In American politics, what comes around usually goes around. But it doesn’t always go around this quickly. Just one year after their decisive victories in the 2008 presidential and congressional elections, Democrats appear to be in serious trouble. Some political commentators believe that Republican victories in the Virginia and New Jersey gubernatorial elections reflect growing discontent with President Obama’s performance and may predict substantial GOP gains in next year’s midterm elections.
On occasion, victories in Virginia and New Jersey have been followed by big gains in the midterm elections. In 1993 Republicans won both of these off-year contests; one year later they gained 54 House seats and 8 Senate seats and took control of Congress for the first time in almost half a century. Twelve years later, in 2005, Democrats swept to victory in Virginia and New Jersey; a year later they picked up 30 House seats and 6 Senate seats and took control of Congress back from the GOP.
But Virginia and New Jersey haven’t always served as bellwethers for the nation. In 1997 both states elected Republican governors but one year later, Democrats gained seats in the House. And in 2001 both states elected Democrats but one year later Republicans gained seats in the House and Senate.
Before we conclude that Republican victories in Virginia and New Jersey provide an early indication of what is likely to happen in next year’s midterm elections, we need to take a more systematic look at the evidence. To do this, I analyzed the relationship between the results of Virginia’s and New Jersey’s gubernatorial elections and the results of the following year’s U.S. House elections for the 11 elections between 1965 and 2005. I chose 1965 as my starting point because it marked the beginning of meaningful two-party competition in gubernatorial elections in Virginia. This analysis shows that the results of the Virginia and New Jersey gubernatorial contests do not predict the results of the following year’s midterm elections.
Table 1 displays the results of the Virginia and New Jersey gubernatorial elections between 1965 and 2005 along with the results of the midterm elections that took place one year later. The table also shows the party of the president during each of these midterm elections. One of the best known regularities in American politics is the tendency of the president’s party to lose seats in midterm elections, especially in the House of Representatives. Since World War II, the president’s party has lost House seats in 14 of 16 midterm elections with an average loss of 23 seats. Therefore, in order to determine whether Virginia and New Jersey predicted the results of midterm elections, we have to take into account whether there was a Democratic or Republican president at the time of each midterm election.
Table 1. Virginia, New Jersey and Midterm Election Results since 1965
Source: uselectionatlas.org and Congressional Quarterly Guide to U.S. Elections
The data in Table 1 show that in the 11 off-year elections between 1965 and 2005, Republicans won both statehouses three times, Democrats won both four times, and the parties split the two statehouses four times. But there is not a consistent relationship between the results of the two gubernatorial elections and the results of the subsequent midterm election. In midterms following a Democratic sweep, Republicans gained an average of four House seats; in midterms following a split result, Republicans lost an average of 16 seats, and in midterms following a Republican sweep, Republicans gained an average of 13 seats. The correlation between the gubernatorial results and the midterm results is a paltry and statistically insignificant .09.
In contrast, there is a much stronger relationship between party control of the White House and the results of these midterm elections. In seven midterms with Democratic presidents, Republicans gained an average of 28 House seats; in four midterms with Republican presidents, Republicans lost an average of 18 house seats. The correlation between the party of the president and the results of the midterm election is a strong and statistically significant -.75. The correlation is negative because control of the White House is associated with losses in the midterm elections.
To provide a more rigorous test of the predictive power of the Virginia and New Jersey results, I performed a regression analysis with Republican House seat change as the dependent variable and two independent variables: the number of Republican victories in the two off-year gubernatorial elections and the party of president which was coded as 1 for Republican presidents and 0 for Democratic presidents. The results of the regression analysis are displayed in Table 2.
Table 2. Results of Regression Analysis of Midterm Election Results, 1966-2006
Source: uselectionatlas.org and Congressional Quarterly Guide to U.S. Elections
The regression analysis confirms the findings of the bivariate analysis. The party of the president had a significant impact on the results of these midterm elections: Republicans did substantially worse in midterm elections under Republican presidents than in midterm elections under Democratic presidents. However, the results of the previous year’s gubernatorial elections in Virginia and New Jersey did not predict the results of the midterm elections. Not only is the estimated coefficient for the Virginia/New Jersey election variable small and statistically insignificant, but it is in the wrong direction: the better Republicans did in Virginia and New Jersey, the worse they did in the subsequent midterm election.
