Will centrists be the big winners of Tuesday’s elections? Some commentators suggest so, predicting that the results will compel both parties – particularly the Republicans – to move towards the center. This view contends that a Romney loss, which would be the Republicans’ second presidential loss in a row, would force a Republican version of the Democratic Leadership Council (DLC) to moderate the party as Bill Clinton and the DLC did after Democratic losses in the 1980s.
There are some problems with this argument. First, we know from the work of Alvarez and Nagler (1995) that voters in 1992 did not perceive Clinton to be any more moderate than Michael Dukakis four years earlier. This casts doubt on the conventional wisdom that the Reagan-Bush blowouts of 1980, 1984, and 1988 led the national Democratic Party to do some ideological soul-searching and moderate its party platform, which allowed it to regain the presidency in 1992. Indeed, if the national Democratic Party changed at all during this period, it was slightly towards the left as a result of the loss of Southern “Blue Dog” Democrats.
Second, our research suggests that political polarization is a long-term trend in American polarization – one that has continued to expand since the 1970s. Of course, there is an upper limit to how much the parties can polarize along the ideological spectrum; and given that the distance between the parties is now as large as any time since Reconstruction, we may be approaching that limit. However, it is extremely unlikely that the results of one election will dramatically reverse this trend. Consider, for example, the ideological positions of House incumbents running for re-election on Tuesday, which we plot below.
Using the Cook Political Report’s House race ratings to separately plot incumbents in safe and close races. We treat incumbents in races rated “Solid” and “Likely” as safe, while those in “Leaning” and “Tossup” races as vulnerable. The first dimension represents ideological position on the liberal-conservative spectrum, using members’ first dimension DW-NOMINATE scores. Safe incumbents are plotted in dark blue (Democrats) and dark red (Republicans), while vulnerable incumbents and shown in light blue and red.
As can be seen, House incumbents will not be breaking much of a sweat on Tuesday; but more importantly, there is little difference between safe and vulnerable incumbents in both parties. Safe incumbents as a group are more ideologically extreme than vulnerable incumbents, though this difference is much larger among Democrats (with a mean first dimension DW-NOMINATE score of -0.42 for safe Democrats versus -0.27 for vulnerable Democrats) than Republicans (with a mean score of 0.67 for safe Republicans versus 0.65 for Republicans in close races). Of course, this is not wholly surprising, since centrist members are more likely to represent swing districts.
However, even if a large number of vulnerable incumbents are defeated and replaced with centrists, it is also clear that many of the large number (322) of House incumbents running in races rated “Solid” or “Likely” by Cook are also strong liberals or conservatives. Since members of Congress retain fairly static ideological positions throughout their careers, the parties in next (113th) House are likely to remain quite polarized. Ultimately, the decades-long trend of disappearing moderates in Congress is likely to continue in the next session (see the plot below). Indeed, the polarization measure (the difference in the mean DW-NOMINATE scores of the Democratic and Republican caucuses) between the safe incumbents is slightly larger (1.089) than the existing level of polarization in the 112th House (1.071).
We have also previously argued that polarization is likely to slightly expand (or at least not reverse) in the Senate. In sum, If the most likely electoral outcome prevails – Republicans retain control of the House, and Democrats retain control of the Presidency and the Senate – gridlock is likely to continue.
In the plot below, we exaggerate the proportion of the 43 incumbents in safe races so that is is easier to see the distribution of these members along the ideological spectrum.