Work on Central Bank Independence Much of my work has been critical of the idea that central banks should be 'independent' from political control. Some of it seeks to identify flaws in the arguments widely taken as supporting independence, and some of it points to neglected respects in which independence may be harmful. One strand of my critique of the case for independence simply questions the purported evidence that it brings benefits. I argued in that the usual approach of the literature – which depends on measuring independence according to statutory characteristics and comparing the result with inflation outcomes – is flawed since there are several important reasons that statutes do not determine central bank behaviour. I expressed some of the same views in , but also moved on to criticise the attempt to measure 'accountability' by a parallel process of reading statutes – at least when the study in question is devoid of normative argument. In I sought in part to substantiate the view that these arguments make for an important criticism of the 'statute-reading methodology' by showing that the understanding of the 'independence' of the Federal Reserve System has undergone a complete change since that institution was founded in 1913, but that this is not traceable to statutory amendment. Rather, it is a consequence of a development of the political system and of the popular understanding of the appropriate behaviour of various parts of the government. In I made a further criticism of the same group of empirical studies, noting that even if statute-reading is appropriate, there are two crucial weaknesses in the evidence. One is that there are important instances where the authors of these studies have apparently selected their favoured measure of independence by reference to the extent to which it is correlated with inflation – thereby making their argument circular. Another is that, even taking alternative measures finally adopted by different authors, there is little correlation between them, and consequently no worthwhile agreement as to what statutory characteristics are constitutive of 'independence'. In the light of this, agreement that 'independence' matters seems to have little significance. In , I scrutinised one widely cited study of the relationship between independence and inflation, noting what appeared to be errors in the data and the use of the data which, I suggested, created the impression of a much more impressive relationship between inflation and independence than was apparent when the study's approach was implemented correctly. Jacob de Haan (who was not the original author) commented on this paper, and I responded in . A second strand of my work concerns the view that independence improves 'credibility'. One issue is simply what 'credibility' means in economic theory. Many authors, and in particular many advocates of central bank independence, seem to be mistaken about this issue in a variety of ways. In and , I have done my best both to explain what the theory, properly understood, really says, and to argue that, once it is properly understood, it quickly becomes apparent that it has almost no empirical value. In the first of these papers, I was particularly concerned to distinguish 'the credibility problem' from issues of political motivation of policy, and short-termist policy – two things with which it is routinely confused. In the second, I aimed to show that the theory of credibility is implausible, unsupported by evidence, and – perhaps more surprisingly – does not in fact consistently inform the actions of policymakers. Both of these papers suggested that bad policy is likely to be the outcome of poor interpretation of theory, making the question at hand much more than one of academic principle. Thirdly, is mainly concerned with showing that if, in fact, the theory is useful and informative, it in any case does not support the case for central bank independence. In this general vein, I made rather shorter comments on the appropriate interpretation of Rogoff's much quoted idea of a 'conservative central banker' in Forder (1997); and on the distinction between 'commitment' and 'reputational' solutions to the credibility problem in Forder (1999c). Finally, illustrated some of the ways in which misunderstandings over theory are manifested in errors in the analysis and recommendations of the literature, and Forder (2004) pointed to one likely source of continuing confusion in a particular paper which sought to survey opinions on the importance and sources of 'credibility' without indicating what the word was to be understood as meaning. So, neither theory nor evidence supports the case for independence. But independence may also be dangerous. One consideration, hinted at in some of the papers listed above, is that since the case for independence depends so much on the idea – albeit a mistaken one – that independence provides 'credibility', independent central bankers are almost bound to be more influenced than other policymakers by the apparent, but quite possibly misunderstood, demands of that theory, and thereby led into error. In other papers, I have suggested a rather broader case against independence. One consideration is simply that an excessive confidence in the power of 'independence' to solve policy problems might lead to flawed policy initiatives, disappointment, and political discord. In I suggested that misapprehensions as to the nature and benefits of independence are a good part of the explanation of the flawed construction of the institutions of the European Monetary Union. There has also, of course, been a great neglect of the issue of whether independence – particularly in its extreme forms – is democratically acceptable. Economists may be uncomfortable in such overtly normative terrain, but it is also true that the issue is often dismissed, or treated as trivial, rather than merely ignored. In I considered a considerable variety of responses in the literature to the issue of democratic accountability, finding none of them satisfactory, and tracing a good deal of the problem, again, to muddle over what the theory of credibility says. I also developed aspects of this argument, along with some specific commentary on the attitude taken by the European Central Bank to the issue of democratic accountability in . Even more surprising, perhaps, is the neglect of the issue as to whether independent central banks should be expected to pursue the public interest at all: ordinary presumptions of economics would have them pursuing their own. In I argue, again through the medium of a consideration of the ECB, that the older presumptions ought to be returned to centre stage. In Forder (2005a), noting that in certain respects the independence of the ECB is not nearly as secure as one is sometimes led to believe, I considered more fully what behaviour in defence of that independence is to be expected from it. It becomes clear, I think, that once one takes the step of seeing central banks as institutions with interests of their own, it is apparent that their release from public control ought to be a highly controversial matter – particularly in the absence of theory or evidence which properly supports it! These considerations clearly raise the issue of what it is that has made central bank independence so popular, and especially so popular amongst economists. I seek to answer this is Forder (2005b). Forder, J. (2003b) Central bank independence: Economic theory, evidence, and political legitimacy. International Papers in Political Economy. Reprinted in The rise of the market. ed P. Arestis and M. Sawyer. London, Elgar, 2004 pp. 145-180 Forder, J. (2004) "Credibility" in context: Do central bankers and economists interpret the term differently? Econ Journal Watch, 1(3) pp 413-426. http://www.econjournalwatch.org Forder, J. (2005a) The limits of 'independence' and the policy of the ECB. Public Choice, 125 December pp 431-444 Forder, J. (2005b) Why is central bank independence so widely approved? Journal of Economic Issues, 39 December pp 843-865 JAMES FORDER Balliol College, Oxford. PAGE 3