From the NYT:

Closing Education Gap Will Lift Economy, a Study Finds

By PATRICIA COHEN February 2, 2015

Study after study has shown a yawning educational achievement gap between the poorest and wealthiest children in America. But what does this gap costs in terms of lost economic growth and tax revenue?

That’s what researchers at the Washington Center for Equitable Growth set out to discover in a new study that concluded the United States could ultimately enrich everybody by improving educational performance for the typical student.

When it comes to math and science scores, the United States lags most of the other 33 advanced industrialized countries that make up the Organization for Economic Cooperation and Development, ranking 24th, far behind Korea, Poland and Slovenia.

Moving up just a few notches to 19th — so that the average American score matched the O.E.C.D. average — would add 1.7 percent to the nation’s gross domestic product over the next 35 years, according to estimates by the Washington Center, a nonpartisan, liberal-leaning research group focused on narrowing inequality. That could lead to roughly $900 billion in higher government revenue, more than making up for the cost of such an effort.

White Americans outscored the OECD average on the 2012 PISA 518 to 497.

If Americans were able to match the scores reached in Canada, which ranks seventh on the O.E.C.D. scale, the United States’ gross domestic product would rise by an additional 6.7 percent, a cumulative increase of $10 trillion (after taking inflation into account) by the year 2050, the report estimated.

White Americans trailed Canadians by only 4 points. White and Asian Americans combined probably about equaled Canada.

Here’s Anatoly Karlin on 2009 PISA scores for immigrants (first and second generation) versus natives. Canada is much more selective than the U.S. and thus benefits from smarter average immigrants.

The words “immigration” or “immigrant” are not mentioned in the New York Times article.

Robert G. Lynch, an economist who wrote the Washington Center report, explained why he took the trouble to make these what-if calculations.

… One point of this exercise, Mr. Lynch explained, is to show that the added cost of improving educational achievement at the bottom would be more than made up for by the rise in economic output and tax revenue.

The study used math and science scores from the 2012 Program for International Student Assessment, a test widely used around the world for measuring and comparing educational achievement. The average combined score for the United States is 978, while the O.E.C.D. average is 995. The Canadian average is 1,044.

Eliminating the achievement gap in America would require raising the country’s average to 1,080, so that it would rank third behind South Korea (with an average score of 1,092) and Japan (with a 1,083 average). That stunning improvement, according to the center, would raise the total output in the United States by another 10 percent. Lifetime earnings of the poorest quarter would jump by 22 percent in this event.

As I’ve been saying for a long time, the conventional strategy of Closing the Gap by raising the bottom half’s performance by roughly a standard deviation while somehow not allowing the upper half to improve seems like a worse idea than my suggestion that our goal should be to raise everybody’s performance by half a standard deviation.

After all, the reality of diminishing marginal returns suggests that improving a half standard deviation is considerably less than half as difficult and more than twice as plausible as improving a full standard deviation. Moreover, my goal works under various combinations of the importance of nature and nurture, while the conventional wisdom’s goal is only possible if nurture is all important and nature is meaningless.