Central bank independence, a not so new idea in the history of economic thought: a doctrine in the 1920s

Abstract

Central bank independence (CBI) has been a prominent topic for decades, but remains relatively unexplored in the history of economic thought. Relevant literature is scarce and focuses on the post-war period. To extend this literature, I argue that there was a doctrine of CBI in the 1920s. I examine its development from an international recommendation and a principle of the central banking doctrine designed by the British governor Norman to the elaborate doctrine of Kisch & Elkin. The paper’s main contribution is a detailed analysis of the doctrine of CBI provided in their reference book on central banking.

Keywords:

• Central banking
• central bank independence
• history of economic thought
• interwar period
• monetary arrangements

JEL CODES:

• B00
• E42
• E58
• E61
• F33

Acknowledgements

I would like to particularly express gratitude to my doctoral supervisor Mehrdad Vahabi for his support and guidance throughout the research. A special thanks to Jaime Reis, Muriel Dal Pont Legrand, Pierre-Cyrille Hautcœur and Stefano Ugolini for valuable comments and remarks throughout the writing process. I am also very grateful to the three anonymous referees for their contribution. Obviously, the responsibility for the content of this paper is only mine.

Disclosure statement

No potential conflict of interest is reported by the author.

Notes

1 In quotations, unless otherwise mentioned, clarifications in brackets, ellipses, and emphasis are mine.

2 In a nutshell, “central bank independence means that monetary policy is delegated to unelected officials and that the government’s influence on monetary policy is restricted” (de Haan and Eijffinger 2016, 2).

3 Beyond numerous individual central banks, there was also mass adoption of CBI by the twelve central banks that founded the euro zone and by the European Central Bank (James 2012) as well as by the majority of the former socialist and communist countries in their transition to a market economy (Johnson 2016).

4 It was Simons (1936) who framed the debate in terms of rules vs. authorities (discretion). Friedman (1962) developed his mentor’s discussion of the monetary order from a liberal perspective and argued for legislative monetary rules as opposed to discretion (whether by politicians or by independent technocrats).

5 Along with Rogoff’s case, a more “political” one is made by the political business cycle literature, which suggests CBI insulate monetary policy from the political cycles generated by re-election or ideologically-oriented policymakers. In the early 1990s, some authors found a negative correlation between CBI and inflation for industrialised economies and suggested causality running from CBI to lower inflation. Alesina and Stella (2010), Cukierman (2008) and de Haan and Eijffinger (2016) surveyed the vast literature on CBI.

6 Combined with the adoption of inflation targeting, CBI is one of the policy recommendations of the New Consensus Macroeconomics (Goodfriend 2007). It came to be seen as the culmination of an “intellectual revolution” set off by the inflationary bias story (Cukierman 2008, 726–727). When Kydland & Prescott were awarded the 2004 Nobel Prize in Economics, the press release stated they “largely influenced the reforms of central banks (…) in many countries over the last decade”.

7 This is a standard narrative since it is frequently repeated in the economic literature and the views it contains are widely held (beyond already cited works see Bénassy-Quéré et al. 2010, 239; Mishkin 2013, 322; Walsh 2008).

8 In his reviews of the literature, Bibow (2004, 2010) deals with Keynes’ thoughts on CBI and compares them with the modern literature. According to Forder (2005), the arguments for CBI go back as far as Grove (1952).

9 The citation refers to two French idioms “sortir tout armé de la tête de quelqu’un” and “se croire sorti de la cuisse de Jupiter”. The first one, inspired by Athena’s birth, means that an idea is already fully developed and nothing can be added to it. The second one, inspired by Bacchus’ birth, refers to the exceptional nature of something or somebody and particularly to the fact that somebody considers himself to be extraordinary.

10 Bibow (2009) and Dehay (1995) tackled the question of whether or not ordoliberals favoured CBI and whether ordoliberalism explains the German Bundesbank’s exceptional independence in the post-WWII world.

11 Keynes’ major works concentrated on matters pertaining to the conduct of monetary policy and did not provide substantial discussion on the structure of monetary policy and CBI. Keynes mainly discussed these matters as a witness before commissions on Indian currency, as an adviser to the Liberal Party, and as a critic of the Labour Party’s 1932 policy pamphlet on currency, banking and finance (Bibow 2002).

