Hypothetical possible evidence that QE2 didn't work
1) disappointing growth roughly half a year after it is introduced: Check.
2) particularly disappointing investment growth: Check.
3) Especially particularly disappointing housing investment growth: Check.
4) The price of the asset purchased fell when the purchases began: Check.
5) The price of the asset purchased fell from the time before the policy is mentioned and after it is implemented: Check.
6) medium long real interest rates rose when the purchases started: Check.
7) medium long term real interest rates rose from the the time before the policy is mentioned and after it is implemented: They are almost identical.
8) Expected inflation fell over the same period: it increased.
But how can monetary policy affect expected inflation except via real interest rates ?
The pattern seems to me to fit the story that some investors believed that QE2 would work (very likely as many economists did). Then those who put their money where their mouths were discovered they were wrong. Those who just talk kept talking. Investors bought on the rumor and sold on the event. This can happen for many reasons. One of them is that the event has no effect on fundamental values and that this becomes clear when it happens.
On the other hand, stock prices rose. In standard models this is the effect of reduces long term real interest rates (not observed) and increased growth which is the effect of reduced reale interest rates (neither observed).
Consumption grew robustly. In standard models this is the effect of rapid GDP growth which is the effect of ...
Ways in which money supply siders are like supply siders.
They make vague arguments. They say there are many ways that the Fed can affect the economy other than interest rates. They don't say how. They appeal to math but don't write models down.
They assume that all economic agents agree with them and respond quickly to changes in policy.
In data analysis the beginning of the policy they advocated is timed at a trough. There is no justification for the timing. For the QE2ers the trough is a low growth rate not a low level. The acceleration of GDP growth from 2010 Q2 to 2010 Q3 is ascribed to a policy shift which is dated at the very earliest in late 2010 Q3.
"Growth accelerated, from a 1.7% annualised pace in the second quarter to 2.6% in the third quarter and 3.1% in the fourth quarter. "
Avent describes relatively fast growth in 2010 Q3 as evidence that QE2 worked. Is anyone willing to look me in the eye (via Skype) and assert that he would have hesitated for a second to claim that 2010 q3 growth was the standard to which post QE2 growth should be compared if 2010 Q3 growth had been slow ? How is this different from supply siders dating the Kemp Roth Tax cut to 1982 ?
Supply siders made a prediction about a variable and claim it was confirmed by observation of another variable. They claimed that tax cuts would cause an anomalous increase in tax revenues. They then argued that rapid growth (timed from the trough of a depression) confirmed their prediction.
Money supply siders claim that increases in stock market indices and in consumption confirm their predictions about long term real interest rates and investment. Both use any datum to hand.
Oh note that Martin Feldstein is a member of both camps. He doesn't seem to notice any difference.
The policy recommendations are very detailed. Total failure is ascribed to the fact that actual policy as implemented was not identical to policy as recommended.
"Unfortunately, the Fed didn't set a policy target. Instead it opted for the $600 billion figure, thereby inviting critics to judge the Fed on whether $600 billion was, in fact, the right action to take."