behavioural industrial location theory - citeseerx

16
A contribution to the review and critique of behavioural industrial location theory by Martyn Carr During the 1960s geography was characterized by the expansion and consolidation of the paradigmatic switch to a quantitative methodology. Particularly well suited to and necessary for this quantitative methodology was the adoption of behavioural theory, providing a multivariate inductive theoretical base for a multivariate induc. tive methodology. A more significant movement for industrial geography, in particular, than the incorporation of a quantitative methodology was the develop- ment of behavioural industrial location theory during the latter half of the 1960s and the early 1970s. This was based on a rigorous questioning of the assumptions of both classical and neoclassical location theory (Dicken, 1971; Krumme, 1969a; Pred, 1967; Steed, 1971a; 1971e; Townroe, 1969). The adoption continued largely unchallenged and unabated throughout much of the 1970s with the production of a plethora of theoretical discussion and empirical analysis (Cooper, 1975; Dicken, 1976; Erickson, 1972; 1974; Krumme, 1970; Le Heron, 1973; 1976; Leigh and North, 1978; Steed, 1971a; 1974; Townroe, 1972; Watts, 1974). Behavioural industrial location theory was accepted as the justified theoretical base for industrial geography, being considered a step forward to a more realistic base than was pro- vided by earlier normative models. The development of research into behavioural industrial location theory, how- ever, was piecemeal with few attempts to coordinate the work of independent researchers and present a framework to the field. This meant that little consider- ation could be given to the exposition of the overall aims of behavioural industrial location theory or to an evaluation of its contribution to industrial location theory as a whole. Although parallel research was developed on the formulation and application of a different theoretical base (Massey, 1974a; 1974b; 1975a; 1975b; 1979; Massey and Meegan, 1976; 1978; 1979) this has not resulted in a concerted movement to define the faults and inadequacies of the prevailing behavioural orthodoxy. This paper will review contributions made to behavioural industrial location theory in the late 1960s and early 1970s using a simple framework derived from particular papers to help structure the field. The period is considered important as the formative period of behavioural industrial location theory, which involved the development of most of the aims and assumptions of the theory. The paper at PENNSYLVANIA STATE UNIV on May 12, 2016 phg.sagepub.com Downloaded from

Upload: khangminh22

Post on 06-Feb-2023

0 views

Category:

Documents


0 download

TRANSCRIPT

A contribution to the review and critique ofbehavioural industrial location theory

by Martyn Carr

During the 1960s geography was characterized by the expansion and consolidationof the paradigmatic switch to a quantitative methodology. Particularly well suitedto and necessary for this quantitative methodology was the adoption of behaviouraltheory, providing a multivariate inductive theoretical base for a multivariate induc.tive methodology. A more significant movement for industrial geography, in

particular, than the incorporation of a quantitative methodology was the develop-ment of behavioural industrial location theory during the latter half of the 1960sand the early 1970s. This was based on a rigorous questioning of the assumptionsof both classical and neoclassical location theory (Dicken, 1971; Krumme, 1969a;Pred, 1967; Steed, 1971a; 1971e; Townroe, 1969). The adoption continued largelyunchallenged and unabated throughout much of the 1970s with the production of aplethora of theoretical discussion and empirical analysis (Cooper, 1975; Dicken,1976; Erickson, 1972; 1974; Krumme, 1970; Le Heron, 1973; 1976; Leigh andNorth, 1978; Steed, 1971a; 1974; Townroe, 1972; Watts, 1974). Behaviouralindustrial location theory was accepted as the justified theoretical base for industrialgeography, being considered a step forward to a more realistic base than was pro-vided by earlier normative models.

,

The development of research into behavioural industrial location theory, how-ever, was piecemeal with few attempts to coordinate the work of independentresearchers and present a framework to the field. This meant that little consider-ation could be given to the exposition of the overall aims of behavioural industriallocation theory or to an evaluation of its contribution to industrial location theoryas a whole. Although parallel research was developed on the formulation andapplication of a different theoretical base (Massey, 1974a; 1974b; 1975a; 1975b;1979; Massey and Meegan, 1976; 1978; 1979) this has not resulted in a concertedmovement to define the faults and inadequacies of the prevailing behaviouralorthodoxy. _

This paper will review contributions made to behavioural industrial location

theory in the late 1960s and early 1970s using a simple framework derived fromparticular papers to help structure the field. The period is considered importantas the formative period of behavioural industrial location theory, which involvedthe development of most of the aims and assumptions of the theory. The paper

at PENNSYLVANIA STATE UNIV on May 12, 2016phg.sagepub.comDownloaded from

387

will define faults and inadequacies of aspects of behavioural industrial locationtheory in order to point out some of its limitations. In doing so this paper employsthe first stage of the method used by behavioural industrial location theoriststhemselves who criticized classical and neoclassical location theory before advanc-ing what was considered to be an improved theoretical framework. Such criticalevaluation of any theory is essential. In practice, this involves reviewing contri-butions to behavioural industrial location theory to assess its suitability for analysisin industrial geography as a whole.

This paper, therefore, will contribute directly towards the development of thereview and critique of behavioural industrial location theory. To review as broada subject is a large task and so this paper can only reasonably aim at partial coverageof the subject and submit this to what must necessarily become a much larger,comprehensive review of the conceptual and empirical contribution of behaviouralindustrial location theory as an element of the whole orthodoxy of industriallocation theory as it stands today. Therefore a critical review of particular aspectsof behavioural industrial location theory is developed.

I 1 Review and critique

The discussion below covers much of the research undertaken by a core of re-searchers at the University of Washington, and how this was developed by be-haviouralists ; it includes criticism of the theoretical base which they proposed.However, it would be of interest first to deal briefly with concepts which industrialgeographers took from economics in order to assess exactly what the behaviouralcontribution to the understanding of industrial geography has been.

1 Economic concepts

It can be suggested that two forms of behavioural theory were adopted by industrialgeographers. One emphasized the internal characteristics of the company and wasadvocated very early on by Dicken (1971) in the UK. The other analysed the inter-nal characteristics of the company more in relation to the underlying forces of theeconomic environment, which incorporated the former. Dicken’s proposals aremore reflective of the basic concepts used and will be dealt with first. The secondform is more characteristic of the works of the Washington School and is dealt

with more extensively in later-sections.Dicken (1971, 427) argued that behavioural theory was applicable in two main

areas: the organization structure of the firm and the goals and objectives of thefirm.

a Organization stnLCture: One of the reasons for the theoretical shift to be-havioural theory was the desire to use a theory which approximated reality moreclosely than existing marginalist theories. The main problem with this was the

at PENNSYLVANIA STATE UNIV on May 12, 2016phg.sagepub.comDownloaded from

388

nature of the allocative unit involved. Marginalist theory viewed the frm as a small.scale, single-plant, one person production unit which responded instantly to externalchanges which affected its internal allocation. In seeking a more realistic approach,however, the behaviourists argued that important .considerations, if industrialchange was to be understood, were the size and organization structure of theenterprise. It was .argued that all enterprises were not small-scale and that thecomplexity of the organization structure was an important factor. This led tothe separation of the ’single-plant’ and ’multiplant’ enterprise (Watts, 1974) whichwas to be used extensively in growth and rationalization studies.

b Goals and objectives: Most studies performed on various aspects of locationdecision.making have largely justified their questioning or rejection of the margin.alist orthodoxy on its profit maximizing goal assumption alone.

Dicken argued: ’

Normative theory, of course, postulates a single, clearly-defined goal of profit maximiz.ation, but there is now general agreement that no business organization possesses sucha single. unambiguous goal and also, following Simon ... , it is accepted that most organ-izations are boundedly rational satisficers rather than optimizers (Dicken, 1971, 427).

Therefore, the profit maximizing assumption was criticized as not generally applic-able. Dicken (1971, 427-28) argued that business goals were multidimensionalincluding not only profits but sales volume, market share, production costs, assets,prestige, growth and, importantly, survival. However, he also noted that the goalsof managers and shareholders may conflict where ownership and managementwere separate. Simon (1965, 122) had argued an organization’s goals were con-servation and growth.

