Matt Yglesias

In Retrospect…

Nate Silver says it’s too soon to tell if we’ve just witnessed a “realigning” election:

Since the turn of the last century, there have been 11 cases in which the presidency changed parties: 1912 (Wilson), 1920 (Harding), 1932 (Roosevelt), 1952 (Eisenhower), 1960 (Kennedy), 1968 (Nixon), 1976 (Carter), 1980 (Reagan), 1992 (Clinton), 2000 (Bush), and 2008 (Obama). In 9 of the 11 cases, the party winning the presidency had also made substantial gains in the Congress as compared with four years’ earlier (although not necessarily as compared with two years’ earlier). The two exceptions were the last two party changes before Obama: Clinton in 1992, when the Democrats were pretty much treading water in the Congress, and Bush in 2000, when the Republicans were doing likewise.

What ultimately distinguishes the elections that are considered to have been realignments is the efficacy of the governance of the rising party, rather than the force with which said party took office. Ronald Reagan and FDR, famously, had coattails — but so did Warren G. Harding, who brought the Republicans a net gain of 123 (!) seats in the House in 1920.

I think the better way to put the point is simply to agree with David Mayhew that there isn’t really any such thing as a “realigning election.” Or, to put it in perhaps a more Rortian way, to say that questions about realignments aren’t useful questions to ask. If presidential election outcomes were completely random the odds would still overwhelmingly favor the emergence of Democratic and Republican “clusters” and you could choose to interpret the beginning election of some cluster as a “realigning” random outcome. But why would you want to?

Of course election outcomes aren’t random. But they’re determined largely by events and, needless to say, the events of 2010, 2012, 2016 and so forth haven’t happened yet. Whether big Democratic wins in 2006 and 2008 will be followed by consolidation of power or backlash depends entirely on what happens in the future and not at all on factors you can discern by peering into the election results.

2008 Polls Versus 2004 Polls

Something funny happened in 2004 where a lot of progressives convinced themselves near the end that John Kerry was likely to win the election even though he was narrowly behind in the polls. Then a lot of people have gone and misremembered that as thinking that Kerry was likely to win because he was ahead in the polls, which he wasn’t. Thus, many are left unable to believe that Obama’s lead in the polls makes his victory likely.

Fortunately, Chris Bowers has a comparison chart we can believe in:

As you can see, Kerry is running well ahead of where Obama was. The theory behind “Kerry’s gonna win” was that undecideds were likely to break heavily for the challenger — heavily enough to make up his small gap in the polls. That wasn’t a crazy thing to believe, but it clearly had a whiff of wishful thinking about it. At the moment, though, Obama just has a big lead.

Hasen: Let’s Run Elections Properly

Rick Hasen has the correct proposal that would let us avoid these every-two-years allegations of voter fraud and suppression — do voter registration like other modern democracies:

The solution is to take the job of voter registration for federal elections out of the hands of third parties (and out of the hands of the counties and states) and give it to the federal government. The Constitution grants Congress wide authority over congressional elections. The next president should propose legislation to have the Census Bureau, when it conducts the 2010 census, also register all eligible voters who wish to be registered for future federal elections. High-school seniors could be signed up as well so that they would be registered to vote on their 18th birthday. When people submit change-of-address cards to the post office, election officials would also change their registration information.

This change would eliminate most voter registration fraud. Government employees would not have an incentive to pad registration lists with additional people in order to keep their jobs. The system would also eliminate the need for matches between state databases, a problem that has proved so troublesome because of the bad quality of the data. The federal government could assign each person a unique voter-identification number, which would remain the same regardless of where the voter moves. The unique ID would prevent people from voting in two jurisdictions, such as snowbirds who might be tempted to vote in Florida and New York. States would not have to use the system for their state and local elections, but most would choose to do so because of the cost savings.

Since some people drop out of high school, you would need some other entry points. But that’s the basic shape of the system — a nationally administered registration scheme that keeps track of what’s going on, instead of a system that relies on “registration drives” to get people signed up, and a whole bunch of separate authorities to keep track of whether they’re signed up legitimacy.

Counties to Watch

The sophisticated election-watchers can tell a lot about the eventual outcome in a given state by paying attention to certain key counties as leading indicators. I’m not that sophisticated myself, but Marc Ambinder lays some of it out for us.

Eagelburger’s Walkback

I wonder what they More »« Less to get him to retract his Obama endorsement? It would seem very strange for a seasoned veteran of high-level politics and policy to accidentally endorse the wrong presidential candidate.

Foiled by the Blue Dogs

The other day, Eve Fairbanks wrote:

Yes, the Democrats are poised to expand their House majority — but by electing conservative Democrats who, in some cases, have ideologically more in common with John McCain than with Nancy Pelosi. These conservative Democrats — many of whose districts will vote McCain — won’t feel that they owe Obama, will be well-organized as a faction under the “Blue Dog” banner, and, if their actions in the 110th are any indicator, won’t shirk from bucking their party’s leadership.

There’s an important element of truth to this. But one should recognize that it’s not as true as some people seem to think. Substantial Democratic House gains will make for a much more progressive House of Representatives even if the bulk of the new members are relatively conservative Blue Dogs. Note for one thing that according to DW-NOMINATE the parties are currently perfectly sorted and the most conservative Democrat (Rep. Barrow of Georgia) is more progressive than the most liberal Republican (Rep. Gilchrest of Maryland). So replacing any number of Republicans with Barrow clones would still make the House more progressive. And of course not all the new members will be Blue Dogs. In 2006, the conservative-to-moderate Democrats who picked up seats in conservative-to-moderate districts got all the press, but you also had several examples of liberals beating moderate Republicans in moderate districts and that will happen again in 2008.

Beyond that, at the moment the 219th most liberal member of congress — the one who, generically, would be the last vote for progressive legislation — is Rep. Tanner of Tennessee. Tanner is a Blue Dog. And Tanner is also quite conservative. But “more Blue Dogs” and “More Tanners” are not equivalent. At the moment, only 16 Democratic Representatives (of which twelve are Blue Dogs) are to the right of Tanner. By contrast, there are 29 members of the Blue Dog caucus to the left of Tanner. And every new House member, whether Blue Dog or otherwise, who’s to the left of Tanner is pushing the median member to the left.

Senator Who?

Not saying these folks are all going to win, but odds are that out of the set {Mark Begich, Ronnie Musgrove, Kay Hagan, Jim Martin} there are going to be some new Senators come January. Does anyone know anything about these people? I’ve poked around their websites and there’s some interesting stuff there, but it’s a limited source of information — any good profiles? I feel like Al Franken and Mark Warner have been covered to death, and I know Jeanne Shaheen’s record from New Hampshire, but these other characters . . . nothing.

Registration Gap in Washoe

The basic way that politics works in Nevada is that Clark County — i.e., Las Vegas and its suburbs — is Democratic and has a lot of people in it. Then Washoe County — i.e., Reno and its suburbs, plus some rural areas — is politically competitive and has some people in it. Then the rest of the state is very sparsely population and strongly Republican. Normally, though, Washoe breaks for the Republicans and thus Nevada is a Republican-leaning state. But Democrats are narrowing the registration edge in Washoe “county Voter Registrar Dan Burk said Republicans had 87,971 registered and Democrats 84,705, with a backlog of more than 5,000 registration applications awaiting processing. Since registration for the primary closed, the Republican edge has shrunk from 5,648 to 3,266.”