12 Mints (1945, 280) and Chandavarkar (1996, 262) also alluded to Kisch & Elkin as CBI advocates.

13 For instance, in a reference three-volume anthology on central banking, Collins (1993) does not include an excerpt from Kisch and Elkin (1928, 1932) in the volume dedicated to the question of discretion and autonomy. However, in a major paper, Capie, Goodhart, and Schnadt (1994, 63) still referred to Kisch & Elkin’s book as “an authoritative source from the 1920s” on central banking.

14 Doctrine is defined here as a set of held principles/beliefs meant to be followed in a particular domain.

15 Among public finance methods, taxation contributed the least to the war effort. On average, less than one-third of the expenditures of belligerent countries from 1914 to 1918 were financed by taxation (Eichengreen 1995, 75). Given the difficulties involved in collecting taxes, including redistributive conflicts and/or limited State ability to tax, governments turned to borrowing and money creation. Given the limits of bond financing and to meet the most pressing needs, they issued treasury bills and had them discounted by central banks from which advances were also obtained (Ibid., 78–81; Kindleberger 1984, 292–296).

16 See Flandreau, Cacheux, and Zumer (1998) and Flandreau (2003) for insights into CBI before WWI.

17 Inconvertibility was seen as temporary and exceptional. The French term cours forcé (forced circulation) may be a better way to describe this point of view as it underlines the role of the State in imposing banknotes as a means of payment at its face value without possible convertibility. As today this is the norm, one refers to fiat money, which is a more positive term.

18 Keynes roughly identifies managed currency with discretion. He refers to the belief that gold was not managed since its value depended on independent factors but not “on the policy or the decisions of a single body of men”. Discussing the desirability of restoring the gold standard, Keynes rejects the argument that gold provides a stable standard of value. He asserts that even gold became “managed”, because its value is “artificial”, depending on the policies of central banks. In his opinion, the stability of exchange via a gold standard does not guarantee price stability. A “barbarous relic” and an old despot whose “natural” value would prevail, the gold standard could be restored only as a “constitutional monarch” (2013, 133–138).

19 While central banking is usually characterised by the Lender of Last Resort (LOLR) function generally assumed by banks of issue in the late 19th century (Capie, Goodhart, and Schnadt 1994, 5–6; Goodhart 1988), modern central banking is also associated with conducting monetary policy as part of macroeconomic management. Thus, modern central banking can be said to combine two main tasks with corresponding functions, financial stability (management of the payment system, LOLR and banking supervision) and monetary stability (issuance of money and the conduct of monetary policy) (Ugolini 2017, 13).

20 By 1925, a new international gold standard was in place, the gold exchange standard, whose principles were established at the 1922 Genoa Conference (see section 3). Another gold exchange standard was set up after WWII under the Bretton Woods agreements and lasted until 1971. Since then, a pure fiat money regime has prevailed. In reality, these gold standards were increasingly “fiat”.

21 This resolution established the principle of central banking universality (Feiertag 2016, 18). After intense debates in the 19th century, free bankers “had been relegated to fringe status” and most economists and experts were thinking about stable money within central banking, considering money/banking as an exception to free trade principles (see footnote 41 and 46 for the Bagehot position on the issue). In the 1920s, Mises played a leading role among the few authors who argued for free banking (White 1993, vii–viii). Needless to stress that CBI only makes sense in central banking. It should however be noted that the rules vs. discretion debate does not exhaust the question of the neutralisation/depoliticisation of money, which can also be tackled via the opposition between central and free banking revived by Hayek in the 70s (Goodhart 1988).

22 The insistence on private ownership is also characteristic of Kisch & Elkin’s doctrine of CBI (see section 4).

23 They were Pigou (English), Cassel (Swedish), Gide (French), Pantaleoni (Italian) and Bruins (Dutch). Together these economists wrote a joint statement and each of them also wrote a memorandum.

24 Quantity-theory of money was by then dominant and reflected the classical dichotomy between real and monetary analysis (Schumpeter 1986 [1954], 264–265, 1054–1055).

