The Resistible Decline of European Science

Using a data set of highly cited researchers in all fields of science, we show that the gap in scientific performance between Europe, especially continental Europe, and the USA is large. We model the number of highly cited researchers in a sample of countries as a function of physical and human capital and a country-specific, factor-augmenting Hicks-neutral productivity term. We find that differences in productivity between Anglo-Saxon countries and other countries are not solely due to differences in the levels of inputs. Not surprisingly, our results reveal the importance of English proficiency. However, they also show that the governance and design of research institutions that characterize Anglo-Saxon countries, as well as a few other countries that have similar institutions, is another critical factor for research output.