Lack of Merit


I was sadly disappointed by the analysis in The Economist's article
Meritocracy in America
. The article's arguments rely on the assumption that genetic heritability of traits producing financial success is completely negligible. However, the article doesn't try to justify that assumption. In fact, unless I'm suffering from some mental block, the article never mentions that assumption. This is inexcusably silly for an article in 2005 which runs to three pages, and which repeatedly uses that assumption to use statistics about overall inheritance of success to derive conclusions about the strength of social mechanisms for inheritance of success.

Actually, I am predisposed to agree with some of the article's conclusion. It'd be pretty easy to convince me that social mobility has been decreasing (at least in the post-Jim-Crow era -- it would be harder to convince me that barriers today are as godawful as the Jim Crow laws were). It'd also be pretty easy to convince me that some reasonable measures of meritocracy could be declining too (emphatically not including skewedness of income as an indicator of nonmeritocracy or nonmobility, because I don't think professional sports, e.g., are wildly more nonmeritocratic and sclerotic than most of the economy). But many of the arguments in the article just don't support the conclusion unless, for reasons unexplained, alternative explanations are excluded.

(I usually pick up The Economist on Thursday when I visit Barnes and Noble cafe to play the game of Go. Thus, the article has been out for almost a week, and maybe someone else has dumped all over it for me. However, the only notable response I noticed in Google "economist meritocracy america" was here, and I think the article merits more dumping than that. If the definitive mo'-dumping response is out there already, then I'd appreciate feedback about how to improve my web search skillz to find such responses.)

Now, possible genetic heritability needn't necessarily involve psychological characteristics, and needn't even involve economic meritocracy or anything else which might reflects favorably on the system. Consider that The Economist itself within the last few years ran an article on "heightism", reporting a correlation between income and height. Height is strongly genetically heritable. For all that might be known by a thoughtful reader whose worldview is shaped by several years of reading The Economist, the reality behind the success statistics reported by the article might have little to do with social mechanisms for parents to favor their offspring, and much to do with the economic importance of inherited cosmetic factors like height, and/or other less-often-measured factors like complexion and hair. However, an even more glaringly obvious ignored explanation is that of inherited psychological characteristics (and/or inherited physical characteristics, like general health, which affect the mind) influencing success. (This influence needn't necessarily be through the usual meaning of "merit" either; it might be through things like predisposition to lie, ability to lie convincingly, predisposition to take risks, or predisposition to devote little energy to direct pursuit of sexual partners and to trust instead in the indirect outcome of one's single-minded pursuit of financial success.)

Discussion of inheritance of psychological characteristics is sufficiently politically incorrect that I feel little curiosity about the article's curious assumption that this effect is obviously zero. However, this political correctness at this level on this date seems woefully clueless or dishonest. Stephen Pinker's book The Blank Slate is full of evidence for the strength of inherited psychological characteristics and contains some evidence on their effect on economic outcomes. And it was quite a successful book, so it's not as though the ideas aren't widely known in 2005. The book is also sharply critical of solemnly blinkered silliness of research work silently excluding the possibility of genetic inheritance, e.g., on p. 375,

Yes, from the many useless studies that show a correlation between the behavior of parents and behavior of their biological children and conclude that the parenting shaped the child, as though there were no such thing as heredity.
and page 384,
The First Law implies that any study that measures something in parents and something in their biological children and then draws conclusions about the effects of parenting is worthless, because the correlations may simply reflect their shared genes (aggressive parents may breed aggressive children, talkative parents talkative children). But these expensive studies continue to be done and continue to be translated into parenting advice as if the heritability of all traits were zero. Perhaps [Marlon] Brando [who in an interview gave some homely observations about genetic predispositions, and was mocked for it] should be asked to serve on grant review panels.
Taking Pinker's word for the older literature (essentially none of which I've read for myself) I notice that this shoe seems to fit in 2005. Research agenda: Since there's probably little genetic relatedness between 1978 researchers and 2005 reporters, factcheckers, and editors at The Economist, evidently foot size is culturally determined! Now to look into mouth size...

