Scientific and other knowledge could be typically a sooner rather than a later casualty of a deep crisis. What remains might be often a very non-representative part of the knowledge existed: a few baffling mathematical or astronomical observations may stay around, but some simple technological tricks would strangely look missing. How often that occurred through the history? How much can Occam's razor help us to know?

As DeAnander tells, much knowledge is lost in "improving" times as well. One speculative example is Lucio Russo's hypothesis that Helleniac greeks already had a deep understanding of science methodology, but it get lost amid Roman "efficiency".

Booming times are in fact ordained predecessors of busts. Collapses happen under special conditions, and any unusual flourishing very likely has roots in removed brakes against the special conditions. Say, the stock market has higher risk of collapse precisely when it is growing beyond expectations.

In economic jargon, the Western (or capitalist) civilisation is highly leveraged one - it is performing better than any social arrangement most of the time, but there is a "systematic" risk of meeting growth boundaries. Here we can see a conflict of empirical induction and Popper's falsification logic: people tend (or are persuaded) to follow behaviors that appear to be most successful currently, even if those behaviors cannot possibly give the same success to everyone. Eventually, one has to take risks to be most successful. But statistically, the same risks are rewarding to a predictable portion of participators. Or worse yet, the flourishing conditions for "everyone" are destroying themselves under growth. And once conditions unnoticeably change, a "rare event" arises, unpredictable from the empirical time series of merry growth. Then your usual compulsion can hurt and bankrupt you.  

Of course, the "rare events" can be deduced from basic knowledge of the world. But purely logical deduction is dull and inferior to empirical induction in practical importance most of the time. Even if you suspect bad turns, you are pressed not to fall behind in "performance". Staying with the herd looks safe enough.

Most of the institutional investors who thought that risk was mispriced were nevertheless reluctant to invest on that view because of the cost of carrying that trade. Since virtually all such institutional investors are agents and not principals, they could not afford to take a position that involved a series of short term losses. They would appear to be better investment managers by focusing on the short term gains that could be achieved by going with the herd to enhance yield by assuming increased credit risk.