A Better Map for Purple America

Have you ever seen the “Purple America” map?

A “Purple America” Map — showing American counties shaded by % of vote.

It was the result of a freshman computer science project developed by Dr. Robert Vanderbei, an Operations Engineering Professor at Princeton. While it was only meant as a course assignment, it gained popularity by showing that America’s states and counties are really not as different politically as we might think.

For example, North Dakota and Minnesota are very different electorally (they have consistently voted Republican and Democrat, respectively, for the past 44 years), but I’d challenge you to find a clearly-defined political border between the two states on the Purple America Map. What about the political border between West Virginia and Pennsylvania? It’s not as defined as we’d expect.

However, the map we see on election day won’t look like that. On election day, we’ll most likely see a neat vertical line between red North Dakota and blue Minnesota, a sharp mountainous divide between red West Virginia and blue Pennsylvania, and probably a wrinkled valley divide between red Arizona and blue California , like we’ve seen for decades.

That’s a problem.

In our state-based winner-takes-all electoral system, the winner of the popular vote in each state usually* wins the entire electoral slate from that state. That’s why California is always seen as an overwhelmingly democratic state, even though California is home to more Republicans than any other state in America¹. And that’s why Texas is seen as a Republican bastion even though there are more Democrats in Texas than there are voters in 38 other states².

Not only does this system make us overlook large numbers of voters in states that aren’t swing states, it also makes it much, much more difficult for 3rd party candidates to convince voters to support them — after all, it’s near impossible for a traditional 3rd party candidate to see reward for their efforts on an electoral map, unless they flip a majority of voters in a state (which hasn’t happened since ‘68).

Here’s my own solution to the electoral problem — a map where the electoral slate is divided by the percent of popular vote in each state. The map below shows the results for the 2012 election between Barack Obama and Mitt Romney, with the winning party in each state colored on the top, and the losing party on the bottom. You can see the large number of California Republicans and Texas Democrats who are usually considered irrelevant to the presidential election.

The 2012 Electoral College map, if Presidential Electors were awarded proportionally to the popular vote by state. Red represents Republican Electors, Blue represents Democratic Electors, and Yellow represents Libertarian Electors. Credit to FiveThirtyEight for the population-based hex map design.

See that yellow hexagon at the bottom of California? That’s an electoral vote for Gary Johnson, the libertarian candidate in the 2012 election. He gathered 1% of the vote in 2012. Although the 2016 electoral map likely won’t show it, he’s doing 8 — 10 times better than that this year, according to polling³.

A map like this wouldn’t change the outcome of an election (except in 2000, when the winner of the popular vote lost the electoral college), but in an election like 2016, it might help the smaller party and 3rd party candidates in a state prove they can win electoral votes, even if they can’t flip the entire state.

Footnotes and Sources:

The data used in these maps can be found here: https://docs.google.com/spreadsheets/d/1xF_MiCuxLIBIyjAt4f928FGz1HZEDyyt-vn9IjXZ6tQ/edit?usp=sharing.

*the exceptions being Maine and Nebraska, who award electoral votes by Congressional district, as well as for winning the state at-large.

¹ From the Federal Electoral Commission’s results for the 2012 election, 4,839,958 Californians voted for Romney, the most out of any other state.(Texas came in second with 4,569,843 Romney voters)(http://www.fec.gov/pubrec/fe2012/federalelections2012.pdf)

² From the Federal Electoral Commission’s results for the 2012 election, 3,308,124 Texans voted for Obama, more than the entire electorate in states smaller than New Jersey (38 in total). (http://www.fec.gov/pubrec/fe2012/federalelections2012.pdf)

³ From FiveThirtyEight’s current projections for Gary Johnson (http://projects.fivethirtyeight.com/2016-election-forecast/)

Credit to FiveThirtyEight for creating the electoral hex map.