GIS Mapping of Polling Place Data for Northwest Germany, 1893-1933

​​​

As I discussed on my data collection page, my map making began with my work on Diederich Hahn. In 2000, I attended a meeting of the Association for History and Computing where I presented the very preliminary that I had done, with maps drawn and colored in Adobe Photshop. At that time, people asked me if I had thought about loading my work into a GIS. After the session, a very kind graduate student explained to me polygons, shapefiles etc. and directed me to ESRI's ARC-GIS platform. Off I went.

​

This work has evolved into Elections in Northern Niedersachsen. A Database and Historical GIS a georeferenced map of polling and census places in northwest Germany (+3000 hours/work), populated with local election results from 14 national elections between 1893 and 1933 (+500K data points) and data for 761 census places from 5 national censes plus other social/economic data (+2000 hours/work) collected from 42 archives and libraries in Germany, the US, the UK and Israel. I have yet to find the proper venue to archive my data. For a long time the Inter-university Consortium for Political and Social Research (ICPSR) at the University of Michigan housed an impressive collection of German election and survey data. Unfortunately the Michigan people no longer curate the collection and it is nearly impossible to find on their website. The next (or perhaps most logical) place to archive my data is the German Social Science Infrastructure Service (GESIS) at the Leibniz Institute for the Social Sciences at the University of Mannheim.​

​

CREATING THE GIS

​​​​​

My first attempt at creating the GIS was bare bones rubber sheeting. I added a JPG based on a copy of the 1957 Verwaltungsatlas that I made at the Niedersächsisches Staatsarchiv Stade as a base layer of my GIS. I then constructed a polygon layer of my polling places using the atlas’s boundaries. This served my purpose at the time quite well until I inadvertently deleted the shape file in the San Francisco airport coming home from a conference in Berkeley. After much heartache, the GIS librarian at Miami's King Library helped me reconstruct the boundaries although I had to manually rename and renumber them.

​​​​

Further down the road, I took the suggestion of a fellow conference panelist and began to georeference my polygons. Not an easy task unassisted and uninstructed! Many hours later, the map was properly georeferenced. Over this, I created a separate layer of all of my polling places (Wahlorten) and census places that were not Wahlorten using base files of Bing and Google areal maps. I also at this time georeferenced layers of old historical maps, and layers displaying railroads and canals. I got the insane notion to adjust a layer showing all watercourses at 1:10,000 scale, hoping to suggestion ways that farmers could get their goods to market. At this time, I also decided to move from the ARC-GIS to QGIS, widely used by my European colleagues. 

​​

With this put to bed, I saved my work and returned to my traditional narrative. Before I returned to the GIS, the university provided me with a new computer. Unfortunately, they did not save the file on my old computer where my GIS was stored. This would mean another +1000 hours for a second reconstruction. Fortunately - and I thank God for this - in the time between the original creation of my Wahlorten shape file  and this catastrophe the Niedersächsisches Landesamt für Landesvermessung und Geobasisinformation made available online Kreis, Gemeinde, Gemarkung and Flur polygons. The later two have been particularly useful. In 1974, their Land conducted a major Gemeindereform, consolidating many of my polling places. With the aid of the Gemarkung and Flur maps, I have been able to reconstruct old village boundaries. There was only one problem with this: QGIS did not entirely like the polygons constructed by the Landesamt and in some cases refused to divide them. This took a great deal of time working around.

​​

I hope in the near future to add a Bauernschaften and Ortschaften layer, in keeping with my research on Theodor Tantzen and Oldenburg progressivism. I am also trying to integrate land use/crop regime maps in such a way that one can see the extent to which they played a role in political decision making and electoral results were tied to specific agricultural products.

​

THE UTILITY OF MY GIS

​​

Using the GIS, I am able to leverage micro-level data in such a was to challenge existing scholarship and develop new hypotheses. I have already integrated these finding in my monograph on rural antisemitism and am startling to use it to examine rural liberalism in what I hope to be new and useful ways.