How should monetary policy makers respond to the new challenges of global economic integration?

Speech delivered by Professor Eugenio Domingo Solans, Member of the Executive Board of the European Central Bank, contribution presented at the Symposium on "Global Economic Integration: Opportunities and Challenges", sponsored by the Federal Reserve Bank of Kansas City, Jackson Hole, Wyoming, 26 August 2000.

In this contribution I intend to elaborate on some of the new conditions and challenges stemming from the global economic integration and the appropriate monetary policy responses.

I shall also refer to some inappropriate responses: I understand that in monetary policy - as in any other kind of policy, economic or not - avoiding inappropriate solutions is at least as important as having brilliant ideas.

Global economic integration has already resulted in two somewhat conflicting features: more economic convergence and more uncertainty.

More economic convergence implies more interaction. Interaction means that opportunities and risks are interdependent, that we know better than in the past "that a country's potential gain from the choice of monetary arrangements depends on the choices that other countries make" (Metzler, 1997:3). Interaction can lead, although not necessarily, to monetary co-operation, co-ordination and even monetary union as an extreme form of close institutional arrangements, which was the path that most countries of the European Union decided to follow when they joined the euro area.

One monetary policy response to interaction, or an outcome resulting from it, is standardisation. Standardisation implies the adoption of similar monetary policy parameters. Never before have central banks been more closely aligned in terms of their principles, objectives, strategies and instruments of monetary policy, irrespective whether these principles, objectives and strategies have been made explicit or not.

Standardisation, in particular, has meant that the well-known polarisation between rules and discretion in a central bank's practices has faded away. This trend has been reflected in the theory of monetary policy, where nowadays more flexible definitions of rules are accepted involving notions of variable settings of instruments, contingent formulas and similar terms. These more flexible definitions of a rule depart from its original definition, which involved the ideas of automatism or mechanicism ^[1], concepts which do not connote the ideas of contingency or variability. By contrast, we can agree that, as suggested by Taylor, "rule-like behaviour is systematic in the sense of methodical, according to a plan, and not casual or at random" (Taylor, 1993), but this is also the way in which non-rule-based central banks conduct monetary policy, including the European Central Bank (ECB) with its two-pillar monetary policy strategy.

The debate between rules and discretion is nowadays more an academic than a practical one. Given the now blurred differences between the two poles, the discussion probably has more to do with terminology than with content.

To choose an externally based monetary policy approach could be seen as an extreme form of standardisation, i.e. an extreme monetary policy response to convergence and interaction. This choice makes sense above all when the second feature mentioned as resulting from global integration is brought into he picture: uncertainty.

By an externally based approach I mean any institutional arrangement based on an exchange-rate rule, such as a managed float with a limited fluctuation band (this was the approach adopted by the European Union countries that joined the Exchange Rate Mechanism (ERM) of the European Monetary System and is the current arrangement among the European countries participating in ERM II), pegging the exchange rate, establishing a currency board or even creating a monetary union with a common monetary policy, as the euro area countries have done. The latter is certainly the most extreme case of a monetary policy institutional arrangement based on an exchange rate rule within countries participating in the monetary union.

Adopting an exchange rate rule could be considered as an extreme case of standardisation because although the strategy of the pegging central bank differs from that of the central bank chosen as a reference, it imports and eventually develops similar monetary and inflationary results. The means differ but the results are similar.

An externally based monetary policy approach may be, although not necessarily, a good choice for small open economies with a high degree of interaction and convergence with other economies, provided that certain conditions are fulfilled and some costs are accepted. In this case, giving up flexibility and committing to an exchange rate rule could be an appropriate way to obtain credibility.

By contrast with small open economies, however, it is unlikely that adopting an external monetary policy rule (i.e. giving up flexibility) would be the best choice for large and relatively closed economies. Given a lesser degree of integration with other economies and, therefore, less convergence, committing to an exchange rate rule could jeopardise the achievement of internal monetary policy objectives. In the case of large and rather closed economies committing to an external rule could result in lack of credibility. This is why the euro area as a whole, the second largest economy in the world, has chosen an internally based approach and thus a flexible exchange rate.

The degree of integration and convergence with other economies is, therefore, a key factor in deciding whether to adopt an externally based monetary policy approach.

As a consequence of greater uncertainty - the second feature resulting from global economic integration - monetary policy formulation and implementation is more challenging, complex and demanding. The challenges relate to several issues. Let me elaborate on three of them: information, monetary policy instruments and the choice of the best monetary approach.

As regards information, the more uncertain scenario resulting from global economic integration, which involves novelty and sophistication, implies the existence of imperfect observability and increasing difficulties in obtaining sufficient, accurate and timely data. Global economic integration also leads to the existence of new and more complex relationships between variables owing to changes in innovation and productivity, among other factors, which call for continuous revisions of economic models. Global economic integration allows new macroeconomic conditions to be transmitted rapidly and extensively through new channels.

There are several kinds of uncertainty and, therefore, different types of information are also needed. As Issing (1999:23) points out, two types are especially relevant to the field of monetary policy: information on the current state of the economy - the data - and information on how the monetary policy instruments affect inflation and economic activity, in terms of both size and timing - the monetary transmission mechanism.

