It looks like the methodology that they are comparing is very different. The BusinessWeek article cites things like increasing numbers of people achieving in t

It looks like the methodology that they are comparing is very different. The BusinessWeek article cites things like increasing numbers of people achieving in those subject areas. The international surveys look at the average achievement in those subject areas. To simplistically illustrate with made up facts: Let's say BusinessWeek finds that in 1990, 5% of students achieve an excellent result in math and in 2000 10% of students achieve an excellent result in math. That's quite an increase. Let's say that there were 100 students and so the "A" category increased from 5 to 10 students. Awesome! However, that doesn't tell us what the average student is achieving. The OECD and others usually look at the average (mean or median) achievement. It is entirely possible that elite education is increasing while average education is decreasing. To put this into numbers: --1990-----------------2000-- A:5 A:10 B:10 B:10 C:50 C:30 D:25 D:25 F:10 F:25 If you look at those two distributions, you can come to different conclusions based on your methodology. By mean average, the US is declining from a 1.75 score to 1.55. However, the US also has more achieving really excellent scores. One thing to note about a study like this is that the United States tends to have more immigrants than other nations. Immigrants are fine people, but they face disadvantages in integrating into a new society. Often times there are language and cultural barriers that slow learning as, well, part of their learning is spent overcoming those barriers. This could drag down average US scores in academics when comparing them to countries that don't have immigrants that might spend a lot more time learning the spoken language over studying math. If you're dealing with two people of equal ability, but one has to study a second language to become fluent first, that person's achievement won't be as great (or at least will be delayed as they have taken time to study something else). This isn't something bad about immigrants, but it does say that there might be biases in the results of a study if one group includes more immigrants. There are also much greater selection biases in other countries. Take France vs the US as an example. In the US, most students attend a traditional high-school. Yes, there are vocational schools, but those aren't stressed. So, in the US, an academic test is likely to cover a decently random sampling of the population at any given age. In France, secondary education is very segregated. As such, it is possible that there is quite a selection bias for the exams between those that are studying sciences and math and those that are doing vocational studies in who takes the exam. I do think the US should be doing more to promote math, science and engineering, but I wouldn't be so quick to say that the US is fundamentally behind. The US education system does face challenges that many other systems don't - such as absorbing a greater number of immigrants, dealing with historical prejudices that have disenfranchised and disadvantaged many, and trying to academically (vs vocationally) educate the vast majority of our children. I'm not saying that anything in here is correct. I haven't studied international education enough to make any conclusions. This is simply meant to raise the possibility that the statistics aren't the full picture.