business because the cost is lower (the trees are there). The field guy does the same thing with farming. If the field guy were transplanted to the forest, he would choose lumber also. The people are identical, but the costs imposed by the environment may not be.

Jon Murphy writes:

@Brian

The assumption that everyone has identical wealth, ability, and preferences appears to be used only to say that for a given payment, everyone will choose the army with the same probability. It's a simplifying assumption, yes, but doesn't seem to have any essential impact on the result.

I strongly disagree. If you weaken that assumption of identity, then the conclusion disappears. So, it does have an impact on the result, which is the economists' argument.

A simple, two-person example. Let's say the whole world is my twin and I (I am really an identical twin. Yes, there's two of me running around. Scary).

If we assume both my twin and I are completely identical, then yes we will both join the army with equal probability and a lottery draft could make sense.

But let's weaken that assumption. Let's say I have a 0.1% chance of joining given my aversion to physical activity and my twin has a 99.99% of joining given my twin's love of blowing stuff up. In this case, since I have a 50% chance of being drafted, the draft is a huge cost upon society, since my preferences are overridden for a price far lower than I would agree to willing and my twin's preferences are overridden because my twin's highly prefered spot (in the Army) is now occupied by me.

You change the conditions and the whole "draft as optimal" argument disappears. They are not simplifying assumptions but rather necessary conditions.

POST A COMMENT




Return to top