Wednesday, October 8, 2014

Debt Management Discussion at Brookings with a Combative Larry Summers and Others

Many of the events I have attended at Brookings are sober, serious-minded affairs, and it helps to be really interested in the topic being discussed. On September 30, a discussion about debt management started off that way, but then Larry Summers got a chance to defend a policy proposal in the paper he wrote with others, “Government Debt Management at the Zero Lower Bound,” and the event became rather lively, interesting, thought-provoking, and entertaining. That is not to say that I agreed with everything that Summers said, and, combative as he was, he did admit that the discussants had given him and his coauthors issues to think about.
Despite the paper’s title, it really is a compendium of observations and analyses about debt management, some of which have no apparent link to the zero lower bound. For example, of particular interest to me because of work I did at Treasury, there is a discussion of the liquidity premium on Treasury Inflation-Protected Securities, but why this discussion is included in this paper is unclear.
The controversial issue that the paper addresses is the coordination of Treasury debt management policy with Federal Reserve open market operations. This, of course, is not just an issue when the Federal Reserve’s policies are constrained by the zero lower bound. For example, the paper makes reference to the 1961 Operation Twist policy of the U.S. government. This was an attempt to lower long-term rates and to increase short-term rates. The rationale was that short-term rates needed to be higher to protect the value of the dollar under the Bretton Woods system and long-term rates needed to be lower to encourage investment and stimulate the economy. The current paper states:
“Operation Twist is perhaps the best example of the potential for Fed and Treasury cooperation, because the circumstance was, much like the zero lower bound today, that the Fed was constrained in its use of the short rate as a policy instrument. However, unlike in the more recent period, during Operation Twist the Fed was able to complement its own actions with the secured cooperation of the Treasury to alter the maturity structure of new debt issuance.”
In fact, though the Fed and the Treasury were not cooperating, as the paper notes in footnote to its discussion of Operation Twist:

“Long-term interest rates fell on most dates in early 1962 when the initial information about Treasury and Fed policies was released. The only exception was when the Treasury surprised both the White House and the Fed by issuing longer-term bonds on March 15, 1961. This made James Tobin (then a member of Kennedy’s CEA) ‘furious.’ Treasury continued to extend its maturity thereafter and within a year the average maturity had increased by 3.5 months. Thus, Treasury began working at cross-purposes with the Fed, just in as [sic] the current episode.”  
In other words, Operation Twist is not informative about the supply effects of Treasury securities on the yield curve, nor is it an example of cooperation between the Treasury and the Fed. The authors need to revisit this issue if they revise the current paper.

The main contention of the paper is that the Treasury and the Federal Reserve are currently working at cross purposes. The Treasury is extending the maturity of the public debt by selling more long-term securities, while the Fed is taking long-term securities off the market through its open market operations. The authors believe that this selling and buying should stop. Moreover, the authors write that, even in more normal interest rate environments, “because of the importance of debt management for the functioning of financial markets and because of its relation to financial stability, the Federal Reserve should have a more significant advisory role than it does currently.”
Interestingly, the discussants argued that this recommendation could result in an erosion of Federal Reserve independence, though on its face, it would seem to give the Fed greater ability to influence the Treasury. In this connection, one should note that the authors believe that Treasury’s debt management policy is currently in error, not the Fed’s quantitative easing policies.

Nevertheless, I think the discussants are right to have this concern, though I would add that Treasury’s independence from the Fed is also a concern. The authors suggest that the Fed and the Treasury “annually release a joint statement for managing the U.S. government’s consolidated debt,” by which they mean debt held by the public not including the Federal Reserve Banks. One can only imagine the lengthy and likely unpleasant and stressful meetings at various levels as the Treasury and the Fed negotiate this statement.
Whose voice would be controlling would depend on circumstances and personalities. The Fed often has a strong hand in discussions about debt management, partly because they have more staff and other resources to draw upon than Treasury. But a strong-minded and strong-willed Secretary, think John Connally or William Simon, could conceivably bring strong pressure to bear on the Fed if there were a strong disagreement between the two institutions.

This brings me to a final comment about the paper. It reads as if the Treasury and the Fed make policy as independent actors without referencing who might be in charge of those institutions at any particular time. Political appointees to the Treasury and Federal Reserve governors and bank presidents make decisions. Sure, they receive input from staff and other sources, but ultimately they make decisions as individuals.
As way of example, with regard to debt management, the decision in 2001 to stop selling 30-year bonds was made by then Under Secretary Peter Fisher. The argument that various Treasury officials made at the time in support of this decision is that, since the yield curve usually has a positive slope, a shorter maturity structure would “over time” lead to lower cost financing. Of course, the time frame was left ambiguous, and the argument only worked, if it worked at all, if Treasury could maintain this policy over time. As it turns out, it could not. Thirty-year bond issuances resumed in the same administration as Peter Fisher served (George W. Bush) after he had left, though not immediately. The Obama Administration subsequently reversed course from the shortening strategies of both the Clinton and Bush Administrations and decided to lengthen the average maturity of the public debt. In other words, particular individuals do matter, independent of what one might model as in a particular institution’s interest.  

1 comment:

  1. Very nice post and thanks for bringing to our attention this conference. I also found it both lively and interesting.

    As I read your post, I thought back to Graham Allison's book on bureaucratic politics, and the idea that policy problems sometimes originate in the different interests and strategies of different parts of the state. Your story about Tobin is a good example.

    I was also reminded of a 1979 essay by Charles Baker Jr on debt management in which he laments the lack of theoretical work in this area. The topics discussed during the conference were so wide-ranging, it made me think this may be one reason there has been so little analysis about these issues (though I must admit I do not specialize in this area so I may be wrong).

    One last thought: I was quite surprised by Mary Miller's reaction and comment at 1:25 when she remarked it was useful to think of the Fed and Treasury balance sheets as 'consolidated'. Her reaction made it appear she hadn't thought about it this way before. Is that possible?