1 Rules or Discretion in Monetary Policy: National and International Perspectives
- Carlo Cottarelli, and Tomás Baliño
- Published Date:
- December 1994
- GUITIÁN MANUEL
The long-standing debate on whether monetary policy is best conducted under a regime based on rules or, alternately, in a framework where discretion prevails has yet to be settled. Indeed, it is a debate that transcends the realms of monetary policy and of economic management in general. It concerns a basic principle of governance that relates to the method of organizing social and economic behavior: should it be based on a well-defined code of conduct, or should it depend on how those in authority judge and interpret events?
On the monetary front, at the national level, the discussion goes back at least as far as Henry Simons (1936), who strongly advocated the establishment of firm, clear rules to govern the conduct of domestic monetary policy. Since then, differences of opinion on the merits and demerits of rules or discretion in monetary management have continued to surface in a manner that can almost be described as cyclical; that is, over time, there have been periods when opinions have tended to coalesce around a preference for a rule-based monetary policy regime, followed by periods when preferences have clearly favored discretionary monetary policy management.1
On the international front, there was a classical period when a rule-based regime, the gold standard, prevailed. This standard, as is well known, came to an abrupt end with the outbreak of World War I, when countries geared their economic policies to meet the requirements of the military effort. As has been made abundantly clear since then, this development was but an illustration of how international rules frequently tend to be abandoned when the constraints they impose on national economic policies conflict with the pursuit of national goals. And the debate over the advantages or disadvantages of adopting either a rule-based or a discretionary regime in the international monetary context has, in the process, exhibited the same kind of cyclical behavior as the debate at the national level.
The central question in the debate, of course, is to establish what, if anything, in the monetary policy sphere should be decided by rule and what, if anything, should be left to the discretion of the policymaker. In this context, it is important to clarify at the outset a number of issues of terminology. Typically, from the standpoint of policy, two concepts are of fundamental importance: (ultimate) policy objectives and (proximate) policy instruments. Ultimate policy objectives, of course, are aims relevant to economic welfare and include growth, employment, price stability, and balance of payments viability. As far as monetary policy is concerned, a consensus has developed on what its primary objective should be: the attainment and maintenance of monetary stability, and in particular, of domestic price stability.2 Proximate policy instruments are variables controlled directly by the monetary authorities, such as changes in the portfolio of the central bank (through, e.g., open market operations or variations in reserve requirements) or adjustments in the interest rates at which the central bank conducts its operations (such as the discount rate). However, the linkages between these proximate policy instruments and the ultimate policy objectives are both complex and indirect. Thus, it has proven useful to focus on the additional concept of intermediate variables, such as money and credit stocks and their rates of growth, which are believed to have a firm, stable, and therefore predictable link with the ultimate policy objectives.3
These distinctions are relevant in the context of the rules-versus-discretion debate because they help to make clear the variables typically involved in the controversy. The discussion does not generally apply to proximate policy instruments because of their indirect relationship to final aims. Instead, the focus is on the intermediate variables, because of their postulated direct and firm link to the ultimate policy objectives. The central issue, then, is whether rules should be applied to determine the path of the intermediate variable in question, or whether discretion would provide a better policy guide.
The plan of the paper is as follows. First, it briefly discusses some basic considerations behind the preference for either rules or discretion, in both the national and international contexts. Second, it examines the choice between the two approaches and the evolution of the debate from a national perspective. Third, the examination is broadened to include the issues that arise in an international setting. Fourth, it is argued that the national and international aspects of monetary policy must be balanced. Fifth, a number of challenges facing monetary policy currently are discussed, with an attempt to ascertain whether they tend in the direction of rules or of discretion. And finally, some general conclusions are presented.
A common feature essential to both the rule-based and the discretionary approach to the conduct of monetary policy is a belief that an analytical relationship exists between monetary and credit developments and the evolution of nominal variables in the economy—in particular, of the national income or the domestic price level (and the exchange rate or the balance of payments). This belief is based on well-established conceptual and empirical analyses—both theoretical and in the context of specific country economies—of the demand for money over the last four decades.4 This said, however, it has always been generally acknowledged that policy is implemented in conditions of uncertainty, that it operates with variable lags, and that consequently its effects are influenced by expectations which are often hard to predict. In other words, even if the linkage between intermediate variables and ultimate policy objectives is conceptually well established, empirically it is imperfect and difficult to estimate with confidence, at least in the short run.5
In the national context, this imperfection and the potential slippages to which it may give rise between the short-term evolution of monetary and credit aggregates and of the domestic price level have led to arguments stressing the desirability of policy predictability. The underlying philosophy is that in a setting in which economies are subject to unpredictable shocks and the short-run effects of policies are difficult to ascertain ex ante, the best approach is to prevent monetary policy from adding to the uncertainties. The gist of the argument is that the predictability of policy should help offset unpredictability of the environment. This principle acknowledges the relevance of the unpredictability of the timing and effects of unconstrained policy and stresses the importance of time-consistency in formulating and implementing policy.6 The attractiveness of rule-based policy regimes is evident in this context. Not only would the policy rule be perceived as a parameter to which economic behavior would adapt (thus minimizing the short-run divergence between the rule and its objectives), but with persistent implementation, it would be seen as a prospective policy precommitment (thus making it unnecessary to forecast policy changes).7 In this sense, appropriate rules contribute to policy transparence as well as to policy credibility, both of which are generally seen as desirable features for economic management.
The proponents of discretion, in contrast, stress the importance of policy adaptability as a means of keeping an uncertain environment under control. In their view, it is precisely because the economy is subject to uncertain shocks, and because policies can have diverse effects (depending on the nature of those shocks) that discretion in policy implementation is desirable. The underlying philosophy here is that variations in monetary conditions reflect shifts in the demand and supply of money that are the result of a variety of factors. Policymakers can ascertain the nature of these shifts and identify the factors that cause them. Therefore, since the effects of monetary policy depend on the nature of the disturbance, the policy response is best determined discretionally.
