It's the lottery, stupid!
Manuel Bagüés y Berta Esteve-Volart
Following the global financial crisis, incumbent parties in many countries are losing elections. This conforms with a large literature that finds that when economic conditions are bad, incumbent politicians are less likely to be reelected. In this column, the authors explore the nature of this relationship by taking advantage of a unique randomized experiment at the macroeconomic level: the Spanish Christmas Lottery.
According to surveys, Zapatero's Socialist Party is expected to take a historic beating in next year's Spanish elections. This is not unlike recent electoral experiences in a variety of European countries, including the last UK national election, where the Labour Party suffered its worst defeat since 1983, and the last regional elections in France, where Sarkozy's party lost in a landslide. This is not a uniquely European experience: in the recent U.S. midterm election, the Democrats suffered, in Obama's words, "a shellacking". These electoral outcomes are consistent with the phenomenon of economic voting: when economic conditions are good, incumbents tend to obtain relatively more votes. But what is the source of this positive correlation?
The literature has thus far explored two main hypotheses. On the one hand, in a context of asymmetric information, it could reflect voters interpreting economic conditions as a signal of incumbent's ability or effort (Nordhaus 1989). On the other hand, this correlation could also be due to voters' systematic attribution errors, according to which voters aiming to assess competence systematically fail to take sufficient account of background or environmental factors (Ross and Nisbett 1991). Both of these hypotheses rely on voters' lack of information.
It is hard to empirically disentangle why economic outcomes affect voting behavior, because of the difficulty inherent in identifying variations in economic conditions that are independent from incumbents' actions. Even in the case of seemingly exogenous events such as natural disasters, where the incumbent does not have control over the event itself, the incumbent may be plausibly held responsible by voters for either preparation or response. In a recent paper (Bagues and Esteve-Volart 2011), we manage to overcome these problems by exploiting the exceptional evidence provided by a unique randomized natural experiment at the macroeconomic level: the Spanish Christmas Lottery.
The Spanish Christmas Lottery as a randomized natural experiment at the macroeconomic level
The Spanish Christmas Lottery, also known as el Gordo (the Fatty), is a unique event in the world. Unlike most lotteries, where only a small share of the population participates, every Christmas 75% of Spaniards participate in the Fatty. Its economic impact is very large: in 2010, the average Spaniard spent €70 on the Christmas Lottery, and the total expenditure amounted to approximately €3 billion, roughly 0.3% of the Spanish GDP.
One possible explanation for its popularity is that it is more of a social event than a gamblers' lottery. Instead of awarding one big prize to a few individuals, as is the case in most lottery systems, the Fatty awards many relatively smaller prizes to several thousand individuals sharing the same ticket number. Because each number is mostly sold by one lottery outlet, winners tend to be geographically clustered. The draw attracts wide TV audiences, and when the top prize is drawn, TV cameras travel to the winning location to show images of some of the happy winners. All in all, the Fatty constitutes a Spanish Christmas tradition.
Our study uses provincial information on the Christmas Lottery top prizes and expenditure from 1986 through 2008 to identify random variations in annual provincial income, and finds a significantly positive effect on the outcomes of incumbents in national elections. The main winning province receives a mean income shock equivalent to 3% of its GDP, and the incumbent party enjoys a significant increase in the share of votes received in the province of approximately 0.6 percentage points.
What is the magnitude of the effect? Consider the case of the average province in 2008, with an electoral roll of 675,000, receiving Christmas Lottery prizes equivalent to 3% of its provincial GDP (approximately €600 million). According to our results, in that case the incumbent obtains 4,000 more votes than expected. Considering the prizes awarded, this makes a "total cost" of €150,000 per vote.
What is the reason behind the effect of the Christmas Lottery on voting behavior?
Because it is public knowledge that the incumbent cannot affect which province receives the Fatty, our results rule out explanations according to which voters may be rewarding the incumbent. The data also seem to reject the possibility that voters are subject to some type of attribution error. Since winners are well aware of the random nature of lottery, it is unlikely that they are wrongly attributing variation in their economic conditions to the government. Moreover, information from surveys from the same period shows that Christmas Lottery prizes increase the propensity to vote for the incumbent, but they do not affect the respondents' assessment of the government.
The results in our paper suggest that the correlation between economic outcomes and incumbent reelection may reflect something other than voters being uninformed. Given that the effect of lottery prizes on votes seems to vanish after two or three years, our results are suggestive of a temporary happiness effect. This is consistent with the evidence in previous literature finding that, after a period of peak experience following a lottery award, winners return back to their prior level of happiness. In the same vein, perhaps when voters are happier, they become more lenient with the incumbent. This would be consistent with the evidence in Healy et al. (2010), who, using information from a completely different setting, find that the outcome of U.S. local college football games just before an election affects the incumbent's reelection. An alternative explanation is that increases in wealth may reinforce voters' preference for the status quo.
In any case, our results suggest that, given the current economic crisis in Spain (and barring a speedy economic recovery), Zapatero's party is likely to lose in a landslide in the next election---whether or not voters believe that he is to blame for the crisis.
Bagues, Manuel and Berta Esteve-Volart (2011), Politicians' Luck of the Draw: Evidence from the Spanish Christmas Lottery, FEDEA Working Paper No. 2011-1 .
Healy, Andrew, Neil Malhotra and Cecilia Hyunjung Mo (2010), Irrelevant Events Affect Voters' Evaluations of Government Performance, Proceedings of the National Academy of Sciences of the United States of America, Vol. 109(29), pp. 12804-9.
Nordhaus , William D. (1989), Alternative Approaches to the Political Business Cycle, Brookings Papers on Economic Activity, Economic Studies Program, The Brookings Institution, Vol. 20(2), pp. 1-68.
Ross , Lee D. and Richard E. Nisbett (1991), The Person and the Situation (New York: McGraw-Hill).
A translation of this article was published in nadaesgratis.es on December 21st 2010