II. Degenerating Fiat Standard and Credibility
- Carlo Cottarelli, and Curzio Giannini
- Published Date:
- December 1997
Institutions do not evolve in a piecemeal fashion for a number of reasons. First of all, adapting an institutional framework involves collective action at several levels (North (1990)). Second, and perhaps more significant, adaptation must be preceded by widespread recognition not only of the need to reform the existing setup, but also of the set of feasible reform options. Such a choice is fraught with difficulties, especially since often no theoretical benchmark exists against which to evaluate the welfare properties of alternative reform schemes. This happens because the need for institutional reform is typically predicated on imperfections or forms of market incompleteness that prevent the attainment of Pareto optimality. As argued by Demsetz (1967), in such circumstances, it would be incorrect to gauge the relative attractiveness of alternative reform schemes by contrasting them with the unattainable optimum. When dealing with institutions, the proof of the pudding is inescapably in the eating.
All this helps to explain why the adaptation of most real-world institutions seems to follow a recurrent pattern, hinging on three phases: a period of “incubation,” during which pressures mount, without leading, however, to an explicit demand for reform; a “crisis” phase, during which there emerge new phenomena, not easily explained on the basis of existing institutions or conceptual models; and, finally, a phase in which the need for adaptation is increasingly recognized, with the emerging consensus tending, however, to favor parsimonious reform schemes, that is, schemes that retain as much as possible of the extant setup.
The adaptation of the monetary framework is no exception to this by now well-recognized pattern.2 This section shows that the wave of alterations in the monetary frameworks that followed the demise of the Bretton Woods system conforms to this pattern. The wave was set in motion by the emergence of a previously unknown macrophenomenon—stagflation. In the lively debate that ensued, proposals for far-fetched reform eventually gave way to less ambitious schemes, which carried the promise of ridding the fiat standard of its inflationary bias without sacrificing its superior flexibility with respect to other standards.
Degenerating Fiat Standard
After the switch from the gold standard to the fiat standard, which was for all practical purposes completed in the postwar years with the nationalization of most of the world’s central banks, matters related to the monetary standard practically disappeared from both theory and policy discussions. In spite of Irving Fisher’s famous warning against the perils of fiat money, inflation remained for more than two decades at about 2 percent in industrial countries, hardly an unbearable rate. At the same time, the world economy developed at an unprecedented pace (Flood and Mussa (1994)).
The international monetary system created at Bretton Woods, which exerted a disciplining influence on national economic policies, deserves part of the credit for this remarkable performance. But other forces were at play. The first two decades after the war, for instance, were marked by widespread skepticism about the effectiveness of monetary policy. As a consequence, the monetary lever was hardly used: rather than actively stimulating the economy, which was rapidly growing anyway, the authorities preferred to concentrate on stabilizing long-term interest rates (Guitián (1994b)). Thus, in the context of rapidly growing economies, authorities did not need to avail themselves fully of the flexibility the fiat standard allowed, which lay at the root of Irving Fisher’s misgivings.
Things started to go wrong when, toward the mid-1960s, growth rates slowed considerably throughout the industrial world. In the context of lower growth, the fixity of exchange rates mandated de facto, if not de jure, by the Bretton Woods institutions, started to be felt as a constraint on domestic economic policies. As a result, among other things, of monetary policies turning increasingly activist, inflation started accelerating in most industrial countries in 1963–64 (Bordo (1993)).
The history of international monetary relations in the second half of the 1960s is by and large the story of the attempt to fix what had by then become a faltering monetary framework without confronting squarely the fundamental problem, namely its incompatibility with the mounting demand for policy activism. One often-overlooked feature of the period marked by the fall of the Bretton Woods’ exchange rate system is the conflict that developed between central bankers and politicians over the appropriate conduct of monetary policy. One after another, in many industrial countries, central bank governors were either abruptly dismissed or duly subdued (Giannini (1994)).