Similar considerations apply in an international context. But in an international setting, the choice between rules or discretion can also be seen in terms of the importance given to national relative to international interests in the process of monetary policy implementation. In the same way that rules in the domestic context can anchor economic agents’ expectations about the direction of policy, rules in a foreign setting help anchor international expectations by buttressing the predictability of individual country policies. From this standpoint, it can be argued that rules take international interests into account precisely by posing a clear external constraint on national monetary policies.
The desire to contain uncertainty with policy predictability and the conflicting preference for tailoring monetary policy responses to the characteristics of a disturbance are typical of discussions of the choice between rules and discretion in the domestic area and have their equivalents in the international domain. At this level, while rules stress the relevance of external policy constraints in the pursuit of national policy objectives, discretion eschews those constraints. Thus, in the same fashion that discretion in a national setting emphasizes the room for policy maneuvering over the anchoring of expectations, while rules advocate precisely the opposite in an international context, discretionary regimes will highlight domestic relative to foreign interests, in clear contrast to rule-based systems.
As already noted, the rationale for monetary policy is largely predicated on the existence of a well-defined and stable demand for money, which is a function of a set of variables limited in number but critical in character. This demand for money plays an essential role in determining economy-wide nominal variables, such as nominal income and the price level.8 It also provides a key reference point for the design and conduct of monetary policy. In a national setting, the approach to monetary policy over the last half century evolved differently in terms of its perceived effectiveness and of its method of conduct. In terms of effectiveness, an active monetary policy was seen early in the period as neither effective nor desirable in pursuing real or nominal objectives, and therefore monetary policy focused on keeping interest rates low to support economic activity. The adverse impact of this approach on price level performance soon led to the abandonment of the principle of gearing monetary policy toward the maintenance of low interest rates.9 In terms of conduct, the conduct of monetary policy moved from discretion (until the early or mid-1970s) to rules (in the period from that time until the end of the 1980s). Thereafter, the pendulum has been swinging back toward discretionary monetary management.10
Analytically, the early discretionary period had its foundations in the advocacy of active demand management (e.g., fiscal and monetary) policy that followed from Keynes’s General Theory of Employment, Interest, and Money. Two aspects in the thinking underlying this view of macroeconomic policy are worth stressing: first, that monetary policy can and should be used for short-run cyclical purposes; and second, that the government should take responsibility for the stability of the economy at full employment. Fundamentally, the thinking in this period was based on a general confidence in the ability of macroeconomic policies to steer the economy and a corresponding belief in the importance of the government’s role in economic management. For policy purposes, the relevant empirical relationship stressed at the time was that between aggregate expenditure and income; little, if any, importance was given to the link between money and nominal income developments.11 As has often been pointed out, at this time the prevailing view was that growth and employment objectives conflicted with domestic price stability aims and that a sustained trade-off existed between these two goals. Therefore, one essential function of policy was to combine these conflicting objectives effectively, and this task was precisely the basis for the advocacy and pursuit of discretionary monetary policy.
One important reason for abandoning the active use of monetary policy as a countercyclical tool was the inability of macroeconomic policy, and in particular monetary policy, to solve the twin problems of inflation and unemployment, a failure that clearly undermined the rationale for discretionary demand management. This evidence helped to marshal a consensus that there was only a short-term tradeoff between growth and employment performance and inflation, and this consensus in turn contributed to the assignment of domestic price stability as the primary responsibility of monetary policy. As a consequence, the pendulum began to move away from discretion toward rules. The work of Milton Friedman and his associates on the stability of money demand and the close relationship it postulated between money and price level developments provided the analytical underpinnings of a new approach: monetary targeting or (as I prefer to call it) monetary rules.12 Rather than using monetary policy as a lever to influence aggregate expenditure, policy under this approach geared proximate monetary instruments to keeping monetary expansion within a predetermined path (a money growth rule). Apart from the close link between monetary and price-level behavior, this approach also stressed the importance of formulating the role of the monetary authorities in terms of the variables they could effectively control and for which they could consequently be held accountable.13
In the last few years, the emphasis on monetary rules has diminished and, in a number of countries, the practice of setting them as policy guides has been dropped altogether. To a large extent, these developments have reflected growing evidence of a looser connection between money and prices that has been traced to a variety of factors. The most important of these has been the process of liberalizing and deregulating of financial markets that has been undertaken in many economies over the last decade and that, it is argued, has affected the stability of the income velocity of money and weakened the firmness of monetary relationships. In the process, a trend toward discretionary management has once again emerged, but with less robust analytical underpinnings than those behind the Keynesian or the Friedmanian analyses. The weakening link between money and prices has not been replaced by another analytical relationship between proximate monetary policy instruments and intermediate variables with ultimate price performance or other nominal policy objectives. Instead, attention has been focused on setting ultimate policy aims. Currently, the conduct of monetary policy is something of a hybrid, combining discretion with regard to proximate and intermediate policy instruments and rules with regard to ultimate policy objectives. This approach therefore targets inflation—that is, it sets a price level or an inflation rule as the main guide for monetary policy implementation. The presumption is that since the rate of growth in the money stock no longer reliably predicts the prospective evolution of income or prices, the only available option for monetary policy is to focus attention primarily on the final aim: price stability. The relevant indicators typically include (but are not restricted to) variables related to the financial sector (e.g., bank credit and the term structure of interest rates).14 One important unresolved issue with this approach, however, is that it leaves undetermined the analytical linkages between policy instruments, indicator variables, and policy objectives. A practical problem can also arise under such an approach: how to use policy instruments when the indicator variables point in conflicting directions.