Also, were I an economics journalist, not only would I be embarrassed to be caught unaware of the ideas (and characterized by the criticism) in The Blank Slate, I might also be embarrassed to be caught not reading Marginal Revolution. Their recent article pointing to an article about parent-child income correlations in adopted Korean children vs. genetically-related Korean children is not definitive, but it's suggestive enough that in the absence of a considerable body of research definitively leading to other conclusions, I think this single study might be enough all by itself to justify my characterization as "silly."

Also, a particularly dodgy-looking statistic: the article reports "nearly 70% of the sons in 1998 had remained either at the same level [by quintiles] or were doing worse [again by quintiles] than their fathers had." I expect this is technically true, but it looks like "how to lie by statistics" to me. Unless I'm confused -- which does happen, sometimes for publication... -- the absolutely random uncorrelated result would be 60%. I don't believe that the journalists expected their readers to pick up on this, or that they intended the sentence to be a masterfully concise summary of "68% [or whatever] of the sons in 1998 had remained either at the same level or were doing worse than their fathers had; in a country where outcomes were assigned randomly, the figure would be only 60%." After all, it is really very odd to summarize 60-something percent as "nearly 70%" when the baseline uncorrelated result is 60%, so that the difference in significance between 66% and 69.5% is large. It makes little sense as an attempt to be concise or informative, and much sense as a hint to help lead one's readers into the misimpression that the insignificance result is so far below 70% that the rounding is irrelevant.

As long as I'm on the subject of this article, sometimes I'm nostalgic for the old Economist and regret, e.g., how often the current Economist passes up the opportunity to make thoughtful, sometimes irreverently subversive, often quantitative observations. So, after wishing that this article met basic standards of social science and statistical cluefulness, I'll wish also they could've dropped into the article somewhere a paragraph comparing the magnitude of within-the-country inherited success factors to the magnitude of the effect of inheriting US citizenship (compared to inheriting citizenship in a random country).

On the plus (?) side, the new Economist does have "slither of society" -- a witty play on "All Snakes, No Ladders," one hopes.

(Incidentally, since this blog seems not to have a category for "Economic Journalism" at the moment, I got to enjoy some found mordant wit of my own with the "Product Review" categorization of this article. I hope that doesn't make me a bad person; I also think it might be a good idea to add such a more specific category or, I suppose, to declare the this kind of article to be inappropriate.)


Technorati provides a decent summary of bloggers' reactions to the article you cite. (Also, I shall grant you the right to create new categories; please be gentle).

My first reaction to the article was to this sentence, "The United States risks calcifying into a European-style class-based society." Specificially, I thought, "what a load of crap". The risk wasn't zero before, so the author really means to say that the risk of class stagnation (i.e. income polarization) is increasing greatly.

I'm fully willing to admit that the income share of the top income earners is increasing, although I'd more like to know their share of consumption. I'd be willing to admit that relative income inequality is stagnant--the rich are making sure their children are not out-earned by the poor, but the absolute real incomes of the poor are rising, if slower than the authors would like.

But when the article slips into "Americans are clearly mistaken if they believe they live in the world's most mobile society", I find that my disagreement is one of personal and social identity.

I simply do not identify myself with the entire United States; sue me. I don't know if I live in the most mobile society as defined by changes in quintile shares of wealth or income; I do know that I live in the most mobile society for me. I could emigrate to many other places, but my opportunities are best here. This is a real distinction, which seems to be revalidated by a lot of others in a phenomenon called immigration.

Does the author realize that since 1978 the U.S. has grown by 73 million people--an increase of 33% (73/222)?

I'd venture to say that any way you slice 'em, a large section of these people are probably getting a tremendous increase in mobility upon entry. This could also drag down mobility for many already at the bottom...


Also, the article discusses legacy preferences, as if this really makes an enormous difference. Does anybody know if the percent of entrants into the Ivy League who are "legacy students" is increasing of decreasing? I presume it's been weakening--decreasing for a while--and that it is only now that it's possible to destroy this institution...