25 On similar lines, chapter 2 of Keynes’ Tract analysed inflation as a method of taxation.

26 As in the period that followed the break-up of the Bretton Woods’ gold exchange standard, in a world of floating exchange rates, with no nominal anchor and experiencing high inflation, CBI came to be considered as an option. It should be noted that CBI does not provide a nominal anchor. It is an institutional arrangement through which the government delegates monetary policy. Despite the fact that an independent central banker is not necessarily a conservative one, it is usually seen as an indirect way of promoting price stability and signalling a government’s intention not to inflate.

27 Palgrave refers to specific central banks such as the BoE in a general entry for “Banking”. He comes close to the concept of a central bank when he describes the functions of the BoE and defines it as the centre of the banking system as he does in Bank Rate and the Money Market (1903, 8). Covering European banking in the Encyclopaedia Britannica (10th edition in 1902-3, entry “Banking”; 11th edition in 1910-1, entry “Banks and Banking”), Palgrave retained the term “banks of issue”, the usual denomination for those institutions before “central bank” became more common. As de Kock further noted, the concept was used for the first time in a major scientific work when in 1917 Sprague authored a chapter entitled “Central Banks” in the 3rd edition of Dunbar and Sprague (1917) The Theory and History of Banking, a reference book on banking whose 1st edition was published in 1891. Dunbar’s former student and author of a major work on financial crises for the National Monetary Commission, Sprague enlarged the book with three additional chapters, the other two on the Fed and on Foreign Exchange. In its 4th edition, in 1922, the chapter on the Fed was penned by Willis, who was very influential in the drafting of the Federal Reserve Act. That same year, in the 12th edition of the Encyclopaedia Britannica, Willis also covered the US in the entry “Banking” and authored an entry on the Fed. So, he succeeded Conant who had written on the American banking system in the 11th edition, just before the creation of the Fed. In fact, Sprague was a pioneer in the use of the concept “central bank” and in its definition. Reference is made in section 4 to some major previous works on the history of banks of issue also mentioned by de Kock.

28 de Kock lists the growing literature after 1928 in which we find such prominent works as Hawtrey’s (1932), Smith’s (1936) and Willis (1936). de Kock’s book became the next reference on central banking after Kisch & Elkin’s (Plumptre 1940b, 631).

29 It can indeed be said that the work of Kisch & Elkin is still more about central banks than central banking and not yet sufficiently theoretical. It may also be that Norman wanted more significant scientific backing than theirs. See section 4 and the conclusion for an assessment of their doctrine.

30 In June 1921, the South African Reserve Bank (SARB) was the first central bank to be created. Via the British monetary expert Strakosch, the architect of the SARB, the BoE had a significant influence on the founding and on the appointment of Clegg as its first governor (1920–1932). Strakosch later played a significant international role as a leading member of the Financial Committee of the LoN (see end of section 4). India lacked a central bank and the BoE wanted the Imperial Bank of India, a commercial bank created in January 1921, to adopt as far as possible a central banking character (Cottrell 2012, 42; Sayers, Ibid., 203–205).

31 The doctrinal register is obvious in the use of the verb “should” in almost every sentence of the manifesto. Doctrine is meant to be imperative and is meant to guide the action of practitioners.

32 Norman encouraged its publication for wider circulation (Cottrell 1997, 34). Strakosch (1921) eventually published an article in The Economic Journal. It rather presented the newborn SARB, but some references were also made to the general principles of central banking applied in that case.

33 In the spring of 1922, the Genoa Conference laid down the principles for the gold exchange standard, which was in place by 1925 with the return of the UK and other countries to convertibility. This was an exchange standard since central banks held a significant part of their reserves in foreign exchange of centre countries (US, UK, and later France). Internally, it tended to be a bullion standard, as opposed to a specie standard, with no gold circulation and convertibility only possible beyond a certain amount. The new standard was short lived and collapsed in 1931 (Eichengreen 1995).

