Electoral epistemology

From CNBC’s recent Russia dispatch:

Adeline Van Houtte, Europe analyst at the Economist Intelligence Unit, noted Wednesday that the vote will be an important test for United Russia…

“United Russia is now polling at around 30%, substantially lower than in 2016. Despite its poor ratings, it maintains a comfortable lead over its biggest competitors.”

“Polling at around 30%” sounds technically correct – but not particularly meaningful. Only two major pollsters, FOM and VTsIOM, are allowed to publish their findings in Russia. Naturally, both are Kremlin-friendly.

Levada Center was declared a “foreign agent” in 2016 and banned from reporting its polls during electoral campaigns. Always on the brink of closure or persecution, Levada – I suspect – prefers to err in the Kremlin’s favor. Anyway, it gave United Russia 27% in March, and there’s no indication the people have grown more supportive of the ruling party since spring.

Now let’s use some common sense. If Kremlin-friendly and/or dependent pollsters report the Kremlin party’s support level at 30%, and the second-largest party’s, say, at 15%, how much trust would you put in these percentages?

Suppose that four parties in a proportional election get 30%, 15%, 8% and 6% of the vote; the threshold is 5%; of the ten other parties, none crosses it. In that case, the modest 30% share of the vote translates into almost 51% of the seats. But this result is impossible to pull off with 15% or even 20% of the vote.

Of course turnout also matters. A lot of Russians don’t vote because they see elections as a farce with a predetermined outcome. While it is a commonplace in Russia that the Kremlin routinely rigs elections in its favor by a variety of means, some citizens interpret it as a strong reason to vote; others, as a fitting excuse to stay home.

Meanwhile, the Kremlin has significant leverage with millions of government employees. If 30% of all eligible voters supported United Russia and 80% of them showed up to vote – compared with an overall turnout of 60% – the party would win 40% of the vote and, in all likelihood, a comfortable majority of the seats.

This scenario wouldn’t make the election free or fair but would somewhat deflate accusations of outright fraud, such as ballot stuffing. However, if Kremlin-affiliated pollsters had reported the government party’s pre-election rating at 15%, its 40% share of the vote would look more that suspicious, even by Russian standards.

That’s why, in a couple of paragraphs, certain poll results published in Russia should not be mistaken for products of bona fide research.

By the way, only half of the seats in the Russian Duma get filled by proportional, party-list voting. The other half are elected on the first past the post basis.

Leave a Reply

Your email address will not be published. Required fields are marked *