The fallacies of central bank independence

Abstract

Mark Carney, Value(s): Building a Better World for All. William Collins (2021), 608 pp. ISBN: 978-0008463618 (hb, £30.00); 978-0008421199 (pb, £10.99); 978-0008421144 (e-book, £5.99) Ben S. Bernanke, 21st Century Monetary Policy: The Federal Reserve from the Great Inflation to COVID-19. WW Norton & Co (2022), 512 pp. ISBN: 978-1324020462 (hb, £25.00); 978-1324020479 (e-book, £25.00) The Bank of England Act 1998 marked the practical end of an extraordinarily rapid global revolution in the institutions of economic policymaking. By the time the Labour government, elected in 1997, surprised the world just days after the election with a plan for independence of the Bank of England, dozens of other central banks had already been made independent, or created as independent central banks, and Britain had become notorious for lagging behind this global trend. ‘Free at last’ was the headline chosen by The Economist on 8 May 1997 celebrating Labour's decision, as if there had been a great war against an imperial power. ‘Resounding welcome for independence day’ was the choice of the Financial Times on 7 May 1997, with obviously similar reference. Twenty-five years and several global crises later, the arguments can be appraised in a new light, and two recent books by central bankers themselves – Mark Carney (Governor of the Bank of Canada 2008–13 and Governor of the Bank of England 2013–20) and Ben Bernanke (14th Chairman of the US Federal Reserve 2006–14, and then Nobel Laureate in 2022) – offer fruitful study. There is a temptation to suppose that the case for independence grew directly from the inflationary experience of the 1970s – as ‘political’ policy had been shown to be inflationary, so ‘non-political’ had to be tried. Although reference was often made to the 1970s and the lessons that could supposedly be learned, a scrutiny of the facts raises considerable question. It was, after all, the very political Margaret Thatcher whose government controlled British inflation. The inflationary French experiment of the early years of the Mitterrand presidency was reversed by Mitterrand himself. Even in the United States, politics is there in the appointment of the Chairman of the Board of the Federal Reserve, and though credit is often given to Paul Volcker as the ‘independent’ chairman who brought inflation down in the Reagan period, it is to be remembered that it was President Carter whose political decision it was to appoint him in 1979. seek to influence the members of the decision-making bodies of the ECB or of the national central banks in the performance of their tasks.1 Not only was the central bank to have all the power, but it was presumed to have all the wisdom as well, since the member states are apparently enjoined never to present it with a persuasive argument. Another part of the picture in the adoption of ‘independence’ was the related events of collapse of the Soviet system, dissolution of the USSR, and democratisation in eastern Europe. These things created a number of new states, and these and others found themselves in need of institutions of governance, but little or nothing in the way of recent democratic tradition in which to ground them. For them, an executive body to run monetary policy was a convenience, as well as being, increasingly after Maastricht, a sign of modernity. If these circumstances went a long way to defining the world environment in which the consensus on central bank independence appeared, there were also very notable developments in economic research. It is a surprising thing because of the way the academic literature was transformed. Up to the late 1980s, there had been occasional discussion of central bank independence, usually treating both questions as to how it might be achieved or maintained, and those of its effects, as serious matters of scholarship. They might be approached from different perspectives, and some might feel there were clear-cut answers. But there was nothing in the way of the pretence that was to develop later, namely that it was all too easy for words, just a triumph of either the righteous or the commonsensical. So, for example, Friedman (1962) was one who criticised the idea of independence for both its anti-democratic orientation and its likely adverse effects on policy. His views on the former rested in part on what he had learned of the interwar experience from Moreau (1954) – a book by a central banker, portraying the dismissive attitude of central bankers to anyone else's view. On the latter aspect, Friedman emphasised particularly that an independent central bank would very possibly be unduly influenced by the banks. And he feared that “since the banking community is concerned primarily with the credit market, central banks are led to put altogether too much emphasis on the credit effects of their policies and too little emphasis on the monetary effects of their policies” (1962, p. 238). Perhaps it is just a generalised version of that point, but Chant and Acheson (1972) subsequently saw perfectly clearly that in so far as central banks have independence, they can be expected to pursue their own ‘prestige and self-preservation’. Their idea of ‘prestige’ included such things as the