Forgotten Roots and The Problems After World War II

Let me start my address by digging into the past and the destruction of what might have become European political science in the 1920s and early 1930s. Of course, European political science did not exist as a discipline in the early 20^th century, although Aberystwyth University founded a department of international politics in 1919, Lund University inaugurated a political science department in 1877, and Uppsala University Chancellor Johan Skytte established the Skyttean professorship of Eloquence and Government in 1622.

The institutional foundation for political science across Europe and outside Wales and Sweden were nevertheless largely built after World War II. Yet quite a few scientists did modern political science work in the precarious interwar years, but had their careers (or lives) hindered or destroyed by the Nazis. Three colleagues performed pioneering work and who would otherwise be well-known representatives of modern political science.

First, Felix Bernstein (Reference Bernstein1932) was a professor of statistics at Göttingen, but had a keen interest in analyzing elections. He published an early paper on ecological inference 20 years before Robinson (Reference Robinson1950) and Goodman (Reference Goodman1953) were able to place their classic studies in the American Sociological Review. Bernstein, whose article was rediscovered by Shively (Reference Shively1985) and Lohmöller and Falter (Reference Lohmöller and Falter1986) in the 1980s, survived the Holocaust, but he was forced into exile.Footnote ¹

A similar fate befell the Viennese social psychologist Marie Jahoda. Her co-authored study Marienthal: The Sociography of an Unemployed Community (Reference Jahoda, Lazarsfeld and 1933) was a landmark micro-level study of the effects of the Great Depression, and seems timelier than ever during the ostensibly unending Great Recession. Jahoda became a professor at NYU, Brunel and Sussex universities after the war.

Finally, Lewis Fry Richardson made contributions to many fields, notably geography, mathematics and meteorology. His pioneering work on arms races and the correlates of war (for example, Richardson Reference Richardson1935, Reference Richardson1948) mainly gained recognition posthumously through the work of Anatol Rapoport (Reference Rapoport1957), who had returned to Europe in the 1930s to work as a piano player, but fortunately found his way back to the United States and to academia later.Footnote ²

Recent studies show that fascism had a systematic effect on the sciences and was in itself a reason for the growing strength of the United States in academia (Moser, Voena and Waldinger Reference Moser, Voena and Waldinger2014; Waldinger Reference Waldinger2010, Reference Waldinger2012). The emigration of successful chemists, for instance, increased innovation in the United States by about one-third after 1933 in fields in which the émigrés were active (Moser, Voena and Waldinger Reference Moser, Voena and Waldinger2014). The dismissal of highly productive scientists furthermore darkened the prospects for doctoral students both in terms of promotion and research quality. Waldinger (Reference Waldinger2010) shows that the firing of professors for racist or political reasons decreased young academics’ chances to publish their dissertations, cast a shadow on postdocs’ prospects to become full professors and reduced their lifetime citations. The Nazis’ ethnic and political cleansing at German universities soon after Hitler’s seizure of power also negatively affected the academic output of the top departments (Waldinger Reference Waldinger2012). We can be sure that similar trends were also manifest in the much smaller social sciences, including our own discipline.

In the years after World War II, for a long time modern political science could not take off. In this grey era following the black period of the sciences under fascism, political science was often relegated to the status of an ancillary discipline and thus subjugated to the trends in supposedly leading subjects such as history or law. The situation started to change in the 1960s, when many political science departments were founded—sometimes as part of new universities, as was the case at my home institution or the University of Essex. However, the expansion of higher education came at the price of decreasing funding. Plümper and Schneider (Reference Plümper and Schneider2007) have demonstrated this empirically for Germany, linking the shrinking expenses per student to increasing unemployment.

The expansion of higher education also increased postdocs’ chance of receiving tenure at a young age at the end of the 1960s and the beginning of the 1970s. Many of these promotions took place at the same institutions where they received their doctorates. At the Free University of Berlin, these in-house hires were known as Fahrstuhlprofessors (elevator professors), as they were upgraded virtually overnight from assistant to tenured professors.Footnote ³ These colleagues rarely had a decent publication record, thus it is no wonder that their students seldom performed any better. Academic protectionism hence created an atmosphere of coziness, often binding colleagues with similar worldviews and a shared contempt for academic standards closely together. In this context, it is not surprising that grade inflation is more pronounced in disciplines with a qualitative rather than a quantitative focus (Jewell, McPherson and Tieslau Reference Jewell, McPherson and Tieslau2013: 1197).Footnote ⁴

Relying in academic hiring and tenure decisions on criteria other than the applicant’s excellence in research and teaching can have disastrous consequences. Rothgeb (Reference Rothgeb2014, 183) shows that heads of US political science departments who admit that collegiality matters in their decision making are more likely to admit that “[a]t my college/university, tenure has shielded incompetent faculty from dismissal.” The tendency to protect academic underperformers grows with the inclination to rely more on absolute publication counts than on assessments that also take the journal’s quality into consideration. Taylor, Fender and Burke (Reference Taylor, Fender and Burke2006) offer similar evidence for the relationship between individual output and the quality of an academic department. According to their evaluation of US economics departments, individual productivity is closely associated with the number of publishing colleagues.

The mistaken tolerance for laziness that many of us have encountered over the years across European academia can be perpetuated only if achievements are not sufficiently rewarded and if promotion is not exclusively based on scholarly merit. A Nobel laureate, the late Gary Becker (Reference Becker1957), convincingly established that shielding a field or industry against competitors with another ethnic background, age or gender is only possible in imperfect markets.Footnote ⁵ The quite dismal record of political science in gender equality is, in my view, largely due to the impact of unproductive “old boys networks” on our discipline. (Note that fortunately the EPSA leadership now looks much more balanced than some years ago). Unfortunately, I can only report figures about the gender gap in US political science, as no equivalent studies seem to exist for Europe. Hesli, Lee and Mitchell (Reference Hesli, Lee and Mitchell2012, 485) report “a significant advantage for men in the probability of becoming an associate professor, which usually includes tenure,” and Maliniak, Powers and Walter (Reference Maliniak, Powers and Walter2013) add to this a considerable gender gap in citations in international relations, with men citing themselves and other men more frequently than women with regard to their own work and those of their female colleagues.

An institutional barrier to academic progress, at least in continental Europe, is the chair system. It allows colleagues to delegate much of their research to doctoral students who are still often treated as underlings, who have to completely rely in their career on the continuing support of their employers. As beneficial as the division and delegation of labor might be in some cases, relying on assistants for conducting research makes you lazy and uncompetitive in the long run (Schneider Reference Schneider2011). An evaluation of the publication patterns before and after promotion in German business administration revealed that the median chair holder gives up on publishing in journals indexed in the Social Science Citation index once he or she receives tenure (Röbken Reference Röbken2011).

The Changing Landscape of European Political Science

However, times are changing. The chair system is increasingly under pressure, even in as conservative a place as Bavaria, which de facto abandoned it in 2013. Further, funding agencies increasingly require at least some decent publications before splashing out money, and more and more universities have introduced US-style graduate schools. These reforms are root causes of the changes in productivity and efficiency of European political science, which I am happy to report.

Figure 1A shows that the number of articles published by researchers based in Europe is growing. Hence, our field—measured here by counts of articles in journals included in the Social Science Citation index in political science, international relations or public administration—is becoming more productive in absolute terms.Footnote ⁶ This is certainly due in part to the increasing number of quality journals based in Europe, among them of course EPSA’s Political Science Research and Methods, which is too new to contribute to these trends. One might object that absolute productivity does not tell us much about the influence of the articles. According to this adage, the SSCI could be expanding by adding low-impact journals to its list in which mainly Europeans publish. As Figure 1B demonstrates, this does not seem to be the case. Compared to their North American competitors, European-based political scientists have also increased the efficiency of their publication strategies. The increasing average number of citations that articles by European scholars receive is an indication of this trend. The growing h-index of articles by Europe-based authors, summarized in Figure 1C, demonstrates that Europe is also catching up in its ability to produce articles with an eventually high impact. Hence European political science has not only become more efficient, but has also become more excellent. It should be added that this trend happens alongside the continuing domination of the United States and Canada in editing high-impact journals. According to the 2012 ranking, only the Journal of Peace Research was among the top ten political science journals according to the five-year impact factor. A couple of other journals edited in Europe were trailing in the top 20 (European Union Politics, British Journal of Political Science and Scandinavian Political Studies).

