< SWITCH ME >

Photo: www.eurovision.tv
Dima Bilan from Russia won the Eurovision 2008

It's the same procedure every year. On some Saturday in May, the whole of Europe is glued to the television and listens to music acts from 20-25 different European countries.

Those evenings of Eurovision Song Contest could be a wonderful contribution to mutual understanding among Europeans of any age. However, there is one obstacle: at the end you have to vote-and afterwards, only one act has won.

Since the Danish victory in 2000, winners of the Eurovision Song Contest have always come from the eastern part of Europe, with Greece and Finland being the most "western" and Turkey and Russia being the most "eastern" among the winners. The bad results of Western European countries, especially of the Big Three (Germany, France, United Kingdom), have made Western Europeans speculate about the reasons for Eastern European success and Western European failure.

Although many blame the quality of their national acts for Western European failure, more and more claim to see structural forces at work.

The main and most widespread argument is that Eastern European countries have been "logrolling" with their voting points-that is to say, trading favours. Remarks by television commentators in Germany or the United Kingdom explaining votes for Latvia or Serbia by saying they are a neighbouring country of the voting country is one example of how much currency this belief has gained.

But is this assumption, which might appear to be self-evident, really true? Or are we falling for a myth?

To answer that question I conducted a statistical analysis of the Eurovision Song Contest tele-voting in 2008. I compassed logrolling in its simpliest form, assuming that neighbouring countries are significatly more favoured and previliged in terms of voting points than non-neighbouring countries.

This leads to following hypothesis: If a country is a neighbour of another country, it receives significantly more points from that country than from others.

Not surprisingly, the results of my analysis show that this is somewhat the case: Between 10-35 % of aggregated voting results can be explained by referring to the factor "neighbouring country."

Nevertheless, the wide-spread belief is that proximity matters more in Eastern Europe than in Western Europe. As it turns out, though, this is just not the case.  The reality is much more complex. It is true that if you take Eastern Europe on the whole, neighbourhood matters more than in Western Europe.  However, if you take only those countries in Eastern Europe which are members of the EU, neighbourhood actually matters slightly less than in Western Europe.

These results make clear that it is not a question of Eastern Europe versus Western Europe.

Instead, we could wonder why being a member of the EU should make such a difference, or if the results are just a one-off. Controls with the data of the 2000 contest suggest that all figures are quite stable and that EU-effect has to be taken seriously.

Are countries which are further developed also fairer in voting? Do people in EU countries have a stronger sense of belonging and do therefore, not previlige their neighbours? The examples of the highly developed EU countries Finland and Sweden, where over 30 % of their voting behaviour can be explained by proximity, contradicts these considerations.

In the end, one might ask: "If eastern Europeans do not logroll more than Western Europeans, why do they actually win every contest?" Of course, one might argue that Eastern European acts just have higher quality. But most probably, the inconvenient truth is: There are simply more Eastern European states and this makes the Eastern European logrolling more decisive than its Western European counterpart.

It is up to you to decide if you find this bad. Unify the Balkan, let the Balkan countries, Ukraine or Moldau join the EU and hope for effects, convince Italy of joining the contest - there are many possibilities of maneouvre to get the western European acts back in the race!

NEXT ISSUE
IN -824 DAYS