Implication of comparison of international competitiveness for educational policymaking: The case of Taiwan,

全文

(1)教育政策論壇 2005 年 2 月. 第八卷第一期 25. 教育政策論壇，2005 年八卷一期，25-42. 國際競爭力之比較對教育政策之啟示：以台灣為例馬信行國立政治大學教育學系教授. 摘要本研究探討教育相關變項是否與國際競爭力有關。資料是取自美國國家科學理事會及瑞士國際管理發展研究所之統計資料。結果顯出代表教育潛在變項之「每千人中研發人員數」單獨可解釋「競爭力」60%之變異量。本研究對教育政策有所啟示：以台灣為例，台灣與瑞士皆為天然資源稀少且國內市場狹小的小國家，其國際競爭力有賴人力資本，尤其是研發人員之創新。培養更多的研發人力為教育可對競爭力有所貢獻之處。. 關鍵詞：研發人力、理工教育、國際競爭力.

(2) 26 馬信行國際競爭力之比較對教育政策之啟示：以台灣為例. Since 1989 the Institute for Management Development (IMD), Switzerland, has each year published indicators of the international competitiveness of economically important countries.. It has attracted attention of policy makers worldwide.. What. can we learn from a comparison of international competitiveness? International competition is a variety of inter-group competition.. Johnson,. Maruyama, Johnson, Nelson, and Skon (1981) distinguished inter-group competition from interpersonal competition because inter-group competition demands within group cooperation.. To produce a product, all workforces in an enterprise have to cooperate. with each other.. The cooperation is either vertical or horizontal, or both. In their. meta-analysis, Johnson, et al. (1981) found that there was no significant difference in achievement between cooperation without versus cooperation with inter-group competition (mean effect size = 0). And cooperation with inter-group competition promotes significantly higher achievement than do individualistic, no competitive efforts (mean effect size = .50, SD = .37, t (19) = 6.04, p < .01). Qin, Johnson, and Johnson (1995) also discovered from their meta-analysis that mixed cooperation (in-group cooperation and inter-group competition) solved problems better than interpersonal competition (mean effect size = 0.39, SD = 0.56, t (14) = 2.61, p < .05). The results of the meta-analysis carried out by Johnson, Johnson, and Maruyama (1983) indicate that cooperation with inter-group competition promotes greater interpersonal attraction among participants than does interpersonal competition (mean effect size = 0.75, SD = 0.62, t (13) = 3.44, p < .01). The effect size of cooperation without competition is slightly higher than cooperation with inter-group competition but the difference is not significant (Johnson, et al., 1981; Qin, et al., 1995; Johnson et al., 1983).. Because the economy of Taiwan. is export- oriented, its products must compete with products from other countries in the.

(3) 教育政策論壇 2005 年 2 月. world market.. 第八卷第一期 27. Therefore, international competition is inevitable for Taiwan.. In. schools, interpersonal competition might be muted, but cooperation with inter-group competition should be fostered.. By the training of cooperation with inter-group. competition, students can be assisted in forming the habit of following “play rule” in inter-group competitions. Competitiveness is an important element in Social-Darwinian theory claiming that the dynamics of social evolution lie in the principle of “equal opportunity and fair competition.”. If a society is full of competition, its total capabilities will be. accelerated automatically, because during the process of competition, all the competitors will focus their attention on the improvement of their own competitive capabilities, and this is the process of the evolution in society. The Organization for Economic Cooperation and Development (OECD) (1992, cited by Llewellyn, 1996) defines a nation’s competitiveness as: “the degree to which it can, under free and fair market conditions, produce goods and services which meet the test of international markets, while simultaneously maintaining and expanding the real incomes of its people over the longer term” (p. 237), and the World Economic Forum defines competitiveness as “the ability of a nation’s economy to make rapid and sustained gains in living standards” (Llewellyn, 1996: 89). These two definitions are compatible in that expanding real income is a necessary condition for making sustained gains in living standards.. An expansion of real incomes depends on an increase in. production and exportation of goods and services. There would thus be a strong relationship between a nation’s productivity and its competitiveness. some countries grow economically at a faster rate than others.. It is evident that. It can be hypothesized. that the stronger competitiveness a country possesses, the faster its economic growth will be. Llewellyn (1996) remarked that a healthy way to strengthen a country’s.

