VoxEU Column Education

Why it matters who leads research universities

The best US universities outperform their European counterparts. This column says part of the gap is due to how universities choose leaders. Outstanding scholars are more likely to be selected as presidents in the top US universities, a move that is associated with improved research performance.

It is well known that the top European research universities underperform compared to their American counterparts. The evidence is summarised with policy recommendations by van der Ploeg and Veugelers (2008a,b). The authors raise the need for improved governance, in particular that governments should give universities greater autonomy to manage their own affairs. Leadership is an important part of governance, and again Europe lags. US universities choose different types of leaders; indeed, as I will argue, the achievement of America’s top institutions today may be explained partially by the legacy of outstanding scholars who have led them.

Leading the top universities

The top research universities in the world are overwhelmingly located in the US (Jacobs and van der Ploeg 2006; Goodall 2006). US institutions also house most Nobel Prize winners (see Figure 1). A further difference between the US and Europe is that many of America’s great universities are also led by their best researchers.

Figure 1. Nobel Prizes awarded to individuals in institutions in France, Germany, UK, and US, 1900–2007

Nobel Prizes awarded in chemistry, medicine, physics and economics (since 1969).

Rockefeller University's ninth president, the UK biologist Paul Nurse, is a Nobel Prize winner, and he is not alone. David Baltimore, who stood down as President of California Institute of Technology in 2006, is also a Nobel laureate, as is J. Michael Bishop, Chancellor of University of California, San Francisco. Indeed California, with a population of 36 million, is home to many great institutions that are led by some of the most distinguished scholars in the world. John Hennessy, at Stanford, is a prominent computer scientist; Robert Birgeneau, a Canadian who heads Berkeley, is a top physicist. At UC San Diego, Chancellor Marye Anne Fox is an eminent chemist, and at UC Irvine, the renowned atmospheric scientist Ralph Cicerone was chancellor until he left his position in 2005 to head the National Academy of Sciences.

Could it be that having a distinguished scholar as leader improves the performance of universities?

Using the Shanghai Jiao Tong global league table, Figure 2 shows the average lifetime citations (normalised for discipline) of each president by the position of their university in the ranking. The 100 universities are grouped into quintiles (the ‘1-20’ group refers to the top of the Shanghai Jiao Tong table, where 1 is Harvard). As can be seen there is a clear monotonic decline – the higher the university is placed in the global ranking, the higher are the lifetime citations of its leader (Goodall 2006, 2009b). In other words, better universities appear to hire better scholars to lead them.

Figure 2. Presidents who lead universities higher in a world university ranking have higher lifetime citations (in quintiles)

Can the correlation be explained simply by assortative matching or reverse causality? For example, Harvard and Stanford become great and hence can attract and afford presidents who are outstanding scholars. The longitudinal part of my research suggests otherwise (Goodall 2009a,b). I follow a panel of 55 UK research universities in three Research Assessment Exercises. (The British government’s Research Assessment Exercises use peer review to assess each university’s research output and then allocates funding according to quality and quantity.) Using regression equations and controlling for confounding variables such as size of institution, my results show that those universities led, a number of years earlier, by good scholars go on to perform better in the RAE. Figure 3 presents the results in a simple cross-sectional bar diagram that correlates universities’ later performance in the Research Assessment Exercises with vice chancellors’ lifetime citations – normalised for discipline – some years earlier. The focus here is on the leaders of those universities that made the greatest gains, and the smallest gains, in the Research Assessment Exercises between 1992 and 2001. The universities that improved the most were overwhelmingly led by better researchers.

Figure 3. Universities that improved the most in the Research Assessment Exercises between 1992-2001 were led by vice chancellors with higher lifetime citations

Why might top scholars improve the performance of research universities?

Autonomy has been mentioned in the literature as being a necessary requirement for the success of universities (Jacobs and van der Ploeg 2006, Aghion et al 2007, van der Ploeg and Veugelers 2008). This includes financial autonomy, the freedom to set market salaries, and to hire or fire staff. Managerialism, which has become ubiquitous in the UK among other countries, works counter to the principle of autonomy. Implicit in much government policy are the unproven assumptions that experts, such as academics and doctors, either cannot manage or are somehow inferior to professional managers. My research suggests that this position is flawed.

In qualitative interviews with heads of UK and US universities (see Figure 4), the issue of autonomy was raised. Many said that leaders must have the power to select their own top team members and to create their own decision-making bodies. Once again, practices differ across the Atlantic. In the US, these powers are common, and the UK is moving in a similar direction. But in mainland Europe, rectors’ actions are more likely to be controlled by faculty committees and overly democratic processes.

Why might top scholars improve university performance? Four reasons emerged from the interviews. First, a president (vice chancellor, rector, principal) who is a distinguished scholar will have a better understanding of the core business of a university, that of research and teaching. This is key to the idea (promoted in Goodall 2009) of “expert leadership”, that in organisations where the core business relies on expert knowledge – for example, law and accounting firms, in R&D, management consultancies, and architecture practices – the leader must first be an outstanding expert in the relevant area of business. This particularly challenges the ideas of managerialism that would appear to promote management skills above expert knowledge. Arguably, top scholars, engineers or lawyers must also have management and leadership skills, and in my dataset of 400, almost all the leaders had progressed through managerial hierarchies in their institutions prior to the top job.

A second explanation raised by interviewees, one that again relates to expert knowledge, is that a scholar-leader will likely demand higher academic standards. Arguably, it is leaders who should set the standards in any organisation. This message is articulated by a dean in one of my interviews, “leaders are the final arbiters of quality. Therefore it is right to expect the standard bearer to first bear the standard”. A number of UK vice chancellors said that it was easier to encourage others to submit articles to top journals for the Research Assessment Exercise if they had done so themselves.

Top scholars send out important signals to a number of audiences, according to interviewees. They signal a university’s priorities, act as a beacon when hiring other outstanding academics, and are attractive to students and donors. Finally, it was suggested that scholars are more credible leaders. A president who is a researcher will gain greater respect from academic colleagues and appear more legitimate. Legitimacy extends a leader’s power and influence.

Figure 4. Interviews with leaders in universities

Note: I did not interview Shirley Tilghman; instead she was asked questions about my research by the Princetonian Newspaper.

How do economists fare as leaders?

Finally, it is interesting to see how economists perform as university leaders.

Many years ago, humanists dominated as university heads. Today they have largely been replaced by scientists and social scientists, among whom economists are a rising force. One of the most successful, and arguably most influential, presidents of the 20th Century was the distinguished labour economist Clark Kerr, former Chancellor of Berkeley and the founding president of the University of California, who is often credited with its success. Indeed, labour economists are the largest group within the discipline, possibly because empirical work is more helpful in these roles than theory. Two retired economists recognised as having made an important contribution are William Bowen, former president of Princeton and the Mellon Foundation, and Henry Rosovsky, who was an influential dean of the Faculty of Arts and Sciences at Harvard, although he stopped short of becoming a president.

Lawrence Summers was one of the most distinguished scholars at the head of the Top-100 universities in the ranking mentioned above. When I present this work in seminars, the former Harvard president is often mentioned – “he was a top scholar who failed”. But my response to this question is that Summers spent many years in Washington DC prior to becoming Harvard’s president, and unquestionably the world of politics is played differently to that of academe. Summers may have been out of touch with the norms and values of the academy. In my interviews at Harvard, many reported that Summers, despite being controversial and short-lived, had developed a powerful intellectual vision for the institution and been a good fundraiser, although it is fair to say he may not have been a good manager of people.

The longest tenured Ivy League president, Richard Levin at Yale, is also an economist. At Bocconi University in Milan, one of Italy’s top institutions, Guido Tabellini has become rector. Tabellini is a highly cited economist, as is Kim Clark, former dean of Harvard Business School, now president of Brigham Young University. George Bain, another labour economist, is credited with raising the research performance of Queens University in Belfast, and before that he was both dean of Warwick Business School and London Business School. A number of business school deans are economists – Glenn Hubbard at Columbia, Rich Lyons at Berkeley’s Haas, Laura Tyson formerly at London Business School, and David Begg at Imperial. It has been difficult for business school deans to make the jump to university president. But the tide may be turning, and if so, more economists, preferably top economists, will likely take the helm at universities.


Aghion, P., M. Dewatripont, C. Hoxby, A. Mas-Colell and A. Sapir (2007), “Why Reform Europe’s Universities?” Policy Brief, Bruegel, Brussels.

Goodall AH. (2006), “Should research universities be led by top researchers, and are they?” Journal of Documentation, 62: 388-411.

Goodall AH. (2009a), “Highly cited leaders and the performance of research universities” Research Policy, 38: 1079-1092.

Goodall AH. (2009b), Socrates in the boardroom: Why research universities should be led by top scholars. Princeton University Press.

Jacobs, B. and F. van der Ploeg (2006), “Guide to reform of higher education: A European perspective”, Economic Policy, 47, 535-592.

van der Ploeg, F. and R. Veugelers (2008a), “Towards evidence-based reform of European universities”, CESifo Economic Studies, 52, 3, 2-22.

van der Ploeg, F. and R. Veugelers (2008b), “Towards evidence-based reform of European universities”, VoxEU.org, 27 September.

840 Reads