9-11 May 2008, The University of Warwick, CSGR

(in cooperation with Universidad Carlos III, Madrid)

Nikolaus Wolf (University of Warwick)

Joan Ramon Roses (Universidad Carlos III)

• Why Regional GDP, 1900-2000?

Much work has been recently done on the economic development of European regions in the EU-25, including Switzerland, Norway and the Balkans, providing new insights into issues like regional convergence, clusters of activity, or specialisation patterns with remarkable policy implications. However, all this research takes the 1960s as its starting point, because no comparable data has been available for earlier periods. Given that spatial structures of economies tend to be extremely persistent over time, studies based on evidence over just the last 5 decades might be seriously misleading. To give an example: without taking into account the potential impact of the two World Wars and especially the division of Europe, we cannot distinguish between the effects of regional policies and the long-run reconstruction of an earlier European division of labour in the wake of Europe’s post-war and post-communist reunification.

• Focus of the Workshop

The Warwick workshop aims to coordinate the recent work that has been started in various European countries to construct GDP time series evidence at the level of European Regions over the last century. There are basically three areas where this kind of research urgently needs to be coordinated to be used for a long-run analysis of the regional development across all parts of the EU-25: first methods, second time and third geography.

To start with methods, attempts to re-construct regional GDP-levels following a national accounting framework would generally provide the most reliable estimates. However, the data necessary for such an approach is often unavailable. An attractive alternative, which is less demanding in terms of data has been suggested by Geary and Stark (Economic Journal 2002) and extended by Crafts (2004). To illustrate this approach, see the brief methodological paper "How to construct regional GDP series for European regions" (Roses and Wolf 2007). We consider the approach detailed in that short paper as the minimal requirements for usable regional GDP estimates. We aim to discuss how these various methods to construct regional GDP data can be integrated into a common data-base.

Second, even with such a minimalist approach the available data will not allow constructing complete annual time-series for all regions of today’s EU-25. Therefore, we suggest limiting and coordinating efforts on benchmark estimates at about ten-year intervals: 1900, 1910 (just before WW1), 1925 (after the European hyperinflations), 1930, 1939 (before WW2), 1950 (after WW2), 1960, 1970, 1980, 1990, and 2000. When and where these dates should vary across countries, especially for periods before 1950, will be subject to discussion, but we certainly need to have benchmark estimates at very close points in time for a Pan-European study.

Third, it will be crucial to coordinate the geographical units of research. Here the aim must clearly be to find historical units as closely as possible to today’s NUTS-2 regions (for a clarification see the classification system at a national level http://ec.europa.eu/comm/eurostat/ramon/nuts/introannex_regions_en.html). Historical borders do often follow these modern NUTS-2 borders, but we need to discuss how to deal with cases where this is impossible.

We aim at covering all European regions but especially invite scholars that work on South-Eastern Europe (Greece, Romania, Bulgaria, the Balkans) and Scandinavian Countries to send in their papers.

We expect to provide travel funding, lunches, and accommodation, supported by the ESF GlobalEuroNet initiative. The number of scholars who can participate is limited.

Priority is given to participants from countries which financially support the ESF Programme (Austria, Belgium, Estonia, Finland, Germany, Hungary, Ireland, Norway, Spain, Sweden, Switzerland, The Netherlands, and Turkey). However, scholars from other countries are most welcome as well and can also be financially supported to some extent.

The organizing committee consists of Joan Ramon Roses (UC3M, Madrid) and Nikolaus Wolf (Warwick). Papers or abstracts should be sent to nikolaus.wolf@warwick.ac.uk no later than 20th January 2008. All submissions will be acknowledged. Notices of acceptance will be sent to corresponding authors by February 29th 2008.