Bad Political Argumentation

Yesterday I read a book of essays by Alfie Kohn, called What Does It Mean to Be Well Educated?

Kohn makes a particular argument that I found to epitomize bad argumentation, in a way that is depressingly common among partisans on all sides of virtually every social and political issue. Here's the formula:

Side 1: "I think we should adopt Strategy A, with the goal of causing Result B."

Side 2, with the bad argument: "I don't think Strategy A would cause Result B. In fact, I think it's so clear that Strategy A won't cause Result B, and that it will instead cause horrible Result C, that I'll accuse Side 1 of lying when they say they want Result B in the first place. Thus, if they're proposing Strategy A, it's really because of a sinister and maleficent plot to bring about Result C."

Now this isn't necessarily a bad argument: It could theoretically be true that Side 1 is being dishonest when it claims to want Result B. But in most cases, Side 2 is just inventing the maleficence attributed to Side 1. That is, Side 2 wrongly assumes that no one can honestly disagree with Side 2's assessment of whether the means achieve the ends -- and that therefore anyone who claims Strategy A would cause Result B isn't just mistaken, but is pursuing another dishonest and sinister end.

Well, this is all rather abstract, so let me type out the egregiously bad passage from Alfie Kohn's book:

Knowing a lot of stuff may seem harmless, albeit insufficient, but the problem is that efforts to shape schooling around this goal, dressed up with pretentious labels like "cultural literacy," have the effect of taking time away from more meaningful objectives, such as knowing how to think. . . . .

The number of people who do, in fact, confuse the possession of a storehouse of knowledge with being "smart" . . . is testament to the naive appeal that such a model holds. But there are also political implications to be considered here. To emphasize the importance of absorbing a pile of information is to support a larger worldview that sees the primary purpose of education as reproducing our current culture. It is probably not a coincidence that a Core Knowledge model wins rave reviews from Phyllis Schlafly's Eagle Forum (and other conservative Christian groups) as well as from the likes of Investor's Business Daily. To be sure, not every individual who favors this approach is a right-winger [Kohn doesn't mention that the founder of Core Knowledge himself is anything but a right-winger], but defining the notion of educational mastery in terms of the number of facts one can recall is well suited to the task of preserving the status quo. By contrast, consider Dewey's suggestion that an educated person is one who has "gained the power of reflective attention, the power to hold problems, questions, before the mind."

Leave aside, for the moment, the fact that this argument is completely wrong, from start to finish. Cognitive psychologists have known for a long time that creative and critical thinking are impossible unless you're working with a huge base of factual knowledge.

If you don't believe me, try to come up with a critical assessment -- generated by your own brain, not quoted from elsewhere -- of ten-dimensional superstring theory, or Thomas Gold's abiogenic theory of oil formation. Can't do it? Neither can I, and it's precisely because I don't know enough facts about quantum physics and oil formation to be able even to understand what those theories say, let alone figure out how consistent they are with the evidence.

It's mystifying to me that anyone thinks it possible for students to be taught to "hold problems . . . before the mind" without knowing facts about the subject in question. Would it really be productive to ask students to think critically on the following problem: "Was the Civil War a good idea?," if they don't know who Lincoln was, have never heard about Reconstruction or the Missouri Compromise, and generally haven't learned lots of facts about what caused the Civil War, what happened during the Civil War, and what results occurred afterwards? And make no mistake: Americans are often abysmally ignorant of the most elementary facts (consider the 2007 poll finding that 31% of Americans couldn't even name Cheney when asked who was the Vice-President).

But as I say, leave all that aside. The real problem with Alfie Kohn's argument here is not just that he's wrong, but that he assumes that he's so obviously and indisputably right that no one could possibly disagree with him. To Kohn, trying to get kids to learn facts somehow prevents or precludes critical thinking -- and this is so obvious that no one could really think that it's important for students to learn facts. If anyone wants students to know that the earth revolves around the sun rather than vice versa (as 18% of Americans think), it can only be because they hate critical thinking, and are just a bunch of mean right-wingers trying to oppress people by preserving the "status quo."

Kohn's argument is egregiously bad. But similar arguments can be found everywhere, if you look. Environmentalists couldn't really believe that global warming is a danger; therefore their real motive is to destroy the free market. Education reformers couldn't really believe that giving poor people a choice among schools is a good idea; therefore their real aim is to destroy public schools. Examples abound.