(4) 28 馬信行國際競爭力之比較對教育政策之啟示：以台灣為例. competitiveness is to “reduce its prices or costs per unit of output relative to those of its trading partners” or to create new products or improve the quality of its products to meet the demands of other countries. Investment in R&D expenditures and personnel (scientists and engineers produced by the higher education institutions) are important factors for technological innovation, which can be indicated by granted patents. Technological catch-up relies heavily on technological activities. If the technologically advanced countries make further investments in R&D, then the technology gap and economic growth between them and countries following behind them would be persistent, or even enlarged. Patents are outputs of innovation while the expenditure and personnel of R&D are inputs to it. In a review of empirical literature, Fagerberg (1994) found that variables in technological innovation, such as R&D investment, patents, scientists and engineers etc. could have an impact on levels and growth of productivity. The presence of universities with major programs in science and engineering, venture capital, technological and scientific employees in the local population, and a good quality of life are alleged as necessary preconditions of high technology development in particular places, such as in the cases of the production center of semiconductors and computers in Silicon Valley and that of communications equipment, computers and biomedical instruments in Range County (Scott & Storper, 1987). Other variables which might contribute to international competitiveness are mentioned by Boltho (1996), such as improvements in infrastructure, raising the level of education and training of workforce, opening markets to foreign competition to invoke an imagined external threat, deregulating some aspects of economics, and sales and advertising campaigns. The International Institute for Management Development (IMD) (2000) used 290.

(5) 教育政策論壇 2005 年 2 月. 第八卷第一期 29. variables in comparing the competitiveness of 47 countries. Among those variables, some can be classified as end product (outcome) elements of competitiveness, such as Gross Domestic Product (GDP) per capita and exports of goods, which might indicate the “real incomes” of a country’s people; life expectancy at birth, which represents the level of the standard of living; and patents granted to residents, and some as process (driving force) elements of competitiveness, which lead to a strengthening of competitive capability, such as connections to internet; total expenditure on R&D per capita; and total R&D personnel in business enterprises. variables in the IMD study, 139 were official statistics.. Among the 290 competitive All the variables were. categorized into eight factors: domestic economy, internationalization, government, finance, infrastructure, management, science & technology, and people. The purposes of the present study are to investigate: (a) whether education related variables have a relationship with competitiveness, and if it so, by how much? If education is related to international competitiveness, then education-related variables would have a significant association with competitiveness-related variables, (b) the nature of the causal relationship between process and product elements of international competitiveness; what proportion of variance in product elements can be accounted for by process elements of competitiveness?. According to the definition of. competitiveness made by the OECD and the World Economic Forum, stronger competitiveness means higher incomes (to use GDP per capita as a proxy) and higher living standard (to use life expectancy at birth and private consumption expenditure per capita as proxies).. Higher private consumption expenditure per capita stands for. stronger purchasing power. competitiveness.. These three variables are set as end product elements of.

(6) 30 馬信行國際競爭力之比較對教育政策之啟示：以台灣為例. Method The cross-sectional hard data (official statistics) from the database of the International Institute for Management Development (2000) were used. The International Institute for Management Development (2000) includes an indicator of educational infrastructure relevant for producing engineers and scientists. The indicator “higher education enrollment of 1996” (net enrollment in tertiary education for persons 17-34 years old) has however 15 missing values and could not be used in the analysis. Instead, a variable “ratio of total science and engineering degrees to the 24-year-old population: 1997 or most recent year” was adopted by the present author from the Appendix Table 5-18 of the database of the National Science Board (2000), which has only four missing values. To test the hypotheses, Pearson’s correlation, regression, as well as structural equation model (Jöreskog & Sörbom, 1993) were employed.. By the regression and. structural equation model, the step-wise entrance of exogenous variables will be displayed, so that the phenomenon of multicollearity can be clarified.. When the. coefficient of an exogenous variable is significant in the reduced model but turns to be not significant after a new exogenous variable enters the equation, it means that the effect of that variable is partialed-out by the new one.. And in the correlation matrix,. one can find that the first-entered variable has a high correlation coefficient with the later-entered exogenous variable.. Results The matrix of correlations in Table 1 indicates clearly that all the four education-related variables have significant correlation with GDP per capita, and they are interdependent with the exception of correlation between the “ratio of science and engineering degrees to the 24-year-old population in 1997” and the “pupil-teacher ratio.

