Friday, August 17, 2012

The Lost Majority Part II: Political Leanings


Part I about Sean Trende's book The Lost Majority, is here, called "Future Up for Grabs."
To talk about some of the technical data behind recent elections, it will help to know some terminology.
Partisan Voting Index (PVI): This is the term referring to an area’s “lean” toward one party or another. Sean Trende uses the term throughout his book The Lost Majority. He says, “When the book refers to the “lean” of a state, county, or demographic group, it is referring to the PVI. In other words, a state or county that is said to “regularly lean Democrat” in a given set of years is one that regularly has a Democratic PVI.” There will be a number attached. D+6.7, for example, would mean that the designated area voted 6.7 percentage points more Democratic than the nation as a whole.
DW-NOMINATE: This is a term used for measuring the ideology of a legislator. Trende quotes Jay Cost’s description: “DW-Nominate is a very complex methodology that produces a very simple result. Legislators are given an ideological score that ranges from -1.0 to 1.0—with -1.0 being extremely liberal, 0.0 being moderate, and 1.0 being extremely conservative (p. xxx).” The range isn’t usually at the extremes. Trende says that 3.0, for example, is considered quite conservative. Hardly anyone scores beyond 0.5 or -0.5, so that’s where you might find the most extreme legislators. Scores are relative; that is, they are in comparison to other legislators in the same congress, not comparing to particular stands on particular issues.
So, with these two concepts in mind, it might be helpful to look at what went on in 2008 and 2010. This is from the introduction (p. xix):
Democrats who celebrated electing arguably the most liberal president in American history in 2008 frequently overlooked that Barack Obama was running amid two unpopular wars, a nasty recession, and a full-blown financial panic that was consuming 401(k)s and housing equity—not to mention that his opponent was a disorganized, gaffe-prone candidate. In a year when political science models suggested double-digit Democratic wins, and when every conceivable intangible suggested that the Democratic wins should be on the high side of those models’ error margins, Obama won by only seven points and was actually trailing in the polls after the convention season closed. As the Democratic Party enjoyed more success, its elites had pulled it leftward, and in doing so had exposed their right flank. The leftward shift didn’t exact immediately obvious costs in 2008, but the price was steep in 2010.
Trende goes into more detail on this situation in the middle of the book. While Clinton’s coalition had been broad, had been able to turn back some of the coalition components that had turned Republican with the Eisenhower coalition, Obama lost back those areas. Obama’s coalition was “narrow but deep.” And none of it was new. Very little PVI change came with Obama’s election:
Between 2004 and 2008, only three states saw their partisan lean, or PVI, switch. Colorado had leaned one point toward the Republicans in 2004; it leaned one-quarter of one point toward the Democrats in 2008. Nevada had leaned six-hundredths of a point toward Republicans in 2004; it leaned two points toward the Democrats in 2008. And Ohio had leaned two-tenths of a point toward the Democrats in 2004; it leaned two points toward the Republicans in 2008. Only seven states saw their PVI shift more than five points in either direction: Hawaii moved 13 points toward Obama, while Arkansas, Louisiana, Oklahoma, Tennessee, Alaska, and West Virginia all moved toward the Republicans. If realignments do exist, this was not it. It did not resemble 1932, when Roosevelt completely remade the political map and 16 states leaned in a different direction than they had in 1928. Nor was it like 1952, when 13 states changed their partisan orientation. To put it differently, the average change in PVI from 1928 to 1932 was eight points, and from 1948 to 1952 it was six points. From 2004 to 2008 it was three (p. 92).
To summarize, in a year when Obama had supposedly remade the map, only three states switched their lean, and one of those was toward the GOP. Of the seven states that made a significant PVI shift, only one shifted toward Obama; the other six shifted toward the GOP—in the year of Obama’s “historic” win.
According to Trende, when a change of direction happens, it’s going to happen where the differences were small to begin with. And those areas, the ones most recently changed, are also the most vulnerable to being changed back. There was a fair amount of moving into typical GOP territory in 2006, and some more in 2008. Unless the direction of the country actually trended more liberal (and signs are that it didn’t), then those areas were ripe in 2010 to turn back. As Trende explains:
[W]henever a party occupies a large number of seats in the House, especially when a large percentage of those seats are occupied by freshmen and sophomores, that party is probably going to suffer in the subsequent elections. The reason is simple: as a party picks up more seats, it pushes further into marginal territory. If it picks up a huge number of seats, as Democrats had done in 2006 and 2008, it by definition extends itself into hostile territory. These members are therefore vulnerable, and the party is set up for losses (pp. 172-173).
Democrats, by gerrymandering to protect some particular urban areas from ever going Republican, have allowed for more districts to lean GOP than Democrat:
Due to gerrymandering and Democratic packing of minorities and liberals into a handful of overwhelmingly Democratic districts, the median district in the United States leans toward the Republicans by a few points. Put differently, the most Republican district in the nation leans Republican by 29 points. Democrats have created 25 districts—almost 15 percent of their caucus—that are more heavily Democratic than this. That means that Democrats cannot win a House majority without capturing a substantial number of seats that naturally favor Republicans. In 2010, Democrats occupied 73 seats that leaned toward Republicans. As expected, many of these seats were, in fact, occupied by newly elected members of Congress, who had picked up swing seats in the 2006 and 2008 elections. This is also where Democratic losses were concentrated (p. 173).
While the need for Democrats to appeal to Republicans is greater, the response by Democratic leadership is instead more polarized. Here’s where the DW-NOMINATE term comes in:
Perhaps most important, the Democratic leadership in Congress represented some of the most liberal districts in the country. This put a distinctly liberal imprint on legislation. Speaker Nancy Pelosi’s district leans 35 points toward the Democrats, while the chairmen of the critical districts charged with developing domestic policy—Appropriations, Education and the Workforce, Energy and Commerce, Financial Services, Judiciary, Rules, and Ways and Means—reported back to districts that leaned on average 21 points toward the Democrats. They share an average DW-NOMINATE score of -.552, substantially to the left of even this historically liberal Democratic caucus. With these members responsible for crafting most legislation, it was highly unlikely that a centrist agenda would emerge.
            Virtually every assumption regarding the nature of the Democratic Party that had given rise to the Clinton coalition was eroded in 2009 and 2010 (p. 126).  

What I’d like to say is that there is no way on earth that the Democrats can win the presidency with this incumbent, under current circumstances. Trende doesn’t go that far. But there is more data to look at, to help us in our speculation on what can happen, depending on various ifs and thens. So we’ll need a Part III for prognostication.

No comments:

Post a Comment