34 Genoa literally renewed with Brussels recommendations on central banking and CBI (Resolution 2 of the Financial Commission). Nevertheless, Genoa emphasised central bank cooperation. Resolution 3 of that commission advocated a more continuous and formal cooperation and suggested that a preliminary meeting of central banks should be held. Resolution 12 assigned the BoE the mission of calling and organising that meeting (Genoa International Economic Conference 1922). After the conference, Norman went to the US and began working with Strong on the agenda, which, along with the Genoa Report, was soon sent to about a dozen central banks but the meeting was never held (Clay 1957, 138; Sayers 1986, 160–161).

35 All in all, these principles and doctrinal efforts to consolidate a model of central banking to "export" are much more meaningful seen from London’s perspective on the eve of a conference that acknowledged a sterling-based gold standard (Capie, Goodhart, and Schnadt 1994, 21; Ugolini 2017, 15–16).

36 According to Kisch and Elkin (1928, 18), the BoE was very legally independent with private ownership and governing bodies free of political appointees. The 1844 Act was still the core legal framework of the time and it remained so until nationalisation in 1946. Indeed, de jure independence may be misleading, especially when there is great stability of central bank statutes over time, since it does not account for changes in the underlying conditions required for its enforcement.

37 During the war and until April 1921, the reference rate was the Treasury bills tap rate. The British Treasury gradually aligned itself with the BoE thereby restoring room for manoeuvre in fixing rates as budgets balanced. A restrictive policy was gradually implemented, notably with rate hikes in the autumn of 1919 and in the spring of 1920. By the end of the summer of 1920, the post-war boom was over, and deflation was on the march. The BoE can be considered to have recovered de facto independence when on-tap issues ceased (Clay 1957, 115–133; Eichengreen 1995, 112–113; Howson 1974; Sayers 1986, 110–126).

38 This “international professional caste” (Cottrell 1997, 35) seems to correspond to an epistemic community since there is an international network of professionals/experts with shared normative and causal beliefs, notions of validity and a common policy enterprise (Haas 1992, 3).

39 Strong’s biographer also stresses the agreement between the American and the British governors particularly regarding CBI (Chandler 1958, 260–261, 281). Later, Norman also found conceptions very similar to his in Hjalmar Schacht, President of the Reichsbank (1923-30; 1933-39), who had argued that central banks were “constantly hampered by the political authorities” (cited in Feiertag 1999, 364). Strong’s mistrust of political interferences in central banking was obvious when he engaged, to a certain extent, in the debates over the creation of the Fed. He criticised and opposed the Glass-Owen Bill, which eventually became the Federal Reserve Act, because it gave the control of the Federal Reserve Board to political appointees and government officials (Chandler 1958, 33–35). See infra Kisch & Elkin on the Fed’s independence.

40 See Barbaroux, Dal Pont Legrand, and Torre (2016) and Brillant and Rojas (2020) for interwar debates between French, American and British central bankers, and experts on the role of central banks in the money market.

41 CBI was a sort of second-best option for Bagehot. Like many other liberals (see footnote 21), his inclination for laissez-faire made him a natural advocate of free banking but he admitted that money and banking had to be “managed” and, for practical reasons, accepted central banking (1873, 20). Thus, the liberal position turns out to be the refusal of political interference in central banking.

42 Cecil Hermann Kisch (1884–1961), educated at the University of Oxford (Trinity College), entered the Indian civil service in 1908. According to Harrod (1951, 129), when Keynes resigned in June 1908, Kisch took over his position at the India Office. Private Secretary to the Secretary of State for India from 1917 to 1921, Kisch was appointed Financial Secretary to the India Office in 1921. The India Office was a British government department responsible for the government of the Indian Empire. It was headed by the Secretary of State for India, a member of the British cabinet. Kisch later became Assistant Secretary of State (1933–1943) and Deputy Under-Secretary of State (1943–1946) and, from 1938 on, was financial adviser to the LoN and the United Nations (for a short biographical note see Moggridge 1992, 889; for a more detailed biography see the obituary in The Times, 21 October 1961, 12). In 1932, he published The Portuguese Bank Note Case, an essay on the famous fraud of the 1920s and on the legal dispute between the Banco de Portugal and the printing company Waterlow & Sons. Throughout his career, Kisch was a civil servant and a policymaker at the India Office as well as a monetary expert (see below). His co-author and cousin Winifred Adeline Elkin distinguished herself in the study of Economics at the University of Cambridge (Newnham College) (see the review of Kisch & Elkin’s book in The Times, 23 January 1928, 20). A civil servant, Elkin was also a suffragist and penal reformer and penned two books on the English legal system.

