Rather than commenting on the credibility of that in Elster’s specific case (which would depend on knowing more than I do about Elster and about the anti-communists he paid attention to), I’ll remark that there certainly are cases in which most of us here would do likewise. (Not literally zero credit, but extremely little, which I think is also what Elster’s doing.)
No, you give them appropriate credit for their correct predictions, and and appropriate de-credit for their incorrect predictions.
You shouldn’t give credit or discredit directly for correctness of predictions, if you have information about how those predictions were made. If you saw someone make their guess at tomorrow’s Dow Jones figure by rolling dice, you don’t then credit them with any extra stock-market expertise when it happens that their guess was on the nose; they just got lucky. (Though if they do it ten times in a row you may start to suspect that they have both stock-market expertise and skill in manipulating dice.)
No, you give them appropriate credit for their correct predictions, and and appropriate de-credit for their incorrect predictions.
You shouldn’t give credit or discredit directly for correctness of predictions, if you have information about how those predictions were made. If you saw someone make their guess at tomorrow’s Dow Jones figure by rolling dice, you don’t then credit them with any extra stock-market expertise when it happens that their guess was on the nose; they just got lucky. (Though if they do it ten times in a row you may start to suspect that they have both stock-market expertise and skill in manipulating dice.)