Bryan Caplan  

Economic Naivete

Great Moments in Health Econom... Bye Bye Borders?...
If I had to sum up the mind-set of my Princeton econ education in one cliche, it would be, "There are those who would strain out a gnat, but at the same time swallow a camel."  Sometimes, the profs could be incredibly fussy.  Ex: Theorists would scoff at the concept of consumers' surplus in favor of compensating variation and equivalent variation.  Other times, they would spend a semester talking about politics as if the self-interested voter hypothesis were absolute truth.

When I was reading the Handbook of Population and Family Economics, I came across a section that made me feel like I was back in Fisher Hall in 1993.  It's in Van Praag and Warnaar's chapter on the cost of children.  They explain their motivation upfront:
[W]e have to ask why we need cost of children estimates.  The primary answer is political.  In modern (and not so modern) societies it is felt that families with more children need more money to reach a specific welfare level than families with fewer children... Given the fact that in most societies it is accepted that households with children should get some accommodation, the question arises how much.
Next, they explain methods of computing the answer.
In this field there are roughly two branches.  The first and oldest is that of the rather straightforward calculation of normative budgets.  This way is theoretically not sophisticated and even incorrect.  However, it is easy to explain to laymen... The second branch is much more "scientific" and hence sophisticated...  We shall start this section with the first method.

Such methods are based on normative budgets... Now the norm may be laid at a subsistence level, or a modest-but-adequate level... or at any other level.  Mostly a level of subsistence is defined by letting nutritional experts define a satisfactory diet in terms of calories and so on.  This diet is then evaluated at current prices.  This is supplemented (equally arbitrarily) by adding normative amounts for clothing, heating, etc...
I won't bore you by detailing the "sophisticated" method, because by the authors' own admission, no one other than a few academics cares!
In economic literature the [normative budget] methods are hardly mentioned, because they are scientifically not interesting or considered as naive, arbitrary, etc.  This should not obscure the fact however that these "noninteresting" methods are nearly the only ones used in practice.
If I were back in the classroom hearing this, I'm sure I would have snarked, "You began by saying that the whole reason to care about this is political.  So why bother telling us about methods that no policy-maker cares about?"  I suppose they would have replied , "Once we perfect the sophisticated approach, they'll listen."  Then I would have rolled my eyes, and the class would have moved on. 

But now I'd like to travel back in time and say, "Wanna bet?"

Comments and Sharing

COMMENTS (3 to date)
John Thacker writes:

Bryan, no matter how many times you put it in your book, link to it in your notes, or talk about it in the context of disputing the self-interested voter hypothesis, your example of abortion remains unconvincing.

Unmarried men and women are much more likely to be pro-choice than married people of even the same age, which fits with self-interest. (Married people would have an easier time dealing with a pregnancy themselves.) Unmarried men being pro-choice is perfectly self-interested, as unmarried men are interested in convincing women to be willing to have sex with them.

I roll my eyes every time you give abortion as an example in your argument.

ryan yin writes:

I thought I was the only one who felt like that in class.

John Thacker, whenever I hear someone give that explanation, I can't help wondering if it's ex post. If it'd gone the other way, would you have said you were surprised? Besides, it doesn't really do much good for an unmarried man to be pro-choice out of self-interest if the woman he's dating is pro-life.

Tom West writes:

Off topic: My complaint is that Bryan uses self-interested very narrowly.

If I earn an above-average income, it could easily be in my (perceived) self interest to have a contented society without a lot of obvious have-nots who have nothing to lose.

My giving away a larger fraction of my earnings to help achieve this means very little in isolation. (I'm not doing this out of altruism. It's *my* security/contentedness that's the issue.) But if I can elect a government that will force all of us wealthy people to contribute more to achieve my goal, then it's a definite win for me.

I do agree with the idea that group identification means a *lot* to voting patterns, but I suspect that missing this elementary point causes a lot of people to dismiss his hypothesis as a whole without giving it a fair shake.

Comments for this entry have been closed
Return to top