The Outlook for 2010: A Critique of Modern Monetary Policy

Despite the sideshow over financial regulation, monetary policy is the most important element of economic growth within the purview of any regulatory entity. Hence, the Federal Reserve still has a preeminent role to play in the recovery (assuming one has started).

One has to view the Federal Reserve’s monetary policy actions not as following absolute rules of knowledge, but in the context of the history of how economists’ understanding of monetary policy has evolved over time. Recall, contemporary monetary policy targets, using Fed funds rate targets, are only about twenty years old. (A reasonably well-developed Fed funds market, itself, only dates back to the 1960s or so.) Fed funds targeting was developed precisely because of advances in non-deposit funding, like the market for certificates of deposit, that skewed prior historical relationships between monetary aggregates like M1 and economic growth through the 1970s and 1980s. Prior to the focus on M1, the Federal Reserve targeted banks’ free reserves and credit conditions. Only in the 1950s, did open-market operations as we know them, today, become developed into a policy instrument.

The ongoing history of US monetary policy therefore has to be one of a continuing search for monetary targets in an ever-changing institutional context for at least temporarily stable relationships between monetary policy and economic growth. Hence, it makes sense to ask “how is it coming along?” Unfortunately, I think the answer is “not well.” While I present here only one dimension of monetary policy, Fed funds rates, the lesson is straightforward: the recent era of managing the Fed funds rate to reduce the depth and duration of business cycles is failing.

The graph below illustrates both the reason for embarking on the experiment with Fed funds policy, as well as the failure of the policy to date. The graph shows the Fed funds rate with NBER recession dates (monthly) illustrated in gray. Visually, two things are apparent. First, recessions prior to 1990 are pre-dated by rate increases, which can be thought to have precipitated or at least worsened the ensuing recessions. Second, many recessions are followed by further Fed funds rate declines, which can – intentionally or unintentionally – address the lagging effects of recessions (like unemployment), but which risk adding stimulus after the “horse is out of the barn,” so to speak.

mason_2010_outlook_graph_gif_version_512.gif  

-Click image to enlarge

As a result of these and many other observations, monetary policy since 1990 has sought to act differently. Consider the first line of the table lying below the graph, which reports Fed funds movements as the percent change in the rate over the six months preceding each recession date. Prior to the Greenspan era, rates typically rose in those six months by an average of nearly fifteen percent. Moreover, while rates increased just over five percent pre-dating the 1957 recession and just over two percent pre-dating the 1960 and 1969 recessions, they rose almost forty-one percent pre-dating the 1973 recession and almost thirty-four percent pre-dating the 1980 recession.

Given the disruptiveness of those latter two recessions, it is not surprising that policymakers perceived policy mistakes, having effectively tightened into both recessions. (Note, however, that the Fed funds rate was not a target in those periods, so the tightening is not a conscious policy decision.) Contrast that tightening with conscious policy movements prior to post-1990 recessions, where the Federal Reserve, unlike previous periods explicitly targeting the Fed funds rate, began reducing rates in the six months prior to recession. In the 1990 recession – the first application of the principle – rates were reduced by almost two percent. But in later applications, the 2001 and 2007 recessions, rates were reduced about fifteen percent before the recession began. Of course, in neither of those cases did the Federal Reserve avert recession, but they avoided tightening into the recession as in previous episodes. While causality is difficult to verify, the 1990 and 2001 recessions both lasted eight months, shorter than the historical average of ten months for recent recessions.

Historically, the Federal Reserve provided additional relief once a recession had started. (See the third line of the table.) On average, Fed funds rates fell another forty-four percent during recessions. Some notable outliers include 1957 and 2001, where the Fed funds rates fell by over sixty-one percent, and 1981, where it fell almost fifty-two percent. In 1990, the Fed funds rates fell by only about twenty-five percent; in 1980, thirty-four percent.

The fourth and fifth lines of the table demonstrate the historical difficulties the Federal Reserve has always had with exit policy. It appears that the most dramatic after-the-fact loosening took place after the 1957 recession, where the Federal Reserve continued expanding until Fed funds rates had declined another forty-six percent on top of the sixty-one percent decline during the recession. In 1969, policy resulted in a follow-on decline of twenty-four percent, on top of the almost thirty-eight percent during the recession. In 2001, explicit policy reduced rates almost another seventeen percent, on top of the roughly sixty-one percent during the recession.

In the past, as today, overshooting was thought to be a problem. Hence, in the six months following the 1957 recession, rates went up almost one hundred eighty six percent from their recession lows. In 1960 and 1969, rates were fairly stable in the six month window, declining slightly beyond their recessionary lows. Following 1973, Fed funds rates six months after the end of the recession were up roughly thirteen percent over their recessionary lows. Following 1980, they rose almost one hundred percent from their recessionary stimulus levels by six months after the recession, only to be slashed a year later by fifty percent and another eight percent in the 1981 recession, in quick step.

Looking holistically at recent policy, in 1990, the Federal Reserve decreased rates roughly two percent prior to the recession, then another twenty-five percent during the recession, and another almost eight percent after the recession. It wasn’t for another 108 months, until May 2000, that rates again rose above their stimulus lows in the 1990 recession.

Of course, by March 2001, the economy was in recession, again. The Federal Reserve cut rates by over fifteen percent before the 2001 recession, another sixty percent during the recession, and then – like in 1990 – roughly another seventeen percent after the recession. But, like 1990, the Federal Reserve did not raise rates above the stimulative 2001 recessionary lows until December 2004, approximately a year before the real estate bubble burst in California.

By 2007, the economy is in recession yet again. This time, the Federal Reserve cut rates by almost fifteen percent before the 2007 recession, and some ninety-seven percent during the crisis. The problem is that there is no room to reduce rates following the recession, as in all but 1973 and 1980. If you recall, those were not times of widespread economic prosperity.

To me, at least, there is a clear lesson in the Fed funds rate experience and application: while you probably don’t want to raise rates into a recession – 1973 saw a six-month pre-recession increase of over forty percent and 1980 almost thirty-four percent – you don’t want to exclusively wait until during and after a recession to reduce rates. Nonetheless, reducing rates before recession may both add to the speculative fury that will (eventually) necessitate the (potentially larger) downturn and leaves no room to lower rates in order to address lagging effects like unemployment. The lesson, therefore, may be that while we have sought front-end loosening to mitigate recent recessions, we have sacrificed back-end loosening that could be an important component of monetary policy.

In the context of the classic quote from Arthur Burns, if you don’t take away the punchbowl until you are out of punch, your party guests get arrested for drunk driving. But if you do take away the punchbowl, you can have another party next weekend. Unfortunately, our guests got arrested and our punchbowl is dry.