Powerless Conservatives or Powerless Findings?
By Stephen M. Utych, Boise State University
Recently, scholars have argued that research supporting a conservative viewpoint may be less likely to be published in political science journals, a field that is disproportionately liberal. Specifically, some have noted that there was considerable work on bias against women in the 2016 presidential election, but relatively little about bias towards men. I present an alternative explanation of this: that if bias against men did not impact the results of the 2016 election, lack of publication of these results may arise not from bias against conservatives in academia, but from a bias against publication of null or inconclusive findings. I use data from the 2016 American National Election Studies to show that a measure that could show bias against men provides mixed results on candidate support depending on how statistical models are specified. While this does not demonstrate that bias against conservatives does not exist, it provides an alternative explanation for differences in publication patterns. I argue that a more accepting peer review process, one that is willing to publish well-designed studies, even when results are inconclusive, would help to alleviate this issue.