Based upon its most recent Values Survey, the Pew Research Center reports that Americans values have remained fairly stable for the past twenty five years but Republicans and Democrats hardly seem to agree. The figure below, taken from the report, shows the over-time trend in the difference between Democrats’ and Republicans’ responses to twelve values questions over the period 1987-2012. The dramatic upward trend in almost all the plots suggests that Democrats and Republicans are polarizing.
Polarization – the trend by which two groups move in opposite directions on a scale, becoming more extreme over time – is a defining feature of current political conversation. In order for polarization to occur, individuals have to develop more extreme views over the course of their lives or new generations have to hold more extreme and divided political viewpoints than earlier generations, or both. This means that Republicans have to become more conservative and Democrats have to become more liberal, or the people who identify as Republican have to be more conservative than people in earlier periods that identified as Republican (and the same pattern with Democrats).
Both these possibilities present psychological explanations for political polarization, but those explanations are quite different.
The first says something about socialization and development over one’s lifecourse. Unfortunately, polarization between Democrats and Republicans is unlikely to be due entirely to individuals changing their views as they age. Donald Green (previously of Yale, now at Columbia) and coauthors demonstrated in their book, Partisan Hearts and Minds, that individuals’ identifications with a political party are incredibly stable. Similarly, now classical work on political socialization conducted by Kent Jennings and Richard Niemi (University of Rochester) has shown that political views evolve gradually over the lifecourse. Changes in individuals’ views is unlikely to explain polarization.
The second possibility – that individuals who identify as Republican and individuals who identify as Democrats today are different from those who have identified with the two parties in the past – seems far more likely. How we identify ourselves is a product of social, familial, and political context. And, given that the American political landscape is limited to about two political parties, those alternatives are imperfect summaries of every person’s political views. As a result, a party’s constituency is going to be somewhat politically diverse.
Research by Matthew Levendusky (Penn) shows that over the last few decades, however, it appears that people have either been more likely to “sort” into a party, such that conservatives are more likely to call themselves Republicans and liberals call themselves Democrats. This means that people have not individually become more extreme on a liberal-conservative dimension nor have individuals shifted from one party to another. Instead, the parties’ positions and those of voters have gradually aligned, so that Republicans appear to be more extreme because their party is more homogeneously conservative and Democrats appear to be more extreme because their party is more homogeneously liberal.
The Pew report makes a big deal out of the large and apparently increasing differences between Democrats and Republicans in terms of core political values. In light of the sorting process identified by Levendusky, it seems that these differences are largely spurious – not a consequence of meaningful changes in public views, but an artifact of how new voters have come to translate their political views into a party identification. The parties are clearly more distinct, more extreme, and more polarized than they have been during most of our lifetimes, but the electorate itself hasn’t changed much at all.