Is Bernie Sanders a Crypto-Communist? A Bayesian Analysis
By Bryan Caplan
The word “crypto-communist” has a paranoid, McCarthyite connotation. But during the Cold War, numerous communist intellectuals and politicians deliberately concealed their commitment to Marxism-Leninism. Why? To be more successful intellectuals and politicians. A few crypto-communists even managed to become national leaders. Fidel Castro gained power in 1959, but only announced his communism in 1961. Nelson Mandela presented himself as a reasonable democratic reformer. Yet after his death, the African National Congress openly admittedly that Mandela had been on the politburo of the South African Communist Party for decades. Ho Chi Minh joined the Communist Party in 1920, but in 1945 he loudly posed as a moderate democratic reformer – famously quoting the U.S. Declaration of Independence to charm the West. Juan Negrin, last prime minister of Republican Spain, was also very likely a crypto-communist.
Which brings me to my question: What about Democratic presidential candidate Bernie Sanders? Is he a crypto-communist? Sanders has sent decades worth of frightening signals – praising Soviet bloc regimes, honeymooning in the Soviet Union, and such. Indeed, he’s said and done almost exactly what you would expect a sincere Marxist-Leninist who wanted to be a U.S. Senator would say and do. Note, moreover, that Sanders came of political age during the 60s and 70s, when communism made a big comeback in the U.S. radical left.
True, this hardly proves that he’s a closeted communist. Alternately, Sanders could be a communist dupe, or a even a true believer in “finding the good in the bad.” The upshot: We have to settle for a probability that Sanders is a crypto-communist, all things considered.
When constructing such probabilities, Bayes’ Rule is usually helpful. As you may recall, the Rule states that: P(A|B)=P(B|A)*P(A)/[P(B|A)*P(A) + P(B|~A)*P(~A)]. In this case, we want to know the probability that (A) Sanders is a crypto-communist given (B) his track record. Piece-by-piece:
1. What’s the probability of Sanders’ track record if he is a crypto-communist? Here, I’d go high. Most crypto-communists in Sanders’ position would be look like him. I give this 75%.
2. What’s the probability of Sanders’ track record if he isn’t a crypto-communist? Sanders view have long been extremely unpopular, but quite a few non-communists on the radical left would have shared them. So I’ll give this 1.2%.
3. What’s the prior probability of being a crypto-communist? Even during the 60s and 70s, this would be low, but not astronomically low. .3% seems plausible.
4. What’s the prior probability of not being a crypto-communist? 100%-.3%=99.7%.
Plugging in to Bayes’ Rule, I get 15.8% – a low but hardly negligible risk that Sanders is a totalitarian hiding in plain sight. Needless to say, you can alter this final estimate by fiddling with the value of the numerical components. But you’d have to change them a lot to get the probability below 5%.
Which brings us to a big related question: When does the risk of crypto-communism become disqualifying for a presidential candidate? I say even a 1% chance should be totally disqualifying, but I fear that most Democrats – and many non-Democrats – will demur. So what risk would they consider acceptable? 5%? 10%? I don’t know, but plausibly revising (1)-(4) to get below a 5% or 10% threshold is no easy feat.