The causes of the 2008 economic collapse don't seem in much doubt. There was a financial crisis, and there was a housing boom and bust. But those events, dramatic as they were, don't account for the depth and duration of the slowdown.

The question is why the initial fall in demand persisted as long as it did. During the recession, the level of nominal gross domestic product in the U.S. fell on a year-over-year basis for the first time since 1958 and more severely than at any time since the Great Depression. The measure remains somewhere between 5 and 15 percent below its pre-2007 trend, depending on how you define it.

Why would such a huge shortfall in nominal GDP unleash economic chaos? The answer gets to the heart of what a demand-driven recession is: a coordination failure. Disruption and dislocation result from what economists call "nominal rigidities." Prices and costs set in nominal terms are "sticky": They don't adjust instantly, and the result is losses, firings and unemployment, canceled investment and debt-service difficulties.

(David Beckworth, an economics professor at Western Kentucky University, explains the specific pathways that propagate these nominal shocks in a blog post.)

Expectations play a critical role in all this. As nominal GDP fell, forecasters surveyed by the Federal Reserve Bank of Philadelphia decided that further falls were likely. Before the recession, the forecasters thought nominal income would grow at an annual rate of about 5 percent. At the trough of the recession in the winter of 2009, though, expectations sank: The average forecast was for a 4 percent drop in the next quarter. More than half thought it would fall for at least one more quarter. Three-quarters thought full-speed growth wouldn't return for more than a year and a half. As for getting back on trend, just about nobody saw that happening. And they were right (see graph).

These collapsed expectations were at least as much a problem as the initial fall in nominal GDP. They led most businesses to think they couldn't ride out the losses. So they moved aggressively to cut costs, greatly magnifying the initial shock.

With souring expectations came soaring uncertainty. In 2007, forecasters thought alike about where nominal GDP was headed: Half of their estimates fell within 0.8 percentage points of one another. By 2009, the dispersion had tripled. This deepened the recession still more by forcing businesses and households to guard against worst-case scenarios -- through flight-to-safety in capital markets, additional layoffs and higher household saving.

A crucial question, therefore, is how much influence economic policy makers can have over expectations of nominal GDP. Monetary policy, done right, has a lot -- more influence, say, than current-quarter nominal GDP. Even if the Federal Reserve couldn't have avoided the initial crisis, it could have kept the recession shorter and shallower if it had done a better job of anchoring expectations and curbing uncertainty.

Lately, it has been doing just that. Since the Fed clarified its commitment to economic recovery, nominal GDP expectations have normalized and uncertainty has fallen sharply.

Managing expectations and uncertainty are the keys to successful monetary policy.

(Evan Soltas is a student at Princeton University and a contributor to the Ticker. Follow him on Twitter.)