This is not the place to review in detail the enormous literature on the causes of the inflation upsurge of the 1970s. For our purposes, suffice it to note that available evidence suggests “that any reasonably plausible explanation … will include substantial components of ignorance and error … and Government irresponsibility” (Smith (1992)). Both factors, however, would have been harmless without popular emphasis on the need to combat unemployment at the expense of the inflationary consequence of monetary activism. As former Federal Reserve Board Chairman Arthur Burns lamented in his self-defense against the allegation of monetary laxity, the search for monetary flexibility reflected deeper and for the while uncheckable trends in society (Burns (1987)).3
Thus, by the mid-1970s the fiat standard, finally freed from the constraints set by Bretton Woods’ multilateral peg system, had become a degenerating monetary framework: the passive monetary stance of the first decades after its establishment had been replaced by an activist stance, spurred by popular concern that the era of fast growth was in danger.
Debate on Monetary Standard
Support for activist monetary policies started to falter when inflation began to be associated with rapidly mounting unemployment (Lindberg and Maier (1985), Bruno and Sachs (1985)). Stagflation seems to have played the same contributing role as it did during the Great Depression in the 1930s or during the recurring banking crises in the late nineteenth century: that is, it signaled unequivocally both the degeneration of the existing monetary framework and the inadequacy of existing theories to confront the challenge. The impact stagflation made at the time is captured by the U.S. Council of Economic Advisers’ Report for 1974, which, after trying to make sense of the inflation upsurge in the context of mounting unemployment, somberly concluded that “there is no simple explanation for this price behavior, which was the most extraordinary in almost a generation and which confounded the Council and most other economists alike.” That something had gone wrong by the mid-1970s was unquestionable; what precisely it was, and how it was to be put to right, remained to be established. It should therefore come as no surprise that in the second half of the 1970s, the debate on the “monetary standard,” which had lain dormant, with rare and short-lived awakenings, for at least four decades, came back with renewed strength.
There was a novelty, though. Previous bouts of the debate had hinged on the relative merits (and drawbacks) of rule-based and discretion-based monetary frameworks.4 The outcome of the debate had appeared to be clear cut. If one believed that the effects of monetary policy, at least in the short run, were highly unpredictable (Milton Friedman’s famous “long and variable lags”), then activist (discretionary) policies were of little use. If one believed that policymakers had sufficient knowledge of a few key parameters of the economic system, then discretion, guided by all available information and backed by optimal control techniques, was preferable (B. Friedman (1975)). The latter view had appeared more attractive to most economists, partly because of the new econometric techniques developed during the 1950s and 1960s.
This view was shattered by a few influential papers showing that a policymaker unbounded by rules had an incentive to “cheat” the private sector in order to either keep unemployment below its natural rate or increase seigniorage by raising inflation. As argued by Kydland and Prescott (1977) and later shown more forcefully by Barro and Gordon (1983a), since rational agents discount the incentive of the policymaker to engineer surprise inflation, they will adjust their behavior accordingly. As a result, the economy will be subject to an inflation bias: inflation will be above target, but to no avail for the unemployment rate.
Kydland and Prescott (1977) and Barro and Gordon (1983a) saw their demonstration of the inflationary bias of discretionary policy as making the case for a monetary rule, along Friedman lines (Fischer (1995a)). Others reached even more radical conclusions, for instance, that the inflation bias could not be effectively used without taking control of the money supply out of the government’s hands. As a result, at the turn of the 1980s, a vast number of far-reaching proposals for monetary reform appeared in the literature. Hayek (1978) and White (1984), for example, advocated the repeal of the state monopoly in the supply of base money. Other scholars, sufficiently influential to win the establishment of a U.S. Senate Commission to investigate the matter, argued in favor of a return to the gold standard.5 Greenfield and Yeager (1993), Hall (1983), and Fama (1980) suggested instead possible schemes for separating the unit of account from the means of payment.6 Finally, M. Friedman (1984) elaborated on its early proposal for a constant money growth rule, turning it into a plea for “freezing high-powered money.”
A New Consensus?
A notable weakness of these reform plans was that they set about to cure the inflation bias of the fiat standard by suppressing the one feature that had made that standard attractive in the first place—the flexibility of response it allowed in the presence of supply and demand shocks. We have indeed seen that the degeneration of the fiat standard was not unrelated to the aim of increasing the capability of the monetary framework to cope with shocks. This aim, moreover, was not inconsistent with macroeconomic theory, which clearly indicated that a certain degree of flexibility might be desirable if the economy was exposed to exogenous uncertainty.
