Eugene Volokh has an intriguing post about unreliable assurances, with examples that speak for themselves.
I'm looking for examples where (1) the opponents of some proposed law, constitutional amendment, or judicial decision argued "this action will be interpreted in this particular bad way" or "this action will set a precedent that will be used to reach this particular bad result," (2) the supporters assured the public that no, of course this won't happen, and (3) some time down the line — preferably no more than 50 years, just to avoid especially hard questions of causation — the foretold result did take place, despite the supporters' reassurances.
The examples he cites, and many others, explain why libertarians are so wary of even tiny baby steps towards rights restrictions -- the so-called "slipperly slope" argument. Professor Volokh's post is particularly interesting because he's often a harsh critic of slippery slope arguments, rightly pointing out that slopes often can be avoided. However, as these examples demonstrate, they often are not.
This fear of a slippery slope is why I oppose gun registration laws -- registration is a clear first step towards a ban and confiscation. Registration is how, gasp, the Nazis and Communists began disarming their subject people. Likewise, I'm sympathetic to those who oppose greater police power designed to combat terrorism even though, yes, the power can and will be misused at some time. I've tried to account for that inevitable occasional misuse, however, in my decision to support some controversial laws, such as the Patriot Act.