The enlargement of the euro area: differences in relative inflation

This paper investigates the structural determinants of relative inflation (i.e. the inflation of non-tradables vs tradables) in the context of overall inflation differentials in the EU. The analysis is based on the Bergstrand theoretical model. This framework incorporates three alternative hypotheses of relative inflation (Harrod-Balassa-Samuelson, relative factors endowment, and demand effects). Due to the lack of reliable data on capital stocks only a curtailed version of the model is tested here empirically. The various specifications of the model are estimated for the majority of EU countries, using the Pedroni panel group mean FMOLS estimator. In general, relative labour productivity and demand factors turn out to be significant and correctly signed, though evidence in favour of the latter effect seems to be less robust. In addition, differences in the determination of relative prices between the new and old EU Member States are found. They seem to be consistent with theoretical considerations and the transition phenomenon. The estimation results are very sensitive to the definition of non-tradables. The paper also discusses policy implications for overall inflation, stemming from relative price models. It questions the usefulness of relative inflation models for the analysis of overall inflation differentials and practical policy decisions.