Outsourcing leads to declining scientific output

[jay gw]

Over the last two decades, the European Union and Asia Pacific nations have steadily increased their world share of the journal literature in the physical sciences. In fact, these regions now surpass the United States in output of papers in physics, chemistry, and materials science, according to a new Science Watch survey, based on the National Science Indicators database 1981-98, from the Institute for Scientific Information.

The field of physics (encompassing general physics, condensed-matter physics, high-energy physics, and optics and acoustics) offers a particularly striking example of the rise in output by EU and AP nations and the simultaneous U.S. decline.

From its lead position in 1981, the U.S. share of physics papers began its slide in the early 1990s, as the EU's share continued upward. By 1996, the steadily upward-trending AP nations had surpassed the falling U.S. in world share.

By 1998, the EU nations accounted for roughly 37% of published physics papers, compared to the AP's 29%, whereas the U.S. share hit a low of approximately 26%.

As science becomes ever more global as developing nations begin to take their place alongside developed nations one might expect that world-share statistics might start to converge.

After all, world share is a zero-sum notion: if one gains share, another must lose.

But the U.S. decline in world share in physics is accompanied by a decline in actual number of papers as well.

Having reached a peak of 22,971 papers in physics in 1994, the number has declined slightly to 22,159 in 1998. Meanwhile, in 1998, the EU and AP nations reached all-time highs in output of physics papers: 32,178 and 24,941 respectively.

http://www.sciencewatch.com/sept-oct99/sw_sept-oct99_page1.htm

A “troubling decline” in the number of US citizens training to become scientists and engineers is creating an “emerging and critical problem” for the United States, according to a new report by the by the National Science Board, the independent government panel that oversees and sets policy for the National Science Foundation (NSF).

If current trends continue, the number of US citizens capable of filling a growing number of challenging science and engineering (S&E) jobs will be only level, at best, says the report, An Emerging and Critical Problem of the Science and Engineering Labor Force. “These trends threaten the economic welfare and security of our country,” the NSB writes.

http://www.biomedcentral.com/news/20040506/02
______

As the manufacturing in the US/EU goes to Asia and elsewhere, one of the worst consequences is that the research and development that is linked with it goes there too.

People like the above science associations complaining about the emerging problem of lack of skilled labor don't seem to understand that training and jobs are NOT separate things. They are side by side.

The US has been gradually declining in scientific output for the last 20 or so years. The number of new engineers is far lower than it was, and about 80 percent of the graduates are not Americans - they are Asians on student visas.

The average university student will look at the economy, make predictions, and adjust their studies accordingly. As the number of manufacturers outsourcing has increased, the number of native science, math and engineering graduates has declined and channelled into other areas like education/teaching and general business.

So what? Well, if the capacity itself for creating technology declines, nation X becomes dependent on others for innovations. The trade is not balanced, because the only goods to trade are service type goods, which have limits since they are performed on site, many times in front of the customer.

 

[drkitten]

Over the last two decades, the European Union and Asia Pacific nations have steadily increased their world share of the journal literature in the physical sciences. In fact, these regions now surpass the United States in output of papers in physics, chemistry, and materials science, according to a new Science Watch survey, based on the National Science Indicators database 1981-98, from the Institute for Scientific Information.

So what? A smaller share of a bigger pie is better than a bigger share of a smaller pie. Would you rather have 20 percent of a thousand pounds, of 50 percent of fifty pounds?

 

[jay gw]

So what? A smaller share of a bigger pie is better than a bigger share of a smaller pie. Would you rather have 20 percent of a thousand pounds, of 50 percent of fifty pounds?

Maybe you're not understanding the points about:

* The United State is the only developed country that is increasing in population size. The EU and Japan are decreasing in size.
* However, the US is decreasing in the number of new science graduates, and decreasing in the number of papers/other measures of output.

Why would a nation's population be increasing, but the amount of scientific output decreasing? What my op stated is that manufacturing and engineering, science and math are not separate. They are indivisible.

It doesn't matter whether other countries are producing more, and the relative percent of the US share declines. It's the absolute numbers that are declining.

