Distinguished fellow lecture: monetary policy and the benefits and limits of central bank independence

Abstract

Central bank independence is an important development in macroeconomics. It has become a lynchpin of modern central banking. The period since the Global Financial Crisis and the experience of the COVID-19 pandemic era posed new challenges for central banks, with potential implications for central bank independence. This paper reviews the scope and rationale for central bank independence and the reasons it spread internationally after the 1980s. It reviews evidence of the effects of central bank independence on inflation, output volatility, financial stability and fiscal policy. The paper considers political threats and contemporary policy issues that could influence the form of central bank independence and credibility. These issues include matters relating to assessments of monetary policy during the pandemic, the scope of central bank mandates, interpretations of inflation dynamics, the funding of central bank operations, and the interaction between fiscal and monetary policy.

KEYWORDS:

• Central bank
• independence
• mandate
• inflation
• monetary policy
• GFC
• pandemic

JEL Classifications:

• E42
• E52
• E58
• E61

Acknowledgements

This paper was presented as a Keynote Address to the New Zealand Association of Economists Annual Conference on 30 June 2023, held at Auckland University of Technology, Auckland. I am grateful to Laura Austin, Vanessa Balshaw, Chris Bloor, Alan Bollard, Matthew Brunton, Andrew Coleman, John Creedy, Punnoose Jacob, Ross Kendall, Lewis Kerr, Adrian Orr, Kate Poskitt, Adam Richardson, and Grant Spencer for helpful discussions and suggestions while preparing for this paper.

Disclosure statement

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

Disclaimer

The author is an External Member of the Reserve Bank of New Zealand Monetary Policy Committee. The views expressed in this paper are those of the author and do not necessarily reflect the views of other members of the Reserve Bank of New Zealand Monetary Policy Committee or of the Reserve Bank.

Notes

1 See Goodhart and Lastra (2018a) on rising political threats. King (2023) suggests the constituency for low inflation is eroding.

2 These de jure measures (i.e., based on legislation) of central bank independence trace back to Bade and Parkin (1978). Their approach was to specify four categories of CBI. Subsequent approaches derived indices of CBI (Parkin, 2013). Examples are Alesina (1988), Grilli, Masciandaro, and Tabellini (1991), Alesina and Summers (1993), Fry, Julius, Mahadeva, Roger, and Sterne (2000). The development of these indices is reviewed in Eijffinger and de Haan (1996) and Arnone, Laurens, and Segalotto (2006).

3 Crowe and Meade (2007) discuss the evolution of central bank governance since the 1980s and place the evolution of central bank independence in this broader context. De Haan, Bodea, Hicks, and Eijffinger (2018) discuss the development of central bank independence before and after the GFC.

4 Cukierman et al. (1992) first proposed turnover rates of governors. Although often used in subsequent studies, the suitability of turnover rates has been questioned by Dreher, Sturm, and de Haan (2008, 2010) who found the turnover rate is subject to endogeneity issues. Furthermore, the turnover rate of governors does not take account of the degree of independence of other members of a monetary policy committee. An example of measures of CBI based on surveys of central bankers is Fry et al. (2000). This was based on a one-off survey responded to by central bank staff. Such surveys are not easily replicable and rely on the accuracy of responses which could be influenced by what staff believe or would prefer, or would like others to believe.

5 See Cuikerman (1994). Alesina & Summers (1993) suggest the historical experience of hyperinflation in Germany may have raised German public aversion to inflation and its interest in a more independent central bank committed to price stability. 

6 The Federal Reserve Reform Act of 1977 provided more focus to the role of monetary policy by requiring the Federal Reserve to direct its policies toward achieving maximum employment and price stability and report progress to Congress (Federal Reserve System, 2022).

7 Goodfriend & King (2005). Goodfriend (2007) discusses the influence of the Volcker period and developments in monetary theory on modern monetary policy.

8 For example, Barro & Gordon (1983) and Backus & Driffill (1985) and earlier work by Milton Friedman, Edmund Phelps, Robert Lucas and others in emphasising the critical role of expectations. 

9 Evans et al. (1996) and Grimes (1996). The CBI reforms were enacted by the Reserve Bank of New Zealand (RBNZ) Act 1989.

10 The influence of the NZ legislation is mentioned by numerous commentators. Examples are Mishkin and Savastano (2001), Goodhart and Lastra (2018b), and Loungani (2023). Walsh (1995) concluded the principal-agent approach implicit in the novel accountability aspects of New Zealand’s approach was designed to sustain a socially optimal commitment policy.

11 High inflation in the 1970s and 1980s prompted more research into the effects of inflation on economic growth and how those effects are manifest. See for example Grimes (1991) and Johnson (1993) and the extensive literature cited in these papers.

12 Bernanke et al. (2001, p. 4) defined inflation targeting as ‘a framework for monetary policy characterised by the public announcement of official quantitative targets (or target ranges) for the inflation rate over one-or-more time horizons, and by explicit acknowledgement that low, stable inflation is monetary policy’s primary long-run goal’.

13 These include studies by Alesina (1988), Grilli et al. (1991), Cukierman et al. (1992), Alesina & Summers (1993), and in a similar way by Carlstrom & Fuerst (2009).

14 Some of the variables included in these studies to account for other influences on inflation include the exchange rate regime, degree of international openness, fiscal policy, size of inflation rates, and characteristics of labour market conditions (see Klomp & de Haan, 2010a and references therein; Ftiti, Aguir, and Smida (2017); and the references in footnote 15).

15 These various studies can be found in Jácome and Vázquez (2008), Bodea and Hicks (2015), Loungani and Sheets (1997), Acemoglu, Johnson, Querubin, and Robinson (2008), Landström (2011), and Klomp and de Haan (2010b).

16 Although their conclusions are robust to alternative de facto measures of CBI.

17 See discussions and research in De Long (1997), Romer, and Romer (1997), Romer and Romer (2002), Bernanke (2004), and Nelson (2005, 2022).

18 A review of research on contributions to declining output volatility during the ‘Great Moderation’ is in Buckle, Haugh, and Thomson (2003).

19 This point is articulated by Bernanke (2004) who couched his argument within the context of Taylor’s (1979) monetary policy efficiency frontier. The way improved central bank credibility can change inflation dynamics and improve the Taylor efficiency frontier is explained in Clarida, Galí, and Gertler (1999), Orphanides and Williams (2005a), and Buckle (2019) who also reviews evidence of the impact of CBI and inflation targeting on New Zealand’s inflation and output volatility.

20 See Cukierman (1992) and Mishkin (1996), while Berger and Kirsmer (2013) explore the potential implications for financial stability of the different incentives of politically determined and politically independent monetary policy decisions.

21 The basis for considering monetary dominant and fiscal dominant regimes is developed by Leeper (1991) and discussed in the context of recent events by Leeper (2023).

22 These papers and results are summarised by Strong and Yayi (2021, Table 1, p4).

23 These conclusions and qualifications are drawn from studies by Bodea and Higashijima (2017, which includes post-1990 years for 78 countries that include democracies, more autocratic regimes and dictatorships), Burdekin and Laney (2016, which covers Central and South American countries from 1990 to 2012), Strong and Yayi (2021, for 30 African countries from 1990 to 2017).

24 In 2018, he remarked that the Federal Reserve is ‘my biggest threat … Because the Fed is raising rates too fast. And it’s independent … ’ Cited from ‘Trump says Fed is his “biggest threat” because it is raising rates too fast’ (cnbc.com).

25 Demiralp and Demiralp (2019) discuss and analyse the erosion of democratic processes and systematic pressure placed on the Central Bank of Turkey. They find that commentaries from the President and Cabinet Ministers advocating a lower policy interest rate had a significant influence on the Bank of Turkey’s policy reaction function.

