Network Clusters in Modern Politics
This article by the Pew Research Center, titled “The demographic trends shaping American politics in 2016 and beyond”, explains several trends behind why American politics are becoming more segmented, with groups of similar demographics voting in the same direction and adopting the same identities and values. These trends include the concept of “think-alike communities”, demographic transformations, the generation gap, and identity-based hyper partisanship. In the second article, “Behind Trump’s victory: Divisions by race, gender, education” the Pew Research Center summarizes the demographic differences in voting patterns in the 2016 Clinton-Trump election and show the effect that these trends have on the modern divide in politics. Some examples include the large proportion of black and Hispanic voters who voted for Hillary Clinton in contrast to the larger proportion of white voters who voted for Donald Trump; the gender gap in vote choice (women voted for Clinton and men voted for Trump); the education gap (college grads voted for Clinton and non-college grads voted for Trump); and the age gap (young adults voted for Clinton and older voters voted for Trump). Furthermore, during the day of the election, as the poll results were counted from each state, it was easily noticeable that large cities in majority-Republican states had voted for Clinton, while the rural areas favored Trump. All of these examples show clear disparities and gaps in demographics in voting patterns.
The voting patterns presented reflect the patterns we’ve learned about in individual decision-making in respect to one’s neighbors or strong ties. As the Pew Research Center said, Americans have been gravitating towards “think-alike communities”, in which both political viewpoints and demographics align among the members of the community. Demographically, we can separate groups based on those who are likely to interact with each other (forming a cluster), and connect each cluster with a few local bridges or link ties (as we saw before, it’s not likely for there to be two giant components in a network). This analysis is especially applicable in groups that are separated by ethnicity, age, and college education, groups that are less likely to interact with each other across clusters. By separating voters in the United States into these clusters, we can begin to understand why those of similar demographics are likely to vote in the same way. In these networks, although people of different clusters are exposed to limited information from the opposing political party, most of the information they receive is from their neighbors, people of similar demographic. Since adopting a political belief can be considered as adopting a risky viewpoint, we can assume that individuals will require a high threshold before changing or adopting a new belief. Therefore, since most of one’s close ties are within one’s demographic group, the payoff to adopt a new political belief is not high enough for the individual to do so, causing different clusters to adopt the same belief among all of their members. We can also look at this from a geographic perspective, in which each highly populated city and rural town represents a separate cluster, connected to each other by local bridges or weak ties. Especially in the rural towns, the spread of information is sparser than in cities, therefore, the people in the towns are expected to adopt the same belief as the others in the town.