11/15/2013

Monetary Policy Implementation vs. Monetary Policy Strategy: Design Features of Monetary Policy, with Application to the Debate Regarding NGDP Targeting

Introduction

I was (re)introduced to monetary policy operations when I came across the work of Scott Fullwiler a couple of years ago. Reading his papers, it soon became apparent to me that there were some serious discrepancies between what I was taught about monetary policy in college and how Fullwiler explains it is actually implemented by real-world central banks. This was odd and distressing to me. Could it really be true that mainstream monetary policy theory was this divorced from reality? At this point in my research, the answer seems to be “yes” and “no.” Ulrich Bindseil’s book, which I discussed in a previous post, supports this notion. Bindseil is the Deputy Director General of Market Operations at the European Central Bank.

I think it is “yes” in the sense that there are rigorous and well-developed arguments coming not just from Fullwiler but from a fairly extensive body of work that supports the notion that certain lineages of academic monetary policy theory have departed from accurate descriptions of what is feasible in reality. This body of work is not just in the domain of historic central banking experts, such as Walter Bagehot, or fringe economic schools, such as Post-Keynesian economics. Over the past two decades, it has become embraced by central bank authorities around the world and by certain strains of mainstream macroeconomic theory, with econ demi-god Michael Woodford notably one of the leaders of the latter front.

However, I also think it is “no” in two senses. The first sense is that, well, as I noted above, some of the most recent mainstream theory is now relatively on board – i.e., Woodford and company. The problem seems to be that many other scholars have not caught up, and most textbooks have not been updated. The second sense is that a critical distinction needs to be made between monetary policy implementation and monetary policy strategy. While many academics have for a long time conflated the two to the detriment of the coherency of their ideas, they have arguably proposed legitimate theories of the latter (which I am not endorsing or critiquing here) that indeed inform central banking practice to this day. 

In the era of academic, blogospheric, and media debate regarding novel and complex approaches to central banking, such as “NGDP targeting,” these two concepts of strategy vs. implementation continue to be conflated, to the detriment of the dialogue that is taking place. People focused on 'strategy' often talk past people focused on 'implementation,' and the discussion sort of goes nowhere  As such, I aim to define and outline these two concepts in this post, with the hope that it can provide a useful template for thinking about monetary policy.

I borrow the specific framing and terminology below from Bindseil. I presume, given that he is an operations authority within the ECB, that these are common ways central bankers view this topic. However, alternate terminology likely exists, as well as slight modifications to the framework used. Scott Fullwiler has noted that these general concepts are common in the policy sciences, as well as in the business and management literature. Depending on the context, I am sure different terminology is used. For example, I’ve seen Fullwiler specifically refer to a three-pronged framework, which consists of 'tactics,' 'strategy,' and 'policy.' 'Tactics' corresponds to what is described as 'implementation' below, whereas 'strategy' and 'policy' are subdivisions of what is described as 'strategy' below.

Defining Monetary Policy Implementation and Strategy

At a high level, I would describe the relationship between monetary policy implementation and strategy as the following: monetary policy implementation refers to the means by which central banks intervene in markets when attempting to achieve the overarching goals set out by monetary policy strategy. The following bullets define and describe these concepts in more detail.

