Louis Jacobson is a GOVERNING contributor.E-mail: email@example.com
On Election Day 2012, voters in 11 states will choose both a president and a governor. If the current outlook holds, many of them could end up pulling the lever for a Democrat and a Republican.
And if that happens, voters in those states will be following a noteworthy -- and little noticed -- pattern: The state's electoral college votes go to a candidate of one party, while a candidate from the other party wins the governorship. Over the last five presidential election years, voters have split their presidential and gubernatorial votes half the time when the gubernatorial race is competitive.
While political scientists have long tracked the frequency of ticket-splitting, they have usually focused on voters who decided to cast ballots for a president from one party and a Member of Congress from another. I should note that there are two types of methodologies that have been used to study ticket splitting. The first type is based on survey data in which a sampling of individual voters is asked about their voting behavior. The second way looks at who wins an individual jurisdiction, such as a state or a congressional district.
Here's a brief rundown on what they found:
A good summary of the first kind of study was published in The Logic of American Politics, by University of California-San Diego political scientists Samuel Kernell, Gary C. Jacobson and Thad Kousser. Using survey data, they found that ticket-splitting between presidential and House contests hovered just below 15 percent in the 1950s, peaked at about 30 percent in 1972, and has steadily declined since then, settling in at about 17 percent in 2008. The authors' data comes primarily from the American National Election Surveys.
Jeff Stonecash, a Syracuse University political scientist, attributes this rise and fall to a lengthy partisan "realignment" beginning in the 1960s and lasting through much of the 1990s -- particularly the shift of southern, white Democrats to the Republican Party. During the period of realignment, many southern whites voted Republican for president but Democratic for Congress.
"That produced more ticket-splitting," Stonecash said. But that pattern is now ebbing, he said, thanks to the near-disappearance of southern white Democrats in Congress. This means that for many Americans, votes for president and Congress are increasingly in tune.
We were able to locate one recent example of a study that looked at individual jurisdictions' voting patterns -- an analysis of votes for the House and the president in 2008 by the Cook Political Report. Of 435 congressional districts, 82 cast votes for members of different parties for the two offices. That's 19 percent, broadly in line with what the survey data found.
The gubernatorial landscape
I asked several political scientists who have studied the topic whether they were aware of studies comparing presidential electoral college votes to gubernatorial votes, and they said they weren't. So I took my own look.
To keep the data manageable, I decided to look at the past five presidential election years - 1992, 1996, 2000, 2004 and 2008. The downside of this approach is that only 11 states regularly hold gubernatorial elections during presidential years, which keeps the universe somewhat restricted. Still, we decided it was more important to test elections held on the same day.
All told, during those five elections, a state split its votes for president and governor 21 of 56 times, or 38 percent of the time. This suggests a rate of divergence well above what was seen even at the historical peak levels in the previous studies.
And if anything, the trend seems to be growing stronger. If you look at the past three election cycles instead of the past five, the rate of ticket splitting rises to 42 percent.
I also threw another factor into the mix. I focused in on a sub-group of states -- those where the gubernatorial race was competitive between the two parties. (By competitive, I meant that the race ended up being close -- single digits separating the candidates.) Why? I figured that in states where the weaker party had a less-than-credible candidate running for at least one of the offices, voters who might otherwise be inclined to split their ticket would have faced strong pressure not to.
It turned out that my hunch was right. Over the past five cycles, exactly half of the states with a competitive gubernatorial race in a presidential year -- 11 of 22 contests -- ended up going with a governor and a presidential candidate from different parties.
First, let's pause for a moment to think about how stunning this is: When a state has a competitive gubernatorial race during a presidential election year, its choice to fill the two offices is no more highly correlated in party affiliation than a random coin flip.
To make sure that these high rates of ticket-splitting weren't just an artifact of a few states acting oddly, we took a closer look at all 11 states that held gubernatorial elections in each of the five election years during this period. It turns out that eight of the 11 split their ticket at least once during that period, far more than the three states that never did so. This suggests that ticket splitting is not that unusual in most of the states on this list.
For the record, the eight states that split their ticket at least once were Indiana, Missouri, Montana, New Hampshire, North Carolina, Vermont and West Virginia. The four that did not were Delaware, North Dakota, Utah and Washington state.
Why does this happen?
How could this be?
One intriguing possibility is that voters are sophisticated enough to understand that state officials have a certain set of duties and federal officeholders have a different set, and that voting ought to take into account which politician is the best one to handle one set of responsibilities as opposed to the other.
Take Montana, where Democrat Brian Schweitzer was able to win a gubernatorial race twice on days when his state was supporting a Republican presidential candidate. (In 1992, Republican Marc Racicot also won the governorship as Democrat Bill Clinton was winning the state's presidential election, aided by a strong Ross Perot third-party showing in the state.)
In Big Sky country -- as is the case elsewhere in the West -- one of the main things voters want from their federal officeholders is to refrain from imposing the dictates of the central government on the state. By contrast, gubernatorial candidates -- even Democrats whose party affiliation would suggest that they prefer a bigger role for government -- don't carry the same baggage, because they're located far from the partisan fights of Washington.
"I think a skillful candidate can frame the race as based on state issues," said one longtime political observer in Montana. "There is a real difference between the parties on many state issues, and I think Montana voters will tend to key on those differences, rather than who they want to be president."
Something similar has typically happened in North Carolina, said Ferrel Guillory, director of the University of North Carolina's Program on Public Life. In North Carolina, the state has frequently supported Republicans for president even as it elects Democratic governors.
Part of this has to do with North Carolinians' desire for pragmatism among its governors, Guillory said.
