Applying mathematics to study trends in politics bears a resemblance to predicting weather and climate. There are so many unknown and unspecified variables, that at best it is only possible to make statistical guesses about the future.
There have been a number of models to predict presidential election outcomes, models that are generally driven by economic data. These models predicted a sweeping victory by Vice-President Al Gore in the 2000 presidential election. Many in 2000 thought that such models represented political destiny and did not appreciate that the models predict a statistical result. All other things being equal, one would have expected a Gore landslide victory, but all other things are never quite equal.
Nonetheless, political models and mathematical descriptors of political situations, as long as they are swallowed with a suitably large block of salt, can illuminate important and interesting trends.
Jordan Ellenberg in Slate magazine [1] recently called attention to work done by Keith Poole of the University of Houston and Howard Rosenthal of Princeton University who have tried to track the political polarization between the parties using roll call votes from 1879 to the present [2].
If we presume political affiliations are like grapes, people with similar views bunch together, we should be able to find a set of orthogonal axes in an ideological space revealing where Democrats and Republicans cluster. If the parties become more polarized, the distance between clusters of Democrats and Republicans in this space should increase. When polarization decreases, the clusters should begin to overlap. For example, one could find a few Republicans that are closer in their votes to the center of the Democratic cluster of votes than a few Democrats and visa versa.
Poole and Rosenthal found that the most explanatory geometric description of roll call votes rested on two dominant ideological axes. The first axis separating Democrats from Republicans was the traditional split based on belief on the appropriate extent of government involvement in the economy. On the left extreme of the axis would be economic Socialists and on the right extreme would be economic Libertarians. The second axis rested on differences in voting patterns on racial issues. While Republicans were generally on the side of more racial neutrality in government policies, Democrats through much of the twentieth century, were split along a North-South division. Northern Democrats and Republicans resembled each other in voting patterns racial matters, while Southern Democrats generally voted to maintain the social structures separating the races. As a consequence, the net polarization between the parties was smaller.
In recent years, racial issues have tended to show a less obvious divide, at least as represented on roll call votes. Some disputes on racial issues may have been subsumed into economic ones. The split between the parties has come to rest more squarely on issues of how involved the government is in the economy and along these issues, the parties have drifted further apart.
One explanation of this drift a part could be growing economic inequality. Economic inequality in the US was at is lowest point in 1968. Since then it has increased, though in recent years it has leveled off. There are a number of proposed explanations for the change in income distribution including greater rates of immigration increasing the number of low-paid workers, changes in the American social structure producing both more single-parent families and two income families, and the increased premium the economy puts on education. Whatever the cause, a recent paper by McCarty, Poole, and Rosenthal correlate political polarization and income inequality [3]. Although the correlation is not surprising, it is surprising that effect of income inequality on polarization is not has high as one might have expected. Perhaps those who are not as well off economically aspire to be so in the future. Social mobility decreases potential animosity between economic classes.
It remains to be seen, but it appears that attitudes on national security issues may come to split Democrats as much as racial issues did earlier in the century. Since World War II, Republicans have grown to be as hawkish on national security issues as Democrats used to be. Since Vietnam, Republicans have been more consistently hawkish on national security issues, especially since Patrick Buchanan has left the Republicans. However, the Democrats seem split between the Senator Joseph Lieberman-wing of the party who supported the Iraq War and the Vermont Governor Howard Dean-wing [4] wing of the party. Either the parties will become more polarized as Democrats trim away the Lieberman wing, or less polarized as the consensus on national security issues grow.
It will take a few election cycles to determine the direction the Democrats take. Poole and Rosenthal found that ideological distributions in the Congress and Senate did not change as a consequence of current elected officials changing the positions so much as by replacement of those elected representatives by elections.
- Ellenberg, J., “Growing apart,” Slate, December 26, 2001.
- Poole, K. and H. Rosenthal, Congress: A Political-Economic History of Roll Call Voting, New York, Oxford University Press, September 2000.
- McCarty, N., K. Poole, and H. Rosenthal, “Political Polarization and Income Inequality,” 27 January 2003.
- Alternatively called the “Democratic wing of the Democratic Party” or more derisively the “French wing of the Democratic Party.”