The Matrix Approach to Industrial Policy

Abstract

The objective of this paper is to analyse the development of industrial policy in the European Union. The changing concepts and rationale for industrial policy are reported with reference to the treaties and communications of the European Commission. Industrial policy has followed the course of these changes, although in reality, the practical changes have occurred at a slower and less radical pace than the changes in philosophy. A new paradigm of industrial policy has recently been developed by the European Commission in a document that first calls for measures improving the competitiveness in all sectors in terms of rather broad measures, and subsequently acknowledges sector differences in the impact of these measures, calling for complementary sector‐specific measures. We label this new paradigm the ‘matrix approach’ to industrial policy, because it combines horizontal and vertical measures. On the empirical level, we see a divide between countries practising a future‐oriented approach emphasising innovation and knowledge, and countries that take a more defensive stance, with higher subsidies and a greater amount of regulation. The countries pursuing the more future‐oriented approach achieve higher shares of technology‐driven and skill‐intensive industries and excel with respect to the goals of the Lisbon strategy.

Key Words:

• Industrial policy
• Lisbon strategy
• horizontal vs vertical policies
• framework conditions
• competitiveness
• innovation

JEL Classification:

• L52

Notes

1. These shares are based on WIFO calculations from AMECO. Manufacturing is defined as NACE 15 to 36. Wide definitions of manufacturing include construction, energy and mining; this raises the share by about 7 percentage points. According to OECD figures (STAN database), the share of manufacturing in GDP amounts to 19%.

2. 1993 was a year of recession.

3. However, the combined share of manufacturing and business services overestimates the importance of ‘wider manufacturing’, because parts of business services are demanded by other sectors, e.g. the service sector itself.

4. Jacquemin (1975), for similar view see also Stöllinger (2000), Hepperle (2004).

5. The Maastricht Treaty is widely seen as the start of a community‐wide industrial policy. Interestingly, it does not use this term explicitly; the headline of the chapter reads ‘Industry’.

6. A Committee on Industrial Policy was proposed by the Council in the same year, but never set up. In 1973, the Council of the European Community set up guidelines for a European Industrial Policy, which concentrated on removing barriers to production and trade and also published the ‘European scale promotion of competitive advanced technology undertakings’. Article 130 EEA of the Single European Act (1986) formed the basis for a common research and technology policy. Its goals of strengthening the scientific and technological basis of industry and its competitiveness are clearly related to industrial policy.

7. ‘Some key issues in Europe’s competitiveness—towards an integrated approach’, European Commission (2003a). Absolute de‐industrialisation may be defined as the long‐term decline of the manufacturing sector, implying an absolute decline in employment, production, profitability, exports and capital stock in the manufacturing sector and the emergence of persistent manufacturing trade deficits. However, absolute de‐industrialisation should be distinguished from relative de‐industrialisation characterised by a decline in the share of manufacturing in GDP, which reflects a process of structural change towards a service‐dominated economy, as a result of productivity growth in manufacturing, increased real income and therefore a rising demand for services.

8. It may be an interesting task to investigate the relation between the four phases of industrial policy and more general economic and technological trends. The importance of mechanics and electronic technology may have been the cause of the sectoral approach in the 1950s and 1960s. We thank Gunther Tichy for this suggestion.

9. Commission of the European Communities, Communication from the Commission, Implementing the Community Lisbon Programme: A policy framework to strengthen EU manufacturing—towards a more integrated approach for industrial policy, COM(2005) 474 final, Brussels, 5 October 2005.

