I've at times complained about the Reverse Mussolini Fallacy. The Mussolini Fallacy is believing that, because Mussolini made the trains run on time (if he did), that excuses his other acts. The Reverse Mussolini Fallacy is believing that, because Mussolini made the trains run on time, making the trains run on time is bad.

A commenter to an earlier post writes:

Boston was one of the first, if not THE first, city to adopt the language suggested by privately produced DHS guides: "If you see something, say something." Apparently the comparison to Maoist China or Stalinist Russia does not worry these people. Reporting crimes is one thing. Reporting suspicions is quite another.

It's true that totalitarian regimes rely on citizens to report misconduct by their fellow citizens. But so must democratic and liberal regimes: The police can't be everywhere (and shouldn't be everywhere), so they rely on citizens to alert them about what they should investigate.

Nor is the "reporting crimes / reporting suspicions" distinction sensible here. Sometimes -- often -- a citizen observes something that isn't clearly a crime, but that seems suspicious, which is to say that might be evidence of a crime. You might hear screaming from your neighbor's house, but not be sure whether there was a crime or just a nasty argument. You might see someone pull a kicking, screaming child into a car; is that a kidnapping or just someone dealing with a tantrum? You might see someone leave a backpack in a subway; could that be a bomb, or did someone just forget his stuff? You often can't investigate further on your own -- it may be too risky, or outside your competence. Of course you should call the police.

Now there may well be times when you shouldn't call the police. The most obvious example is if the offense shouldn't be a crime in the first instance (if you tattled on a neighbor for saying anti-government things in Soviet Russia, that would be bad even if the statements were the crime, so you were "reporting [a] crime[]" rather than a "suspicion[]"), or if you think the punishment is likely to be grossly disproportionate to the crime. Sometimes demands of loyalty to friends or family may outweigh the demands of preventing harm to possible future victims of those friends or family; that's a really tough call, but I do suspect that -- rightly or wrongly -- I wouldn't turn in a family member for petty shoplifting. If you think that the police are bad enough, you might not want to help them at least unless the crime is very severe. And finally, if you have only the vaguest of suspicions, the possible crime isn't very severe, and you're afraid that the costs of bringing in the police (the hassle to the likely innocent target, plus the waste of time for everyone) exceed the benefits (the possibility that the target is indeed guilty), that might be reason not to call.

But the distinction can't, I think, be simply reporting crimes vs. reporting suspicions; nor can I fault an "If you see something, say something" program, at least unless the program is chiefly aimed at genuinely innocent conduct. (Critics of the war on drugs may fault such programs on this latter ground, but that's a separate argument from the one the commenter was making.) Helping the police catch criminals is generally good, and urging the public to do this is good, too. It becomes bad chiefly to the extent that the law being enforced is a bad law; and if that's so, then that should be the target of our argument, not the attempt to promote citizen reporting of crimes as such.