Where do universities recruit researchers?
Vít Macháček and Martin Srholec
Think-tank IDEA of the Economic Institute of Czech Academy of Sciences
Study No. 2/2020
Top universities hire researchers from the global labour market. In such institutions, internal candidates could be even ineligible to apply for open positions to limit the so-called 'academic inbreeding'. Attracting talent from outside brings new ideas, approaches and collaborations, and is vital for sustaining research excellence.
How could we measure the extent to which universities hire researchers from outside? Does the tendency to employ researchers originally from the same place markedly differ across universities from different countries? How does this tendency differ across disciplines and over time?
From the author affiliations in the Scopus citation database, we found how many researchers are currently based at the same university where they were affiliated at the beginning of their research careers. The researchers who published at least one article with affiliation to their current university during the first twelve months of their publication activity are marked as originating from the same place. If their initial articles were published under a different organization, we traced whether this was in the same country or abroad.
Based on this data, we divided current researchers in each university and discipline into three mutually exclusive groups:
The comparison includes evidence from eighteen universities in fourteen countries, including the new EU member countries of the Visegrad group. The results are presented for eleven large disciplines, two groups of researchers by seniority and the initial location given by the specific university or only a city, where it is located.
This analysis is original and its results are not available elsewhere. The results are presented in an interactive manner that allows readers to customize the analysis. The findings should be of interest not only to academics and students, but also to policy-makers and broader public by that matter.
The results presented do not reflect academic inbreeding per se due to limitations of the methodology. We do not know whether the researcher only published one of her initial articles with an affiliation to the university in which she is currently based, or whether this is also her alma mater. If she actually graduated from her current academic employer, this is inbreeding in the true sense, but otherwise it is not.
Therefore, it is important to acknowledge that our indicator conflates the tendency to inbreeding in the sense of a researcher being based at the same university from which she graduated, and career persistence in the sense of remaining employed at the same university where the researcher took her initial job. This is a general indicator of researchers’ (im)mobility, which, depending on circumstances, more or less correlates to inbreeding.
To test, how significant is this limitation, we conducted a robustness test on a sub-sample of 90 researchers in three disciplines. We directly found out from publicly available sources, from which university they graduated and compared this information with the affiliations in their initial articles. The conclusions with regards to inbreeding were the same using both approaches between 77 % to 90 %, which indicates that the presented results are in fact fairly robust to this kind of limitations. Nevertheless, we proceed with caution when interpreting the results, as the match of affiliations in current and initial articles is only an indirect indicator of inbreeding.
Tip: For more details on the methodology see pop-ups with definitions in the introduction.
When employing researchers, the most inward-looking prove to be the national flagship universities of the Visegrad countries: the Czech Republic, Slovakia, Poland and Hungary. More than half of their current researchers are shown to have linkages to the same university already at the start of their careers.
In contrast, hiring researchers originally from outside is most prevalent in the leading universities in the United States and United Kingdom, including Princeton and Oxford. Fewer than a quarter of their current researchers began their research careers at the same university.
Nevertheless, the Visegrad universities are similar in this respect to KU Leuven, the University of Vienna and Lund University in many disciplines. The main dividing line does not seem to follow the traditional 'East vs. West' differences, but rather tends to highlight the gap between those institutions at the top of global university rankings and the rest.
Tip: Switch disciplines in the dropdown menu above the chart to see how the picture is changing.
As can be expected, the flipside of employing researchers whose research careers began at the same university is low internationalization. Cosmopolitan universities in smaller countries, in particular ETH Zürich, maintain the highest shares of researchers with origins from abroad, while a strong national focus is the hallmark of universities in the Visegrad countries.
If the initial location indicator is changed from the university to the city, the tendency to hire from outside naturally appears smaller, especially for the universities that already had a high tendency to hire from inside, and those located in large cities. These sometimes have more than three-quarters of their current researchers originating from the same city.
Nevertheless, a comparison of junior and senior researchers does not reveal a clear overall trend. Several universities noticeably decreased their propensity to employ researchers originally from inside, including KU Leuven, Lund University and the University of Warsaw. However, there is also an opposite tendency, most notably at the University of Szeged and Comenius University.
Tip: Select a group by seniority or a base location by city in the upper menu to examine robustness of the main results.
Next, we complement the overall picture with details on where the researchers who started their careers abroad originated and information on the difference between seniors and juniors in this regard. Hence, it also measures changes over time.
It is important to realize, however, that the researchers originating from abroad may not necessarily be foreigners. It may well be that some are in fact citizens of the country, in which the respective university is located, who began their research abroad and eventually returned to their homeland.
The Visegrad universities continue to hire researchers trained abroad to a fairly limited extent. In contrast, the Western universities not only employ far more researchers with foreign backgrounds, but also increasingly more of them from the junior category. In this respect, the gulf between the East and West is deepening.
As can be expected, in the Western universities most of the flows come from one Western country to another, especially within Western Europe. However, the advent of researchers originating from other parts of the world is striking, most prominently from China and other developing countries.
In the Visegrad universities, the dominant category are also researchers originally trained in Western countries, who in this context are quite likely to be returnees. Interestingly, Eastern Europe represents a relatively limited source of researchers, not only for the Western universities but also for the Visegrad area, despite geographical, historical and cultural proximity.
Initial affiliation abroad by region (% of total current researchers)
Overall, this study demonstrates that relevant insights about academic mobility can be derived from evidence on the affiliations of researchers, which is regularly recorded in published journal articles. This is important because this type of evidence is very difficult, if not impossible, to obtain from other existing sources, especially in a broad international and interorganization comparison.
Jones and Sloan (2019) found that the share of professors at top economics departments in the United States who received their Ph.D.s. at the same workplace was only 5 % at Yale University, 7 % at the University of California, Berlekey, 11 % at the University of Chicago, 12 % at Princeton University, 24 % at Stanford University and 28 % at Harvard University and MIT, which is broadly in line with the proportions identified in this study based on bibliometrics.
Science Europe (2013) is a rare example of an earlier study that engaged in a similar exercise to this study, albeit only at the country level, dealing only with international mobility and disregarding differences between disciplines. Their results showed that during the period 1996-2011, the most 'sedentary' researchers were those employed in the new EU member countries of Eastern Europe, including the Visegrad group, and southern European countries; this is also very much in line with our findings.
Some of the flagship universities in Visegrad countries, which produce the best research in their respective national contexts, could arguably gain little from hiring researchers from elsewhere in the same country. However, it is likely that they would benefit from drawing more on the best available talent from abroad, perhaps from countries with similar (or even lower) wage levels in Eastern Europe or in the developing world.
The main barriers to more hiring from abroad are likely to be twofold. First, the Visegrad universities do not always announce calls for new positions internationally, so foreign candidates may be unaware of openings. Second, the Visegrad universities may not be attractive to foreigners not only because of their relatively low salaries, but also because of outdated human resources practices, including conditions for early career researchers, rigid career development paths and teaching requirements, and administration in local languages.
The best practices of merit-based and international hiring are slowly proliferating at Visegrad universities, but progress remains limited to individual workplaces and positions, rather than constituting a broad trend that would be noticeable in the overall data. Indeed, system-wide changes would require bold structural reforms of the national researcher's labour markets, which are notoriously difficult to design and implement.
Tip: Spend more time with the app and explore in detail universities and disciplines of your interest.
Please cite as: Macháček, V. and Srholec, M. (2020) Where do universities recruit researchers? IDEA Study 1/2020. Institute for Democracy and Economic Analysis (IDEA), CERGE-EI, Prague.
See the references in the list of literature.
Acknowledgement: Financial support from the research programme Strategy AV21 of the Czech Academy of Sciences is gratefully acknowledged. All usual caveats apply.
This study represents only the views of the authors and not the official position of the Economics Institute of the Czech Academy of Sciences or the Center for Economic Research and Graduate Education (CERGE), Charles University. We gratefully acknowledge the comments and insights from Štěpán Jurajda, Daniel Münich, Sergey Slobodyan, Šimon Stiburek a Aleš Vlk.
Hiring from within, or 'academic inbreeding', refers to a tendency of universities to recruit their own graduates for academic positions.
Excessive hiring from within is considered unhealthy in the academic environment, because it restricts the inflow of fresh ideas and leads to imitation or 'parroting' rather than originality.
How much of hiring from within is natural and from what point it becomes detrimental varies between disciplines, the maturity of the particular research topic and the developmental level of the university system.
However, engaging in cutting-edge research increasingly requires cross-fertilization of knowledge from variegated origins, which is why top universities are continuously on a quest to attract new talent from outside.