Celebrity Video Pledge

This celebrity video pledge to Obama is getting lots of press. The part I found most touching -- aside from Jason Bateman's pledge to flush only on number two -- was when Cameron Diaz joined in the pledge to get to know her neighbors by saying, "I'm gonna give them a smile." Well, it looks like Diaz's neighbor Paris Hilton recently moved, but I'm sure the current neighbors are in desperate need of Diaz's assistance.

Citation Formats

I think someone has actually managed to come up with a few citation formats that haven't been discussed ad nauseam by the Harvard Bluebook editors.

(Via Crooked Timber).

Education Week Report

I have a guest post at Jay Greene's blog, critiquing a recent report that attempts to rank education in all the states.

Prime Obsession, and Robert Shiller

Two books that I recently read:

1) Prime Obsession, by National Review's John Derbyshire. Prime Obsession has a mix of chapters: about half address the biography and background of Bernhard Riemann, the eminent German mathematician, as well as many of his forebears, contemporaries, and successors. The other half of the chapters are a layman's explanation, starting with some elementary math, of the Riemann Hypothesis.

It's a brilliant book. Derbyshire has a wonderful knack of explaining complicated mathematical concepts from the ground up. He does this so well that even a non-mathematician has at least a glimmering of a thrill when, in an anecdote, the physicist Freeman Dyson is told of the statistical distribution for the non-trivial zeros of the Riemann zeta function, and his response is, "That's the form factor for the pair correlation of eigenvalues of random Hermitian matrices!" What this means (so far as I understand it) is that the distribution of prime numbers has a completely unexpected parallel to the energy levels observed in quantum mechanical experiments. (It's a bit dismaying, to be sure, when after a chapter written in language like that, the next chapter begins with the warning: "You might find the following few sections challenging." But Derbyshire is up to the task of making it all comprehensible.)

2) Robert Shiller's The New Financial Order: Risk in the 21st Century. Shiller is a clever and creative economic thinker, and he presents several bold ideas in this book (including standardized units of currency, income tax schedules based on inequality, and much more).

I was intrigued by his idea of insurance for "livelihoods." Shiller provides an example of a young man thinking of a career in biochemistry. The risk is that his highly specialized knowledge might be displaced or become unusable in the employment market, for some reason. So there's a lot of uncertainty that might cause talented individuals to forego a socially valuable career. Shiller's answer is a livelihood insurance policy that "would be designed to pay him a regular supplement to his income over the years in the event of a decline."

The problem with such insurance plans is two-fold:

a) Insurance companies wouldn't have enough information to price and sell such plans for many careers, particularly new careers as to which there is no historical or national data on income patterns;

b) Even where such insurance plans are possible, the very people who would most benefit would also be least likely to be able to afford a policy. Journalists and GM workers would love to have an insurance policy right now that paid them a lifetime income in the event of job loss. But for that very reason, if such policies were sold at all, they would be exorbitantly expensive for journalists and GM workers (for the same reason that it would be highly expensive for a death row inmate to buy life insurance).

Now if such insurance policies were publicly subsidized in some fashion, that would be a different matter. But then Shiller's argument devolves into a case for a government-sponsored minimum income, except one that varies by career. While Shiller does point out that because of moral hazard, livelihood insurance policies might have to pay only for aggregate drops in the average income for a certain profession, there would still be some moral hazard (I think) for people selecting whether to enter a career where the government's minimum income was higher.

New Study: Free Health Care Didn't Improve Outcomes

A new study from a randomized experiment in Ghana showed that when primary health care was made free, people used health care more often but weren't any healthier:

Delays in accessing care for malaria and other diseases can lead to disease progression, and user fees are a known barrier to accessing health care. Governments are introducing free health care to improve health outcomes. Free health care affects treatment seeking, and it is therefore assumed to lead to improved health outcomes, but there is no direct trial evidence of the impact of removing out-of-pocket payments on health outcomes in developing countries. This trial was designed to test the impact of free health care on health outcomes directly.

