Category Archives: Economic education and methods

A Question about Inequality Within States

Posted on by

Salil Mehta writes,

Let’s start by looking at this chart below. It shows the differences in state-level ratios, contrasting the typical incomes at the top 1% versus the typical incomes at the bottom 1%… the Economic Policy Institute (EPI) chart above has a clear concordance between income dispersion and the population size itself.

Pointer from Tyler Cowen.

My question is this. Suppose that we ignored the actual geography of states, and instead we produced artificial pseudo-states by taking random samples of all U.S. data. We took one sample the size of Texas, and called it pseudo-Texas. Another the size of Vermont, and called it pseudo-Vermont. etc. My guess is that the pattern of inequality across pseudo-states would look a lot like the pattern across actual states. If that is true, then there is not really much information in the pattern of inequality across states.

Sentences I Could Have Written

Posted on by

From James Manzi.

If we really could build regressions that would reliably predict what the impacts of various policies would be, it would be a powerful argument against certain political and economic freedoms. Why go to all the trouble of having a messy and expensive market, or states as laboratories of democracy, when we could just have a couple of professors build us a model?

Judith Rich Harris 1, Nurture Assumption 0

Posted on by

Kevin M. Beaver and others write,

The role of parenting in the development of criminal behavior has been the source of a vast amount of research, with the majority of studies detecting statistically significant associations between dimensions of parenting and measures of criminal involvement. An emerging group of scholars, however, has drawn attention to the methodological limitations—mainly genetic confounding—of the parental socialization literature. The current study addressed this limitation by analyzing a sample of adoptees to assess the association between 8 parenting measures and 4 criminal justice outcome measures. The results revealed very little evidence of parental socialization effects on criminal behavior before controlling for genetic confounding and no evidence of parental socialization effects on criminal involvement after controlling for genetic confounding.

Pointer from Jason Collins. A caveat is that this is an example of the statisical fallacy of using absence of evidence to imply evidence of absence.

How to Fix Economic Education

Posted on by

In this essay, I argue that it is badly broken.

Unfortunately, the conventional misrepresentation takes seriously the economics of a camping trip. There are resources to be allocated, including the time that campers have available to pitch a tent, light a fire, cook a meal, and so forth. And there are goods to be allocated, including sleeping bags, food, and water. Thus, from the conventional standpoint, there is nothing wrong with using the camping trip as a metaphor for the economy.

Read the whole thing.

Family Structure Matters

Posted on by

Timothy Taylor looks at various meta-analyses of studies of the possible causal role of the absence of a father on outcomes for the children. He quotes a meta-analysis by Sara McLanahan and others

The research base examining the longer-term effects of father absence on adult outcomes is considerably smaller, but here too we see the strongest evidence for a causal effect on adult mental health, suggesting that the psychological harms of father absence experienced during childhood persist throughout the life course. The evidence that father absence affects adult economic or family outcomes is much weaker. A handful of studies find negative effects on employment in adulthood, but there is little consistent evidence of negative effects on marriage or divorce, on income or earnings, or on college education.

Read Taylor’s whole post. He and the authors he cites are quite aware of the difficulty of distinguishing correlation from causation in this sort of research.

Andrew Gelman is Too Glib

Posted on by

Not in general, but in this post, where he writes,

I’d like to flip it around and say: If we see something statistically significant (in a non-preregistered study), we can’t say much, because garden of forking paths. But if a comparison is not statistically significant, we’ve learned that the noise is too large to distinguish any signal, and that can be important.

Pointer from Mark Thoma. My thoughts:

1. Just as an aside, economists are sometimes (often?0 guilty of treating absence of evidence as evidence of absence. For example, if you fail to reject the efficient markets hypothesis, can you treat that as evidence in favor of the EMH? Many have. Similarly, when Bob Hall could not reject the random walk model of consumer spending, he said that this was evidence in favor of rational expectations and consumption smoothing.