10. Darmer & Kuyper (2000, p. 4): As an example of a broad definition, see Johnson (1984): ‘Industrial policy means the initiation and co‐ordination of governmental initiatives to leverage upward the productivity and competitiveness of the whole economy and of particular industries in it.’ For a narrow definition, see for example Tyson & Zysman (1983) ‘Industrial policy … means government policy aimed at or motivated by problems within specific sectors’ or Krugman & Obstfeld (1991): ‘Industrial policy is an attempt by a government to encourage resources to move into particular sectors that the government views as important to future economic growth.’ There is not even agreement on whether the term ‘industry’ in industrial policy indicates manufacturing—as is more usual in Europe—or whether industrial policy should include measures of producer related services, finance, consulting, and transport (or even construction and energy, which would conform to the US interpretation of ‘industries’ as sectors)

11. Aiginger et al. (1998, 2001). The respondents were economists in the field of industrial organisation, competition policy and innovation, with the majority working in academic institutions and a minority in government institutions and competition authorities. In a survey among economists and policy advisors, 68% of the respondents disagreed with the claim that ‘the best industrial policy is no industrial policy’ while only 24% agreed. Rejection is somewhat weaker among the US experts, but even here a slight majority rejects laissez faire. As to the success of industrial policy, the opinion is mixed: the experts are evenly split on the question whether ‘industrial policy is subsidisation in disguise’, and whether ‘recommendations by experts have an influence on industrial policy’. The results of these surveys, which were carried out in the late 1990s, tell us that the importance of industrial policy is widely acknowledged, while its success is seen as mixed.

12. For an overview of the French approach, see Alcouffe (2005). Outstanding examples of the ‘French approach’ are Airbus, Concorde, Minitel, HDTV, and TGV. Often, supporting the space industry and energy is also project‐oriented, or follows a specific strategy with a strong top‐down component.

13. Denmark recorded the highest share of state aid to manufacturing, amounting to more than 4% of value added; shares above 2% are reported for Greece, Luxembourg and Spain. The smallest share—less than 1%—was measured for the UK, Sweden and Finland.

14. It is labelled ‘dynamic product market regulation’ in Aiginger (2004b), because it is available over time and for more than just one year.

15. The overall EPL indicator is basically a sum of the two indicators for regular and temporary contracts. These are constructed by using factor analysis to aggregate the basic indicators that were calculated according to an in‐depth review of existing regulations. For regular contracts, indicators of procedural requirements (duration, notice and severance pay and standards and penalties for unfair dismissals) were reviewed. For temporary contracts, the objectives according to which temporary contracts can be offered, the maximum number of successive renewals and cumulated duration are reviewed. For further details see Nicoletti et al. (2000).

16. Eurostat Website Structural Indicators, Explanatory Texts: data on public procurement are based on information contained in the calls for competition and contract award notices submitted for publication in the Official Journal of the European Communities (the S series). The nominator is the value of public procurement, which is openly advertised. For each of the sectors—works, supplies and services—the number of calls for competition published is multiplied by an average based, in general, on all the prices provided in the contract award notices published in the Official Journal during the relevant year.

17. The share is higher in the EU than in the USA, although the number is still higher in the USA (because of the higher overall number of graduates).

18. Participation rates are only a crude indicator, as no statement regarding the quality or the amount can be made (European Commission, 2004b).

19. Broadband penetration measures the number of broadband connections in relation to the population (100 × number of broadband access lines/number of inhabitants). Broadband lines are defined as those with a capacity equal or higher than 144 kbits/s (Eurostat Website Structural Indicators, Explanatory texts: http://europa.eu.int/newcronos/reference/sdds/en/strind/innore_bp_sm.htm).

20. In this paper, the term ‘Scandinavian countries’ refers to Sweden, Denmark and Finland, the terms ‘Scandinavian’ and ‘Nordic’ are therefore used interchangeably.

21. Eurostat Website Structural Indicators, Explanatory texts: (http://europa.eu.int/newcronos/reference/sdds/en/strind/innore_vc_base.htm).

22. European Commission (2004a). This work includes no figures on raising venture capital in the largest markets; the largest markets for raising venture capital funds are the UK, Sweden and the Netherlands.

23. For a definition of technology‐driven industries, see Peneder (2001).

24. Flexicurity is the technical term for a regime combining flexibility for firms (easy dismissals) with security for employees (new job offers or training).

25. With the exception of the Netherlands, which succeeded in attracting many headquarters because of its geographical position and favourable tax regimes.