Milton Friedman is probably the most famous proponent of the view that appropriate monetary policy can always end recessions. But even he took the view that monetary stimulus works with "long and variable lags." That's never made a ton of sense to me.
Once you stop thinking of monetary policy as being about trying to stabilize the supply of base money and instead being about trying to shape nominal expectations, you'd expect the impact to be felt very quickly. Firms and individuals don't make decisions instantaneously, and there are implementation lags even once decisions have been made, but if you're trying to shift expectations up to higher equilibrium, it'll either happen quickly or else it won't work. Since no central bank does conduct monetary policy by trying to stabilize the monetary base, talking about "long and variable lags" sounds like an effort to evade accountability.
So how are we doing?
Pretty good it seems. September auto sales reached their highest seasonally adjusted level in four years scoring a 13 percent year-on-year increase and beating even the "cash for clunkers" surge. And the ADP report out this morning says there was a net gain of 162,000 jobs in September (PDF). Now that's all preliminary and maybe it won't hold up. But this—rather than long and variable lags—is what I'd expect to see from QE 3. An unusually strong fall as durable goods orders and construction starts surge for a few months, pushing the economy up to a higher level before returning to its more sluggish trend.