The outcome of the 2012 presidential election was a letdown not just for Mitt Romney, FOX News, and numerous election forecasters, but also for proponents of economic voting. Incumbents are supposed to lose, not win reelection when economic times are bad. The country had suffered the worst recession since the Great Depression, recovery was tepid, unemployment stubbornly high, and the federal budget mired in trillion-dollar deficits. The general public was in a sour mood about the economic situation. Indeed, forecast models that relied heavily on economic variables predicted the incumbent president's defeat. So how did Barack Obama escape that fate?
Without denying the contributions of other factors, such as foreign policy, the personal qualities of the candidates, as well as the campaign, we propose an answer that keys on a critical condition of economic voting: the attribution of responsibility. This is not an unfamiliar concept in the study of the economic vote (for the latest overview, see Stegmaier and Lewis-Beck 2013). The standard assumption is that the incumbent party/president bears the responsibility for the current state of the economy. But this may not hold when the current state is, to a considerable extent, the legacy of the previous administration. Was Obama in 2012 seen as the major culprit for the bad economy? Or was it the Bush administration, under which the economic meltdown began? Bad economic times notwithstanding, the U.S. economy did show signs of recovery during Obama's first term. So voters would be able to credit the president with improving those conditions. Did Obama gain more from voters with a favorable view of the economy than he lost among those with an unfavorable view?
Using the American National Election Studies (ANES) (1) Winter 2012 survey ("Evaluations of Government and Society Survey"), we have examined these questions, taking special care to control for the effects of partisanship as well as other confounds. Our main conclusion is that the American public was reluctant to blame Obama for the bad economy, being far more inclined to blame his immediate predecessor, George W. Bush. It was almost as if Obama was running against Bush, not Romney in 2012. By saddling Bush with heavy responsibility for bad economic times voters largely shielded Obama from electoral damage. Obama was also credited for whatever recovery voters saw during bad times: voters were far more inclined to rely on positive views of the economy than on negative ones. While this type of asymmetric voting runs contrary to the more common type, where negative views carry more weight, it is consistent with the relative lack of blame placed on Obama for the bad economy. In addition, we probed the hypothesis that voters may operate in a prospective manner, as "bankers" rather than "peasants," to use a familiar pair of metaphors. It may not have mattered to voters that economic conditions were poor; what mattered was what economic conditions would be in the future. While there is some support for prospective voting, this type of orientation did not trump retrospective voting in 2012; nor did a majority see the economic future in rosy colors. In the end, it was the combination of blame and credit for economic conditions that helped Obama win with a bad economy.
The standard model of economic voting treats the electorate, in V. O. Key's memorable phrase, as a "rational god of vengeance and reward" (1964, 568). On Election Day voters reward incumbents for good economic conditions and punish them for bad ones (For the latest overview of economic voting studies, see Stegmaier and Lewis-Beck 2013; see also Duch 2007; Lewis-Beck and Stegmaier 2000; Norpoth 1996a). Candidates have proved quite adept at prompting voters to employ this calculus. Many may remember the notorious question Ronald Reagan used during the 1980 campaign, "Are you better off now than you were four years ago?" In his seminal work on economic voting, Kramer (1971, 134) put it this way: "If the performance of the incumbent party is 'satisfactory' according to some simple standard, the voter votes to retain the incumbent governing party in office, while if the government's performance is not 'satisfactory,' the voter votes against the incumbent." The standard model is all about the incumbent side of the electoral equation, the perspective on the economy is retrospective, and voters punish bad performance as much as they reward good performance. All that matters is whether economic times are good or bad, not if the incumbent government really deserves the credit or the blame for the state of the economy. Support for this model comes from a long list of studies that have probed electoral choice or presidential approval (e.g., Fiorina 1981; Kramer 1971; Lewis-Beck 1988; Lewis-Beck et al. 2008; Norpoth 1996b). It also pertains to congressional and gubernatorial elections (Atkeson and Partin 1995; Kone and Winters 1993; Stein 1990, Svoboda 1995). For the most part, it is assumed that voters rely more on their views of the state of the economy rather than their own personal experiences (Kiewiet 1983; Kinder and Kiewiet 1979).
The standard economic model has proved quite popular for election forecasting (for a comprehensive overview of election forecasting, see Stegmaier and Norpoth 2013). In the 2012 election, however, that model fared poorly. This can be seen in Table 1, where the electoral forecasts featured in PS: Political Science & Politics (Campbell 2012) are arranged by the magnitude of their Obama vote forecast. The first five of those models predicted less than 50% of the two-party vote for Obama--thus an Obama defeat. All five rely heavily on economic indicators that are consistent with the standard economic model. In fact, the Fiscal Model (Cuzan) and the Bread and Peace Model (Hibbs) were almost 90% certain that Obama would lose. Among the models that predicted an Obama victory using economic indicators along with other measures, most offered their forecasts with little certainty. Only two models attached a high degree of confidence (more than 80%) to the forecast of an Obama victory; one of them, the Primary Model (Norpoth and Bednarczuk) made no use of the economy at all. For the most part, forecast models that relied on the economy either got it wrong in 2012 or else were highly uncertain about their forecasts. How come these models were so bearish on Obama's electoral prospect and bullish on Romney's?
The main reason is that economic times were bad in 2012. The U.S. economy was still showing the scars of the Great Recession. Recovery was tepid, at best; unemployment, one of the most visible indicators of economic health, stayed above the 8% mark throughout most of the election year; such a level, conventional wisdom says, dooms the White House party. Additionally, gross domestic product growth remained anemic, well below a robust 3% level; consumers were bogged down with debt, still reluctant to spend, and the housing market was waiting for a sustained recovery (CNN Money 2013). Beyond macroeconomic indicators, perceptions of the economy were highly unfavorable in 2012.
Voters on Election Day 2012 rated the economy hardly any better than they did four years earlier, when the incumbent party lost the White House; in contrast, victory for the incumbent party in 2004 went hand in hand with a more favorable view of the economy. This can be seen in Table 1, which compares the responses of voters to common economic questions asked on the National Exit Poll from 2004 to 2012. In 2012, three-quarters rated the economy as not so good or poor, which is close to, although slightly below, the level in 2008. With regard to their own financial situation, more felt worse off than better off in 2012, just as in 2008. And as then, the economy was the most important issue for the majority of voters. Given the sour mood about economic conditions in 2012, it was not unreasonable to expect that the incumbent would be defeated again.
To see what kind of economy it takes for incumbents to win reelection, consider the 2004 election. That year, according to Table 2, about half of the voters gave the state of the national economy an excellent to good rating. They were also more inclined to say yes rather than no in response to the Reagan question: 32% felt better off, while 28% felt worse off than four years ago. And only one in five thought the economy/jobs was the most important issue in the 2004 election. Such a pattern of economic opinions, which predicts incumbent victory, was missing in 2012. So what secured incumbent victory in 2012 under economic conditions that were seen to be just as bad as those in 2008, when the incumbent party lost the White House? For an answer, we turn to the attribution of responsibility: who was mainly responsible, in the eyes of the voters for the bad economy? Did others get more blame than Obama? And did those attributions of blame sway vote choices?
Attribution of Responsibility
Economic voting requires attribution of responsibility for economic...