Chart 3 Krugman specialisation index
1996-2001
.......... 1990-1995
1985-1989
×∕∕∕× 1985-2001
UK SE BE DK NL IT FR EU FI EA DE GR AT ES PT LU
Sources: OECD, NCBs and ECB calculations.
Note: See Annex 4.1. 1 for data details. The figures for the EU and
the euro area refer to weighted averages of country indices.
measures at the national level, while we note a
slight upward trend towards the end of the
1990s in smaller EU countries or, where large
countries are concerned, at the regional level23.
By and large, the cross-country heterogeneity
of the Krugman index remains when looking
separately at the manufacturing sector and the
utilities and business sector services.
Comparing manufacturing with utilities and
business sector services shows that EU
countries tend to be more specialised within
manufacturing.24 While the degree of
specialisation within manufacturing has tended
to decrease over time, utilities and business
sector services have shown a tendency to
become more specialised, albeit at a lower level
than in manufacturing. One reason for these
stronger cross-country differences of the
Krugman index in manufacturing could be the
higher tradability of manufacturing products.
Specialisation is a driving force of industry
concentration across countries, as shown in
Table 2. This table presents Balassa indices25
for all country-industry pairs on a
disaggregated level. These indices reflect the
weight of a sector in the production of a
particular country, relative to the weight of that
sector in total EU production. A value greater
than one indicates a relative specialisation of a
country in a particular industry. Based on these
Balassa indices, (relative) industry
concentration can be assessed, calculating the
weighted standard deviation across countries
(last column in Table 2): the higher the indicator
2 SECTORAL
SPECIALISATION:
CURRENT SITUATION
AND EVOLUTION
23 See Hallet (1999), op. cit. Hallet uses the Eurostat REGIO
database. Maps 1 and 2 in an annex of the paper illustrate the
point. For instance, the pace of increase in the specialisation
index is in the same range for, say, Finland and Ireland on the one
hand, and nine out of the 22 French regions on the other.
24 Partly, this difference in the cross-country variance of the
Krugman indicator between manufacturing and business sector
services can be related to a different degree of sectoral detail:
while 11 manufacturing sectors were available, only eight
utilities and business sector services were taken into account for
the Krugman indicator in the third column of Chart 3.
25 After B. Balassa (1965), “Trade liberalization and ‘revealed’
comparative advantage”, The Manchester School of Economic
and Social Sciences, no 33, pp. 99-123.
ECB
Occasional Paper No. 19
July 2004