The oscillation between rules and discretion in monetary policy that has characterized the post-World War II period in national economies has been evident at the international level for over a century. In the international sphere, the traditional setting for monetary relationships has been provided by the gold standard.15 The gold standard was a rule-based regime under which imbalances among country economies were, in principle, redressed by endogenous and automatic gold flows from deficit to surplus economies. The policy rule under this standard was simple: to establish and maintain fixed values of national currencies in terms of gold. National economic policies inconsistent with this rule were not sustainable, because they led to external imbalances and consequent movements across countries of (exhaustible) gold stocks.
The rules of the game under this standard required that the gold outflows caused by excess demand and the gold inflows caused by surpluses of savings over investment be allowed to work their way through the respective economies. From the standpoint of monetary policy, this meant that gold outflows (which reduced the money stock) and gold inflows (which increased it) would be allowed to affect the economy at large—that is, they would not be sterilized. Actually, not only were they not to be sterilized, but policy actions could be undertaken to reinforce their effects in order to speed up the process of adjustment. For example, gold outflows could trigger policy responses, such as contractions in the central bank’s domestic assets or increases in the interest rate, that would increase pressure to reduce the money stock and would thus help stem gold outflows.16 Logic suggests that the opposite responses were called for in economies receiving gold inflows—that is, action to expand the central bank’s domestic assets or to lower interest rates. But clearly, such responses in surplus countries would essentially have been redundant and unnecessary if decisive action had been taken in the deficit countries.
As noted at the outset, the gold standard was abandoned with the outbreak of World War I and was never restored in its full-fledged version. Arrangements in the interwar period tended to favor discretion in national monetary policy; as a result, flexible exchange rate arrangements generally prevailed, albeit with different degrees of management or government intervention.17 The experience during this period contributed to a generally negative impression of flexible exchange rate arrangements, which were viewed more as a source of disturbance than as an element of support for economic adjustment.18
In fact, since the abandonment of the gold standard, the history of international monetary arrangements seems to have involved a continuous search for a code of conduct that combines an important measure of the discipline of a rule-based regime with a provision for a margin of discretion in national monetary policies. Early attempts to create such a code were made soon after World War I, when negotiations were undertaken at Genoa in 1922 that restated the traditional tenets of the gold standard (e.g., a unified world monetary system based on national currency parities fixed in terms of gold). But in addition, the notion was developed there that not all countries needed to hold their international reserves in gold; some could hold them in foreign exchange. A gold-exchange standard was thus envisaged. The monetary turbulence that characterized the 1920s prevented the adoption of these principles, but support for the gold standard continued, even though it was tempered by concern about the effect on domestic economic conditions. By the time a new international conference was convened in London in 1933, views varied about the relative importance that would be accorded to domestic economic recovery as opposed to external stability. Profound differences over fundamental economic policy priorities undermined the London discussions, and no global strategy or framework emerged from the conference. The world economy entered a de facto discretion-based period that left countries free to pursue national aims. Arguments (surprisingly similar to those that would be heard much later in the context of other attempts at orderly international arrangements) were voiced that the Genoa framework had proven “too rigid” to handle international monetary tensions. These were soon followed, however, by counterarguments (also remarkably modern) asserting that the alternative discretionary arrangements were “too flexible” to maintain an adequate measure of international order.19
The experience of the interwar period, with its shifts between the attempts to re-establish international order through rules and the need to face the reality of national priorities with discretion, paved the way for the adoption of a formal international arrangement that sought to balance these conflicting tendencies. This arrangement was embodied in the Bretton Woods Agreement of 1944, which laid out an institutional and legal framework for the conduct of international monetary affairs.20 The Agreement sought to strike a balance between the constraints required by the establishment of an international order and the scope for action countries would need to pursue national economic aims. In contrast to previous attempts, Bretton Woods represented a deliberate effort to guide ex ante the evolution of international monetary relations. The agreement was based on the re-establishment of fixed exchange rates (the rule), which in certain circumstances could be adjusted (the discretion).
Bretton Woods constituted the first formal code of conduct for international monetary and exchange relationships. The code of conduct was based on well-defined rules which, along with establishing currency par values, included the elimination of restrictions on payments for current international transactions and transfers. On this front, however, the Bretton Woods regime also sought to strike a balance by codifying the goal of current account convertibility while allowing an element of discretion in the use of capital controls.21 Within the Bretton Woods system, gold retained a relatively limited role, in that the par value of national currencies was to be expressed “in terms of gold as a common denominator or in terms of the United States dollar of the weight and fineness in effect on July 1, 1944.”22 In practice, this meant that the United States committed itself to maintaining the convertibility of the U.S. dollar into gold at a fixed price and that the rest of the countries undertook to keep their exchange rates fixed in terms of the U.S. dollar. Thus, the dollar-exchange standard came into being. Bretton Woods lasted for a quarter of a century and witnessed a widespread liberalization of trade and current account transactions, as well as the integration of many new countries into the international monetary system.
Over time, this rule-based regime became progressively vulnerable, as the accumulation of U.S. dollar assets abroad gradually outgrew the official gold holdings in the United States, and, more fundamentally, as economic policies in the major economies became less and less consistent with those required to maintain the rules of the game. And as usual, tensions led to another change in the rules, this time in the form of the progressive establishment of flexible exchange rate arrangements as the organizing principle for international monetary relations.23 A period thus began in the 1970s that was characterized by the prevalence of national economic policy discretion. Inconsistencies among national policies were to be redressed through appropriate fluctuations in exchange rates. These developments led to the current setting, which is characterized by a combination of formal and ad hoc arrangements and within which elements of both rules and discretion coexist. At present, there are two formal arrangements on the monetary front. One is global and discretion-based (i.e., the code of conduct laid out in the IMF Articles of Agreement), and the other is regional and rule-based (i.e., the European Monetary System (EMS)). Globally, the international monetary system is still operating on the basis of flexible exchange rate arrangements, which provide a degree of freedom in national economic policies, at least conceptually (hence, the system is described as discretion-based). Regionally, in Europe, a number of countries have bound their economies by a commitment to fixed (though adjustable) exchange rates, thus making their economic policies endogenous to the requirements of such a commitment (hence, this system’s categorization as rule-based).24
Since the mid-1970s, the major industrial countries have been involved in an ad hoc arrangement consisting of annual (and, on occasion, more frequent) summit meetings to discuss their economic and other policies. During their first decade, these meetings served as a vehicle to coordinate economic policy positions and economic performance assessments. While national policy discretion was initially the standard for these gatherings, in the mid-1980s a more rule-based perspective emerged, as evidenced by the Plaza Agreement (September 1985) and the Louvre Accord (February 1987), which, among other things, called for a measure of coordination in intervention in foreign exchange markets. In recent years, the approach has moved again toward discretion, although it is still constrained by the need for policy coordination.
Generally speaking, though, arrangements at present fall well short of those that would characterize a well-ordered system. And yet, during this period of national policy discretion, when a measure of domestic economic policy autonomy has been clawed back from the interdependent network that characterized the Bretton Woods par value regime, important decisions have been made by many countries to open up their economies and liberalize financial and capital markets. As a result, the last decade has witnessed the globalization of international capital markets, a phenomenon that in itself has imposed and will continue to impose strict limits on the margins for national policy discretion. Paradoxically, the very decisions that contributed to the establishment of global capital markets have in effect eliminated frontiers among national economies. And yet those frontiers are essential if national economic policy discretion is to be effective.
Actually, the prevalence of generally liberalized world capital markets raises a question about the appropriateness of continuing to accept capital controls in the international code of conduct. In this context, a strong argument can be made for updating the code to include the barring of restrictions on international capital flows—in other words, for bringing the code in line with reality and economic logic by focusing on full convertibility, rather than on current account convertibility.25
Balance or Conflict
The oscillation between rules and discretion in international monetary arrangements is to some extent inherent in the operation of these arrangements. Accepting common international rules that offer scope for national economic policy action, and, in particular, adopting a fixed exchange rate imply a willingness to share the spillover effects of any national policy actions. Consequently, participation in an integrated system carries with it potential benefits and costs that can be exploited and incurred, respectively, by individual countries. That is, the costs caused by domestic policies can be “exported” to the system at large, and the benefits derived from appropriate policies elsewhere can be “imported.” Unless the commitment to the rules is unshakable, the possibility of exporting costs and importing benefits can weaken the incentive to maintain appropriate domestic policies internally. Thus, interdependence can be abused, and if so, it will lead to dispersion as countries seek to avoid externally induced costs. Experience shows that commitments to rules are hardly unshakable and that tensions develop in rule-based regimes, typically leading to the replacement of rules with norms that stress discretion for national policies. Thus, national considerations become predominant, and with them efforts to increase the scope for discretionary action and the exercise of judgment. Uncertainty arises in the process, as does the potential for policy inconsistencies. Developments such as these bring with them incentives for policy coordination and the re-establishment of rules. Such oscillations, though, are only a reflection of the periodic conflicts that emerge between national and international aims, and therefore the search for balance in international regimes is likely to be a perennial endeavor.
In the monetary domain, the very aim of policy provides a clear illustration of these tensions. In its broadest formulation, the fundamental responsibility of monetary authorities is to maintain domestic financial stability in the economy. As already noted, the most common interpretation of this responsibility, at present, is to equate it narrowly with attaining and maintaining domestic price level stability. The emphasis on price level stability is based fundamentally on the empirical evidence mentioned earlier, which suggests that a close relationship exists between monetary conditions and the development of nominal variables in the economy, and in particular, between monetary expansion and the evolution of the price level. Thus, in most modern discussions of the role and functions of central banks, such price stability is seen as the primary—if not the single—objective of monetary policy. Consequently, in recent years many central bank charters have identified this objective as the main responsibility of the monetary authorities, an approach that to some extent reflects the prevalence in the international economy of flexible exchange rate arrangements. But is domestic price stability broad enough to be the monetary authorities’ central aim? A strong argument can be made that what the monetary authorities must ensure is the maintenance of what can be called monetary stability in the economy, one of the fundamental aspects of which is, admittedly, stability in the domestic price level. But as has often been noted, this objective is broader and encompasses aims such as relative external stability as well as sound financial conditions.26
This broader objective calls for stability in the value of the currency—that is, stability in the value of money. A stable level of domestic prices ensures only that the internal value of money is maintained. But the external value of money must also be preserved, and this task calls for exchange rate stability, which must, therefore, also be an essential aspect of the aim of monetary policy. From a fundamental standpoint, the twin aims of stability in the internal and external values of money should be complementary and mutually supportive. After all, the absence of inflation will most definitely be a contributing factor (all other things being equal) to a stable exchange rate. And a stable exchange rate will, in turn, help contain domestic price level variations.27
This broadening of the primary objective of monetary policy encompasses a host of age-old issues that revolve around divergent views of the relationship between exchange rate and domestic price stability. Maintaining both the internal and external values of money is often seen as two competing rather than complementary aims. However, the perception that these are competing objectives—which parallels similar perceptions about other economic goals, such as the difficulty of simultaneously achieving growth and controlling inflation, or the trade-off between adjustment and financing—is just as questionable as those other perceptions. Nevertheless, this focus on the possibility of a conflict between internal and external monetary stability does suggest a number of issues relevant to the conduct of monetary policy. For example, the external value of money will depend not only on domestic monetary policy, but also on external factors, specifically including other countries’ monetary policies. If the latter are unstable, a dilemma will develop as to whether the internal (price level) or the external (rate of exchange) value of money should be maintained. Considerations of this nature, and especially the dilemma they pose, are typically behind the preference for domestic price stability as the sole aim of monetary policy.28
There can be no doubt that conflicts of this nature have arisen and will continue to arise in the interaction between national economies. But assuming that all countries value the stability of their currencies, the conflict will not last, and therefore the essential question is whether it is appropriate for monetary policy to pursue one dimension of monetary stability at the expense of the other. If the external environment is reasonably stable, the frequency and extent of the conflict between the two aspects of stability will depend on the conduct of domestic monetary policy (abstracting from the possibility of changes in real domestic economic conditions, which raises other fundamental questions, such as the degree of flexibility in product and factor markets in the economy). In these circumstances, the central focus of monetary policy should be on maintaining stability of both the internal and external values of money, because price stability cannot be sustained with an unrealistic exchange rate and exchange rate stability cannot prevail for long with domestic price instability.
Here again, the challenge is to find an appropriate balance between rules and discretion in order to handle the temporary conflicts that may arise between the dual aspects of the aim of monetary policy. Some leeway in this respect can be temporarily provided by appropriate foreign borrowing or lending and by using or accumulating international reserves. More fundamentally, however, an essential prerequisite of monetary stability and, in particular, of price level stability is a substantial measure of both upward and downward flexibility in the prices of goods and factors of production. Price stability and the maintenance of international competitiveness in a dynamic economy will require domestic price and wage flexibility. Indeed, the price level can hardly be stable and competitiveness can hardly be maintained in circumstances where prices and wages can go up but not down. Flexibility in prices and wages is arguably among the most desirable features for many, if not all, economies, but unfortunately, it is also among the most difficult to attain and maintain. Actually, the difficulty of ensuring domestic cost-price flexibility looms large as a factor behind conflicts between maintaining the internal and external values of money. It is also often the basis for arguments favoring exchange rate flexibility or, equivalently, variations in the external value of money (on the grounds that maintaining a certain exchange rate will conflict with the protection of competitiveness).
Challenges to Monetary Policy
It is somewhat paradoxical that as international monetary arrangements have shifted toward discretion, many countries have implemented policies aimed at deregulating and liberalizing financial sectors, thus facilitating the integration of these sectors into the global capital market. This trend has posed and will continue to pose important challenges for the design and conduct of monetary policy.
On one level, the challenges focus on the appropriateness of domestic monetary targets as guides for national monetary policy, as well as on the relevance of such targets in a setting where national monetary borders have become progressively porous. With domestic financial deregulation, the boundaries between banking and other financial activities have blurred, compounding the difficulty of identifying a monetary variable with a behavior stable enough to anticipate the evolution of other nominal variables in the economy. There are two aspects to this problem. One is that deregulation may have affected the stability of the demand for money in its various traditional definitions and therefore may have impaired the usefulness of money demand forecasts as a basis for monetary policy implementation. Some of the arguments made in this context focus on the impact of eliminating financial sector restrictions on the demand for money. In itself, this factor should be a onetime event that ends once deregulation is complete. Another aspect of the issue is the increasing degree of substitutability in private portfolios between banking liabilities and those of other financial intermediaries. Rather than affecting the demand for money (however defined), this argument centers on the growing complexity of the definition of money, which possibly can no longer be confined to central bank or banking system liabilities.
These considerations pertain to the national policy domain. But with the internationalization and globalization of financial markets, another challenge has arisen for monetary policy implementation. While the first challenge underscores the blurring of the boundaries between banks and other financial intermediaries, the second stresses the growing disappearance of national economic frontiers. The issue at stake here is the relevance and appropriateness of national monetary variables in an environment where currencies are increasingly substitutable. This phenomenon, which acquires particular strength in the context of regimes such as the European Monetary System (EMS), is valid on general grounds, as the experience with dollarization in a number of economies aptly illustrates.
Rather than posing a conceptual question, these challenges are of an empirical nature. They relate to the identification of the national monetary variable with the most stable behavior (in relative terms). And they concern the question of the relative stability and predictability of national and international monetary aggregates. These challenges add to the complexities of actual monetary policy implementation. The blurring of boundaries between banks and other financial intermediaries calls for definitions of monetary variables that correspond to the financial sector at large. As for the obsolescence of national monetary aggregates, the important implication to draw is the constraints it imposes on domestic monetary policy. It essentially allows for little, if any, margin for error or inconsistency in the conduct of policy, as any such error or inconsistency would soon be corrected by international financial flows. Specifically, the pursuit of unduly restrictive or expansionary monetary policies no longer yields the consequences that could be expected in a more segmented international economic setting. Rather, such policies only stimulate net financial inflows or outflows, thus confirming that market forces are the main determinants of the stance of monetary conditions.
The challenges that worldwide capital markets pose for national monetary policies and international monetary arrangements apply whether rules or discretion prevail. The fundamental implication is that a powerful constraint has been placed on the scope for monetary policy implementation—that is, on the room for discretionary maneuvering—as well as on the operation of policy rules. This constraint is, of course, market forces, which subject monetary management to market discipline either under rules or under discretion.29
In turn, the blurring of identities of banks and other financial intermediaries reopens the question of whether earlier ideas concerning fundamental distinctions in financial activities and institutions are worth revisiting, if not reconsidering. These ideas include those behind the Chicago Plan for banking reform and the notion of “narrow” banking, which essentially advocated the separation of banks’ deposit-taking functions from their lending activities.30 Internationally, and with regard to economies now in the process of establishing market-based systems, a similar line of reasoning at the central banking level is behind the proposals for currency boards.31 These proposals focus clearly on one definite concept of money and may become useful, from the point of view of policy, to the circumstances of reforming economies.
