Taking Seriously the Excess Money Demand Problem: A Reply to Robert Murphy
[David Beckworth offers this reply to Robert Murphy:]
Robert Murphy recently questioned my claim made in a National Review article that there is a conservative case for QE2. He finds my premises and conclusion wanting. Unfortunately, in his analysis Mr. Murphy misses and mischaracterizes some important points I made about QE2. His article, therefore, does not truly represent to readers of Mises.org the case I made for QE2. This note provides as an opportunity for those interested to see what I actually argued. While I do not expect everyone to agree with me, I do hope it will allow for a more productive conservation among all of us about the nature of monetary problems.
Before addressing the concerns raised by Mr. Murphy, let me summarize the key argument I made in my article. During 2008 there emerged a surge in money demand as the housing fiasco began to unfold. This spike in money demand got even more pronounced in late 2008 with the uncertainty created by the financial crisis. Given that we have a central bank — and this is not an endorsement of the Fed — its job should be to offset and stabilize such money demand shocks. The Fed failed on this count and, as a result, what should have been an ordinary recession got turned into the “Great Recession” of 2007-2009. Yes, this Fed failure — like its failure to raise the federal funds to its natural rate level sooner in the 2002-2004 period — is another indication the Fed is flawed. Nonetheless, we are stuck with this monopoly producer of money and have to work with it. This means the Fed should have done more to prevent the surge in money demand. Because it did not, the Fed effectively tightened monetary policy in 2008. Moreover, despite the large increases in the monetary base to date, money demand remains elevated. From this perspective, then, monetary policy is still relatively tight. QE2 is an attempt — a flawed one as I will discuss later — to address it.
So how do we know there is still an excess money demand problem? For starters, various money velocity measures — which indirectly reflect money demand — have fallen and remain depressed as seen in the figure below:
This unchecked decline in velocity is why total current dollar spending has not returned to trend as can be seen here. Another indication of continued excess money demand comes from flow of funds data. This data shows that for the combined balance sheets of households, non-profits, corporations, and non-corporate businesses the percent of total asset that are liquid ones (i.e. cash, checking accounts, saving and time deposits, and money market funds) is still relatively high. In other words, households and firms are still holding a disproportionate share of their assets in liquid form. This can be seen in the figure below:
Since this excess money demand problem is central to the argument made in my article, I found it odd that Mr. Murphy not even once grapples with this issue in his critique. I am not sure why he has ignored it, but the fact is that there are two sides to value of money: supply and demand. Once one realizes this and its implications for monetary disequilibrium — even a pure gold standard can be distortionary if there is sudden spike in money demand — then one has to make a more guarded argument for a commodity standard. Appreciating the importance of money demand shocks also helps explain why conservative economists like Scott Sumner, Bill Woosley, Josh Hendrickson, and I are sympathetic in spirit (if not in form) to QE2. It would do all hard-money advocates some good to wrestle with the monetary disequilibrium literature and its implication for a commodity standard. It is worth noting that there are prominent Austrians like George Selgin and Steve Horwitz who take the monetary disequilibrium seriously.
Let me now turn to the specific concerns raised by Mr. Murphy in his article. His first one is that I have failed to check my premises. Here is Mr. Murphy:
Beckworth gets into trouble early in his article, when he says matter-of-factly, “One reason for this confusion is a failure by some conservative commentators to understand the real purpose of QE2. It is not solely about lowering interest rates, increasing bank reserves, and encouraging bank lending, though all of those will occur.”
I grant that QE2 will increase bank reserves. However, the other two alleged results — lower interest rates and more bank lending — are not so obvious.
The phrase “It is not solely about…” in the above excerpt of mine seems to have been lost on Mr. Murphy. It implies something else will happen. Thus, on the issue of interest rates I actually said that interests will start increasing if QE2 is successful later in the article:
Note that lower long-term interest rates are not the key to QE2 working. Yes, long-term interest rates may initially drop as the Federal Reserve buys up long-term Treasury securities to increase the monetary base. But this effect will be fleeting if QE2 is successful. Once the economy starts recovering, interest rates will start increasing.
Somehow Mr. Murphy missed this and argues that that I said QE2 will only cause interest rates to fall. As regards to bank lending, I also note later in the article that as QE2 unfolds and address the excess demand problem, an economic recovery will take hold and lead to more lending. Though there is some evidence that interest rates are doing exactly what I described above, it is too early in the QE2 cycle to be certain and pass judgment on these so called premises of mine.
Mr. Murphy is also incorrect to assert that I claimed households aren’t spending enough and that there needs to be more consumer spending. What I said is that creditors, who are sitting on their idle money balances and are the cause of the excess money demand problem, need to spend more. I also argued that the debtors, who should be saving more, should not be spending more. Here is what I said:
[F]or every debtor there must be a creditor. Thus, for every debtor who is cutting back on spending in order to pay off his debts, there is a creditor receiving money payments. In principle, these creditors should be increasing their money spending to offset the decline in money spending by the debtors — but if that were happening, there would have been no decline in overall total current-dollar spending.
This is not a Keynesian directive for more mindless spending. This is a description of the monetary disequilibrium problem where creditors are hoarding money and causing an unnecessary disruption to economic activity. I do argue, though, that the Fed could create the right incentives through QE2 for the creditors to quit hoarding their money and help restore stable nominal spending. Again, this is a case of working with the institutions we have. Along these lines, it is worth remembering that stabilizing nominal spending was seen as a desirable policy objective by Frederick Hayek in a second-best world of central banks and fiat money.
Finally, Mr. Murphy fails to acknowledge that though I support QE2 in principle as a means to address the excess money demand problem, I also acknowledge its imperfections in the article. Here is what I said:
Unfortunately, QE2 may not live up to its potential for several reasons. First, QE2 was not implemented in an optimal fashion. It should have been enacted in a rule-based fashion, with a clear, explicit objective…This would have increased long-term certainty about the price level without requiring an explicit dollar commitment up front from the Federal Reserve. Instead, the Federal Reserve did the opposite. It committed to a large, ad hoc amount of $600 billion that will be used to maintain some vague inflation target. Second, the Federal Reserve continues to pay banks to sit on excess reserves, a policy that keeps money demand elevated at banks… Finally, the Federal Reserve is not immune to political pressures. It is possible the current criticism aimed at the Federal Reserve will make it more timid in implementing QE2.
Ultimately, my case for QE2 boils down to a pragmatic desire to fix the excess money demand problem with existing institutions we have. Thus, contrary to Mr. Murphy’s assertions I understand the problems with fiat money and the limits to intervention. Ultimately, though, I would like to see U.S. monetary policy gravitate toward a more market-friendly approach, such as see Scott Sumner’s proposal to have the Fed target the nominal GDP forecast coming from a (yet-to-be-developed) nominal GDP futures market. This approach would go along ways to restoring long-term macroeconomic stability.