I was chatting with a friend about the state of the economy, and I brought up how the monetary expansion that began in 2008 wasn’t actually stimulative because the Fed started paying interest on excess reserves. My friend expressed a little skepticism that excess reserves mirrored the monetary expansion one-for-one. But behold:
No, it’s not literally one-for-one, but it’s close. Non-circulating money gives me the vague impression of a tree falling in the woods with no one around to hear it.
Here is a case where your eyes can deceive you. From 8/08 to 10/10, MB went up by $143 Billion more than excess reserves. This translated into a $364 Billion increase in M1, or 26% over roughly the past two years. The YOY changes in 2009 are the largest in the series going back to 1959, with the 18% increase to June 2009 being the single biggest one year increase in the money supply.
Jeremy, here is a graph of the base minus excess reserves. I agree that it’s clear that the base went up by more than excess reserves.
However, if you take the pre-2005 trend in the monetary base and extrapolate it to 2010, you end up almost exactly where we are now. There is nothing stimulative about this; circulating hard money has grown at trend for the last five years, though not smoothly. Indeed, I believe that the dip in base money growth below trend in 2006-2007 was one of the reasons for the recession, or at least for why the recession happened when it did. But the Fed needs to take circulating hard money above trend growth for it to count as stimulative, in my opinion.
Sure, I’ll grant that the growth of both MB and M1 was slow from 2005-2007, and this was due in part to tightening beginning in mid-2004. But, in my opinion, it seems overly restrictive to not call policy since 2008 stimulative. Is this semantics or economics? I’m not sure I can tell the difference anymore.
Anyway, eyeballing M1 it is now back to the 2000-2004 extrapolated trend. M2 looks pretty close to. But perhaps my eyes are deceiving me this time.
Like you, I don’t want to get lost in semantics, and it sounds like we don’t really disagree about any of the facts. In my view, stimulative monetary policy means getting the base above trend to push NGDP back to trend. But I can see how people might use “stimulative” to mean faster-than-trend growth even if it doesn’t mean taking the base above trend.
In any case, I still found the graph in my post pretty striking, because if all you see is the blue line, you assume that policy has been unbelievably stimulating. The red line tells an important story.
Yes, I agree with you last point. Seeing the MB spike alone is misleading. But once we know both MB and ER have spiked, it might be useful to look at what happened to other aggregates, since this is money actually being used by the public and likely to impact prices.
Which brings me to my disagreement with you (or rather, my confusion). Which monetary aggregate should we focus on? I’m confused why we should care about MB, especially looking backwards. Don’t we care more about M1 and M2, when judging stimulative-ness? We don’t know in advance how much a change in reserves will affect the money supply. But we do know after the fact, like, now.
I think there are a couple reasons to prefer narrow definitions of money when doing monetary economics. First, the narrower your definition, the more true it is that the government has a complete monopoly on the supply side. So if we take the government to completely control M0, we can consider differential changes in M1 and M2 exclusively on the demand side, and we have a nice, tractable model of supply and demand. If you’re primarily considering M2, say, then the government does not control the supply, and it becomes fuzzy whether a particular change is supply-side or demand-side. It gets very confusing very quickly.
Second, monetary economics is a theory of the price of the medium of account, not of the medium of exchange. People use all kinds of things as media of exchange, including T-bills in some large financial transactions. T-bills have some money-like properties, but only narrower definitions of money are constant-value in terms of the medium of account. Even money market funds can “break the buck.”
It’s not that M1 and M2 are not important; I just prefer to think of them as demand-side epiphenomena. It’s cleaner that way. You can do coherent monetary economics using M1 or M2; it’s just much more complicated.
Eli, I am in agreement with much of what you say here. The Fed has direct control over MB, but only indirect control over broader monetary aggregates. So in some sense, how stimulative their stance is should be viewed through MB.
On the other hand, M1 is going to be much more closely related to P from the quantity theory (I assume, but can never find the relevant literature). And in the Sumnerian view, stabilizing the growth rate of PY is key (am I right here?). So if I’m right on these two counts, in another sense how stimulative monetary policy is may be better viewed through M1.
Now, how a given change in MB impacts M1 and consequently P is unknown a priori. But in the end, it would seem to be what matters most. If The Fed is bad at targeting PY (because they’re bad at targeting M1), the case for trying to do it becomes weak, in my view.
If by “more closely related to P” you mean that velocity of M1 is more stable than velocity of the base, then I think that is true. I’m willing to concede that broader monetary aggregates are better statistical predictors of PY, and that the Fed should use a variety of aggregates if it wants to understand the monetary economy.
What I think that the broader aggregates tell us is what is happening to money demand. If MB goes up 10% and M1 goes up 5%, then we know that demand for the base is going up. Unlike in other markets where you can’t empirically decompose supply and demand, in the money market you can because we have these multiple simultaneous observations (of very close substitutes). A complete picture of how stimulative monetary policy is depends on how appropriate a change in the supply is for what is going on with demand.
On the targeting point, I think the Fed can hit pretty closely any single target it wants. The problem is that we don’t know what they’re targeting, and for all I know it is something like bank profits. I think the recent failure of the Fed to keep anything stable is an indictment of discretion, not of targeting.
Your comment that The Fed may be targeting bank profits is a very public choice view, and of course one that I’m inclined to agree with. But if this is true, I don’t see why we’d expect The Fed to behave like Scott Sumner or anyone else wants without major institutional reform.
I think that merely announcing a target provides a sort of minimal constraint. New Zealand’s central bank adopted an explicit inflation target several years ago and that has worked out well.
Even if an explicit target doesn’t work, I’m not sure what sort of major institutional reform is likely to work. I’m all for free banking, if there is a way to get there from here. The problem is that institutional reform goes through Congress, which is even more of a public choice quagmire than the Fed is.