Abstract :
The paper presented here by Ron Johnston is not only a valuable contribution to the literature of bias in district-based elections, but also a timely call for a re-invigorated spatial analysis of electoral systems. The metric Johnston adopts for measuring this bias has the advantages of being relatively simple in both its application and the interpretation of its results, and appears promising enough to warrant its future use. In these comments, I discuss areas for possible improvement, particularly in applying the methodology to the particular circumstances of U.S. elections, and briefly examine how his findings inform more general questions of intent in partisan and racial gerrymandering. Finally, I second his call for political geographers to pay more attention to the spatial analysis of elections, a topic of considerable interest to politicians, social scientists and voters.