Fig. 1

Productivity and efficiency of European and North American political science. 1A: Number of published articles. 1B: Citations per article. 1C: Hirsch index for articles published in a given year

The online appendix demonstrates that these developments are consistent across different subfields such as electoral behavior and armed conflict. Finally, the emancipation of European political science has also reached the field of political methodology. While the substantive focus of methods-based research differs across the Atlantic, European scholars are increasingly conscious that methodological innovations are key for substantive progress.

We are thus witnessing a transformation of political science in some European countries. Figure 2 shows that there are considerable differences in productivity and efficiency across the continent. The Netherlands and the Scandinavian countries top the list of science systems that produce above-average prolific and well-cited scholars. While the United States does very well in absolute terms (and relatively well in relative terms), other large countries such as Germany, Italy and Spain are underperforming. This dismal performance is, at least in the German case, a consequence of the tradition to bury main findings, if they exist at all, in edited volumes or non-English journals that are only read within a narrow internal market and which have, accordingly, a vanishing impact factor. In an evaluation of research on armed conflict, articles on this topic by Germany-based scholars received the smallest average number of citations across the 20 states included in the analysis.Footnote ⁷ It is therefore not surprising that the h-index, which measures the ability of a national system of political science to produce high-impact articles, also differs greatly among countries. Note particularly that the small Scandinavian states have very high h-indices, given the size of their populations, while some large science systems have difficulty coming up with influential research.Footnote ⁸

Fig. 2

Number of articles published 2005–09, the citations they received and the H-index per country. Note: The h-index only refers to four years because of limitations of the Web of Knowledge in dealing with a large number of countries.

Determinants of Scientific Effectiveness

While this descriptive evidence is encouraging, we do not fully understand the factors that contribute to the increasing productivity and efficiency of European researchers. Aghion et al. (2010: 46) show in a comparison of US and European institutions of higher education that universities’ productivity hinges on both the autonomy they are granted and the competition they face: “… by giving more generous stakes for research competitions, governments can make universities use their funding better, use their autonomy better, and respond more productively to local competition.” This suggests, in their view, the need to increase the mobility of students and faculty, and the value of “yardstick competition” programs such as the German excellence initiative or the British Research Excellence Framework.Footnote ⁹

The academic environment in which we operate makes a difference, and particularly affects our ability to do important research. Williams et al. (Reference Williams, Rassenfosse, Jensen and Marginson2013) show in a highly aggregated analysis that financial resources, as well as the quality of the environment, increase the productivity of national academic systems.Footnote ¹⁰ To see whether this also holds true for political science, I have calculated bivariate correlations between the quality of the environment indicator and the average number of citations articles published between 2005 and 2009 have received so far. The correlation amounts to 0.33, and the relationship between the Hirsch index (2005 to 2008) and the environment is of a similar magnitude. Although the scarcity of system-level data on the quality of higher education systems does not allow me to make causal statements, it seems clear that sound policy making can make a difference and improve the lives of educators and students alike.

Also note that universities and their departments can deal with this environment that policy makers set at the national or regional level in different ways. McCormack, Propper and Smith (Reference McCormack, Propper and Smithforthcoming) show the considerable variation in university and departmental management across British higher education. They particularly confirm that autonomy in hiring and rewarding productive scholars increases departments’ productivity, while hierarchical incentives (setting of targets, monitoring) mattered less. As continental universities have a lot less financial autonomy, researchers’ mobility is reduced. This makes initial hiring decisions much more important. I also note, however, given the increasing requests to write letters of recommendation for universities in a wide variety of countries, that there has been an emergence of a European labor market for political science. I am sure our field would experience another boost in competitiveness if all universities renounced the practice of hiring their own doctoral students for postdocs or even permanent positions.

Obviously, the educational background of individual scholars is also important. Hesli and Lee (Reference Hesli and Lee2011) show that the rank of a Ph.D. program improves the productivity of US political scientists. This suggests, in the absence of a similar study in Europe, that the introduction of doctoral programs across the old continent will most likely pay off. However, we are well advised not to base our hiring decisions too much on the prestige of the doctoral school, as tardiness in completing the doctorate strongly decreases scholars’ productivity.