Truths and Myths about the 2008 Election, Part IV

by John Sides on November 6, 2008 · 1 comment

in Campaigns and elections

4. THERE IS NO REALIGNMENT
This notion permeates post-election analysis. For example, there is the John Judis piece I noted earlier. This New York Times piece says: “In a striking measure of potential political realignment, the gap between self-described Democrats and Republicans has grown.” Since Election Day, 29 different news stories contain the words “realignment” and “election.”

The notion of “realignment” has become fairly amorphous. At times, journalists and commentators use the term simply to mean “something big has happened.” This Boston Globe piece refers to “a profound political and social realignment in America.” At other times, it seems to mean “the winning candidate’s supporters will form a permanent majority coalition.” Sometimes this new coalition is discussed in terms of territory. This Washington Post editorial refers to a “geographic realignment.” Some times it is discussed in terms of groups of voters. The New York Times piece mentioned above states: “The exit poll…suggested that Mr. Obama had won over many of the groups who had been aligned with the Republicans in recent years, including suburban voters, parents and voters in the Midwest.”

In political science, the term means this. For a realignment to occur, there has to be a dramatic and permanent shift in the party coalitions. That shift then ushers in an extended period of party control, which in turn brings with it a notable shift in policy.

However, the concept of realignment isn’t in such good standing anymore. As David Mayhew notes, nailing down evidence of these shifts is quite hard. In the post-war era, party control of government has oscillated back and forth. See more here in Phil’s post from 2005. And note the context of his post: people were portraying the 2004 election as auguring a Republican realignment. That gives you a sense of how casually this idea springs forth, no matter how narrow the winner’s margin of victory.

More importantly, there is little evidence that the 2008 election constitutes a realignment. Why?

  • First, even if it did, we wouldn’t know for a while. Years, in fact. As Andrew Busch writes here, “No one can tell whether a particular election is a realigning election until the long-term arrives and one can look back.”
  • Second, things didn’t change that much. This is why I posted the maps below. Yes, Obama won states that Democrats hadn’t won in a while, and perhaps demographic trends suggest Democrats will have a chance to win those states in the future. But these small shifts in state-level vote margins don’t signal any wholesale change in partisan loyalties or party coalitions.
  • Indeed, if you look at the exit polls (see Phil’s earlier post), Obama does better than Kerry among most every demographic. His vote share among the young and Latinos stands out, but the results don’t suggest a reordering of the party’s coalitions. Instead, it looks more like voters of all stripes were displeased with the economy and President Bush and so voted for the opposing party’s nominee.
  • The final nail in the coffin is this analysis from Larry Bartels. He compares the 2008 results in each state to the 2004 results. He finds remarkable continuity across these two elections. His further comparison of 2008 to 1932, where you in fact do see big shifts, is further evidence.

There is a natural tendency to want to make elections seem like harbingers of big changes, as Lee has noted. I do believe that the election of an African-American president is historic and important in many ways. But it doesn’t necessarily portend a realignment. Nor, in fact, does it need to in order to be historic.

The Obama team’s take? Yesterday, his pollster said: “I think it’s too soon to talk about realignment.”

{ 1 comment }

Hans Noel November 7, 2008 at 9:48 am

It’s definitely true that this is no realignment yet, and probably isn’t one at all. But the pundits can perhaps be forgiven for not knowing what a realignment is, since we don’t really know either.

Even John’s definition has a lot of wiggle room. He writes

“For a realignment to occur, there has to be a dramatic and permanent shift in the party coalitions. …”

How should we measure what the party coalitions are? By who votes for them? By who identifies? By who is in the governing coalition? By what interests are aligned with them? By their platform differences? Activist preferences? I think we usually mean voters here, but that’s potentially just outright inaccurate. On many policy changes (the New Deal, civil rights, abortion), the voters are a lagging indicator.

“… That shift then ushers in an extended period of party control, …”

This is usually part of our definition, but it’s orthogonal to the concept. If, say, half the Democratic voters changed places with half the Republican voters, this would be a change in the alignment of voters. But it would not lead to a change in the dominance of any party.

“… which in turn brings with it a notable shift in policy. …”

Assuming there is a change in dominance, and assuming the parties differ at all, this could happen. But does it follow from the voters, or did it precede them? If the voters are not responsible for the change, then no election is a critical realigning election.

In other words, maybe the problem is that “realignment” is a sloppy concept, even in political science. I’m usually not very tolerant of journalists not getting it, but this confusion might be our fault.

Comments on this entry are closed.

Previous post:

Next post: