Meritocracy: Picking Our Elites at Random?
Given the enormous length of my Meritocracy package—over 35,000 words including sidebar, endnotes, and appendices—it’s hardly surprising that certain parts have received a great deal of discussion, while others have not.
For example, my suggestion that our top universities now operated more as hedge-funds than as educational institutions was widely distributed and discussed, as was my analysis of the strong statistical evidence pointing to the existence of “Asian quotas.” However, I more generally characterized our current elite admissions system as “a complex mixture of diversity, meritocracy, favoritism, and corruption,” and proposed a possible replacement, which has received very little attention at all.
I would fully concede that my case for the benefits of my suggested elite admissions lottery is quite speculative compared to my documentation of the flaws of the current system; but I feel the idea is sufficiently original and intriguing that it should receive a bit more discussion. Fortunately, this may now be starting to occur, with an Indian academic recently exploring the applicability of my admissions proposal to his own country’s elite universities, while being much less interested in the bulk of my article.
Should we admit students through a lottery, Dheeraj Sanghi
Another Asian-American blogger also found my suggestion quite appealing, while detailed rebuttals and alternate suggestions had previously been suggested by professors at Harvard and at George Mason University.
Should Harvard start admitting kids at random?, Andrew Gelman
College Admissions Markets, Robin Hanson/OvercomingBias
Since so many of my readers may have nodded off long before reaching twenty-five or more pages into my printed text or its web equivalent, perhaps I should briefly recapitulate the analysis presented in the last portion of my article.
My basic argument is a simple one. To the extent that academic ability and performance follow a normal distribution curve, there are relatively few individuals at the very high end, but names grow much thicker in density as we move below that top tiny slice. Therefore, a purely meritocratic system—admitting applicants in exact rank-order based on some set of objective academic measures such as standardized exams—is quite precise and accurate in selecting the topmost students but far less so below that level, with small differences generally separating the accepted from the rejected.
As a direct result, such a system places enormous pressure on students to spend weeks or years of their lives cramming to gain those few extra points necessary for success, inevitably leading to an arms-race situation, which forces all of their ambitious peers to do likewise. The result is the widespread immiseration of generations of students, often with little true academic gain, a situation already found in the societies of Japan, South Korea, and China, as well as in certain heavily immigrant school systems in our own country.
Worse still, given such enormous bunching of performance markers at the crucial threshold, it is quite likely that in many cases careless errors or other extraneous factors may play the largest role in success or failure, leading to a quasi-random admissions process for the majority of students rather than one based on true intrinsic ability and potential.
Therefore, my rather original suggestion is to cut this Gordian Knot of pure meritocracy and replace such a costly and sometimes capricious system with a simple direct lottery among those students who are below the top tier of exceptionally talented applicants but still above the academic threshold necessary to perform reasonably well and benefit from the classes at a Harvard or a Yale. And as an automatic byproduct, such an admissions lottery among the vast number of reasonably qualified applicants would ensure a truly randomized national “diversity” in the elite student bodies, rather than the fake “pseudo-diversity” produced by the corrupt and biased current admissions system.
And if a Harvard admitted (say) its top 300 students based on purest academic merit but filled the 1300 remaining students of its class by lot from among 30,000 or more reasonably qualified applicants, the selection process would require just a tiny fraction of current administrative effort and the overall class would probably be much stronger than what results from the current system, in which former admissions officers have claimed that just 5% or fewer of the students gain entrance on purely academic criteria.
Meanwhile, only the tiniest slice of America’s most absolutely talented students would even be in the running for those few hundred top slots, allowing the remaining 99.99% to relax quite a bit along with their parents, since losing a lottery is hardly a cause for personal for familial shame. And the rather more representative mix of student abilities and backgrounds at elite universities might also temper the arrogance of those who do find themselves within the winners’ circle while forcing employers to look more closely at individual records rather than merely relying upon the name of the college on the diploma. Leading businesses would necessarily draw their applicants from a far wider range of universities, thereby reducing the increasingly incestuous nature of our national governing elites.
This last point is a very serious one. My article had referenced a book by cultural anthropologist Karen Ho, suggesting that in recent years leading firms on Wall Street selected their employees only from a tiny handful of America’s most prestigious colleges, but I have now come across a much more quantitative and detailed academic journal article by Lauren Rivera, a management professor at Northwestern, making exactly the same case, and extending it to the elite portions of American business and law as well.
In effect, Rivera documents how the commanding heights of the American economy are increasingly closed to those who do not graduate from a few elite schools like Harvard, Yale, Princeton, and Stanford, while employers don’t particularly care how students actually did in college after admission. Combined with a biased admissions process, such an apparent system of elite selection seems almost tailor-made for negative national consequences, and physicist Steve Hsu had already noticed and discussed the Rivera paper when it first appeared early in 2011:
Credentialism and Elite Employment, Steve Hsu
Credentialism and Elite Performance, Steve Hsu
Now the proof of the pudding is in the eating, and if the actual real-world results produced by our national elites over the last generation or so had been reasonably good, or even merely acceptable, then we might just grit our teeth and accept the totally unfair and biased system used to pick them. But instead our national trajectory has been dreadful, though rarely brought to public attention by our elite media, whose controlling leadership is now usually drawn from much the same background and selection process.
