Today Qatar was the surprise winner in the bid to host the FIFA World Cup in 2022, beating Japan, The United States, Australia, and Korea. It’s an interesting procedure by which the host is decided consisting of multiple rounds of elimination voting. 22 judges cast ballots in a first round. If no bidder wins a majority of votes then the country with the fewest votes is eliminated and a second round of voting commences. Voting continues in this way for as many rounds as it takes to produce a majority winner. (It’s not clear to me what happens if there is a tie in the final round.)
Every voting system has its own weaknesses, but this one is especially problematic giving strong incentives for strategic voting. Think about how you would vote in an early round when it is unlikely that a majority will be secured. Then, if it matters at all, your vote determines who will be eliminated, not who will win. If you are confident that your preferred site will survive the first round, then you should not vote truthfully. Instead you should to keep bids alive that will easier to beat in later rounds.
Can we look at the voting data and identify strategic voting? As a simple test we could look at revealed preference violations. For example, if Japan survives round one and a voter switches his vote from Japan to another bidder in round two, then we know that he is voting against his preference in either round one or two.
But that bundles together two distinct types of strategic voting, one more benign than the other. For if Japan garners only a few votes in the first round but survives, then a true Japan supporter might strategically abandon Japan as a viable candidate and start voting, honestly, for her second choice. Indeed, that is what seems to have happened after round one. Here are the data.
We have only vote totals so we can spot strategic voting only if the switches result in a net loss of votes for a surviving candidate. This happened to Japan but probably for the reasons given above.
The more suspicious switch is the loss of one vote for the round one leader Qatar. One possibility is that a Qatar supporter , seeing Qatar’s survival to round three secured, cast a strategic vote in round two to choose among the other survivors. But the more likely scenario in my opinion is a strategic vote for Qatar in round one by a voter who, upon learning from the round 1 votes that Qatar was in fact a contender, switched back to voting honestly.
12 comments
Comments feed for this article
December 3, 2010 at 12:57 am
ryan
there was also an extra voter in round one, so maybe the eleventh qatar vote got sick or something?
December 3, 2010 at 9:10 am
jeff
oops, a mistake in the table. japan had only 3 votes in the first round. so one vote on net switched away from japan and one vote on net away from qatar.
December 3, 2010 at 1:25 am
Name: Mark
The Democratic House leadership’s gambit on middle class tax cuts worked, Republican lies about tax cuts for middle class families being a job-killer and un-American and a travesty notwithstanding. The bill just passed, 234-188, with three Republicans voting for us, and 20 “Dems” voting for the millionaires.
December 3, 2010 at 3:22 am
Vinnie
I’m surprised by the support South Korea considering how recently they hosted. Anyway, the whole process seems a little sketchy.
December 3, 2010 at 6:10 am
David Barry
Japan only got 3 votes in round one.
December 13, 2011 at 6:11 pm
Lois
Just the type of iinshgt we need to fire up the debate.
December 14, 2011 at 5:52 am
gprgvuwrat
RrC8JM nxkznyhhmhxq
December 16, 2011 at 6:15 am
yzyznanm
Mak40x gvhwvkmmoicb
December 3, 2010 at 6:25 am
Strategisch stemgedrag bij de FIFA | eco.nomie.nl
[…] De 2 stemmen van Engeland in ronde 1 zijn in ronde 2 allebei naar Rusland gegaan. Tot zo ver is er nog niets aan de hand. Maar twee leden die in ronde 1 nog voor Nederland en Belgie stemden, kozen in ronde 2 voor Rusland. En dat is vreemd. Als Nederland/Belgie van alle opties hun favoriet was, waarom was dat dan niet meer het geval toen er nog maar 3 opties in de race waren!? Door deze leden is er dus strategisch gestemd, in ronde 1 dan wel in ronde 2. (zie ook hier) […]
December 3, 2010 at 6:28 am
Marco
This is even clearer in the 2018 vote, the results of which were as follows. Round 1: England 2, NL/Bel 4, Spa/Por 7, Russia 9, England eliminated. Round 2: NL/Bel 2, Spa/Por 7, Russia 14.
Hence the two England voters from round 1 switched to Russia, which makes sense. However, 2 members that voted for Netherlands/Belgium in Round 1 switched to Russia in Round 2, which clearly reflects strategic voting.
England was apparently the favorite to win this election. So it may very well be that the only reason that they were eliminated in round 1 is that almost everyone strategically voted for their second preference, under the assumption that England would survive this round anyhow…
December 3, 2010 at 10:51 am
gabriele
The type of strategic voting you mention is often referred to as “push over”: I prefer candidate A, but I vote for candidate C because in the second round A would defeat C, but lose against B.
What conditions do you need for this? First, you need that A is far from winning in the first round, but well ahead of B and C. Second, you need that C would defeat A in the second round, but A would win against B. Third, that you expect B and C to have similar vote shares in the first round.
The second condition is the most important! It implies that if A and B pass to the second round, most C voters would vote for B (otherwise, how can A lose against B?). This means that they actually prefer B to A (their preferences are C>B>A).
Suppose that I, as a voter preferring A, vote for C. What would voters with preferences C>B>A do? If they vote for C, they will make so that A and C pass to the second round and A will win. If they vote for B, they will make so that A and B pass to the second round and B will win. Hence, since they prefer B to A, they will vote for B. Then, why should I vote for C. Indeed I should not: I should (and will) vote for A.
Yes, you can have push over in equilibrium in some edge cases (all voters preferring C are indifferent between A and B), but in general it does not seem possible.
December 3, 2010 at 12:21 pm
Hardy24
In the FIFA voting rules if two bidders are tied in the final round, the President has the casting vote.