Monetary policy strategy consists of two main elements:
  1. A macroeconomic model of the transmission mechanisms of monetary policy
    • One element of monetary policy strategy is the central bank’s macroeconomic model of the transmission mechanisms through which it believes monetary policy is supposed to impact the economy. In other words, this model captures how the central bank’s final target is linked to its operational targets, indicator variables, and intermediate targets.
      • The final target of monetary policy is the economic variable(s) that the central bank ultimately aims to impact. Transmission of the central bank’s strategy starts with its operational target, which may then impact its intermediate targets, which then impacts the final target. The final target is often referred to in a central bank’s statute, such as pursuing high employment and stable prices, however defined.
      • Operational targets will be defined below in the context of monetary policy implementation.
      • Intermediate targets are variables a CB believes they can control with a reasonable time lag and with a reasonable degree of precision and which is in a stable or predictable relationship with the final target of monetary policy. Examples may include medium term interest rates, long term interest rates, the exchange rate, and monetary aggregates. Bindseil reports that the concept of ‘intermediate targets’ have fallen out of favor and are often referred to as indicator variables.
      • Indicator variables provide data that inform how the central bank may need to adjust its operational target variable to ultimately impact the final target. Bindseil offers examples of ‘exogenous’ variables such supply side shocks and technological innovations. As noted above, intermediate targets could be included in this category as well.
  2. Rules for operational adjustment
    • The second element of monetary policy strategy consists of the manner in which the central bank adjusts its operational targets in response to new information and communicates this to the public in order to achieve its final target.
Monetary policy implementation consists of three main elements:
  1. The operational target of monetary policy
    • The operational target is something that the central bank wants to and can control on a daily basis through use of its monetary policy instruments as a means of achieving the goals laid out by the chosen monetary policy strategy.
      • Today, there is broad consensus among central banks that the short-term inter-bank interest rate is the appropriate operational target. However, in the past and still today, many academics argue that quantitative variables, such as the monetary base, are preferable. That said, arguably the most popular mainstream macroeconomic theory, New Keynesianism, takes the interest rate view.
  2. The framework for controlling the operational target
    • The CB controls the operational target using monetary policy instruments
      • The main monetary policy instruments used today are standing facilities (i.e., interest on reserves and lending facilities), open market operations (OMOs), and reserve requirements
  3. Daily use of instruments to achieve the operational target
    • This may include conducting OMOs, changing standing facility rate, and changing reserve requirements
One thing you may be wondering is where the roles of communication policy and forward guidance fit into this. I’m only half way through Bindseil’s book, but it seems these issues aren’t addressed very much (as I noted in my previous post, an updated version that addresses the post-financial crisis world of central banking would be nice). This may be a deficiency, but I think there are a couple reasons for this. First, things like communication and forward guidance may fit more within the realm of monetary policy strategy, which is suggested in the summary above, and Bindseil's book focuses on implementation. Secondly, as a central banker and someone intimate with the modern mathematical models describing how monetary policy implementation works, Bindseil places a great deal of importance on taking action to not just back up the credibility of central banks but more so to actually achieve their operational goals. In other words, at least with respect to targeting interest rates, Bindseil shows how the best models out there demonstrate that simply announcing your intentions is not sufficient without use or adjustment of one of the policy instruments (including the standing facilities). Fear not heterodox readers, Bindseil still appears consistent with the Post-Keynesian descriptions of interest rate targeting that I've read, as I’ll describe in a future post. And fear not NGDP targeters (at least in some respects), as we’re talking about implementation and interest rate policy here, not the broader strategy.

The Nexus Between Monetary Policy Strategy and Operations

Bindseil conceives of there being a hand-off from monetary policy strategy to monetary policy implementation. In other words, using the central bank's model of the economy, the strategists determine what the operational target should be at any given moment, and then communicate this to the implementation experts, who then work to hit the operational target in the real world. Indeed, from lecture notes online, Bindseil observes there is often a division of labor and expertise between these two areas within central banks:

“Monetary policy can be said to have a Sollbruchstelle (a predetermined breaking point), which is where  the monetary macroeconomists hand over the short term interest level, identified as optimal to achieve the ultimate target, to the implementation experts. While the first function belongs to the area of “white-collar” central banking and is usually performed by the economics department (with involvement  of the research department), the second function belongs to the “blue collar” areas, for instance, in the  European Central Bank it is carried out by a department called “Market Operations”. In principle,  monetary macro-economists in central banks do not need to understand monetary policy  implementation and, symmetrically, implementation experts do not need to understand much about  monetary policy strategy and the transmission mechanism. Actually, the resulting efficiency gains that  specialization allows are normally realized in central banks, i.e. there are no attempts to make  implementation experts at the same time macro-monetary economics specialists, and vice-versa. This  segregation between the two functions may also be called a “dichotomy” between monetary  macroeconomics and monetary policy implementation. Others have referred to it as the “separation  principle” of monetary policy implementation. One may also speak about a “hierarchy” organizing  monetary policy, with the choice of the ultimate target at the top, followed by the analysis of the  transmission mechanism and the associated identification of the optimal level of the operational target of monetary policy, and with monetary policy implementation finally achieving this target through  monetary policy operations.”

Interestingly, judging by what Bindseil writes in his book, he would probably go further to suggest that prior attempts of “white collar” economists to delve into the realm of “blue collar” central banking is what led to 20th century confusion regarding how monetary policy actually works. As such, I'm a bit surprised that Bindseil would use such seemingly disparaging terms, although perhaps he is reporting the reality.* EDIT: In my opinion, this is not to say the two groups shouldn't understand what each other is doing. To the contrary, it seems the lesson from the '20th century confusion' is not that the two groups should stay separate, but that they should be intimately familiar with each other's domain.

Without getting too much into the history here (for another post), Bindseil notes that the term “instrument” was understood in an ambiguous way in the influential article of Poole (1970), namely to designate both operational and intermediate targets (specifically short-term interest rates and monetary aggregates). This, he believes, has contributed to a persistent misunderstanding among many academics that the central bank can reliably use the monetary base as an operational target. Instead, Bindseil strongly believes it is fundamentally interest rates that CBs can control as operational targets, whereas targets such as the monetary base are at best intermediate or final targets. This is, by the way, consistent with what Post-Keynesians argue.

