I find this extremely plausible. Most homo sapiens of my acquaintance are notably unenthusiastic about being reminded of their past misjudgments. However, Henry Farrell demurs:
I'd be prepared to bet a significant amount of money that the number of people who believe that they supported the war back in 2003 is far lower than the number of people who actually did support the war back in 2003. Indeed, I suspect that the number of people who believe that they supported the war back in 2003 is a minority of the US public. Since the Cassandra-backlash effect that Tabarrok is talking about is contemporaneous, and presumably depends on people's current beliefs about what they thought in the past, this makes me think that something else is going here (and that this something else has to do with the desire of elite actors in the commentariat to hold onto their privileged position in the public discourse).I find this extremely plausible too. It's like asking people if they voted in the last election: 80% say they did, even in elections where we know for a fact that the turnout level was only 60%.
But this is testable, no? In March 2003, most polls showed support for the war at about 65% (though it depended somewhat on how the question was worded). If you asked people now about their original view, how much lower would that number be? 10 points? 20? Zero? Maybe the next time Gallup or the New York Times or some other pollster does one of those massive 100-question surveys that they sometimes do, they should find out.