Threadneedle street meets Lombard street: Bagehot and central bankers in the aftermath of the great recession

Abstract

The paper concerns the reception among central bankers of Bagehot’s ideas on the lender of last resort (LLR). For the period 1999–2017 we construct a database of speeches by G7 central bankers that mention Bagehot. On the basis of this database, we show that there was a boom in mentions of Bagehot during the recent financial crisis. Using the Journal of Economic Literature (JEL) codes, we indicate that this increase in the references to Bagehot among central bankers mainly relates to the idea of the LLR, and specifically the so-called Bagehot rules. However, a variety of interpretations of these rules can be observed among central bankers. We argue that these differences in interpretation can be explained by the fact that Bagehot’s rules are used by central bankers either to justify, or, on the contrary, to impose limits on the liquidity injections envisaged under the unconventional monetary policies adopted in reaction to the crisis. We conclude that these diverging preferences on liquidity injections among central bankers allow us to distinguish two main “types” of receptions of Bagehot’s rules: liberal versus conservative.

Keywords:

• Lender of last resort
• central bankers
• financial crisis
• Walter Bagehot

JEL CLASSIFICATION CODES:

• B1
• E58
• G01
• G18
• G28

Acknowledgements

We thank Ghislain Deleplace, Sylvie Diatkine, André Lapidus and Laurent Le Maux for their valuable comments. We also thank the referees for their helpful suggestions. Nonetheless, the usual caveat remains.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 Kuttner (2010) sees the unconventional monetary policies of quantitative easing to be examples of lender-of-last-resort policies.

2 Note that this amounts to a return to the original historical mandate of financial stability of the US Federal Reserve (Reinhart and Rogoff 2013).

3 Prior to the crisis, however, a few economists, notably in the history of economic thought, were questioning the importance of Bagehot. De Boyer Des Roches and Solis Rosales (2003) and Laidler (2003) underline that the phrase “lender of last resort” seems to have been coined by Baring, long before Bagehot. And as for the theory of the LLR itself, Thornton appears to have been the first to develop it, before Bagehot (Diatkine 2003). Finally, Thornton can be considered as having anticipated some of Bagehot’s ideas on the LLR (Goodhart 1999).

4 On the FRASER database we found around twenty speeches from members of the Board of Governors or from Federal Reserve Bank presidents mentioning Bagehot for the period 1913–1998 (i.e., prior to our own sample period). This means that he was largely absent during the 20^th century, so that the surge of Bagehot’s mentions in the aftermath of the 2007 crisis cannot properly be viewed as a “return.” We are grateful to a referee of this journal for pointing out that the idea that there was a historic return to Bagehot requires significant qualification.

5 In this hawks versus doves typology, individual central banker types are based on central bankers’ actions, often with a formal treatment based on individual Taylor rules (Bennani et al. 2018). For example, a “hawk” is often characterized by votes (or dissents) for higher interest rates or interest rate votes for tighter monetary policy (Eijffinger et al. 2018). The literature using this classical typology of hawks versus doves is rather scarce regarding the LLR. Nijskens (2014) is an exception, but he does not link hawks and doves with Bagehot.

6 In addition, JEL coding helps us to manage an already large database of speeches, facilitates its further possible extension, and has a particular relevance in the analysis of the dynamics of evolution of the reasons why Bagehot is mentioned.

7 “Rule R7 constrains rule R1” is not a general statement regarding the LLR function. This statement is valid under the gold standard system, and also according to the Monetarist school, which recommends limiting money injections to avoid inflation and more generally to prevent conflicts between the monetary policy and LLR functions of the central bank (Humphrey 1975).

8 To our knowledge, this sentence is absent from Bagehot (1873).

9 As underlined by a referee, this question of stigma was not present in Bagehot’s original doctrine. Moreover, the importance of this problem of stigma should be tempered by the fact that it mainly concerns discount window operations, and not all liquidity injection procedures, as stressed by Bernanke (2009).

10 For Haltom and Lacker (2013, 12): “If failures threaten to hurt other firms or the economy at large, Bagehot said the central bank should continue to protect the money stock through liberal lending.” They also add, “In fact, much of Lombard Street was about that need, not panics. However, he wrote, if the large gold stock wasn’t enough to allay panic, the Bank of England should follow the ‘brave plan’ and lend liberally.”

11 “Even when it is appropriate for a central bank to function as a lender of last resort, it should follow a rule‐like or systematic approach. This suggests announcing in advance the criteria that will be used to lend and who will be eligible to participate. Economic and financial stability would be best served by establishing such guidelines in advance and committing to following them in a crisis. That commitment is hard to deliver on, but institutional constraints can help tie the hands of policymakers in ways that limit their discretion” (Plosser 2011, 8).

12 Plosser (2008) defines the notion of being rule-like as follows: “This means making policy decisions using available information in a consistent and predictable manner. This does not mean one can know what the future holds or what future policy decisions will be. You often hear Fed officials speak of policy being data dependent, and indeed it is. But the data should feed into a decision-making process in a mostly systematic way. Some people like to think of this systematic part of policy as a “reaction function” or contingency plan with parameters that are largely stable over time.”