Sabatos Crystal Ball


Disgruntled GOP moderates could impact party unity

Alan I. Abramowitz, Special Guest Columnist March 6th, 2008


While the Crystal Ball continues to process the results of this week’s primaries, we are happy to welcome back Dr. Alan Abramowitz for another guest column this week. Alan is one of the most distinguished and best known political scientists in the United States today, and he has a rare talent for distilling data into appealing, fascinating accounts about American politics. The Alben W. Barkley Professor of Political Science at Emory University, Alan is the author of Voice of the People: Elections and Voting Behavior in the United States (2004, McGraw-Hill). –Larry J. Sabato, U.Va. Center for Politics

A lot of Republicans are unhappy with their party this year. Some conservative Republicans, following the earlier lead of talk show hosts such as Rush Limbaugh and Sean Hannity, have been threatening to sit out the November election or vote for a third party candidate because they don’t consider their party’s presidential nominee, John McCain, to be sufficiently conservative.

Since emerging as the clear frontrunner for the Republican nomination on Super Tuesday, Mr. McCain has been working hard to win the support of conservatives by stressing his hawkish views on Iraq and his conservative positions on social issues such as abortion. In a further effort to ease the concerns of conservatives, McCain recently promised to oppose any tax increases during his term as president.

But a careful examination of the evidence from the 2006 midterm elections as well as voting patterns in recent primaries indicates that it isn’t conservatives who pose the biggest threat to Republican unity in the fall. It’s moderate-to-liberal Republicans who represent the biggest challenge to John McCain in uniting his party against the Democratic nominee, especially if that nominee is Barack Obama.

While conservatives may continue to complain about McCain, they will almost certainly end up voting for him against a much more liberal Democrat. But a large number of moderate-to-liberal Republicans could actually defect to the Democratic nominee if they perceive McCain as moving too far to the right in his effort to appease party conservatives.

In the 2006 midterm elections, defections by moderate-to-liberal Republicans contributed to the Democratic takeover of the House of Representatives and were largely responsible for Republican defeats in three major Senate races in states that had voted for George Bush by wide margins in 2004: Missouri, Montana and Virginia.

According to national exit poll data, 14 percent of moderate-to-liberal Republicans in districts with competitive House races voted for Democratic candidates. This was more than four times the three percent defection rate among conservative Republicans in these districts. And defections by moderate-to-liberal Republicans played a crucial role in narrow Democratic victories in those key contests (MO, MT and VA) that enabled Democrats to seize control of the Senate in 2007. According to the exit polls in these states, the defection rate among moderate-to-liberal Republicans was13 percent in Virginia, 16 percent in Missouri and 17 percent in Montana. In contrast, the defection rate among conservative Republicans was only 3 percent in Virginia, 2 percent in Missouri and 7 percent in Montana.

The reason that moderate-to-liberal Republicans defected to the Democrats at such a higher rate than conservative Republicans was that they were much more dissatisfied with the performance of President Bush in general and with the war in Iraq in particular. According to the 2006 national exit poll, 25 percent of moderate-to-liberal Republicans disapproved of President Bush’s job performance and 31 percent disapproved of the war in Iraq. In contrast, only nine percent of conservative Republicans disapproved of Mr. Bush’s job performance and only 13 percent disapproved of the war in Iraq.

These results suggest that John McCain’s efforts to woo GOP conservatives by stressing his support for the war and his determination to continue President Bush’s policies if he is elected are likely to cost him support among moderate-to-liberal Republicans in November. Further evidence of this danger to Mr. McCain can be seen in turnout patterns in some recent presidential primaries.

Turnout in the Democratic presidential primaries this year has greatly exceeded turnout in Republican presidential primaries. Moreover, evidence from exit polls indicates that in states with open primary laws that make it easy for voters to cross party lines, a good many Republicans have been casting ballots in Democratic primaries. Evidence of this can be seen in two states with open primary laws that held their presidential primaries in February: Virginia and Wisconsin.

