Labour Productivity and Technology Gap in European Regions: A Conditional Frontier Approach

Abstract

Filippetti A. and Peyrache A. Labour productivity and technology gap in European regions: a conditional frontier approach, Regional Studies. A conditional frontier approach is proposed to capture the role of the technology gap in explaining labour productivity differences in 211 European regions in eighteen countries over the years 1995–2007. Labour productivity growth is driven by capital accumulation and technical change. In lagging behind regions, productivity growth is mainly driven by capital accumulation. The technology gap does not play a role in driving labour productivity growth and remains stable across regions in the considered period. Cohesion policy seems more effective in terms of fixed investment rather than technological capabilities, while technology gap remains a source of unused potential productivity growth.

Filippetti A. and Peyrache A. 欧洲区域的劳动生产力与技术差距：边界方法，区域研究。本研究提出有条件的边界方法，用以捕捉解释劳动生产力差异中技术差距的角色。自1995年至2007年间，欧洲十八个国家的211个区域中，劳动生产力的成长受到资本积累与技术革新所驱动。在落后的区域中，生产力的成长主要受到资本积累所驱动。技术差距在驱动劳动生产力成长中并不具任何角色，并且在研究时程中，于各区中皆维持稳定。相较于科技能力而言，凝聚政策似乎在固定投资方面较具有作用，而技术差距则维持做为未使用的潜在生产力成长之来源。

Filippetti A. et Peyrache A. La productivité du travail et le fossé technologique dans les régions d'Europe: une approche frontière conditionnelle, Regional Studies. On propose une approche frontière conditionnelle pour saisir le rôle du fossé technologique comme facteur explicatif des écarts de productivité du travail en 211 régions européennes situées dans dix-huit pays pendant une période allant de 1995 jusqu'à 2007. La croissance de la productivité du travail est alimentée par l'accumulation du capital et par la mutation technologique. Dans les régions retardataires, la croissance de la productivité est largement alimentée par l'accumulation du capital. Le fossé technologique ne joue aucun rôle quant à l'alimentation de la productivité du travail et reste stable à travers les régions au cours de la période considérée. La politique de cohésion semble plus efficace en termes des investissements fixes plutôt que des capacités technologiques, alors que le fossé technologique reste une source de croissance potentielle inexploitée de la productivité du travail.

Filippetti A. und Peyrache A. Arbeitskraftproduktivität und die technologische Lücke in europäischen Regionen: einen konditionalen Frontier-Ansatz, Regional Studies. Wir präsentieren einen konditionalen Frontier-Ansatz zur Erfassung der Rolle der technologischen Lücke bei der Erklärung von Unterschieden hinsichtlich der Arbeitskraftproduktivität in 211 europäischen Regionen von 18 Ländern von 1995 bis 2007. Die Arbeitskraftproduktivität erhöhte sich im Zeitraum aufgrund von Kapitalansammlung und technischen Veränderungen. In Regionen mit verzögerter Entwicklung hing das Produktivitätswachstum in erster Linie von der Kapitalansammlung ab. Die technologische Lücke spielte als Faktor für das Wachstum der Arbeitskraftproduktivität keine Rolle und blieb im untersuchten Zeitraum in den Regionen stabil. Die Kohäsionspolitik scheint hinsichtlich der festen Investitionen eine größere Wirkung zu erzielen als hinsichtlich der technologischen Kapazitäten; gleichzeitig ist die technologische Lücke weiterhin eine Quelle von ungenutztem potenziellem Produktivitätswachstum.

Filippetti A. y Peyrache A. Productividad laboral y la brecha tecnológica en las regiones europeas: un planteamiento fronterizo condicional, Regional Studies. Proponemos un planteamiento fronterizo condicional para captar el papel de la brecha tecnológica a la hora de explicar las diferencias de la productividad laboral en 211 regiones europeas de dieciocho países entre 1995 y 2007. El crecimiento de la productividad laboral está condicionado por la acumulación de capital y el cambio técnico. En las regiones menos desarrolladas, el crecimiento de la productividad está impulsado principalmente por la acumulación de capital. La brecha tecnológica no desempeña un papel importante a la hora de impulsar el crecimiento de productividad laboral y sigue estable en las regiones durante el periodo analizado. La política de cohesión parece más eficaz en términos de inversión fija que las capacidades tecnológicas, mientras que la brecha tecnológica sigue siendo un recurso no utilizado de un posible crecimiento de productividad.

