10 May 2000 • Do Growth Rates Differ in European Manufacturing Industries? • Michael Pfaffermayr

The determinants of industry growth are of significant relevance to industrial policy. On the one hand, manufacturing growth can be viewed as part of the overall growth process. More important, however, differential growth across countries or industries contributes to the shaping of Europe's entire industrial structure.

According to the theoretical literature, growth at the industry level is either determined by demand side factors (e.g., income growth or demand for new varieties), by supply side factors (e.g., cost shifts or new technologies) or by structural factors (e.g., market structure or economic policy). Empirically confirmed factors responsible for substantial long-run differences in industry level growth rates are, from the supply side, the forces of catching up, technological factors, and endowment growth (such as growth in human capital through education); on the demand side, these factors include income and price elasticities.

On the NACE 3-digit level, the average rate of nominal growth in a typical industry amounted to 2.1 percent during the 9-year period 1989-1997. The standard deviation of 6.3 percentage points reveals the high level of variation between EU countries and between industries. 47 percent of the variation can be explained by country, sector, and combined sector and country effects. The variation across countries is more pronounced than the sector effects, indicating that the country-specific environment, economic policy and macro-economic development have a significant impact on industry growth. This picture is consistent with the view that European manufacturing is not yet fully integrated. However, as the regressions below illustrate, there is a strong tendency towards catching up and deeper integration. Most of the variation in average growth rates can be attributed to combined country-industry effects, suggesting that country-specific environments combined with industry-specific determinants common throughout the entire EU – such as demand growth – are the ingredients of long-run performance.

Furthermore, we find that specialisation and geographical concentration have a significant impact on growth. Geographical concentration is significantly negatively associated with subsequent average industry growth. This implies that the industries of those countries already holding a relatively large share of value added compared to their size grow considerably more slowly. Geographical concentration is decreasing and industry structure is becoming more equally distributed across countries. In general, specialisation has no effect on growth, with the notable exception of agglomerations.

Significant catching up has been taking place in the European manufacturing industry, with the result that industries in countries with low labour productivity have grown markedly faster. Comparative advantages based on productivity differences have levelled off. Although the direct measure of specialisation has not shown that there has been any impact, the findings on productivity can still be interpreted in this way. Convergence goes hand in hand with a decreasing degree of specialisation and thereby a more equal dispersion of industrial structures across European countries. This process is expected to deepen integration. Empirical evidence, however, shows that not all countries have managed to catch up.

Besides these forces, overall market growth, defined as the growth of apparent consumption in the triad, plays an important role in fostering industry growth. Furthermore, there is the negative impact of globalisation, measured as the ratio of exports plus imports over apparent consumption in the triad. This reflects the impact of global competition on structural change, with the more exposed industries growing on average more slowly.

Additional empirical evidence suggests that industries in which intangible assets are important (i.e., skill and R&D-intensive industries) grow faster, even after accounting for the fact that demand in these industries is more dynamic. Economic policies enhancing the skills of the workforce and research will foster growth. The remaining fixed country effects suggest that other policies, which could not be tested econometrically, will play an important role. An informal analysis suggests that liberal market regimes, flexible labour markets, and programs for small firms seem to be the most effective ingredients for promoting growth, but the best policy mix may differ from country to country and depend on its income per capita. Denmark and Ireland, which have above-average growth performance, may provide benchmark examples of successful policy mixes.