Tipping Point Analysis and Power Matrix Visualization of 11 Swing States in 2012 Presidential Election

Inspired by this 2012 article in the NY Times’ FiveThirtyEight blog, I used Nate Silver’s tipping point methodology to identify key swing states in the last presidential election.

The outcome of a presidential election depends on each of the states that together cast a combined total of 270 or more electoral votes for a candidate. But, only in some states is the race close enough to actually be considered competitive and not predetermined based on ingrained party loyalty. In addition, thanks to the electoral college, some states are given more weight than others.

The basic gist is that the “tipping point” state, the one that cast the 270th electoral college vote, is the king-maker president-maker.

Why is this relevant?

It seems to me that identifying this “tipping point” state and, let’s say, ten of its nearest neighbors (based on margin of victory and electoral votes) is a way to identify battleground states for the next election.

This attempt at playing political moneyball, of course, carries with it all sorts of assumptions. One of the biggest is using the past to predict the future. Another is looking at a single presidential election and making broad generalizations. We are assuming stable demographics. We are assuming more of the same, politically speaking, by disregarding the possibility of an inspiring candidate who can transcend partisan divisions or one who can grow the electorate.

Getting started

To get started, I gathered the following data:

Using these two sources and a bit of copying and pasting, I built a quick spreadsheet containing the source data.

My next step was to calculate the margin of victory for each state.

It is the difference in votes between the two parties divided by the number of total votes. This allows us to order the states by those most in favor of the Democratic presidential candidate on one end and those most in favor of the Republican presidential candidate on the other.

I populated a new column titled “Margin of Victory,” using the following formula:

Those states that have the potential to determine the election and provide a presidential candidate with the 270th electoral vote are the swing states.

And since margin of victory was calculated in the previous step, I sorted my spreadsheet by that column in ascending order – from the most Democratic states to the least.

Then it was a simple matter of counting up electoral votes for a candidate and identifying the state that cast the 270th electoral vote.

The results resemble Nate Silver’s analysis from a few days after the November 2012 election using preliminary election results available at that time.

Power matrix of the 11 most swing-prone states

Countless times I watched organizers at draw power matrix visualizations on butcher paper to help analyze relationships with allies, opponents, and decision-makers.

In this power matrix each of the 11 most swing-prone states has been plotted with the number of electoral college votes along one axis and the margin of victory for the two leading political parties’ candidates on the other axis.

7% 6% 5% 4% 3% 2% 1% 1% 2% 3% 4% 5% 6% 7% 8%Margin of victoryElectoral votes 28 24 20 16 12 8 electoral votes is national median 4 Eleven Swing States by Electoral Vote and Margin of Victoryof Top Two Candidates for US President, 2012

Pick six

The goal of this visualization is to encourage viewers to take a close look at the data and draw their own conclusions easier than, say, looking at a table full of numbers. This power matrix illustrates power, as measured by electoral votes, alongside degrees of partisan support.