This works until something breaks the spell

The concept of a preference cascade seems to be getting broader discussion. It is a subtle but I believe real phenomenon. It was originally described in 2002 by Glenn Harlan Reynolds in Patriotism and Preferences.

The muting of open patriotism after the Vietnam era may have been a case of what social scientists call "preference falsification": One in which social pressures cause people to express sentiments that differ from those they really feel. As social scientist Timur Kuran noted in his 1995 book Private Truths, Public Lies, there are all sorts of reasons, good and bad, that lead people not to show how they truly feel. People tend to read social signals about what is approved and what is disapproved behavior and, in general, to modify their conduct accordingly. Others then rely on this behavior to draw wrong conclusions about what people think, and allow those conclusions to shape their own actions.

[snip]

This illustrates, in a mild way, the reason why totalitarian regimes collapse so suddenly. (Click here for a more complex analysis of this and related issues). Such regimes have little legitimacy, but they spend a lot of effort making sure that citizens don't realize the extent to which their fellow-citizens dislike the regime. If the secret police and the censors are doing their job, 99% of the populace can hate the regime and be ready to revolt against it - but no revolt will occur because no one realizes that everyone else feels the same way.

This works until something breaks the spell, and the discontented realize that their feelings are widely shared, at which point the collapse of the regime may seem very sudden to outside observers - or even to the citizens themselves. Claims after the fact that many people who seemed like loyal apparatchiks really loathed the regime are often self-serving, of course. But they're also often true: Even if one loathes the regime, few people have the force of will to stage one-man revolutions, and when preferences are sufficiently falsified, each dissident may feel that he or she is the only one, or at least part of a minority too small to make any difference.

This relates to a concept in economics (and I think in sociology/psychology as well), revealed preference. People talk a good game about what they value and what they want. But if you really want to know what they value, look at what they do. There are all sorts of reasons why there is a disconnect between stated intentions and actions and the dynamics behind a preference cascade is one special case of the general truism.