American Politics: The Past 100 Years


CS171 Spring 2015

Welcome to American Politics: The Past 100 Years! We hope you've enjoyed exploring the visualizations so far. This page is mainly here to clarify some technical points about the data, for those interested in the finer details.

First, an important note on third parties: in any given U.S. election, many third parties actually run (and get a small number of votes) that most Americans have never heard of. As a result, it was necessary to apply a standard for which third parties actually got included in the visualizations. After all, a line graph with 100 practically irrelevant parties is overwhelming and inartful. Therefore, we chose to include only those third parties which won a majority of the electoral votes in at least one state. This excludes some third parties which won substantial portions of popular votes without getting any electoral votes, as well as some third parties which received (for example) 1 electoral vote from a faithless elector. (That is, a member of the electoral college who decides to vote for a candidate based on his or her own preferences, rather than those of the voters.)

The 1960 election in particular also deserves special attention. Harry Byrd was a Democratic candidate who recieved quite a few electoral votes from unpledged electors. Those are mechanically different from faithless electors, but the effect is basically the same: a candidate gets electoral votes that (s)he doesn't "deserve" based on the popular vote. Harry Byrd doesn't have any states colored in on the choropleth because he got all his votes from unpledged/faithless electors, and didn't win the plurality of the popular vote in a single state.

One last thing: our main voting data goes back to 1912, but our voter eligibility data only goes back to 1980, and our median household income data only goes back to 1984. This is an unfortunate consequence of assembling data from many sources, some of which are more inclusive than others. We hope that you can still find some meaningful trends in the few decades that are included for these latter sources!

Furthermore, even though the main data goes back to 1912, certain states do stop earlier. Hawaii and Alaska were added to the Union in 1959, so their data begin at 1960. Likewise, Washington D.C. didn't have presidential voting rights until the 1964 election.

Incidentally, you may have noticed that household income is the only data category not directly linked to elections. We included this because of our intuition that the wealth levels in a state may influence the way voters behave. Were we right? You decide!

For more information about the evolution of our design, see our process book here!