• Labour productivity growth
• Technology gap
• Data envelopment analysis (DEA)
• European Union Cohesion policy

• 劳动生产力成长
• 技术差距
• 数据包络分析 (DEA)
• 欧盟凝聚政策

• Croissance de la productivité du travail
• Fossé technologique
• Analyse d'Enveloppement des Données
• Politique de cohésion de l'Union européenne

• Wachstum der Arbeitskraftproduktivität
• Technologische Lücke
• Dateneinhüllanalyse (DEA)
• Kohäsionspolitik der Europäischen Union

• Crecimiento de la productividad laboral
• Brecha tecnológica
• Análisis envolvente de datos (AED)
• Política de cohesión de la Unión Europea

JEL classifications:

• C14
• O3
• R11

Acknowledgements

The authors would like to thank Prasada Rao for useful suggestions; and the participants of the Schumpeter Society Conference 2012. Part of this work was carried out during the visiting period of Antonio Peyrache at the Research Council of Italy (CNR); and of Andrea Filippetti at the University of Queensland, Centre for Efficiency and Productivity Analysis. The authors are indebted to the Centro Ricerche Economiche Nord Sud (CRENoS) and Eurostat for support on data issues. The authors are also grateful to two anonymous referees and the Editor. The usual disclaimer applies.

Notes

1. This argument has been put forward by several scholars from different perspectives. Economic historians have argued that developing countries may exploit their backwardness position by imitating and adopting new technologies produced in advanced economies, emphasizing the cost of imitation and the requirement of technological capabilities (Gerschenkron, 1962; Landes, 1969; Abramovitz, 1986). Technology gap models show that imitation and the catching-up process are expected to be more rapid when the potential provided by a large technology gap can be exploited by means of a well-developed absorptive capacity (Castellacci, 2011; Benhabib and Spiegel, 2005). A related stream of research points to the development of technological capabilities as a precondition for countries to generate and manage technical change, in terms of both the introduction of new technology as well as the adoption of existing technology (Lall, 1992; Filippetti and Peyrache, 2011).

2. Cohesion policy ‘aims to promote harmonious development of the Union and its regions by reducing regional disparities’ (Article 174 of the Treaty).

3. Henderson and Russell (2005) put forward a quadripartite decomposition of productivity changes. Their decomposition modifies the measure of labour used by incorporating an augmentation for human capital; then they decomposed ex-post introducing an additional component accounting for human capital accumulation. This is not incompatible and is indeed complementary to what is being done in the present paper. Moreover, consistent with the technology gap argument, a measure of human capital more related to the capacity of mastering technologies is used, rather than standard measures of human capital (for example, years of schooling) which are usually employed in growth-accounting studies.

4. In order to account for price differences across regions, a price deflator must be used. The International Comparison Program (ICP, The World Bank) strongly recommends the use of PPPs as a price deflator instead of simple market exchange rates. This is because exchange rates are – at best – a reflection of the purchasing power of the tradable goods of the economy and are subject to erratic movements which reflect capital flows (financial volatility). On the contrary, PPPs are based on a basket of goods and services that is more representative of the structure of the economy. Therefore, PPPs better reflect price changes (that is, they are not subject to the volatility of exchange rates). The authors are extremely grateful to Professor Prasada Rao for a useful clarifying discussion on this specific point.

5. Data can be accessed from the Centro Ricerche Economiche Nord Sud (CRENoS) website (http://crenos.unica.it/crenos/en/content/databases) (see also Dettori et al., 2012).

6. A reliable alternative option is represented by R&D expenditure as an input of innovation activity, but data are not available for the present sample.

7. The cited study by Enflo and Hjertstrand (2009) takes Germany at NUTS-1 level as well. They showed that the distribution of labour productivity is not significantly affected.

8. New Member State Objective 1 regions are those belonging to the East European countries, namely: the Czech Republic, Hungary, Latvia, Lithuania, Poland, Romania and the Slovak Republic. Old Member State Objective 1 regions are instead those belonging to the Western European countries (EU-15) (see Table A2 in Appendix A).

9. It was decided not to incorporate the whole exercise to avoid overburdening the exposition of the main results. The results from the bootstrap are available from the authors upon request.

10. The authors are indebted to two referees for these points.

11. Contrary to other econometric methods (for example, stochastic frontier analysis – SFA), DEA does not assume the efficiency scores to be independent of the inputs. Moreover, DEA allows for spatial correlations of the efficiency scores. Therefore, for example, if there is a positive spillover due to co-location, this will be reflected in a higher efficiency score for the recipient region. In other words, the efficiency scores could be correlated with the matrix of geographical distances.

12. The authors are indebted to an anonymous referee for this point.