Tuesday brings yet another dramatic development in the story of Russia’s relationship with the West. Monday, it was the possibility of improving relations in the wake of French President Francois Hollande’s decision to visit Moscow to seek solidarity in the fight against the Islamic State. Tuesday, it is the specter of renewed conflict following the downing of a Russian fighter jet by Turkish armed forces. Tomorrow, it may be something else.
No matter the direction in which relations are trending, new research (ungated) presented at last week’s Association for the Study of East European and Eurasian Studies Annual Meeting by political scientists Timothy Frye, Scott Gehlbach, Kyle L. Marquardt, and Ora John Reuter suggests that the West will be dealing with a leader who is genuinely popular at home.
On the one hand, this news is hardly earth-shattering. As data from the Levada Center in the figure below shows (click on it if you want a larger version), Russian President Vladimir Putin has long enjoyed approval ratings that just about any Western leader would envy:
However, as the authors note, there are some reasons for thinking these numbers might be artificially inflated. Pollsters might be under pressure to produce numbers that are favorable to Putin (although Western researchers report similar numbers). Alternatively, respondents might not share their true preferences with pollsters. This could be because they are somehow afraid there will be negative repercussions from expressing anti-Putin sentiment or, more benignly, because they simply want to tell the pollster what they think the pollster wants to hear, a phenomenon known as social desirability bias.
Fortunately, new techniques have been developed in recent years by survey researchers in order to measure these kinds of “sensitive” opinions known as “list experiments” (or, more formally, item-count experiments). The way this works is as follows. First, survey respondents are randomly split into two groups. The first group is asked how many — but not which — of a certain list of items they like (or dislike, or disapprove of etc., depending on the needs of the study). The second group is presented with the same list of items plus the sensitive one.
So to give an example, if we thought people were sensitive about saying they didn’t like cats, we might ask the first (control) group how many of the following animals — dolphins, dogs, and chickens — they like, and then the second (treatment) group would be asked how many they liked out of dolphins, dogs, chickens and cats. Since we expect that support for dolphins, dogs, and chickens will on average be equal across the groups (since people will be randomly assigned to each group), the difference in the average number of items people state they like across the two groups gives us an estimate of proportion of the population that likes cats without forcing anyone to specifically answer a question about whether or not they like cats.
This is the method the authors employed in Russia. They ran two surveys, one in January and one in March of 2015, and in each survey embedded two experiments: one where Putin was included alongside a group of contemporary Russian leaders, and another where Putin was included alongside more historical leaders.
The net result, though, was the same across all of the studies. While the list experiments suggested that Putin was indeed a bit less popular than one would conclude from the direct questions normally used to measure his approval, his underlying support remained very high. More specifically, while direct questioning suggested Putin’s approval was in the high 80s, the list experiments suggested it was closer to 80 percent.
Three points are worth noting from these findings.
First, the experiments suggest that Putin was indeed very popular among Russians at the time of the survey, and, perhaps even more importantly, that most of the support reported for Putin using standard questions to measure presidential approval was, at least in that point in time, genuine.
Second, there is some evidence suggesting that some Russians might be answering direct questions about Putin’s popularity in the affirmative when they may not actually like the job he is doing. The actual size of this bias could be affected by a number of different factor (see the paper for details and additional analysis), but it is likely more than 5 percent and less than 10 percent. So while this should not affect our overall conclusion that Putin is genuinely popular, it is interesting to observe that this is taking place.
Finally, as the authors of the paper note, the study has nothing to say about the depth of the support for Putin. It might be that the 80 percent of Russians whom the list experiment suggest support Putin are fervent supporters who will stick with him through thick and thin, and it might be that they are fair-weather supporters who will desert him at the slightest provocation.
Taken together, the findings suggest both that the West will likely be dealing with a popular leader when it interacts with Putin in the days to come, and, at least earlier in the year, that the headline approval numbers generated using traditional survey techniques provide a (fairly) good approximation of that support.