Finally, one argument Bindseil sets out to defend in the book is the following:

“Monetary policy strategy is not of key relevance to choosing an appropriate approach to monetary policy implementation: monetary policy is necessarily implemented through control of short-term interest rates, irrespective of strategy.”

I won't attempt to explain why Bindseil believes this yet. Again, this will take at least another post, and I need to first broach the topic of interest rates vs. quantities. I'll leave this as food for thought for now, without insinuating any degree to which this may be wrong or right.

Application to NGDP Targeting

Recently, there was a strong and partially successful push for the Fed to alter its prior strategy of inflation targeting (guided by concepts such as the Taylor rule) to an approach called “NGDP targeting.”

This advocacy campaign has its roots both in the blogosphere and more formal domains of academia (on the relevancy and benefits of the two, see, for instance, Krugman here). Bentley professor Scott Sumner and his fellow “market monetarists” are largely responsible for popularizing the idea on the blogosphere and in mainstream economics news sources. This publicity has certainly reached a wide audience, including economists. However, NGDP targeting and related variants have also been discussed within academia for at least a decade. The crescendo of consensus on NGDP targeting reached a considerable volume when FOMC voting member Charles Evans and former CEA chairwomen Christina Romer both endorsed the idea in 2011. The push for NGDP targeting reached a high point when in September 2012, Michael Woodford presented a paper advocating NGDP targeting at the annual Jackson Hole conference hosted by the Federal Reserve Bank of Kansas City. Finally, in December 2012, the Fed made a historic shift in its approach to monetary policy by adopting the ‘Evans Rule,’ which moved Fed policy much closer to the concept of NGDP targeting.

NGDP targeting is definitely still a somewhat controversial idea, even though it has a broad base of support. It's controversial among mainstream as well as heterodox economists. Probably the most common critique is: "Where is the transmission mechanism?"

Again, I reserve my right not to get into this debate here. However, what I would like to do here is implore supporters of NGDP targeting to explain how this monetary policy approach maps to Bindseil's policy framework above. After all, if this idea is to be used by central banks, then it will probably need to be mapped in such a manner one way or another. So, within the context of policy implementation, what are the operational targets? What instruments will be used to achieve these operational targets, and how will this be done? Within the context of strategy, what are the indicator and final variables? What rules for operational adjustment and communication policies will be used to achieve the goal NGDP targeting?

Many of the answers to the strategy questions above have been repeatedly provided by NGDP supporters, albeit not within the context of the framework laid out above. However, I find the answers on the transmission mechanism and implementation sides lacking. Much is said about forward guidance, but there needs to be more, as the economy may not automatically do what the central bank claims it wants to happen (said in an admittedly tongue in cheek fashion).

----------------------
*As a side note, I find it interesting that the connotations of these terms parallel a common grievance of heterodox economists and bloggers, which is that mainstream economists often express (an unjustified) scorn to an accounting and flow of funds oriented approach to understanding the economy, or what Perry Mehrling might call “the money view” approach to economics.

6 comments:

  1. That's a very useful set of vocabulary. Not sure if I should buy the book or just follow along as you write posts on the subject.

    "...not that the two groups should stay separate, but that they should be intimately familiar with each other's domain"

    Couldn't agree more.

    ReplyDelete
    Replies
    1. Thanks! You can try following along at first. I do intend to summarize the book to the best of my ability. An alternative is that you can try a local library. Also, I haven't dug deeply through the link to the online lectures notes, but it does seem to summarize much of the book.

      Delete
  2. The blue collar-white collar distinction is interesting. The first concerns itself with how to achieve something, the second with what to achieve. We can add a third group to the mix that tries to determine if the whole effort is legal and/or politically feasible. Even if the white collar strategy can be implemented by blue collar techniques, does it lie within the set of rules that governs the central bank? NGDP targeting needs not only a legitimate blue-collar transmission mechanism but also a legally viable one.

    ReplyDelete
    Replies
    1. Agreed. JKH and Fullwiler have also thrown another distinction of 'policy' into the mix. They may be using it differently, but I think I could see use for an additional dimension like this. Will continue to think this over.

      Delete
    2. Here's an example. Miles Kimball already has the white and blue collar issues behind his negative rate scheme worked out, now he's going through the legalities:

      http://blog.supplysideliberal.com/post/63059715906/legal-issues-relevant-for-the-transition-to-electronic

      Delete
  3. "Interestingly, judging by what Bindseil writes in his book, he would probably go further to suggest that prior attempts of “white collar” economists to delve into the realm of “blue collar” central banking is what led to 20th century confusion regarding how monetary policy actually works."

    I think that the current monetary policy system is unnecessarily complex needing a division between white collar and blue collar central banking. But if the this counterproductive complexity was done away with strategy and operations should be efficiently managed by the same people.

    ReplyDelete