Bryan Caplan  

The Unsung

We Are So Different... Why Haven't They Been Fired?...
GDP is an agnostic statistic.  If someone spends money on something, it counts as GDP.  This agnosticism helps statisticians avoid controversy.  But it's hard to see any other epistemic benefit.  If we really want to measure output, we have to try to cope with some undeniable facts.  For starters:

1. Some "output" is actually destructive.  At minimum, the national "defense" of the bad countries you think justifies the national defense of all the other countries.

2. Some "output" is wasted.  At minimum, the marginal health spending that fails to improve health.

3. Some "output" doesn't really do what consumers think it does.  At minimum, astrology.

Note: None of these flaws have any definitional libertarian component.  Even if there's no good reason for tax-supported roads, existing government roads really are quite useful.  Still, coercive support is often a credible symptom of pseudo-output: If the product is really so great, why won't people spend their own money on it?

Once you start passing output through these filters, the world seems full of pseudo-output.  Lots of military, health, and education spending don't pass muster.  Neither does a lot of finance.  Or legal services.  In fact, it's arguably easier to name the main categories of "output" that aren't fake.  Goods with clear physical properties quickly come to mind:

  • Food.  People may be mistaken about food's nutritional properties.  But they're not mistaken about its basic life-preserving and hunger-assuaging power - or how much they enjoy the process of eating it.
  • Structures.  People may overlook a structure's invisible dangers, like radon.  But they're not mistaken about its comfort-enhancing power - or how aesthetically pleasing it is.
  • Transportation.  People may neglect a transport's emissions.  But they're not mistaken about how quickly and comfortably it gets them from point A to point B.

Lest this seem horribly unsubjectivist, another big category of bona fide output is:

  • Entertainment.  People may be misled by entertainment that falsely purports to be factual.  But they're not mistaken about how entertained they are.

None of this means that people in the First World don't experience incredible standards of living.  Once you stop being agnostic about GDP, though, you realize that large shares of measured output contribute little to that experience. 

If I'm right about education, for example, then educators contribute a lot less to our living standards that our share of GDP suggests.  Of course, this doesn't mean that educators' living standards are especially low.  The point, rather, is that educators consume a lot more value than they add. 

Who makes up the difference?  The unsung workers who pull more than their weight: farmers, constructive workers, transportation workers, entertainers, and many more.  Be careful, though, to define these occupations broadly: In essence if not in name, Norman Borlaug may have been the greatest farmer of all time.

COMMENTS (9 to date)
Brandon writes:

And then there's also the stuff that *isn't counted at all*, like transactions that happen on the black market or that happen outside the market, like non-cash deals between family and friends...

Mark writes:

And we can include in the wasteful spending part a big chunk of government spending. And let's not forget this great example.

Even for projects that aren't complete wastes, since there are no market prices for government spending, we use costs. Projects and services we might deem somewhat useful (like you noted, a nation's defense) it is likely to cost far more than its value to society.

Bill Woolsey writes:

Much of the things you count as wasteful are being measured by their opportunity cost anyway.

How much useful stuff could have been produced if the resources weren't used for these activities than don't produce goods and services that people buy.

If you want to measure welfare (which you keep pointing to," then GDP doesn't do it.

If you are interested in how much is being produced and how much could be produced, then GDP is a reasonable approach.

The question of whether it would be better to cut defense spending and have more private goods and services is different from the question of what happens to the productive capacity of the economy over time or how an excess demand or supply of money impacts the actual production of goods and services.

Jeff writes:

I agree with the gist of this post up to a point, but I think you are going a tad too far in some cases. For instance, you list a couple of types of expenditures for services that don't do what consumers think, like astrology. Isn't it likely that astrology just falls under the category of entertainment for many of the customers of astrologers? Even if a product or service doesn't do what it says it will, that doesn't render it worthless. On the contrary, the fact that people continue to purchase the product or service in question suggests that it has some value to them.

Likewise, if you and Robin Hanson are correct about the signalling model of education and healthcare, doesn't the signal itself still have some value? I guess in the case of spending on healthcare procedures that actually harm the patient, like unnecessary surgeries, the answer is probably "no," but for unnecessary tests, for instance, I would say there's still some value generated in terms of peace of mind for the patient and his/her family, perhaps a placebo effect, if you're lucky, satisfaction on the part of healthcare providers who enjoy their work, etc.

Ditto for education spending. Of course, having the sticky fingers of government involved so heavily in healthcare and education probably increases the waste, but I guess my point is that to write off so much output as "fake" is going a step too far for me. Some of it unquestionably is, but just how much is going to be a hard question to answer.

RPLong writes:

Prof. Caplan, I think the thrust of your point is essentially a rejection of subjective value. People who believe that value is subjective believe that "output" can only be measured by those who derive benefit, and only according to their own subjective values.

I mean, you might as well say something like, "I don't understand why brussels sprouts count toward GDP - they taste terrible!"

jb writes:

I definitely think that if entertainment is a legitimate expense, then astrology is also legitimate - it provides both entertainment and peace of mind.

And, then, when you think about it, a lot of the "wasted" money spent on education and heatlh care also provide peace of mind. Apparently it's the best kind of entertainment, given how much people are willing to spend on it.

And lastly, of course all that money spent on defense should be considered entertainment. Haven't you ever heard of "security theater"?

Daublin writes:

Very well written, Bryan, and a point that deserves much more attention. If one wants to think about national economic policy, then one really needs some sort of grasp of what output really is valuable.

I would be interested if you or readers can point to attempts to do this. I'm sure there have been.

To the posters opposing the "astrology" example, my reading of it is that there is a component for entertainment and a component for improving one's life through better information. Similarly, for lottery tickets, there is a component for the thrill of having a ticket, and there is a component for the possibility of actually winning the lottery.

In both cases, I think it is fair to say that many consumers of these products are getting less value than they think they are, and therefore the price tags overestimate the true value. They are getting the entertainment component for sure, but they are over-valuing the non-entertainment components.

ajb writes:

Once you start worrying about opportunity cost and destructive value then you have to start judging interpersonal utility. For example, what about products that offend people enough that they are willing to pay to avoid it or are willing to cause harm? (Think of imagery that enhances certain identities or causes others to react destructively or violently to other imagery). Or what about people who leave a community? The lost physical value (say of a home) may underestimate the subjective utility of a loss. (It could be for reasons as banal as poor schools or as controversial as aversion to different groups). And how do we judge benefits of competitive politics vs. the disutility to people of competing preferences? Some people might rather live in a community with lower economic growth but monolithic worldviews, with NO possibility of competition.

andy writes:

Regarding radon - in this document the (unknown) author claims that Radon studies were done mostly on miners and then extrapolated. The author thinks that the 'mine' atmosphere contains too many polutants to allow such interpolation and that it's practically impossible to do a good population study on radon exposure.

Comments for this entry have been closed
Return to top