perceived importance of their goals, and the public's and others' opinion of their performance in achieving them. ‘Self-preservation’ promoted the minimisation of conflict with those who diminish its status, and the maintenance of defensive possibilities lest conflict should arise. These ideas, they thought, could throw much light on the operation of monetary policy. Others placed more weight on the historical study of actual central banking behaviour and relations with governments. In the course of his study of the Federal Reserve, Woolley (1984) noted a potentially important distinction between the ‘independence’ of policy in the sense of the central bank being able to set its desired policy and its ability to achieve its desired goals. In general, no agency controls all the policy levers relevant to its objectives, so even a central bank which is ‘free’ may still need the cooperation of other actors, and hence not be so ‘independent’ as it might seem. There was plenty to argue about in these kinds of studies, but they were based on analyses of the incentives faced by agents and the history and institutions of policymaking in the countries in question. That would be a natural way, one might think, of addressing so complex a matter in political economy. All that sort of analysis, though, was swept aside, when in the late 1980s three developments seized economists' minds. One was that a collection of studies seeking to measure independence according to the letter of central bank statutes was treated as providing a means of testing propositions about ‘independence’. Alesina and Summers (1993) has remained a much cited, though tremendously overrated, example of the genre. A second was a conviction that the 1980s saw the discovery and acceptance of the hitherto unappreciated proposition that high rates of inflation would not systematically either maintain employment or spur growth. And the third was that the ‘credibility’ of policy – that is, of monetary policy specifically – was both essential to its success, and somehow created or enhanced by the granting of independence to a central bank. None of these would have stood up to critical commentary at all, and the fact that they were so readily accepted shows, more than anything else, that there was very little in the way of criticism, and no willingness at all on the part of the advocates of independence to consider it. Perhaps it was just that the zeitgeist of central bank independence demanded acceptance. But it is difficult altogether to resist the view that the idea of opening up places on policymaking committees to career academic economists with neither banking nor policymaking experience may have helped to dull the sceptical senses of some ambitious souls. The idea that the statutes of a central bank could be read and scored-up to discover the degree to which policy was set without regard to political or short-term interests was a true classic of what McCloskey (1990) called the ‘lawyerly mistake’. It is the structure of incentives and constraints that determines the effect of any rule, not the legislators' intent in drafting it. One waits for the advocates of independence to tell us that we can tax an industry of perfect competitors without the cost being passed on to consumers if only the law will make clear that it is the producer who is to pay. It would be an illiterate proposition, but what of the one that we know the elected government cannot control monetary policy because the law says the central bank does that? When something with a behavioural aspect was introduced to the assessment of central bank independence, it sometimes turned out to be only an opportunity for circularity, such as when it was discovered by Cukierman et al. (1992) that in developing countries inflation was unrelated to the statutory measures that had proved so satisfactory in making the case for independence in developed countries, but was related to the turnover rate of central bank governors. That showed that in developed countries independence was measured by a statutory index, and in developing countries by the turnover rate of governors. In both, the hypothesis that independence reduces inflation was thereby confirmed! Of the idea that it had ever been believed that a systematic inflation could actually reduce unemployment, I shall pass over with a mere note,2 but the renewed emphasis on the idea that it could not, which appeared in the 1980s, carried with it a further proposition of very dubious value. This is that there was nothing monetary policy could usefully do, other than control the rate of inflation. Asset prices were simply forgotten; almost no one seems to have thought of the possibility that in a small open economy the volatility of the domestic price level and of the exchange rate may be substitute evils, needing to be traded-off as such; and then there is the point that monetary policy might have distributive effects. The crucial point was that the supposed absence of these things from the list of monetary concerns made it seem viable for central banks to be treated as agencies with a single objective on which they could be assessed. On this basis, so it was said, concerns about democratic legitimacy could be set aside: the bank had one job; policy