One thing I have noticed is that just based on my wading through economic cycle data over the years I have tended do discount the notion of economic confidence as irrelevant and concepts like multiple-equilbria-in-macroeconomic-aggregates as completely unlike anything we observe in the real world.
Yet, my fellow economists, particularly older ones take the notions quite seriously.
Michael Forli suggests that times may have changed
A final, somewhat technical, implication is that the conditions for self-fulfilling prophesies in the macroeconomy may no longer exist. The idea that there could exist virtuous/vicious circles between real economic outcomes and confidence (or asset prices, effectively the same thing) was first formally advanced by David Cass and Carl Shell. They referred to these virtuous/vicious circles as “sunspot equilibria;” George Soros dubbed the property “reflexivity.” Subsequent applied macroeconomists appealed to the procyclicality of productivity as evidence of the type of increasing returns to production sufficient to generate confidence feedback loops. If labor is no longer a quasi-fixed factor of production this may eliminate one type of non-convexity in production, thereby reducing the likelihood that the economy has multiple equilibria and is subject to self-fulfilling prophecies. While it is hard to say much definitively, it is interesting to observe that over the last two years the economy has been subject to large swings in investor sentiment and asset prices, and yet actual growth outcomes have been remarkably stable.