The Great Depression: demand or supply shocks?

The fact that we are in a big recession has renewed interest in the Great Depression, and this has revived the questions about its origin. In particular, the eternal question on whether demand or supply shocks have driven it is back.

This time it is asked by Mark Weder, who runs a horse race between tow versions of a real business cycle model: one with only shocks to total factor productivity (supply shocks) as measured by Solow residuals, one with only preference shocks (demand shocks), measured as residuals of an Euler equation. The latter, though, are not associated with monetary or fiscal variables. Both types of shocks are the evaluated in their ability to forecast what happened to GDP, and none is a clear winner.

But is this really the best one could do? Clearly the models are way to simple to 1) forecast anything, 2) to capture the changing policy environment during this period, as highlighted by Milton Friedman, Anna Schwartz, Harold Cole, Lee Ohanian and many others. Also, the relative importance of the two shocks may have shifted over time, something that would have been worthing looking at.