26 Based on official statistics. By April 2023, the rate had reduced to 43.68%. In the early 2000s, under CBI, annual inflation in Turkey fell from between 60% and 105% to below 10%. Evidently two factors are helping prevent a more severe currency crisis following the interest rate reductions and soaring inflation: an influx of funds from Russian depositors unable to deposit in other countries, and the government is seeking large transfers from Arab nations (Hardie, 2022).

27 For example, the Secretary to the Australian Treasury remarked in 2020 that in forecasting the effects of the pandemic, falls in Australian GDP of 20 per cent were being ‘seriously contemplated’ (Kennedy, 2020). Another example of the degree of uncertainty was the reaction of equity markets. The percentage falls in the S&P500 index on 12 and 16 March 2020 were similar to the falls on 28 October and 8 November 1929 (Hawkesby, 2020, p. 7).

28 The unique economic circumstances and considerable uncertainties faced by central banks during the outbreak of COVID-19 and the policy responses and thinking behind those responses are discussed in various central bank speeches and papers such as for example Orr (2020) and Debelle (2020b) for small open economies, and for the USA see Clarida, Duygan-Bump, and Scotti (2021).

29 Although the comparison of US and UK policies by Tenreyro (2021) illustrates there were significant difference among countries in the characteristics of the fiscal measures and the resultant employment outcomes.

30 For example, Rogoff (2022) has criticised the timing of the US Federal Reserve Board’s increase in interest rates and curtailing of quantitative easing. Wheeler and Wilkinson (2022) claim that central banks became complacent about their ability to maintain low inflation and are critical of the delays reversing accommodative monetary policies. Borio (2023) suggests central banks initially underestimated the strength and persistence of inflationary pressures.

31 An example of the benefits of ex post reviews is the evaluation of Australian fiscal policy during the COVID-19 pandemic available in Borland (2023).

32 Shapiro (2020, 2022) for example, identifies demand and supply shocks to inflation by classifying demand-induced components of inflation as those where an unexpected change in price moves in the same direction as the unexpected change in quantity in a given month; supply-induced components are identified as those where unexpected changes in price and quantity move in opposite directions. The New Zealand Treasury (2023) and Reserve Bank of Australia (2023) have replicated this approach using New Zealand and Australian data respectively. Some critics have used this type of decomposition to judge the performance of fiscal and monetary policies during the pandemic era.

33 This type of challenge is not new. Central banks struggled to understand the reasons for persistent low inflation and undershooting of inflation targets during the decade following the GFC.

34 Examples are Reserve Bank of New Zealand (2022) and Kashkari (2023). The recent review of the Reserve Bank of Australia (Australian Government, 2023) assessed the monetary policy framework, decision-making processes, use of macroeconomic models, and forecasting methods and has made sweeping recommendations.

35 The persistence of very low inflation and the challenges this posed for inflation targeting central banks is discussed in Tarullo (2017), Constâncio (2017), and Spencer (2017). This period prompted suggestions that inflation was a concern of the past and inflation targeting and CBI was no longer relevant, a suggestion that has been refuted by several, including Draghi (2018).

36 Most notably the Federal Reserve Board (FRB) and the Reserve Bank of Australia (RBA). The FRB approved a new policy framework on 27 August 2020 with the Statement on Longer-Run Goals and Monetary Policy Strategy. This included five elements including one stating ‘The Committee expects to delay liftoff from the ELB until PCE inflation has risen to 2 percent on an annual basis and other complimentary conditions, consistent with achieving this goal on a sustained basis and to be discussed later, are met’. The statement includes other elements suggesting monetary policy would remain accommodative for some time (Clarida, 2022). The RBA initially provided state-dependent forward guidance in March 2020, stating that ‘The Board will not increase the cash rate target until progress is being made towards full employment and it is confident that inflation will be sustainably within the 2–3 per cent target band’. The RBA added a time-dependent component in October 2020 and November 2020, and reiterated this in February 2021 by indicating a time horizon of three years over which the cash rate would be unlikely to change (Australian Government, 2023, pp. 41–42).

37 Some of the theoretical and empirical literature is contained in Sheshinski and Weiss (1993). See also Ball and Mankiw (1994). An example of empirical evidence that the frequency of price rises increases with inflation is in Buckle and Carlson (2000).