"This state has voted much more ideologically for president and senator than for governor," he said. "North Carolina likes its governors, whether Democratic or Republican, to be buttoned-down and ready to work on schools and highways."
Here are some other explanations for ticket splitting that we heard when we asked a range of political observers:
--The rise of independent voters. While increased partisanship in both parties has driven down the likelihood of ticket splitting by individuals (the phenomenon analyzed by the academic studies cited earlier) a parallel trend has, somewhat paradoxically, made states as a whole more open to divided results. The reason: independent voters.
The number of independent voters has been increasing for years. Most recently, according to an analysis by USA Today, the number of independents has risen by more than 400,000 since 2008, an increase of almost 2 percent.
Almost by definition, voting behavior by independents is less predictable than that of partisans. To the extent that independents are now common enough in many states to serve as kingmakers on Election Day, they can promote exactly the divergent results we're seeing in our analysis. "We have a lot of ticket splitting in New Hampshire because we have a lot of independent voters," said Linda Fowler, a Dartmouth College political scientist.
--Historical legacies. It's hard to win the governorship unless you can groom home-grown political talent and push it up the political ladder. And in at least two of the states on this list -- North Carolina and West Virginia -- the Republican Party has been burdened with being second-rate in lower-level state office races even as its federal candidates have won races consistently.
By now in North Carolina, this transformation is almost complete, as the GOP in 2010 won control of the state legislature and is favored to win the governorship in 2012. (Ironically, GOP gains at the state level have coincided with the party's weakening hold on the state's presidential vote, thanks in part to moderates and independents in the state trending toward the Democrats. It's too soon to say how the topic of gay marriage, currently dominating the news cycle, will affect North Carolina voting just yet.)
In West Virginia, by contrast, the Democrats remain solidly in control of state-level offices -- as they have for decades -- even as the party has fallen deeply out of favor in presidential races. (So much so that a felon behind bars in Texas recently took 42 percent of the Democratic primary vote away from Obama.)
That's how Democratic governors like Joe Manchin and Earl Ray Tomblin have kept hold of the governor's mansion despite difficult headwinds at the top of the ticket.
"Despite attempts by state and national Republicans to pin Obama and his policies to the state slate of Democratic candidates, it really hasn't worked," said Steven Allen Adams, a political observer and blogger in West Virginia. "West Virginia voters, particularly Democrats, are able to vote overwhelmingly for state Democrats and vote against Obama."
--Candidate quality matters. In several cases, states have rewarded gubernatorial candidates who exhibited unusual political skill and a willingness to buck their party, even as voters spurned the candidate's party in presidential races. Indeed, because a gubernatorial race is a statewide contest, it's less likely to be "nationalized" into a generic party mold than either the presidential race or congressional races, which tend to be highly centralized. This has allowed individual politicians to break out of the crowd.
We've seen this most commonly with moderate Democrats, notably Jim Hunt in North Carolina, Evan Bayh in Indiana, Schweitzer in Montana and Jay Nixon in Missouri. All have hewed closely to their state's conservative leanings despite their party affiliation. "Missouri voters have frequently split tickets because they vote more the person than the party," said Ken Warren, a political scientist with Saint Louis University.
The idea that "candidates matter" is especially true in small states. Voters in small-population states like Montana, New Hampshire, Vermont and West Virginia may either know the gubernatorial candidate personally or have multiple opportunities to meet them in person. This personal contact can make more voters willing to overlook ideological differences they may harbor with that politician's party.
Such close contact is what helped elect and reelect Republican Jim Douglas as governor of solidly blue Vermont. "Somewhere between 25 percent and 35 percent of Vermont voters chose" Douglas as they voted for a Democratic presidential candidate in 2004 and 2008, noted Eric L. Davis, an emeritus political scientist at Middlebury College.
The likeliest example of a heterodox Republican winning in a Democratic state could come this fall if Rob McKenna can win the governorship, even as Washington state backs Obama for president.
--A lack of a straight-ticket-voting option. Fourteen states have some type of option that enables voters, with just one ballot mark, to choose all candidates from a given party, according to the National Conference of State Legislatures. A 15th state, North Carolina, offers the option for all races except for president, which is why it's not included here.
However, only three of the states in our analysis (excepting North Carolina) have a straight-ticket option -- Indiana, Utah and West Virginia. So voters in the other eight states we're looking at don't have an easy option to choose a straight ticket.
What this means for 2012
If there was already a tendency toward ticket-splitting in presidential and gubernatorial races, 2012 is shaping up to be a banner year. Five of the 11 states with both contests on tap for this year seem unlikely to split their tickets -- the Democratic states of Delaware and Vermont and the Republican states of Indiana, North Dakota and Utah. But six states, or more than half the total, could end up doing so.
In two of these states -- North Carolina and New Hampshire -- both the gubernatorial and presidential races are competitive, meaning that almost any combination of winners and party affiliations is possible.
In one state, Missouri, the presidential race leans Republican but the gubernatorial race leans pretty solidly Democratic, as Nixon seeks a second term.
In the three other states, the presidential races are not competitive, with Montana and West Virginia likely to go Republican and Washington likely to go Democratic, but the gubernatorial candidate from the other party has a reasonable shot at winning.
Ultimately, it's not impossible to believe that the ticket splitting in competitive races could double from its current 50 percent to 100 percent this fall.
And that would be a result for the history books.
GOVERNING Politics is the place for news and analysis on campaigns and elections. If there's a ballot measure in California, a legislative election in Alabama, a mayoral election in Anchorage or a governor's race in Rhode Island, GOVERNING Politics probably is writing about it. We love everything about state and local politics, from polls and campaign ads to policy debates and demographic trends.