Skip to main content
ABC News
Gaming Senate Fights in Connecticut and North Dakota

‘Tis the season for retirements from the United States Senate.

Less than a week after the Republican Kay Bailey Hutchison of Texas announced that she would not seek re-election in 2012, a Democrat, Kent Conrad of North Dakota, who has represented the state since 1987, said today that he wouldn’t seek a new term either. A third senator, Connecticut’s Joe Lieberman, will announce his plans tomorrow, but The New York Times and other outlets are reporting that he is likely to retire as well.

My focus, naturally, is in figuring out who their successors might be. We discussed Hutchison’s case last week, so let’s look at North Dakota and Connecticut now.

Getting the lay of the land in an open race for a United States Senate seat is not terribly complicated. We essentially want to look at three factors: the national partisan environment, the local partisan environment, and the quality of the potential candidates.

Any guess as to what the national environment will look like in November 2012 would be highly speculative. Yes, Barack Obama’s approval ratings are up slightly. But historically, there has been almost no correlation between what a President’s standing looks like at this stage in the political cycle, and what it will look like the following November. Almost no pollsters, meanwhile (other than Rasmussen Reports, which has some kinks to work out) have yet issued generic ballot polls.

We can perhaps come to some conclusions, however, about factors more specific to North Dakota and Connecticut, respectively.

North Dakota

North Dakota has a Partisan Voting Index (PVI) of R+10, which means that in an average year the Republican candidate for president will win 10 percent more of its vote there than he would in the country as a whole. So, for instance, in a year in which the Democratic and Republican candidates each received 50 percent of the vote nationwide, North Dakota would give the Republican 60 percent, and the Democrat 40 percent, all else being equal.

That’s a powerful advantage for Republicans: what amounts to a 20-point head start given average candidates in an average year. In fact, even in a good Democratic year, in which Democrats led the generic ballot by 10 points nationally, the Republican would still probably be favored.

This indicator, Partisan Voting Index, is based on presidential voting: is there any reason to conclude that voting for the Congress is systematically different? I looked at this in some detail last year and came to the conclusion that the only demographic variable that is of much help in addition to PVI in predicting the outcome of Congressional races is income: poorer states and districts tend to vote a little bit more Democratic for Congress races than they do for the Presidency, and vice versa. North Dakota is slightly poorer than average, so this helps Democrats a bit (under the alternative version of PVI that I use, which I call Partisan Propensity Index, it rates as more like a R+9 than an R+10). Still, this is a minor difference and it remains strongly Republican territory.

But systematic differences — those which you might divine by building a statistical model — are one thing: there could also be localized differences that are peculiar to one state or region. North Dakota, for instance, despite having voted Republican for President every year since 1964, has frequently elected Democrats to the Congress (the same is true for South Dakota). Is there something particular about the state’s political culture that begets such outcomes?

I am skeptical about this, because there is not much of a tradition — as there is in states like Arkansas or West Virginia, for instance — of North Dakotans voting Democratic at the statewide level. Its state senate has been in Republican hands since 1995, and in all but 12 years since the Second World War. Likewise, its state house has been Republican in all but four years over the same time period. And Republicans outnumber Democrats in the state — by a margin of 38-28, according to exit polls conducted in 2008.

If North Dakotans vote Republican for president, and Republican for state and local offices, then it might be something of a fluke that they’ve often voted Democratic for Congress. The state, with its economy dependent on agriculture, might sometimes have different priorities than the rest of the country (particularly since agricultural cycles do not always mirror those in the broader economy): that might contribute to uncertainty and the occasional idiosyncratic result. But particularly given that Republicans swept the state in 2010 — disposing of its Democratic representative, Earl Pomeroy, and electing a Republican, John Hoeven, to the Senate seat vacated by the retiring Democrat Byron Dorgan, I am not sure that there is any sort of below-the-radar advantage for Democrats there, especially in an open-seat race.

That does not mean Democrats will concede the state — nor, necessarily, should they, given that they already face a difficult map for retaining control of the Senate after 2012, with about twice as many of their seats up for election. But the person who would seem to be their strongest candidate, Mr. Pomeroy, has taken a job as a lobbyist and does not seem interested in running, so they may lack a top-tier candidate.

Republicans, meanwhile, are likely to have a competitive primary fought between candidates like former Governor Ed Schafer and Lieutenant Governor Drew Wrigley. The danger to them might be in nominating a too-conservative candidate rather than one of these establishment alternatives. Although North Dakota is Republican, it also has some tradition of Lutheran moderation, and its elderly population and vulnerability to agricultural cycles make it relatively sympathetic to safety-net programs like Social Security. It’s also a highly educated state with high rates of political participation, so winning candidates there need to appeal to the broader spectum of the electorate. If Republicans were to nominate a Tea Party candidate that resembled Kentucky’s Rand Paul, they might squander their advantage there.


Connecticut has a PVI of D+7, making it slightly less Democratic than North Dakota is Republican. It also has a very wealthy population, meaning that it might vote slightly more Republican in a race for the Congress than it would for the presidency.

Still, Democrats held all five of its House seats in November, as well as winning the race for the Senate seat formerly occupied by Chris Dodd, and the governorship being vacated by the retiring Jodi Rell. And the last Republican the state elected to the U.S. Senate was Lowell P. Weicker, Jr., who won his race in 1982.

The scariest possibility for Democrats would be if Ms. Rell decided to run for the seat; although her approval ratings have fallen some from their peak, they were nevertheless 59 percent in the most recent Quinnipiac poll, conducted in October. But having declined to run for either another term as governor or for Mr. Dodd’s seat in 2010, Ms. Rell is probably not a favorite to run in 2012.

The Republicans might also have something of a downside case in the form of Linda McMahon, the former World Wrestling Entertainment chief executive who was their nominee for Senate in 2010. With Ms. McMahon having spent tens of millions of dollars on advertising but nevertheless losing her race by 12 points in a strong Republican year (albeit to the fairly popular Democrat Richard Blumethal), Republicans should have some questions about her electability. Still, she is thought to be a potential candidate, and would have a lot of money to spend in the primary against other potential candidates like Tom Foley, the former ambassador to Ireland who closed strongly in the gubernatorial race last year.

Democrats, meanwhile, have a fairly deep bench in Connecticut, including several candidates who have been preparing to run for some time on the assumption that Mr. Lieberman might run as an independent, or could lose a Democratic primary. Susan Bysiewicz, until recently Connecticut’s Secretary of State, has already declared for the race, and at least one of the state’s U.S. Representatives — most likely Chris Murphy, who represents the 5th district — is likely to run as well. The primary will be hard-fought — and annoying to New York City television viewers who don’t like seeing political commercials from neighboring Connecticut creeping into their broadcasts of the Yankees game — but Democrats should emerge with a competent candidate.

So no real surprises here: if I had to set an early betting line, I would make the Democrats something like 3-to-1 favorites in Connecticut, and the Republicans 4-to-1 favorites in North Dakota. In both cases, the Republican primaries are worth watching: nominating Ms. McMahon in Connecticut could make the state less competitive, while a Tea Party candidate in North Dakota would make it significantly more so.

Nate Silver founded and was the editor in chief of FiveThirtyEight.