Thread by @ulleryatintell

Last week, we took a look at new voters over the past month using Pennsylvania's voter export. @James_McGinnis wrote this story on the lead Democrats have across the state bit.ly/3ou6EyO and the map below is one I made last week showing party control by county 1/

I often consider countywide maps like these not quite as informative as I think voter breakdowns should be. You don't really get the nuanced analysis that you'd probably want to see. For instance:

Democrats outnumber Republicans in all of these southeastern Pa. counties. You can get an idea of which counties have the most Democrats, but that's about all you can really get from a countywide heatmap

This map, however, can tell us a lot more about these counties. This map shows us the difference in percentage points between Republican and Democratic voters (each party as a percent of total voters).

Of course, we're talking about a database of registered voters that's just about 9 million strong right now. So, mapping out all of the ~2,560 towns in Pa. is a lot more difficult than mapping out just these 67 counties.

There isn't a uniform naming convention shared across the board by all counties in the voter database. Then the map file I've used has it's own naming convention, which is another major challenge. It would take an insane amount of time and energy to do that.

And I'm a bit insane. This is really just the start of what is going to be a fun election night to cover. I am missing about 13,000 voters across the state in a handful of towns (usually the ones that sit between two counties). The big missing spot over Erie is water btw.

I'm planning a few different twitter data threads. We can look at overall election history, age, registration date. We don't have any racial data in there, but just about any other demographic dimension should be here. What do you want know before or on election day?

As always, I'm using @PythonPr language, @matplotlib @geopandas @pandas_dev in a jupyter notebook.
@threadreaderapp unroll