Two broad sets of conclusions flow from the analysis in the paper. In general, monetary policy choices are not between rules or discretion; the choice is one of more or less relative reliance on one or the other policy principle. The issue at stake is one of balance, and this balance will likely differ across countries as well as over time.
At the national level, the emphasis on rules, or monetary targeting, that prevailed from the 1970s well into the 1980s was indicative of a policy balance biased toward enhancing the credibility and predictability of monetary policy. Nevertheless, discretion was generally exercised. The current lack of emphasis on policy rules reflects a shift toward discretion brought about by the weakening of the link between intermediate variables and their objectives. But rules have not been discarded altogether; instead, they themselves have been formulated in terms of the ultimate policy objectives themselves (price level or inflation targets). The advantage of formulating the rules in terms of policy instruments (even if these are intermediate in nature) is that such guidelines provide a transparent signal of the conduct and direction of policy. In contrast, rules formulated in terms of ultimate policy objectives or policy outcomes provide little or no information on how policy is or will be actually conducted, even if they clearly indicate the aim of policy. But in effect, the rules approach also passes on to economic agents definite information on policy intentions. Since the setting of intermediate variables is dependent on the price objective, the policy rule approach gives signals about both proximate and intermediate policy instruments, as well as about ultimate policy aims. The alternative approach, discretion, either provides less information or limits its signals to the final policy objective.
At the international level, regimes like the gold standard or the Bretton Woods par value system tended to be balanced toward rules. But again, they exercised discretion as well, through gold price variations under the gold standard (which could occur between the so-called gold-points) and through exchange rate adjustments, capital controls, or temporary current account restrictions (all of which were available options) under the Bretton Woods system. Thus, although national monetary policies were bound by the international rule, some scope for discretion remained. A fundamental aspect of the Bretton Woods order was its self-enforcing nature: the constraints it imposed were administered by the countries themselves. The system was also international, in that the exercise of discretion was administered internationally.32 In contrast, regimes like the prevailing one are clearly balanced toward national discretion. Indeed, one of the main attractions of such regimes has been precisely that they were thought to allow countries to pursue their domestic objectives independently, according to their own lights. International discretion, however, has remained bound by the national need to keep domestic policies in line with policy objectives. Unlike rule-based regimes, this type of system is neither self-enforcing nor international in nature. On the contrary, in such a setting discretion is administered nationally, and the rest of the world is left to administer external constraints.
The second set of conclusions that can be drawn from the analysis follow from the deregulation and liberalization of capital and financial markets. In essence, the predominance that such evolution has given to market forces will be useful if it serves to clarify the narrow limits within which monetary policy and, more broadly, economic policy as a whole must operate. This essentially means that, to be effective, policies, whether based on rules or on discretion, should not run counter to fundamental market forces. But neither should they accommodate market developments when these are not in accordance with fundamental forces. Yet market developments that do not reflect fundamental forces are often allowed to influence policy, and to that extent, the market is not subject to its own discipline. One significant consequence of policy adaptations in these circumstances is the moral hazard they create; markets internalize the freedom they are granted from their own discipline, and as a result governments assume the risks and costs of developments that do not reflect fundamental trends. In this context, rule-based regimes may have an advantage over discretionary policy management if their design and implementation are consistent with the existence of global markets. To the extent that they restrict policy adaptations, they will also contain the tendency to yield to market pressures and thus will help contain moral hazard risks. On the other hand, discretionary systems, which underscore the importance of flexible policy implementation and therefore are less transparent, are likely to find it difficult to resist market pressures and to incur the consequent risks.
In sum, the issue in the selection of monetary regime prototypes is not so much the stark choice between rules or discretion but the need to strike an appropriate balance in the reliance on one or the other of these two basic principles. As the saying goes, there is an exception to every rule. The essence of rule-based regimes is to ensure that departures from the rule remain the exception—that is, minimal. Discretion-based regimes, in turn, require only a modicum of rules in the form of a set of guiding principles relative to which the exercise of discretion can be measured.
“Money Versus Credit: The Role of Banks in the Monetary Transmission Process,” inFrameworks for Monetary Stability (Washington: International Monetary Fund, 1994).
“Regulatory Power Helps Fed Fulfill Its Monetary Role,”American Banker (January 6, 1994).
“Currency Boards: Issues and Experiences,” inFrameworks for Monetary Stability (Washington: International Monetary Fund, 1994).
Monetary Policy Under the Gold Standard: 1880–1914 (New York: Federal Reserve Bank of New York, 1959).
A Retrospective on the Bretton Woods System (Chicago: University of Chicago Press, 1993).(eds.),
The Reconstruction of the International Monetary System: The Attempts of 1922 and 1933, Princeton Studies in International Finance, No. 33 (Princeton, New Jersey: Princeton University, 1973).
The International Monetary System: Essays in World Economics (Cambridge, Massachusetts: MIT Press, 1987).
“Should an ‘Independent’ Central Bank Control Exchange Rate Policy?,” in Frameworks for Monetary Stability (Washington: International Monetary Fund, 1994).
“Rules Versus Discretion in Monetary Policy,”unpublished paper presented at an International Conference on A Framework for Monetary Stability (Amsterdam: De Nederlandsche Bank and Tilburg University, October21–23, 1993).
The International Monetary Fund, 1966–1971: The System Under Stress (Washington: International Monetary Fund, 1976).
“International Policy Coordination in Historical Perspective: A View from the Interwar Years,” in International Economic Policy Coordination,ed. by WilliamH. Buiter and RichardC. Marston (ed) (Cambridge: Center for Economic Policy Research and Cambridge University Press, 1985).
Elusive Stability: Essays in the History of International Finance, 1919–1939 (Cambridge: Cambridge University Press, 1989).