Concerning the need for additional information related to the data, a central bank will have to produce more complete, accurate, frequent and timely statistics. In addition, it might consider changing the weights given to the different sources of information. For example, by virtue of convergence financial markets have become deeper, more flexible, more standardised and more liquid - in two words, less imperfect - and, therefore, the information coming from them has become more relevant to monetary policy decisions, although it has to be treated with special caution because of the risks of circularity and asymmetric perception (Issing, 1998: 20-21). Changing the weight given to information from the markets would not, by any means imply, "following the markets". In the monetary policy game, the central bank must play the role of leader and, as Blinder (1998: 59-62) indicates, be also independent of the financial markets.

Concerning the need for additional information related to the monetary policy transmission mechanism, the appropriate answer for central banks is to enhance research in order to develop more accurate models. It goes without saying that the contribution of the academic world is of the utmost importance.

In conclusion, the need for additional information concerning the data and the monetary transmission mechanism calls for the enhancement and, where necessary, the upgrading of the statistical and research functions within the internal organisation of the central bank.

Financial innovation and the change in market conditions, which relate to global integration, can reduce the central bank's control of the monetary base. This could be a good argument in favour of using interest rate instruments in order to implement monetary policy. In practice, although not in theory (McCallum, 1999: 1505-15), central banks are inclined to use an interest rate instrument. Again the debate is more academic than practical.

In the field of monetary policy implementation, inappropriate solutions would involve an "excess" of regulations. "Excess" would mean, in this context, that the possible enhancement of the effectiveness of monetary policy owing to the regulation would be outweighed by the distortions created. An example of this kind of inappropriate monetary policy response would be setting reserve requirements at differential levels for different types of assets in order to influence credit allocation (Schaberg, 1999: 138). It goes without saying that, although both affect banks, this kind of monetary policy regulations has nothing to do with prudential supervision regulations.

Global integration, and the uncertainty resulting from it, call for a monetary policy response which must, above all, be credible, i.e. realistic and effective.

Monetary policy design is confronted with many alternatives: activism versus no activism, aggressiveness versus smoothness, automatism or mechanicism versus judgement, rigidity versus flexibility, precommitment versus absence of precommitment, time consistency versus time inconsistency, transparency and accountability versus opacity, simplicity versus complexity, etc.

In an atmosphere of global integration and greater uncertainty, the best choice among these alternatives, i.e. the most appropriate monetary policy answer, would be no-activism, smoothness, judgement, flexibility, precommitment, time consistency, transparency and accountability. Such a choice, unavoidably, implies complexity.

In general terms, although the issue remains open, I am inclined to think that activism is not the appropriate monetary policy answer in an atmosphere of uncertainty. Activism connotes fine-tuning and short-term perspective, which I understand is not the approach monetary policy should take. Besides, a central bank which takes into account the fact that its actions affect learning may choose to be less active than a central bank which ignores learning effects (Ellison and Valla, 1999).

Activism versus no activism and aggressiveness versus smoothness, although related, are different alternatives. The first relates to "how often" while the second has to do with "how much". Concerning aggressiveness versus smoothness, although there are good arguments for choosing either option (Brainard, 1967, Goodhart, 1999 and Söderström, 2000), most central banks favour smoothness, especially in a scenario of uncertainty, for reasons of caution, consistency and credibility - the "three letters "C" argument", as we could call it. Gradualism is not activism: it simply means to divide a move, aggressive or not, in more than one step.

Judgement and flexibility can be better achieved through discretion A pure monetary rule, which implies automatism, rigidity and simplicity, would not work in an uncertain environment and would, therefore, be an inappropriate monetary policy response. I wonder who among you would prefer to rely on a simple, rigid, mechanical autopilot rather than a judicious, experienced human on a plane which happened to be flying in "uncertain" conditions. If the degree of novelty and uncertainty is very high, even contingency rules might not provide an appropriate answer.

The basic reasoning underlying the preference for discarding automatism, rigidity and simplicity in an uncertain scenario is that in order to offset the unpredictability of the environment, policy-adaptability is better than policy-predictability, to put it in terms that Guitián (1994: 22) would have used.

The best monetary policy choice also implies precommitment, time consistency, transparency and accountability. These conditions require discipline, which is closer to a rule than to discretion.

The way out of this awkward spot should be a formula able to combine the judgement and flexibility of discretion with the discipline that is achieved in the case of rule-based policy design through the automatic feedback between the target and the instrument variables. This formula, based on "bounded discretion", would imply substituting the automatic feedback of the rule with "reputational forces" and institutional constraints, in line with Barro and Gordon's approach (1983a; 1983b). Bounded discretion is something different from pure discretion or contingency rules and is also far from arbitrariness.

The main elements of this "boundness" would be:

In this contribution I have proposed a broad outline of a monetary policy approach based on "bounded discretion", able to respond to the new challenges of global economic integration, which corresponds - did you guess? - to the Eurosystem's monetary policy framework.

The appropriate response, i.e. the most credible, realistic and efficient one, should involve a lack of activism, combined with smoothness, judgement, flexibility, precommitment, time consistency, transparency and accountability. These requirements are better fulfilled by a discretionary policy design supported by central bank independence, pre-established prioritised objectives, a clearly specified encompassing strategy, enhanced communication and democratic control.