1. Why the US lost leadership in scientific papers, and will soon lose in other indicators R. D. Shelton and Lance Miller WTEC Sponsored by NSF coop agreement ENG-0844639.

2. Outline: Some Q and A Problem: US publication plateau (or share decline) despite huge and increasing R&D investments Solution: Shelton Model reveals that the driver is world share of R&D funding, not its absolute value Refinement: The government part of R&D funding, and higher education spending are even better predictors, which can explain why the US lost leadership to the EU. Shelton-Leydesdorff Model. Generalization: Models of other indicators (patents, PhDs in science, Hi-tech market share) also help explain US declines. The drivers are identified by regression over the OECD data, and provide policy levers for remedies.

3. 200 80 320 GERD in $Billions Q: The American Paradox: Declining SCI paper shares despite increasing R&D (GERD*) Why? *Gross domestic expenditures on R&D (OECD Constant $ PPP), fractional paper share from NSF National Science Indicators.

4. Clue: US paper share depends more on its share of GERD than its total GERD

5. A: Shelton (2006) model of publication system $ Inputs US EU AT ROW Papers Published National Research Systems -- Fairly Independent Highly Interdependent Paper Selection Journal Editors g1g1 p1p1 G (total)P (total) w i = g i /G GERD share m i = p i /P Paper share Model: m i = k i w i

6. Performance of Shelton Model ( m i = k i w i ) in forecasting paper share from 2005 to 2010 Based on forecasts from 2005 data of GERD and k i. The relative efficiencies k i are fairly constant for these three. Accuracy is not bad, but PRC is growing slower than forecast. Whole counts in the WoS. (Shelton & Leydesdorff, 2011)

7. Q: This model can account for the recent US & EU decline of due to the rising GERD share of China, but, why did the US lose the lead in the mid-90s?

8. Relative efficiency k i is this ratio normalized by 39 OECDg country values. In 1990 EU and US had the same k i, but curves diverged in 1990s. (Foland & Shelton, 2010) Clue: EU passed the US because it sharply increased its ratio of papers/R&D $.

9. A: The EU increased paper share in the 90s by sharpening focus on sectors that maximize papers Government funding instead of industry University instead of business R&D spending Civilian instead of military R&D Multiple regression shows these input components are more effective in producing paper outputs—first two much more so.

10. Regression analysis of which GERD components best account for paper outputs Year = 1999, Constant $ PPP series used Dependent Variable (DV) = papers in SEI, fractional counts Independent Variables = two components of R&D funding (IV1, IV2)—several types N = 39 countries in OECD Group, sometimes fewer P is significance probability of IV; if p < 0.05, variable is important R 2 > 96% always – IVs are very good predictors

11. Government vs. Industrial funding of R&D IV1 = government funded part of GERD IV2 = industry likewise Much smaller components omitted IV1: P = 0.000 (very significant) IV2: P = 0.838 (not significant) Regression equation: Papers = 2.73 IV1 – 0.031 IV2 + 1957

12. In 1990s, both shifted R&D funding from government to industry, but this change was much smaller in the EU Government funding is much more likely to produce papers. Paper advantage: EU.

13. Patterns are almost identical with a small lag. The slight shift of EU is due to its HERD focus.

14. Regression identifies drivers for other indicators Patents. Complementary drivers: industrial funding and business spending are best. (Shelton & Leydesdorff, 2011) PhDs in S&E: Current stock of researchers High-tech market shares: stock of researchers, but this is a proxy for a real underlying variable, perhaps location of research—correlated with location of manufacturing.

15. National positions (Shelton & Foland, 2009) Stop press updates in red In 2005 data the leader was:  US: GERD, researchers, impacts, patents, hi-tech exports  EU: papers in SCI, S&E PhDs, Nobels  PRC: trade balance But forecasts show the PRC will gain: Lead hi-tech exports and researchers by 2010  Pass EU in GERD by 2015 (Still a good bet) Pass US in S&E PhDs by 2015 (Probably in 2008)  Lead in papers in SCI by 2017 (Chinese growth slowed, but it leads in INSPEC, Scopus, et al.)

