Our recent study on the U.S. healthcare debate suggested that extreme voices on the issue were more influential than more moderate voices. Given these results, it may seem like the U.S. government has never been more polarized. But just how polarized is it? And what does it suggest about the stability of the U.S. government in general?
The news is better than you think, at least according to a new social network study from mathematicians at UC San Diego, Caltech, UNC Chapel Hill and University of Oxford. They recently used a concept called modularity to measure polarization in the United States Congress. Their results not only show that modularity is a significant predictor of party changes in the House and Senate, but that turnover is more prevalent at medium levels of modularity versus high.
What is modularity all about?
Modularity is a way to compare the strength of ties within a group of people to the strength of ties between each group. In the case of the U.S. Congress, the groups we’re interested in are Democrat and Republican (or Whig, or Jacksonian, depending on what century you’re in). As the ties within each group become stronger, the ties between them become tighter, or more “modular.” Conceptually, this is what we mean by polarization – in a polarized legislature, people stick staunchly to the party line.
What did they find?
The study used roll-call voting data to measure polarization in Congress for the 1st – 109th congress (1788-2004). Their values are consistent with several historical facts about polarization in Congress, for example, the spike in fractionalization at the turn of the 20th century caused by the end of Reconstruction. It also reflects the recent spike in fractionalization since the 95th congress.
High Polarization = A Sign of Stability?
The study also looked at the position of individual legislators within the group structure using measures of “divisiveness” and “solidarity.” That is, how central is the individual within their community and to what extent does their voting behavior align with the group’s voting behavior?
What’s interesting is that these individual results suggest that there are significant differences in party-attachment across different levels of polarization.
- In low-polarized Congresses, characterized by the politics of logrolling and poorly-organized leadership, legislators are expected to pursue constituency interests first and foremost. The result is a less stable environment.
- In partially polarized Congresses, where group structures exist but are either not fully established or are breaking down, legislators face a complex environment in which their choices are subject to greater error and greater risk. But this risk requires broader thinking, resulting in a more stable environment where the benefits and perils of alliance accrue to those in the largest community.
So what does this say about today’s environment?
Interestingly, the modularity measure suggests that modern-day polarization is high but not to a greater extent than in many other periods. In fact, low-polarization seems to be the exception rather than then norm. And if high polarization is a sign of greater stability, then perhaps the U.S. government is more stable than we thought.
In theory, modularity may be able to detect early warning signals of changing group dynamics. By comparing the modularity values of party divisions, we could potentially identify environmental and strategic conditions under which particular organizational concepts are likely to succeed and fail. But that’s left to future research. Stay tuned.