faced no difficult trade-offs; and the performance of the individuals put in charge would be easy to assess. So far, so silly, but the incorporation of the ‘theory of credibility’, usually said to be derived from Barro and Gordon (1983), into the case for independence created some sort of sheen of academic sophistication, as well as a route to a semblance of an explanation of why economists who suddenly saw nothing to limit enthusiasm for this old idea had not voiced support for it earlier than the 1980s. How could they have, when it was only this newest of theories which made the case? The importance of the idea of credibility arises from the fact that monetary policymakers face a difficulty in controlling the expectations of economic agents such as wage bargainers. The general shape of the problem is that the central bank may wish to influence expectations by undertaking or even promising that policy will be set in some specific way in the future, but – in the cases of interest – if that undertaking is believed, the resulting private sector behaviour will create a situation in which it is not optimal to set the policy that had been promised. Consequently, if the private sector understands these things, it will not believe the undertaking, and it lacks ‘credibility’. A much-cited specific example is that a central bank will promise to deliver low inflation in the hope of inducing the private sector to set low wage increases; but if it does, the central bank can, by allowing inflation to rise, achieve a desirable fall in unemployment. Hence the private sector is led in fact to anticipate that inflation, and sets high wage increases. In that case, even the possibility of controlling inflation may depend on willingness to accept recession. Despite the fact that the theorists of credibility, going back to Kydland and Prescott (1977), and certainly including Barro and Gordon (1983), cast the problem as one faced by ‘the policymaker’, and clearly had the Federal Reserve in mind, the idea that changing the identity of the policymaker dissolved the problem took root.3 So it was said that independence would achieve, and was the only practical way to achieve, ‘credibility’. This mistake could be so widely accepted no doubt partly because the mathematics of the theory is none too easy to understand and partly a result of a fashionable contempt for democracy generally. But it surely also was that just because it became so routine in intellectual circles to agree on the importance of ‘credibility’, whatever that meant, and on the importance of independent central banks in achieving it, however they did that. And as to the question of the effectiveness of monetary policy in achieving any goal other than the control of inflation, that was really just a matter of writing down models which included no other possibilities. If one's model has no role for asset prices; nor financial stability; nor, for that matter, achieving a ‘green transition’, then of course it will not suggest that monetary policy be used to achieve any of these. The fact that some objectives now being pushed onto central banks are absurd does not make it less shocking that there was so little comment on the point that the case for independence was made on the basis of models of such narrow scope. Mark Carney's Value(s) is a book about more than just central banking, but not so much more than the author may have hoped. Playing on the idea that ‘value’ is something well understood in economic terms, but ‘values’ demand a broader perspective, he writes 100 pages on the history of money; 150 on ‘Three Crises of Value(s)’ – these being the financial, Covid, and environmental crises – and another 100 on ‘Reclaiming our Values’. There are certainly countless examples of pride coming before the fall in finance. Think of those who dubbed the period before the global financial crisis the ‘Great Moderation’. Or the four most expensive words in the English language, ‘This time is different’. (p. 343) Indeed, think of them. It was Bernanke (2004) who made the idea of the ‘Great Moderation’ famous when speaking as a member of the Board of Governors of the Federal Reserve. There, he argued that the idea that outcomes had been so satisfactory because of the good luck of the world being free of shocks was understating the quality of policy that had been pursued: that is, though he did not quite say so, the quality of policy that had been pursued since central bank independence had taken such a prominent role in thinking. If one were looking for an explanation of way the feeling that ‘this time is different’ took hold in the years before the financial crisis, it is surely to be found here. That time it was different because central banks had become so good at running the economy, and governments so wise in letting them, one might have been led to think. Bernanke's 21st Century Monetary Policy is a history of monetary policymaking in the United States over nearly the whole period since 1945, but with increasing detail as time goes on. In its conception its goal is to describe policy in the light of three developments Bernanke perceives. These are that the relationship between inflation and unemployment has changed; that the normal interest rate is in long-term