Published online by Cambridge University Press: 22 June 2016
In many contemporary urban spaces, political information accrues to high status neighborhoods. This might exacerbate political inequality as the information-rich and information-poor each talk primarily with others like themselves. When information is specific and broadly diffused through the media, however, the convenience and low cognitive costs of everyday conversation could be especially helpful for the disadvantaged. This article shows how political conversations intensify or ameliorate spatial knowledge gaps, using a six-wave panel survey in fifty Brazilian neighborhoods between 2002 and 2006. Multilevel models demonstrate that conversation was more frequent in high education neighborhoods, but had a greater impact on specific, factual knowledge in low-education neighborhoods, leading to shrinking knowledge gaps. However, conversation slightly widened spatial gaps in socially perceived general knowledge.
Iowa State University (email: aesmith2@iastate.edu). This research was funded in part by a Mellon Fellowship and an NSF Doctoral Dissertation Research Improvement Grant. A previous version of the analysis was presented at the 2010 Networks in Political Science Conference at Duke University. Thanks to Barry Ames, Steve Finkel, Jonathan Hassid, Jon Hurwitz, Jay McCann, Scott Morgenstern, Dave Peterson, and Robert Urbatsch as well as to participants in the Duke conference for very helpful feedback. The first three wave of the data analyzed here were collected under a National Science Foundation grant to Professors Barry Ames, Andy Baker, and Lúcio Rennó, while Waves 4 through 6 were collected under the direction of Professor Barry Ames with the research funds from the University of Pittsburgh Andrew Mellon chair. I am grateful to Professor Ames for the use of the data from the latter waves. Data replication sets are available at http://dataverse.harvard.edu/dataverse/BJPolS and online appendices are available at http://dx.doi.org/doi: 10.1017/S0007123415000721.