Based on the overall turnout in the Democratic and Republican primaries in these states and estimates from the exit polls of the size of the crossover vote in each party’s primary, we can calculate that about 16 percent of Republican voters in Virginia and 25 percent of Republican voters in Wisconsin cast their ballots in the Democratic primary. In contrast, only 2 percent of Democratic voters in Virginia and 3 percent of Democratic voters in Wisconsin cast their ballots in the Republican primary.

Both Virginia and Wisconsin are likely to be battleground states in the November election. The fact that one seventh of Republican voters in Virginia and one fourth of Republican voters in Wisconsin chose to participate in the Democratic primary should be a clear warning signal to the McCain campaign, especially if Barack Obama is the Democratic nominee. According to the exit polls, 72 percent of Republicans who voted in these Democratic primaries cast their ballots for Obama. Obama’s ability to lure large numbers Republican crossover voters in these Democratic primaries indicates that there could be a high rate of defection to Obama among moderate-to-liberal Republicans in the November election, especially if John McCain continues to focus on shoring up his support among GOP conservatives.

March 13, 2008 Addendum: Are Stay-at-Home Conservatives a Threat to John McCain?

Several alert readers of last week’s Crystal Ball article (LINK) have contacted me in the past week to question my claim that the greatest threat to Republican unity in 2008 comes from moderates, not conservatives. These readers point out, correctly, that the evidence I cite of greater defections by moderate Republicans in the 2006 midterm elections does not address the question of turnout. Indeed, since this evidence comes from exit poll data, it cannot be used to address the question of turnout. However, there is other evidence that can be brought to bear on the question of whether stay-at-home conservatives pose a serious threat to John McCain in 2008. This evidence comes from the 1992 presidential election when many conservative Republicans were also very unhappy with their party’s candidate, President George H.W. Bush, because he had broken his “no new taxes” pledge. It has become an article of faith among conservative commentators that the main reason why the first President Bush lost his bid for reelection in 1992 was that disgruntled conservatives either stayed at home or voted for the independent candidate, H. Ross Perot.

The latest version of this “stay at home conservatives” theory was set forth in the Washington Post’s Outlook section last Sunday by long-time conservative activist L. Brent Bozell. According to Bozell, unhappy conservatives cost the first President Bush a second term in the White House in 1992 just as they cost Republicans control of Congress in 2006. And if John McCain doesn’t follow the strict conservative line in 2008, Bozell argues that conservatives will again stay at home even if it means throwing the election to Hillary Clinton or Barack Obama.

Bozell’s explanation of why George H.W. Bush lost the 1992 presidential election makes for a dramatic story. But there’s one problem with it. It isn’t true. In order to test the “stay at home conservatives” theory, I used data from the 1992 American National Election Study which surveyed a national sample of eligible voters before and after the election. I compared the rates of turnout and the candidate choices of moderate-to-liberal Republicans with those of conservative Republicans. The results, displayed in the table below, clearly do not support the “stay-at-home conservatives” theory. In 1992, conservative Republicans turned out at a much higher rate than moderate-to-liberal Republicans, 90 percent vs. 76 percent, and they were less likely to defect to either Ross Perot or Bill Clinton than moderate-to-liberal Republicans. The bottom line here is that 69 percent of conservative Republicans went to the polls and voted for President Bush compared with only 44 percent of moderate-to-liberal Republicans.

Figure 1. Turnout and Support for George H.W. Bush among Moderate-to-Liberal and Conservative Republicans in 1992

Ideology Did Not Vote Voted for Clinton Voted for Perot Voted for Bush
Moderate/Liberal 24% 11% 22% 44%
Conservative 10% 5% 16% 69%

Source: 1992 National Election Study

Contrary to the claims of conservative commentators like L. Brent Bozell, it was low turnout and defections among moderate-to-liberal Republicans that did the most damage to George H.W. Bush in 1992 and it is the threat of low turnout and defections among moderate-to-liberal Republicans that poses the greatest danger to John McCain in 2008.

Alan Abramowitz can be contacted via email at The Crystal Ball extends its appreciation for authoring this guest column.