On what do we base our decision to elect a particular leader? In general, are the voters gullible or in academic terms semi-rational? Is there an advantage of such limited-rationality in voters? What effect does such voting behavior have on the institutions of democracy? Can we explain the election of dictators in democracies? These are all interesting and extremely important questions, Edward Glaeser and Giacomo Ponzetto try to address in their Barcelona GSE Working Paper (No. 985), “Fundamental Errors in the Voting Booth.”
The elementary feature of this paper is modeling what is called in the psychological literature the Fundamental Attribution Error (FAE). The fundamental attribution error was named by Ross (1977) who defined it as a “general tendency to overestimate the importance of personal and dispositional factors relative to the environmental influences”. In simple words, it implies that, for example, voters who experience economic success will tend to give too much credit to their political leaders. Alternatively, voters who observe good behavior from a politician will tend to think that politician is inherently good.
In the literature, many authors have recorded evidence of the Fundamental Attribution Error. For instance, governors of oil-producing states are more likely to be re-elected when oil-prices are increasing globally. And many CEOs of companies are also rewarded for luck rather than just their performance. In this paper, Glaeser and Ponzetto study the impact of Fundamental Attribution Error on political outcomes in two standard models of political economy. The two models being the signal-jamming model of Alesina and Tabellini (2008) and the signaling model of Besley (2007).
Dictators start as messiah of prosperity
One of the most striking results of the paper is that FAE leads voter to think that the current good performance of an incumbent represents a permanent characteristic of the politician. Consequently, they should be more enthusiastic about turning a temporarily elected leader into permanent, dictatorial ruler. The intuition behind the result is that voters ignore the fact the politicians are opportunistic and are only responding to electoral incentives. In a situation of constitutional crisis, voters are more likely to choose the incumbent who has performed well, sowing seeds of dictatorship, rather than a new leader. Voters with the FAE will also favor endowing well-performing leaders with too much power, failing to recognize that their behavior will change when they no longer face electoral incentives.
According to the authors, the FAE can also explain why independent institutions such as a free press can lose credibility among voters. Glaeser and Ponzetto note that the FAE makes voter under-estimate the value of such institutions because they believe that what they already observe reveals more than it does about the politician’s own contribution. This result may explain the limited protest against state subversion of the press in Turkey and Russia in recent times.
On the other hand, their work also confirms that imperfect rationality can have positive effects on democracy by changing the behavior of politicians (Ashworth and Bueno de Mesquita 2014). Voters attribute good behavior by politicians to permanent characteristics, not temporary incentives, and consequently, good behavior carries more clout in the voting booth, which in turn creates stronger incentives for good behavior.
In this paper, the authors explore the political consequences of the key finding in social psychology that individuals underestimate the importance of circumstances and overstate the importance of enduring personal characteristics. This FAE naturally leads voters to make mistakes about re-election, such as overweighting skills in areas where politicians actually have little impact relative to skills in areas where politicians really make a difference.
All in all, the FAE could be potentially useful in reducing corruption in some instances but the potential risks for institutional design suggest that careful limitations on institutional change would be welfare improving. Slow-moving constitutions that require lengthy debate before a change would seem to provide a check on the errors created by the FAE.