Rates of Return on Plant and Machinery in the Regions of the UK, 1968‐1991
| Author | Barry P. Andrew,Richard I.D. Harris |
| DOI | http://doi.org/10.1111/1467-9485.00164 |
| Date | 01 August 2000 |
| Published date | 01 August 2000 |
{Journals}sjpe/47_3/x180/makeup/x180.3d
ScottishJournal of PoliticalEconomy, Vol. 47, No.3, August 2000
#Scottish Economic Society 2000. Published by Blackwell PublishersLtd, 108 Cowley Road, Oxford OX4 1JF, UK and
350 Main Street, Malden, MA 02148, USA
RATES OF RETURN ON PLANT AND
MACHINERY IN THE REGIONS OF THE UK,
1968± 1991
Richard I.D. Harris and Barry P. Andrew
ABSTRACT
This paper provides an update of an earlier study undertaken by one of the authors and
extends the analysis by modelling the determinants of rates of return, using an
appropriate cointegration based approach. A second major extension was to test
formally whether rates of return have been converging over time. The results from
testing for convergence show that there is little evidence of catch-up, either because
long-run convergence already existed during this period, or because there was
divergence (away from the East Anglia=South East region). With regard to what
determines rates of return, the peripheral regions were generally more responsive to
changes in total factor productivity than were the `southern' regions. Similarly, `capital
deepening' resulted in larger responses in the peripheral regions. In contrast, the
impact of a change in labour's share in value-added on rates of return had no obvious
spatial pattern. Our results also indicated that the largest source of interregional
differences in rates of return was regional differences in the underlying determinants,
rather than differences in model elasticities per se. Of the changes in the underlying
determinants, increases in capital deepening over the period were predominant.
II
NTRODUCTION
Estimates of rates of return for manufacturing industries in the UK regions
covering 1968± 78 were provided by Harris (1982a). This paper updates the
earlier information and extends the analysis in several ways. Firstly, and most
importantly, a model of the determinants of rates of return is presented and
estimated, using an appropriate cointegration based approach. A second major
extension is an attempt to test formally whether rates of return have been
converging over time. A time series approach to testing for convergence is used,
and some interesting results are obtained, although it is recognised that the data
only covers a relatively short period of time.1
304
University of Portsmouth
1Coincidentally, there also appears to have been a number of underlying shifts in the data,
which further complicate testing for convergence.
{Journals}sjpe/47_3/x180/makeup/x180.3d
Before discussing the approach and results obtained, it is useful to briefly
consider the importance of this topic vis a vis the (regional) literature dealing
with the manufacturing sector and regional industrial policy. In a relatively early
attempt to model rates of return, Day (1973) notes that:
... the motivation for a firm to expand output in a particular market or to
enter a market or indeed to move any resources into a market, is taken to be
the rate of return that the firm expects to earn in that market ... (Thus) it may
be presumed that firms will only enter markets promising above average rates
of return (Day, 1973, p. 136).
However, little overt attention seems to have been given to the importance of
rates of return with regard to such issues as what determines regional investment
(in situ or inter-regional, or foreign inward investment), and related matters such
as the location and movement of industry. Indeed, as noted by Richardson (1978,
p. 65) the earliest theoretical models of industrial location did not assume profit-
maximisation behaviour since the optimal location was conceived as the
minimum-transport-cost site (later broadened to the least-cost location, which is
only equivalent to the profit-maximising location when demand is spatially
invariant). More recently, studies of industrial movement have usually taken only
a rather generalapproach to recognising that the relative attractivenessof a region
is important as a determinant of interregional movement (cf. Armstrong and
Taylor, 1993, pp. 344± 45), but few have concentrated on financial indicators
outside of government inducements to lower the cost of capital in the periphery
via tax incentives and capital grants.2A partial exception is Sant (1975), who
included the percentage change in manufacturing output in his multivariate
analysis of interregional movement between 1945± 65, finding this to be positive
and significant for regions like the North and Scotland, but this `accelerator' type
approach to modelling investment is only loosely linked to rates of return.3
In general terms, estimates of rates of return on capital employed might be
thought an important factor in determining various aspects of (regional)
investment, including the attractiveness of a region (and=or industry) for inward
investment, as well as providing evidence of the extent to which entry (and exit)
barriers have an impact on spatial competition. In addition, and since fixed
capital is only mobile at the margin, undertaking investment at various locations
involves long-term `sunk costs', and this suggests that if there is any tendency for
regions to converge towards an equilibrium rate of return across industries
(perhaps with government help), this will inevitably be a slow process.
2Moreover, government policy to encourage industrial movement (and generally improve the
attractiveness of the peripheral regions) was not meant to boost profit margins per se, through
grants and subsidies, but rather to lower the `user' cost of capital. Regional financial incentives
that leak away into profits was deemed to be evidence of `dead-weight loss' in the 1980s, and a
prime reason for abolishing automatic capital subsidies.
3Equation (6) below shows that increases in output will (cet. par.) raise the rate of return, but
only if factor inputs (principally labour, given that changes in the capital stock tend to be
relatively small when compared to the overall stock of capital in the short-run) do not
simultaneously increase costs to such an extent that the rate of return is lower.
RATES OF RETURN ON PLANT AND MACHINERY 305
#Scottish Economic Society 2000
To continue reading
Request your trial