Wednesday, December 05, 2007

Ah Something I actually know something about.

Kevin Drum has 2 questions

He notes

The average science score of U.S. 15-year-olds lagged that of students in 16 of 30 [industrialized] countries....U.S. students were further behind in math, trailing counterparts in 23 countries.

and asks

if that's the case, I asked, why does the American economy continue to do so well? ... If American kids are getting mediocre educations, and if they've been getting these mediocre educations for several decades now, shouldn't this have long since shown up in the business world, the tech world, and the financial world?

A very striking feature of the data reported in an earlier round of PISA and in similar studies is that, while average performance is fair to middling in the USA inequality across students of performance is enormous in the USA compared to other countries.

The top 10% of US students do very well on the tests. Sad to say, it might be that they are the ones relevant to "the business world, the tech world, and the financial world."

One would expect that the US has enormous income inequality too. Check. We are used to the idea that one should not choose a slightly higher average in exchange for much higher inequality.

However, the measure which would matter if we didn't care about inequality isn't necessarily the average. With formal education it could be the effect on the performance of professional, managerial and technical workers who were mostly top students in junior high. I am glad to say that this is not true outside of the range of basic education in OECD countries. Studies of a broader cross section of countries show that enrollment in primary and lower secondary school is strongly correlated with subsequent growth. In OECD countries enrollment rates are all at the maximum.

Also thinking about the cause of the immense inequality in US educational outcomes might help one understand why the effects on innovation and GNP growth are ambiguous.
The US has unequal education partly because of unequal funding and partly because of decentralised curriculum decisions. As George Bush and Ted Kennedy agree, this is very unfair to children victimized by the soft bigotry of low expectations (those gentlemen are experts on the subject although Kennedy is, at least, hard working).
However, a uniform national curriculum can reduce the variety of mental approaches to problems. I am often struck by the fact that Italians have read lots of books (compared to people in the USA) and that they are the same books. That there are things that everyone knows and everyone knows all reasonable people agree on which sound crazy to me (obviously this is partly culture shock).

Consider, for example, Japan. Excellent performance on standardized tests, but some decades ago (just before the fall) it seemed like everyone in Japan simultaneously become concerned that the love of consensus prevented original thinking in Japan. The claim might be nonsense, but the fact that many people made it at the same time (months after a wave of Japan is number 1 triumphalism) lead me to think that maybe they all had made a valid point simultaneously.

Of course it is also true that, while education is important for economic performance, it is not the only thing. Even US liberals who think they are open to regulation and redistribution and such would, if informed how things are done in Europe without being told that real countries are being discussed, probably imagine that the unfortunate economies would be in total collapse. Instead they are doing OK.

Also economic power and economic performance aren't the same thing. US leadership has something to do with people working very long hours in the USA. That is a choice not a result of greater ability or efficiency.

Finally, I have a wish of a hope of an explanation of why US students are not dumb but perform poorly on the tests. In the USA there are various tests which, like PISA , are used to follow general educational progress and not to rank students. Maybe kids in the USA are more aware of the fact that they have no material incentive to answer PISA questions correctly. That is maybe more students decide to not bother with a test which will have no effect on their grades.

As to my claim to know something, see Giorgina Brown, John Micklewright, Sylke V Schnepf, and Robert J. Waldmann (2007)"Cross National Surveys of Learning Achievement: How Robust are the Findings," The Journal of the Royal Statistical Society Series A vol 170, pp. 623-646.

also IZA wp 1652

No comments: