ISI KANDUNGAN Bil 6/2021

Assalamualaikum Warahmatullahi Wabarakatuh dan Salam Sejahtera

YBhg. Datuk/ Datin / Prof/ Tuan/ Puan

SDI@PTAR adalah salah satu perkhidmatan penyebaran maklumat terpilih yang disediakan oleh Perpustakaan Tun Abdul Razak, UiTM Shah Alam untuk ahli Mesyuarat Senat UiTM. Perkhidmatan ini bertujuan untuk menyalurkan maklumat terbaharu mengenai isu-isu semasa di dalam dan luar negara yang memberi nilai tambah serta impak kepada pengajaran, pembelajaran dan penyelidikan UiTM ke arah menjadi Universiti Terkemuka Dunia.

Untuk keluaran kali ini, SDI@PTAR menampilkan artikel teks penuh mengenai University Transformation. Diharapkan maklumat ini memberi manfaat kepada YBhg. Datuk/ Datin/ Prof/ Tuan/ Puan.

Sebarang cadangan dan maklumbalas mengenai perkhidmatan ini boleh disalurkan kepada En. Mohd Ismail bin Abidin, Timbalan Ketua Pustakawan (e-mel [email protected]) dan Puan Nik Zatihulwani binti Jamaludin, Pustakawan, (e-mel [email protected]), Bahagian Penyelidikan, Pembelajaran dan Rujukan, Perpustakaan Tun Abdul Razak, UiTM Shah Alam.

Sekian. Terima kasih.

Bahagian Penyelidikan, Pembelajaran & Rujukan Jabatan Perkhidmatan Perpustakaan Perpustakaan Tun Abdul Razak Utama UiTM Shah Alam

Bil 6/2021

ISI KANDUNGAN BIL. TAJUK SUMBER h-INDEX PENULIS 1 Does the Asian catch-up model of world- Educational 11 Lee, J., Liu, K., Wu, Y. class universities work? Revisiting the zero- Research for University at Buffalo, The sum game of global university rankings and Policy and State University of New York, government policies Practice Buffalo, #346 World University Ranking

2 Comprehensive Internationalization in the Higher 2 Sanders, J.S. Pursuit of ‘World-Class’ Status: A Cross-Case Education Temple University, Japan Analysis of Singapore’s Two Flagship Policy Campus, Tokyo, Japan Universities #301-350 World University Ranking

3 Engineering a World Class University? The Higher 2 Fu, Y.-C., Baker, D.P., Zhang, Impact of Taiwan’s World Class University Education L. Project on Scientific Productivity Policy National Taipei University of Technology, Taipei, Taiwan #488 World University Ranking

Assalamualaikum Warahmatullahi Wabarakatuh dan Salam Sejahtera

YBhg. Datuk/ Datin / Prof/ Tuan/ Puan Bil 6/2021

Does the Asian catch-up model of world-class universities work? Revisiting the zero-sum game of global university rankings and government policies Educational Research for Policy and Practice https://doi.org/10.1007/s10671-020-09261-x

ORIGINAL ARTICLE

Does the Asian catch-up model of world-class universities work? Revisiting the zero-sum game of global university rankings and government policies

Jaekyung Lee1 · Keqiao Liu2 · Yin Wu3

Received: 31 October 2019 / Accepted: 6 March 2020 © Springer Nature Singapore Pte Ltd. 2020

Abstract This study examines international brain race for world-class universities as measured by the QS World University Rankings (QS) and the Academic Ranking of World Universities, particularly in the context of Asian nations’ institutional competition and benchmarking against American counterparts. Applying mixed methods with statistical analysis of time- series data from 59 nations and case study of four selected nations, the study examines the cross-national trends of global university rankings and the roles of government policies during the era of performance-driven accountability in higher education. Tracking the zero- sum game of global university rankings over the past decade, the study accounts for key driving factors that produced divergent trajectories among the nations: how and why the USA as established leader and Japan as early catch-up leader become losers, whereas China and Korea as fast followers become winners? Although the Asian catch-up model of world-class university development, high-stakes institutional competitions and targeted funding with STEM priorities contributed to their rapid growth of research productivity and rankings, it reveals major limitations and problems. Drawing cross-national lessons and implications, we discuss new directions of higher education policies and global university ranking measures.

Keywords Higher education · University rankings · World-class universities · Accountability · Research productivity

Growing competition in the new global knowledge-based economy is accompanied by an international brain race in higher education, particularly in science, technology, engineering, and math (STEM) fields. In the midst of an international brain race, global university rankings became more visible and affected government policies in higher education (Hazelkorn 2015). Particularly, policymakers in China, Japan, and South Korea (Korea in short hereafter), where

B Jaekyung Lee [email protected]

1 Graduate School of Education, University at Buffalo, Buffalo, NY 14260, USA 2 School of Public Finance and Public Administration, Jiangxi University of Finance and Economics, Nanchang, China 3 Department of Psychiatry, McGill University, Montreal, Canada 123 J. Lee et al. their research universities were viewed as key driving forces of economic development, focused on improving the performance of their higher education systems as measured by international university rankings such as the Times Higher Education (THE), the QS World University Rankings (QS), and the Academic Ranking of World Universities (ARWU). Using American or other Western top-tier research universities as benchmarks, those Asian countries made strategic investment in higher education with STEM priorities toward the goal of building or adding more of their own world-class universities (Altbach 2004; Paul and Long 2016;SadlakandCai2009; Yonezawa and Shimmi 2016). Among the top-ranked universities in East Asian countries, there have been significant changes in the global university rankings of those in China, Japan, and Korea since early 2000s (see Table 1). Over the period of 2003–2013, for example, the number of Japanese universities in the top 500 as ranked by ARWU dropped from 36 to 20, whereas the number of corresponding Korean and Chinese universities in the top 500 increased from 8 to 11 and from 9 to 28, respectively, in the same period. At the same time, the number of US universities in the ARWU top 500 ranks dropped from 161 to 149. While the dominance of both American and Japanese universities has weakened, the representation of other Asian universities, particularly from China and Korea, has increased rapidly. Previous studies did not examine nuanced differences among these Asian nations in their policy strategies and funding priorities for developing world-class universities based on their initially different positions in the global university ranking systems. Our study focused on the comparison of three East Asian nations that include Japan as an early catch-up leader and China and Korea as fast followers, all of which benchmarked their progress against world-class American research universities as established leaders. The purposes of this study were to understand the international trends of global university rankings and contributing factors, specifically, the role of performance-driven, high-stakes accountability policy and funding toward building world-class universities (see Table 1). Given the zero-sum game of university ranking trends, this study also discussed why the Asian universities’ ranking advances are restricted largely among top 200 and 500 levels (but not 100 level) and what these rapidly grown Chinese, Japanese, and Korean universities are still missing in comparison with established competitive universities in the USA. The study would help understand the effects and limitations of Asian catch-up model for world-class universities and inform discussion about the improvement of higher education policy and university ranking systems.

1 Review of prior research and policy issues

On the one hand, prior research raised questions about the validity of international univer- sity rankings such as ARWU as well as domestic rankings such as US News. The ranking measures have been criticized for favoring universities with reputations (so-called academic beauty contests) and rewarding institutions with expensive research programs and sponsored research activities (Safon 2013; Volkweinand Grunig 2005). Moreover, the current university ranking measures tend to ignore or undervalue the importance of teaching quality and student learning outcomes (Liu and Cheng 2005; National Research Council 2012). In response to the university rankings, institutions tend to change resource allocations (Kim 2018). Private universities typically respond more quickly to market-like competition pressures that do their public counterparts, and those strategic behaviors such as globalization efforts tend to achieve higher status in world rankings systems such as the ARWU (Cantwell and Taylor 2013;Tay- lor and Cantwell 2014). While global and domestic university rankings continue to influence

123 Does the Asian catch-up model of world-class universities… , and the USA 149 (in 2013) (in 2014) 23.89 (in 2012) (in 2015) funding targeted for world-class university development; America COMPETES Act (2007–); Performance-based university funding and accountability policies enacted by many states (since 1990s–) Down from 161 (in 2003) to Down from 109 (in 2008) to 98 Down from 40.78 (in 1996) to Up from 2.44 (in 1996) to 2.79 No national policies and 2013) 2014) (in 2012) (in 2015) (1999–2005 for phase I; 2006–2012 for phase II); World Class University Program (2008–2012) Up from 8 (in 2003) to 11 (in Up from 11 (in 2008) to 14 (in Up from 0.54 (in 1996) to 2.16 Up from 2.24 (in 1996) to 4.23 Brain Korea 21 Project (BK21) (in 2013) (in 2014) 3.45 (in 2012) (in 2015) Of Excellence (COE21) (2002–06); the Global Centers of Excellence (Global COE) (2007–); the World Premier International Research Center (WPI) (2007–); Top Global University Project (2014–) Down from 36 (in 2003) to 20 Down from 29 (in 2008) to 17 Down from 5.68 (in 1996) to Up from 2.69 (in 1996) to 3.28 Twenty-First-Century Centers 2013) 2014) (in 2012) (in 2015) phase I, 2001–2005 for phase II), approximately US $2.2 billion for phase I; 985 Project (1999–2011); Double World-Class Project (2015–) China Japan Korea USA Up from 9 (in 2003) to 28 (in Up from 13 (in 2008) to 19 (in Up from 0.80 (in 1996) to 8.62 Up from 0.56 (in 1996) to 2.07 211 Project (1996–2000 for Snapshots of changes in world-class university rankings, research citations, R&D expenditures, and government policies among China, Japan, Korea counts, ARWU 2003–2013 counts (country rank), QS 2008–14 share), 1996–2012 expenditures (% of GDP), 1996–2015 improving university research and performance measures Table 1 Top 500 world-class university Top 500 world-class university Research citations (% world Research and development Policies/programs for 123 J. Lee et al. institutional practices, less is known about the role of national policies (Hazelkorn 2015). The US universities tend to be quite insular in terms of their perspective on peer institutions in other nations such that they are not actively competing in this race; very few US universities benchmark themselves by using the ARWU or the QS rankings. On the other hand, prior research examined various factors that account for cross-national variations in international university rankings (Pietrucha 2017; Taylor and Braddock 2007; Williams et al. 2013). For example, they considered key factors such as national economic wealth (GDP per capita) and research funding (R&D expenditures). However, previous stud- ies had limitations in that they mostly examined cross-sectional data for international ranking differences as opposed to time-series data for ranking changes. While previous studies exam- ined the effects of relatively stable (time-invariant) factors such as social/institutional factors (e.g., political stability and government effectiveness by Pietrucha 2017) and language factors (e.g., language bias of citation-based ranking measures by van Raan et al. 2011), they did not examine differences in education policy (nation/state-level) factors. Further, when previous studies examined national aggregate research productivity as measured by publications and citations, they also did not consider differences in policy support and funding among different types of universities and fields (e.g., STEM vs. arts and humanities). This study builds on prior research and fills the knowledge gap by examining the zero- game nature of global university rankings and government policies, particularly Asian nations’ catch-up model of world-class university development and benchmarking approach. It is worth noting that these international competitions for world-class university rankings occurred during the times of performance-driven educational accountability policy movement as well as increasing marketization of education sector (Lee 2016). New public management (NPM) movement has spread based on the premise that market-oriented and performance- driven management of the public sector will lead to greater cost efficiency for governments (Osborne and Gaebler 1992). Benchmarking became a popular business strategy in higher education as well in response to market demands for quality and global best practice (Levy and Ronco 2012). Previous studies found that conditional on institutional autonomy and resources, performance-driven incentives and accountability were associated with higher fac- ulty productivity (Aghion et al. 2010; McCormack et al. 2014). However, the Asian model of building world-class university relied heavily on government funding and central planning without creating an environment for intellectual autonomy and sustainable innovations (Alt- bach 2000;Li2018). Although the role of the government is crucial for advancing national competitive advantage, it is questionable whether the Asian catch-up model of higher educa- tion helps develop an environment for learning and innovation (Porter 1998) and for social mobility and equality (Marginson 2018). Altbach (2000) argued that the university rankings presume a nonexistent zero-sum game by arbitrarily creating limited spots for top ranks. There are always winners and losers in this system even if all make progress. Zero-sum game, a term from game theory (von Neumann and Morgenstern 1944), refers to a situation in which resources gained by one party are matched by corresponding losses to another party and thus overall competition always produces winners and losers with no total gain or loss. Zero-sum bias may occur among those contestants who judge a situation to be zero-sum when it is actually nonzero sum (Barclay and Stoller 2014; Meegan 2010); nonzero sum refers to any other situations including a win–win situation when both parties may gain. Under the growing external pressure of zero-sum gaming under performance-driven government funding and accountability as well as market-based global university rankings, the question is how we can resolve the tension between competition and collaboration and change zero-sum game mind-set to win–win approach in higher education.

123 Does the Asian catch-up model of world-class universities…

Those external policy and market forces bring new challenges and threats in higher education across the nations. There are challenges for Asian countries’ public universi- ties that operate under many central government restrictions over their finance, research, and educational autonomy (Altbach 2000; Hammond 2016). Further, the language biases of international publication/citation records pose difficulties for non-English-speaking nations to catch up with the US or UK counterparts (van Raan et al. 2011; van Leeuwen et al. 2001). Under the pressures of high-stakes accountability and rankings, there is also growing risk of gaming the system. China’s publication growth was allegedly inflated by manipulations, and the government’s funding bias toward elite universities and STEM fields led to uneven uni- versity development (Horta and Shen 2019; Hvistendahl 2013). As high-stakes educational accountability policies and university rankings always generate both winners and losers, the zero-sum game competitions can have uneven results and impacts for different nations and different types of universities and fields. In light of these concerns, this study examined the trends of global university rankings and contributing factors including cross-national higher education policy forces, particularly in the context of East Asian nations’ high-stakes catch-up policies toward world-class uni- versity development. Which countries emerged as the winners or losers of competitions for world-class university rankings? What factors contributed to the cross-national variations in global university ranking trajectories? Did China and Korea as fast followers replicate the earlier success of Japan’s catch-up with the USA? Did Japan as early catch-up leader manage to advance their university reputations and rankings further? Has the Asian catch-up model of world-class universities worked? If yes, how? If no, why not? This study addressed these questions and shed new light on the Asian catch-up model. Further, the study drew international policy implications for higher education reform and the improvement of global university ranking measures.

2 Data and methods

This study applied mixed methods for investigating the cross-national linkage between higher education policies and global university rankings together: (1) statistical analysis of inter- national data about world-class university rankings in relation to research expenditures, citations, and STEM priorities, and (2) case study of higher education policies among four selected nations based on the literature review and content analysis of policy documents. Given that the trends of global university rankings diverged among the nations over the past decade, we examined a statistical model of world-class university entries based on the indicators of research investment, productivity, and concentration. Building on the results of statistical analysis across nations, we further explored the role of government policies in China and Korea as fast followers and Japan as early catch-up leader in comparison with the USA as established leader. First, our study used two readily available and widely used sources of global univer- sity ranking data (see “Appendix Table 4” for descriptive statistics): (1) ARWU data from ShanghaiRanking Consultancy, including world-class university ranks among 45 nations (i.e., nations that have at least one college included in the top 500 world rankings) during the 2003–2013 period (data retrieved from http://www.shanghairanking.com), and (2) QS data for world-class university ranks among 59 nations (i.e., nations that have at least one college included in the top 500 world rankings) during the 2008–2014 period (data retrieved from http://www.topuniversities.com); years 2003 and 2008 are treated as baselines, the earliest years when each of the two datasets is available for analysis. The Times Higher Education

123 J. Lee et al.

(THE) rankings data were not appropriate for trend analysis due to their changes of the metrics. Second, we used other sources of international data to explain the variations in rank- ings (see “Appendix Table 4).” Each nation’s time-series data on GDP per capita and R&D expenditures were obtained from the World Bank database; R&D expenditures are current and capital expenditures (both public and private) and cover basic research, applied research, and experimental development. For citation shares and STEM concentration variables, we derived data from the SCImago Journal & Country Rank, a publicly available portal (http:// www.scimagojr.com) that includes the journals and country scientific indicators developed from the information contained in the Scopus® database (Elsevier B.V.). For citation shares variable, we used the number of citations by the documents published during the source year and computed each country’s relative contribution (%) to world citation output; Citation data were drawn from over 21,500 titles from more than 5000 international publishers and country performance metrics from 239 countries worldwide (SCImago 2007). Because the measures of publications and citations were highly correlated with each other, we chose to include only the measure of citations as an indicator of research productivity and impact; this choice helps alleviate multicollinearity in regression analysis. For STEM concentration variable, we computed citation shares within each field first and then the ratio of citation shares in science and engineering fields (including health and medical sciences) to citation shares in arts/humanities and social science fields for each country. For the analysis of global university rankings, the total sample sizes were 495 records for ARWU (i.e., 11 years times 45 nations) and 413 records for QS data (i.e., 7 years times 59 nations). Given the nested data structure (i.e., repeated ranking measures within nations), we applied Hierarchical Linear Modeling (HLM), multilevel growth modeling approach in particular, to track individual nations’ changes of global university rankings and examine the sources of international variations in those trends (Raudenbush and Bryk 2002). The following HLM models were used for tracking and explaining changes in each nation’s world-class university counts based on ARWU and QS rankings as outcome variables. Level 1 Model (within-nation analysis):

Yti  π0i + π1i (Time)ti + eti

Y ti is natural logarithm of world-class university entry counts (among top 100, 200 and 500) for nation i at year t;(Time)t is the number of years subtracted from 2013 at year t; Level 2 Model (between-nation analysis):   π  β β β 0i 00 + 01 GDP per capita i + 02 (R&Dexpenditures)i + β03 (citation shares)i + β04 (STEM concentration)i + r0i     π  β + β GDP per capita gain + β R&D expenditures gain 1i 10  11  i  12  i β β + 13 citation shares gain i + 14 STEM concentration gain i + r1i

π 0 is the current status of outcome at year 2013; π 1 is annual growth rate during the period (i.e., outcome gain per year); and GDP per capita is gross domestic product per capita based on purchasing power parity (PPP) (in US $1000) in each country i. R&D expenditures is research and development expenditures as percent of GDP in each country i. Citation shares is the percent world shares of citations for all published documents across fields in each country i. STEM concentration is the ratio of citation shares in science and engineering fields versus arts/humanities and social science fields in each country i. 123 Does the Asian catch-up model of world-class universities…

At level 1 (i.e., within-nation), both current status (as of 2013) and growth rate (annual gain) parameters are estimated to keep track of each nation i’s research outcome variable Y at year t; initial status (π 0i) and growth rate (π 1i)areassumedtovaryasafunctionamongthe nations. At level 2 (i.e., between-nation), we account for cross-national variations in ranking trends by using research funding (R&D expenditures), impact (citation shares), and priority (STEM concentration) measures as key predictors; economic wealth (GDP per capita) is used as a control variable. If we were to use ordinary regression model with this kind of international time-series data and find that citation shares are significant for rankings, we would not be able to determine whether this is due to cross-country differences (e.g., USA always has a higher share of research citations than China) or if increases within each country lead to gains in rankings. Our HLM model above addresses this problem by using a different type of level 2 predictors for level 1 intercept (current status) versus slope (growth rate). When we modeled between-nation differences in terms of the current status of rankings, we used the status values of citations shares, GDP per capita, etc., in corresponding year. In contrast, when we modeled between- nation differences in terms of the growth rate (changes) of rankings, we used the within- country gain measures as opposed to their current status measures. Given such different predictors for initial status versus growth rate, we reported separate results on their effects (e.g., “citations shares status” versus “citation shares gain”). Further, we conducted a supplementary analysis by adding data for nations that did not have any institutions in the top 500 to the dataset. The total sample size was 93 nations including 48 nations that have zero count for ARWU top 500 rank and 34 nations that have zero count for ARWU top 500 rank. We applied what is known as a “hurdle model”: hurdle models are a class of models for count data that help handle excess zeros and overdispersion (Cameron and Trivedi 2013). We applied logistic regression model to the full data (N  93 nations) where we first explained how the predictors relate to being ranked in top 500 at all, and then linear regression model to the subset data (N  45 nations for ARWU and N  59 nations for QS) where we explained how the number of institutions ranked is related to the predictors for those that pass the “hurdle” of having at least one highly ranked institution. Our study has limitations and caveats. The primary unit of analysis in our study is nation, not individual university. There are possible variations and fluctuations of the rankings among universities within each nation over time, but we focus on the aggregate national pattern of changes; it is beyond the scope of study to track individual universities. A subsequent study may address this information gap via three-level HLM design (level 1: time, level 2: university, and level 3: nation). We also choose to restrict our analysis to nations with top 500 ranked universities simply because those were common thresholds for global university rankings; there were no publicly available ARWU data about universities that fall below top 500 ranks. A subsequent study may address this potential sample bias by expanding data collection and including more developing nations. Finally, universities in the UK and other European countries are not included in this study. Our primary focus in this study was East Asian countries, and we chose the USA as a reference which had been a primary benchmarking target for those Asian nations.

123 J. Lee et al.

3 Results

3.1 Trends and variations in world-class university rankings

At the aggregate national level, world-class university ranking orders appears to be highly steady. The high correlations of world-class university counts (i.e., the number of universities in top 100, 200 and 500 ranks within nations) across all years suggest that the cross-national rank order has hardly changed over the long period. For example, the USA still remains as the highest ranked nation with the largest number of top 100, 200, 500 universities in the world in spite of continuing drops in those counts. Notwithstanding this overall ranking stability at the aggregate national level, however, there exist substantial variations in terms of growth rates among the nations. Random effects analysis of the ARWU data (unconditional model without level 2 predictors) reveals significant variations in the growth rate of world-class university entries among all 45 nations. To illustrate those cross-national variations, Fig. 1 shows the trends of ARWU 2003–2013 world-class universities in top 500 ranks for China, Japan, Korea, and USA. The trends are divergent among the four nations, with bifurcations of either positive or negative changes. The USA and Japan made some setback, while both Korea and China have made progress to catch up with them. Figure 1 also shows the trends of QS 2008–2014 world-class university entries in top 500 ranks for China, Japan, Korea, and USA. While the actual counts of the universities in top 500 ranks for these four countries are different between ARWU and QS ranking systems, their trends in the two ranking systems are similar. China made significant increase of world-class university entries in the top 200 and top 500 ranks, but not in the top 100. Korea made significant gain only in the top 500 ranks, but not in the top 100 and 200. In contrast, Japan had significant loss of their university entries

ARWU Top 500 Count QS Top 500 Count

180 120 160 100 140

120 80 100 60 80

60 40 40 20 20

0 0 2003 2004 2006 2012 2013 2007 2008 2010 2011 2005 2009 2012 2014 2008 2010 2009 2011 2013 US Korea US Korea Japan China Japan China

Fig. 1 Trends of top 500 world-class university counts as ranked by ARWU 2003–2013 (left panel) and QS 2008–2014 (right panel) among China, Japan, Korea, and the USA 123 Does the Asian catch-up model of world-class universities…

Table 2 Results from the HLM World-class university counts analyses of ARWU 2003–2013 world-class university entry Top 100 Top 200 Top 500 trends Current status (as of 2013) Intercept 0.52*** 0.90*** 1.82*** GDP per capita 0.00 0.01 0.00 R&D expenditures 0.12 0.14 0.11* Citation shares 0.48*** 0.68*** 0.79*** STEM concentration − 0.45*** − 0.49*** − 0.37*** Growth rate (annual gain) Intercept 0.00 0.01 0.02 GDP per capita gain 0.00 0.00 0.00 R&D expenditures gain 0.00 0.02 0.02 Citation shares gain 0.01 0.04 0.38*** STEM concentration gain − 0.01 0.03 − 0.34* *p <.05, **p <.01, ***p <.001 in the top 500 ranks. Similarly, USA made significant losses of their university entries in the top 200 and top 500 ranks. Consequently, the gaps in world-class university entries have narrowed between the USA and Japan as initial leader and China and Korea as fast follower. In the midst of US and Japan setbacks, China has already surpassed Japan and it is also closing the gap with the USA fast. Although the USA still maintains the first place overall in comparison with the rest of the world, particularly with elite universities (i.e., a few top-tier public universities as well as private universities including the Ivy League), it is noteworthy that the loss of world-class universities as measured by ARWU and QS happened mostly outside of top 100 ranks, among largely second-tier public universities such as nonflagship state universities. In fact, some American universities from top 101–200 ranks descended to 201–500 ranks, and at the same time some of those that were originally in top 201–500 ranks moved out of ranking at all.

3.2 Contributing factors: research funding, citations, and STEM concentration

Table 2 summarizes the estimates of HLM final model (conditional model with level 2 pre- dictors) that account for between-nation variations in ARWU world-class university entries. As expected, the average growth rates of all nations (i.e., intercept for growth rate) are zero, revealing the zero-sum game nature of international ranking competitions. This finding con- firms the fact that the ranking system always produces equal number of winners and losers. With control for the nations’ economic wealth (GDP per capita), research funding as mea- sured by R&D expenditures has only partially significant relationships with the status of top 500 university entries, but not top 100 and 200 levels. In contrast, citation shares show consistently significant positive effects at all top 100, 200, and 500 university levels for status and at top 500 level for growth (Table 2). For the 2013 status of ARWU entries as outcome, one percent increase in world research citation shares was associated with about one-half to three-quarter percent increase in top-ranked university counts; the regression coefficient estimate of their relationship gets relatively stronger for the less selective league (β03  .48, p <.001 for top 100; β03  .68, p <.001 for top 200; β03  .79, p <.001 for top 500). Figure 2 demonstrates one of such positive relationships between citations and rankings at the national level. For the growth rate of ARWU entries as outcome, 123 J. Lee et al.

Fig. 2 Scatterplots of top 500 world-class university counts versus research citation shares (both variables are shown in natural log, 2013 data)

Fig. 3 Scatterplots of top 500 world-class university counts versus STEM concentration (both variables are shown in natural log, 2013 data) a similar pattern of the relationship, albeit relatively weaker, was found; one percent yearly gain of the nations’ citation shares was associated with a four-tenth percent annual gain of their top-ranked university entries (β13  .38, p <.001 for top 500). Conversely, STEM concentration, as a proxy indicator of research productivity/impact in science and engineering fields vis-à-vis arts/humanities and social sciences, is negatively associated with the status of ARWU world-class university entries (β04 −.45, p <.001 for top 100; β04 −.49, p <.001 for top 200; β04 −.37, p <.001 for top 500). Figure 3 123 Does the Asian catch-up model of world-class universities…

Table 3 Results from the HLM World-class university counts analyses of QS 2008–2014 world-class university entry Top 100 Top 200 Top 500 trends Current status (as of 2014) Intercept 0.50*** 0.83*** 1.63*** GDP per capita 0.01* 0.01* 0.00 R&D expenditures 0.09 0.10 − 0.02 Citations shares 0.25*** 0.39*** 0.54*** STEM concentration − 0.21 − 0.29* − 0.35** Growth rate (annual gain) Intercept 0.00 0.00 0.01 GDP per capita gain 0.00 0.00 0.00 R&D expenditures gain 0.02** 0.02 0.00 Citations shares gain 0.02 0.14** 0.24 STEM concentration gain − 0.02 − 0.06 − 0.18 *p <.05, **p <.01, ***p <.001

20% 20% 20% 60% China Japan Korea US

50% Arts & 15% 15% 15% Humanies 40% Engineering 10% 10% 10% 30%

20% Natural Sciences 5% 5% 5% 10% Social Sciences 0% 0% 0% 0% 1999 2005 2011 2008 1996 2002 2005 1996 1999 2002 2005 2008 2011 1996 1999 2002 2008 2011 1996 1999 2002 2005 2008 2011 Fig. 4 Trends of research citations (% world shares) in arts and humanities, engineering, natural sciences, and social sciences among China, Japan, Korea, and the USA illustrates one of such negative relationships between STEM concentration and world-class university counts; those East Asian countries including China, Japan, and Korea stand out as outliers in this scatterplot. Despite some differences in the level of significance, similarly positive or negative patterns of the relationships emerge between citation share and STEM priority as predictors and QS world-class university entries as outcomes (see Table 3). Once those explanatory factors are taken into account, countries such as China and Korea that pushed for improving research productivity and university rankings still perform significantly better than expected, whereas their counterparts such as Japan and USA perform worse than expected. HLM residual analyses show that gaps in adjusted top 500 university count growth rates get smaller but some of them remain significant. Figure 4 provides a closer look at the trends of research citation (% world shares) as a proxy indicator of research impact by field for China, Japan, Korea, and USA. As with the world-class university ranking trends, these citation trends are divergent among the four nations, with bifurcations of either positive changes for China and Korea or negative changes for Japan and the USA. For China and Korea, the growth of research citation shares occurred 123 J. Lee et al. much faster in science and engineering fields than in arts/humanities and social sciences; consequently, their STEM concentration has increased over the period in terms of research productivity and impact relative to the world. For Japan, the decline in research citation shares occurred much faster in science and engineering fields than in arts/humanities and social sciences. For the USA, the decline in research citation shares occurred equally across all fields so that STEM concentration has not changed significantly. The above analysis does not include nations that did not have any highly ranked institutions. To address that limitation, our “hurdle model” analysis examined the same set of factors to differentiate nations with zero count (coded zero) versus nations with at least one count in the top 500 (coded one). The results of binomial logistic regression analysis for ARWU and QS rankings were consistent, showing that research citation shares were both statistically and practically significant predictor of the outcome, that is, whether the nations pass the hurdle of being counted at all by those world-class university ranking measures. The estimate of citation share effect on being ranked in logit coefficient was 12.12, p <.01forARWUand 25.78, p <.01 for QS; even tiny increase in research citations would drastically increase the chance of being ranked. The other predictors in the model, including GDP per capita, R&D expenditures and STEM concentration variables, did not significantly account for whether nations have any institutional entry in the top 500; those factors only mattered once nations passed the threshold of their institutions being counted at all.

3.3 Higher education policies and university rankings

To account for the underlying policy forces of aforementioned university ranking trends and contributing factors, we considered the broader social and educational policy contexts that may have influenced those cross-national variations and changes. In the following section, we examined the cases of China, Japan, Korea, and the USA in terms of their performance- driven higher education policies and potential impact on their university rankings. This case study features the Asian catch-up model of world-class university development initiatives.

3.3.1 The case of China

In China, the government launched 211 Project and 985 Project in the 1990s to improve the stature of Chinese universities in the world, particularly boosting top-tier Chinese universities’ world rankings (Yang and Welch 2012). In 1995, the 211 Project was formally initiated for purpose of realizing this goal (Ministry of Education of the People’s Republic of China 2008). As of 2011, 112 universities were selected for funding in this project. From 1995 to 2011, around 3 billion dollars (USD) were provided to those 112 universities from the central government alone (People’s Daily 2008). When also taking account of the funding from the local governments and other related agencies, more than 6 billion dollars (USD) were distributed (Sohu News 2012). The funding was used to build up the key disciplines, to create the public service systems of the higher education, and to enhance the infrastructure of universities (Ministry of Education of the People’s Republic of China 2008). Building upon the 211 Project, China launched the 985 Project in 1999; the aim of this project was to put selected Chinese universities on the list of the world’s top research univer- sities by the year 2020 (Ying 2011). In total, from 1999 to 2015, about 14.7 billion dollars (USD) were invested in the project (China News 2012); it allocated an additional 10–20% of research expenditures to a set of 39 universities, among which the top nine of them were 123 Does the Asian catch-up model of world-class universities… also funded through the earlier 211 Project. Funding priorities of these government projects concentrated on STEM fields. The evaluation of the 211 and 985 projects has been mixed. During the past two decades, those two projects created significant research capacity and contributed to improvements in the global rankings. First of all, they helped boost publications and citations as a result of the government’s project refunding criterion; for example, the research publications as produced by the “Project 211” universities increased by 94% during the first phase (Lixu 2004). The ARWU rankings of Chinese universities that received government support under the 211 and 985 projects improved significantly, while some of them newly joined among top 500 ranks (see “Appendix Table 5).” Stepping up the pace of developing world-class universities, the Chinese government launched the World-Class 2.0 project in 2015, aimed at becoming a global higher education center. However, the lack of institutional autonomy and academic freedom, academic corruption, and intellectual property infringement remain critical challenges for China to overcome (Li 2018; Eesley et al. 2016).

3.3.2 The case of Japan

In Japan, the Ministry of Education led national university restructuring initiatives since 2001, including the reorganization and integration of universities and fields, the introduc- tion of private-sector management methods, and performance-driven third-party evaluation system (Yonezawa 2003). Unlike China and Korea, however, Japan’s primary challenge was to build on the earlier catch-up success in STEM fields and further advance university internationalization; they already had 36 universities among top 500 ranks as of 2003. The government’s support to the universities was more targeted, program-based; in the midst of debates and criticisms, the original government plan of supporting top 30 universities changed to the plan of supporting research units across institutions. Through competitive grant applications, selected programs within universities rather than the whole universities received the government funding. A set of funding opportunities were available, including the Twenty-First-Century Centers of Excellence (COE21), the Global Centers of Excellence (Global COE), and the World Premier International Research Center Initiative (WPI); these initiatives were seen as trials of performance-based university funding system (Yonezawa 2003). COE21 was established in 2002. It was designed to “cultivate a competitive academic environment among Japanese universities by giving targeted supported to the creation of world-standard research and education bases” (Japan Society for the Promotion of Science 2001). Ph.D.-granting departments of graduate schools were eligible for funding (ranging from 0.1 million to 4.8 million dollars per year) that would last for 5 years. Later, Global COE started in 2007 for establishing education and research centers that elevate the international competitiveness of the Japanese universities (Japan Society for the Promotion of Science 2001). Compared with the COE21, Global COE allowed a wider range of disciplines to apply for the funding, but less programs were approved (Qi 2011). In 2007, in addition to the Global COE, Japan launched another competitive funding program entitled “World Premier International Research Center (WPI) Initiative” to expedite the development of globally visible research centers through concentrated support to a limited number of proposals for a period of 10 years. While WPI was more selective than the COE21 and Global COE (Qi 2011), it strives for “building top world-level research centers staffed by the highest caliber of core researchers” (Japan Society for the Promotion of Science, 2011. For the FY2007 grant, only five projects were selected; they were limited to STEM fields and all but one came from former imperial universities (Oba 2008). In 2014, Japan launched 123 J. Lee et al. another initiative, “Top Global University Project” which funds step-up efforts for university reform and internationalization (Shimmi and Yonezawa 2015). In spite of these government efforts, the world rankings of many Japanese universities that received government support under the COE and WPI projects hardly improved or even dropped among top 500 ranks (see “Appendix Table 5).”

3.3.3 The case of Korea

The Korea Ministry of Education and Human Resources, in response to concern over the relatively low standing of the nation’s universities and researchers, launched the Brain Korea 21 program (BK21) since the late 1990s (Seong et al. 2008). BK21 sought to make Korean research universities globally competitive and to produce more high-quality researchers in Korea. The Korean government invested approximately $1.2 billion (USD) in selected uni- versities during the 1999–2005 period; about three-quarters of the BK21 budget was used to support graduate students in the form of stipends, scholarships, and research infrastruc- ture (Kim 2007). Funding model was similar to Japan, for which competitively selected team of researchers in the top universities are supported on a project basis. In Phase I of BK21 (1999–2005), approximately 90% of funding was allocated to 48 research groups in 14 universities in the areas of science and engineering, and in Phase II (2006–2012), top ten universities received 80% of the total budget (Um 2013). The evaluation of these Korea government initiatives was favorable in terms of increased enrollment of graduate students and enhanced faculty research productivity (Kim 2007; Moon and Kim 2001; Seong et al. 2008;Shin2009;Um2013); top universities showed rapid increases in the quantity of publications including SCI-indexed journals in science and engineering fields. The world rankings of Korean universities that received government support under the BK21 projects improved or newly joined top 500 ranks (see “Appendix Table 5”). The Korea government also launched the World Class University (WCU) program for attracting talented and renowned foreign scholars from overseas, with the realization that the quality and impact of publications were not satisfactory in spite of rapid gains in numbers during BK21 period (Um 2013). The budget for WCU was USD $617 million for the 5-year period (2008–2012). The WCU project funding was also concentrated in STEM fields. In terms of building human resources, the program has not reached the target level required for attracting foreign scholars or students (Um 2013). Moreover, the Korean government’s BK21 priorities on advancing the economic contributions of universities were not highly compatible with the institutional goal of nurturing academic excellence (Brewer et al. 2002).

3.3.4 The Case of the USA

The focus of higher education accountability policy in the USA shifted from college inputs to performance outcomes since late 1990s. However, unlike East Asian counterparts that focused on global university rankings, research and graduate education, the US states’ policy initiatives prioritized undergraduate education issues and did not pay attention to global university rankings. In fact, state-by-state college performance report cards hardly include research (except for an indicator of sponsored research) and consider graduate studies only in terms of total degree attainment. In fact, the most popular indicators used by the states for accountability purposes tended to focus on student outcomes such as graduation, retention, and job placement rather than research productivity (Burke 2005). Recent study shows that 123 Does the Asian catch-up model of world-class universities… performance funding policies may have heterogeneous effects on graduation and retention rates, disproportionately benefiting institutions already positioned to perform well (Favero and Rutherford 2019). On the other hand, top research universities, including elite private universities and selec- tive public universities, continued to focus on improving research excellence. Both federal and state governments continued to invest billions of dollars in university-based research since research capacity remains as the primary factor distinguishing US universities from others in the rest of the world (Cole 2010). However, the USA has lost international competitive edge in STEM fields. The national report, Rising Above the Gathering Storm (NAS/NAE/IOM 2007), raised concerns about the growing US deficiencies in STEM fields, which led to the passage of the America COMPETES Act of 2007. Performance-driven, high-stakes account- ability policies using the indicators of graduation rates and research expenditures may have brought unintended negative consequences such as curtailing arts/humanities education and narrowing institutional mission to more fundable areas such as STEM (Dougherty and Reddy 2011). Although the USA has maintained the first place in terms of world-class university entries for both ARWU and QS ranking, it lost 12 institutions for ARWU over 2003–2013 period and 11 for QS over 2008–2014 period; the losers were mostly second-tier public institutions, while the reputations of elite private and top-tier public universities remained intact. These divergent patterns of ranking changes imply highly stratified system of higher education in which improving the university’s research reputation (and thus rankings) among world- class research universities was only a concern among a few elite private or top-tier public universities, whereas a majority of public universities scramble for institutional survival with declining student enrollment and government support (Powell et al. 2012).

4 Discussion

International brain race for building “world-class research universities” became more intense due to increasingly competitive, performance-driven government funding and accountability policies as well as increasingly well-publicized, global university rankings. The results of this study suggest that international rankings may have had powerful influences on national higher education policies and benchmarking strategies, particularly in Asia. Here, we need to note nuanced differences between Asian nations and the USA in their approaches. Since higher education policy in the USA is largely formulated and implemented differently in each of the 50 individual states, international rankings are never cited or discussed as part of policy discussions; instead, domestic rankings such as the US News play a role. In contrast, Asian nations where central governments control higher education policies as part of globalization strategies are very sensitive to the international rankings. In the case of China, the AWRU has a powerful influence as a state-sponsored benchmarking tool that allows Chinese universities to compete on the international stage. The study reveals the zero-sum game nature of international university ranking competi- tions among the nations over the past decade, including the USA and Japan as initial leader (both becoming losers by top 500 rank losses) versus China and Korea as fast followers (both becoming winners by top 500 rank gains). The growth of China and Korea as opposed to the setbacks of Japan and the USA gives sharp contrast in terms of world-class university entry counts as measured by both ARWU during the 2003–2013 period and QS during the

123 J. Lee et al.

2008–2014 period. While all of these four nations improved R&D expenditures relative to GDP,their varying paces of investment affected research productivity and university rankings. The findings of this study raise questions: What are the different strategies that “state–cen- tered” Asian systems have from other Western systems? In addition, are there any differences between those three Asian countries in their policy approaches? How were China and Korea able to pull rapid growth (China outpacing Korea) while Japan was not? First of all, the Asian higher education systems feature strong government regulations and central funding mechanisms. which uniformly affect all colleges and universities in the nation, including private institutions. Under the banner of national campaign for building world-class univer- sities, institutional resources and manpower were mobilized across the nation for achieving the new government priorities. China and Korea both attempted to replicate the earlier suc- cess of Japan’s “catch-up” strategies through their governments’ performance-driven targeted funding with STEM priorities which raised international publications/citations, which in turn helped improve their global university rankings. While a primary reason for Japan’s struggle with global university rankings should be the matured system’s difficulty to maintain their status, the differential pace of growth among the three Asian nations might be also attributable to systemic differences in their conditions as well as strategies for developing world-class universities. Japan’s relatively smaller uni- versity/faculty pool as well as weaker financial incentive of the policy may have undermined their growth in comparison with China. China focused more on lifting whole institutions (i.e., targeting preselected universities and schools), whereas both Japan and Korea concen- trated on funding research projects and centers. Moreover, given very big faculty salary gap (purchasing power adjusted) between Japan and China, even similar amount of government funding for university and faculty incentive program was likely to have much different impact on their productivity; Chinese universities also actively recruited faculty from overseas by offering very high Western-style salaries to select elites (Altbach et al. 2012). Finally, differ- ences among the three nations in the intensity of their major universities’ STEM emphasis also may have contributed to the gaps in their ranking gains; the percentage of top 500 uni- versities with STEM focus, priority, or orientation was 100% in China, 62% in Korea, and 35% in Japan (Cheng and Liu 2006). In spite of rapid growth of the university rankings, this study reveals inherent limitations and potential biases of this Asian catch-up model for world-class university development. First, the growth of Chinese and Korean universities was hardly seen among top 100 universi- ties, limited largely to the second and third tiers (top 200 and 500 levels). This finding might be also related to the nonlinear pattern of relationship (i.e., diminishing returns) between citations and rankings; the effects of citations are relatively stronger for top 500 level than for top 100 level (Tables 2, 3). Second, the decline or stagnation of Japanese universities among top 500 ranks also reveals that the Asian model of centralized performance-driven university development (research and development in general) could run out of its utilities after certain stage; it may work better for the early stage of development for catch-up but may not work for advanced stages of development for innovation and leadership. In the same vein, Korea and China may fall into the trap of benchmarking, thus following Japan’s suit if they fail to evolve from the “catch-up” model to “first mover” strategies for leading innovations.

123 Does the Asian catch-up model of world-class universities…

Another potential pitfall of the Asian model is relative imbalance of institutional support for multiple disciplines due to the stronger STEM dominance in research publi- cations and citations. Among top 500 universities, there are more STEM-oriented (vis-à-vis arts/humanities/social sciences-oriented) institutions in Asian higher education systems than in Western counterparts (Cheng and Liu 2006). Although STEM-oriented funding priori- ties and incentives helped build their competitive niche, the Asian model is not congruent with the international pattern of the relationship between STEM concentration and world- class university entries; Western nations with lower STEM concentration (i.e., more balance between STEM and arts/humanities/social sciences) tend to have more top-ranked universi- ties (Tables 2, 3). For the Asian nations, their unique pattern as outliers may not only reflect their national strategic priority in STEM fields for rapid economic development but also it implies their English language barriers that restrict comparable access to international schol- arly networks and journals, particularly in non-STEM fields. The government policy that favors top national research universities in STEM fields may risk exacerbating existing gaps and inequalities within the Asian nations. Given these limitations and biases, it is questionable whether the Asian catch-up model of high-stakes funding competition and performance-driven accountability policy approach is a viable choice for long-term success. How can Asian universities become more like American universities? This question may have guided East Asian nations that attempted to benchmark success against American top universities as a role model. Although bench- marking strategies helped them rapidly modernize universities and advance economic development, performance-driven and ranking-driven policies drive zero-sum game biases which undermine real improvement and innovation. When performance-driven, high-stakes accountability policies relied more heavily on short-term incentives (e.g., competitive rewards and grants) as a policy instrument rather than long-term capacity-building and academic free- dom, they are likely to induce perverse incentives such as gaming the system for rewards without genuine and long-term improvement. From this critical perspective, it would be wise for Asian nations to reframe a question for world-class university development: “How should we distinguish us from American counterparts?” This strategic differentiation approach is more likely to create win–win results when each nation pursues more comprehensive yet distinctive world-class universities. Then, the challenge is not so much catching up with established leaders as distinguishing ones from peers. We recommend alternative approach that involves win–win strategies for major universities across the nations. Unlike the notion of “world-class university” that is almost exclusively identified with global university rankings, the concept of “flagship university” is more fluid and influenced by national context and local needs (Douglas 2016). The USA has one or more “flagship” public universities for each state, which tends to be a land grant university with the highest research profile and most doctoral programs. This traditional model of flagship university can be extended to public universities that play other important roles such as public service and regional development beyond research and teaching. Tied to geographic and cultural representation, even smaller regional or local university may also develop strategic strengths in selected fields. Further, more regional collaboration and integration efforts in research may help produce win–win results (Hammond 2019).

123 J. Lee et al.

To reform the current university ranking system, putting more emphasis on the indicator scores as opposed to the overall rankings also would help transcend the research-dominant criteria of world-class university and dispel the myth of “one-size-fits-all” university rank- ings. The breakdown of performance measures by subfields also can reveal different areas of strengths and weaknesses. Our study demonstrated relative strength of Asian systems in STEM but weakness in non-STEM fields by breaking down publication/citation records by fields. In a similar vein, analyses of the various performance indicators may be more infor- mative and useful. These changes will also help universities develop independent academic culture and internal accountability system for self-improvement.

Appendix

See Tables 4 and 5.

Table 4 Descriptive statistics of ARWU and QS variables Variables ARWU data QS data (N  45 nations) (N  59 nations)

Mean SD Mean SD

Top 100 entries 2.23 8.04 1.70 4.80 Top 200 entries 4.45 13.39 3.39 8.10 Top 500 entries 11.14 24.40 8.43 15.47 GDP per capita 35.72 15.92 31.92 17.53 R&D expenditures 1.76 1.01 1.51 1.13 Citations shares 2.13 3.81 1.65 3.43 STEM concentration 1.51 1.06 1.62 1.32 Top 100 entries, top 200 entries, and Top 500 entries are total number of university entry counts among top 100, top 200, and top 500 ranks, respectively, within each ranking system GDP per capita is gross domestic product per capita based on purchasing power parity (PPP) (in US $1000) R&D expenditures is research and development expenditures as percent of GDP Citation shares is the percent world shares of citations for all published documents across fields STEM concentration is the ratio of citation shares in science and engineering fields versus arts/humanities and social science fields

123 Does the Asian catch-up model of world-class universities…

Table 5 List of top 500 world-class universities and their government funding program sources in China, Japan, and Korea Nation Count Universities Rank in 2003 Rank in 2013 Changes in Government listed in ranking in funding 2003–2013 2003–2013 programs

China 1 Tsinghua 201–250 151–200 Improved 985 Project University 2 Peking 251–300 151–200 Improved 985 Project University 3 Fudan 301–350 151–200 Improved 985 Project University 4 Nanjing 301–350 201–300 Improved 985 Project University 5 University of 301–350 201–300 Improved 985 Project Science and Technology of China 6 Zhejiang 351–400 151–200 Improved 985 Project University 7 Shanghai Jiao 401–450 151–200 Improved 985 Project Tong University 8 Jilin University 451–500 301–400 Improved 985 Project 9 Shandong 451–500 301–400 Improved 985 Project University 10 Sun Yat-sen No rank 201–300 Added 985 Project University 11 Beijing Normal No rank 301–400 Added 985 Project University 12 China No rank 301–400 Added 985 Project Agricultural University 13 Harbin Institute No rank 301–400 Added 985 Project of Technology 14 Huazhong No rank 301–400 Added 985 Project University of Science and Technology 15 Sichuan No rank 301–400 Added 985 Project University 16 Xian Jiao Tong No rank 301–400 Added 985 Project University 17 Beihang No rank 401–500 Added 985 Project University 18 Central South No rank 401–500 Added 985 Project University 19 Dalian No rank 401–500 Added 985 Project University of Technology 20 Lanzhou No rank 401–500 Added 985 Project University

123 J. Lee et al.

Table 5 continued Nation Count Universities Rank in 2003 Rank in 2013 Changes in Government listed in ranking in funding 2003–2013 2003–2013 programs

21 Nankai No rank 401–500 Added 985 Project University 22 Peking Union No rank 401–500 Added See Note1 Medical College 23 South China No rank 401–500 Added 985 Project University of Technology 24 Southeast No rank 401–500 Added 985 Project University 25 Tianjin No rank 401–500 Added 985 Project University 26 Tongji No rank 401–500 Added 985 Project University 27 Wuhan No rank 401–500 Added 985 Project University 28 Xiamen No rank 401–500 Added 985 Project University Japan 1 The University 19 21 Dropped COE21, Global of Tokyo COE, WPI 2 Kyoto 30 26 Improved COE21, Global University COE, WPI 3Osaka 53 85 Dropped COE21, Global University COE, WPI 4 Tohoku 64 101–150 Dropped COE21, Global University COE, WPI 5Nagoya 68 101–150 Dropped COE21, Global University COE, WPI 6 Hokkaido 102–151 101–150 Same COE21, Global University COE, WPI 7 Kyushu 102–151 151–200 Dropped COE21, Global University COE, WPI 8 Tokyo Institute 102–151 101–150 Same COE21, Global of Technology COE 9 University of 102–151 151–200 Dropped COE21, Global Tsukuba COE, WPI 10 Kobe University 201–250 201–300 Same COE21, Global COE 11 Hiroshima 251–300 301–400 Dropped COE21 University 12 Keio University 251–300 301–400 Dropped COE21, Global COE, WPI 13 Niigata 251–300 401–500 Dropped COE21, Global University COE

123 Does the Asian catch-up model of world-class universities…

Table 5 continued Nation Count Universities Rank in 2003 Rank in 2013 Changes in Government listed in ranking in funding 2003–2013 2003–2013 programs

14 Chiba University 301–350 Dropped COE21, Global COE 15 The University 301–350 401–500 Dropped COE21 of Tokushima 16 Tokyo 301–350 Dropped COE21 Metropolitan University 17 Waseda 301–350 301–400 Same COE21, Global University COE 18 Gunma 351–400 Dropped COE21, Global University COE 19 Juntendo 351–400 Dropped COE21 University 20 Kanazawa 351–400 401–500 Dropped COE21 University 21 Osaka City 351–400 401–500 Dropped COE21, Global University COE 22 Shinshu 351–400 Dropped COE21, Global University COE 23 Yamaguchi 351–400 Dropped Global COE University 24 Ehime 401–450 Dropped COE21, Global University COE 25 Gifu University 401–450 Dropped COE21 26 Kumamoto 401–450 Dropped COE21, Global University COE 27 The Graduate 401–450 Dropped University for Advanced Studies 28 Tokyo Medical 401–450 301–400 Improved COE21, Global and Dental COE University 29 Jichi Medical 451–500 Dropped COE21 School 30 Jilin University 451–500 Dropped 31 Kagoshima 451–500 Dropped University 32 Kinki University 451–500 Dropped COE21, Global COE 33 Mie University 451–500 Dropped 34 Nara Institute of 451–500 Dropped COE21, Global Science and COE Technology 35 Science 451–500 Dropped COE21, Global University of COE Tokyo

123 J. Lee et al.

Table 5 continued Nation Count Universities Rank in 2003 Rank in 2013 Changes in Government listed in ranking in funding 2003–2013 2003–2013 programs

36 Tokyo 451–500 Dropped COE21 University of Agricultural and Technology 37 Okayama No rank 301–400 Added COE21 University 38 Nagasaki No rank 401–500 Added COE21, Global University COE Korea 1 Seoul National 152–200 101–150 Improved BK 21 University 2 Yonsei 251–300 201–300 Improved BK 21 University 3 Korea Advanced 301–350 201–300 Improved BK 21 Institute of Science and Technology 4 Pohang 301–350 301–400 Same BK 21 University of Science and Technology 5 Hanyang 401–450 301–400 Improved BK 21 University 6 Korea University 401–450 301–400 Improved BK 21 7 Kyungpook 401–450 401–500 Same BK 21 National University 8 Sungkyunkwan 401–450 201–300 Improved BK 21 University 9 Catholic No rank 401–500 Added BK 21 University of Korea 10 Kyung Hee No rank 401–500 Added BK 21 University 11 Pusan National No rank 401–500 Added BK 21 University The following government programs are included: 985 Project (China); COE21 (See http://www.jsps.go. jp/english/e-21coe/04.html), Global COE (See http://www.jsps.go.jp/english/e-globalcoe/04_selection.html), WPI (See http://www.jsps.go.jp/english/e-toplevel/04_centers.html) (Japan); and BK21 (Korea). The USA is not included here since there are no government funding programs for world-class university development at the national level 1Peking Union Medical College is a special case of the 985 Project. As a university, it is not on the 985 Project list. But, it has connection with Tsinghua University, and on its official Web site, it considers itself belonging to the 985 Project

123 Does the Asian catch-up model of world-class universities…

References

Aghion, P., Dewatripont, M., Hoxby, C., Mas-Colell, A., & Sapir, A. (2010). The governance and performance of universities: Evidence from Europe and the US. Economic Policy, 25(61), 8–59. Altbach, P. G. (2000). Asia’s economic aspirations: Some problems. International Higher Education, 19, 7–8. Altbach, P. G. (2004). The costs and benefits of world-class universities. Academe, 90(1), 20–23. Altbach, P. G., Reisberg, L., Yudkevich, M., Androushchak, G., & Pacheco, I. F. (Eds.). (2012). Paying the professoriate: A global comparison of compensation and contracts. New York: Routledge. Barclay, P.,& Stoller, B. (2014). Local competition sparks concerns for fairness in the ultimatum game. Biology Letters, 10(May), 1–4. Brewer, D. J., Gates, S. M., & Goldman, C. A. (2002). In pursuit of prestige: Strategy and competition in US higher education. New Brunswick, NJ: Transaction Press. Burke, J. C. (Ed.). (2005). Achieving accountability in higher education: Balancing public, academic and market demands. San Francisco, CA: Jossey-Bass. Cameron, A. C., & Trivedi, P. K. (2013). Regression analysis of count data. Cambridge: Cambridge University Press. Cantwell, B., & Taylor, B. J. (2013). Global status, inter-institutional stratification, and organizational seg- mentation: A time-dynamic Tobit analysis of ARWU position among US universities. Minerva, 51(2), 195–223. Cheng, Y., & Liu, N. C. (2006). A first approach to the classification of the top 500 world universities by their disciplinary characteristics using scientometrics. Scientometrics, 68(1), 135–150. China News (2012). http://www.chinanews.com/edu/2012/12-27/4443375.shtml. Retrieved 01 Sept 2018. Cole, J. R. (2010). The great american university: Its rise to preeminence, its indispensable national role, why it must be protected. New York, NY: Public Affairs. Dougherty, K. J. & Reddy, V. (2011). The impacts of state performance funding systems on higher education institutions: Research literature review and policy recommendations. CCRC working paper. Douglas, J. A. (2016). The evolution of flagship universities: From the traditional to the new. UC Berkeley Center for Studies in Higher Education Research & Occasional Paper Series: CSHE.11.16. http://cshe. berkeley.edu. Retrieved 01 Sept 2018. Eesley, C., Li, B., & Yang, D. (2016). Institutional change and high-tech entrepreneurship? Evidence from China’s Project 985. Organization Science, Articles in Advance, 27, 1–16. Favero, N., & Rutherford, A. (2019). Will the tide lift all boats? Examining the equity effects of performance funding policies in U.S. higher education. Research in Higher Education, 61, 1–25. Hammond, C. D. (2016). Internationalization, nationalism, and global competitiveness: a comparison of approaches to higher education in China and Japan. Asia Pacific Education Review, 17, 555–566. Hammond, C. D. (2019). Dynamics of higher education research collaboration and regional integration in Northeast Asia: A study of the A3 Foresight Program. Higher Education, 78, 653–668. Hazelkorn, E. (2015). Rankings and the reshaping of higher education: The battle for world-class excellence. London: Palgrave Macmillan. Horta, H., & Shen, W. (2019). Current and future challenges of the Chinese research system. Journal of Higher Education Policy and Management, 20, 1–21. Hvistendahl, M. (2013). China’s publication bazaar. Science, 342(6162), 1035–1039. Japan Society for the Promotion of Science. (2001). https://www.jsps.go.jp/english/e-21coe/index.html. Retrieved 01 Sept 2018. Kim, K. (2007). A great leap forward to excellence in research at Seoul National University, 1994–2006. Asia Pacific Education Review, 8(1), 1–11. Kim, J. (2018). The functions and dysfunctions of college rankings: An analysis of institutional expenditure. Research in Higher Education, 59(1), 54–87. Lee, J. (2016). The anatomy of achievement gaps: Why and how american education is losing (but can still win) the war on underachievement. New York: Oxford University Press. Levy, G. D., & Ronco, S. L. (2012). How benchmarking and higher education came together. New Directions for Institutional Research, 156, 5–13. Li, S. (2018). From 985 to World Class 2.0: China’s Strategic Move. Inside Higher Ed. https://www. insidehighered.com/blogs/world-view/985-world-class-20-chinas-strategic-move. Retrieved 01 Sept 2018. Liu, N. C., & Cheng, Y. (2005). The academic ranking of world universities. Higher Education in Europe, 30(2), 127–136. Lixu, L. (2004). China’s higher education reform 1998–2003: A summary. Asia Pacific Education Review, 5(1), 14–22.

123 J. Lee et al.

Marginson, S. (2018). Higher education, economic inequality and social mobility: Implications for emerging East Asia. International Journal of Educational Development, 63, 4–11. McCormack, J., Propper, C., & Smith, S. (2014). Herding cats? Management and university performance. Economic Journal, 124(578), 534–564. Meegan, D. V. (2010). Zero-sum bias: Perceived competition despite unlimited resources. Frontiers in - chology, 191, 1–7. Ministry of Education of the People’s Republic of China. (2008). http://old.moe.gov.cn/publicfiles/business/ htmlfiles/moe/moe_846/200804/33122.html. Retrieved 01 Sept 2018. Moon, M., & Kim, K. (2001). A case of Korean higher education reform: The Brain Korea 21 Project. Asia Pacific Education Review, 2(2), 96–105. National Research Council. (2012). Improving measurement of productivity in higher education. Washington, DC: The National Academies Press. Oba, J. (2008). Creating world-class universities in Japan: Policy and initiatives. Policy Futures in Education, 6(5), 629–640. Osborne, D., & Gaebler, T. (1992). Reinventing government: How the entrepreneurial spirit is transforming the public sector. Reading, MA: Addison-Wesley. Paul, A. M., & Long, V. (2016). Human-capital strategies to build world-class research universities in Asia: Impact on global flows. In M. Chou, I. Kamola, & T. Pietsch (Eds.), The transnational politics of higher education (pp. 142–167). New York: Routledge. People’s Daily. (2008). http://paper.people.com.cn/rmrbhwb/html/2008-03/26/content_48323392.htm. Retrieved 01 Sept 2018. Pietrucha, J. (2017). Country-specific determinants of world university rankings. Scientometrics. https://doi. org/10.1007/s11192-017-2634-1. Porter, M. E. (1998). The competitive advantage of nations. New York: Free Press, MacMillan. Powell, B. A., Gilleland, D. S., & Pearson, L. C. (2012). Expenditures, efficiency, and effectiveness in U.S. undergraduate higher education: A national benchmark model. The Journal of Higher Education, 83(1), 102–127. Qi, W. (2011). A discussion on the 985 Project from a comparative perspective. Chinese Education and Society, 55(5), 41–56. Raudenbush, S. W., & Bryk, A. S. (2002). Hierarchical linear models: Applications and data analysis methods (2nd ed.). Newbury Park, CA: Sage. Sadlak, J., & Cai, L. N. (Eds.). (2009). The world-class university as part of a new higher education paradigm: From institutional qualities to systemic excellence. Cluj-Napoca: UNESCO European Centre for Higher Education and Shanghai Jiao Tong University, Cluj University Press. Safon, V. (2013). What do global university rankings really measure? The search for the X factor and the X entity. Scientometrics, 97, 223–244. SCImago. (2007). SJR—SCImago Journal & Country Rank. http://www.scimagojr.com. Retrieved 21 July 2015. Seong, S., Popper, S. W., Goldman, C. A., & Evans, D. K. (2008). Brain Korea 21 phase II a new evaluation model. Santa Monica, CA: Rand. Shimmi, Y., & Yonezawa, A. (2015). Japan’s “top global university” project. International Higher Education, 81, 27–28. Shin, J. C. (2009). Building world-class research university: The Brain Korea 21 project. Higher Education, 58(5), 669–688. Sohu News. (2012). http://news.sohu.com/20120109/n331631055_1.shtml. Retrieved 01 Sept 2018. Taylor, P., & Braddock, R. (2007). International university ranking systems and the idea of university excel- lence. Journal of Higher Education Policy and Management, 29(3), 245–260. Taylor, B. J., & Cantwell, B. (2014). Global competition, US research universities, and international doctoral education: Growth and consolidation of an organizational field. Research in Higher Education, 56(5), 411–441. Um, M. (2013).Center of excellence as a tool for capacity building case study: Korea. OECD Programme on Innovation, Higher Education and Research for Development (IHERD) report. https://www.oecd.org. van Leeuwen, T., Moed, H., Tijssen, R., Visser, M., & van Raan, A. (2001). Language biases in the coverage of the science citation index and its consequences for international comparisons of national research performance. Scientometrics, 51, 335–346. van Raan, A., van Leeuwen, T., & Visser, M. (2011). Severe language effect in university rankings: Particularly Germany and France are wronged in citation-based rankings. Scientometrics, 88, 495–498. Volkwein, J. F., & Grunig, S. D. (2005). Reputation ratings in higher education: Double, double, toil and trouble. In J. Burke (Ed.), Achieving Accountability In Higher Education: Balancing Public, Academic, and Market Demands, Chapter 11 (pp. 246–274). San Francisco, CA: Jossey-Bass.

123 Does the Asian catch-up model of world-class universities… von Neumann, J., & Morgenstern, O. (1944). Theory of games and economic behavior. Princeton, NJ: Princeton University Press. Williams, R., de Rassenfosse, G., Jensen, P., & Marginson, S. (2013). The determinants of quality national higher education systems. Journal of Higher Education Policy and Management, 35(6), 599–611. Yang, R., & Welch, A. (2012). A world-class university in China? The case of Tsinghua. Higher Education: The International Journal of Higher Education and Educational Planning, 63(5), 645–666. Ying, C. (2011). A reflection on the effects of the 985 Project. Chinese Education and Society, 44(5), 19–30. Yonezawa, A. (2003). Making “world-class universities”: Japan’s experiment. Higher Education Management and Policy, 15(2), 9–23. Yonezawa, A., & Shimmi, Y. (2016). Transformation of university governance through internationalization. In N. C. Liu, Y. Cheng, & Q. Wang (Eds.), Matching visibility and performance. Global perspectives on higher education. Rotterdam: SensePublishers.

Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

123

Assalamualaikum Warahmatullahi Wabarakatuh dan Salam Sejahtera

YBhg. Datuk/ Datin / Prof/ Tuan/ Puan Bil 6/2021

Comprehensive Internationalization in the Pursuit of ‘World-Class’ Status: A Cross-Case Analysis of Singapore’s Two Flagship Universities Higher Education Policy, 2018 Ó 2018 International Association of Universities 0952-8733/18 www.palgrave.com/journals

Original Article Comprehensive Internationalization in the Pursuit of ‘World-Class’ Status: A Cross-Case Analysis of Singapore’s Two Flagship Universities Justin S. Sanders Graduate School of Human Sciences, Osaka University, Suita, Japan. E-mail: [email protected]

This paper uses John Hudzik’s (2015) Comprehensive Internationalization as a con- ceptual framework to examine how two Singaporean universities, the National University of Singapore (NUS) and Nanyang Technological University (NTU), leverage internationalization as a means to achieving ‘world-class’ status. Document analysis, semi-structured interviews and campus observations are used to provide a cross-case analysis of the universities’ strategies related to partnerships, student and staff mobility, and internationalization of the curriculum. This analysis is placed against the backdrop of wider state policy to explore the relationship between institutional approach and national context. Overall, the findings suggest that a global frame of reference is embedded in the two universities’ visions, and internationalization activities are aggressively pursued to achieve those visions. Factors related to each university’s approach to comprehensive internationalization are explored in the discussion section. Higher Education Policy (2018). https://doi.org/10.1057/s41307-018-0117-5

Keywords: Internationalization; Singapore; Strategic planning; Comparative analysis; World-class Universities

Introduction

Internationalization is now often seen as an imperative for universities (Hudzik and Stohl, 2012; IAU, 2012), especially for those aiming at world-class status (Altbach and Salmi, 2011; Wang et al., 2013). Although the exact definition of a ‘world- class’ remains contested and elusive, characteristics include the ability to attract highly qualified faculty and students, globally impactful research, effective leadership, well-equipped facilities, and international recognition (Altbach and Salmi, 2011). Global rankings often serve as a proxy for world-class status, and despite the many issues with rankings, most highly ranked universities would indeed match the above description. Justin S. Sanders Comprehensive Internationalization in the Pursuit of ‘World-Class’ Status

Proponents of internationalization see it as contributing to the purpose, functions and delivery of higher education (Knight, 2008), by improving an institution’s academic quality, competitiveness and ability to contribute to local, national and global society (Gao, 2015). Rationales and approaches to internationalization vary (Knight, 2004; Stier, 2010; Seeber et al., 2016), but generally it is characterized by an openness to flows in ideas and people, the production of globally relevant knowledge and research, and preparing internationally competent students (Marginson et al., 2011). Although international faculty and students are what get measured in the rankings, they are far from the entirety of internationalization. Because of widely varying circumstances, there is no ‘best’ way to internation- alize, and an adequate contextual analysis is critical to understanding any institution’s approach (Koh, 2011). Case studies, appropriately positioned within their unique context, can provide valuable insights for administrators and policymakers in other settings. Indeed, they are a favored mechanism of many edited volumes related to the internationalization of higher education (IoHE). However, rarely is a deep comparative analysis provided, which may lead to influential contextual factors being overlooked. To address this, this article provides a cross-case analysis of how Singapore’s two flagship comprehensive research universities, The National University of Singapore (NUS) and Nanyang Technological University (NTU), leverage internationalization in their pursuit of ‘world-class’ status.1 Positioning the experience of each university against the other helps to highlight how different internal and external mechanisms influence approach, as well as how a shared policy framework can be interpreted and implemented differently. Singapore stands out as an example of a country that uses internationalization to drive its higher education sector (Altbach et al., 2010), and NUS and NTU are perhaps among the most internationalized universities in the world. Despite some concerns over what some see as an overly commercialized approach, Singaporean higher education has clearly made significant progress in recent decades. Through study of this unique context, there are perhaps important lessons to be gained with regard to investment in higher education and global engagement. Using John Hudzik’s (2015) ‘comprehensive internationalization’ as a concep- tual foundation, this article focuses on the actual practice of internationalization in the cases. This paper assumes that although there are potential pitfalls and negative consequences associated with internationalization (Knight, 2015; IAU, 2012), overall, higher education and thus society can benefit from greater international- ization, and thus, study of different models of practice is worthwhile. The article hopes to offer some practical insight into the approaches the two cases take to integrating international dimensions while embedding themselves as key nodes in the global knowledge network. Additionally, it seeks to add some further examples from the Asia–Pacific region to the existing practice-based resources that largely emerge from and focus on Western contexts.

Higher Education Policy 2018 Justin S. Sanders Comprehensive Internationalization in the Pursuit of ‘World-Class’ Status

Comprehensive Internationalization While internationalization activities are often ad hoc and uncoordinated, it is increasingly common for internationalization to be an integrated aspect of institution-wide strategy and planning. Hudzik (2015) describes this ‘comprehen- sive internationalization’ (CI) as

the means by which higher education institutions respond to widening and more complex expectations to connect globally across all missions to better serve students, clientele, and society in a twenty-first century context. In brief, comprehensive internationalization sees to mainstream access of all institutional clientele to international, global, and comparative content and perspective (ibid, 1). Elsewhere, he defines the term as the

commitment confirmed through action to infuse international and compar- ative perspectives throughout the teaching, research and service missions of higher education. It shapes institutional ethos and values and touches the entire higher education enterprise…[It] not only impacts all of campus life but the institution’s external frames of reference, partnerships and relations (Hudzik, 2011, 6). Importantly, Hudzik (2015) notes, CI attempts to internationalize all aspects of the university’s unique mission and should be tailored to local realities and priorities. This results in differences across HEIs, but he lists a set of common aspirations inherent in CI:

• Mainstreaming internationalization and expanding faculty and student engagement. • Integrating internationalization into core institutional missions. • Expanding who supports and contributes beyond the international office. • Interconnecting activities to produce synergies across the institution.

CI should be addressed on both the macro/strategic and the operational/ programmatic levels, with long-range planning combined with short- and medium-term goals, sequencing, prioritization, allocating resources, and action. Critical is communicating a well-articulated vision, garnering stakeholder buy-in, developing staff and faculty capacity, and identifying and removing barriers. To move from concept to reality, Hudzik assumes strong and effective leadership capable of guiding the institution as well as governance structures that allow for change.

Higher Education Policy 2018 Justin S. Sanders Comprehensive Internationalization in the Pursuit of ‘World-Class’ Status

Managing the interplay of the diverse interests and priorities of a compre- hensive internationalization strategy becomes an organizational challenge, often requiring the establishment of institutional governance mechanisms to guide the process and mediate across interests…Organizational change in the form of increased bureaucracy may be a consequence of attempts to encou- rage and coordinate more complex and larger-scale forms of international- ization (ibid, 39). Such a view may be particularly well suited to the North American context, but with the spread of New Public Management (NPM) it is becoming increasingly applicable globally and is gaining traction in universities within Asia–Pacific. Through interviews with university leaders throughout , China and Singapore, Gao (2015) found that ‘comprehensive’ and ‘holistic’ are frequently used to describe a deliberate, systematic and coherent institutional approach that touches on policies, curricula, collaborations and international perspectives for students and staff. The university leaders felt that internationalization helped to achieve academic excellence by recruiting capable researchers, generating impactful publications, preparing globally competent graduates and gaining a competitive edge and reputation.

Methodology This research draws on a thorough review of the university websites, over 60 institutional documents, 13 semi-structured interviews with administrative and faculty leadership, and on campus observations, to chronicle how the cases attempt to utilize CI to contribute to their institutional visions and missions. Both university-level and college/school-level data were gathered to examine how strategy and practice operated between levels. From the document review, a general framework was established, and key personnel were selected to contribute to a more comprehensive understanding of the rationale behind particular strategies and activities. Data were collected between June and September 2017. Content analysis (Weber, 1990; Creswell, 2009) was preferred over a deep thematic analysis as the target was institutional strategy and approach, rather than the construction of internationalization as a concept. While the universities are talked about in general terms here, it is acknowledged that universities are complex institutions with a wide range of stakeholders with their own priorities, directions and agency, and important perspectives and nuances will unfortunately be lost in such a short space.