If historically, up to that last 20 or so years, you produce 50 papers/measure of output per 10,000 people, and every year that starts to decline, and now you produce 40 per 10,000 there must be a reason for it.

The reason I put forth in the op is that if the potential labor force knows beforehand that an industry is becoming obsolete, why would anyone study it?

None of the developments are isolated.

 

[drkitten]

* The United State is the only developed country that is increasing in population size. The EU and Japan are decreasing in size.
* However, the US is decreasing in the number of new science graduates, and decreasing in the number of papers/other measures of output.

Not significantly. According to your numbers, the number of papers "declined" from 22.971 to 22.159. That looks like noise to me.

If historically, up to that last 20 or so years, you produce 50 papers/measure of output per 10,000 people, and every year that starts to decline, and now you produce 40 per 10,000 there must be a reason for it.

Yes, but if historically you produce 50 papers per 10,000 people and now you produce 49.9, there need not be.

Come back to me when you learn about "clinical significance."

 

[jay gw]

Come back to me when you learn about "clinical significance."

Come back after you read previous posts, particularly the one about the National Science Foundation.

I especially like this part:

Even if action is taken today to change these trends, the reversal is 10 to 20 years away. The students entering the science and engineering workforce in 2004 with advanced degrees decided to take the necessary math courses to enable this career path when they were in middle school, up to 14 years ago.

The students making that same decision in middle school today won't complete advanced training for science and engineering occupations until 2018 or 2020. If action is not taken now to change these trends, we could reach 2020 and find that the ability of U.S. research and education institutions to regenerate has been damaged and that their preeminence has been lost to other areas of the world.

http://www.nsf.gov/sbe/srs/nsb0407/start.htm

 

[drkitten]

Come back after you read previous posts, particularly the one about the National Science Foundation.

Read it. It has nothing to do with outsourcing, and in fact, specifically contradicts your claims.

If the trends identified in Indicators 2004 continue undeterred, three things will happen. The number of jobs in the U.S. economy that require science and engineering training will grow; the number of U.S. citizens prepared for those jobs will, at best, be level; and the availability of people from other countries who have science and engineering training will decline, either because of limits to entry imposed by U.S. national security restrictions or because of intense global competition for people with these skills.

We're not outsourcing technical jobs; we are creating technical jobs faster than American students are willing to be trained to take them.

The number of jobs requiring S&E skills in the U.S. labor force is growing almost 5 percent per year. In comparison, the rest of the labor force is growing at just over 1 percent.

Outsourcing is in fact a good solution to this problem; by dumping low-paying, non-technical jobs on foreigners, American students will be encouraged to study science and engineering (S&E). The United States is, in fact, "insourcing" scientific and engineering talent, drawing substantial numbers of foreign workers to contribute their talents to the US economy.

Growth in the S&E labor force has been maintained at a rate well above the rate of producing S&E degrees because a large number of foreign-born S&E graduates have migrated to the United States. The proportion of foreign-born students in S&E fields and workers in S&E occupations continues to rise steadily. Persons born outside the United States accounted for 14 percent of all S&E occupations in 1990. Between 1990 and 2000 the proportion of foreign-born people with bachelor's degrees in S&E occupations rose from 11 to 17 percent; the proportion of foreign-born with master's degrees rose from 19 to 29 percent; and the proportion of foreign-born with PhDs in the S&E labor force rose from 24 to 38 percent.

The problem isn't with outsourcing. People don't particularly want to do the sort of technical work you're talking about in their home countries -- instead, they're coming to the US to do it.

But this has nothing to do with the central theme, which is this mythical "decline" in research productivity, which you are extrapolating wildly based on a single-year sample of a deeply flawed study.

First of all, American research productivity is, when measured by five year moving averages, actually increasing, up 28.9% over the ten year period between 1988-1998 (see this paper for some sources). The approximately 3% change seen in 1998 is dwarfed by the overall running trend. As the Japanese report put it :

During the period shown in the figure, the U.S. consistently had an overwhelming share of over 30% of world output, despite a decline starting in the 1990s. The output itself has leveled off, but the reason for this decline in share is not so much because of a reduction in the output of papers in the U.S., as it is an increase in the output by other countries.

