It is an undeniable fact that the Electoral College system makes some voters more important than others depending on where they live. We all have a pretty good idea whether our vote "matters" depending on our home state. The most important voters live in Pennsylvania, Florida, etc.
Defenders of the Electoral College offer several arguments in favor of this distortion, including a warning that a popular vote would be dominated by big cities. Often implicit in this argument is that "urban" (read: black and brown) votes are less legitimate than those of Real Americans™ in the Heartland. The Electoral College, they say, protects rural voters.
It can be argued whether such a system is desirable. I, for one, think that if 60 percent of the country lives in cities, then 60 percent of voting power should go to individuals who live in cities. However, such a discussion is irrelevant if the underlying factual assumption turns out be wrong in the first place. If the Electoral College doesn't give more power to rural voters, the question of whether it should becomes as important as whether it should protect left-handed voters or Capricorn voters.
In order to see whether, as a matter of fact, rural voters get a boost over city dwellers, we need to measure the impact of the Electoral College on the importance of each vote. For that purpose, I developed the Voter Importance Index, which measures the likelihood than any individual voter in a state will change that state's electoral votes, adjusted for the number of electors representing that state. This way, it captures the benefit given by the Electoral College to small states and the benefit given to swing states.
The Voter Importance Index is calculated by dividing a state's electoral votes by the number of decisive voters in that state — that is, how many it would take to flip the state from one candidate to another. The higher a state's Voter Importance score, the more impact individual voters in that state have over the Electoral College. I have already used the Voter Importance Index to show that, contrary to popular belief, the Electoral College is bad for small states. All Voter Importance Index data from the past 20 elections can be found here.
Using this Index, we can determine whether the Electoral College actually benefits rural voters over urban ones. Of the top ten least urban states, only two (New Hampshire and Maine) make it to the top ten of this list for 2016. West Virginia, Kentucky, and Alabama, three of the most rural states, find themselves among the least important here. Upon first glance, it does not appear that a rural population is any help in the Electoral College.
Let's consider all the states. Below I have plotted each state's Voter Importance score against its share of urban population.
Here we find a clean spread of states, both urban and rural, at the very bottom of the importance scale. Both the most rural place in the Electoral College (Vermont) and the least (DC) have citizens whose votes hardly matter. The trend line shows the absence of any pattern between the two variables, only explaining 0.5 percent of the variation among states.
The chart above is made hard to read by the outliers of Michigan and New Hampshire. If we remove the from the list, the resulting chart looks like this. Notice the slightly positive trend and the group of important, urban states (Pennsylvania, Florida, Nevada, Arizona), both suggesting that rural voters don't get much of a boost.
This absence of a pattern was not just present in 2016. Below is a chart with data from all elections since 2000. To preserve comparability across different elections, I used the z-scores of each state's Voter Importance rather than raw scores.
Expanding the sample size produces the same lack of a pattern that we saw before. There is still a flat trend line, this time explaining 0 percent of the variance. The heavily urban perennial swing state of Florida takes the cake here in term of importance. If you direct your attention to the bottom left corner, you will also discover that all ten of the most rural states are below average importance.
Finally, we can also establish that this is not a modern anomaly. Take the example of the 1972 election, plotted below, with its winner as the heavily urban Rhode Island.
So if fans of the Electoral College want a voting system that elevates the voices of rural voters, then fine. But it's not the one we have now. Year in and year out, our presidential elections boost voters in swing states to the detriment of everyone else. The numbers are in: the Electoral College does not protect rural voters.
Photo credit William Hochberg, Creative Commons,
Voter Importance Index data