LEADERSHIP IN SCHOLARSHIP: Evidence from the American Economic Review of editors' influence on the profession's narrative

Editors of the American Economic Review (AER) are appointed to steer good papers in growing topics towards the journal. There is no indication of any abuse of power: it sees that any steering of the topics of the AER towards topics studied by the editor is explained by the differences in topic coverage before the appointment.

These are among the findings of new research by Ali Sina Önder, Sergey Popov and Sascha Schweitzer, presented at the annual congress of the European Economic Association (EEA) in August 2020.

In many fields in academia, careers and research agendas depend on publications in peer-reviewed journals. A lot in that process depends on the editor: desk rejection depends on the editor’s first read of the paper, the choice of referees for review is frequently up to the editor, and the final decision, while frequently coinciding with referees’ recommendations, sometimes gets overruled by the editor.

Since editors affect the distribution of scarce resources – pages in the journal – economists study the impact of editors on who gets published. Some papers study whether having an editor as co-author helps you publish in journals; others evaluate whether having an editor colleague is good for your publication chances. 

This study addresses a novel question: ​what​ gets published. Editors are researchers, and the content of their research comes into consideration when evaluating candidates for editorship positions.

Somewhat luckily for the research design here, writing a paper in economics takes years, so the appointment of a new editor can be used as a natural experiment, and randomness, among other things, comes from the topic compositions of the editor.

So the question is: if a new editor, who works in the field X, becomes an editor, does it make it easier to publish papers on topic X in that journal? On the one hand, the new editor might take papers from their field favourably; on the other hand, the competition for space in the journal might go up. 

The researchers use the textual analysis method provided by the Latent Dirichlet Allocation (LDA) to analyse quantitatively the topics of published manuscripts. LDA makes it possible to estimate both the topics that are covered by the papers published in AER and other journals, and each individual paper’s composition of topics.

The online library JStor provided data needed for classification of texts, and the authors are able to contrast trends in topics of the AER, the flagship journal in economics, and in topics of four other top journals, and see how changes in these trends correlate with topics of newly appointed editors.

Depending on the specification, there were around 20 new editor appointments in the time span of 30 years (some had to be dropped because they did not stay long enough), which made it possible to proceed with the analysis. 

There is some correlation between topics of the incoming editor and the shift in the topics of the AER. But the editors don’t shift topics much beyond what could be predicted by knowing the topics of the AER and the rest of the ‘Top Five’: the AER tends to publish things that were popular in the rest of Top Five and underrepresented in the AER in the previous period. 

The topics of the incoming editor seem to be strongly correlated with topics covered by other top journals, so it seems that whoever decides on the appointment of the editor tries to boost AER’s representation of topics popular in comparable journals. 

Results persist doing the same analysis using JEL codes; citations seem to not matter, except a high number of citations in topics of the incoming editor seem to be associated with low citations in the rest of the Top Five before the appointment.



Ali Sina Önder (University of Portsmouth), Sergey V. Popov (Cardiff University), and Sascha Schweitzer (University of Bayreuth)