decline; and that systemic risk to the financial system is growing. But really, although those developments provide the context for Bernanke's discussion of policy, it is much more as if the real objective of the book is to describe an idea of the ‘independence’ of the Federal Reserve, and justify it. The history is fascinating. The earlier parts describe the post-war rise of inflation, its common attribution to cost-push causes, the inadequate monetary response, and Volcker's much more determined approach to bringing it under control. That story is familiar enough, but Bernanke's account moves along very well. The Greenspan era gets more attention, but there is even more on the response to the financial crisis and the pandemic, with Bernanke himself having been Chair through the former. In these later parts, the detail and obvious command of the material that Bernanke brings to the account make the book a valuable resource and certainly essential for serious study of the issues. He, too, seems not really to have comprehended that the case for independence might be in doubt. That much is apparent from remarks he makes about responses to challenges to independence that come up at various points. To take one case, the ‘Audit the Fed’ proposal of 2010 came and went, being briefly a “nightmare” (p. 164) for the Fed, and in the end showing that “the Fed cannot ignore politics” (p. 165). What he means is that the Fed cannot ignore the politics of defending its own independence. And it all shows that “we needed to make our case to the broader public” (p. 167). There even comes a point where Bernanke notes that an argument made against a certain policy was that some possible outcomes “could provide powerful ammunition to our numerous critics on Capitol Hill” (p. 187). It is not only, though, that all this suggests Bernanke does not see that there is a real argument to be had, but it also makes it very clear that the Fed is very far from being ‘independent’ in the sense of being above politics. It may have behaved in a way that was ‘non-political’ in the sense of not favouring a particular party or particular candidates, but party politics is only one kind of politics. Quite clearly, and on Bernanke's own account, the Fed is very much a political actor on its own account, and acting in its own interest. A further point of note is how far from bearing out the case that was made for independence Bernanke's account is. For one thing, there is a great emphasis on the personalities involved. So, to take a notable case, of Volcker, Bernanke says that good economic performance required credibility, and that required independence, and that because he had had the support of Carter and Reagan, Volcker had enjoyed more independence than his immediate predecessors (p. 43). So here we simply throw away Alesina and Summers (1993) and all their kind. It is not the central bank statutes we need to read, but either the personalities or the political goals of the president. It is having the support of the president that makes the central bank ‘independent’! Evidently, it is also not the case that, in the view of central bankers, monetary policy has no role except the maintenance of price stability. That is surely an uncontroversial lesson of the quantitative easing experience. That these extreme measures are presented as having been absolutely necessary does not leave any room for the naive pronouncements of the 1980s literature. That being so, of course, the questions of accountability and how one assesses the quality of policymaking regain their position at the centre of discussion of the institutions of policymaking. No longer can it be said that central bankers, with just one goal, an easy means to achieve it, and easy ways for everyone else to assess their success should be free of the ordinary processes of democratic accountability. Numerous specific issues might be raised – the question of how many bailouts there should have been in the financial crisis is obviously one. But, of course, Lehman Brothers attracts attention. On this, Bernanke's account seems to raise some questions. He explains that the failure of a widely interconnected firm would induce panic throughout the market, and that was the reason for organising the rescue of Bear Stearns in March 2008 (pp. 125–6). However, attempts to save Lehman by a similar approach of having it taken over by another bank led nowhere, so it was allowed to fail in September, whereupon the panic grew worse. Bernanke puts much emphasis on the experts' opinion that Lehman was in fact insolvent rather than merely short of liquidity, and says that, for this reason, the Federal Reserve was not permitted to act as lender of last resort. So there was no option but to let it fail. In a fuller account of the same events, Bernanke, Geithner, and Paulson (2019, p. 70) say “We would have rescued Lehman if we could have”, explaining the fact that Paulson (Secretary of the Treasury 2006–09) had said that there would be no such rescue as being merely a “negotiating tactic” (p. 65) to try to induce a private sector rescue. There is obviously a muddle somewhere in these accounts since, if the Fed was prohibited from contributing funds, then Paulson's remark was just a statement of fact – though, of