TL;DRussia Weekend Roundup
24 September 2022: Russia mobilizes, plus texts and a requiem
There are hours that contain days, and weeks that seem to last for years. This was one of them. I can hardly remember the Queen’s funeral — and it was only Monday. The march down the Mall, which, to my own surprise, I watched live on the BBC, has been firmly supplanted in my head by yet more Russians marching off to war.
I was lucky enough to have the opportunity to share some thoughts about Putin’s ‘mobilization’ — lengthy thoughts, actually — with the Washington Post and the New Yorker (any other week, an interview with the New Yorker would have put me over the moon), so I won’t rehash those thoughts here. And there was a Twitter thread on the same subject, which kind of blew up.
Interestingly, though, the Twitter thread only blew up in one of the two usual ways: while it got a lot of attention, it got almost no negative attention. An argument about how the draft is falling heavily on minorities and other marginalized communities inevitably involves some empathy with those communities — and empathy has been understandably in short supply of late. I was expecting to get mountains of responses along the lines of “they’re getting what they deserve”. I didn’t. I’m not about to claim that a sense of humanity has returned to Twitter, but even if it’s just for a few days, I’ll take it.
What I’m thinking about
The Washington Post piece I published on Wednesday was not meant to be about Putin’s mobilization of 300,000 reserves (or maybe 1.2 million) — that was pure serendipity. I had sat down on Tuesday to dig through a dataset that the independent Russian polling agency the Levada Center and the German political sociologist Heiko Pleines had posted for public access last weekend, making available for the first time (partial) raw data from Levada’s war-related surveys conducted in March through June of this year.
Consistently since the war began, Levada had been reporting that approximately 80 percent of respondents supported the war — a number that a lot of people questioned. A number, frankly, that I questioned. The objections were numerous: people opposed to the war were unlikely to respond to survey requests; wartime samples were unlikely to be reliable; whatever answers were given were unlikely to be truthful; and so on. Levada and Prof. Pleines released the data in order to give people like me (and especially people with better data skills than me) the chance to poke around and form our own conclusions — and that is a tremendous service.
The data, of course, don’t tell us anything about response rates, but Levada also published a series of experiments and validity checks, which suggest that, if there are shifting response patterns, they’re unlikely to be having much of an effect on the overall picture. As a result, I didn’t approach the data with any particular agenda, other than to see if I could dig up some kind of pattern that might go beyond what we all knew: that the vast majority of Russians say they support the war, and that younger Russians are somewhat more likely to oppose it.
I did, I think, find something — and that’s what went in the WaPo piece. In a nutshell, I dove into the questions that Levada asked people about their emotional response to the war, including positive emotions (pride, hope, happiness) and negative ones (fear, shame, depression, anger and shock) — but I was particularly interested in those respondents who said they had no particular emotional response to the war. Now, I have a bit of history with this kind of question: years ago, Max Ananyev and I drafted a paper (which we really need to publish) showing that non-response to politically sensitive survey questions in Russia can be a powerful indicator of underlying political discontent.
Lo and behold, the Levada data revealed the same thing: people who said they had no emotional reaction to the war were unusually likely to avoid answering the question about whether they supported the war. A bit more number crunching, and I found that young married men — and particularly young married men who expressed no emotion about the war — were biting their tongues significantly more frequently than anyone else in the survey. Bootstrapped regression analysis (to make up for the small numbers of young, married, emotionally stoic men in the survey) gave me the confidence I needed to make the argument publicly in the Post. It’s not hard to see the link between that finding and Putin’s mobilization drive.
But I’ve spent much of the rest of the week thinking about the bits of data that didn’t make it into the Post. Specifically, the fact that there are vanishingly few other reliable correlates of support for the war. Most surprisingly (to me, at least), material welfare doesn’t correlate at all. This is surprising not because I think that poorer or richer people should have different opinions, but because a ‘rally ‘round the flag’ effect should give people an increased sense of material welfare, as Graeme Robertson and I showed after the annexation of Crimea. In this case, that’s not happening. So, I need to go back and revisit my (admittedly superficial) conclusion that Russia is in fact experiencing a ‘rally ‘round the flag’.
That, though, is not the only thing giving me pause. Other than emotions and age, almost all of the other (limited) data available in what Levada published behave, well, oddly. Occupational status is the only other variable that has a statistically significant relationship to opinion about the war in all four months of data published. Gender is significant in May and June, but not in March and April. Marital status matters in March, April and June, but not in May. Education is only significant in May. And even for age, the strength of the statistical relationship fluctuates noticeably.
Someone smarter than me might be able to help me find other solutions, but for the moment I’m only seeing two possible stories. One is that the actual nature of sentiment about the war is changing from month to month, such that the kinds of people who supported it in April might not have supported it in June, and vice versa. That’s possible, I suppose, but not really plausible; at least, I can’t come up with a good theoretical reason why gender should be a significant predictor of support for the war in one month, and not in another. It’s the same war, and the same people.
Unless, of course, it’s not the same people — and that’s the other possible story. To be clear, I know that it’s not the exact same people: these survey’s aren’t what sociologists call longitudinal, and so the pollsters recruit new cohorts of respondents each month. But that recruitment is supposed to hit the same categories, and any individual differences are meant to be smoothed out by a probabilistic sampling strategy. If, however, response rates are in fact shifting and either the population or the sampling strategy have yet to settle on a stable pattern, then the relationships between these basic variables and approval of the war might be expected to fluctuate from survey to survey.
In short, I need to keep thinking, but my doubts haven’t gone away.
What I’m reading
Once Putin announced the mobilization of reserves, the only person I really wanted to read was Michael Kofman, and he didn’t disappoint. His initial analysis — on Twitter, of course — of the impact that the troop call-up is likely to have on the battlefield remains the clearest thing I’ve read on the subject thus far.
Searching for Michael’s tweets, meanwhile, brought me to an excellent podcast he recorded — before the mobilization announcement — with the inimitable Lawry Freedman, on what history teaches us about how Russia fights. (I know it doesn’t count as reading, but it’s not music, so it’s staying in this section. Sue me.)
What really caught my eye this week, though, was the analysis conducted by the New York Times of a trove of data leaked from one branch of Russia’s internet censorship agency, Roskomnadzor. The team at the Times did an excellent (computer-aided) job not only of sifting through mountains of data (even of those data are only a small part of the picture), but connecting those data to what happens in ‘real life’ — showing how online surveillance leads to harassment and arrests and underpins the entire system of fear and political compliance that helps keep Putin in power.
When you’re done reading that, click over to an earlier analysis of the same leaked data by the Russian news website Meduza — and note their finding (which I had forgotten) that the Kremlin’s surveillance of anti-war movements began almost two years before the war. Hmm.
Because war and authoritarianism aren’t depressing enough, I highly recommend taking the time — though maybe not on a Saturday morning — to read this report from the United Nations Institute for Disarmament Research on the impact that this war is likely to have on nuclear arms control and (non-)proliferation. Long-time readers (both of them!) will recall that I wrote back in June that this war had the potential to spur states around the world to seek and obtain their own nuclear deterrents. Lewis Dunn, a retired senior diplomat and the author of the report, concludes that the danger of that is real — if, happily, not as acute as I had worried. The best news in the report is Dunn’s conclusion that the war is not making the use of nuclear weapons more likely, and the idea that it might spur Washington and Beijing to sit down and talk. But the war’s overall impact on the already crippled nuclear arms control and disarmament process will most likely be nothing short of devastating.
What I’m listening to
The only music appropriate for this week is a requiem — and for me, it’s Gabriel Fauré’s, performed by the Cambridge Singers, under the direction of John Rutter.