The Econometrics of Conclave
I’ve never figured that econometrics research about the Conclave could be possible. Econometrics requires data, but are there any data on the choice of the Pope? As a rule there should be no detailed information since the secrecy of the vote is determined at the Vatican. Wait a bit, unless it slips out…
Yes, that’s exactly what happened. I found an excellent research (pdf) by the econometrician JT Toman (via David Gibson) about “not-allowed-information” examining the voting patterns of seven conclaves. Some interesting results were found.
First of all, the article observes that cardinals act more and more strategically over the course of conclave. The author classifies the cardinal electors in two kinds: the sincere voters who vote in one candidate round by round regardless of the probability of success, and the strategic voters who change their choice along the events.
The table shows the behavior of these two kinds of voters for each studied conclave. The first line of each election represents the percentage of sincere voting, and the second line is strategic voting. According to the table, the percentage of sincere voters of each conclave falls depending on how long the conclave lasts. Similarly there is an increase of strategic voting by each round. We can ponder several reasons for this, transmitting cohesion within the Church to the faithful may be one reason for not extending much the conclave’s decision.
Second, it was found that most strategic voters behave like a bandwagon effect during the course of the rounds. In other words, votes converge to cardinals who received the most votes during the previous tally, and especially on those whose number of votes is increasing.
Third, a negative peculiar effect of “nightly conversations” was identified. The number of votes of the cardinal leading drops considerably the morning after the last ballot. A reasonable explanation for this is that the nightly conversations allow the cardinals to combine and to hinder the election of a cardinal who had been leading.
Finally, we should also consider that not every cardinal is eligible in practice, many are unknown so just a select subset are taken into account. This poses the problem of heterogeneity (common in economic models whose agents are not identical). Tucker pointed out that the author’s research seeks to understand the existence of heterogeneity using a curious mix of religion and econometrics:
“Finally, as is seen by the uneven voting activity among the cardinals, there is unobservable heterogeneity among the cardinals that is constant across time. Only a select subset are considered serious contenders for the position of the Papacy. One interpretation is that this is the Word of God. The voters do not perceive the cardinals as equal candidates. More cynically, one might argue that the differences observed are the result of the pre-conclave maneuverings. However, why would the effects of these be constant over time? We could expect these effect to die off after the observations of the first vote count. However, as God is omnipresent and all-knowing, having His effect to be constant over time is less of a surprise. Either way, this unobservable heterogeneity is accounted for in the estimation process through the inclusion of fixed effects parameters”.