Methods and Findings

2,194 households containing 2,592 Ghanaian children under 5 y old were randomised into a prepayment scheme allowing free primary care including drugs, or to a control group whose families paid user fees for health care (normal practice); 165 children whose families had previously paid to enrol in the prepayment scheme formed an observational arm. The primary outcome was moderate anaemia (haemoglobin [Hb] < 8 g/dl); major secondary outcomes were health care utilisation, severe anaemia, and mortality. At baseline the randomised groups were similar. Introducing free primary health care altered the health care seeking behaviour of households; those randomised to the intervention arm used formal health care more and nonformal care less than the control group. Introducing free primary health care did not lead to any measurable difference in any health outcome.

The authors include the obvious caution that their results are not generalizable to other countries. Still, the results do suggest that researchers shouldn't automatically assume that giving people more health care leads ineluctably to greater health:

Although the findings of this trial may not be generalizable to other countries, they nevertheless raise the possibility that providing free health care might not be the most cost-effective way of improving health in all developing countries. Importantly, they also suggest that changes in health care utilization should not be used in future trials as a proxy measure of improvements in health.

On a broader note, Abhijit Banerjee and Esther Duflo have a very interesting paper here discussing a wide range of randomized trials as to all sorts of governmental policies.

UPDATE: This post seems to have caught some attention: Megan McArdle, The American Scene, Overcoming Bias, Marginal Revolution.

Elena Kagan to Serve as Solicitor General

Congratulations to Elena Kagan -- currently dean of Harvard Law School (and teacher of my administrative law class) -- on being named as the nominee for Solicitor General. I'm reminded of the previous Harvard Law professors who turned down the chance to be nominated as Solicitor General -- Felix Frankfurter turned down Roosevelt, and Paul Freund turned down Kennedy. In fact, I seem to recall reading at one point that Frankfurter (by then on the Supreme Court) advised Freund that a Harvard Law professor should view any government position except the Supreme Court as a demotion. Elena (like Charles Fried) apparently doesn't share that rather self-serving view.

UPDATE: Despite Elena Kagan's lack of appellate experience, I would anticipate -- based on her class -- that she would be superb. She was one of the top two or three professors I had in law school. Unlike most other professors, she taught in a pure Socratic style: lots of random questioning of students, with no opportunity for students to pass.

No matter what a student said (ranging from tough and highly perceptive questions to complete irrelevancies), or whether the student was prepared to say anything at all, Elena Kagan had an unbelievable ability to take the student's questions or answers and then weave it all into a productive classroom discussion. Granted, the Supreme Court's Justices are a bit more prepared and intense than the typical classroom, but I'd guess that Elena Kagan would be quite capable at taking the Justices' questions and, on the fly, weaving her answer into a skillful presentation of whatever argument she was presenting. And incidentally, it was John Roberts' skill at that same thing that made various Supreme Court Justices say (prior to his nomination) that he was the best Supreme Court litigator in the country.

Losing Sleep Makes You Drunk

Well, not literally drunk, but you might as well be. A study in Nature* found that merely being awake for 17 hours -- which any of us might experience on a daily basis -- decreases hand-eye coordination the same as having a .05 blood alcohol concentration:

Forty subjects participated in two counterbalanced experiments. In one they were kept awake for 28 hours (from 8:00 until 12:00 the following day), and in the other they were asked to consume 10-15g alcohol at 30-min intervals from 8:00 until their mean blood alcohol concentration reached 0.10%. We measured cognitive psychomotor performance at half-hourly intervals using a computer-administered test of hand-eye coordination (an unpredictable tracking task). Results are expressed as a percentage of performance at the start of the session.

* * *

Equating the two rates at which performance declined (percentage decline per hour of wakefulness and percentage decline with change in blood alcohol concentration), we calculated that the performance decrement for each hour of wakefulness between 10 and 26 hours was equivalent to the performance decrement observed with a 0.004% rise in blood alcohol concentration. Therefore, after 17 hours of sustained wakefulness (3:00) cognitive psychomotor performance decreased to a level equivalent to the performance impairment observed at a blood alcohol concentration of 0.05%. This is the proscribed level of alcohol intoxication in many western industrialized countries. After 24 hours of sustained wakefulness (8:00) cognitive psychomotor performance decreased to a level equivalent to the performance deficit observed at a blood alcohol concentration of roughly 0.10%.

