- agriculture;
- industry including energy (hereafter
industry);
- construction;
- trade, restaurants, transport and
communication (trade);
- finance and insurance (finance);
- real estate, renting and business activities
(business activities);
- community, social and personal services
(public services).
Confirming the initial remarks, Table 6 shows
that the variance of industry consistently
accounted for the largest share of the overall
variance between 1988-2002 followed by the
variances of trade, construction and business
activities. These last two service sectors are
less volatile43 than industry (partly on account
of the absence of the inventory cycle) and have
been rapidly expanding their share of total value
added since the 1980s. By contrast, agriculture
and finance, despite having the highest degree
of relative volatility per country (as will be
discussed below, see Table 7), contribute little
to EU-wide volatility. This is possibly related
to the fact that these last two sectors are
relatively more exposed to country-specific
fluctuations.
The analysis looks in particular at the sub-
periods 1993-97 to capture the effect of the
Single Market and 1998-2002 to capture the
effect of the introduction of the single currency.
We find that only very little change occurred
between these two sub-periods, with the seven
sectors considered contributing slightly less to
the overall variance of the euro area cycle in the
period 1998-2002. In addition, the covariance
of industry and trade and industry and business
activities has tended to become more important
between these two periods, partly due to the
increased linkages between these two sectors as
a result of the outsourcing of many services
from manufacturing.
The sectoral breakdown of aggregate volatility
for the euro area cycle supports the view that
sectoral composition is important for the
analysis of aggregate volatility. In particular,
the strong differences in the contributions of
industry compared with business activities and
public services may give rise to country-
specific business cycle patterns that are
determined in part by differences in sectoral
composition.
3.2 SECTORALVOLATILITY
Differences in volatility across countries and
sectors are a characteristic element of aggregate
business cycles. Documenting the sectoral
sources of aggregate fluctuations in EU
countries can therefore provide information
about the driving forces for business cycles.
This section focuses on comparing the relative
sectoral volatility across EU countries over the
period 1980-2001 on the basis of the
disaggregation into seven main sectors, as
described above. Moreover, in the light of the
important sectoral changes that some EU
countries have undergone over the last two
decades, the impact of change in sectoral
composition in EU countries on the change of
aggregate production volatility is also
documented44.
3.2.1 RELATIVEVOLATILITYANDITS
EVOLUTION OVER TIME
In order to calculate the relative volatility of the
sectoral indicators with respect to each
country’s total GDP volatility, the business
cycle component was extracted from aggregate
and sectoral gross value added.45 As can be seen
from Table 7, substantial differences exist
between sectors, with agriculture and finance
being the most volatile sectors, followed by
construction. Service sectors such as trade and
business activities show minor volatility, with
the notable exception of the UK where volatility
in trade almost reaches the level of volatility
experienced in agriculture.
43 For sectoral relative volatility measures, see Table 7.
44 See Annex 4.2.2 for the methodology that was used to extract the
business cycle components.
45 Volatility measures were constructed on the basis of an 11-
quarter symmetric rolling window.
3 SECTORAL
COMPOSITION
AND BUSINESS
CYCLES
ECB
Occasional Paper No. 19
July 2004