Scopus is a worldwide citation database containing predominantly journals articles. Scopus is more suitable for the purpose of this study than the Web of Science, because it covers roughly twice as many journals, hence it also provides data on more researchers, and offers more detailed publication histories. The data was downloaded at the end of June and beginning of July 2019.
The study covers researchers who published at least one article in a Scopus-indexed journal with affiliation to the respective university and which was categorized in the respective discipline in 2018. To prevent doctoral students who have not yet entered the labour marker for researchers from driving the results, we include only those who published their first article in 2012 or earlier.
First, all articles published in 2018 that are associated with each university and discipline were found with the help of Advanced Search on the Scopus website, and downloaded as a CSV file. The following query was used:
AF-ID(affiliation-id) AND SUBJAREA(ScopusCode) AND DOCTYPE(ar) AND PUBYEAR = 2018
where affiliation-id is the unique ID of the university, ScopusCode is the unique code of the discipline derived from the code from the Scopus Subject Classification, ar indicates the document type “article” and 2018 is the year of publication. The CSV files contain names, affiliations and the unique Scopus AuthorID of all authors. Due to the Scopus CSV download limit, we only obtained the first 2,000 articles sorted in descending order by the date of publication in each search (this limit applies to only three university-field combinations covered in this study).
In the next step, the authors affiliated with the relevant universities were separated from their co-authors who are affiliated with other organizations. For this purpose of indentifying whether the authors worked at the respective university, we manually created a list of name variants, which contains not only different versions of the university names in English and the local languages but also shortcuts, department names and addresses. We also created a list of city name variants, for cities in which the universities are located.
By gradually adding university name variants into the list, we have successfully confirmed the associations of more than 99.5 % of the articles downloaded with the relevant university. We have only included workplaces directly incorporated into the university structure, not those which are affiliated but formally independent workplaces, such as university hospitals.
Some of the current researchers are simultanously affiliated to multiple workplaces, and it is difficult to determine where they are based. However, these are a minority. Our data show that about 80 % of current researchers have only one affiliation and this share is fairly consistent across universities or disciplines.
For each researcher in the dataset, we determine whether she started her research career in the same university (or city) as her current affiliation. The researchers who published at least one article with affiliation to their current university during the first twelve months of their publication activity are marked as originating from the same place. In other words, they had some initial connection to the institution where they are based now.
To determine this, we downloaded the full journal article publication track record of all researchers currently associated with the universities and disciplines covered in this study using the Scopus API query as follows:
AU-ID(authorID) AND DOCTYPE(ar)
where authorID is the unique Scopus ID of the researcher and ar indicates the document type of “article”.
In the next step, we selected all articles published within twelve months after the author’s first publication. If at least one of these articles included affiliation with the author’s current university (or city), she was marked as originally from 'inside'. The same list of name variants was used as when we identified current researchers. Since this list only contains existing affiliations, and not closed or renamed workplaces, any linkages to the latter are missed. In addition, we determined whether any of the initial articles included affiliation with an organization in the same country where the author is currently based.
Only results based on data available for at least 30 researchers in the respective university and discipline combination are reported.
This study is based on the official Scopus field classification. Due to the large amount of data, we have only considered 11 fields across the scientific spectrum, which are common to the universities analyzed:
The Scopus ID of the discipline is in brackets. The classification refers to "Subject Areas", which are disciplines at the middle level of disaggregation. The only exception is Social Sciences, which combines five subject areas - Business, Management and Accounting, Decision Sciences, Economics, Econometrics and Finance, Psychology and Social Sciences not specified elsewhere, such as Education, Social Geography, Law, Sociology and Political Science.
According to the year of publication of their first article, the researchers are divided into two groups by seniority:
Only researchers who published their first article in 2012 at latest are included in the study to filter out current doctoral students.
The study provides comparisons of eighteen universities from both sides of the former iron curtain and from overseas. Only major national universities that perform research across a broad spectrum of research disciplines are included:
The Scopus ID of the university and country is in brackets.
According to the place of publication of their initial articles, researchers originally from 'inside' are determined by:
If the city is used as the base, the tendency to hire locally is measured regardless of the mother organization, which could be insightful particularly if the researchers are initially affiliated with workplaces connected to the university, but which are formally independent, such as university hospitals.
