The Monkey Cage

I thought this was an interesting idea, looking at areas of influence according to individuals’ perceptions of their place identification. Rather than presuppose that a specific set of officially defined geographic boundaries is a correct way of capturing a citizen’s context, let study respondents define the boundaries for themselves, through a series of steps capturing what towns and cities people identify with most, from a local scale to a more global one.

While one might quibble with the methodology, the results are certainly intriguing so far. Based on the responses of over 50,000 contributors since 2005, the current map shows a fascinating patchwork of influence-areas that are not highly consistent with well-known geographic conventions such as state boundaries.

The geographic patterns probably most resemble DMAs – the Nielsen Company’s description of “Dominant Market Areas,” or media markets. Even with a rather small number of respondents, I found the expansive pattern of influence covered by Denver to be remarkably consistent with what I knew growing up in western Nebraska where we regularly watched Denver television. For some purposes, particularly as a consumer of mass media, I wasn’t really a Nebraskan much of the time. Texas officially has 20 media markets, but 6 fewer than that appear on the common census influence map. Boise and Salt Lake City encompass vast multi-state areas of influence, as do Minneapolis and Atlanta.

The comparison of this map with maps in which the geographic units are provided to us predefined, raises important questions about the way citizens conceptualize space, how social scientists operationalize context, and how investigators across numerous fields define hierarchical units for examining behavior and attitudes with multilevel modeling. Just how valid is it to cluster voters or citizens by state or by media market, as I have done in some of my own work, when citizens’ ‘perceptual’ regions may look so different? Surely this is a question worth further investigation, as it bears on the validity of key constructs.

In a related article titled, “Geographic Proximity Versus Institutions: Evaluating Borders as Real Political Boundaries,” published in the November 2008 issue of the journal, American Politics Research, Wendy Tam Cho and Erinn Nicley, demonstrate that in some locations state borders have real meaning for political behavior, but in other locations they do not.

Specifically, local populations that are otherwise similar, and highly proximate, but divided by a state border, often behave very differently — clearly influenced by the pull of state politics. In other cases, however, the border doesn’t really matter much at all. The authors cite similar research and findings in the literature on international borders. Cho and Nicley discover that larger counties extend their influence across borders, but that significant rivers along a border create political distinctiveness by dampening cross-border interaction. State borders do exert real influence on constituent populations, as we would expect, but that influence is not uniform. I don’t think this research invalidates the various studies in which states have been used as measures of political context, but it does point to a geographic heterogeneity in the everyday formation of place identities that should not be easily dismissed.