There are 10 million mexicans living in the US, the average household income in mexico is about $22,000 (the average immigrant is probably poorer than this), the average household income in the US for Hispanics is $30,735 (the average immigrant is also probably poorer than this). There's a reason for the illegal immigration problem. Most social mobility studies ignore the foreign factor. Since most immigration in Europe is intra-European, sitting here from my armchair i'd say the US has the most social mobility among developed countries by far.

I'm willing to admit that excluding immigrants mobility has decreased, though.

The average household income in Mexico is not $22,000. According to the 2000 Profile of the Foreign-Born population of the United States, median household income of Mexicans living in the U.S. is $27,345 (it was $22,400 in 1997). I would say that the average household income in Mexico is roughly $5000, maybe double that after adjusting to purchasing power (just an estimate, if anyone would like to correct this). But I know it's nothing like $22,000. But simply saying that the average Mexican immigrant to the U.S. increases his/his family's income is not the equivalent of social mobility in the U.S. This simply indicates that the U.S. has a healthier and wealthier economy, where low-skill services are more highly rewarded than in Mexico.

I would argue that the real value of the article is make more people of the fact that income inequality is getting worse in the US. I would bet that the overwhelming majority of Americans would deny that fact. The question I would relate it to is what is the optimal rate of inequality. A certain level of inequality is good and makes everyone better off. But too much inequality is bad, and the economy is harmed. The really interesting point about the article, and the facts behind it is that the break in trend occured about 1980 -- from WW II to about 1980 income equality was relatively stable and after 1980 the share of income going to the top 1% grew from about 15% TO 22%. At a first brush one would argue that economic growth has been slower since 1980 and savings has fallen sharply. Since the strongest argument in favor of inequality is that it leads to greater savings the first brush conclusion is that greater inequality has damaged the economy.

Probably the biggest single reason for greater inequality has been the growth in CEO paypackages. The rational for paying executives more is that with greater incentives they would work harder and improve returns. But if you look at business profit growth it was roughly 7% from WW II to 1980 and has been 7% since 1980. Higher executive compensation has not lead to greater corporate profits, so why has giving management a larger share of the pie made the country as a whole better off?

There are other reasons inequality has risen, but the article does not address the question of whether or not it is a good thing.

Although higher management salaries have not necessarily helped company's bottom lines, the fact that U.S. companies have stayed competitive despite/because of these higher salaries means that, in theory, this shift was requisite in order to simply maintain profitablility. Also, I was under the impression that CEO compensation had a lot to do with the enormous risks that the board/stockholders want to give the CEO the incentive to take - part of the "golden parachute" theory, but also a reason why executive pay has risen - more creative, and inherently riskier solutions have been needed. I would also say that, because the U.S. has been able to keep up tremendous economic growth (considering its size and its prior development) during the past 25 years, this income inequality has NOT been an overall negative in the economy.

The article claims that the correlation between parent-income and child-income is increasing.

Without asking what the optimal correlation would be, or what factors besides unmeritocratic inheritance could result in a given correlation, if the correlation is increasing that would seem to support a conclusion that barriers between inherited classes are becoming more difficult to overcome through mobility.

It would seem to me that every issue brought up in The Economist article was discussed in "The Bell Curve". I believe I got my copy in 1994 or so.

Intelligence is heritable. In a meritocracy, economic stratification is based on intelligence, this is why PhDs make more money than MS's, who make more money than BS's, etc.

It should not surprise anyone that smart people have smart kids. It also should not surprise us that smart people make more money, and that the kids of smart people make more money. None of this is a bad thing.

Good comments on immigration, by the way.


Powered by Movable Type 5.02

About this Entry

This page contains a single entry by published on January 7, 2005 10:57 AM.

Hello, World was the previous entry in this blog.

Food Choice and Income is the next entry in this blog.

Find recent content on the main index or look in the archives to find all content.