Not surprisingly, therefore, radical reform proposals were soon outcompeted by less ambitious ones, all hinging on the brand-new notion of credibility. Elaborating on Kydland and Prescott’s seminal 1977 paper, a number of authors argued that what was really missing was an institutional mechanism for convincing private agents that the flexibility inherent in the concept of “managed money” would not be exploited either to achieve short-run, that is, transitory, welfare gains, or to pursue the private interests, such as electoral gains or seigniorage revenue, of those in charge of monetary policy. In other words, one needed a credible framework that would make the private sector assign low probability to the event of being cheated through unexpected inflation.
In the course of the 1980s, the idea that the monetary framework that had emerged from the demise of Bretton Woods was lacking “credibility” spread so widely as to become commonplace. All standard accounts of the degeneration of the fiat standard came to rely on some form or other of Kydland and Prescott’s story (see, for example, Bruno and Sachs (1985), Nordhaus (1990), and Smith (1992)). Analogously, the need to strengthen monetary institutions so as to ensure the credibility of monetary policy became a leitmotiv in central bankers’ public speeches (Ciocca (1987), Volcker and Gyohten (1992)).7
The rise of the notion of credibility dealt a severe blow to all rule-flavored reform proposals, in that it helped to see that the announcement of rules or policy paths was in itself neither a necessary nor a sufficient condition for price stability. It was not sufficient, because announced rules were not necessarily credible. Nor was it necessary, because credibility of purpose could be achieved by directly altering the incentives faced by (or the utility function of) the policymaker. Until then, advocates of rule-based regimes had never entertained the possibility that a policy rule may be incentive incompatible, and therefore not believed by the private sector. Concern for the credibility of the restoration of the Gold Standard had been voiced in the early 1920s, but this was seen more as a contingent problem raised by the dismal state of European economies in the aftermath of the Great War than as a possible shortcoming of rule-based monetary frameworks. Be this as it may, the notion of credibility had never made its way into the theoretical literature (e.g., the notion of credibility is completely disregarded in the collection of papers on the standard edited by Yeager in the early 1960s).
While it is relatively easy to see why the notion of credibility is analytically promising, it is far less obvious how credibility can be achieved in practice. Indeed, credibility is a rather elusive notion, always in danger of turning into a truism, as has the notion of money’s “acceptability.” Just as one might be tempted to define a given form of money as acceptable if it is indeed ordinarily accepted as a quid pro quo, it would be equally tempting to define as credible those actions of the policymaker that are shown to have been believed by the private sector. Such an approach, however, would yield very little insight, if at all, on how to establish credibility in the first place.
As it happens, no general theory of credibility has ever been spelled out, despite the current popularity of the term. All one can safely say, given the state of our knowledge, is that credibility can be pursued in at least three different ways (Schelling (1982)). First, a hypothetical social planner could try to devise a commitment technology, that is, a mechanism that would give those in charge of monetary policy an incentive to stick to the announced policy course. Since governments find it understandably difficult to credibly constrain themselves, this will typically take the form of a delegation of power. Second, monetary authorities could try to establish a reputation for dependability, by arranging to have their determination publicly tested at an early stage after coming into office. Finally, one could try to influence directly the private sector’s expectations by increasing the visibility of the policy process, that is, by discussing, regularly and openly, the theory and the data underlying policy actions.
Clearly, these “three paths to credibility” are not mutually exclusive. For instance, even if the commitment technology were temptation proof, authorities would benefit from the private sector sharing the same theory of how the economy works, since, as noted by Schelling (1982), “translating even a confidently shared expectation of government action into a share expectation of results requires that a decisive subset of economic agents confidently shared a theory … relating program inputs to inflationary outputs.” Conversely, even if a large section of the population shared with the authorities a given theory, there would be no guarantee that the correct actions be taken if the authorities were not working under the appropriate incentives. Elster (1989) provides further insights on credibility buildings.
In spite of its theoretical indeterminacy, the notion of credibility lies at the heart of all the reform schemes implemented in the last decade or so. Indeed, as we shall see in Section III, one can trace behind all of them the attempt to achieve credibility of purpose through delegation, reputation building, increased transparency, or a blend of these mechanisms.