and Morrill (1959), including several of his
soon-to-be-famous graduate students (the
‘‘space cadets’), combined German location
theory with demanding statistical analysis in
their studies of highway development in
Washington State.
Haggett’s influential (1965) book, Loca-
tional analysis in human geography, summarized
and solidified the new approach to location
theory (see alsolocational analysis). It was
to be formal, empirically rigorous, and under-
girded primarily byneo-classical economics
(both indirectly via the German school and
regional science and directly from mainstream
economics). A central topic was the firm, a
staple of neo-classical micro-economic analy-
sis and underpinned by assumptions of
rational choice theoryand perfect compe-
tition (many small producers, none of which
control price). The theory of the firm was
translated directly into spatial terms by Isard,
and later in economic geography by Smith
(1981a [1971]). But chinks in the armour of
classic location theory were soon revealed.
First, the determinacy of rational choice was
shown to be unrealistic, and ideas of ‘bounded
rationality’, or ‘satisficing behaviour’ were
preferred, producing a literature on thebehav-
ioural geography of firm location (Pred,
1967, 1969). Second, perfect competition
was demonstrably imperfect for understand-
ing the contemporary world oftransnational
corporations, leading to an alternative litera-
ture on the ‘geography of enterprise’ (later
‘corporate geography’) focused on firms as
institutions, decision-making hierarchies and
branch plant economies (McNee, 1960;
Hayter and Watts, 1983).
The disquiet only got worse. David Harvey,
who earlier in his career had drawn on classical
and neo-classical models of agricultural loca-
tion, including von Thu ̈nen’s work, and ex-
plored general models of the evolution of
spatial patterns, turned tomarxism, and par-
ticularly Marx’s own writings, in the early
1970s and pursued a radically different ap-
proach. His subsequent analysis of the histor-
ical geography of capitalism, brilliantly
realized inThe limits to capital(Harvey, 1999
[1982]), did not so much ignore the questions
of traditional location theory as pose them in
another register altogether. His emphasis was
less on accounting for the locations of
economic activities than in disclosing the con-
nections between thecrisis-ridden dynamics
of capitalism and itsuneven development
as aspace-economy. Doreen Massey’s work
around the same time was indebted to Marx-
ism too, but it had a much more direct impact
on location theory. Like Harvey, she was
trained in the orthodox approach, even study-
ing briefly with Isard. Her reformulation
began in the early 1970s with a detailed
internal critique of orthodoxy arguing that
applying neo-classical economic theory to lo-
cation generates fundamental and irreparable
logical contradictions. The alternative that she
constructed over the next decade or so com-
bined the theoretical postulates ofpolitical
economywith the philosophical armature of
realism, culminating in her watershed book,
Spatial divisions of labour(Massey, 1984). This
was crucially important because it demon-
strated that location theory need be neither
formal nor neo-classical, and because it con-
siderably expanded both what could count as
factors bearing on location (in her case,class
andgender, politics and culture) and what
could constitute theoretical explanation (in
her case, a ‘geological metaphor’, taking the
form oflayers of investment; Barnes, 2001).
Massey’s larger point was that the object of
location theory – sites of economic activity –
could not be separated from the wider, geo-
graphically variegated social setting of capital-
ism. Her argument was starkly reinforced by
the setting of her own study: the massive
deindustrializationof the UK during the
late 1970s and early 1980s (see alsodivision
of labour; localitystudies).
A similarly inspired argument emerged
shortlyafterwardsamonga group of economic
geographers in California, although their
context was utterly different: the study of
vibrant capitalist production associated par-
ticularly with high-tech and high-end eco-
nomic activities. Scott (1988b) developed
what he called ‘neo-Weberian’ location theory
drawing eclectically on the work of neo-
MarxisteconomicstPieroSraffa(seeneo-
ricardian economics) and that of institu-
tional economists Ronald Coase and Oliver
Williamson (see institutional econom-
ics). Storper and Walker (1989) elaborated
an ‘inconstant geography of capitalism’
roiled by technological change and social
conflict. They stressed the tightly inter-
linked – ‘network’ – character of activities
in which social and economic institutions
cross-cut and interleaved within specific
regional formations. The consequence was
a rediscovery of the idea of industrial
districts, first recognised by Alfred
Marshall at the turn of the twentieth
century: dense geographical nodes of
tightly knit complementary manufacturingGregory / The Dictionary of Human Geography 9781405132879_4_L Final Proof page 427 31.3.2009 2:44pmLOCATION THEORY