2. I think that a simpler way to make Gelman’s point would be to say that passing a statistical significance test is a necessary but not a sufficient condition for declaring the evidence to be persuasive. In particular, one must also address the “selection bias” problem, which is that results that pass significance tests are more likely to be written up and published than results that fail to do so.

The Omniscient Voyeur

Posted on by

Bryan Caplan writes,

In the GSS, males report an average of 14.19, women an average of 4.76.* If you mean the median, then males report a median of 3, woman a median of 2.

In my statistics course, I use this as a classic example of biased statistics. Here is why I suspect bias.

Suppose that the number of men is M and the number of women is W. Suppose that an omniscient voyeur can count all of the heterosexual relationships. Call this number n. Then the average number of sex partners for men is n/M. For women, it is n/W. Assuming that W and M are about the same, then the omniscient voyeur will know that the averages are about the same. So if the reported averages are different, then the reported averages are biased statistics.

Heritability of g

Posted on by

Plomin and Deary write,

(i) The heritability of intelligence increases from about 20% in infancy to perhaps 80% in later adulthood. (ii) Intelligence captures genetic effects on diverse cognitive and learning abilities, which correlate phenotypically about 0.30 on average but correlate genetically about 0.60 or higher. (iii) Assortative mating is greater for intelligence (spouse correlations ~0.40) than for other behavioural traits such as personality and psychopathology (~0.10) or physical traits such as height and weight (~0.20). Assortative mating pumps additive genetic variance into the population every generation, contributing to the high narrow heritability (additive genetic variance) of intelligence.

Pointer from Kyle Griffin. Read the whole thing. The authors’ discussion of Genome-wide Complex Trait Analysis, a new method of examining genetic influence, was new to me. Deary is the author of Intelligence: A Very Short Introduction, which doubles as an excellent introduction to statistical methods in research.

In my appearance in St. Louis in six weeks, I will be talking about the forces that cause stratification of earnings to accelerate. One of those factors is assortative mating based on g.

Taleb, Evolution, and GMOs

Posted on by

The podcast with Russ Roberts is here.

Taleb’s point is that “science” cannot prove that GMO’s are safe. We know that when organisms evolve using a trial-and-error process of gradual tweaking, that process is safe. But direct intervention to create GMO’s is not the same process.

Taleb says.

The big risk is what can happen when you have two things going together–which is, what happens, Soviet style, is a combination of monopoly of some plants over others, that it’s too large a system; and of course creation of other species that will themselves also be too powerful and then you may kill the GMOs or one may kill the other and you may have huge imbalances in nature. And these imbalances in nature can produce large deviations.

I think I get Taleb’s point. But it strikes me that to cause catastrophic harm, a GMO has to be both weird enough to cause unprecedented things to happen but not so weird that it fails to function as a living organism. That may be an impossible combination.

Arrow’s Impossible-to-Understand Theorem

Posted on by

Tyler Cowen writes,

Can one economist in forty properly define the “independence of irrelevant alternatives” axiom behind the Arrow Impossibility Theorem

That one I think I understand. What I do not understand is the “nondictatorship” axiom. Does it mean that ex ante no one is given the privilege of selecting for the group? Or does it mean ex post that no one’s preferences are perfectly satisfied? My impression, which may be wrong, is that it is the latter, in which case I think that nondictatorship is a misleading term for it. [UPDATE: commenters point out that I am wrong, and that it is an ex ante assumption]

As an undergraduate, I spent a lot of time trying to understand the Arrow Impossibility Theorem. I was much more fluent with it then than I am now, and yet I never thought that I had it. I think that every attempt to explain it in simple terms is wrong.

I agree with Tyler that much of modern finance is not well understood. The Portfolio Separation Theorem is very important, and yet few can explain it. My attempt is here.

One of my long-standing complaints is that nobody is sure whether aggregate demand slopes up or down. In undergraduate macro, it slopes down. In graduate macro, with inflation on the vertical axis, it slopes up (although in graduate macro nobody draws it–they just write down equations.)