“International Monetary Instability Between the Wars: Structural Flaws or Misguided Policies?,” in The Evolution of the International System: How Can Efficiency and Stability Be Attained?,ed. by YoshioSuzuki, JunichiMiyake, and MitsuakeOkabe (ed) (Tokyo: University of Tokyo Press, 1990).
Golden Fetters: The Gold Standard and the Great Depression (New York: Oxford University Press, 1992).
“Optimal Monetary Policy Design: Rules vs. Discretion Again” in Intermediate Targets and Indicators of Monetary Policy: A Critical Survey (New York: Federal Reserve Bank of New York, 1990).
“Formal Targets for Inflation Reduction: The Canadian Experience,”unpublished paper presented at an International Conference on A Framework for Monetary Stability (Amsterdam: De Nederlandsche Bank and Tilburg University, October21–23, 1993).
“Intermediate Targets Versus Information Variables as Operating Guides for Monetary Policy,”unpublished paper presented at an International Conference on A Framework for Monetary Stability (Amsterdam: De Nederlandsche Bank and Tilburg University, October21–23, 1993).
“A Monetary and Fiscal Framework for Economic Stability,”American Economic Review,Vol. 38 (June1948), pp. 245–64.
Essays in Positive Economics (Chicago: University of Chicago Press, 1953).
Studies in the Quantity Theory of Money (Chicago: University of Chicago Press, 1956).(ed.),
A Program for Monetary Stability (New York: Fordham University Press, 1959).
The Optimum Quantity of Money and Other Essays (Chicago: Aldine Publishing Company, 1969).
A Monetary History of the United States: 1867–1960, National Bureau of Economic Research (Princeton, New Jersey: Princeton University Press, 1963).
The Political Economy of International Relations (Princeton, New Jersey: Princeton University Press, 1987).
“Bretton Woods and Its Precursors: Rules Versus Discretion in the History of International Monetary Regimes,” in A Retrospective on the Bretton Woods System,ed. by MichaelD. Bordo and BarryJ. Eichengreen (ed) (Chicago: University of Chicago Press, 1993).
“The ERM Experience” inFrameworks for Monetary Stability (Washington: International Monetary Fund, 1994).
“Credit Versus Money as an Instrument of Control,”Staff Papers,International Monetary Fund, Vol. 20 (November1973), pp. 785–800.
1992a), Rules and Discretion in International Economic Policy, Occasional Paper, No. 97 (Washington: International Monetary Fund).(
1992b), The Unique Nature of the Responsibilities of the International Monetary Fund, IMF Pamphlet Series, No. 46 (Washington: International Monetary Fund).(
1993a), “The Role of Monetary Policy in IMF Programs,”unpublished papers presented at an International Conference on A Framework for Monetary Stability (Amsterdam: De Nederlandsche Bank and Tilburg University, October 21–23).(
1993b), “Towards a Code of Conduct for International Finance,”unpublished paper presented at the City of London Second Central Banking Conference (London, December 7–8).(
1993c), “From the Plan to the Market: Banking and Financial Reform Aspects,” in The New Europe: Evolving Economic and Financial Systems in East and West,ed. by D.E.Fair and R.Raymond (ed) (The Netherlands: Kluwer Academic Publishers).(
1993d), “The European Monetary System: Recent Intellectual History—A Commentary,” in The Monetary Future of Europe (London: Center for Economic Policy Research).(
1993e), “The Issue of Capital Account Convertibility: A Gap Between Norms and Reality,”unpublished paper presented at a Seminar on Currency Convertibility (Marrakech, Morocco, December16–18).(
1993f), “Currency Convertibility: Concepts and Degrees,”unpublished paper presented at a Seminar on Currency Convertibility (Marrakech, Morocco, December 16–18).(
“The Role of the Central Bank in Stabilization and Adjustment,”a paper forthcoming in the proceedings of a seminar on Inflation, Stabilization and Growth in Honor of Felipe Pazos (Washington: World Bank, 1994).
Currency Boards for Eastern Europe (Washington: Heritage Foundation, 1991).
The International Monetary Fund, 1945–65: Twenty Years of International Monetary Cooperation (Washington: International Monetary Fund, 1969).
International Monetary Fund, Articles of Agreement (Washington: International Monetary Fund, 1944).
“Rules Rather than Discretion: The Inconsistency of Optimal Plans,”Journal of Political Economy,Vol. 85 (June1977), pp. 473–91.
Rules Versus Discretion: An Essay on Monetary Policy in an Inflationary Environment, BIS Economic Papers, No. 3 (Basle: Bank for International Settlements, April1981).
League of Nations, with Ragnar Nurkse as primary contributor, International Currency Experience: Lessons of the Interwar Period (Geneva: League of Nations, 1944).
What Should Banks Do? (Washington: The Brookings Institution, 1987).
International Payments and Monetary Policy in the World Today, Wicksell lectures 1961 (Stockholm: Almqvist and Wiksell, 1961).
“Independent Currency Authorities: An Analytic Primer,”Staff Papers,International Monetary Fund, Vol. 40 (March1993), pp. 202–16.
“Convertibility: An Indispensable Element in the Transition Process in Eastern Europe,” in Currency Convertibility in Eastern Europe,ed. by JohnWilliamson (ed) (Washington: Institute for International Economics, 1991).
“Rules Versus Authorities in Monetary Policy,”Journal of Political Economy,Vol. 44 (February1936), pp. 1–30; reprinted in his Economic Policy for a Free Society (Chicago: University of Chicago Press, 1948).
“Financial Innovation and Deregulation in Perspective,”Bank of Japan Monetary and Economic Studies,Vol. 3 (September1985), pp. 19–29.