(7) 教育政策論壇 2005 年 2 月. 第八卷第一期 31. in secondary education.” Table 1.. Correlations Between GDP and Education-Related Variables. 1. GDP per capita 2. Total R&D (N=47). personnel nationwide /1,000people (N=42). 1 -2. .79** --. 3. 3. Ratio of science & 4. Secondary engineering school degrees to the enrollment/ 24-year-old relevant age population in 1997 group (N=43). (N=43). .68** .63** .70** .73**. -.45** -.47**. --. .59**. -.17. --. -.37*. 4. 5. Pupil-teacher ratio in secondary education (N=45).. * p < .05 ** p < .01. Table 1 shows that the variable “Total R&D personnel nationwide /1,000 people” has strongest relationship with the variable “GDP per capita”.. In Table 2, the variable. “R&D personnel per 1,000 people” accounts for 61.66% of the total variance in the dependent variable “GDP per capita”.. As the other educational variables were added. to the regression equation, no increments to the adjusted R2 were observed. Consequently, the variable “R&D personnel per 1,000 people” was selected as a representative variable for education-related variable contributing to GDP per capita. Table 2. Coefficients From the Regression of GDP Per Capita on Education-related Variables Model 1 R&D personnel. Model 2. .79**. S & E degrees. .62**. Model 3. Adjusted R2. F (1,40) = 66.93** .6166. F (1,41) = 25.45** .3679. Model 5. .67**. .54**. .46*. .15. .18. .22. .13. .07. F (3,33) = 16.18** .5585. -.18 F (4,31) = 12.21** .5616. Enrollment Pupil-teacher ratio F values. Model 4. F (2,37) = 28.23** .5827. Note. R&D personnel = Total R&D personnel nationwide / 1,000 people; S&E degrees = Ratio of science and engineering degrees to the 24-year-old population in 1997; Enrollment = Secondary school enrollment / Relevant age group; Pupil-teacher ratio = Pupil-teacher ratio in secondary education * p < .05 ** p < .01.

(8) 32 馬信行國際競爭力之比較對教育政策之啟示：以台灣為例. In order to test the causal relationship between process and product elements of international competitiveness, different models of LISREL were tried and displayed in Table 3 Because of multicollinearity, it is necessary that variables be introduced step by step to show the effect of each process element on the product elements of international competitiveness. The parameter specifications used by Jöreskog & Sörbom (1993) were applied, ) but with minor amendment in the present study. Instead of λ(ijX ) and λ(Y , λij was ij. employed to stand for the standardized effect of a latent variable on an observed variable.. In doing so, the serial number of the first latent endogenous variable must. follow that of the last latent exogenous variable, and the serial number of the first observed endogenous variable has to come after that of the last observed exogenous variable. International competitiveness is a latent variable (an abstract construct) and must be indicated by measurable and observable variables.. It was indicated by three. observed variables: life expectancy at birth, GDP per capita, and private consumption expenditure per capita. In Table 3, the number of the subscript of a Roman alphabet letter denotes the serial number of an observed variable. The number of the subscript of Greek alphabet letter denotes the serial number of a latent variable. δ and ε represent the error term of an observed independent (X) and dependent variable (Y) respectively. Because of length restriction of the article, the error terms in Table 3 were omitted. They were all not significant. The outer number of the subscript λ is the serial number of a latent variable, and the inner number is that of its indicator. The non-significance of error term and the significance of λ term of an observed variable indicates that the observed variable is suitable to be the proxy of that latent variable,.

(9) 教育政策論壇 2005 年 2 月. 第八卷第一期 33. e.g. the significances of λ22, λ32, λ42 in Model 1 mean that the observed variables Y2 (Life expectancy at birth), Y3 (Private consumption expenditure per capita), and Y4 (GDP per capita) are appropriate to be used as the indicators for the latent dependent variable η2 (competitiveness). Table 3. Different Structural Equation Models Demonstrating Effects of Determinants of Competitiveness Model 1. Model 2. Model 3. Model 4. Model 5. X1=RDPERSON. X1=RDPERSON. X1=RDPERSON. X1=RDPERSON. X1=RDPERSON. Y2=LIFE. X2=RDEXPEND. X2=RDEXPEND. X2=RDEXPEND. X2=RDEXPEND. Y3=GDP. Y3=LIFE. X3=COMPUTER. X3=COMPUTER. X3=COMPUTER. Y4=CONSUMPT. Y4=GDP. X4=HANDY. X4=HANDY. X4=HANDY. ξ1=Educat. Y5=CONSUMPT. Y5=LIFE. X5=ADVERTIS. X5=ADVERTIS. η2=Compet. ξ1=Techinno. Y6=GDP. Y6=LIFE. Y6=PATENT. λ22=0.80*. η2=Compet. Y7=CONSUMPT. Y7=GDP. Y7=LIFE. λ32=1.00*. δε23=-0.06. ξ=Techinno. Y8=CONSUMPT. Y8=GDP. λ42=0.97*. λ11=0.85*. ξ2=Techinfr. ξ1=Techinno. Y9=CONSUMPT. ζ2=0.40. λ21=0.96*. η3=Compet. ξ2=Techinfr. ξ1=Techinno. γ21=0.77*. λ32=0.80*. λ11=0.88*. ξ3=Market. ξ2=Techinfr. χ(1,N=42)=0.002,. λ42=1.00. λ21=0.93*. η4=Compet. ξ3=Market. λ52=0.97*. λ32=0.96*. λ11=0.88*. η4=Patents. GFI=1.00. ζ2=0.17. λ42=0.83*. λ21=0.93*. η5=Compet. RMR=0.00. γ21=0.91*. λ53=0.82*. λ32=0.94*. λ11=0.88*. χ(3,n=42)=3.69. λ63=0.99*. λ42=0.81*. λ21=0.95*. p=0.30. λ73=0.96*. λ53=1.00. λ32=0.93*. GFI=1.00. ζ3=0.07. λ64=0.82*. λ42=0.80*. RMR=0.0085. φ12=0.89*. λ74=0.99*. λ53=1.00. γ31=0.21. λ64=0.97*. λ64=1.00. γ32=0.78*. ζ4=0.03. λ75=0.82*. χ(10,n=42)=3.61,. φ12=0.91*. λ65=0.99*. p=0.96. φ13=0.83*. λ95=0.96*. GFI=1.00. φ23=0.91*. ζ4=0.26. RMR=0.031. γ41=0.04. ζ5=0.02. R2=0.93. γ42=0.64. φ12=0.91*. γ43=0.33. φ13=0.81*. χ(14,n=42)=4.66. φ23=0.92*. p=0.99. β54=0.01. p=0.96. 2. R =0.60. 2. R =0.830.

(10) 34 馬信行國際競爭力之比較對教育政策之啟示：以台灣為例 Table 3. (continued) Model 1. Model 2. Model 3. Model 4. Model 5. GFI=1.00. γ41=0.82*. RMR=0.028. γ54=0.01. 2. R =0.97. γ52=0.55 γ53=0.47* χ(19,n=42)=14.2 p=0.77 GFI=1.00 RMR=0.026 R2=0.98. Note. RDPERSON= Total R&D personnel nationwide / 1,000 people RDEXPEND= Total expenditure on R&D per capita COMPUTER= Number of computers / 1,000 people HANDY= Cellular mobile telephone subscribers / 1,000 people ADVERTIS= Advertising expenditure per capita PATENT= Number of patents in force / 100,000 inhabitants LIFE= Life expectancy at birth. GDP= GDP per capita. CONSUMPT= Private consumption expenditure per capita Techinno= Technology innovation activities Market=Market activity. Techinfr=Technology infrastructure. Patents=Granted Patents. Compet=Competitiveness. * p < .05. In Model 1 of Table 3, “R&D personnel per 1,000 people” was selected as an indicator for education. γ21 = 0.77* in Model 1 indicates that education makes a significant contribution to competitiveness. γ21 is the standardized regression coefficient of ξ1 on η2. The outer number of the subscript γ is the serial number of the independent latent variable, and the inner number is that of the dependent latent variable. R2= .60 designates that education alone can explain 60% of the variances in competitiveness.. This result supports the hypothesis that the education-related. variable “R&D personnel per 1,000 people” does have a significant association with competitiveness-related variables..

(11) 教育政策論壇 2005 年 2 月. 第八卷第一期 35. In Model 2, “Total expenditure on R&D per capita” was combined with “R&D personnel per 1,000 people” to form indicators for the latent variable “Technological innovation activities”. The latent variable “Technology innovation activities” explained 83% of variances in competitiveness. γ21 = 0.91* denotes that “Technology innovation activities” is a significant determinant of competitiveness.. In order to. improve fitness of model, the error term δε23 = -0.06 was generated by a statement: “set the errors between RDEXPEND and LIFE correlate” in the LISREL 8 program, as the Maximum Modification Index located at this term.. After modification, the fitness. of the model improved. In Model 3, the latent variable “Technology infrastructure” was indicated by observed variables “Number of computers/1,000 people” and “Cellular mobile telephone subscribers/1,000 people”.. In Model 4, the latent variable “Market. activity” was indicated by the observed variable “Advertising expenditure per capita”. In Model 3 and Model 4, the latent variables “technology infrastructure” and “market activity” were introduced stepwise. They both brought about increment of R2. “Granted patents” was inserted to Model 5 as an intervening variable between innovative activity and competitiveness. The non-significance of γ31 (the effect of technology innovation activities on competitiveness) in Model 3 was due to multicollinearity. The effect of technology innovation activities on competitiveness was originally significant in Model 1, but after stronger latent variable (Technology infrastructure) was introduced, its effect was partialed out.. Similar cases are the non-significances of γ41, γ42, γ43 in Model 4 and. γ52 and γ54 in Model 5. γ41 = 0.82* means technology innovation activities have significant influence on number of granted patents.. γ54 = 0.01, γ52 = 0.55, γ53 = 0.47*. mean that “Market activity” has a significant effect on competitiveness, but the effect of “Technology infrastructure” and “Granted patents/1,000 people” on competitiveness.

(12) 36 馬信行國際競爭力之比較對教育政策之啟示：以台灣為例. is partialed-out by the stronger latent independent variable “Market activity”. ψij marks the covariance of two independent latent variables. ζ is the error term of a latent dependent variable. ζ5 = 0.02 in model 5 hints that R2 = 1- .02 = .98. It indicates that the three latent independent variables (ξ2, ξ3, and η4 )can explain 98% of variances of the dependent variable “Competitiveness” (ξ5). Have the results of the present study any implication for the educational policy makers of small countries such as Taiwan? The situation of Taiwan is analogous to that of Switzerland. They are both small countries with scarce resources and limited domestic market. Their competitiveness hangs on human capital, especially on innovations by R&D personnel. To cultivate more and more creative R&D personnel is the contribution education can make to competitiveness.. From Table 4, it can be seen that R&D personnel per 1,000 people. in Switzerland is about 1.5 times as many as in Taiwan.. The overall rank of. competitiveness of Switzerland improved from ninth in 1996 to fifth in 2000, and the R&D personnel per 1,000 people in Switzerland ranked second in 2000.. To catch-up. to the level of Switzerland in this respect would be a reasonable goal for the educational policy of Taiwan, if it seeks to improve its international competitiveness.. Discussion Among education-related variables which may have relationship with international competitiveness, “R&D personnel per 1,000 people” is the variable most suitable to be chosen as the indicator for the latent variable “Education,” because it has stronger correlation with GDP per capita than other variables, such as “Ratio of science and engineer degrees to the 24-year-old population,” “Ratio of secondary school enrollment to the relevant age group,” or “Pupil-teacher ratio in secondary education.” “R&D personnel per 1,000 people” standing alone, can explain 60% of variance in.

(13) 教育政策論壇 2005 年 2 月. 第八卷第一期 37. international competitiveness, which was represented by three observed variables: “Life expectancy at birth,” “GDP per capita,” and “Private consumption expenditure per capita.” Table 4. Comparison of the Competitiveness Related Indicators Between Taiwan, Switzerland, and the Best countries Variable. Value of Rank of Taiwan Taiwan. Value of Switzerland. Rank of Value of best country Switzerland. GDP per capita. $13,111. 25. $36,071. 2. Private consumption expenditure per capita Number of computers/1,000 people Cellular mobile telephone subscribers/1,000 people Advertising expenditure per capita Total expenditure on R&D per capita Total R&D personnel /1,000people Number of patents in force/100,000 in habitants Life expectancy at birth. $7,973. 26. $18,260. 5. $44,424(Luxembourg) . $21,953(USA).. 260.1. 23. 408.3. 10. 538.9(USA).. 493.60. 10. 441.65. 13. 679.10(Finland).. 151.1. 19. 346.53. 13. 419.41(USA).. $242.80. 20. 1142.3. 1. 1143.2(Switzerland).. 4.662. 14. 7.11. 2. 7.401(Sweden).. 686.8. 7. 1342.2. 1. 1342.2(Switzerland).. 73.7. 31. 79.1. 5. 80.3(JAPAN).. Data source: From “The world competitiveness yearbook” by International Institute for Management and Development, 2002, Lausanne, Switzerland: IMD.. The significant correlation between the “Ratio of secondary school enrollment to the relevant age group” and GDP per capita found in the present study confirms the result of Mankiw, Romer, & Weil’s (1992) study which demonstrated that adding human capital, with the “Ratio of secondary school enrollment to the relevant age group” as a proxy, to the exogenous variables (saving and population growth) of regression equation led to a significant increment of .2 in R2. A direct way to expand the number of R&D personnel is to increase the scale of doctoral programs in Science and Engineering. The National Science Board (2000: Chapter 4) describes the worldwide effort to expand doctoral programs in science and engineering.. The major Asian countries, China, India, Japan, South Korea, and.

(14) 38 馬信行國際競爭力之比較對教育政策之啟示：以台灣為例. Taiwan, awarded science and engineering degrees in an average annual increment of 12% from 1993 to 1997.. In Western countries, by individual country, the USA has. the highest number of doctoral degrees in Science and Engineering fields in 1997. But the combined doctoral Science and Engineering degrees of the three largest European countries (the UK, Germany, and France) recently surpassed that of the USA. Endeavors in strengthening and expanding doctoral education in science and engineering are to develop the capacity for high quality research leading to technological innovation and to build up the knowledge-based economy, and in the end to gain strength in competitiveness. If Taiwan seeks to improve its international competitiveness, to catch-up to the level of R&D personnel per 1,000 people of Switzerland would be a reasonable goal for the educational policy of Taiwan.. However, it is easier for Mainland China to. enlarge the number of R&D personnel than Taiwan.. Under the leadership of the. central government, Mainland has put various long-term and home-based projects into force, such as “The 863 Research Project”, and “The Hundred, Thousand and Ten Thousand Talent Plan” to train scientist and engineers at home and attract oversees Chinese scientist and engineers from abroad (Cao, 1996). In contrast, education in pluralistic Taiwan has been facing a value-conflicting controversy in emphasizing the education of scientists and engineers on the one hand, and pay attention to social sciences and humanities on the other. Since the 1990s, the Council for Economic Planning and Development has been criticized for manipulating the enrollment of higher education through manpower projection and planning in gearing for the nation’s economic development.. In the final report on the educational reform, it is declared:. “any quantitative planning for the expansion of higher education is inevitably self-assertive” (Educational Reform Evaluation Committee, 1996: 68). As a consequence, every higher education institution may put forward proposals to establish.

(15) 教育政策論壇 2005 年 2 月. 第八卷第一期 39. up to five new departments or graduate schools to the Ministry of Education for approval. But after a few years of boost, the Ministry of Education could not afford to it and regulated that the burden of finance and personnel of newly established department or graduate school lie with the university self. Therefore, it would be more difficult than before to establish new departments of science and engineering to nurture R&D Personnel.. Although the number of graduates with a PH. D. in science. and technology has increased yearly since 1987/88, it decreased from 103,300 in 2000/01 to 101,700 in 2001/02, and the proportion of such graduates to the total number of graduates with a PH. D. decreased from 80% in 1994/95 to 69.5% in 2001/02 (Council for Economic Planning and Development, 2002: Table 22). Whether the result of the present study can be used as argument to promote the cultivation of R&D personnel will challenge the intelligence of educational policy makers.. References Boltho, A. (1996). The assessment of international competitiveness. Oxford Review of Economic Policy, 12 (3), 1-16. Cao, X. (1996). Debating “brain drain” in the context of globalization. Compare, 26 (3), 269-283. Council for Economic Planning and Development. (2002). Manpower Indicators (Taipei, Council for Economic Planning and Development). Retrieved May 30, 2004, from http://www.cepd.gov.tw/people/english/indicator/t22.pdf Educational Reform Evaluation Committee. (1996). The fourth report of the evaluation of educational reform. Taipei: Author. Fagerberg, J. (1994). Technology and international differences in growth rates. Journal of Economic Literature, 32, 1147-1175..

(16) 40 馬信行國際競爭力之比較對教育政策之啟示：以台灣為例. International Institute for Management Development. (2000). The world competitiveness yearbook 2000. Lausanne Switzerland: IMD. Johnson, D.W., Johnson, R.T., & Maruyama, G. (1983). Interdependence and interpersonal attraction among heterogeneous and homogeneous individuals: A theoretical formulation and a meta-analysis of the research. Review of Educational Research, 53(1), 5-54. Johnson, D.W., Maruyama, G., Johnson, R., Nelson, D., & Skon, L. (1981). Effects of cooperative, competitive, and individualistic goal structure on achievement: A meta-analysis. Psychological Bulletin, 89 (1), 47-62. Jöreskog, K., & Sörbom, D. (1993). Structural equation modeling with the SIMPLIS command language (LISREL 8). Hillsdale, NJ: Lawrence Erlbaum Associates Publishers. Llewellyn, J. (1996). Tackling Europe’s competitiveness. Oxford Review of Economic Policy, 12 (3), 87-96. Mankiw, N. G., Romer, D., & Weil, D. N. (1992). A contribution to the empirics of economic growth. The Quarterly Journal of Economics, May, 407-437. National Science Board. (2000). Science and Engineering Indicators, 2000. (NSB-00-1.Arlington, VA, National Science Foundation. Available online). Retrieved May 30, 2004, from http://www.nsf-gov/sbe/srs/seind00/ (downloaded May 2001) Organization for Economic Cooperation and Development. (1992). Technology and the economy: The key relationships. Paris: OECD. Qin, Z., Johnson, D., & Johnson, R.T. (1995). Cooperative versus competitive efforts and problem solving. Review of Educational Research, 65(2), 129-143. Scott, A. J., & Storper, M. (1987). High technology industry and regional development: A theoretical critique and reconstruction. International Social Science Journal,.

(17) 教育政策論壇 2005 年 2 月. 第八卷第一期 41. 112, 215-232.. 2004 年 05 月 31 日收件 2004 年 10 月 28 日修正 2004 年 11 月 03 日接受.

(18) 42 馬信行國際競爭力之比較對教育政策之啟示：以台灣為例. Implication of Comparison of International Competitiveness for Educational Policymaking: The Case of Taiwan* Hsen-Hsing Ma Professor, Department of Education, National Chengchi University. Abstract This study investigates whether education-related variables have a relationship with international competitiveness. Analyzing data-bases statistics from the National Science Board and the International Institute for Management Development, the present study shows that: “R&D personnel per 1,000 people,” which represents the latent variable “education,” standing alone, can explain 60% of variance in competitiveness. educational policymaking. that of Switzerland. domestic market.. The result of the present study has implication for The situation of Taiwan, as an example, is analogous to. They are small countries with scarce resources and limited Their international competitiveness hangs on human capital,. especially on innovations by R&D personnel. To cultivate more and more creative R&D personnel is the contribution education can make to competitiveness.. Keywords: international competitiveness; R&D personnel; education of science and engineering. *. This research was supported by grants from the National Science Council, Taiwan ( NSC 90-2413-H-004-007). The assistance of part-time assistants, Miss Chiu, Yun-Jing and Miss Gao, Yu-jing is appreciated..

(19)