Has the Republican Party entered an unusually fractious era? That’s the conclusion many are drawing from Rep. John Boehner’s abrupt resignation as Speaker of the House of Representatives, Kevin McCarthy’s equally surprising decision to pull out of the race to replace him, and the tortured negotiations in which Paul Ryan engaged with various factions of the GOP caucus prior to his election as Speaker on Thursday.
As Nate Silver notes at FiveThirtyEight, the Republicans’ difficult fight over who should sit in the speaker’s chair joins other indicators — including the rise of the tea party and two chaotic presidential nomination contests in a row — to make it seem like the party is in disarray.
Common sense suggests that an ideologically diverse party should be more difficult to lead, and political science research agrees. As Hans Noel explains over at Mischiefs of Faction, under the political science theory of “conditional party government,” lawmakers drawn from a heterogeneous party are much more reluctant to grant their leaders power over the agenda than those from a more homogeneous party. Hence, a diverse flock of Congressional sheep can be more difficult to keep in line than a unified one.
But a look at the data suggest that, in fact, the Congressional GOP is no less unified now than at any point since World War II. Political scientists at Voteview.com have archived every Congressional roll-call vote and have employed a data algorithm called “DW-NOMINATE” that uses these votes to place each member of Congress on the primary axis of conflict in American politics at the time. These days, of course, the primary conflict is between liberals and conservatives, and DW-NOMINATE assigns each lawmaker a score on this liberal-conservative scale.
Thus one way to get a sense of the level of internal division among a party’s members of Congress is to examine the extent that scores vary internally by party. As anyone who took (and remembers) AP Stats can tell you, a good measure of a set of numbers’ variation is their standard deviation. (Bear with me if you’re not statistically inclined; this will be over soon.) A party whose roll-call scores exhibit a low standard deviation is relatively unified; higher standard deviations indicate more heterogeneity and (potentially) within-party fractiousness. Of course, these scores capture just one important aspect of intra-party conflict in Congress; there are many others. But they are a good place to start if we are trying to objectively measure within-party polarization, rather than simply reach conclusions with anecdotal evidence.
In the graph above, I plot the within-party standard deviation of DW-NOMINATE scores among Democratic and Republican members of the House of Representatives from the 80th Congress (elected in 1946) to the present. The steady descent in the variation of Democrats’ scores is unmistakable. This of course is largely due to the fact that the party’s members from the South, who were Democrats in name but conservatives in practice, have been replaced by even more conservative Republicans. What was once a relatively heterogeneous Democratic group is now a more unified (and liberal) party than at any time in recent history.
More remarkable is the essentially flat line tracing the pattern of variation among Republican members of Congress over time. These data say the GOP is no more internally polarized now than it’s been at any time since WWII. The party is certainly less unified than the Democrats: The difference between the two parties’ internal variation now is about the same as it was in the mid-1950s, when the Democrats were the more heterogeneous party.
But compared historically, the Republicans in Congress are not an unusually diverse party.
Those seeking a reason for why the GOP appears to have so much trouble governing lately will need to look beyond the explanation that its members are more internally polarized than usual.