I have analyzed the issues connected with the periodic moves from rules to discretion in a general economic policy context in Guitián (1992a). For other recent discussions of the subject in a national and an international context, see Crockett (1993) and Giovannini (1993), respectively.
There are a variety of levels at which the fundamental objective of monetary policy can be formulated: financial stability, which encompasses the establishment and maintenance of sound conditions in the financial sector; monetary stability which focuses on the soundness of conditions in the banking sector; and domestic price stability, which focuses on the maintenance and safeguard of the internal value of money; see, for further elaboration. Manuel Guitián (1994). There is, in addition, exchange rate stability, which concerns the protection and maintenance of the external value of money: see Cottarelli (1994).
These variables are often referred to as intermediate targets. I am not persuaded that “target” is the right term to apply to them, if only because the term conveys the notion of “objective.” It is true that, in a sense, intermediate variables have characteristics common to policy objectives themselves (after all, they are the variables to which proximate policy instruments are aimed); but, more importantly, they also conform closely to the concept of policy instruments (indeed, their relevance is derived from the role they play in the attainment of ultimate policy objectives). Emphasis on this latter aspect of intermediate variables would lead to classifying them as policy instruments; see Guitián (1973 and 1993a): see also Alexander and Caramazza (1994).
Much of the impulse behind these analyses came from research conducted at the University of Chicago: see, in particular, Friedman (1956). But over time, this impulse gained general acceptance and the basic principles derived tram these analyses became influential factors in policy implementation. At present, though, serious challenges to these principles have emerged: see, in particular, the discussion on challenges to monetary policy below.
There is an ample body of technical literature examining the rules-versus-discretion debate from the standpoint of the design of an optimal monetary policy discussing these imperfections. See, for example, Englander (1990) and the references listed therein.
In the extreme, the predictability of policy rules will depend on how rigidly they are followed. In Simons’ words. “…we obviously need highly definite and stable rules of the game, especially as to money. . . . Once established, however, they should work mechanically, with the chips falling where they may” (1948, p. 169).
The view of monetary policy as an ineffective economic tool reflected early post-World War II concerns about the prospect of secular stagnation. The shift in policy from interest rate pegging to an active stance is well illustrated in the U.S. experience of the early 1950s, and in particular, by the Treasury-Federal Reserve Accord, under the terms of which the latter abandoned its policy of fixing the price of government securities (that is, interest rates). For elaboration, see Lutz (1961) and Friedman and Schwartz (1963).
Until the last stage, the evolution was paradoxical in that in the early period, national monetary policy discretion was prevalent, and yet the international system was rule-based (the Bretton Woods par value regime). In contrast, when national monetary policies moved toward rules, the move was accompanied by a shift to a discretionary system (flexible exchange arrangements) in the international economy. The present predominance of discretion, though, prevails at the national and international levels. For further discussion, see Lamfalussy (1981). Guitián (1992a), and Crockett (1993).
In a different context and for a different purpose, Polak (1991) discussed the attitudes that prevailed at the time with regard to the role of government. For a succinct description of the concepts behind Keynesian demand-management policy, see Lamfalussy (1981). As pointed out in note 4 above, studies suggesting a stable relationship between money and income were being conducted particularly in the University of Chicago, Friedman (1956).
It is to be noted that Milton Friedman, a strong advocate of rules in domestic monetary policy, is also an advocate of discretion in the international monetary sphere. His is a consistent position, of course, because the pursuit of a domestic monetary rule requires discretion at the international level. See Friedman (1959) and his “The Case for Flexible Exchange Rates” in Friedman (1953); see also Guitián (1993d).
See, for further detail, Crockett (1993) and Freedman (1993). Benjamin Friedman (1993) also discusses in detail the use of indicators, or, in his terminology, information variables, as guides to monetary policy implementation.
There are numerous examinations in the literature, but a classic source on the gold standard remains Bloomfield (1959). But see also, for more recent treatments, Cooper (1987), Eichengreen (1985, 1992), and Gilpin (1987).
The automatic nature of the operation of the gold standard is less evident in practice than in theory. But its basically rules-based nature cannot be contested; see Guitián (1992a).
For elaboration, see the influential study of Nurkse for the League of Nations (1944).
For a detailed analysis of this period, see Clarke (1973).
Article IV of the original Articles of Agreement of the International Monetary Fund; see International Monetary Fund (1944).
This revision, formalized in the second Amendment of the Articles of Agreement of the International Monetary Fund, remains in effect.
On August 2, 1993, a decision was made to increase the margins of fluctuation of EMS currencies around their central parities from 2.25 percent to 15 percent. The decision has changed the original balance between rules and discretion in favor of the latter to such an extent that it could be argued that the regime has become virtually discretionary; although this may be an accurate description in principle, however, it is not in practice, at least so far. For a discussion of the experience of the exchange rate mechanism of the EMS, see Griffiths and McDonald (1994).
See, for further discussion, Volcker, Mancera, and Godeaux (1991). Monetary stability is a somewhat narrower concept than domestic financial stability, in that the former relates to the banking sector and the latter encompasses the financial sector at large. But the difference is diminishing as the boundaries between banks and other financial intermediaries become blurred.
For an extensive discussion of these issues, see Cottarelli (1994).
A classical examination of these questions will be found in Keynes (1924); see, in particular, his discussion of alternative aims in monetary policy, which remains valid in today’s circumstances.
A related topic of importance in the context of the progressive predominance of world market forces is the subject of bank supervision, the relevance of which, at the national and the international level, has heightened in the context of financial deregulation and liberalization. Global financial market forces have important implications for the monetary policy aim of maintaining sound financial conditions and for the exercise of the function of lender-of-last-resort. These subjects are beyond the scope of this paper, but for a brief discussion, see Axilrod (1994).
See, for a more complete discussion, Guitian (1992b).