Sabato's Crystal Ball

Did the Wall Street Meltdown Change the Election?

A Reply to James Campbell

Alan I. Abramowitz February 5th, 2009

facebooktwittergoogle_plusredditpinterestmail

Editor’s Note: Several weeks ago, the Crystal Ball published a 2008 election analysis by Prof. Larry Sabato, which concluded in part that Democrats were destined to win the presidential election, given prevailing conditions. Then Prof. Jim Campbell of The University at Buffalo took issue with that conclusion, arguing that the mid-September financial meltdown had derailed McCain. Now comes a third academic, Prof. Alan I. Abramowitz, a frequent Crystal Ball contributor and the Alben Barkley Professor of Political Science at Emory University, to take issue with Prof. Campbell’s theory. We are delighted to present these diverse viewpoints in our forum. As always, our readers can make up their own minds about last year’s historic contest.

According to James Campbell, the mid-September financial crisis hit the 2008 presidential election like a bolt out of the blue, transforming it from a horserace in which John McCain had a real chance of victory into a one-sided contest in which Barack Obama enjoyed a decisive advantage. Campbell’s claim is another version of the Wall Street meltdown theory that has been advanced by a number of conservative commentators since the November election. It is easy to understand why this explanation appeals to conservatives: it implies that Barack Obama’s victory did not result from any long-standing, deep-seated disillusionment with the Bush Administration or the Republican Party, but was simply an emotional reaction to an unforeseen crisis. Moreover, the Wall Street meltdown theory has a particular attraction for Professor Campbell–it provides him with a convenient explanation for the failure of his own forecasting model which, based largely on the results of a single trial heat poll taken a few days after the Republican convention, predicted a decisive victory for John McCain.

Campbell claims that it was impossible to accurately predict the outcome of the 2008 presidential election because the mid-September financial crisis dramatically altered the electoral landscape and rendered all previous forecasts inoperative. But the evidence from the academic forecasting literature indicates that the outcome of the 2008 election was actually highly predictable. In fact, Campbell was the only prominent academic forecaster to predict a McCain victory. The large majority of forecasts presented at the 2008 American Political Science Association meeting and elsewhere predicted an Obama victory based on factors that were known long before the election–the very same factors that have allowed political scientists to accurately forecast the results of presidential elections in the past, such as the incumbent president’s approval rating, the state of the economy, and the length of time the president’s party has held the White House. The 8 forecasts presented at the APSA meeting on August 29 predicted an Obama victory by an average margin of between 6 and 7 points–very close to the actual results of the election. My own “time for change” model predicted that Obama would win 54.3 percent of the major party vote-missing his actual 53.7 percent share by less than one percentage point. As I wrote in late August, before the national nominating conventions had taken place, “…the combination of an unpopular Republican incumbent in the White House, a weak economy, and a second-term election make a Democratic victory in November all but certain (PS: Political Science and Politics, October 2008, p. 695).”

Contrary to Campbell’s claim that the Wall Street meltdown changed everything, an examination of public opinion data from the spring and summer of 2008 shows that the Wall Street meltdown merely reinforced the public’s already overwhelmingly negative opinions about the performance of the incumbent president and the condition of the economy. According to the Gallup Poll, George Bush’s approval rating was hovering in the vicinity of 30 percent and Americans’ opinions about both their own financial situation and the condition of the national economy were overwhelmingly negative throughout the spring and summer of 2008. We now know that by Election Day, the U.S. economy had been in recession for almost a year. Hundreds of thousands of jobs had already been lost and record numbers of homes had already been foreclosed on by the time the stock market began to nosedive. For ordinary Americans, the Wall Street meltdown was not a turning point, but rather one more sign of the dire condition of the economy and the failure of the Bush Administration’s policies.

To James Campbell, John McCain’s brief lead in the polls following the Republican Convention was the most significant event in the entire 2008 presidential campaign. He argues that if the financial crisis had not occurred in mid-September, Senator McCain would have had a very good chance of holding onto this lead all the way until November 4th. However, this argument ignores the fact that Barack Obama had held a consistent but modest lead in national polls throughout June, July, and August and that the size of his lead did not expand substantially until well after the onset of the financial crisis. In the Gallup tracking poll, Obama led McCain for 19 of 21 weeks between the end of the Democratic primaries and the November election. McCain led Obama for exactly one week–the week immediately following the Republican Convention. Obama’s lead averaged 3 points in June, 4 points in July, 3 points in August, and 3 points in September. During the last two weeks of September, immediately following the onset of the financial crisis, Obama’s lead still averaged only 4 points. It was not until October that Obama’s lead expanded to an average of 8 points.

McCain’s rise in the polls in early September was a direct result of the Republican Convention and his selection of Sarah Palin as his running mate. But was it the Wall Street meltdown that brought McCain and Palin’s honeymoon with the electorate to an end, or was it simply the normal erosion of the Republicans’ convention bounce? The timing of the decline in support for the GOP ticket in the polls clearly indicates that it was the latter. An examination of data from the Annenberg National Election Study, which was tracking voter preferences on a daily basis during the campaign, shows that support for the Republican ticket began to decline within a few days after the conclusion of the Republican Convention and well before the onset of the financial crisis in mid-September (the vertical line on the graph). This evidence strongly suggests that the brief period during which the McCain-Palin ticket held a narrow lead in the polls was nothing more than a short-term public response to the overwhelmingly positive media coverage of the Republican Convention and the Palin selection–a response that was bound to fade once the media and the public began to scrutinize the Republican ticket more critically and to refocus on the major issues in the election–the performance of the Bush Administration and the condition of the American economy.



The Republican Party’s decisive defeat in the 2008 presidential and congressional elections was a result of shifts in public sentiment that long preceded the Wall Street meltdown. One of the most important of these was a shift in party identification within the electorate–a shift that had been going on for at least five years. According to annual data compiled by the Gallup Poll, between 2003 and 2008 the proportion of Americans identifying with or leaning toward the Democratic Party rose from 45 percent to 52 percent while the proportion identifying with or leaning toward the Republican Party fell from 45 percent to 40 percent. The Democrats’ 12 point advantage in party identification in 2008, up from 10 points in 2006 and 11 points in 2007, was the largest since the Gallup Poll began measuring the party leaning of independents in 1991.

These data, and similar data from other polling organizations, indicate that the American public’s disillusionment with the Bush Administration and the Republican Party began years before the Wall Street meltdown. And because party identification is by far the strongest predictor of candidate choice in U.S. elections, the Democrats’ advantage in party identification was a major factor contributing to their victories at all levels of government in 2006 and 2008, including Barack Obama’s defeat of John McCain in which, according to the national exit poll, 90 percent of party identifiers cast their ballots along party lines. Given the Democratic advantage in party identification, it was clear from the beginning of the 2008 campaign that John McCain was facing a very difficult uphill battle against Barack Obama.

In conclusion, evidence from academic forecasting models, economic trends, and public opinion polls all points to one inescapable conclusion: Barack Obama’s decisive victory in the 2008 presidential election was a predictable result of forces at work in the American electorate long before the onset of the Wall Street meltdown. To suggest otherwise is, indeed, poppycock.

I would like to thank Richard Johnston for providing these results from the Annenberg National Election Study.