Mapping the 1950 Census in Grand Forks

I have a few hobby projects that keep my occupied during my down time. In most cases, these project serve as a way to forestall burn out while still having the potential for being professionally productive. In other words, they’re a distraction, but not of the most frivolous kind.

This weekend, I decided to see what would happen if I took the addresses found in the 1950 census, which some work study students generously normalized, and connected them to the addresses found in the parcel map downloadable from the city’s website. 

There are a bunch of problems with doing this, of course, and many of them are a bit complex to solve, but my first draft look does not appear entirely implausible.

Unfortunately this map only shows about 65% of Grand Forks residents in 1950. This means about 10,000 residents are homeless. Some of the missing residents are from neighborhoods lost during the flood. Lincoln Drive for example probably accounts for around 10% of them or so. Another 10% might well come from the campus area where students in dormitories, especially the post-war temporary dorms, are not in this map. Another group of residents not on this map come from areas lost to urban renewal. 

I’ve started to digitize the various lost neighborhoods (which are in blue on this map) and this should maybe account for a few thousand residents whose addresses no longer exist in the parcel map.

The real challenge is that the Lincoln Drive neighborhood has to be reconstructed from 1994 topo maps which are still being scanned and then have to be georeferenced and digitized.

Finally, another group of Grand Forks residents lived in addresses throughout the city that simply no long exist. I suspect that downtown will be particularly filled with these addresses as the city has lost a number of multifamily buildings downtown. (The map above, then, is biased toward single family residences and underrepresents population density around campus and downtown!). Some lost addresses will likely be single family homes replaced by apartments or commercial buildings. These addresses might represent ephemeral kinds of settlements like trailer parks or other forms of temporary post-war housing. In some cases, houses were lost during the 1997 flood or isolated fires and never rebuilt. One of the interesting things that we’ve learned is that trying to map minority groups, such as Native Americans, we’ve found that many of the addresses where they lived in the 1950s no longer exist. It’s a bit too hard for me to know exactly why this is and the sample is pretty small, but it’s still worth interrogating! 

Over the next month or so, I’ll be trying to fill in the lost addresses and integrate them with existing maps. 

To make things just a little more fun, I’ve been doing this in both ArcGIS Pro and in QGIS as a way to develop my GIS skills just a bit more. As the old saying goes: we don’t have hobbies before they’re fun! We have hobbies to put the frustration that we encounter in our daily jobs in better perspective.