Mapping Racial Segregation

Ethnicity of population from US Census, by Census block, for the NYC metropolitain region. By Eric Fisher.

Major United States cities are racially segregated.  That’s the conclusion that’s meant to be drawn by a beautiful series of mappings of US Census ethnicity data of most major U.S. cities, by Eric Fisher, posted as a comprehensive set on Flickr.  This origins of this project are in a map of Chicago by Bill Rankin, plotting the same Census ethnicity data.  Rankin’s Radical Cartography website is a great place to dive into this type of project.

The segregation reading of these maps is perhaps too simple to be useful – I think we knew this already anyway right?  What’s interesting to me is to analyze what constitutes a racial-neighborhood boundary.  Much of the geometry present in these maps is actually dictated by the data set – the Census-block level information.  In dense cities these are small geographic areas – for instance in Manhattan a Census block is typically only a few city blocks – but their geography is not to be confused with the physical geography of cities.  In less dense areas (check out the Midwestern and Southern cities mapped in the set) the resolution of  a Census block becomes unwieldy.  It is no wonder that major streets, freeways, waterways, and railways divide these swaths of colored dots because this is remnant from the Census geography.  What if the dots were given more autonomy or weight – what if we simply drew a shape around all the blues or reds, or tried to connect seemingly autonomous swaths of color.  In New York, for instance, we might see lines of infrastructure like a subway line, or historic reshaping of the city by Robert Moses, which connect ethnic populations.  It will be interesting to watch the tags on flickr that users have attached to these maps – hopefully they can start to inform the graphics with deeper understandings from people familiar with the social and historical significance of these places.  I hope they will be put into a better interface – perhaps more like Google Earth – that will allow people to zoom in and layer other information over them.  Right now they are beautiful images, but for us to really interpret them, we need to interact with them as a geographic tool.

Advertisements