Higher Education Policy 2018 Justin S. Sanders Comprehensive Internationalization in the Pursuit of ‘World-Class’ Status

Singapore background Singapore was among the first countries to identify the trend toward IoHE years ago (Alfaro and Ketels, 2016), and it is somewhat of a natural extension of the state’s historically outward-oriented policies (Daquila, 2013). With few natural endowments other than location and human resources, in just over 50 years Singapore has risen to one of the world’s highest GDPs per capita; largely through policies that actively embedded itself deeply in the world system and providing a welcoming environment for human resources and multi-national corporations (MNCs) (Goh and Gopinathan, 2006). Unsurprisingly, education has been a key component of this strategy (Mok and Tan, 2004; Sidhu et al., 2011; Lin and Lin, 2014). Singapore began developing its tertiary sector in earnest in the 1990s (Gopinathan, 1999). In 1991, the government created an international academic advisory panel, made study visits to overseas universities, and began to focus on partnerships, exchange and attracting foreign talent (Mok and Tan, 2004). After the 1997 Asian financial crisis, in a push toward an innovation-driven economy less dependent on MNCs, universities were to drive entrepreneurship, and research and innovation capacity took on new significance (Kam and Gopinathan, 1999; Cheung and Sidhu, 2003; Gopinathan, 2007). The World Class University Program (WCUP) was introduced in 1998 with the aim of recruiting 10 elite foreign universities (ERC, 2002; Tan, 2010). This was expanded to the Global Schoolhouse (GSH) Program in 2002, with the additional aims of tripling the number of international students to 150,000, recruiting talented researchers, encouraging innovation and entrepreneurialism among locals, enhanc- ing the capacity and reputation of national universities, and improving the private sector (Gopinathan and Lee, 2011; Daquila, 2013; Tan, 2016). Owned by the Economic Development Board, GSH was largely seen as a talent development strategy, continuing Singapore’s commitment to build capacity through foreign expertise (Sidhu et al., 2011). In 2006, significantly increased research funding from the newly created National Research Foundation (NRF) and incorporation pushed national univer- sities further toward becoming generators of knowledge and economic capital (Mok, 2010; Gopinathan and Lee, 2011; Alfaro and Ketels, 2016; Sharma, 2017). Andersson and Mayer (2015) credit the increased research funding with the success of the two universities in the global rankings. Although incorporation provided autonomy in many areas meant to drive efficiency, the universities are still very much seen as publicly accountable institutions with important national roles (Ministry of Education, 2007; Ng and Tan, 2010). Although GSH has generally been characterized as a successful policy framework, there have been setbacks. In 2011, the 150,000 international student goal was abandoned amidst public protests, and the government imposed a 10–15%

Higher Education Policy 2018 Justin S. Sanders Comprehensive Internationalization in the Pursuit of ‘World-Class’ Status cap on international undergraduates in public universities. Additionally, the high- profile withdrawals of several foreign universities and trouble retaining top academics have led to some shifts in engagement and recruitment strategies (Rumbley et al., 2012; Daquila, 2013; Ng, 2013; Tan, 2016). Several also question the system’s sustainability, noting the tension resulting from what some see as an overly commercial approach, and worry that sector development may be outpacing its learning maturity (Mok, 2010; Ng, 2013; Reyes and Gopinathan, 2015).

Comparison The following section outlines several key features of the two universities. In general, the two universities were found to have remarkably similar approaches to internationalization, but there were some key differences that are explored in the discussion section.

History and Development Originating in 1905 as a campus of the British King Edward VII College of Medicine, NUS went from a branch of the University of Malaya to the independent University of Singapore before merging with Nanyang University to become the National University of Singapore in 1980. Since, it has grown to a globally recognized comprehensive research university ranked among the top in Asia. As Mukherjee and Wong (2011) and others (Mok and Tan, 2004; Salmi, 2009) have noted, NUS’s rise can be contrasted against its former counterpart, the University of Malaya, which is still largely considered a local institution with somewhat less international recognition. NTU is a clear example of a young university taking advantage of the global knowledge network to quickly achieve global recognition. The Nanyang Techno- logical Institute (founded in 1981) was merged with the National Institute of Education in 1991 to form NTU. Since, it has grown into a research-intensive university with a range of colleges and institutes across disciplines. NTU maintains a heavy emphasis on science and technology and has become the country’s most research-intensive institution, which has been instrumental in attracting capable and productive international partners and faculty. NTU’s rise in the rankings has also been dramatic, climbing more than 200 spots in Shanghai’s Academic Ranking of World Universities (ARWU) between 2010 and 2017.

Vision and Strategy In the late 1990s, in conjunction with shifting national policy, the Harvard trained President began remodeling NUS into a ‘Global Knowledge Enterprise,’ empha- sizing entrepreneurship, global engagement, innovation and the production of

Higher Education Policy 2018 Justin S. Sanders Comprehensive Internationalization in the Pursuit of ‘World-Class’ Status globally competent graduates (Wong et al., 2007; Sidhu et al., 2011; Reyes, 2016). By 2007, the vision had evolved into becoming ‘a leading global university centered in Asia, influencing the future.’ This vision maintains the global aspirations, but with a stronger regional emphasis, positioning itself as a bridge between Asia and the rest of the world. This vision is widely visible throughout the campus and the University’s promotional material. To achieve this vision, NUS’s 2015 Global Strategy rests on three pillars:

• Leadership in global education through distinctive global programs and deep engagement with China, India and Southeast Asia. • Strengthened recognition and branding. • Improving production and communication of impactful research.

This is to be accomplished through:

• Strategic engagements with key partners in the region. • Deep and multifaceted collaborations. • Various types of education-, research-, and enterprise-related initiatives and degree programs. • Multi-institutional coherent and/or cohort programs. • Targeted regional student recruitment. • Differentiated and coherent student mobility programs. • Deeper and more structured academic and internship exposure. • Broadened cultural experience in at least two regions.

Xavier and Alsagoff (2013) concluded that NUS’s internationalization is more concerned with strengthening economic reach and reputation through association than with teaching and learning. While the above are certainly priorities, the document review and interviews found a fairly consistent emphasis on student experience and preparing ‘global-ready’ graduates, a wide variety of student enrichment opportunities, and an overall commitment to teaching. Much of the collateral emphasized student development as a core rationale for other activities.

I believe the most important purpose of our university is to create distinctive and compelling value for our students, for Singapore and the wider global community…In particular, we must constantly ask how well we are helping our students to maximise their potential to be critical thinkers and effective problem solvers, and who also have a strong ‘‘can do spirit’’, who ‘‘can connect’’ well with others, and are continually learning (President Tan Chorh Chuan, State of the University Address, 2015).

Higher Education Policy 2018 Justin S. Sanders Comprehensive Internationalization in the Pursuit of ‘World-Class’ Status

Established in 1991, NTU employed strategic partnerships to build up its faculties and programs from the very beginning. Many of NTU’s first partners came from the vice-chancellor’s personal network. Internationalization continues to be an integral part of the University’s ethos:

‘‘[Internationalization is] embedded in everything we do in NTU. It is part of our DNA’’ (President Bertil Andersson, NTU OIA, 2016, p. 1). NTU’s vision to become ‘a great global university founded on science and technology, nurturing leaders through research and a broad education in diverse disciplines’ also demarks global aspirations, but has a disciplinary research rather than a regional grounding. The document review also revealed a strong emphasis on faculty and research. Overall, there is a strong push toward interdisciplinary research to address global issues as well as priority areas for Singapore and an expressed belief that basic and applied research benefitted from international collaboration. Although absent from the vision, the ‘East meets West’ rhetoric also features prominently in NTU promotional material. NTU’s previous strategic plan for internationalization (2010–2015) emphasized international faculty, broadening the learning experience, international opportuni- ties for students, improved residential experience and compatibility with other universities. For the current (2016–2020) plan, internationalization is embedded rather than a discrete section, but there are priorities set around deeper regional engagement, study abroad rates, international partnerships, and international recruitment. NTU has no published global engagement strategy, but the interna- tional brochure identified the following objectives:

• Continuous expansion of international opportunities for students. • Growth of network of reputable partners. • Emphasis on faculty exchange and research collaboration. • Dedication to diverse and talented faculty. • Focus on attracting top students from around the world. • Global curriculum that facilitates collaborative cross-cultural learning. • Cultivation of global alumni network. • High international visibility.

Management Both universities have fairly centralized corporate-style management structures, allowing senior management significant influence. In this respect, Hudzik’s model may be more applicable here than elsewhere in the region. Management structures for international activities are fairly similar between the two universities, with a vice president-level senior international officer (SIO) overseeing separate offices

Higher Education Policy 2018 Justin S. Sanders Comprehensive Internationalization in the Pursuit of ‘World-Class’ Status

State

Board Partnerships

Faculty and researcher President recruitment

Graduate admissions Facules Student exchange Other Senior Programs administrave Internaonal VP of Research Enterprise unit and units Officer (VP) curriculum Overseas Strategy and Student Research immersion partnerships exchange partnerhsips programs

Support services, HR polcies, facilies, undergraduate admissions, website, etc

Figure 1. Model of the management structure for internationalization. for partnerships and strategy and student exchange. Responsibility for other dimensions of internationalization, such as faculty and student recruitment and research partnerships, are distributed throughout the institutions and their schools and colleges (Figure 1). NUS’s Vice President of University and Global Relations (UGR) overseas UGR staff, the International Relations Office (IRO) and Communication Relations, with the teams working fairly independently. UGR staff, organized by region, are responsible for developing the university’s international strategy, helping to keep leadership informed of important global trends and developing university-wide partnerships. The IRO, established in 1996, manages exchange agreements, provides international opportunities for students, promotes NUS internationally and develops international programs and activities on campus. NTU’s Vice President of International Affairs office has two main units. The Office of International Affairs (OIA), established in 1991, mainly handles university-wide partnerships/networks, strategy and operations, develops regional portfolios, monitors engagements and opportunities, and hosts visitors to campus. The Global Education and Mobility Office (GEM) handles exchange partnerships and the administration of programs for inbound and outbound students. Again, other international activities are generally dispersed throughout other institutional work streams.

Higher Education Policy 2018 Justin S. Sanders Comprehensive Internationalization in the Pursuit of ‘World-Class’ Status

Partnerships Wang et al.(2013) note the important role partnership plays in the development of many world-class universities. They are really at the heart of internationalization for both cases and the foundation for many other international activities. Partnerships are seen as a tool to develop strategic areas of interest, improve research, expand educational offerings, enhance reputation, diversify the under- graduate population, provide opportunities for students, secure funding, attract external expertise and recruit faculty and students. There are essentially five overlapping types of partnerships:

• Student exchange, providing cost-neutral international opportunities for students. • Educational, offering joint or dual degrees and programs, and sometimes entire colleges. • Research, taking advantage of facilities and expertise in other locations. • Industry, supplying funding for research and student internship opportunities. • Associations and networks, allowing for multi-partner engagement.

Both have a history of working with foreign universities to build up their own faculties and programs. As the former NTU President Cham (in Leong, 2002, 2–7) expressed it: ‘We can’t afford to be world leaders in every area. We will be leaders in niche areas and we will achieve this by working with world renowned universities and good industry partners.’ Now, they each boast several hundred university-wide exchange and research partnerships and thousands at the individual faculty level. It is clear that numerical targets have been driving things, but interviewees from both indicated a desire to move toward more qualitative measures and expand existing partnerships rather than pursue new ones. Each university maintains a database of faculty partnerships, which is used in determining whether to pursue broader institution-wide agreements. Interviewees repeatedly stressed the critical role faculty play in both developing and maintaining partnerships, and one NTU interviewee noted the importance of having strategic plans for partnerships at both the university and faculty levels. Interviewees at the faculty level confirmed that their schools or colleges also maintained distinct strategies around partnerships. The universities and their faculties generally look to pursue partnerships when there is something they feel they cannot provide locally, considering fit, location, cost and funding, reputation and expertise, opportunities for students, and language. Partners tend to be highly ranked Western institutions, although the number of regional partners is starting to grow. Typically, partnerships are pursued when they feel they have something to gain, and capacity building of others is less of a priority, although this too is beginning to shift. Occasionally, when the state wants to develop a particular sector, the government will initiate large-scale partnerships,

Higher Education Policy 2018 Justin S. Sanders Comprehensive Internationalization in the Pursuit of ‘World-Class’ Status such as the medical schools with Duke University and Imperial College London or the Yale-NUS College. Deep commitment on both sides is seen as critical in the success of these large-scale initiatives. Interviewees at both universities cited bandwidth as a significant challenge in maintaining such large numbers of partnerships and saw networks as an increasingly effective means to reduce travel time and costs and engage with several partners at once. Networks were also seen as a way to ‘give back’ by sharing their experiences and developing partners with different types of educational experience for their students.

International Talent Singapore’s view of its higher education policy primarily as a talent development strategy is well documented (Ng, 2013; Lee, 2014; Sharma, 2017; Sanders, 2018). International faculty and students are typically brought to enhance institutional capacity and advance the medium- and long-term economic interests of the city- state through research and innovation production and contribution to the workforce. Diversity and introducing varied cultural perspectives are seen as added benefits, but not the main drivers. At the same time, local students and scholars are forced to compete with this foreign talent, which creates significant and perhaps excessive pressure, but is seen as necessary to enhance local capacity and competiveness.

Why do we have international students here? Not just because we need their talent. But because it helps to prepare the Singaporean students for the rest of the world, and the rest of their journey. – NTU Dean.

Students The universities have broadly similar portions of international students. At the undergraduate level, this is due to the government cap, but because of the high rates of exchange there is a larger portion of international undergraduates on campus than the cap would suggest. About half of graduate students are international (Figure 2). NUS’s higher portion of international graduate students may reflect its more established international reputation, but may also be related to the popularity NTU’s engineering programs among career-oriented local students. While international students can be an important revenue generator, they tend to be thought of in broader economic terms and contributions beyond just the university budget. Both NUS and NTU actively recruit international students with attractive marketing material and various incentives. Singapore’s attractiveness as a business and financial center, reputation, financial support and scholarship and grant schemes, English, program options, job prospects, accommodation, partner- ships, facilities, and a supportive policy environment are important recruitment

Higher Education Policy 2018 Justin S. Sanders Comprehensive Internationalization in the Pursuit of ‘World-Class’ Status

40,000

35,000

30,000

25,000

20,000

15,000

10,000

5,000

0 NUS NTU

UGs Intl UGs Grads Intl Grads

Figure 2. Relative portions of domestic and international graduate and undergraduate students. tools, and the universities maximize these to the extent that they are able (Sidhu et al., 2011; Kell and Vogl, 2012; Waring, 2014). Undergraduate admissions criteria for international students are clear and easy to understand, and the universities accept a wide variety of secondary qualifications. Despite many references to attracting ‘the best and brightest from around the world,’ the vast majority of students come from China, India and SEA. While the government does not release official numbers, this is verified through the interviews and observations on campus. Rather than purposeful strategy, this is a result of external factors such as the demand for higher education in the region and the reluctance of North Americans and Europeans to study outside of their regions. However, there are strategies in place to both capitalize on these external realities as well as diversity the student body as much as possible through targeted recruitment activities. To promote local and international student interaction, the universities, in partnership with the MOE, have administrative staff tasked with student integration, purposefully mix residence halls, and have activities and mentoring/buddy schemes to facilitate interaction.

Faculty and Staff NUS is around 60% foreign faculty, while NTU is around 70%. However, so many ‘locals’ have degrees and experience from abroad, that there is not a clear binary. Leadership tends to be primarily local Chinese, although this is slightly less true for NTU (most notably the President). NTU’s higher portion of international faculty is most likely a result of its rapid expansion and the past president’s pro-active international recruiting. Promoting international faculty to leadership positions allows the University to benefit from both their experience elsewhere and their

Higher Education Policy 2018 Justin S. Sanders Comprehensive Internationalization in the Pursuit of ‘World-Class’ Status familiarity working within Singapore, and sends a clear signal to international faculties that they can have successful long-term careers within the institution. English, a welcoming cosmopolitan environment, attractive packages and greater flexibility in hiring procedures introduced since incorporation are important factors enabling the universities to recruit such faculty, and in some cases entire teams:

One of the reasons the NTU has risen so fast in some of the key areas is because we were able to recruit not just significant individuals, but almost entire groups for research or in related research areas. – NTU senior administrator. When you bring in good people, they are already working on world-class level projects…. So we benefit by leveraging the external ecosystem they bring in. So we can become better. So the best attract the best. – NUS senior administrator. However, the lack of competitive local candidates is seen as a concern by both the government and universities, and there are schemes in place to encourage more locals to pursue PhDs. Interviewees noted the challenges in striking the right balance between institutional development and local capacity building and succession planning. Although nationality is not weighted heavily in hiring, all else being equal they prefer to hire Singaporeans. However, this is complicated by the increasing competitiveness of the universities:

In my time, if I’m a Singaporean, there weren’t that many international people applying, so my chances of getting a faculty position with a PhD and some research experience were much higher than now. Now a Singaporean coming back with a PhD from an Ivy League may not get a job. – NUS Leadership.

International Opportunities for Students Both NUS and NTU send an impressive 70% of undergraduates on at least one overseas experience, with over 30% going for a semester or more. Both have targets to raise participation to 80%. To achieve this, both heavily promote study abroad around campus, offer a diverse range of opportunities and have funding programs in place. Programmatic changes, such as the introduction of electives, have also been introduced to make it easier for student to study elsewhere and help facilitate exchange. International experiences are seen to prepare students for careers working with international industries, bring more international perspectives into the classrooms and help develop strong ties with partner universities.

Higher Education Policy 2018 Justin S. Sanders Comprehensive Internationalization in the Pursuit of ‘World-Class’ Status

If you look at what is happening in the 21st century, especially for Singapore as a small city-state, we need to be able to work with many diverse cultur- es…We have to provide this kind of opportunity for our students – NTU senior administrator. Exchange agreements with more than 300 partners at each university are used to facilitate exchange on a cost-neutral basis. The baseline criteria for exchange agreements are use of English (to ensure students get the most out of content while earning credit abroad), accreditation, and the ability to match courses. The most popular study abroad destinations are in Europe and North America, and there are relatively few agreements with universities in Asia, which may be a factor in the strategic objectives to strengthen regional ties. In addition to semester or year-long exchange programs, both universities provide a range of short-term opportunities in diverse locations, including summer/ winter programs, internships, industrial placements, research attachments, volun- teer opportunities and funding for overseas conferences and competitions. The enterprise units of each university also offer programs which embed students in start-up companies in entrepreneurial ‘hot spots’ around the world. This variety of programs allows students to pursue experiences according to their particular interests and circumstances, as well as contributing to the diversity of experiences and perspectives they bring back. Challenges of such high rates of exchange include program and course matching, disruptions to course planning, monitoring and evaluation, and measuring the quality of international experience. Interviewees at both universities noted the desire to and the challenges of shifting to more qualitative indicators. However, neither yet systematically integrates international experiences into programming or teaching and learning nor is it an explicit component in faculty development.

We are trying to encourage students to think about study abroad not as a parenthesis outside of their eight semesters…we are really trying to get them to think about it in an integrated way. – NUS Dean.

So just because we as a university believe that everyone should have this type of experience, the impact and outcome is different for different students. And we’re still trying to find ways to measure that. – NTU senior administrator.

Programming, Curriculum and Environment Internationalization of the Curriculum (IoC), or the ‘incorporation of an interna- tional and intercultural dimension into the content of the curriculum as well as the teaching and learning processes and support services’ (Leask, 2009, 209), may be

Higher Education Policy 2018 Justin S. Sanders Comprehensive Internationalization in the Pursuit of ‘World-Class’ Status the most complex aspect of IoHE. NTU and NUS primarily rely on the large numbers of international faculty and students, programs and curriculum based on foreign models or designed delivered jointly with foreign institutions, the high rates of domestic students with overseas experience, historically small domestic publishing sector and Singapore’s naturally international character. In addition, each university went to lengths to provide an international climate on campus through events, clubs and even dining options. The emphasis on preparing ‘future- ready graduates’ is also seen as inherently tied to preparing students for a globalized world.

So the whole idea is that…education today has to be global in nature. Not just in terms of going and spending a semester or year abroad, but really in terms of your core training, your thinking, your skills, and then of course the contacts that you make, and so on. And beyond that, you need to understand the institutional and political culture of these other countries where you’re going to be doing business. – NUS Professor. However, few supports are in place to ensure that faculty are equipped to explicitly integrate international perspectives and pedagogies into the classroom. Nor is internationalization considered in teaching evaluations. Additionally, although international experiences are considered a signature part of the NUS and NTU experiences, there is no systematic approach to helping faculty incorporate the students’ international experience into their coursework nor is there training for faculty on how to deal with students’ departure for a semester or more. This suggests that IoC may be one of the most difficult aspect of CI to implement, but also raises the question how purposeful it needs to be if the environment itself is sufficiently international.

Discussion The above comparison sought to explore the cases’ approach to comprehensive internationalization in pursuit of world-class status, or in other words, whether they mainstreamed internationalization by embedding it into core institutional missions, expanding participation, and aligning activities across the institution in order to attract highly qualified faculty and students, produce globally impactful research, and gain international recognition. Indeed, the scope and scale of the universities’ international activities are impressive, and in almost all instances these activities were implemented with the express aim of raising their research profiles, delivering outstanding educational programs and opportunities to their students, and enhancing their international reputation. From the cases, it seems clear that excellence within a global frame of reference is an integrated part of the universities’ visions, and the evidence suggests that

Higher Education Policy 2018 Justin S. Sanders Comprehensive Internationalization in the Pursuit of ‘World-Class’ Status strategies such as partnerships, international recruitment and study abroad play an important role in achieving those global visions. The deep and pervasive rhetoric of internationalization was evident throughout university documents, and interviewees were supportive of the endeavor. This does not mean that internationalization was universally supported throughout the institutions, and some interviewees did note some tensions among faculty. Nonetheless, overall the university leadership appears to have succeeded in developing and effectively communicating clear visions and creating enough institutional engagement and buy-into move toward those visions.

Determinants of Approach to CI While there are important differences between the two universities, their strikingly similar approaches to CI signal the importance of national context in the Singapore case. In general, the Singaporean approach to development rests on a foundation of global engagement and leveraging the international to serve the local, consistent with Hudzik’s (2015) assessment that CI does not remove a university from its national context or responsibilities. Singapore’s cosmopolitan environment, use of English, and generous government financing are seen to significantly enhance the universities’ ability to engage internationally. Cultural aspects such as Singapore’s high tolerance for ambiguity (Hofstede Insights, 2018) due to its history as a multicultural trading nation may also help to remove some of the common barriers to CI (Hudzik, 2015). The country’s small size also allows for greater concentration and coordination of resources around a small number of institutions and strategic areas, which may be more difficult in larger national contexts. Age is another important determinant. Developing largely in a ‘post-nationalist’ era (although continuing to serve national interests), both universities have benefited from a ‘late development effect,’ and neither had to deal with the ‘drag of mature experience’ (Hudzik, 2015, 83). Both were able to internationalize as a way to build and develop their research and education profiles through a process of guided evolution rather than drastically reorienting themselves as some universities in other contexts are struggling with. This may be more pronounced for NTU’s case. While neither university is standing still, the much younger NTU has had more ‘room to grow’ over the last 20 years and has depended more heavily on partnerships and foreign faculty to fuel that growth. Strong leadership is an underlying assumption in Hudzik’s model, and this may be especially relevant for Singapore, which is considered to have a relatively high power distance dynamic and respect for hierarchy and leadership (Hofstede Insights, 2018). The combination of this cultural element, the relatively centralized and top-down management structures, and the high levels of international faculty may make it easier for the leadership to pursue an international agenda. Local faculty on the one hand may be less likely to (and have less agency to) resist

Higher Education Policy 2018 Justin S. Sanders Comprehensive Internationalization in the Pursuit of ‘World-Class’ Status directions from leadership, and international faculty may be more supportive of such initiatives as they themselves are international agents. They may also be less inclined toward resistance out of deference and respect for ‘local’ ways of doing things. The impact of international faculty on a university’s ability to internation- alize could be an interesting area for further study. Finally, NTU’s more concentrated focus on science and technology may also play in their approach to internationalization. Given the focus, there was a stronger emphasis on industry partnerships and a stronger prioritization of research. This concentration may also play a role in its ability to quickly improve its international reputation rise in the rankings. As the more comprehensive university, NUS has more offerings in the social sciences, humanities and liberal arts, which benefit from more local expertise and factor less prominently in the rankings.

Vertical and Horizontal Strategic Alignment Creating alignment and synergies between various international strategies and activities is a complex process, and is often not explicit. In managing complex institutions such as universities which have multiple missions and many semi- autonomous units, leadership do not necessarily have the ability to coordinate at all levels. Rather, what is essential is to set the overall direction, build the framework that determines how to advance in that direction, provide the supports and allocate the resources that allow movement in that direction, and create an institutional culture where individual agency aligns with overall direction. The evidence suggests that the semi-autonomous units within the universities do indeed take both national and university-wide direction and policy into account when developing their own strategies, take advantage of institutional and national level supports, and work closely with university leadership on direction setting. This is facilitated by a constant dialogue across the various levels, and the reward and promotion systems also work to foster alignment of faculty leadership over time. Vertically, the state remains heavily involved in university operations by setting overall national direction and priorities and then investing heavily in areas that support those priorities (Mok, 2015). However, rather than overly prescriptive policies that may not translate effectively at the lower levels, the state shows a preference for broad policy direction, allowing for autonomy within a set framework. Allowing the universities more flexibility in hiring practices and tenure systems in order to attract top faculty is one example and encouraging and often funding large-scale partnerships is another. Thus, the universities are better able to pursue internationalization strategies that best suit their capabilities and circumstances. In turn, the evidence suggests that members of the university communities are keenly aware of their role as national institutions, and though not always explicitly, generally take state interests into consideration when planning.

Higher Education Policy 2018 Justin S. Sanders Comprehensive Internationalization in the Pursuit of ‘World-Class’ Status

It doesn’t happen explicitly, but because we are so deeply tied to the state and deeply tied to the interests of Singapore, it’s always in the back of our minds. And it just so happens that it does fit quite well with the university’s mission and vision. – NTU senior administrator. Importantly, the dialogue flows in both directions. As the same NTU administrator notes:

To their credit as well, the state does consult with us, and discuss with us, and take a lot of our feedback into account when they are developing policies and strategies around education, higher education, research, and even internationalization. Horizontal alignment is also more implicit than explicit, although when examining the rationales for most international activities it is clear that they are seen as contributing to multiple mutually supportive ends simultaneously. Partnerships, for example, do not directly contribute to the global ranking metrics, but they are viewed as critical to building institutional reputation, enhancing research capacity, and attracting talented faculty and students and are thus an important strategic element in their drive toward high-ranking status. Thus, internalization is seen more as a web rather than discrete list of isolated activities. A faculty-level exchange agreement, for example, is not seen as an isolated agreement, but rather part of a larger framework of activity that has the potential to contribute to multiple institutional objectives and perhaps grow into something greater in scope. This is evident in the evolving strategies to deepen and expand existing partnerships rather than create new ones. Similarly, international faculty are not recruited simply because they factor into the ranking metrics, rather they are seen as part of an ecosystem and virtuous cycle. In addition to the research they produce, international faculty are seen to keep abreast of developments in their home countries and be familiar with practices abroad, which are helpful in forming partnerships and informing practice at the host institutions. They also help to create a more international environment on campus and introduce more varied perspectives into the curriculum, which in turn is seen as beneficial for local students. The more internationalized environment is also considered an important recruitment tool for other international faculty and students. An awareness of such benefits is an important element in the decision by national and university leadership to attempt to raise the level of local talent to global standards rather than forgo these potential benefits by limiting the numbers of foreign faculty or introducing quotas similar measures. While they wish to hire locally as much as possible, the overarching policy is to get the best people available.

Higher Education Policy 2018 Justin S. Sanders Comprehensive Internationalization in the Pursuit of ‘World-Class’ Status

Adaptive Strategy Finally, there is a need to acknowledge the relationship between position and strategy. Both domestic and international contexts and priorities are constantly shifting, and an effective internationalization strategy needs to evolve accordingly. What may have been effective strategy throughout the 1990s is not likely to be as applicable today. For example, key to both NUS and NTU’s relative success has been the effective partnering with prestigious Western universities. However, it seems both are reaching a saturation point and are now targeting deeper regional engagement. Asia, and in particular China, has historically served primarily as the source for international students, but in light of shifting global dynamics and seeing the potential for further growth in other areas, they are setting up increasing numbers of joint programs and facilities with institutions in China and elsewhere. Similarly, after the government introduced the cap on international undergraduates in 2011, the universities found themselves operating under a new paradigm and turned to increased levels of exchange and recruitment of international graduate students as means to maintaining the benefits of a diverse student body and to fulfilling their talent attraction roles for the country. Thus, an effective strategy can help get to where you want to go, but once you get there, a new strategy will be needed to reach the next stage. For universities with majority international faculty and graduate students, hundreds of international partners, and the vast majority of undergraduates studying abroad, ‘where do we go from here?’ is a natural question to ask, and helps explain why the current strategies are contain terms such as ‘deep,’ ‘differentiated’ and ‘multifaceted.’

Any good institution today is going to be a globalized institution. It’s a question of getting the balance. And that’s where we are now struggling. Where to get that balance. – NUS Professor.

Concluding Remarks By presenting these two cases of comprehensive internationalization in the Singaporean context, this article has hoped to highlight some of mechanisms by which internationalization strategies can be used toward broader institutional visions and goals. These cases help to demonstrate the powerful transformative potential of partnerships and may encourage a more interconnected global higher education network where universities around the world leverage one another’s strengths. At the same time, they illustrate the potential benefits as well as potential tensions of a heavily international faculty body. However, they also drive home the importance of adequate funding and resourcing and demonstrate the need for effective strategy to consider and navigate shifting contexts both at home and afield.

Higher Education Policy 2018 Justin S. Sanders Comprehensive Internationalization in the Pursuit of ‘World-Class’ Status

This is not to say that either NUS or NTU have reached the end of their internationalization journey. Nor should they necessarily serve as architypes. There remains a need to shift from quantitative to qualitative measures of success, but some uncertainty of how to go about that. As Hudzik (2015) notes, ‘without being able to establish some level of evidence about cause and effect there is no objective way to establish the outcomes, results, or impacts of international programming’ (ibid, 106). Additionally, both have room to improve in supporting faculty in dealing with the challenges of such high rates of exchange and large numbers of international students. More could also be done to integrate students’ international experiences into programming and curricula. There are also a number of important philosophical questions about their approach to CI, which include asymmetric partnerships and a tendency to favor prestigious western institutions over those from within the region, the siphoning of human resources from the region, an emphasis on competition over collaboration, too much attention paid to ranking indicators, and perhaps favoring the international over the local. The long-term implications of such practices should be thoroughly considered by both national and institutional policy makers. Nonetheless, over the last twenty or so years, both universities have managed to significantly increase their research performance and output, vastly expand their program offerings to students, make important economic and social contributions to the nation, and firmly embed themselves in the global knowledge network. Importantly, they have also positioned themselves as key nodes connecting people and ideas from the Eastern and Western arenas and thus serve an important bridging function. While the cases should not be seen as blueprints for other universities, the hope is that through providing such windows into the actual practice of CI, they may be instructive in how others might leverage their unique context to more effectively internationalize to support their institutional missions.

Acknowledgements The author would like to express sincere thanks to all of the participants who were very generous with their time. Without them, this research would not have been possible.

Notes

1 In January 2018, both universities brought on new presidents who quickly began implanting changes in strategy and structure. This research represents the situation at the universities through October 2017.

Higher Education Policy 2018 Justin S. Sanders Comprehensive Internationalization in the Pursuit of ‘World-Class’ Status

References

Alfaro, L. and Ketels, C. (2016) The ‘‘Global Scholhouse’’: Singapore’s Higher Education Aspiration, Cambridge, MA: Harvard Business School. Altbach, P.G., Reisberg, L. and Rumbley, L. (2010) Trends in Global Higher Education: Tracking an Academic Revolution, Paris: UNESCO. Altbach, P.G. and Salmi, J. (2011) The Road to Academic Excellence: The Making of World Class Universities, Washington, DC: The World Bank. Andersson, B. and Mayer, T. (2015) ‘Internationalization of Higher Education – The Nanyang Technological University Story: Perspectives and Experience from the Lion City’. In J. Hudzik (ed.) Comprehensive Internationalization: Institutional Pathways to Success, New York: Routledge, pp. 175–182. Cheung, W. and Sidhu, R. (2003) ‘A tale of two cities: Education responds to globalisation in Hong Kong and Singapore in the aftermath of the Asian economic crisis’, Asia Pacific Journal of Education 23(1): 43–68. Creswell, J.W. (2009) Research Design: Qualitative, Quantitative, and Mixed Methods Approaches, Thousand Oaks, CA: Sage Publications. Daquila, T.C. (2013) ‘Internationalizing higher education in Singapore’, Journal of Studies in International Education 17(5): 629–647 ERC (Economic Review Committee) (2002) Developing Singapore’s Education Industry, Ministry of Trade and Industry, Republic of Singapore. Gao, Y. (2015) ‘Constructing internationalisation in flagship universities from the policy-maker’s perspective’, Higher Education 70(3): 359–373. Goh, C. and Gopinathan, S. (2006) ‘Education in Singapore: Development since 1965’, in B. Fredrickson and J. Tan (eds.) An African Exploration of the East Asian Education, Washington, DC: World Bank, pp. 80–108. Gopinathan, S. (1999) ‘Preparing for the next rung: Economic restructuring and educational reform in Singapore’, Journal of Education and Work 12(3): 295–308. Gopinathan, S. (2007) ‘Globalisation, the Singapore developmental state and education policy: A thesis revisited’ Globalisation, Societies and Education 5(1): 53–70. Gopinathan, S. and Lee, M.H. (2011) ‘Challenging and co-opting globalisation: Singapore’s strategies in higher education’, Journal of Higher Education Policy and Management 33(3): 287–299. Hofstede Insights (2018) Country Comparison: Singapore. https://www.hofstede-insights.com/country- comparison/singapore/, Accessed 24 July 2018. Hudzik, J.K. (2011) Comprehensive Internationalization: From Concept to Action, Washington, DC: NAFSA. Hudzik, J.K. (2015) Comprehensive Internationalization: Institutional Pathways to Success, New York: Routledge. Hudzik, J.K. and Stohl, M. (2012) ‘Comprehensive and Strategic Internationalization of U.S. Higher Education’, in D. Deadroff, H. de Wit, J. Heyl, and T. Adams (eds.) The SAGE Handbook of International Higher Education, Thousand Oaks: Sage, pp. 61–80. IAU (2012) Affirming Academic Values in Internationalization of Higher Education: A Call to Action, Paris: International Association of Universities. Kam, H.W. and Gopinathan, S. (1999) ‘Recent developments in education in Singapore’, School Effectiveness and School Improvement 10(1): 99–117. Kell, P.M. and Vogl, G. (2012) International Students in the Asia Pacific, Dordrecht: Springer Netherlands. Knight, J. (2004) ‘Internationalization remodeled: Definition, approaches, and rationales’, Journal of Studies in International Education 8(1): 5–31.

Higher Education Policy 2018 Justin S. Sanders Comprehensive Internationalization in the Pursuit of ‘World-Class’ Status

Knight, J. (2008) Higher education in turmoil: The changing world of internationalization, Rotterdam: Sense Publishers. Knight, J. (2015) ‘Internationalization brings important benefits as well as risks’, International Higher Education 46(1): 8–10. Koh, A. (2011) ‘Singapore’s ‘‘global assemblage’’: Digging into the culture of education policy making’, Critical Studies in Education 52(3): 267–278. Leask, B. (2009) ‘Using formal and informal curricula to improve interactions between home and international students’, Journal of Studies in International Education 13(2): 205–221. Lee, J.T. (2014) ‘Education hubs and talent development: Policymaking and implementation challenges’, Higher Education 68(6): 807–823. Leong, C. (2002) The NTU Story Part II. Coming of Age, Singapore: Nanyang Technical University. Lin, C.C. and Lin, G. (2014) ‘Analysing innovation policy dimensions and contexts: In the empirical cases of Taiwan and Singapore’, International Journal of Organizational Innovation 6(4): 203–216. Marginson, S., Kaur, S. and Sawir, E. (2011) Higher education in the Asia-Pacific: Strategic Responses to Globalization, Rotterdam: Springer. Ministry of Education (2007) International Academic Advisory Panel Sixth Meeting, Singapore: Singapore Government Press Release. Mok, K.H. (2010) ‘When state centralism meets neo-liberalism: Managing university governance change in Singapore and Malaysia’, Higher Education 60(4): 419–440. Mok, K.H. (2015) ‘The quest for global competitiveness: Promotion of innovation and entrepreneurial universities in Singapore’, Higher Education Policy 28(1): 91–106. Mok, K.H. and Tan, J. (2004) Globalization and Marketization in Education: A Comparative Study of Hong Kong and Singapore, Camberley, UK: Edward Elgar. Mukherjee, H. and Wong, P.K. (2011) ‘The National University of Singapore and the University of Malaya: Common roots and different paths’, in P.G. Altbach and J. Salmi (eds.) The Road to Academic Excellence: The Making of World-Class Universities, Washington, DC: The World Bank, pp. 129–166. Ng, P.T. (2013) ‘The global war for talent: responses and challenges in the Singapore higher education system’, Journal of Higher Education Policy and Management 35(3): 280–292. Ng, P. and Tan, C. (2010) ‘The Singapore Global Schoolhouse’, International Journal of Education Management 24(3):178–188. Reyes Jr., V.C. and Gopinathan, S. (2015) ‘A critique of knowledge-based economies: A case study of Singapore education stakeholders’, International Journal of Educational Reform 24(4): 136–159. Reyes, C.N. (2016) ‘Framing the entrepreneurial university: The case of the National University of Singapore,’ Journal of Entrepreneurship in Emerging Economies 8(2): 134–161. Rumbley, L., Altbach, P. and Reisberg, L. (2012) ‘Internationalization within the higher education context’, in D. Deadroff, H. de Wit, J. Heyl, and T. Adams (eds.) The Sage Handbook of International Higher Education, Thousand Oaks: Sage, pp. 3–26. Sanders, J. (2018) ‘National internationalisation of higher education policy in Singapore and Japan: Context and competition’, Compare: A Journal of Comparative and International Education. Published online 8 Jan 2018. https://doi.org/10.1080/03057925.2017.1417025. Salmi, J. (2009) The Challenge of Establishing World Class Universities, Washington, DC: The World Bank. Seeber, M., Cattaneo, M., Huisman, J., and Paleari, S. (2016) ‘Why do higher education institutions internationalize? An investigation of the multilevel determinants of internationalization rationales’, Higher Education 72(5): 685–702. Sharma, Y. (2017) ‘The story of how Singapore became a research nation’, University World News, 15 December. Sidhu, R., Ho, K.C. and Yeoh, B. (2011) ‘Emerging education hubs: The case of Singapore’, Higher Education 61(1): 23–40.

Higher Education Policy 2018 Justin S. Sanders Comprehensive Internationalization in the Pursuit of ‘World-Class’ Status

Stier, J. (2010) ‘International education: Trends, ideologies and alternative pedagogical approaches’, Globalisation, Societies and Education 8(3): 339–349. Tan, C. (2010) ‘Educational policy trajectories in an era of globalization: Singapore and Cambodia’, PROSPECTS 40(4): 465–480. Tan, J. (2016) ‘Singapore’s ‘‘Global Schoolhouse’’ aspirations’, International Higher Education 87: 9–10. Wang, Q., Cheng, Y. and Cia Liu, N. (2013). Building World-Class Universities: Different Approaches to a Shared Goal, Rotterdam: Sense Publishers. Waring, P. (2014) ‘Singapore’s global schoolhouse strategy: Retreat or recalibration?’, Studies in Higher Education 39(5): 874–884. Weber, R.P. (1990) Basic Content Analysis, Thousand Oaks: Sage Publications. Wong, P.K., Ho, Y.P. and Singh, A. (2007) ‘Towards an entrepreneurial university model to support knowledge-based economic development: The case of the National University of Singapore’, World Development 35(6): 941–958. Xavier, C.A. and Alsagoff, L. (2013) ‘Constructing ‘‘world-class’’ as ‘‘global’’: A case study of the National University of Singapore’, Educational Research for Policy and Practice, 12(3): 225–238.

Higher Education Policy 2018

Assalamualaikum Warahmatullahi Wabarakatuh dan Salam Sejahtera

YBhg. Datuk/ Datin / Prof/ Tuan/ Puan Bil 6/2021

Engineering a World Class University? The Impact of Taiwan’s World Class University Project on Scientific Productivity Higher Education Policy, 2018 Ó 2018 International Association of Universities 0952-8733/18 www.palgrave.com/journals

Original Article Engineering a World Class University? The Impact of Taiwan’s World Class University Project on Scientific Productivity Yuan-Chih Fua,c , David P. Bakera and Liang Zhangb aThe Pennsylvania State University, State College, PA, USA. E-mail: [email protected] bNew York University, New York City, NY, USA. cPresent Address: Center for Innovative Research on Aging Society (CIRAS) and Graduate Institute of Education, National Chung Cheng University, Chiayi County, Taiwan.

High profile, university excellence initiatives are becoming increasingly popular national R&D policies globally, with 23 nations implementing some form of an excellence ini- tiative from 1995 to 2013. While differing in tactics, their common strategy is to fund accumulated advantage in science productivity, intentionally stimulating a world-class university among selected best universities to lift their capacity to compete internationally with the world’s most research-intensive universities. Here a ten-year difference-in-dif- ferences statistical evaluation of rates of scientific publications of Taiwan’s much-pub- licized, university excellence initiative ‘‘World Class University Project’’ yields findings unanticipated by the policy design. Significant financial resources provided to the selected and designated ‘‘World Class Universities’’ (WCUs, treatment group) did increase their volume and rate of publication of scientific papers overall and in more prestigious jour- nals, but not at the policy’s anticipated accelerated rate. Also, interestingly the non- selected, less research-intensive universities (control group) also increased their publi- cation rate in parallel and sometimes more than the WCUs over the same period. A possible diffusion effect of institutional norms, greater inter-institutional competition, and isomorphism around science productivity generated by the implementation of the policy is examined and discussed in light of the global spread of university excellence initiatives. Higher Education Policy (2018). https://doi.org/10.1057/s41307-018-0110-z

Keywords: university excellence initiatives; research university; scientific production

Introduction Experimenting with new approaches to higher education policy aimed at improving the knowledge production capabilities of a nation’s universities is an increasingly popular R&D strategy. Bound up with international competition in science and participation in the knowledge economy, these ‘‘university excellence initiatives’’ represent a departure from traditional approaches to regulate and fund public Yuan-Chih Fu et al. Engineering a World Class University? institutions (Salmi, 2009). As displayed in Figure 1, over the nearly two decades from 1995 to 2013 twenty-three countries implemented some form of an excellence initiative with the intent of transforming selected institutions into internationally recognized research-intensive universities. Such initiatives are concentrated among Asian and European nations, where dramatic recent increases in the growth rate of global scientific production have occurred (Zhang et al., 2015). The use of these initiatives is driven not only by a growing volume of scientific knowledge and its assumed value to national economies, but also because of greater proliferation of public rankings of research productivity of universities and the ensuing fascination with the idea of a ‘‘World Class University’’ (Altbach and Balan, 2007; Deem et al., 2008). Although the world-class university is an imprecise and disputed notion, most agree that a high-quality research university requires at least three complementary assets: (1) a high concentration of talented scientists, (2) abundant resources to offer a rich environment for conducting advanced research, and (3) favorable governance enabling institutions to manage research resources in a flexible and effective fashion (e.g., Geiger, 2004; Salmi, 2009). In light of the well-established association between funding and productivity, and given that of first two components of the successful research university rely on the availability of R&D funds, most new initiatives experiment with awarding supplemental funds for approaches ranging from forming research clusters to improving graduate schools to investing in infrastructure (Wood, 1990). Regardless of specific tactics, the overarching strategy is to select the most promising

Figure 1. Global Diffusion of University Excellence Initiatives, 1995–2013. Note: Authors’ compilation of timing of R&D programs.

Higher Education Policy 2018 Yuan-Chih Fu et al. Engineering a World Class University? institutions and then provide them with significant supplemental R&D funds. The goal is to intentionally apply new funds to past accumulated advantage in science productivity among a selected few of a nation’s best research-intensive universities to assist them in achieving global competitiveness on par with the highest producing universities across the world. The strategy is to engineer and sustain world-class universities, where the highest reputation, often already most resourced, institutions are supplied with additional significant resources in order for their scientists to achieve an even higher level of output. Following this logic, Figure 2 illustrates the hypothetical desired divergence in growth of the science productivity, measured by journal paper publications, of selected and funded universities versus other eligible, but non-selected, institutions. Over the course of the initiative, two outcomes are expected to occur: first, the selected universities will continue to publish at a higher rate than the non-selected; and second and essentially, the selected universities’ rate of production will significantly increase and diverge away from the non-selected universities. What might occur among the non-selected universities after the university excellence initiatives are implemented in the higher education system is usually not considered. Whether China’s 985 project, South Korea’s BK 21, Japan’s 21st COE, France’s Operation Campus, or, as examined here, Taiwan’s World Class University Project (hereafter WCUP), the assumption behind the selection-and-supplemental funding strategy is the same: reallocation of R&D funds to invest in a nation’s best to transform them into a world-class research university. Detractors of the logic

Figure 2. Hypothetical desired effect on scientific journal publications by scientists at selected and non-selected universities in a university excellence initiative program.

Higher Education Policy 2018 Yuan-Chih Fu et al. Engineering a World Class University? behind the excellence initiative strategy point out the inherent inequalities of spending limited funds on already better funded universities. Realistic or not, fair or not, the popularity of these programs makes then a likely feature of the world’s R&D policy environment for the foreseeable future. Although a popular option among nations’ science policy portfolios, to date only four evaluations of long-running university excellence initiatives programs have been published, and the results are mixed. An assessment of France’s Operation Campus found that the program failed to improve funded institutions’ standings in world rankings (Lacroix and Maheu, 2015). The other two programs’ evaluations focused on publication rates and found a more complex pattern of outcomes, where the first intended outcome occurred but not necessarily the second. Because programs selected and funded are already research-intensive universities, it is not surprising that over the course of Korea’s BK21 and China’s 985 these institutions continued to have high publication rates (Shin, 2009; Zhang et al., 2013). But, in the case of China 985 this trend was not led, as expected, by the nation’s initially most research-intensive institutions, but rather by initially less successful ones that also receiving supplementary funding. During Korea’s BK21 program there was no divergence as the publication rate among unselected research-intensive universities grew in parallel with funded ones (Seong et al., 2008). And in the same vein, an earlier study of the first year of the WCUP found that selected universities receiving the least total funds performed better than most of the higher funded ones (Chang et al., 2009). Therefore, the purposeful creation of an accumulated advantage among a few universities for the intense global competition is more challenging than proponents of excellence initiatives originally envisioned. At the same time, these types of programs are likely not necessarily failures and may have unexpected beneficial R&D impacts on whole systems of universities. An evaluation from 2005 to 2010 of Taiwan’s much-publicized WCUP aimed at the same two policy expected-effect goals charted in Figure 2, offers new and more detailed evidence about the effectiveness of excellence initiatives.

World Class University Project Taiwan’s WCUP is a quintessential university excellence initiative program. In 2005, under the auspice of Taiwan’s Ministry of Education (MOE, 2005), 32 out of the nation’s 145 universities applied to a program offering significant funding to increase science production. Based primarily on their past level of scientific publications, a small number of universities were selected, and with little apparent irony were immediately publically designated by the MOE as ‘‘World Class Universities’’ (hereafter WCUs). Over the ensuing five years, the WCUs were awarded approximately a total of 1.5 billion US dollars in extra funding, an

Higher Education Policy 2018 Yuan-Chih Fu et al. Engineering a World Class University? augmentation ranging from 11 to 25% of these universities’ annual R&D expenditures. Developed by the MOE and like China’s 985, WCUP funding criteria gave considerable weight to past research performance and future potential of the whole university (Zhang et al., 2013), while Japan’s 21st COE (Shima, 2017) and South Korea’s BK21 (Seong et al., 2008) only included criteria about special research centers and doctoral programs. An early objective was to invest extensive resources on a small number of the best universities to place at least one of them among the world’s top 100 within ten years, and then in top 50 a decade after that (MOE, 2005). While the WCUP can also be evaluated on teaching and mentoring criteria, given the program’s main focus on evaluating scientific research, an evaluation of numbers of publications and impact level of journals are the most central outcomes (Hazelkorn, 2015). In addition to significant supplemental funding, the MOE also gave the presidents of selected WCUs financial autonomy and flexibility in developing tactics to increase research. Across the WCUs, various approaches were employed such as recruitment of talented STEM plus health (hereafter STEM +) faculty, the formation of university-based research centers, and other collaborative arrange- ments. Like most university excellence initiatives, the overarching goal was to lift the nation’s best universities into international prominence by providing them with substantial new funding to yield an acceleration of their science publication rate. Although similar in its design and goal to other national university excellence initiatives, the Taiwanese context influenced the motivation and impact of the WCUP in unique ways. Before the policy’s development, even the most research- intensive universities obtained research funding through individual faculty grants, and it was rare for universities to organize faculty to compete strategically for external research funding. Just prior to this period, however, two driving forces reshaped the landscape of higher education. First, as the nation achieved universal tertiary enrollments, the growing educated middle-class was keen on having a set of superior universities within the higher education system offering the possibility of differentiating their children’s educational attainment. Stoked by the popularity of the World Class University ranking system and Taiwan’s unique geo-political position, the second was an intense interest in using inter-university competition among the nation’s universities to achieve competitiveness with Chinese univer- sities (Tai, 2006; Song and Tai, 2007). For the first time in the nation’s development of higher education, the combination of these factors provided motivation and legitimation for the national government to create new policy to generate inter-university competition as a strategy for the allocation of state appropriations and to bolster the R&D capacity of a few universities. The WCUP transformed the R&D environment of the WCUs as their academic leaders and administrators were charged with guiding their departments to organize among other tactics interdisciplinary research groups. The new funds facilitated the

Higher Education Policy 2018 Yuan-Chih Fu et al. Engineering a World Class University? formation of research clusters and their competitive funding within institutions. Also, and important to interpreting the program’s outcome, being designated a WCU yielded considerable prestige and public acknowledgement as Taiwan’s answer to the global trend toward world-class universities and their assumed heightened scientific discovery. Information about universities’ applications and selection outcomes, made available by the MOE, enables a methodologically rigorous evaluation of the aforementioned two-part goals of the WCUP by comparing pre- and post-initiative scientific publication rates for selected WCUs with initially WCUP eligible, but judged less promising research-intensive and hence non-selected universities (hereafter non-WCUs). The WCUP was not a randomized experimental design, but even with what amounts to a policy of intentional selection-effect bias (on past performance) this comparison affords a reasonable assessment of its impact. Outcome measures used in previous evaluations have been criticized for overreliance on total publications without any differentiation by quality. Beyond comparing rates of all scientific publications, here rates by the quality (impact score) level of scientific impact of journals are included in the evaluation. After examining the impact of the WCUP, the new findings, along with those from other similar initiatives, are used to inform a broader discussion of R&D policy aimed at engineering an accumulated advantage among research-intensive universities.

Analytic Method Design, data, and variables The evaluation of the WCUP program is accomplished through a difference-in- differences (DD) estimation of a treatment effect. Thirty-two of Taiwan’s universities applied to the initiative in 2005, and out of these, 10 of the most promising were judged and designated as WCUs, awarded the significant supplementary research funding, and form the treatment group.1 Out of the remaining 21 non-WCUs, the 14 initially most research intensive, measured by journal publications, are employed here as the control group. Since application to the WCUP was open to all universities in the country regardless of their research capacity, excluding the least research-intensive from the control group provides the most rigorous test of the program’s outcomes. Also during the same period the MOE administered several institutional-based funding programs in addition to the WCUP that modestly increased state appropriation to unselected universities, including the non-WCUs in the control group. These additional funding resources could cause some contamination of the comparison group, and therefore we also control for total research funding for both WCU and non-WCU in estimating the equation. This scenario is hard to avoid,

Higher Education Policy 2018 Yuan-Chih Fu et al. Engineering a World Class University? particularly in the area of public policy, and while the impact is probably minor, possible influence on the results is considered in the discussion (Seong et al., 2008). The policy treatment is the addition of WCUP funding to the WCUs that began in 2006 (WCUP funds in millions NTD used in R&D, Mean = 249; S.D. = 179). Figure 3 shows the levels and growth in R&D funds for both groups of universities, including from WCUP (policy treatment) among the WCUs. Policy treatment effects were measured on STEM + publications extracted from the Web of Science annually from 2001 to 2010. Six dependent variables were created from these data: the total number of STEM + publications from each university, and the number of publications in quintiles of journals distributed along a standard impact factor, with the 1st Impact Tier being publications of journals with the impact factor being ranked at the top 20% of its assigned academic field. DD equations are conditioned on the size of a university’s faculty and total R&D funds. A Cobb–Douglas production function, as often used in research on universities, is conceptually defined here as the output based on the combination of the human capital and R&D resources, where the former is operationalized as the size of the faculty and the later as the R&D funding (Cobb and Douglas, 1928; Shin, 2009; Zhang et al., 2013). These data were collated from The Digest of Education Statistics published by the MOE (2016), the National Science and

Figure 3. Total STEM + R&D Funding Implications of Taiwan’s WCUP University Excellence Initiative Awards beginning in 2005. Note 1: R&D funds adjusted in 2011 constant dollars. Note 2: Total indicates the R&D funds from all resources, while Gov. indicates the R&D funds from governmental agencies including WCUP. Source: MOST (2013). The National Science and Technology Activities Survey.

Higher Education Policy 2018 Yuan-Chih Fu et al. Engineering a World Class University?

Technology Activities Survey conducted by the Taiwan Ministry of Science and Technology (hereafter MOST) (2013). Means and standard deviations for all variables are shown in Table 1. In line with the goal of creating upward divergence rate of publications, the ten WCUs were best positioned to accelerate their research output from the program’s supplementary funding. Baseline research resources of R&D funding and faculty of the WCUs were substantially greater than the non-WCUs, and the number of STEM + publications were significantly larger in total and across all journal impact tiers for WCU Universities.

Analytical model Analysis of STEM + publications yields DD’s estimates between WCUs and non- WCUs. In particular, the model estimates whether the growth in publications from the baseline period to the period of supplemental funding is greater among WCUs than non-WCUs. As is common in studies of research production, dependent variables were lagged by 2 years to account for the time lag between research funding and publications (Seong et al., 2008; Zhang et al., 2013). Formally, the

Table 1 Means (standard deviations) for variables in analysis. Source: Own calculations based on WOS, MOST, and MOE

Periods Baseline (2001–05) Follow-Up (2006–10)

Group WCU Non-WCU T WCU Non-WCU t

STEM + publications Total publications 906 226 8.7 1530 438 9.1 (641) (105) (961) (247) 1st impact tier 359 63 8.4 667 151 8.2 (290) (33) (505) (116) 2nd impact tier 205 55 8.2 371 113 9.0 (149) (30) (228) (60) 3rd impact tier 125 39 8.8 213 71 9.7 (79) (17) (113) (38) tier 94 33 8.4 140 53 7.8 (57) (17) (86) (28) 5th impact tier 62 21 7.0 73 31 7.8 (47) (10) (38) (19) Control variables Faculty 489 239 6.5 547 272 6.5 (304) (90) (329) (106) R&D funds 1728 318 7.6 2407 420 7.6 (1544) (137) (2172) (193)

R&D funds adjusted in 2011 constant dollars.

Higher Education Policy 2018 Yuan-Chih Fu et al. Engineering a World Class University? proposed estimation equation and variables with their definitions used for the following data analyses are as following: 0 yit ¼ a0 þ bðÞþWCUi à Post di þ ht þ Zitn þ lit where yit is a log number of publication for institution i in year t; WCUi is a binary variable indicating whether an institution is part of the WCU group; Post is a dummy variable indicating the period of 2006-2010; di is a set of institutional fixed effects, while ht is a set of year fixed effects; Zit contains a set of covariates for institution i in year t. The covariates include: lnðR&DitÞ: The R&D funds expended by school i in year t. The variable is the natural log of university’s R&D funds in STEM + fields each year. lnðFacultyitÞ: The full-time faculty registered in school i in year t. The variable is the natural log of university’s faculty size in STEM + fields each year.

Results Given the design of the WCUP, it is expected that the number of STEM + pub- lications from the already heavily resourced WCUs would continue after the supplemental funding treatment, and this is illustrated in the first row of Table 1. The mean total publications after the start of the funding 2006 to 2010 and publications in each journal impact tier from WCUs are substantially higher than non-WCUs. After receiving the supplementary funding, the volume of publications from WCUs is almost three and one half greater than non-WCUs. The WCUP achieved its first goal, but for the policy’s main goal to be met there should be evidence of an increasing divergence in the growth rate of the WCUs. The mean log growth of all STEM + journal publications for both WCU and non-WCUs is displayed in Figure 4. The solid line displays the difference between these two groups of universities publications, which decreased from 1.31 in 2005 to 1.26 in 2006. During the remaining years, except for a slight rebound observed in 2008 (logged value = 1.24), a declining, not increasing, difference between the two groups is evident; instead of WCUs accelerating their publication more than the non-WCUs, this control group’s rate accelerated equally fast, and in some years faster. Figure 5 shows the logged publication trajectories of WCUs and non-WCUs across the quintiles of journal citation impact scores. The same pattern holds across all levels of journal quality. For example, while WCU scientists published more papers in journals with the highest impact scores after the WCUP funding, the post- treatment publication rate among non-WCUs was larger. It is possible that universities with small numbers of publications in the initial stage may enjoy higher growth rates than universities with large numbers of publications, resulting in a negative relationship between initial stage and growth rate regardless of policy change. In other words, the reduction in the relative gap in

Higher Education Policy 2018 Yuan-Chih Fu et al. Engineering a World Class University?

Figure 4. Comparison of all STEM + publications (logged) by scientists at WCU’s and non-WCU’s. publications could be an artifact of the logarithm transformation of the number of publications. Time trends in Figures 4 and 5, however, do not seem to support this scenario. Figure 4 indicates that the relative gap between WCUs and non-WCUs has slightly increased between 2003 and 2005, and then changed its trend right after 2005. The first panel in Figure 5 clearly demonstrates that the decrease in the relative gap only occurred after 2005. As displayed in Table 2, the estimated DD equation of the treatment effect (WCUP*Post), conditioned on faculty size and R&D total expenditures, does not change the bivariate results. The statistically insignificant treatment effect coefficients indicate that there is mostly no difference in the growth in the publication rate between the treatment and control groups. However, as for the rates of publication in the highest impact journals, the statistically significant negative coefficient indicates that the control group non-WCUs grew faster by 16.5% compared to their peers in the treatment group WCUs during the period of policy intervention. In accordance with the design of DD equation, the observed ongoing convergence among two groups could be directly attributed to the implementation of WCUP. The result derived from the DD estimate confirms that on average the growth rate of publication in the highest impact journals in the non-WCUs increased faster than the WCUs. Turning to the effect of the policy on publishing in the highest impact journal, Figure 6 indicates the national share of articles in the highest impact journals from

Higher Education Policy 2018 Yuan-Chih Fu et al. Engineering a World Class University?

Figure 5. Comparison of scientific publications (logged) by scientists at WCU’s and non-WCU’s by journals’ mean citation impact, 2001–2010. (1st impact tier is journals with highest impact score).

Table 2 Difference-in-difference estimates of growth rates of STEM + publications between WCU’s and non-WCU’s, total publications and by publications in citation impact score

DV(logged) (1) (2) (3) (4) (5) (6) Total 1st Tier 2nd Tier 3rd Tier 4th Tier 5th Tier

WCUP*Post - 0.084 - 0.165* - 0.097 - 0.040 - 0.115 - 0.168 (- 1.27) (- 1.82) (- 0.99) (- 0.65) (- 1.10) (- 1.31) ln(funding) 0.244* 0.280 0.0617 0.231** 0.433 0.791** (2.12) (1.36) (0.48) (3.36) (1.98) (3.28) ln(faculty) 0.496 0.319 0.324 1.159* 0.441 0.237 (1.31) (0.54) (0.71) (2.34) (1.01) (0.57) Institution fixed effects Yes Yes Yes Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Constant 1.322 0.855 1.937 - 4.009 - 1.606 - 3.271 (0.71) (0.30) (0.78) (- 1.35) (- 0.76) (- 1.50) R-square 0.87 0.76 0.59 0.83 0.79 0.69 Observations 240 240 240 240 240 240

Standardized beta coefficients; t statistics in parentheses. *p \ 0.10, **p \ 0.05, ***p \ 0.01.

Higher Education Policy 2018 Yuan-Chih Fu et al. Engineering a World Class University?

Figure 6. Change in publication in the highest impact STEM + journals as share of national publications, by WCU’s and non-WCU’s.

WCUs and non-WCUs. The x-axis indicates the national share attributed to each university during the period of 2006–2010, and the y-axis indicates the change in the national share across two periods. Among the non-WCUs, nine raised their national share, ranging from 0.01 to 2.05%, while three WCUs raised their share but another seven WCUs lost share from - 0.27 to - 1.63%. The finding indicates that the participating universities regardless from which group, once in the middle of ladder, were more likely to be affected by the competition and strive to publish in the best journals. This finding provides detailed evidence indicating that competition itself is driving the change in the landscape of the highest impact journals.

Conclusion and Policy Implications After the 2005 policy intervention of supplementary R&D funding the WCUs treatment group did continue to publish a greater volume of papers, but their publication rate did not accelerate faster, and at times was slower than the control group non-WCUs. This is true for publishing in all journals as well as for publishing in journals with different levels of scientific impact. While the first hypothesized treatment did occur, the crucial second one did not. The results from this evaluation of Taiwan’s WCUP parallel other evaluations of university

Higher Education Policy 2018 Yuan-Chih Fu et al. Engineering a World Class University? excellence initiatives in terms of increases in scientific publications. Funding provided by such initiatives generates growth in scientific publications among the selected universities, but not with the much-anticipated extreme accumulated advantage that would lift these universities well beyond where they were at the baseline. The intended effects on the university level are apparently more difficult to engineer than assumed. Given that the selected universities did continue to increase their publication rate, the funds were neither misspent nor are these kinds of policies an abject failure. But now as is clear in the Taiwanese case, the intriguing result across many such initiatives is what happens among the control group of non-selected universities. One explanation of the high performance of the control group is the aforementioned awarded extra-funds to non-WCUs during the treatment period. While the quality of the administration of the WCUP meant high treatment fidelity, these funds could have contaminated the control obscuring a treatment effect. They could also be an alternative explanation to the control group’s publication outcomes. The national survey on university’s R&D funds administered by MOST (2013), shown in Figure 3, however, verifies that these funds used in the non- WCUs for R&D activities very modest, and unlikely a source of much contamination. The unexpected high performance among the non-WCUs suggests a broader process that might be in part set off by the excellence initiative project itself. These kinds of programs come with much public fanfare and discussion; they are not quiet insider tinkering of R&D policy of interest only to a small number of administrators and scientists. In Taiwan, for example, the WCUP was part of a broader and intense discussion of the future of the nation’s higher education system, and some increase in overall R&D support for all institutions. Such publicity charges the idea of selecting the ‘‘best’’ of the nation’s university with extra status and celebration. This institutional normative message is likely not lost on those institutions not chosen, particularly if they have aspirations to be among the best. The public reinforcement of a differentiation of prestige among universities creates the image of a reputation ladder for universities to mark themselves on, particularly in more centralized funding systems with less overt rankings. When that happens, it is the most avid status-seeking institutions in the middle of the ladder whose leadership is particularly motivated to be recognized as being among the ‘‘best’’ and clearly demarcate themselves from others lower on the prestige ladder (Harwood, 2010). This interpretation is also in line with the earlier study showing that over the first year the less funded WCUs performed better than most of the higher funded universities (Chang et al., 2009). Further too, among the share of publications in highest impact journals, not only did WCUs positioned at the bottom of the group grow, but so did the majority of the non-WCUs positioned at the top of the group. This in combination with the DD estimates suggests that a change in the publishing

Higher Education Policy 2018 Yuan-Chih Fu et al. Engineering a World Class University? behavior of all universities (treatment and control) can be attributed to the policy intervention. Qualitative research on what occurred at both WCUs and non-WCUs after the treatment funds were awarded further verifies a diffusion process (Fu, 2017). Explicit strategies and specific actions taken by research administrators to facilitate research and boost its growth at both WCUs and non-WCUs become more similar over time. The responses extracted based on the interviews with the senior academic officers serving at non-WCUs indicates that the WCUP also transformed the non-WCUs’ commitments to research. These transformative changes include the formation of the research clusters, the adoption of performance-based compensation for faculty, and the provision of seed funds for the establishment of the laboratory. These tactics were relatively common on WCUs but rare on non- WCUs prior to the implementation of WCUP. University excellence initiatives may ignite intra-national competition more than they catapult a few universities into the ranks of the world’s most productive. Additionally, what the WCUP facilitates was the institutional-level competition that permeated the whole campus. Therefore, the change at the institutional level was more apparent. The impact of competition at the institutional-level behavior makes Taiwan’s case even more dramatically compared to other East Asian countries where only implemented the program-level competition. There is likely also coercion drawing the competent non-selected universities into the competition. For example, because of Taiwan’s centralized low tuition policy, both its public and private universities heavily rely on governmental appropriation and subsidies for operation. For any university, losing out on a significant new source of funding represented by the WCUP sends a stark message about research productivity and future funding. A kind of aggregated ‘‘publish more or perish’’ message pervades the environment of the excellent initiative. The findings point to the conclusion that these kinds of programs do not produce the intended outcomes in their most narrow sense. Unlike perhaps among individual scientists, engineering a world-class institution among universities is difficult. Indeed, the Taiwanese results as well as those from other large excellence programs suggest it may be impossible, at least at the levels of supplemental funding invested so far. The unexpected effect on non-WCUs echoes arguments proposed by less-known organizational theory: the behaviors of universities can also be driven by the pursuit of prestige as well as the expectations imposed by the resource providers (e.g., Bowen, 1980; James, 1990; Harwood, 2010). Universities, like all formal organizations, are sensitive to their environment and accountability of performance for their legitimation, this chain of factors creates strong pressure toward adopting successful models of competition, and also because a model at a certain time becomes the norm for being a successful university (Mohrman et al., 2008). A possible diffusion effect of institutional norms, greater inter-institutional competition, and isomorphism around science productivity should be investigated

Higher Education Policy 2018 Yuan-Chih Fu et al. Engineering a World Class University? in further evaluations of excellence initiatives and equally should be considered by planners of new programs. The usual excellence initiative may be too mechanical and narrowly behavioral to achieve their goals. There is likely no magic formula for how many and which of a nation’s competent research universities to select for additional funding, and policy-makers should probably dismiss this notion. Perhaps the main message is that inclusive R&D policies accountable to outcomes like publications in high- quality journals that encourage competition among scientists for research funding regardless of their university are the best way to achieve better universities. This is more the historical environment in which the best American research universities thrived (Feller, 1996; Geiger, 2008). In the Taiwanese case the overall outcome was nevertheless positive, and perhaps with adjustment could be made into a broader policy that helps to lift all willing institutions. This could be a good thing for the health of national innovation and science system. The rising and flourishing of these new competitors add to the nation’s R&D, and they also push the most elite institutions to innovate around research management and incentives.

Note

1 Initially, there was a total of 12 WCUs, but one is excluded here because its research portfolio was highly centered on the humanities and social sciences, not STEM. Another one is excluded because it was moved off from the WCUs two years after the WCUP was initiated. To avoid contaminating the following estimation, both of two universities are not included in the group of the WCUs as well as among the non-WCUs.

References

Altbach, P.G., and Bala´n, J. (2007) World class worldwide: transforming research universities in Asia and Latin America, Baltimore: Johns Hopkins University Press. Bowen, H. R. (1980) The costs of higher education: how much do colleges and universities spend per student and how much should they spend? San Francisco: Jossey-Bass. Cobb, C.W., and Douglas, P.H. (1928) ‘A theory of production’, The American Economic Review 18(1): 139–165. Chang, D.F., Wu, C.T., Ching, G.S., and Tang, C.W. (2009) ‘An evaluation of the dynamics of the plan to develop first-class universities and top-level research centers in Taiwan’, Asia Pacific Education Review 10(1): 47–57. Deem, R., Mok, K.H., and Lucas, L. (2008) ‘Transforming higher education in whose image? Exploring the concept of the ‘world-class’ university in Europe and Asia’, Higher Education Policy 21(1): 83–97. Feller, I. (1996) ‘The determinants of research competitiveness among universities’, in A. H. Teich (ed.) Competitiveness in academic research, Washington, DC: American Association for the Advancement of Science, pp. 35–72.

Higher Education Policy 2018 Yuan-Chih Fu et al. Engineering a World Class University?

Fu, Y.C. (2017) ‘Does the project of world class university (Phase I) spur scientific productivity in STEM fields? A study of policy impact in Taiwan’, Ph.D. dissertation, Department of Educational Theory and Policy. The Pennsylvania State University Geiger, R.L. (2004) Knowledge and money: research universities and the paradox of the marketplace, Stanford: Stanford University Press. Geiger, R.L. (2008) Research and relevant knowledge: American research universities since World War II, New Brunswick, NJ: Transaction Publishers. Harwood, J. (2010) ‘Understanding academic drift: on the institutional dynamics of higher technical and professional education’, Minerva 48(4): 413–427. Hazelkorn, E. (2015) Rankings and the reshaping of higher education: the battle for world-class excellence, New York: Palgrave Macmillan. James, E. (1990) ‘Decision process and priorities in higher education’, in S.A. Hoenack and E.L. Collins (eds.) The Economics of American Universities: Management, Operations, and Fiscal Environment, Buffalo: State University of New York Press, pp. 77–106. Lacroix, R. and Maheu, L. (2015) Leading research universities in a competitive world, Ithaca: McGill- Queen’s University Press. MOE (2005) The project of world class universities and research centers for excellence in Taiwan, Taipei: Ministry of Education. MOE (2016) The digest of education statistics, Taipei: Ministry of Education. MOST (2013) The National Science and Technology Activities Survey (restricted data: 2001–2012). Accessed November, 2013. Mohrman, K., Ma, W., and Baker, D. (2008) ‘The research university in transition: The emerging global model’, Higher Education Policy 21(1): 5–27. Salmi, J. (2009) The challenge of establishing world-class universities, Washington, DC: The World Bank. Seong, S., Popper, S.W., Goldman, C. A., and Evans, D. K. (2008) Brain Korea 21 phase II: a new evaluation model, Santa Monica, CA: Rand Corporation. Shin, J.C. (2009) ‘Building world-class research university: the Brain Korea 21 project’, Higher Education 58(5): 669–688. Song, M.M., and Tai, H.H. (2007) ‘Taiwan’s responses to globalisation: Internationalisation and questing for world class universities’, Asia Pacific Journal of Education 27(3): 323–340. Shima, K. (2017) ‘Changing science production in Japan: The expansion of competitive funds, reduction of block grants, and unsung heroes’, in J. J.W. Powell, D.P. Baker, and F. Fernandez (eds.) The century of science: the worldwide triumph of the research university, Bingley, UK: Emerald Publishing, pp. 113-140. Tai, H.H. (2006) World class universities: excellence and innovations, Taipei: Higher Education Press. Wood, F. (1990) ‘Factors influencing research performance of university academic staff’, Higher Education 19(1): 81-100. Zhang, H., Patton, D., and Kenney, M. (2013) ‘Building global-class universities: assessing the impact of the 985 Project’, Research Policy 42(3): 765–775. Zhang, L., Powell, J.J., and Baker, D.P. (2015) ‘Exponential growth and the shifting global center of gravity of science production, 1900–2011’, Change: The Magazine of Higher Learning 47(4): 46-49.

Higher Education Policy 2018