* Drew Dawson and Kathryn Reid, "Fatigue, Alcohol and Performance Impairment," Nature, vol. 388, July-August 1997.

Blood Sugar Undermines Human Memory

The New York Times reports on new research showing how even "moderately elevated" blood glucose reduces the ability to form memories, and -- because blood glucose regulation declines with age -- this may help explain why our brains decline as we age.

Oddly, the only recommendation in the article is this:

Since glucose regulation is improved with physical activity, Dr. Small said, “We have a behavioral recommendation — physical exercise.”

I'm all in favor of exercise, to be sure, but I also immediately thought of diet. Studies have shown that low carbohydrate diets dramatically lowered blood glucose (see here, here, and here). To be sure, there's one small study suggesting that "severe carbohydrate restriction" somewhat impairs memory in the very short term (i.e., one week). It would be interesting (albeit practically impossible at this point) for someone to do a study on how cognition is affected after a few decades of a carbohydrate-restricted diet.

Lone Letter-Writers and the SEC

A long and interesting article by Michael Lewis and David Einhorn making a good case that the SEC has suffered from regulatory capture. One thing in particular jumped out at me, because it resembled an earlier NY Times story on a 2004 SEC rule change:

To that end consider the strange story of Harry Markopolos. Mr. Markopolos is the former investment officer with Rampart Investment Management in Boston who, for nine years, tried to explain to the Securities and Exchange Commission that Bernard L. Madoff couldn’t be anything other than a fraud. Mr. Madoff’s investment performance, given his stated strategy, was not merely improbable but mathematically impossible. And so, Mr. Markopolos reasoned, Bernard Madoff must be doing something other than what he said he was doing.

* * *

Harry Markopolos sent his report to the S.E.C. on Nov. 7, 2005 — more than three years before Mr. Madoff was finally exposed — but he had been trying to explain the fraud to them since 1999. He had no direct financial interest in exposing Mr. Madoff — he wasn’t an unhappy investor or a disgruntled employee. There was no way to short shares in Madoff Securities, and so Mr. Markopolos could not have made money directly from Mr. Madoff’s failure. To judge from his letter, Harry Markopolos anticipated mainly downsides for himself: he declined to put his name on it for fear of what might happen to him and his family if anyone found out he had written it. And yet the S.E.C.’s cursory investigation of Mr. Madoff pronounced him free of fraud.

Similarly, in that earlier NY Times story, the SEC in 2004 allowed the big investment banks to be more highly leveraged. No commenters opposed the rule, except for one guy:

A lone voice of dissent in the 2004 proceeding came from a software consultant from Valparaiso, Ind., who said the computer models run by the firms — which the regulators would be relying on — could not anticipate moments of severe market turbulence.

“With the stroke of a pen, capital requirements are removed!” the consultant, Leonard D. Bole, wrote to the commission on Jan. 22, 2004. “Has the trading environment changed sufficiently since 1997, when the current requirements were enacted, that the commission is confident that current requirements in examples such as these can be disregarded?”

He said that similar computer standards had failed to protect Long-Term Capital Management, the hedge fund that collapsed in 1998, and could not protect companies from the market plunge of October 1987.

Mr. Bole, who earned a master’s degree in business administration at the University of Chicago, helps write computer programs that financial institutions use to meet capital requirements.

He said in a recent interview that he was never called by anyone from the commission.

“I’m a little guy in the land of giants,” he said. “I thought that the reduction in capital was rather dramatic.”

To be sure, regulatory agencies are often inundated with crank letters from random individuals -- which inevitably means that such letters are received with a presumption of irrelevance -- and it's easy in retrospect to point out warning signs that an agency should have seen (check out the over 168,000 comments received in the FCC's media consolidation proceeding, Docket 06-121).

Still, in both of these cases, the SEC brushed off a letter that was, on its face, rational and credible, and that wouldn't have been lost in a stack of thousands of other comments. And it's rather remarkable that the supposed regulatory experts are being shown up, time and again, by a single person who is simply paying closer attention to what's going on.

The Supreme Court's Basketball Games

I'm surprised none of the legal bloggers have picked up on this hilarious essay (via Andrew Gelman). It's apparently from a McSweeney's book; it's by Jim Stallard, and is titled, "No Justice, No Foul: Everything You Didn't Know That You Were Afraid To Know About the Supreme Court."