43 Attention will also be paid to chapter IV entitled “Administration of Central Banks” and to chapter VI entitled “Relations between the Central Bank, Commercial Banks and the Money Market”. The book, published in January 1928, was reprinted in the following May (2nd edition) and in 1930 (3rd edition). A revised and enlarged edition (the 4th) appeared in 1932. The latter will only be referred to in the case of new and substantial content.

44 In the acknowledgements, the authors thank central banks and other national authorities for their collaboration during their research. They also thank Frederick Goodenough, Chairman of Barclays Bank, Otto Niemeyer, former Controller of Finance recruited in 1927 by the BoE as adviser to the governor, and Strakosch who read the book before publication. Both Niemeyer and Strakosch were close to Norman, and members of the Financial Committee of the LoN.

45 Along similar lines, Obst (1923, 285–286), a German economist, identified four types of central banks regarding ownership and the degree of government control over administration: pure state Banks (Russia), pure private Banks (BoE) and two mixed situations. In pure models, the situation is straightforward since ownership gives control. In mixed situations, the State can intervene in the administration even if the central bank is privately owned (most European central banks) or it can be involved only as a shareholder. For another German work on the history of banks of issue and central banks see Dierschke and Müller (1926),

46 “Money is too important to be left to the central bankers”, a famous remark attributed to French statesman Raymond Poincaré, is a good representation of that idea.

47 Bagehot made similar arguments against government interference in the management of the central bank and the French example of it (1873, 70–71): “(…) in theory, there is much to be said for this plan. The keeping the single banking reserve being a national function, it is at least plausible to argue that Government should choose the functionaries. (…). As that system exists, a logical Frenchman consistently enough argues that the State should watch and manage it. But no such plan would answer in England. (…). And the practical result for which we do care would in this case be bad. The governor of the Bank (…) would change as chance majorities and the strength of parties decide. A trade peculiarly requiring consistency and special attainment would be managed by a shifting and untrained ruler. In fact, the whole plan would seem to an Englishman of business palpably absurd; (…) that there are specious arguments of theory for it, would not be sufficient to his mind.” Palgrave (1903, 58–60) briefly refers to and shares Bagehot’s arguments.

48 See Plumptre (1940a, 23–29) for a critical assessment of Kisch & Elkin’s arguments in favour of CBI.

49 This is coherent with Goodhart’s (1988) argument that the emergence of true central banking assuming the LOLR function is associated with a non-profit-maximising and non-competitive behaviour. On that matter, the authors mention Keynes’ pamphlet The End of Laissez-Faire (2010 [1926], 272–294).

50 Concerning the term of office, the authors stress the fact that political appointees may be under pressure to align with the government if the term is too short and re-appointment is desired. A long term increases independence but may enable unsuitable governors to remain in office. The authors refer to different sources of intervention, from supervisory and veto powers to conflict settlement, that come under the heading of what is called operational independence. Limitations on accommodation to the government refer to restrictions on direct or indirect financing by the central bank.

51 The 1913 Act provided for a Federal Board composed of two ex-officio members and five (six after an amendment in 1922) members, including the governor, appointed by the President of the US with the approval of the Senate. Alongside this powerful nominated component, the elected component of the Boards of the Federal Reserve Banks was also strong with six out of nine Directors elected by the shareholders, member banks, i.e. the commercial banks which hold the entire stock of the Fed. Originally designed to be a bankers’ bank, in both senses of the term, a central bank acting as a LOLR and a central bank owned and controlled by the commercial banks, the Fed ultimately included a considerable degree of public control in its management. The Fed’s relative lack of independence from political authorities resulted from a compromise between political/government and private/banking influences. Hence, its particular mix of nominated and elected members. As stressed by Forder (2003), the political control was intended to counteract the banking influences and guarantee independence from banking. The strong nominated component was introduced by the Democrats and President Wilson who mistrusted the banking community.

52 Hawtrey (1928, 440–441) calls it the orthodox view according to which needy governments use central bank accommodation, an expedient only acceptable in an emergency and not as a proper way of conducting public finances. He doubts the accuracy of this point of view. Later, he argued along the same lines (1932, 267–268).

53 For instance, it was extensively used, especially the passages supporting CBI, during parliamentary discussions on a plan for the establishment of a central bank in Australia in 1930 (Hawkins 2010, 88).

54 Throughout his career, he was inevitably in touch with Keynes, another specialist of Indian currency. In 1919, their paths crossed at the Indian Exchange and Currency Committee - Kisch was secretary and Keynes a witness (Chandavarkar 1990, 73). More interestingly, Keynes corresponded with Kisch after the Royal Commission on Indian Currency and Finance of 1926 - both were witnesses and Kisch also submitted several notes/memoranda. Keynes commented on the commission’s outcome, the appropriate institutional structure for a future central bank and relative independence from imperial authorities (Ibid., 82, 106–107). See Bibow (2002) for Keynes’ proposals regarding the relations between central banks and the government in the UK and India.

55 Ulrich’s (1931) publication of a much less known French counterpart also corroborates the existence of an established doctrine by the turn of the decade. His first chapter on CBI also treats both political interference and the influence of particular groups, and specific criteria are also discussed.

56 As a “money doctor”, Kemmerer exported the Fed model by adding a high dose of independence from political control (Drake 1989; Kisch and Elkin, 1928, 20–21). As noted by Gomez Betancourt (2010a, 2010b), he was also a respected expert in monetary issues at home and actively participated in the debates over the conception of the Fed. Kemmerer (1918/1950) authored The ABC of the Federal Reserve System, a very popular book, of which twelve editions were published, prefaced by Strong.

57 In light of the available data, it could be argued that there was a wave of adoption in the 1920s, comparable to that in the late 20th century. This matter is beyond the scope of the present paper (see Do Vale, in press).

58 An important organ of the economic/financial multilateral oversight machinery initiated by the Brussels Conference, the Financial Committee arranged stabilisation plans following the conditionality principle. Such plans and the consensus at the time can be seen as the forerunners of the International Monetary Fund‘s adjustment plans and the underlying Washington Consensus (Pauly 1996).

59 Norman originally proposed the idea of a foreign governor, which was rejected by the national authorities since the General Commissioner supervising the whole plan was already a foreigner (Cottrell 2012, 53). Norman had considerable influence during the negotiations of some of the plans especially the Austrian one. This plan provided the LoN formula of central bank reform (Brown 1940, 347–349; Cottrell, Ibid., 53; Sayers 1986, 168). Not surprisingly, that central bank design very much followed the British model.

60 General provisions of the kind were almost non-existent and inevitably proclaimed subordination rather than CBI. The trend for increased formalisation became stronger in the post-WWII period (Kriz 1948, 565–566).

61 Addis, a Director of the BoE, considered that Basel was “to fulfill the dream of Genoa by the gradual development of the BIS into a cooperative society of Central Banks” (cited in Toniolo 2005, 20). The meeting announced at Genoa was never held but Norman did not abandon the idea of more institutionalised cooperation and, in September 1925, envisioned a central bank’s bank: “I rather hope that next summer we may be able to inaugurate a private and eclectic Central Banks’ ‘Club’, small at first, large in the future” (cited in Ibid., 30).

62 According to Mints (1945, 280), the vast majority of economists,at least in the US, still championed CBI in the 1930s. As a matter of fact, Mints refers to the removal of the two ex-officio members from the Fed’s Board, brought about by the Banking Act of 1935, as evidence for lasting strong advocacy. He also mentions the protest of the Economists’ National Committee on Monetary Policy, headed by Kemmerer and Willis, against the first version of that bill that provided for more political control over the Fed via greater presidential nomination powers. In the end, the government refrained from increasing political control and Fed’s independence may be said to have been reinforced. Nevertheless, Willis (1936) and Kemmerer (1936, 258–266) continued to be concerned about the increased centralisation of the system and excessive government influence.

63 There has been a parallel economic literature on regulatory capture in general. However, the literature only recently began to deal with the potential capture of the central bank by the banking/financial sector focusing on the composition of central banks’ boards of governors and revolving doors.