16. US scientific publication decline can be explained by examination of two decades. In the 1990s the US was passed by the EU. In the 2000s the shares of both were decreased by the rise of China. In the 1990s, US R&D funding shifted from 1/2 government & 1/2 industry to 1/3 & 2/3. Since regression shows that industry-funded R&D produces far fewer papers, US paper production sagged. The EU also spends more on R&D that produces papers, particularly in universities, making it the world leader. This explains the American Paradox: the decline of US paper share as it increased its R&D investment. In the 2000s the PRC curve was far below those of the US and EU, but models based on its rapidly increasing share of world R&D investment, forecast that it will pass both. Similar data and models for patents, PhDs grads in science and engineering, and high-tech market share show continuing decline of the West with the rise of China. Conclusions

17. For more information

18. Foland, P & Shelton RD (2010) Why is Europe so efficient at producing scientific papers, and does this explain the European Paradox? 11th International Conference on S&T Indicators, Leiden, Sept. 2010. Leydesdorff, L., & Wagner, CS. (2009a). Macro-level indicators of the relations between research funding and research output. Journal of Informetrics, 3, 353-362. Leydesdorff, L., & Wagner, CS. (2009b) Is the United States losing ground in science? A global perspective on the world science system Scientometrics, 78, 23-36. Shelton, RD., (2008), Relations between national research investment and publication output: Application to an American paradox. Scientometrics 74, 191-205. References-1

19. Shelton, RD. & Foland, P. (2010) The race for world leadership of science and technology: Status and forecasts. Science Focus 5, pp. 1-9 (Feb. 2010) in Chinese. Also, Proceedings of the 12th International Conference on Scientometrics and Informetrics, pp. 369-380, Rio de Janeiro, July, 2009. Shelton, RD and Leydesdorff, L (2011a) Bibliometric evidence for an empirical trade-off in national funding strategies. Proceedings of the 13th International Conference on Scientometrics and Informetrics, pp. 763-774. Durban, July, 2011. Shelton, RD and Leydesdorff, L (2011b) Publish or Patent: Bibliometric evidence for an empirical trade-off in national funding strategies. Journal of the American Society for Information Science and Technology, in press. References-2 Posted at itri2.org/s/ or www.leydesdorff.net

20. University vs. Business Expenditure of R&D IV1 = HERD, higher education part of GERD spending IV2 = BERD, business part likewise IV1: P = 0.009 (very significant) IV2: P = 0.000 (very significant, but coefficient is much smaller) Regression equation: Papers = 2.53 IV1 +0.58 IV2 + 2136

21. Higher Education Expenditure on R&D Despite much smaller overall GERD, the EU spends more on university R&D. Paper advantage: EU.

22. In 2010 PRC increased at 26%, EU declined slightly. If this continued, curves would cross in 2015. Latest GERD data from OECD

24. Total International Patents from Derwent

25. w i ’ is national share of OECDg government R&D funding; k i ’ is the relative efficiency for this component. Uses 5-year average of rates of Government R&D funding increases. EU and PRC fit well, but US is worse than forecast, because its rate of Government increase has plummeted to near zero. (Shelton & Leydesdorff, 2011) Performance of Shelton-Leydesdorff Model m i = k i ’w i ’ in forecasting from 2005

26. This is almost a zero-sum game. Or an unstable queue with arrival rate > service rate. It’s even worse, because the arrival rate is increasing. Thus the delay time and backlog must increase without bound. USPTO is failing to keep up. Although users pay fees that could hire more examiners, the Congress siphons that off to pay other bills

27. One consequence is that US applications are being crowded out by foreign applications. In 2008, there were more foreign grants than domestic ones. The trends are slow, so it will take years before no Americans get US patents.

28. The US invests far more in R&D than the EU Q: But, after 1995, the EU publishes slightly more scientific papers than the US. Why?

29. Because US investment mainly comes from industry, and industry funded R&D is far less likely to result in papers And the EU overcomes the US slight advantage in government funding by spending more in the higher education sector that also produces more papers.