Steed (1971b; 1971c) dealt with organization learning and adaptive behaviouras goals are reformulated on the basis of environmental perception. He suggested.that goal formation was a function of the external environment, the internal socialsystem, motives of the individuals in power and the bargaining process. Objectivescould be short or long term and involve adaptive behaviour to feedback, althoughthese could be distorted by perception and long-term uncertainty. Short-termobjectives were argued to be ’operational’; however, although profit was includedas a short-term objective it was not generally recognized as operational. Short-run objectives were partial and satisficing. Long-term objectives comprized com-pany strategy. Krumme (1969a; 1969b) dealt further with corporate goals in

relation to locational adaption, as considered below. ’

In criticizing the goal of profit maximization, to an extent, behavioural theorysuffered from the interpretation of other theories by the evaluative criteria internalto its own theory. In this way it viewed the profit maximization assumption oforthodox theory as a realistic goal of business. Part of Machlup’s (1967) defenceof marginalist theory was the argument that profit maximization was a simpleassumption of use in a theoretical construct (the theory of the firm) which had noobservable concrete meaning and was, by admission, unreal. The misunderstandingof profit maximization as an artificial construct led to behavioural researchers

at PENNSYLVANIA STATE UNIV on May 12, 2016phg.sagepub.comDownloaded from

389

(Cooper, 1975; Townroe, 1971) attempting to ’prove’ satisficing behaviour whenthe idea of proving it or profit maximization are both meaningless. One of theaims of these satisficing studies was to provide evidence against the idea that an

. industrialist would consider every possible location to find the proper profit-maximizing location - an unrealistic belief for a school of thought arguing realism!Also, in most cases, no questions are asked about how goals are formed. Simon(1965, 198) implied that goals could be exogenous to those of the capitalist systemof which the organization was a part, although this is questionable. It would bereasonable to suggest that limits exist to the formation of goals. Behavioural indust-rial location theorists gave little consideration to this.

Overall, behavioural studies attempted to produce empirical work on the goalsand objectives of an organization. Krumme’s advice that research should aim tounderstand the effect of an action and not the goal behind it would appear to bebetter directed.

2 The geography of enterprise: the Washington School

The behavioural industrial theoretical approach was first applied in a substantiveresearch context, true to the advocacy of behavioural theorists in economics,in studies produced as part of and subsequent to Thomas’s ’Growth pole andregional development’ project at the University of Washington in the late 1960sand early 1970s (Krumme, 1969a; 1969b; 1970; Steed, 1971a; 1971b; 1971c;1971d; 1971e; 1974; Erickson,, 1974; Le Heron, 1973; 1976). This involved aseries of papers which applied behavioural theory mainly to the analysis of largefirms in oligopolistic production structures and provided a solid foundation, interms of conceptual application and empirical work, for a behavioural theory nfindustrial georgaphy. The work was advanced in other contexts involving largefirms (e.g. Chapman, 1974; Rees, 1978; Watts, 1980) but the behavioural approachwas also extended to the analysis of firms in general.

Simon (1959, 278) had commented that the first applications of behaviouraltheory would be to those topics where the assumptions of static equilibrium wereleast tenable. The work stemming from the Washington School seems to havefollowed this closely in dealing with the problem areas of innovation, technologicalchange and economic development, all of which require a dynamic framework foranalysis. For example, as part of the analysis of the growth aspects of the regionaleconomy, Erickson (1972) developed the ’lead firm’ concept and Le Heron (1973;1976) looked at the effects of best practice technology, both requiring a non-staticframework.

The aim of the following sections is to introduce aspects of the ’geography ofenterprise’ included by and stemming from the Washington School. The structureof the following sections has been taken from Krumme (1969a) whose categoriesof investment decision-making, locational ’range of tolerance’ (i.e. the area of

profitability around the optimum location) and locational adjustment are used incategorizing contributions in the general field of the locational behaviour of firms.

at PENNSYLVANIA STATE UNIV on May 12, 2016phg.sagepub.comDownloaded from

390

Locational and investment decision-making will be considered together. Thecategory of the locational ’range of tolerance’ will involve a discussion of the

behavioral debate relating to Weber. As the forms of locational adjustment re-cognized by Krumme (1969a) are very limited (including very short-term, non-spatial adjustment) and have been drastically expanded in another paper (Krumme,1969b), the category of locational adjustment will be considered here with muchbroader terms of reference, directly encompassing locational adaption strategiesof varying scales and over varying time periods as were developed by the growing’geography of enterprise’ school. In addition, a section which compares behaviourand structure has been included, which relates mainly to Steed’s work.

a Investment and location decision-making: One important insight which wasprovided by the development of the behavioural approach, stemming from its

origins in management and administrative studies, was the recognition that thelocation decision could be understood more holistically as part of the generalinvestment decision-making of the enterprise. Consequent applications of be-havioural theory therefore involved relating location decision-making and invest-ment decision-making to varying degrees. However, some authors continued to viewthe location decision independently from other investment decisions. This divorceof investment and location decisions limited the understanding of industrial geo-graphy in that often only new plant location was considered and decision-makingpsychology in locating the new plant became an overresearched causal element(e.g. Stafford, 1974).

