The “2008 financial crisis” and the accompanying “great contraction” have severely challenged the prior consensus about the risks of such events. Before the recent turmoil, many economists were convinced that our understanding of economic fluctuations, and of policy makers’ ability to manipulate the economy, made the risk of major downturns all but irrelevant. They were wrong.

The crucial lesson for the future is therefore humility: whatever we think we “know” is far less certain than most textbooks and policy pronouncements presume. This does not mean we know nothing useful; but we should recognize that our current “understanding” might have major flaws. More than 80 years after the Great Depression, for example, controversy still rages over the causes and consequences of that episode.

The principal reason for our imperfect understanding is that macroeconomists and policy makers do not get to “run experiments” with the economy, and that is the only way to learn its true structure and determine how policies affect its performance. Stated differently, we never get to observe the counterfactual in which Fannie and Freddie do not expand mortgage lending, the fiscal stimulus does not take place, the Treasury does not undertake TARP, or the Fed does not engage in quantitative easing. So, maybe these policies helped, maybe they hurt, or maybe they did some of both; we will never know with great confidence.

How should economists and policy makers incorporate this inevitable uncertainty into their actions? It is hard to formulate a general rule, but one lesson seems clear: neither group should provide strong assurances that they have it all figured out. Some consumers and businesses will respond by doing more to protect themselves (for example, by saving more for that rainy day), and this will reduce the risks of the worst-case scenarios.