In this study, we compare the globalization of science across countries and disciplines over the 2005-2017 period. The more often researchers publish in the same journals as their peers abroad, the more globalized their research is considered to be.
The results show that science in advanced countries has traditionally been highly globalized, regardless of the discipline. In recent years, China has profoundly globalized its science system, gradually moving from the lowest globalization rates to a strong position on the world stage. In contrast, Russia and other former members of the Soviet bloc start and remain low, with science that continues to be relatively isolated from the rest of the world.
In Local journals in Scopus, we analyzed local academic publishing in selected European countries over the 2013-2016 period.
The results reveal a strong tendency to publish locally in the former communist countries. Local journals are prevalent in Croatia, Romania, Slovenia, Lithuania, and Czechia, but are rather rare in comparable advanced countries.
In Czechia, for instance, nearly one fifth of all indexed results are concentrated in Czech journals, with a high percentage (>33%) of articles by domestic authors. About half of authors contributing to Czech journals are based in Czechia, and another tenth in Slovakia.
In contrast, the vast majority of articles that appear in journals published in comparable advanced countries are written by foreign authors. The publishing of local, or at best regional, journals appears to be a distinctly Eastern European phenomenon.
In this study, we mapped patterns of predatory publishing across the globe over the 2015-2017 period.
The analysis is based on Beall's lists of "potentially predatory" journals and publishers, of which we found 3,218 journals in Ulrichsweb and 405 journals in Scopus.
The results show that predatory publishing has become most widespread in middle-income countries in Asia and North Africa.
However, the analysis also indicates that Beall’s lists need to be used with caution, as some of the implicated journals may not be necessarily fraudulent in the strict sense.
Macháček, V. and Srholec, M. (2016) Predatory Journals in Scopus. Institute for Democracy and Economic Analysis (IDEA), Study 2/2017.
The study compares quantity and influence of publication output of countries in Web of Science (WoS) by category over the 2012-2016 period.
The quantity is given by the number of scientific articles and the influence is inferred from Article Influence Score (AIS) of the journals they are published in.
For each WoS category, the output and influence of the country is benchmarked relative to the chosen set of other countries.
The results reveal that both the size and influence of research in most Central and Eastern European countries continue to lag behind Western Europe.
This gap is especially strong in Social Sciences, Medical and Health Sciences, and Arts and Humanities
See the interactive application.
Jurajda, Š., Kozubek, S., Münich, D. and Škoda, S. (2017) Scientific publication performance in post communist countries: still lagging far behind. Scientometrics, 112(1). p. 315-328
Publishing practices matter. In disciplines in which frequency of publishing is high and doctoral students customarily publish under the heading of their alma mater, (typically natural sciences) our indicator is close to measuring inbreeding. Conversely, in disciplines in which publishing frequency is low and doctoral students often publish only after finishing their studies (typically social sciences and mathematics), the difference is likely to be notable.
Waaijer et al. (2016) show that for U.S. doctorate recipients in 1950-2010, the median publication of the first article was before the year their doctorate was granted in astrophysics, chemistry, genetics and psychology, and after that year in economics. Based on a sample of the U.S. doctoral students receiving doctorates between 1965 and 1995, Lee (2000) found that the share of those who had already published at least one journal article during their studies ranged between 64 and 86 % in analytical chemistry, 45 and 60 % in experimental psychology, but only 7 and 35 % in American literature. In a database of all doctoral students in Quebec over the 2000–2007 period, Lariviére (2012) found that the share of those with at least one published paper was 63 % in health sciences, 40 % in natural sciences and engineering, 11 % in social sciences and 4 % in arts and humanities. The differences between disciplines are indeed significant.
In addition, the acceptable practice for claiming affiliations in papers when entering the labour market after graduation matters. In some disciplines, notably economics, it is customary in the global (or American) labour market to publish a dissertation paper only under the heading of the first employer, even when the research was conducted elsewhere. In these cases, our indicator will significantly overestimate inbreeding.
Finally, it is important to take into account how much doctoral students are pressed to publish articles in indexed journals. If this is required (explicitly or not) before a student is allowed to defend her thesis, our indicator captures the extent of inbreeding. In contrast, in environments that are more relaxed with regards to publishing demands on doctoral students, we may be far off the mark for identifying a link to the alma mater. Because the pressure to publish increases over time (not only on doctoral students), it is interesting to compare our results for senior and junior researchers.
