There is a famous saying, by H.L. Mencken, that “every complex problem has a solution that is simple, neat – and wrong”. This is more than ever applicable to the complex problem of trying to describe the political landscape of the United States. The simple solution that the Twitterverse, one-sided cable-news talking heads, and some of the more extreme politicians on either side of the political spectrum want us to buy into, is that there are only two buckets of people: with us and against us.
The simple, neat, and wrong binary split of America and Americans is wrong because it denies reality. When you sit down with people to talk about what they really believe in, you will find that things are way more complex. Some people will be pro-choice and pro-gun. Others will be against trade tariffs and for a southern border wall. Yet others will want a $15 minimum wage, but don’t want single-payer healthcare. So, if we cannot just sort people in the red or blue bucket and label them good or evil depending on which bucket we belong to ourselves, what can we do instead?
How Politicians Stack Up
Enter the ActiVote Political Matrix.
The Political Matrix plots every legislator and voter in a 2-dimensional matrix, where people are closer to each other in the matrix if they have more similar political opinions. Let’s look at one matrix depicting the 4 recent candidates for President in 2020, representing the Democrats, Republicans, Libertarians and Green Party, plus the 2 main Vice Presidential candidates.
Among these six candidates, unsurprisingly, on the left-to-right spectrum, we see 3 candidates on the left and 3 on the right. Also, unsurprisingly, Howie Hawkins is the most left-wing candidate of the three, while Joe Biden is the most moderate, with Kamala Harris the more progressive part of the Democratic ticket. On the other side of the aisle we see 3 clearly right-wing/conservative candidates, spread out vertically. Libertarian Jo Jorgenson, as expected, is hugging the “Individual Focus” axis, while Donald Trump is much closer to the “National Focus” axis. Former Vice-President Pence is plotted between them, not illogical given his focus on state rights, individual rights, and more. Check out how each candidate got to their position if you want to see the math in action.
If we look at larger collections of politicians, things become even more nuanced. The following picture shows every single State Representative, State Senator, US Representative and US Senator currently serving in their legislature. Blue dots are Democrats, red dots are Republicans and yellow dots are all others:
There are some legislators on the far edges of the political spectrum, both left-and right, but about ⅓ of all legislators are actually closer to the center than to the edges, showing a gradual spectrum of opinions.
How This Works
By now you may wonder how people are plotted on a particular position and whether this is done reliably. So, let’s look at the exact science behind the inexact problem of plotting people in a political matrix.
ActiVote uses two types of information. For those who actively participate with ActiVote: a survey of questions on a wide-variety of topics, each with 5 possible answers on a spectrum ranging from (very) progressive to (very) conservative. For legislators, we use their full legislative record through all the votes they have cast. While some votes don’t say much about their position at all (e.g: the unanimous vote to name a post office), other votes are very telling if some legislators support bipartisan legislation, while others do not.
Based on all this information we create a “political fingerprint” which indicates for each of the 18 policy categories where they stand. Finally, we take the political fingerprints of all those legislators and feed it into a data science algorithm to find the best 2-dimensional representation of their positions such that people are closer to each other if they have more in common and farther away if they share less.
While we did not tell the algorithm anything about the meaning of “left” and “right in our political discourse, in the results the horizontal axis clearly has a lot in common with the standard left-right spectrum. Nor did we dictate anything to the algorithm about “government focus” and “individual focus” or “global vs. national focus”. However, that is the grouping the data science algorithm came back with.
So, positions are calculated without any bias and can and will change over time: the more survey questions people answer and votes they cast, the more fine-tuned their “political fingerprint” becomes and the more accurately they will be plotted in the matrix. Also, when the political winds change, positions may change too. For instance, the position on trade held by the major parties today is very different from what it was just five years ago.
Party Caucuses
When looking at Congress it is interesting to look at specific subgroups of which we would have a mental picture of where we think they would be. Such as “The Squad” (far left in the matrix?) the Tea Party (far right in the matrix?) and the Problem Solvers Causes (somewhere in the middle?).
The Squad and the Freedom Caucus show our division, and the Problem Solvers Caucus shows the wide group of legislators working to find compromise:
We believe that the Political Matrix is a much better tool to depict what people really think than the two-bucket, simple, neat – and wrong, binary answer. The Political Matrix shows nuance and it shows that many people are closer to the center than to the edges of the political spectrum. It can be the basis for dialogue, not just about where we disagree, but especially also about what we agree on.
You can find the overall depiction of the state legislatures and Congress at the bottom of this article. If you want to see where you personally stand, (by answering survey questions) or explore the candidates running in your elections and legislators representing you, download the ActiVote app.
