have a regional component, being limited by transport costs, depending upon the spatial
configuration of the supply chain. They may also be subject to institutional factors, such as
regional brand loyalty.
As noted above, there are a number of reasons for the occurrence of localisation
economies, which can occur when firms in the same industry concentrate in a region. These
include the presence of a local pool of skilled labour, pecuniary external economies arising
from specialisation and from economies of scale in supplying firms, non-traded local inputs,
such as infrastructure and, as argued below, knowledge spillovers, which are pure
technological externalities. Urbanisation economies accrue to firms in different industrial
sectors, which cluster together in urban areas. These economies arise from the existence of a
large and varied labour market, economies of scale in provision of infrastructure and public
services, a variety of business services and again, knowledge spillovers, this time between
different industries.
The literature presents conflicting evidence concerning the relative importance of
localisation and urbanisation economies. One reason for the difficulty is that urban areas will
typically exhibit the effects arising from both types of economy. Henderson (1986) found
some evidence for the existence of localisation economies for some US manufacturing
sectors, using wage levels as a surrogate for productivity data and McCann & Fingleton
(1996) provided evidence for the existence of localisation economies in Scotland’s Silicon
Glen. Glaeser et al (1992) found evidence that local competition and urban variety promote
employment growth in US cities, rejecting the hypothesis of the existence of localisation
economies in cities, and presenting evidence for the existence of knowledge transfers between
industrial sectors, supporting the hypothesis of Jacobs (1969) on the role of urbanization in
innovation. Henderson et al. (1995) found evidence of urban externalities for high-tech
industries, consistent with the role of exchange of ideas in cities. They also found that mature
capital goods industries seemed to promote and benefit from localisation economies. In
addition, endogenous growth theory (Romer 1990) has provided a further theoretical
foundation for the existence of localisation economies.
Distinction is made between static and dynamic agglomeration economies. Static
economies result in a one-off downward shift in industry LRAC as shown in figure 1.
Dynamic economies imply a continuous downward shift over time of the LRAC as localised
industry output increases cumulatively. Whilst it seems that local pecuniary external
economies related to input supply for a (localised) industry could be one cause of dynamic
economies, the principal explanation for these economies is related to the production and use
of knowledge, in particular the effects of knowledge spillovers, there being a public good
element in the production of knowledge. The effects of knowledge spillovers on innovation
and technical change, economic growth, employment and productivity in clusters, have
become the central policy issue (OECD 1999, 2001) .
2.1 Pecuniary and technological externalities
It is important to distinguish between the two types of externality, pecuniary and
technological. These differences are illustrated by the case of a change in the transport system
involving changes in transport costs. Pecuniary externalities arising from changes in transport
costs have their origins in both the commodity and factor markets. Technological externalities
have their origins in direct interactions, outside the market, between producers (p) and
consumers (c), giving in all four different combinations (p-p, c-c, p-c and c-p).
Pecuniary externalities can be illustrated in relation to the factor market in figure 2.