Consider, for example, the recent research findings by Edward Wolff, a highly-regarded finance professor at NYU, which focused on the wealth distribution of American households. Not only did he show that by 2010 the combined net worth of our top 1% had grown nearly as large as that of the bottom 95%, but that the largest component of this change had been the massive impoverishment of the overwhelming majority of Americans, with the median American family having lost at least 47% of their wealth over the last few years. Even this figure is almost certainly a substantial underestimate, since home equity represents the bulk middle class family assets, and housing prices have generally continued their decline since 2010.
Some of Wolff’s other results are just as striking, demonstrating that in real terms average American wealth is now lower than it was in the late 1960s: the typical American family is poorer today than their counterpart of the Kennedy-Johnson era. Indeed, in 2010 nearly a quarter of all American households were totally destitute, with zero or negative total net worth.
And once again, the likely picture in 2013 is probably bleaker since the 2010 figures showed that over 37% of all households had net worth of under $10,000, implying that two years of further housing decline have probably pushed many more families over the brink of financial ruin. Wolff’s historical estimates only stretch back to 1962, and the 2010 rate of financial destitution was far worse than at any point since that year. So it actually appears that the “American dream” has become perhaps fifty years of net impoverishment, an absolutely astonishing development.
Wolff’s research has been reported here and there around the Internet, but has been almost totally ignored by the New York Times or the other major newspapers which I read every morning, let alone received the appropriate banner headlines, presumably on the astute Brezhevian theory that if you draw the blinds on your stalled train, you can then just pretend it is still moving. But even those who quote Wolff may have missed some of his most depressing comparative implications.
Even in a country as stable and well-documented as the U.S., performing historical wealth calculations is a major research undertaking, and I doubt that similar estimates would be easy to produce, or even necessarily meaningful, for many other countries around the world given the endless wars and political upheavals of the last half century. But estimates of the rise in real per capita GDP do exist, along with the standard Gini measure of inequality. Taken together, these provide some insight into how mean wealth might diverge from the median, and the likely trends in the latter, giving us a grim benchmark by which to evaluate relative American economic performance.
As mentioned above, it appears likely that the average American family is now poorer today than had been the case fifty years ago, and this puts them in quite rarified international company. Outside of Myanmar or Afghanistan, I cannot think of a single country located anywhere in Europe or Asia for which such a dismal economic trajectory would seem likely. So by this not unreasonable measure, the peoples of Albania and Bangladesh have probably done better over the last half century than have our own, and the same is true for most countries in Latin America.
Obviously, ordinary Americans had been the world’s most affluent people in the early 1960s, and it is much easier for the poor to catch up than for the rich to stay ahead. But if the average American is poorer today than he was fifty years ago, while the citizens of the overwhelming majority of European, Asian, and even Latin American countries have grown far richer, we must ask ourselves if our ruling elites have performed adequately in their duties. Perhaps random names picked out of the phone book would not have done any worse.
P.S. For those who might believe I am too harsh in my criticism of our elite media, allow me to provide a single illustrative example.
Last week, Floyd Norris, chief financial correspondent of the New York Times, published a year-end quantitative analysis of our comparative economic condition, carrying the soothing title “In a 5-Year Comparison, the U.S. Recovery Fares Well.” Given that reading large font headlines is easy while examining data less so, I suspect that all too many NYT readers took a quick glance, clucked their tongues at the endless gloomy naysayers of the Left and Right, and turned the page feeling much more comfortable about America’s future.
Unfortunately, the actual data presented in the charts told a somewhat different story. Norris focused on GDP growth rates during 2008-2012, and divided the major economies of the world into the good, the so-so, and the bad. Unsurprisingly, China came out as the best of the best, having grown over 50% during that period, while America was at the bottom of the so-so group, with a total growth rate of only 2.5%, although avoiding the net recessionary outcomes of Japan, France, and several other major European nations. So perhaps the headline writer had ingested a bit too much holiday cheer, but at least we ended up on the positive side of the economic ledger.
Or did we? Few Americans fully realize that our population growth rate is exceptionally high for a developed nation, being double that of China. When considering national prosperity, the relevant statistic to consider is obviously per capita GDP not raw GDP, and adjusting Norris’s figures for increase in national population would reduce our own net growth rate to negative levels, pushing us firmly into his “No Recovery” group of economic laggards, and rendering his cheerful headline even more ludicrous.
The whole notion of failing to adjust economic growth for population growth seems widespread in our media, perhaps because it helps hide our economic failures, yet it makes absolutely no sense, unless we wish to claim that India is really three times richer than Switzerland. Perhaps from the NYT’s perspective, there is a quick and easy solution to our impoverished middle class and stagnant economy, one which would deliver an immediate 8% jump in our national GDP—merge with Mexico!