Sectoral specialisation in the EU a macroeconomic perspective

has tended to be greater than in the other
sectors.

Moreover, the comparison between the two
sub-periods 1985-1990 and 1996-2001 shows a
slight tendency for manufacturing productivity
to increase, with the exception of Spain,
Portugal and the UK, while labour productivity
developments in business sector services were
much more diverse; the EU and euro area
average for these sectors decreased despite the
fact that labour productivity held up in these
sectors for Belgium, Spain, the Netherlands and
the UK.

Box 3

PART-TIME EMPLOYMENT AND MEASUREMENT OF SECTORAL SPECIALISATION

One of the ways of assessing the extent of sectoral specialisation is the Lilien indicator, which
measures the variation of the speed of structural adjustments. This measure is based on the
number of persons employed per sector. In countries where part-time work is relatively
abundant, the number of persons employed in a sector may not accurately represent the share of
that sector in the total number of hours worked. This is because the measure does not take into
account the possibly differing average number of hours worked across sectors. More
specifically, counting heads instead of hours may overstate the weight of sectors where the
number of hours worked is low and vice versa. Furthermore, if the evolution of working hours
varies across sectors, sectoral labour productivity growth is likely to be distorted when
measured as the number of persons employed rather than number of hours worked.

The extent of these potential measurement problems depends on the distribution of working hours
across sectors. Distortions regarding the level of productivity are, ceteris paribus, likely to be larger
in countries where part-time work is abundant in specific sectors than in countries where the number
of hours worked does not vary significantly across economic sectors or in the course of time.

For most countries, detailed sectoral data on the number of hours worked is unavailable.
Furthermore, comparing national data on hours worked is hampered by conceptual issues.
SNA93 guidelines suggest the use of full-time equivalent employment for international
comparisons. Full-time equivalent employment represents the actual (as opposed to contractual)
number of hours worked, divided by the average actual annual hours worked in full-time jobs
within each respective economic territory. This measure is less prone to errors than data on the
total number of hours worked and insensitive to differences in the average actual annual hours

Part-time work: deviation from full-time equivalent

(1999, percentages)

Sources: OECD, NCB calculations.

Note: Data for Portugal refer to 1998.

ECB

Occasional Paper No. 19

July 2004

<<   <   26   27   28   29   30   31   32   >   >>