The author's obviously tongue-in-cheek thesis: That all of the Supreme Court's 5-4 decisions are decided on the basketball court. The essay consists of elaborate and laugh-out-loud descriptions of the basketball games by which Earl Warren, Harry Blackmun, etc. won this or that controversial decision.

The opening paragraphs:

Whenever I hear some historian on PBS prattling about the Supreme Court, I have to step outside for air. I know it's a matter of seconds before the stock phrases -- judicial review, legal precedent, activist court -- will start rolling out, and I'll feel my blood coming to a boil as I hear the scamming of yet another generation.

Are you sitting down? Everything you were taught about the Supreme Court and its decisions is bunk. For most of the 19th Century and all of the 20th, our biggest, most far-reaching legal decisions have been decided not by careful examination of facts and reference to precedent but by contests of game and sport between the justices. The games varied through the years -- cribbage, chess, horseshoes, darts -- even a brief, disastrous flirtation with polo. (Now do you understand Plessy v. Ferguson?) But ever since 1923, basketball has been the only game, and as the years rolled by and the decisions came down, the whole thing has settled nicely into place.
* * *
Holmes brought the idea back to Washington and pitched it to Chief Justice Taft. The corpulent chief had been lobbying for Greco-Roman wrestling, but he was starting to realize none of his colleagues would go for a sport in which they might be killed. (The Fatty Arbuckle incident was fresh on everyone's minds). Taft finally agreed that basketball offered a superior form of jurisprudence.

Read the whole thing.

Clayton Christensen's Theory of Disruptive Innovations in Education

I recently read Harvard Business professor Clayton Christensen's book Disrupting Class: How Disruptive Innovation Will Change the Way the World Learns, written with Michael Horn and Curtis Johnson. The book extends Christensen's famous theory of disruptive innovation to education (just as another book does as to healthcare).

The "disruption" will be computers. The problem with education today, say Christensen/Horn/Johnson, is that it is too standardized and cookie-cutter, even though students all "learn in different ways." Computers will be more tailored to students' learning styles and capabilities.

To amplify, Christensen/Horn/Johnson give credence to Howard Gardner's theory of multiple intelligences (such as linguistic, logical, spatial, kinesthetic, musical, etc.), which they rapidly conflate with the claim that everyone has a different learning style. Their main evidence as to learning styles consists of an anecdote about a sixth-grader who didn't read or write, until her perceptive teacher realized that she was a kinesthetic learner and needed to create a special "dance" for each letter of the alphabet, whereupon the girl "practiced all her spelling words through dancing." (I think this anecdote is of limited generalizability at best.) Thus, they claim that education should be customized to everyone's learning style, something that can allegedly be accomplished via computer programs.

But as noted educational psychologist Dan Willingham of Virginia has pointed out, there is really no evidence that "learning styles" exist, not in the strong sense that is often claimed. Yes, some people are indeed better at learning kinesthetic skills (such as dancing or sports) than they are at math. That does not mean that if math is taught with a kinesthetic element, such people will learn math better than anyone else. When a kinesthetic demonstration is helpful -- such as using blocks to teach kindergarteners the concept of subtraction, or conducting a chemistry experiment -- it will be helpful for everybody, not just for so-called "kinesthetic learners."

On top of that, even if people really did have different "learning styles," and even if it made sense to think of subjects such as history or literature being taught in a "kinesthetic" manner (or perhaps via "music" or "math," for people gifted in those areas), I'm dubious that computers would be suited for delivering such instruction. Computers aren't really kinesthetic at all -- the motions of clicking a mouse or typing on a keyboard don't have much of anything to do with the actual content of a math or physics problem that might be presented. Nor, for that matter, would a typical student find a computer of much use in dancing the letters of the alphabet.

All of that said, I do think that Christensen and his co-authors are right that computer-delivered instruction can have quite significant benefits in certain subjects and in certain respects. The most obvious example would be a math program that provides math questions based on each individual student's past pattern of right/wrong answers, thus providing instruction that is perfectly tailored to what that student needs to practice. Such programs already exist, and a new study by a few respected economists provides evidence that they are very effective.

On the whole, it's a valuable and thought-provoking book.