Over the past couple of years Nature has hosted an intense debate on the state of Spanish science (see, for example, refs 13). Two issues are repeatedly raised: the lack of sufficient funding and the existence of social networks that, regardless of the candidates' scientific merit, systematically award positions to one of their members. While the former is a priori easily quantified1,2,3, allegations of “blatant endogamic practices”2 have thus far remained untested.

An indication of inbreeding can be obtained by determining whether the address of a scientist's first publication coincides with their current address as a faculty member. Using this measure, we compared possible inbreeding in Spanish universities with the situation in three other countries: the United States, the United Kingdom and France.

We collected data from the Web of Science (ISI, WoS version 4.3, last October on 160 randomly sampled researchers — 40 from each country — holding permanent faculty positions in science departments. In Spain they held the position of 'profesor titular', in the UK 'lecturer', in the United States 'assistant professor' and in France 'maître de conference'.

In order to determine that the first publication to appear in WoS (which spans 1981–2001) was indeed the first paper published by the researcher, we only included individuals whose first recorded publication was in 1984 or later. To avoid biases caused by recently created universities, where all the faculty are necessarily external, we only considered faculty working in universities that had existed for more than 50 years.

The percentage of external candidates that obtained a permanent faculty position in each country is revealing: 93% of candidates to posts in the United States were external, as were 83% in the UK and 50% in France. In Spain, by contrast, only 5% of lectureships were given to individuals who had published their first paper while working in another institution. Differences are very significant (Kruskal-Wallis, p<0.0001), the percentage registered in Spain being at least ten times lower than in the other countries.

These observations are consistent with allegations that lectureships at Spanish universities are almost exclusively awarded to individuals who started their scientific careers in the same institution. Alternative explanations, such as a reluctance among Spanish researchers to move from a geographic region, are unlikely as, in such a scenario, some degree of exchange of scientists between institutions within the same region would have been expected. Spain's Ministry of Education currently lists more than 60 universities ( and the Spanish Research Council ( lists more than 90 science-related research institutes spread across the country.

Our observations thus suggest that Spanish universities are almost completely impermeable to external candidates, effectively preventing the movement of researchers and thus the exchange of ideas and expertise which is one of the keystones of scientific progress.

Attempts to revitalize the state of Spanish science are being made by both Spanish scientists and institutions. These include such welcome initiatives as the Manifesto for a Social Pact for Science and Technology4, and innovative plans by the Catalan government to create research contracts with periodical evaluations of scientific performance. But these efforts will be wasted if researchers are employed on grounds other than their scientific merit.