(Emphasis mine). A smaller piece of a bigger pie is still bigger.

More to the point, Americans are still doing better science. Even your (flawed) report pointed this out :

t should be emphasized that these output figures do not consider the citation impact of published papers. According to the measure of impact (that is, citations per paper), the comparative influence of U.S. papers in physics and other fields—regardless of any decline in world share—remains robustly healthy. In 1998, for example, U.S. physics papers were cited, on average, at a level 51% above the world average in the field, while EU physics papers exceeded the world average by 20% and AP physics papers scored an impact 24% below the world mark. The numbers are similar in chemistry, where the impact of U.S. papers surpassed the world figure by 53%, compared to an impact mark 10% above the world average for the EU and 25% below for the AP.

The Japanese report confirms this :

The U.S. has the largest share of papers and its citation share is higher than its paper share, with almost half of the papers cited throughout the world being of U.S. origin. Thus once can see the enormous impact of U.S. papers.

Furthermore, the impact trend of US papers is increasing, as you can see from figure 6-1-4 of the cited report. US-based research is not only "better," but it's continuing to improve.

The following chapter of the Japanese report shows much the same thing regarding patents (as another measure of technical progress). As before, US-based patents are increasing in number, in quality, and in impact.

They're also increasing in connectedness between basic science and application, which is also a mark of a good technological infrastructure. Again, the US leads the field and continues to improve its lead:

Figure 6-2-10 shows the trends for the science linkage of U.S. patents by nationality of the applicant. All of the values are rising, which seems to indicate a strong relationship between patents and scientific papers. Examining the trends by country, we see that the U.S. values are the highest, increasing remarkably.

Finally, if you are really interested in the technological economy, the spot to look would be in the economics itself, for example, in the imports and outputs of technology. Again, the Japanese have been kind enough to quantity those for us:

U.S. technology exports are overwhelmingly high, with an amount 8 times higher than that of Japan in 1997. The trends show a prominent increase from the late 1980s, and although the technology imports increase on an annual basis, they are small relative to the value of exports, leading to substantial gaps in the technology trade incomes and expenditures. The ratio of technology exports to imports is 3.58 in 1997, which is exceptionally large among the countries shown in the figure, and shows the strong influence of U.S. technology.

We can expect a decrease in the US ratio over the next extended period of time simply because of the "smaller share of a bigger pie" effect, but the long term trends about the absolute difference between technology imports and technology exports is expected to be hugely in America's favor.

In terms of labor productivity caused by technological advances, "[c]omparing the five countries from the perspective of absolute levels of labor productivity of added value, the superiority of the U.S. remains unchanged, with the figure rising from 8.46 million yen to 9.90 million yen over the 16 years from FY1981 to FY1996."

So the effect that you are worried about doesn't exists, and what you identify as the root cause of the nonexistent effect is actually what a rational person would identify as a potential cure if the effect existed.

 

[jay gw]

Once again, I am forced to point out:

A “troubling decline” in the number of US citizens training to become scientists and engineers is creating an “emerging and critical problem” for the United States, according to a new report by the by the National Science Board, the independent government panel that oversees and sets policy for the National Science Foundation (NSF).

http://www.biomedcentral.com/news/20040506/02

And....another:

U.S. Is Losing Its Dominance in the Sciences
New York Times

More troubling to some experts is the likelihood of an accelerating loss of quality scientists. Applications from foreign graduate students to research universities are down by a quarter, experts say, partly because of the federal government's tightening of visas after the 2001 terrorist attacks.

Shirley Ann Jackson, president of the American Association for the Advancement of Science, told the recent forum audience that the drop in foreign students, the apparently declining interest of young Americans in science careers and the aging of the technical work force were, taken together, a perilous combination of developments.

"Who," she asked, "will do the science of this millennium?"

http://www.nytimes.com/2004/05/03/s...7a1c973069b2a8&ex=1398916800&partner=USERLAND

So DrKitten's remarks are directly contradicting what the American scientific community is saying, in unison, that the US science funding and programs are in jeopardy of decline, and are declining right now. Just needed to point that out for everybody's benefit.