I'm tempted to conclude this is something that happens so rarely you may as well count the times it happens as a rounding error.

It's a bit like saying government will just work if we only vote for the right candidate.

Everyone's tempted to conclude things like that, but there's a subtle fallacy happening when we do. Generally speaking, we notice government programs of all stripes when they get in our way. When they don't get in our way, we don't. It's very easy for someone paying attention to the news to be able to rattle off incidents where the FDA and FEMA and the SEC screwed up, because the screw-ups are in the news. But when they do their job correctly we never hear about it. Basically, the dataset we carry around in our heads about How Well Government Works is substantially biased toward instances of failure.

The only way to really measure how useful a regulation is or isn't is to determine what it is the regulation is supposed to address and examine a substantial set of "before and after" data points; trotting out vivid but essentially anecdotal examples of WHEN REGULATORS GO BAD makes for great editorializing, but it doesn't actually tell us very much.