allocates employment to the sector to which the establishment is classified by its principal
production activity. So railway employees are allocated to the transportation sector, whether
the transportation supplied is a service to final consumers, or involves the movement of fuel to
power stations or steel to car plants. The input-output approach, by contrast, attributes to the
sector the employment contributing to the final output across all the stages of production.
Where the steel is delivered to car plants the transportation involved and the associated
employment are attributed to the final output of motor vehicles; and similarly for the inputs into
later stage inputs into final production, as with the fuel to electricity generation to the final
consumer. On the same perspective, efficiency gains are not confined to the point of delivery
but can be achieved throughout the supply chain. This perspective has been designated by
Pasinetti the ‘vertically integrated sector’ (Pasinetti, 1973). It will be the key concept in much of
the remainder of the paper.
The structure of the paper is as follows. Section 2 gives a more formal development of the
concept of the vertically integrated sector (VIS), and illustrates the re-attribution of employment
on this basis. Section 3 then reviews the employment-intensities of individual industries on the
VIS basis. These are compared over time and across the six economies. In Section 4 we examine
the relative employment-intensity of manufactures as against services within this framework, and
find the familiar generalisation about the labour-intensity of services to have little empirical
support. Section 5 estimates the contribution to employment growth which can be attributed to
the changing pattern of demand in each country. We do this for the product mix both in overall
final demand and within the consumption basket, and find that the evolving product mix within
final demand has been mildly employment-friendly in European countries. In Section 6 we
estimate a set of counterfactuals across countries, deriving the employment levels which would
have resulted in the US had final demand and consumption followed the mix from each of the
European economies and, conversely, had each of the European economies followed the more
service-oriented demand and consumption patterns of the US rather than their own. Again the
implications for the level of employment are found to be very limited. In Section 7 we take up
the analysis of employment change within the six economies, attributing this across the three
proximate sources of the growth of final demand, structural change as encapsulated in input-
output relations, and labour productivity growth. This reveals some striking empirical
regularities. Structural change is on occasion job-creating, on occasion job-destroying, and on
occasion job-transferring, for example through outsourcing. But its overall contribution to