However, the limitations of a purely locational approach were recognizedearly on by Krumme (1969a, 32). He argued that every investment decision hasa locational aspect so they are interrelated. Krumme, however, saw the locationdecision to involve only aspects of new discrete investment including relocation,branch plant establishment and new location (although not additive investmentto existing locations - i.e. expansion - and not disinvestment - i.e. closure).Closure was only implied within the context of adoption/adaption in discussingthe ’range of tolerance&dquo;.

The strongest case for the integration of the investment decision and the locationdecision was presented by Townroe (1969; 1971; 1972) who used postulatescentral to behavioural theory such as goals, imperfect knowledge, etc. in an argu-ment akin to that of Krumme (1969a; 1969b) noted above. Townroe’s (1969)central focus involved the advocacy that the location decision must be understoodwithin the whole investment decision process and not just as the last decision in aseries. The location decision may not be the key decision within the whole invest-ment decision.making process and, in particular, the problem of where to locate

~. a manufacturing plant could well be marginal and arise rarely.Townroe defended the study of the location decision by arguing that its sig-

nificance is increased by widening the scope of the location decision question toinclude the whole investment decision. Decision-making would then include all

aspects of the location of new productive capacity within the whole investment

at PENNSYLVANIA STATE UNIV on May 12, 2016phg.sagepub.comDownloaded from

391

process, encompassing plant extension and productivity increase on existing floor-space, not just new plant location and plant relocation which in themselves wereconsidered to be relatively unimportant aspects of the changing distribution ofindustry.

However, Townroe’s defence of the study of the location decision would seemto require a research emphasis more upon the investment process itself. In a be-havioural context an answer to the question why a plant is located at one placerather than another would be of less consequence, if Townroe’s comments were tobe accepted, than the question of whether or not investment was to take place atall. Tle study of the financial processes affecting investment would be necessary,in particular, in the context of industrial decline where the location decision wasnot recognized. Townroe’s argument implied that productivity decline and plantclosure are not within the scope of the location decision question, are not

part of the firm’s general investment process and so are outside the realm of appli-cation of behavioural theory. This is at odds with theorists who specified industrialdecline (e.g. Steed). Therefore Townroe’s behavioural perspective would appearto be limited also.

In a further attempt to overcome the problem of relating the location decisionto general investment decision-making, Rees (1974) exemplified the limited pers-pective of behavioural industrial location theory in analysing the processes of firmgrowth, totally disregarding contraction and closure. An investment decision.

making framework was used in which growth was viewed as an accumulation oflocation decisions, particularly new plant location and adjustment at existinglocations, occurring even during a recession. This positioning of the study of thelocation decision within the context of firm growth added more specificity to therelationship between investment - decision-making and location decision-makingand was more closely focused than the cursory, non-specific links discussed byTownroe. However, the scope of a firm growth model is limited simply becauseall firms do not grow particularly during a recession.

Overall, most attempts to relate the investment and the location decision werepartial and saw the location decision to be the focus of attention. The gradualrecognition of this as being part of the investment decision gave a wider under-standing. Later studies attempted to focus more on the question of investmentdecision-making having consequences for location.

b Weber: Two schools of thought were developed by behavioural industriallocation theorists over the role of weberian theory in industrial geography. Oneschool of thought rejected Weber’s traditional position as the theoretical basisof industrial geography (Wood, 1969; Taylor, 1970). This was countered by asecond school which defended Weber arguing that behavioural industrial locationtheory could find starting points in existing theory (Bater and Walker, 1972;Dicken, 1977; Krumme, 1970; Smith, 1970).

