G.O.P. Poised to Control 30 Governor Seats
September 3, 2010, 5:09 pm
<!– — Updated: 5:14 pm –>By NATE SILVER
Republicans are on track to control approximately 30 governor seats after the Nov. 2 election, according to the FiveThirtyEight gubernatorial forecasting model. And they are likely to do particularly well in the swing states of the Midwest.
Such an outcome would reverse the current state of the nation’s governors’ mansions, which are now held by 26 Democrats, 23 Republicans and 1 independent.
Thirty-seven states are holding elections for governor this year; an unusually high number, 24, are open-seat races in which the incumbent was either barred from running again because of term limits, or chose to retire. (One incumbent, Jim Gibbons of Nevada, was defeated in the Republican primary.) Many of these open races, including California, Florida, Oregon and Wisconsin, are too close to call. But in 11 of these contests, including Michigan and Pennsylvania, the Republican candidate has at least an 80 percent chance of victory, according to the model. Conversely, Democrats are clear favorites in only four open-seat races — Colorado, Connecticut, Hawaii and New York, where Andrew Cuomo is almost certain to become the next governor.
Incumbency -– normally a powerful advantage in gubernatorial elections -– may provide few protections for Democrats this year. This is especially so in three Midwestern states. In Ohio, Gov. Ted Strickland trails his Republican opponent, former Representative John Kasich, by double-digits in some polls, and with the state’s unemployment rate also in the double digits (10.4 percent), he is running out of time to catch up. Although Iowa, like some other agricultural states, has weathered the recession relatively well, the Democratic incumbent there, Chet Culver, has had poor approval ratings for some time and has run into a challenging opponent in Terry Branstad, who served as the state’s governor from 1983 to 1999. And in Illinois, the Rod R. Blagojevich scandal is still casting a shadow on his successor, Pat Quinn, a Democrat, and seems to be outweighing any influence that President Obama might have there.
Democrats are also underdogs to knock off any of the six Republican governors who are running for re-election, although they retain a fighting chance in Texas, where they have a strong candidate in Bill White, the soft-spoken former mayor of Houston.
In most election cycles, governors’ races are predominately local affairs, and can sometimes produce counterintuitive results: there are Democratic governors in Oklahoma and Wyoming, for instance, but Republican ones in Hawaii and Vermont. (All four of the incumbents in these states are retiring.) And, because of the large number of retirements, many analysts had warned of a difficult year for Democrats in gubernatorial races long before the national mood soured on their governance.
When one party has a sufficiently strong political wind at its back, however, gubernatorial elections sometimes get caught up in the wave. In 1994, for example, Republicans claimed 12 governorships from Democrats, while the only Democratic pickup was in Alaska, where Walter Hickel, then of the Alaskan Independence Party, was retiring. And in 2006, Democrats won six Republican-held seats, while retaining all of their own.
Should the political momentum favor the Republicans this year, as seems likely, they have picked an auspicious time for it. This year’s gubernatorial elections are unusually important because of the role that some governors play in redistricting, the process of dividing the nation into 435 Congressional districts that occurs after each Census. Redistricting is often a contentious process, but it is especially so in states that are bound to lose seats in Congress. The result can resemble a game of musical chairs, with too many incumbents vying for fewer districts. Among the states expected to lose seats are Iowa, Michigan, Ohio and Pennsylvania – all states where the Republican nominee has emerged as a clear favorite.
Gubernatorial races, especially open races, are often quite dynamic until the last hours of the campaign. Thus, most Democrats who are now trailing have a chance to come back. In Michigan, for instance, where Virg Bernero, a Democrat and the mayor of Lansing, is running against the venture capitalist Rick Snyder, the number of undecided voters is high and the race is likely to tighten, although perhaps not enough for Mr. Bernero to secure victory.
Nevertheless, both the math and the clock are working against the Democrats, and even if the Republicans do not do win 30 seats, they have a 95 percent chance of controlling at least an outright majority of governor’s mansions, according to the model.
This is the debut of FiveThirtyEight’s gubernatorial forecasts, which will be updated approximately once a week until the elections. The methodology behind them is similar to FiveThirtyEight’s Senate forecasting model, in that they are informed by a rigorous statistical analysis of election outcomes since 1998. However, the gubernatorial model also contains a number of important differences: economic variables like unemployment, which have greater predictive power in gubernatorial races than in Congressional races, play a more explicit role, for instance. In addition, gubernatorial races have a greater statistical tendency to regress to the mean -– that is, to tighten as the election draws closer –- even if the polling is initially lopsided. These distinctions will be discussed at more length in subsequent articles.