course, it would be an oddity that he made it sound so much like a policy decision rather than making reference to the law. Apart from anything else, that would have been a better negotiating tactic. But there is a further question raised by the work of Ball (2018). His detailed analysis of Lehman's balance sheet and treatment of the events aims to show that the Federal Reserve did have the power to lend to Lehman, and argues that, had it done so, a long-term rescue might well have been possible. It is a complex matter, and there must be room for debate. But Ball clearly does show that there were inconsistencies in official accounts of what was done, and that the law is – at a minimum – not clear. It is a hard one to adjudicate, but Ball's argument can hardly be said to be negligible. And yet it receives no response from Bernanke. There are many questions there, but the question for the advocates of independence is whether it is to be argued that if the Fed made a mistake or – as Ball suggests – acted politically, its lack of accountability for that is just a matter of public indifference. The executive agency staffed by ‘experts’ takes the decisions, and the rest of us just live with it as it happens, and even after the fact can only write books about it? On the question raised by Woolley (1984) all those years ago, Bernanke is on the side of the independence sceptics, though he does not seem to recognise it. The outstanding example this time is the necessity – as Bernanke saw it – of persuading Congress to take legislative steps to address the fall-out from the financial crisis. Precisely the point was that the Federal Reserve simply did not have the powers to achieve its goals – it could not be ‘independent’ in that sense. On the question of ‘credibility’, it is hard to see that the question is even understood in the terms in which the point was so firmly made in the past. Fretting over hypothetical attempts to surprise the labour market with bursts of inflation has, thankfully, disappeared, but the issue of ‘forward guidance’ takes its place. This idea makes its first explicit appearance in 2003, when the Fed started to drop hints that the market was overestimating the likelihood of interest rate increases, and in due course moved to indicating that it expected interest rates to remain low for some time. The difficulty with this, from the point of view of the idea of credibility, is that it is one of these statements about future policy which, if it has the intended effect, will not be one the central bank wants to implement. Here, if the promise of low interest rates brings recovery, that will be what makes the central bank raise interest rates. Such thoughts are commonplace in the academic literature, such as Svensson (2003). But it is clear from Bernanke's treatment that the Fed was far from thinking that persuading anyone of the sincerity of its commitment was its problem. It would hardly have started with a remark as difficult to appreciate as: “the probability, though minor, of an unwelcome substantial fall in inflation exceeds that of a pickup of inflation from its already low level” (p. 96). On Bernanke's account, what it then did was to try to make its meaning clearer. Quite what the problem was is none too clear. Why should the central bank be, as it seems it was, trying to achieve a precise calibration of the strength of its hints? If it wanted the markets to understand its thinking, what about just telling them what it was? Anyway, there is nothing there, hint or otherwise, to suggest that the Fed thought its problem was that the markets were understanding what was being said about the future and not believing it. It is all much more like the bad old days of central bankers trying to obscure their meaning so as to enhance their status by magnifying what Goodfriend (1986) called the ‘mystique’ of central banking. There is not much in plain English, after all. So, whether it be a good or bad idea, the case for central bank independence cannot be found in the arguments that so completely persuaded the economics profession in the 1980s. We can be sure that central bankers think it is a good thing, but that is Chant and Anderson all over again. In all this, Bernanke is complacent too, though he may not realise it. His book is full of discussion of threats to the independence of the Federal Reserve, and there are occasional signs of some anxiety about how best to respond. No complacency there. But like Moreau, Bernanke is so completely convinced of the case for what he calls ‘independence’ that it is hard to see that he has realised there could be legitimate doubts about it. The view is on every page and his attitude can be summed up by a remark about how his successor, Jay Powell, responded to President Trump's habit of expressing his own views about monetary policy and its makers. In Bernanke's words, Powell “regularly and publicly emphasized that independence allows the Fed to make decisions in the public interest based on objective data and analysis and free of short-term political considerations” (p. 230). Powell emphasised that independence allows … He did not put that case, or emphasise that opinion, or express that view. He emphasised the fact that independence allows … Quite