Wood (1969) criticized the role of Weber as the theoretical basis of industrialgeography advocating what was essentially a linkage approach. A single factor

at PENNSYLVANIA STATE UNIV on May 12, 2016phg.sagepub.comDownloaded from

392

(transport cost) optimum location model was considered unsuitable as a generalapproach and to have only a small part to play in industrial geographical theory.This attack was based on Weber’s inability to explain the complexity of materialsand information linkage. Weber was relegated to a minor position within a hier-archy of four behavioural research directives where the locational variable wasconsidered to be of relevance. However, placing emphasis on the linkage structurealone is limited and ignores questions of the ownership and production structure.

Defence came, in particular, from Smith (1970) who argued weberian theory tobe a special case of a more general n-cornered locational figure which allowed atotal cost surface to be produced from the sum of individual input-cost surfaces.This could be related to a revenue surface to define ’spatial margins to profitability’within which suboptimal location decision-making could be accommodated.

The reformulated weberian model failed to take off as a basis for industrial

geographical theory. Consequently, most behavioural geographers viewed thelocation question as one aspect within a very broad behavioural framework whichwas particularly adaptable to case studies and classification. This was somethingto which Smith’s (1970) model was unsuited, being more applicable to aggregates.This proved to be more of an aside as far as industrial geography was concerned,providing weak resistance in the face of the behavioural onslaught. However,although the purely behavioural approach has taken precedence in industrial

geography, Smith’s aggregate approach survived as the basis of a more generalwelfare approach, which perhaps is an acknowledgement of the reformulatedweberian model’s unsuitability for industrial geography.

The behavioural debate of weberian theory, however, revealed a much moresignificant problem. This involved whether or not existing location theory wasto be rejected outright and totally replaced by behavioural concepts or whether ornot it had a place within the dominantly behavioural framework. Certain authorsargued that although weberian theory was not suitable as a basis for industrialgeography it should not be rejected out of hand but recognized for the partialcontribution that it could make. In general, traditional concepts were consideredto be essential but insufficient (Krumme, 1970, 318). In particular, Smith’s re-formulated model was considered to be relevant to branch plants only (Taylor,1970); this was developed in an argument which applied weberian theory to thelocation of branch plants by multinational companies (Dicken, 1977). Also,Hamilton (1974) argued that although orthodox theory was normative it stillhad application for industrial geography in nineteenth century, seedbed and thirdworld studies. So, it seems that although behavioural geographers deposed weberiantheory from its position as the basis of industrial geography, the theory was notrejected totally but limited to certain topics where its application was consideredjustified.

c Locational adaption: In discussing change in the production activities of firms,Alchian (1950) distinguished between adoptive and adaptive processes. Earlystudies on behavioural industrial location theory had included the adoptive element

at PENNSYLVANIA STATE UNIV on May 12, 2016phg.sagepub.comDownloaded from

393

only very generally, emphasis being placed upon adaption, presumably as this wasmore practicable to research. Very often this led to a situation where adoptiveforces went unrecognized. The following discussion will concentrate on earlystudies emphasizing the adaptive element. ’ .

The initial general framework for adaption was presented by Krumme (1969a;1969b) and Steed (1971a; 1971b; 1971c) and related adaption to the conditionsof the external economic environment. The emphasis, although including quali-tative organization change, was very much on change at the level of production anda shift to the analysis of elements of organization without their production processimplications came later (e.g. acquisitions in Leigh and North, 1978). Some laterstudies did relate organization change to the production process level (Smith,1979). However, nearly all of the later behavioural studies were to ignore thecrucial role of the external economic environment.

,

Adaptive behaviour by a firm was considered by Krumme to result in deviationfrom the optimum location. Consequently, the concept of the static optimum wasconsidered to have little analytical value. Two reasons were advanced to explainthis: non-linearity in the production function and the substitutability of inputsinto and outputs from the production process. These vary owing to indivisibilities,scale economies, market conditions, etc. which are themselves subject to change,especially with changes in technology.

The principle of substitution was also considered to be applicable to interfirnlrelationships outside the immediate input-output sphere of the individual firm’sproduction process. Within this only subcontracting, production and assemblycooperation and joint research and development (i.e. largely the sharing of infor-mation) were included. It was noted that more opportunities were likely to beavailable for Iarger firms and that adaption would vary according to the size ofthe organization. This aspect has been extended and broadened considerably insubsequent work by Rees (1978) and Watts (1978; 1980).Krumme (1969b, 16-17) presented a classification of locational adaption

based on three criteria: space (whether adjustments are spatial and involve re-location or branch plant location; or are non-spatial and involve changes in product,product mix, raw materials, production stages, technology, etc.); organization(varying the previously constant form of the economic unit by including decision-making at the level of manager, legal structure or power structure of the enterprise.In the classification itself this seems to involve a simple distinction between singleand multiplant firms); and tirrte (short-run, medium-run, long-run).

