Poole brings math, politics together

Editor’s note: This article was initially published in The Daily Gazette, Swarthmore’s online, daily newspaper founded in Fall 1996. As of Fall 2018, the DG has merged with The Phoenix. See the about page to read more about the DG.

Yesterday, Dr. Keith Poole, Kenneth Lay (yes, that Ken Lay) Professor of Political Science at the University of Houston, gave a talk entitled “Political Polarization and Economic Inequality”, a discussion of the content and methods used in the book of the same title written by himself, Nolan McCarty, and Howard Rosenthal.

Before getting into the actual data, Poole showed how mathematical techniques (now performed by computers) can create visual models from data tables by showing a table of driving distances between US cities and then showing the map that a computer made from that data. It looked quite similar to an actual map of the United States. Many highly technical mathematical terms were used in Poole’s explanation, and Professor of Mathematics Stephen Maurer, who was in the audience, seemed intrigued by the ideas that were in play.

Poole started work on this project 20 years ago. He has entered every single Senate and House of Representatives roll call vote into a computer program that uses advanced mathematical techniques to create a visual model of how each candidate has voted over time. Poole then examined the visual map, applied his knowledge of political science, and determined which areas of the map correspond to a liberal voting record and which show a legislator to be conservative. The end result is a number line with each legislator positioned somewhere between -1 and 1, with zero being neither conservative or liberal. A higher negative value means that a congressman is more liberal, and a higher positive value is an indicator of conservatism. This model can then be used to predict the results of any given vote with a high degree of accuracy.

Poole added a second dimension to his analysis to account for issues that cut across party lines, such as the 1960s civil rights issue that split the Democrats into northern and southern contingents. When no such issue existed, the dots (each dot representing a senator or representative) were virtually without exception split into two groups, with democrats on the left and republicans on the right. During the era of civil rights votes, however, the southern democrat dots moved rightwards and sometimes even crossed the line onto the republican side. Those dots were also high up on the diagram to show that they consistently voted against civil rights bills, while those who voted the opposite were near the bottom.

As the Democratic Party was divided over a major issue, this was not an era of political polarization. The more polarized a political system is, the more members of each party tend to vote together. Poole presented graphs of every Congress in United States history, showing how political polarization has changed over time. We are currently in a very polarized era, with levels of polarization not seen since early in the 20th century.

Poole now came back to the main thrust of his argument by showing a graph of income inequality in the US, measured by the percentage of wealth held by the top 1% of Americans. He presented a line graph of this data, and then overlaid a line graph of the degree of political polarization onto the screen. The two graphs were very similar, with a Pearson coefficient of .81. The relationship between polarization and the Gini coefficient of income inequality is even stronger at .94.

Why does this occur? The large screen in SCI 101 displayed Poole’s main point, which was: “Political polarization and the distribution of income should move in tandem because the primary dimension of American politics has been the role of government in the economy). Historically, world wars have both caused tax rates to increase (reducing income inequality) and unite rather than polarize Congress. Political polarization increased sharply in the 1970s and 1980s, as the neoconservative movement took hold and brought the Republican Party further to the right. Ronald Reagan’s tax policies then caused an increase in income inequality. The economy can drive politics and politics can also drive the economy.

Audience questions were taken both during and after the presentation. Poole became very animated when describing how the recent settlement between several US states and tobacco companies amounted to a tax on poor people (who now have to buy their cigarettes at higher prices) as the money paid to states is not being used to help the poor. Poole apologized repeatedly for going “too long,” but it was clear that he had invested a lot in the project and wanted to share his work.

Leave a Reply

Your email address will not be published. Required fields are marked *