Übers Teilen würden wir uns freuen: @nilsweidmann.bsky.social @humanrights-fau.bsky.social @alicja-plkwcz.bsky.social @krajewskimarkus.bsky.social @argohdes.bsky.social

Werde Teil unseres internationalen Teams und profitiere von interdisziplinären Austausch, u.a. am Exzellenzcluster „Transforming Human Rights".

We find an average treatment effect of a 20% decline in research performance when academic freedom declines substantially. But we do not find any statistically significant effect on the top 10% or top 1% publications, as measured by citation counts in their subject category and publication year. 8/

However, we also explored the effect of academic freedom decline episodes on national performance. We thereby tested whether the current global crisis in academic freedom affects research performance. 7/

Results can be summarized as follows: Whereas differences between countries in enabling AF are related to differences in their national performance, changes in enabling AF within a country are not reflected in annual performance differences to that extent as the differences between the countries. 6/

We found a positive and statistically significant effect of the ‘between AF’ variance on bibliometric indicators with a pooled effect size of 1.7969 (95%-CI = 1.4647; 2.1290]). ‘within AF’ has a positive pooled effect size of 0.0266 but is statistically not significant at conventional levels. 5/

We applied three different identification strategies, namely two-way fixed effects regressions, within-between random effects regression, and difference-in-differences estimators, and used different bibliometric indicators to calculate the effect of AF on national performance. 4/

To determine whether China is an exception and to explore the relationship between academic freedom (AF) and national knowledge production, our study examines the association between an index measuring national AF and bibliometric indicators, such as publication counts. 3/

The remarkable increase in publications from China over the past two decades raises questions about the assumed dependence of effective science production on academic freedom, which is supposed to protect science from external influences. 2/

Results can be summarized as follows: Whereas differences between countries in enabling AF are related to differences in their national performance, changes in enabling AF within a country are not reflected in annual performance differences to that extent as the differences between the countries. 6/

We found a positive and statistically significant effect of the ‘between AF’ variance on bibliometric indicators with a pooled effect size of 1.7969 (95%-CI = 1.4647; 2.1290]). ‘within AF’ has a positive pooled effect size of 0.0266 but is statistically not significant at conventional levels. 5/

We applied three different identification strategies, namely two-way fixed effects regressions, within-between random effects regression, and difference-in-differences estimators, and used different bibliometric indicators to calculate the effect of AF on national performance. 4/