Tuesday, December 28, 2010

Redlining Durham

More than two years ago I posted a striking map of Durham from the 1930s which purported to be from the public works department and which coded playgrounds, parks, public areas, and streets either  "White" or "Negro."

 I remember wondering at the time about what exactly the impetus behind the creation of this particular map had been. Later at a lecture given by the wonderful Trudi Abel of Digital Durham (which also features the above map), several audience members asked the same question. Though I don't blog very frequently about  Durham anymore I thought I would share a set of maps which helps answer this question.

First I should say that I would never have known where to go looking for these maps without the work of the T-RACES project at UNC run by Prof. Richard Marciano et al. I was introduced to their work on California mortgage redlining maps at a wonderful web conference whose complete proceedings are available here (thanks especially to Pam Lach and Molly Bragg). In the T-RACES project, Prof. Marciano and others have used maps of several California cities generated by the Home Owners Loan Corporation (a Federal agency of the New Deal era) to display patterns of residential/commercial segregation through a google maps interface.

The HOLC, whose mandate it was to subsidize and secure mortgages, commissioned maps  of cities around the country to guide and restrict the availability of their mortgage subsidies by neighborhood. The HOLC and other banking organizations coded neighborhoods and areas by how desirable they felt they would be for investment. Those areas of a city with working class and non-white residents would be outlined and coded in a way to prevent mortgage underwriting while those with wealthier and largely white residents would be coded as open for federal support.  In their extensive documentation on the project, the T-RACES team  included a list of all the cities nationwide for which these maps exist. Not surprisingly, I was excited to see that Durham was on the list. While I believe the T-RACES team plans to extend their project outside of California to include Asheville, Charlotte, Durham, Greensboro and Winston-Salem (the NC cities covered by HOLC maps), they are not yet online and have been only available at the National Archives to date.

I decided then while on a trip to DC to take a few snapshots of the maps (I will leave it to UNC to produce the high-res scans and detailed overlays!). The Durham maps below were created in the mid-1930s by consulting with local real estate agents and bankers about their perceptions of different parts of the city.  Using these descriptions, HOLC staff created discrete zones handily coded by their desirability for investment. Sample descriptions of these neighborhoods are below:

This neighborhood (close to the Morehead area) gets a C ranking (C-6), the second to worst and notes "Infiltration of: Negroes - gradual."
The Hickstown area above  is coded D (D-4) virtually excluding buyers there from receiving federal underwriting. Note that the HOLC examiners have singled out its unpaved roads, location near the RR tracks, population of "Mill Workers, laborers, mechanics," and its "many" families needing government relief as reasons for exclusion.

Below is a first draft of the map using a standard Durham street map (the same as that used in the public works map mentioned at the beginning of the post) and featuring hand drawn boundaries for neighborhoods with their designation.

This draft map was then officially printed (below) with slight alterations for use by HOLC officials and bankers. Note the A (or those deemed most underwriting worthy) areas: of Watts-Hillandale (A-1) and Trinity Park (A-2) whose houses were secured by racially restrictive covenants and which have received willing support from lenders up through the present day.

I can't say for sure but I suspect that the black/white coded map mentioned above was produced as part of a contemporaneous data-collection project to solidify race and class segregation. As such this set of maps should be seen together as examples of just one of the ways cartography can encode and maintain social inequality.

* For more on HOLC and residential redlining see Amy Hillier's excellent article in a 2005 issue of Social Science History (available to all as a pdf here ). The maps and documents from HOLC relating to Durham and reproduced above are stored at the National Archives (College Park, MD) in Record Group 195 MLR # A1-39 (Box 12).


Rik said...

I remember working in the home stereo business in Durham in the 1960's and when applying over the phone for a customer to finance an expensive system, we were always asked if it was a 1 or a 2. The 1 being white, and the 2 being black.

Andrew Edmonds said...

Hey, Mitch:

I gave a Preservation Durham Lunch & Learn presentation about Durham cartography two weeks ago and unveiled the same Durham redlining map to the audience (via Richard Marciano, who was kind enough to let me download a copy).

I also referenced your blog as a place to go for other interesting maps, so I hope you'll get some increased traffic.

Marsosudiro said...

@Rik -- thank you for the personal glimpse into history. I'm curious whether you were ever tempted to reverse the numbers, for whatever reasons?

@Andy and Mitch -- great minds think alike, and in the same month! I have appreciated both Andy's presentation and Mitch's post.

@Mitch -- "not writing much about Durham anymore"? But why? Oh the pain! Andy mentioned in his presentation that Durham Maps doesn't publish often but it's always excellent.

Mitch Fraas said...

Thanks all - I've heard now that Prof. Marciano has the Durham redlining maps digitized and coded but that they're only available in-house at UNC for now - I can't wait for the NC redlining project to go online to the public. Special thanks to Andy too for all the great work he's doing with PD - ever thought about doing a blog?

Andrew Edmonds said...

Marciano's Google mashup of the HOLC map may now be found on UNC's Main Street, Carolina website, here: