From the responses to my remarks last week on monetary policy, I see that my words were interpreted by some readers differently than I’d intended, for which I apologize. Let me try again.
Let me begin by summarizing the two main points I was trying to make. First, I pointed out that, in the current environment, large-scale asset purchases by the Fed are essentially a swap of short-term for long-term government debt. I asked whether those who criticize the Fed for being too timid would also criticize the Treasury for likewise being too timid in keeping so much of its borrowing long term. I answered for them (and on this I did not see any of the Fed’s critics dispute my proposed answer) that there were clear fiscal management dangers if the Treasury were to aggressively move the federal debt into extremely short-term securities. I therefore suggested that if we would see some potential danger associated with such a strategy if adopted by the Treasury, then we should also see some danger about such a strategy if adopted by the Fed.
Second, I pointed out that the direct stimulative effects of a debt maturity swap were decidedly minor.
The conclusion I draw from these two observations is that we might have to push on this lever extremely hard to get anything accomplished, and that pushing on the lever is not without its own dangers. My position is therefore that the Fed is correct in viewing this particular tool as one that should be used with caution.
Now for some of the responses. Brad
DeLong observes that although Federal Reserve deposits at the moment are essentially equivalent to short-term Treasury obligations, once we get away from the liquidity trap they will not be. Insofar as some of today’s created reserves turn into M1 instead of T-bills at that future date, Brad argues, the time path would be more stimulative and, through an expectations effect, that could make a difference today. This is in fact my view as well, and one of the reasons why I insist that the main mechanism by which large-scale asset purchases could make a difference is not through their direct mechanical consequences for interest rates, but instead comes from their value in sending a credible signal for future Fed policy.
And that’s why I think the main task for the Fed is not to decide how much LSAP we need, but instead to articulate a clear framework within which those measures are implemented that is also mindful of the particular fiscal management challenges that these policies tie us into. I suggested that the current policy, which I read as not allowing inflation to fall below 2%, works well for both objectives.
Scott Sumner is dismayed at my focus on the concrete steps the Fed would need to take to implement more stimulus. Scott is a strong advocate of the view that if the Fed were simply to announce a particular target, everything would fall in place to make sure it happens. For clarification, I fully agree with Scott and Brad that it would be possible for the Fed to achieve a higher inflation rate than 2%. My concern is with the robustness of the particular strategies for implementing such an objective. In particular, the fiscal management problem I am referring to is one of a rapid flight from dollars and loss of confidence in the Fed and Treasury. As I explained in my original post, this is the exact opposite of deflation. A strong Fed commitment to avoid deflation should not provoke market fears about the ability of the Treasury and Fed to manage their short-term liabilities. The question on which I and some of the Fed’s critics may differ is whether there is a similar robustness to the Fed’s announcing that what they’re trying to produce is, say, 3.5% inflation. My fear is that the practical instruments available to the Fed are too blunt, and the stability of the expectations equilibrium too tenuous, for us to be confident that the Fed could manage a multi-trillion dollar balance sheet if financial participants develop sudden doubts about how it will play out.
Notwithstanding, I agree that there is room for the Fed to look for a more expansionary objective. To repeat my earlier conclusion:
Perhaps there is a clear way to communicate an alternative, more ambitious goal, such as keeping nominal GDP growth above 5%, or temporarily focusing on getting unemployment down to 7%. If articulated narrowly and with some caution, these might allow the Fed to do more while still preserving confidence in what I have described as the logistics of managing potentially volatile short-term government debt.
But unlike many of my fellow academics, I worry about those logistics and am convinced that it is a mistake to ask too much from monetary policy.