دلالتهای ضمنی تنوع در رفتار میزان مصرف برای انتخاب قواعد سیاست پولی در اروپا

Differences in economic structures across countries have potentially important implications for the conduct of monetary policy in the Euro Area. One facet of this lies in consumer expenditure behaviour. Our objective is to analyse the policy implications of assuming maximal and minimal differences between European economies using the National Institute's Global Econometric Model. We assess the performance of three possible ECB monetary policy rules under these different scenarios, using measures of the volatility of prices and output. We take as our benchmark a fully heterogeneous Europe, where individual country consumption functions are estimated separately. We estimate a homogeneous model for core European countries, incorporating countries into the core where it is statistically justified to pool them. We also estimate a fully homogeneous Europe where all Euro Zone countries are pooled. We find that the two ‘pillar strategy’ adopted by the ECB dominates other monetary policy frameworks.

Given the recent introduction of a single Euro Area monetary policy, it is important to ask whether the differences between European countries are so substantive that they influence the European Central Bank's implementation of the ‘one size fits all’ interest rate policy. A pessimistic view is given in Cecchetti (1999). One central difference amongst European countries lies in consumer behaviour, with some more sensitive to particular shocks than others. Differences may, for instance, manifest themselves due to asymmetries in the availability of credit facilities, with the various European countries at contrasting stages of development in their financial systems. The extent of credit restrictions, and their impact on consumer expenditure, is analysed in Bacchetta and Gerlach, 1997, Campbell and Mankiw, 1991a, Campbell and Mankiw, 1991b and MacLennan et al., 1998. It is of course the case that differences between European countries may depend upon the policy environment. Dornbusch et al. (1998) argue that asymmetries in the monetary transmission mechanism are mainly due to heterogeneous financial systems and these will diminish over time. Frankel and Rose (1998) suggest that business cycles are more correlated for countries that trade more, and Rose (2000) finds that members of a monetary union trade more with each other than similar countries who are not part of a monetary union. Hence entry into a monetary union may reduce differences in business cycles. However, Krugman (1991) and Hughes Hallett and Piscitelli (1999) suggest that the effect of European economic integration on institutions may not be great. Whether or not more convergence takes place it is clear that the structure of monetary policy responses is important. Clarida et al. (2000), using a stylised model of the macroeconomy, suggest that both output volatility and inflation variability is affected by the strength with which central banks respond to inflation and the output gap. There is evidence to support this proposition, and for instance Bernanke et al. (1997) demonstrate that the reaction of the US Fed to the oil price shock of the early 1970s was central to economic outcomes. In this paper we consider whether differences between European economies can be found in the data and whether they are important for the choice of a monetary policy rule. We test for statistically significant differences between the Euro Area countries and also the UK. We estimate a pooled model using seemingly unrelated regression estimation (SURE) with fixed effects and impose homogeneous coefficients across countries where it is statistically justifiable to do so. We find that this is only possible for a small group of core countries. We also estimate a completely homogeneous model which imposes cross country restrictions. We do not find that it is statistically valid to impose the constraint that all countries are members of a European core, but this is useful in our analysis. This model represents a fully integrated European economy, which, conceivably, will result from implementing a single currency and monetary policy. We compare these models to a completely heterogeneous model which is based on the existing equations from the National Institute's Global Econometric Model (NiGEM). We go on to consider whether European asymmetries matter for the conduct of monetary policy. Using our three models (an unrestricted model; a fully restricted model; and a statistically justifiable pooled model) we assess the performance of three possible monetary policy rules which the ECB could pursue. These rules are a nominal aggregate rule, an inflation targeting rule and a two pillar strategy, or combined rule (CR). Under the nominal aggregate rule, interest rates are raised and lowered with reference to the deviations of a nominal aggregate (e.g. money GDP or a monetary aggregate) from some target level. Likewise with an inflation targeting rule short term rates are changed in response to deviations of inflation from a target level. The combined nominal aggregate and inflation targeting rule is an approximation of present Euro Area monetary policy of relying on ‘two pillars’ in determining the appropriate level of interest rates.1 We examine the volatility of core real and nominal macroeconomic variables using these rules within the context of the NiGEM and by using stochastic simulations. These simulations are a means of replicating the uncertainty that economic systems face and hence of considering the performance of our monetary rules under multiple economic shocks. The present remit of European monetary authority to pursue price stability is particularly important in our final analysis. We find that a combined monetary policy rule is important for reducing price volatility whatever the structure of the Euro Area economies. A ‘two pillars’ approach to monetary policy, in our analysis, always reduces price volatility in the Euro Area. If we do not believe the process of European integration will have a substantial impact on real economic structures, a ‘two pillars’ CR dominates other possible monetary rules with respect to output, inflation and price variability. If European economies become more similar in the way suggested by our econometric analysis then it is possible that a nominal aggregate rule may be better if the objective function contains more than stabilising the price level. The UK would, in most circumstances, be better off moving to a monetary rule which contained more than an inflation target. The rest of this paper is set out as follows: Section 2 briefly reviews the theoretical and empirical literature on consumption; Section 3 considers the estimation methods used in our analysis; Section 4 examines our estimation results; Section 5 reviews monetary policy responses and NiGEM; Section 6 introduces the stochastic simulation used to assess the various policy rules and European consumer behaviour; 7 and 8 considers the results from our stochastic simulations; Section 9 concludes.

In this paper we compare European economic structures in terms of consumer behaviour and monetary policy rules. In particular, we consider whether economic asymmetries in Europe matter for the conduct of monetary policy. Three different approaches to estimation of consumption functions for a cross section of European countries are implemented on the macroeconometric model NiGEM: an unrestricted model, representing a heterogeneous Europe; a fully restricted model, representing a homogeneous Europe; and a statistically justifiable model, pooling coefficients where they are sufficiently similar. The unrestricted model is based on the existing equations of NiGEM. The pooled models are estimated using SURE methods, with individual country fixed effects. We find that we cannot statistically justify pooling all countries because differences matter. Pooling all countries is nevertheless useful as an experiment as it gives some information about outturns if monetary union produces similar economic structures for all European consumers. However, we can only find a statistically acceptable model if we pool using the core countries of Germany, the Netherlands, Austria and France. We utilise stochastic simulations with a macro-econometric model to represent the uncertain economic environment and the shocks Europe would face. We conclude that there appear to be no circumstances where the ECB would choose inflation targeting as its preferred approach. The ECB has the remit of maintaining price stability in the medium, and it has set out a strategy to achieve this. The ECB two pillar strategy can be described as following a combined monetary rule focussing on a nominal aggregate and on inflation. If the ECB is only concerned with the variability of the price level then the two pillar strategy dominates both inflation targeting and targeting a nominal magnitude on its own, whatever the structure of the world. This strategy also dominates others whatever variables we include in the welfare function if the world is as on our baseline model. It is possible that the ECB might choose to move to a nominal targeting strategy if Europe became more homogenous, but this move would depend upon the elements it includes in its welfare function and on the structure of the world. We also conclude that the UK should not have adopted an inflation target unless the only variable of concern to the Chancellor was the variability of inflation. The UK would indeed, in most circumstances, be better off moving to a rule with a nominal anchor in it.