If all the economists in the world were laid end to end, they still wouldn't reach a conclusion.
--George Bernard Shaw
When I ask two economists a question I get two different answers, unless one of them is Lord Keynes, in which case I get three.
So complete is the consensus which now exists over the desirability of central bank independence that it is possible to forget how quickly it emerged. Yet it was only in 1988 that Alberto Alesina wrote, "This paper argues tentatively that independent Central Banks have been associated with a lower average inflation rate and may have been responsible for reducing politically induced volatility of monetary policy and inflation. (1)
In 1989, the Reserve Bank of New Zealand became independent. In the same year the Delors Report proposed giving unprecedented independence to the projected European Central Bank, and very soon after-in 1992-these proposals were largely implemented in the Maastricht Treaty. Meanwhile, Vittorio Grilli, Donato Masciandaro, and Guido Tabellini (1991) reaffirmed Alesina's claim that independence is associated with low inflation and further claimed that this benefit brings no cost in lower output. It did not then take long for Otmar Issing (1996), addressing a meeting of economists, to hold these propositions to be "two of the established findings of our discipline," and only a little after that the depth of feeling underlying this new consensus was made apparent by the language chosen by Rudiger Dornbusch, Carlo Favero, and Francesco Giavazzi (1998, 28), who said of the European Central Bank, "The idea of a political body that interacts with the ECB is shocking: Europe and the world have moved a healthy distance from short-term political control of monetary policy."
When the Bank of England became independent, in effect in 1997 and by law in 1998, a worldwide revolution in macroeconomic policy making was complete. In just ten years Alesina's tentative suggestion had become an "established finding" and been acted on all round the world, (2) and mere compromises of independence-let alone, one presumes, outright opposition to it-could be described as "shocking" by leading economists in an academic journal.
The speed of this development is perhaps further highlighted by comparison with the time taken for acceptance of other ideas from economic theory like those considered by William Baumol and Gerald Faulhaber (1988). They noted that none of marginal analysis, the idea of net present value, and peak-load pricing achieved quick acceptance although they all might have been expected to be of interest to individual profit maximizers. (3) One might have expected the requirements of the legislative process and the relation of the issue to macroeconomic policy, where controversy is often so fierce, both to have impeded the progress of central bank independence. (4) A closer comparison might be with free trade, which policy makers have never implemented with the enthusiasm with which they took to central bank independence. Another could be with the adoption of Keynesian policies-Herbert Stein (1969, 3), for example, noted that the "fiscal revolution" took thirty-one years in America. (5)
In light of these things, it is perhaps also surprising that there has been so little attention to explaining the speed with which independence has been adopted. There has been some analysis of the incidence of central bank independence, but only a limited amount of it has been directed to the explanation of the wave of reform since 1988. Although insight is certainly available from such work as there has been, it addresses policy reform rather than the creation of consensus among economists. This is unfortunate not least because the best explanations of reform would seem to depend on the prior existence of a professional consensus. An obvious candidate to explain that consensus may be that the benefits of independence have, as Issing implied, simply been established as scientific fact. However, there are a number of reasons to suppose that this will not do. In the light of this, I offer, like Alesina, only tentatively, some further suggestions as to why this idea has appealed so suddenly and so powerfully to so many economists.
Explanations of the Spread of Independence
One approach to finding an explanation of the wave of reform would be to identify the factors which explain differences in independence before 1988 and enquire whether any of them has undergone a radical transformation since. In fact, however, none of these lines of enquiry seems likely to be fruitful.
For example, John Goodman suggested that a government foreseeing an imminent loss of office might seek to hamper its successor with an independent central bank (1991). But there has been no great increase in the turnover rates of governments since then. William Bernhard pointed to the role of the central bank as provider of information as to the effects of alternative policies and went on to suggest that where distrust is most likely to arise and affect the relationship between the various groups supporting a government, central bank independence might be promoted as a way of avoiding disputes between them (1998). It is perhaps difficult to see that left-wing groups in coalition governments would really have welcomed such an arrangement throughout the postwar period since it requires them to believe that central banks have no constituencies to serve or interests of their own. But, in any case, there has not been the great change in the temper of coalitions that would be required to account for policy developments.
Alternatively, one might look to the work of Jakob de Haan and Gert-Jan van't Hag (1995). They tested a number of hypotheses, some of them suggested by Alex Cukierman (1994), to the effect that independence is associated with certain subsequent political and economic outcomes. If, as the authors suggest, these are to be understood as candidates to explain differences in independence, it would seem that the designers of the central banks must have been influenced by considerable foresight. That in itself may seem unlikely, and furthermore the only one of this group for which de Haan and van 't Hag find statistical support is the idea that independence is related to infrequent changes of government. That would seem to be particularly hard to interpret as the cause of a prior grant of independence. The wider survey conducted by Sylvester Eijffinger and de Haan (1996) cast even more doubt on the availability of helpful results from these kinds of enquiries.
Perhaps more plausibly, de Haan and van 't Hag also found some support for the idea that high inflation between 1900 and 1940 was associated with greater independence after that period, although they did not comment on the extent to which this result is dependent on the German experience. Nevertheless, there might be a suggestion that the trend toward independence in the 1990s is in some way linked to the previous inflation. Although this view may be superficially attractive, it should be noted that whatever the relation between the two is, it can hardly be a simple one since one of the outstanding features of the movement to independence is that it did not occur until after inflation ceased to be an immediate problem. (6) Whatever the source of the policy-making revolution was, it clearly cannot have been a response to inflationary crisis. Indeed, as Kathleen McNamara (2002) noted, the lag between the conquest of inflation and the adoption of independence would, if anything, seem to deepen the puzzle about what explains independence.
A rather different approach focuses on the identification of the interests which favor independence. One suggestion is simply that it is elected governments that benefit since as, for example, argued by Edward Kane, it allows them to avoid blame for policy failure (1980). Geoffrey Miller made a quite different argument on governments' interests (1998). He suggested that putting the control of inflation outside the power of elected governments makes it easier for them to strike credible bargains over real values where indexation is infeasible. Therefore, both governments wishing to strike such bargains and the lobby groups representing the bargaining interests favor independence. Alternatively, there may be particular groups with an interest in sound-money policies. Keith Bain noted that those in secure employment may favor independence (1998), and there is a long line of argument suggesting that labor or the poor generally benefit from inflation while capital and the rich favor price stability. Philip Arestis and Malcolm Sawyer argued in this spirit with specific reference to the question of who benefits from central bank independence (1996). Along the same lines Gerald Epstein and Juliet Schor considered a wide range of class interests as being among the determinants of independence before they wrote (1988). None of these, however, seems well placed to explain why independence should be so much the creature of the 1990s. That point was taken up by Gustavo Piga (2000), who, attributing the same view to Jean-Paul Fitoussi (1995), suggested that the aging of populations has promoted creditor interests. Even here the question of the speed of reform is problematic: populations are aging and class interests may change, but they do not do so suddenly. Yet central bank independence did arrive very suddenly.
Among arguments pointing to the interests served by central bank independence, however, those which emphasize the interests of the financial sector are perhaps the best known. Adam Posen (1993, 1995), taking up a theme well known from, for example, Coakley and Harris 1983, suggested that before the wave of reform central bank independence was explained by a certain measure of the power and interests of the financial sector. Others--notably Paul Bowles and Gordon White (1994)--suggested...