38 Reserve Bank of New Zealand (2023a) shows the sensitivity of inflation expectations to current inflation has been increasing.

39 This concern has also been raised in the context of the growing tendency for central banks to undertake a financial regulatory and supervisory role. Lastra and Wood (1999) expressed concern that regulatory supervision independence would give bank supervisors to some extent ‘the coercive power of the state against private citizens’. Quintyn and Taylor (2002) consider this is a power that has no equivalent with independent monetary policy.

40 For example, Bank of England, Bank of Canada, Sveriges Riksbank, Norges Bank, and the European Central Bank.

41 Notably the US Federal Reserve and Reserve Bank of Australia, whose legislation pre-dates formal inflation targeting. The Reserve Bank of New Zealand originally had a sole primary objective of price stability. This changed in December 2018 when legislation introduced in addition to price stability, the objective of supporting maximum sustainable employment.

42 For a more extensive discussion of these issues, see Rotemberg and Woodford (1997), Clarida et al. (1999), Blanchard and Galí (2007), Reserve Bank of New Zealand (2023b), and Jacob and Özbilgin (2021).

43 See for instance, Clarida et al. (1999), Blanchard and Galí (2007), and Jacob and Özbilgin (2021). Typically, these papers consider the implications for representative households and firms and not potential differences between different types of firms and households. In principle, a demand shock generating this ‘divine coincidence’ property may not be ‘divine’ for all sectors. But monetary policy instruments are typically not designed to differentiate between sectors and hence it is often more efficient to deal with adverse effects of demand shocks for some sectors using other (fiscal or labour) policies.

44 For example, Jacob and Özbilgin (2021) find that in the presence of significant cost-push shocks to price and wage inflation, welfare is enhanced when the central bank focusses solely on stabilising inflation.

45 Orphanides and Williams (2005a) evaluate related issues.

46 Coleman and Dasgupta (2023). Also relevant to the weighting scheme is the extent to which the behaviour of the objectives is correlated. The degree of correlation will depend on the nature of the shocks. 

47 A lexicographic ordering is applied by the Bank of England. Similarly, price stability is the primary objective of monetary policy for the Bank of Canada, Norges Bank and the European Central Bank, and consideration of employment is conditioned by the ability to satisfy the primary objective.

48 This concern has been raised by Goodhart and Lastra (2018b) in the context of the growing tendency for central banks to undertake financial regulatory and supervisory roles in addition to the traditional monetary policy role.

49 Alternative policy instruments are discussed by Lilley and Rogoff (2019) and include raising inflation targets, counter-cyclical fiscal policy, ‘helicopter money’, negative interest rate policy (NIRP), and forward guidance on interest rates.

50 Negative interest rates were used from 2012 in Europe, Denmark, Sweden, Switzerland and Japan. Reviews by The World Bank (Arteta, Kose, Stocker, & Taskin, 2016) and IMF (Brandão-Marques, Casiraghi, Gelos, Kamber, & Meeks, 2021) conclude NIRP in these countries supported demand and influenced exchange rates in ways similar to conventional interest rate cuts. Prior concern about the consequences for bank profitability, excessive risk taking, and financial instability did not eventuate materially, although they acknowledge these effects could become more evident if rates were substantially negative or used for longer periods. These reviews stressed the effects of NIRP and the extent they can be deepened are likely to depend on the structure of financial systems. Banks and banking systems more reliant on retail-oriented deposit funding for instance, could be more vulnerable to negative interest rates.

51 COVID-19 initially triggered a rise in demand for liquidity. In addition to their monetary policy decisions, central banks with financial stability responsibilities bought bonds on secondary markets to improve liquidity and mitigate risks to financial stability. For example, see Hawkesby (2020).

52 Hooley et al. (2023, p. 7), who also report that only a few central banks provided direct support to the non-financial sector, and for those that did, the size of support was relatively small in most cases.

