Do governments of host countries get an Olympic boost?

Since ancient times, rulers have used festivals and games as distractions for the masses; an entertained populace is likely to be a content one, or at least a more content one than it would otherwise have been. Does this feed through to a government’s re-election chances? Surprisingly, it seems it might do.

In the 116-year history of the modern Olympic games, only three have been held in wholly non-democratic countries: Berlin 1936, Moscow 1980 and Beijing 2008. That leaves no fewer than 23 Games hosted by democracies of one form or another – a fairly large group to look at.

There are of course all sorts of definitional problems in determining whether a government was re-elected or not. How are minority or coalition governments counted? Do you look at presidencies or parliaments in hybrid systems? Should state or national elections count in federal countries? How democratic does a country have to be to be included? Generally, the rule I’ve gone with is if a national government is substantially of the same party after the election as it was at the time of the Games, it’s been re-elected – but we shouldn’t take any of it too seriously.

Of the 23 cases, 15 governments were re-elected and eight fell either at or before the next election: a 65% ratio. To know how good that is, we need a baseline, which I’ve taken as all the general elections held in the host countries twenty years either side of the Games, excluding undemocratic periods.

That gives a general re-election rate for governments at and around the time of each Games as 61%, so only marginally worse.

A much greater disparity opens up in the post-WWII figures (where the data is also more straightforward). Here, no fewer than twelve of the fourteen governments of the host countries were re-elected. The exceptions are Finland, where the government from 1952 had already fallen following a crisis prompted by a proposed austerity programme, and Canada, which briefly ejected Trudeau from office in 1979. Every other government was subsequently re-elected, a success rate of 86%. The contextual general re-election rate is 68% – a substantially lower figure. The gap grows even larger if Mexico 1968 is excluded as not sufficiently democratic.

Does any of this matter? Some of the sample sizes are small and obviously there are other, generally far more significant events going on at the time. That said, the gap in the ratios is quite striking and the ability to put on a successful Games (something Montreal – one of the two exceptions – notably failed to do), can lead to a country feeling good about itself and, at the margins, about its government.

Obviously, the success or otherwise of the London Games will not determine the government’s fate in 2015. They might, however, not be completely meaningless either.

David Herdson