Why did the Republicans Win?
Did the Republican victories in Virginia and New Jersey reflect voter discontent with President Obama’s performance–discontent that could produce Republican gains in 2010 if it persists or intensifies? Both states voted for Barack Obama in last year’s presidential election–Virginia by a margin of 6 points and New Jersey by a margin of 15 points. This year, however, turnout was way down, as it always is in off-year elections. The number of votes cast fell from 3.7 million to less than 2 million in Virginia and from 3.9 million to 2.3 million in New Jersey. And the evidence from exit polls in both states shows that the Republican base was more energized this year than the Democratic base. This year, 18-29 year-old voters, who played a key role in Obama’s victory in 2008, made up only 10 percent of the electorate in Virginia and only 9 percent of the electorate in New Jersey compared with 21 percent and 17 percent respectively in 2008. In Virginia, African-Americans and other nonwhites made up only 22 percent of the electorate this year compared with 30 percent in 2008. And in both Virginia and New Jersey, Republicans and conservatives comprised a larger share of the electorate this year than in 2008. The difference was particularly striking in Virginia where Democrats outnumbered Republicans by 39 percent to 33 percent last year but Republicans outnumbered Democrats by 37 percent to 33 percent this year. Although Barack Obama won Virginia by 6 points last year, among this year’s voters, John McCain defeated Obama by a margin of 7 points.
The turnout patterns seen in Virginia and New Jersey are fairly typical for off-year elections although the differences between the presidential and off-year electorates appear to be somewhat larger than usual. In off-year elections, supporters of the opposition party and those who are dissatisfied with the status quo in Washington are usually more energized than supporters of the president’s party and those who are satisfied with the status quo. This helps to explain why candidates from the president’s party have lost the last six gubernatorial elections in Virginia and New Jersey. But this doesn’t mean that these elections should be viewed as referenda on the performance of the president. Even with the Republican skew of the electorates in Virginia and New Jersey, the exit polls showed that President Obama was fairly popular in both states: 48 percent of the voters in Virginia and 57 percent of the voters in New Jersey approved of his job performance.
Despite the low turnout of Democratic base voters in Virginia and New Jersey, if the voters had cast their ballots based on their opinions of the President’s performance, Creigh Deeds would have lost a nail-biter in Virginia and Jon Corzine would have coasted to a second term as governor in New Jersey. In fact, both Creigh Deeds and Jon Corzine did relatively poorly among voters who supported the President. While almost all of those who disapproved of President Obama’s performance voted for the Republican gubernatorial candidates, only 80 percent of those who approved of the President’s performance in Virginia voted for Deeds and only 73 percent of those who approved of the President’s performance in New Jersey voted for Corzine. It was Jon Corzine’s unpopularity and Creigh Deeds’ weakness as a candidate, rather than a negative reaction to President Obama, that largely explains the Democratic defeats in these elections.
Discussion and Conclusions
The Virginia and New Jersey gubernatorial elections get a lot of national media attention because no other states hold their gubernatorial elections in the year between the presidential election and the midterm election. It is therefore understandable that political observers frequently speculate about the implications of the results in these two states for the upcoming midterm elections and interpret the results as referenda on the performance of the incumbent president. However, the evidence presented in this article provides no support for the belief that the Virginia and New Jersey results predict what will happen across the entire nation next year or that these elections constituted referenda on President Obama’s performance. Instead, the Democratic defeats in Virginia and New Jersey reflected a combination of normal turnout patterns favoring the out-party in off-year elections and the weaknesses of the Democratic candidates in both states. It is very likely that Democrats will lose seats in the House of Representatives in the 2010 midterm election, but that is because there is a Democratic president in the White House and the president’s party almost always loses House seats in midterm elections. How many seats they lose will depend on how voters in Virginia, New Jersey and the other 48 states evaluate the performance of President Obama and the Democratic Congress a year from now.