The test covered three disciplines: i) Biochemistry, genetics and molecular biology; ii) Physics and astronomy; and iii) Social sciences. In each discipline we randomly selected 30 researchers, for which we manually found out from publicly available sources (such as ORCID, personal web pages, etc.), whether they are currently based on their alma mater. As the alma mater we considered university, in which the respective researcher obtained her highest academic degree, in the vast majority of cases a Ph.D. diploma. For 8 researchers we cannot find this information, hence replaced them by another random draw.
The classification of researchers, whether they are currently based on their alma mater, was the same using both approaches by 77 % in biochemistry, genetics and molecular biology, 90 % in physics and astronomy and 87 % in social sciences. The initial affiliations thus turned out to provide a relatively reliable indication of where the researcher graduated. The differences between disciplines are small, which is surprising given the large differences in the propensity of doctoral students to publish, but perhaps this confirms that even if dissertation research is published only after finishing the studies, the researchers still tend to acknowledge their alma mater in their publications.
The classification differed for 14 researchers. Of those there was the same number of 7 false positives and false negatives, which tend to offset each other, thus the impact on the presented ratio to total researchers was even more limited. As far as disciplines are concerned, 7 were from biochemistry, genetics and molecular biology, 3 from physics and astronomy and 4 from social sciences, and regarding the affiliations, 4 were from American and British universities, 6 from western European continental universities and 4 from the Visegrad universities. Hence, it seems that the bias tends to be relatively evenly spread across the board.
There were three reasons for the different classification. First, the researcher published her initial articles with affiliation to her employer, not the university from which she graduated (5 cases). Second, the researcher had a relationship to her current university, hence published with affiliation to it, before or during doctoral studies elsewhere (3 cases of false positives). For example, the researcher published already during master studies, then obtained a doctorate at other university and came back. Third, there are measurement errors, such as a joint Scopus ID for two authors with a similar name or inaccuracies in matching the initial and current affiliations (4 cases of false negatives).
Tato studie vznikla díky podpoře AV ČR v rámci Strategie AV21. Za cenné připomínky děkujeme Danielu Münichovi a xxx. Veškeré názory, případné nepřesnosti, opominutí nebo chyby však jdou pouze na vrub autorů.
Doporučujeme předchozí studie think-tanku IDEA na související témata:
Sidorkin, O. and Srholec, M. (2017) Do direct subsidies stimulate new R&D output in firms? A comparison of IMPULS, TIP and ALFA programmes. Studie 8/2017. Think-tank IDEA při NHÚ AV ČR.
Palguta, J., and Srholec, M. (2016) Stimulují přímé dotace soukromé výdaje firem na VaV? Metoda regresní diskontinuity. Studie 17/2016. Think-tank IDEA při NHÚ AV ČR.
Macháček, V. a Srholec, M. (2016) Transfer znalostí do praxe podnikajícími akademiky v České republice. Studie 8/2016. Think-tank IDEA při NHÚ AV ČR.
Podívejte se na videa ze seminářů k těmto publikacím:
15. 6. 2017 - Seminář "Vedou státní dotace firemního výzkumu a vývoje k novým výsledkům?"
Na stránkách think-tanku IDEA najdete i mnoho dalších studií.
Interactive application is not suitable for mobile phones. We recommend viewing the study on a desktop computer, laptop, or most tablets. Or try maximizing your browser.
If you have a suitable device, maximize the screen and refresh the page (F5).
To keep reading for later, try copying the Internet address and sending it to your PC.
Vít Macháček has worked at the Institute for Democracy and Economic Analysis (IDEA) since 2015. He obtained his Mgr. degree from the Institute of Economic Studies at Charles University in 2016, where he is currently pursuing his Ph.D. Until 2018, he also worked as a Macroeconomic and Finance Analyst at Česká spořitelna. Vítek's research interests include the Economics of Innovation, Scientometrics, and European Integration.
Martin Srholec obtained his Ph.D from the Faculty of Economics at University of Economics and the Centre for Technology, Innovation and Culture at the University of Oslo. Since 2010, he has worked as a researcher at CERGE-EI in Prague. Between 2011 and 2017, he also worked at the Centre for Innovation, Research and Competence in the Learning Economy (CIRCLE) in Lund. Martin has worked at the Institute for Democracy and Economic Analysis since 2013. His research interests include Economics of Innovation, Innovation Systems and Innnovation Policy.