so. For Bernanke it is just a fact. It is a fact that some challenge, but they are ill-motivated, or possibly stupid, or I suppose perhaps just mistaken. There are plenty of them, and the Fed must be on its guard – but the value of defending what is the Fed's cannot be doubted. As Volcker (1990) said, at the end of the 1980s, central banks had come to be highly regarded. He tied that, perhaps too closely, to the control of inflation in the preceding years. The ‘Great Moderation’ did nothing to damage enthusiasm for them. That moderation failed to prevent, and may have helped to cause, the crisis of 2007–08, but on most views the central banks' response to the crisis was appropriate; tolerably well-managed, all things considered; and sometimes quite brave. Are we going to say the same of the handling of Covid? In the fall-out from the financial crisis, the effect of quantitative easing made monetary policy loose. The crisis was to a large extent a crisis of liquidity, and policy provided liquidity. This created circumstances where aggregate demand fell, and on one reasonable view a collapse of confidence could have led to a very severe recession. And in the fall-out, banks were required to hold much greater reserves than before, so it was opportune for the policymaker to provide them. Covid also saw a large amount of quantitative easing. But there was no crisis of liquidity – far from it. Sure enough, demand fell, but that was for much the same reason as supply fell: people were told to stay at home. Did that create a danger of a crisis of confidence, leading to a great depression? And if it did, was increasing the money supply a sensible way of addressing it – how would that bring locked-down workers back to work? The response to Covid was therefore oddly similar to the response to the financial crisis. Indeed, Bernanke describes the Fed's Covid response as a return to its “2008 playbook” (p. 260). Yes, but inflation was well under way before the war in Ukraine, and the question is whether the same playbook was appropriate to a quite different problem. It is difficult to see that the monetary policy of the Covid period should escape a good share of the blame for inflation, and it is going to be hard to argue that it arose because of ‘political interference’ – particularly as we have Bernanke's testimony that it was the central bank's own playbook they were using. The case for central bank independence, as it was made in the economics literature in the 1980s, was a scholarly disgrace. It was, first of all, intellectually a disgrace. The much-vaunted ‘credibility theory’ did not even support the proposal; the ‘lawyerly mistake’ of reading statutes to discover behaviour flew in the face of even a first-year undergraduate's understanding of the significance of incentives and constraints; the simple correlations in which such measures of independence were used to demonstrate its effect on inflation would be rejected out of hand by examiners for a Master's degree, even without the circularity of selecting measures according to their correlation with the variable to be explained. It was a professional disgrace, too, in eliciting so little critical or contrarian comment. There was a gigantic groupthink. And it does not alleviate the matter that there were so many economists who might have thought they had something to gain from the acceptance of the ideas. Complacency about the desirability of independence followed the near-unanimity of opinion that developed. That is threatening, and sometimes terrifying, in itself. But complacency has its dangers. One of them is that one forgets the arguments supposed to support the case. The case for rejecting the arguments of the 1980s should have been plain from the beginning, but they certainly are now. Bernanke has written a fine book in the tradition of the studies of central banking of the early post-war period. It is fantastically well-informed, insightful and clever about the policy of the Federal Reserve, and about the challenges it has faced. It is a fine work of political economy. But in being such, it shows in its own way the irrelevance of so many of the arguments of the 1980s. At each point, Bernanke's testimony or his evidence shows that the points made to support independence are hollow and vacuous, and should never have gained credence. It is complacent in another way. It is complacent about all being well while central banks are running the show. It was completed just when the first challenge in the form of resurgent inflation was appearing. Independent central banks had surely very much caused that; they certainly failed to realise it was happening. It is no longer true, as had previously been said to be so important, that central bankers had only one job – they had acquired or been given so many others. But it is still true that one job – controlling inflation – is the very essence of the advocacy of independence. We will see what the future brings, but one can guess what Milton Friedman would say, since his concern had been precisely that central bankers, left in charge of policy, would be too much concerned with interest rates, and not enough with the quantity of money, and hence, as he thought, would be ineffective in the control of inflation.