53 Bullock (2022) discusses reasons for choosing quantitative easing during the pandemic, including a summary of the macroeconomic effects and issues relevant to assessing the impact on the Australian Government’s fiscal position. See also Reserve Bank of New Zealand (2022). 

54 Debelle (2020a) and Hawkesby (2020) discuss in detail how the RBNZ and RBA balance sheets changed during the pandemic, and how deposits of financial institutions at the central banks are remunerated.

55 The financial risks considered here are those that could potentially influence the choice of monetary policy instrument. They do not relate to a central bank’s operational budget. The fiscal risks referred to are limited to the direct effects on the central bank and therefore consolidated government balance sheet. There can be other ‘quasi-fiscal’ risks arising from arising from the effect of quantitative easing on the market for public debt and the efficiency government bond programmes (see Hooley et al., 2023).

56 Many central banks (including the Federal Reserve, European Central Bank, and Bank of Japan) value them at the amortised value and do not recognise unrealised profits or losses because they anticipate holding them to maturity (Hooley et al., 2023).

57 Bell et al. (2023, p. 4) note that in advanced economies, historically the net income of central banks has been a modest source of government revenue. But this can vary over time. When policy rates were continuously low following the GFC from 2010 to 2019, central bank income and corresponding transfers to government were relatively large. However, during the pandemic period, net interest losses for several central banks have led to smaller transfers to the fiscal authority or none at all, and in some cases probably for years to come (see Chart 1D, p1).

58 Unlike commercial businesses, central banks are not at risk of insolvency because they are typically protected from court-ordered bankruptcy, and are underwritten by government. Moreover, central banks have the ability to create ‘base money’ to finance operations. There are limits to the scope to generate their own funding despite a monopoly over base money creation, because there is a limit to the demand for liquidity, and because the extent of the recourse to seigniorage required to safeguard a central bank’s solvency might be of a scale that could undermine monetary policy objectives (Buiter, 2008; Reis, 2013).

59 These and other risks are reviewed and discussed by Nordstrom and Vredin (2022).

60 For example, the Bank of England is fully indemnified by UK Treasury via its subsidiary the Asset Purchase Facility. The RBNZ and Bank of Canada have indemnity agreements for specific operations without a subsidiary. At present, these are negotiated on a case-by-case basis which does not necessarily imply that indemnity will be provided in every situation.

61 See Archer and Moser-Boehm (2013) and Bell et al. (2023). The RBA for example, does not rely on indemnities. Recent RBA losses (including unrealised mark-to-market valuation losses) have been partially funded by drawing on reserve funds accumulated from previous retained profits and by moving into negative equity. The RBA is able to continue to operate via its ability to create base money, and has indicated to Government it expects future profits will be retained until its capital is restored (Bullock, 2022).

62 An example is Kengmana (2021).

63 There were several arguments against the active use of fiscal policy for macroeconomic stabilisation, including potential offsetting private behaviour, implementation lags generating pro-cyclical effects and accentuating cycles, incentives of politicians and time-inconsistent discretionary fiscal policy. These are perennial issues (Alesina, 2000; Barker, Buckle, & St Clair, 2008; Brook, 2013; De Haan & Gootjes, 2023).

64 In the United States for example, the federal government provided large and highly concentrated support in the form of ‘stimulus checks’ sent directly to households (Brunnermeier, 2023). The fiscal support programme implemented in Australia is discussed in detail in Borland (2023).

65 The fact there was a combination of fiscal and monetary responses provoked suggestions that a form of ‘fiscal dominance’ was at play. For example, Lingle (2021) described central bank responses to the GFC and the COVID-19 pandemic as indicating they were becoming a ‘handmaiden to profligate fiscal policy’. In contrast, Haldane’s (2020, p. 17) interpretation of the circumstances in the UK is that the combination of fiscal and monetary responses did not imply ‘one policy driving the other, much less of monetary financing of government deficits’.

66 This point has also been made by the review of the Reserve Bank of Australia (Australian Government, 2023) which suggests there should be increased joint work between Treasury and RBA on the relative roles of fiscal and monetary policy.