After a week of struggling to find geographic breakdowns of census records for the early 1900s, we did it! Actually Ann Schaenzer did it. Thea and I went to her today to ask if she knew of any good resources, and she pointed us to Social Explorer. Not only does this database provide maps of all US Census records, but you can filter out different kinds of information. AND you can upload your own csv files as a layer. AND you can save your maps after you log in with your St. Olaf Google account. AND you can add collaborators to each project. AND you can customize the colors, symbols, and overall layout of your layers (including the way that the census data is presented). AND you can embed your maps on other webpages, or download them as image files. The map below contains our data from “Where Burleigh Performed,” on top of the racial census records from 1910. The green dots represent white population density, and the orange dots represent black population density. Here is the link to view it online. (It is misleading, but the title of the map is in the upper left hand corner, whereas the title of the selected data is presented on the top).
As stated before, this map provides us with 1910 census data, organized into population density by race. Each orange or green dot represents 10 people, and each blue circle represents one Burleigh performance. What is unclear is how the census data knows where to place each dot. Does it use the addresses provided by the census to show an average location of one race? I would have to ask Ann if she knows how it works. In any case, I think the most interesting way to use this map is to look at a Burleigh location using a very narrow scope, about a block or two in width. Looking this way may help us understand who would have been in proximity to Burleigh, and therefore who his audience was. I think it is also very interesting to look at each county’s census data. For example, Erie County, Pennsylvania was over 99% white. The total population was 115,517 people, and African-Americans accounted for only 392 of them. I did a very quick run through the data and found something pretty interesting. I looked at all of the cities where Burleigh did at least three performances. I then looked at the counties with which those cities belonged to see the racial makeup of the county. Every major county that Burleigh performed in was at least 95% white, except for one: District of Columbia County to which Washington, D.C. belongs, where the black population was only 71% white. I’m not sure if this was obvious to any of my teammates, but I had not fully comprehended this fact before. To continue this line of research, it might be interesting to compare Burleigh’s press coverage in D.C. to that in NY (98% white in New York County and Kings County), Chicago (98% white in Cook County), or Erie (99% white in Erie County).
I think that the ability to upload data to Social Explorer is pretty promising. One downside is that we can’t figure out how to export the data. On the website, the census records can be visualized in various different ways, including shading and dot density. But, as of now we can’t figure out how plug their polygons into one of our pre-existing maps — we’ll have to bring our data to them! Overall, I think it is worth considering embedding this kind of map onto our website.