Looking for Voter Fraud (in old elections) with Data Visualization

The 2020 US election is finally over, and all the election excitement of the last week reminded me of something I had been meaning to look into: Sergey Shpilkin’s data visualizations that clearly show fraud in Russian elections.

I generated the same visualizations for all US presidential elections from 2000 to 2016. The result is that I can’t find any evidence of fraud in any of those elections. But the visualizations show clear evidence of voter suppression of democrat voters. On closer inspection that turns out to be the effects of the electoral college system, which leads to a very interesting conclusion: You might already know that if the president was elected by popular vote, the US would only have had four years of republican presidencies from 1992 to 2024, with the rest being democrat. But these visualizations suggest that just looking at the popular vote actually underestimates the distortion of the electoral college. It also acts akin to voter suppression of democrats, without which national politics would swing even stronger to the left than the popular vote suggests. But lets start by looking at Sergey Shpilkin’s work:

Read the rest of this entry »