Lilien indicator only partly reflects the
adaptability of production in a country to
changes in domestic and international demand,
given that considerable adjustment occurred
within, rather than between, sectors30.
Comparing the speed of structural adjustments
between manufacturing on the one hand and
utilities and business sector services on the
other, Chart 5 reveals that over the last 20 years
sectoral re-allocations have been more rapid
within manufacturing than within utilities and
business sector services. Moreover, while
structural adjustment significantly slowed
during the 1990s within utilities and business
sector services, it only slowed significantly
from the end of the 1990s in manufacturing,
after a strong acceleration at the beginning of
the 1990s.
2.3 SECTORAL COMPOSITION AND
PRODUCTIVITY
Sectoral specialisation and its evolution are
likely to affect the labour productivity growth
experienced by EU countries. This sub-section
provides an overview of labour productivity
growth across sectors and countries over the
last 15 years for EU countries and for the
United States. Moreover, it determines for the
manufacturing sector the contribution to labour
productivity growth of different sub-sectors,
clustered as high, medium-high, medium-low
and low technology intensity sectors. Finally, it
determines to what extent aggregate labour
productivity growth has been driven by shifts in
sectoral composition, carrying out a shift-share
analysis of labour productivity growth.
2.3.1 SECTORAL LABOUR PRODUCTIVITY
DEVELOPMENTS
Table 3 shows labour productivity growth,
measured as the gross value added at constant
prices divided by total employment, for EU
countries as well as the EU, the euro area and
the United States over the period 1985-2001
and for the sub-periods 1985-1990 and 1996-
2001. The table displays data for the aggregate
economy, for the manufacturing sectors and for
the utilities and business sector services.
Caveats concerning the difficulties of
measuring output in the service sectors must be
borne in mind and caution is required when
interpreting labour productivity growth
developments in services, especially in some
sub-sectors31. Moreover, given the lack of data
availability, it was not possible to measure
sectoral labour productivity per hours worked.
Labour productivity growth may therefore be
underestimated in some countries or sectors,
where the proportion of part-time workers has
increased notably32.
The table reveals that differences in
productivity growth seem to be as significant
across sectors as across countries, meaning that
country and sector characteristics are both
important in determining labour productivity
growth.
Regarding sectoral labour productivity
developments of main aggregates in the EU
(euro area), Table 3 shows that on average
labour productivity growth performance was
the highest in the electricity, gas and water
supply sector when measured over the entire
period. Moreover, labour productivity growth
was on average twice as high in the
manufacturing sector as it was in business
sector services. Within the manufacturing
sector, labour productivity growth was the
highest in the high technology intensity sector,
and here in particular in the “machinery and
equipment” sector. The very high productivity
growth in this sector in the United States
further highlights the importance of ICT
investment for productivity. Within business
sector services, labour productivity growth was
by far the highest in the post and
telecommunications sector and, to a lesser
extent, in the financial intermediation sector.
Overall, in the EU and the euro area labour
productivity growth in the network industries
30 See OECD (1987), “Structural adjustment and economic
performance”, Paris.
31 For more details on the measurement of real value added, see
Annex 4.1.3.
32 See Box 3 for a discussion of this issue.
24
ECB
Occasional Paper No. 19
July 2004