. Krumme suggested that a variety of possible adaptions to changes in environ-mental conditions can occur within different organization structures over time.These are not limited to locational adjustments but can include adjustments at apoint in space as well as qualitative and quantitative shifts in space. The followingsimple groups resulted: .

i) The single-plant firm can adapt in the short run in the form of the intensityof use of investment goods and changes in input factor combinations as longas new investment is not necessary. Short-run spatial adaption is allowed to a

at PENNSYLVANIA STATE UNIV on May 12, 2016phg.sagepub.comDownloaded from

394

more radical extent in the multiplant firm having a wider range of alternativesavailable to it than the single-plant firm;

ii) New or abandoned investment, which does not involve the addition or removalof totally independent production units, is a feature of medium-run adaption.Again, the scope of adaption is less for the single-plant firm. Medium-run adap-tion can include agreements and cooperation between firms which may involveproduction integration, subcontracting, research and development collaboration;

iii)Long-run adaption involves locational adjustment at the production level in-’

eluding disinvestment, relocation, branch plant location as well as ownershipchange (including merger and takeover) or new partners. The scope of adaptionis wider for the larger-scale, multiplant firm planning over the longer term.New discrete investment, such as relocation and branch plant location, was

considered to be a form of long-run adaption and this appears to have been thefocus of British studies with, in particular, the decision-making considerations oflocation adaption being an important focus of analysis (e.g. Cooper, 1975; Townroe,1969; 1971; 1972). Short-run adaption involving linkages following locationalchange is discussed by Moseley and Townroe (1973).

Steed (1971b) illustrates and develops aspects of Krumme’s classification onthe theme’ of locational adaption, particularly long-run adaption. He argues thatfirm survival is only possible with continuous adaption, which is not necessarilyspatial, in response to changes in the external environment. An important com-ponent of adaption is the management decision-making process which involvesfour interrelated concepts: quasi-resolution of conflicts; uncertainty avoidance;problemistic search; and organization learning. An organization’s ability to adaptdepends partly upon management’s perceptual and information processing capacity(i.e, its ability to learn and restructure its goals from information feedback fromthe environment) and its ability to impose corrective decisions. The latter is partlya function of the diversity and flexibility of internal resources influencing theextent to which adaption can take place.

Steed goes on to argue that the nature of the external environment will influencethe time span of adaption, medium-run and longer-run strategies becoming morenecessary the more turbulent the external environment. This requires direct knowl-edge by management decision-makers of their environment and the ability toimplement strategic, longer-run decisions which can change the external environ-ment. Four types of environment are distinguished: ’placid, randomized’ (wherestrategy and tactics are not distinct); ’placid, clustered’ (where survival becomeslinked to a separate strategy involving knowledge of the environment); ’disturbed-reactive’ ; and ’turbulent’.

In the category of ’turbulent’ environment the short-run adaptive responseof a firm is inadequate to ensure survival owing to the complexity of interactionwithin the environment and the increased organizational dependence upon environ-ment. This means that organizational change is increasingly externally induced.The implication is that survival would be more likely for the larger firm havinglonger-run and broader adaptive strategies. As a consequence, the larger firm and

at PENNSYLVANIA STATE UNIV on May 12, 2016phg.sagepub.comDownloaded from

395

the accelerated trend towards the larger firm become characteristics of the ’dis.turbed-reactive’ and ’turbulent’ types of environment. The corollary is that firmfailure is a consequence of the overemphasis on short-run adaption which provesinadequate. However, Cyert and March (1963) considered short-run adaption to beimportant for a firm which is unable to influence an unstable environment.

The perception and ability of management in charting a suitable path of changefor the firm is, therefore, given an important position in Steed’s interpretation oflocational adaption. This is particularly relevant to the larger, multiproduct firm,possibly a multinational or a company operating in monopolistic or oligopolisticconditions, involved in medium-run and long-run strategy formulation in turbulentenvironments rapidly changing in complexity in which short-run adaptive responsesare likely to be inadequate for survival.

The studies of Krumme and Steed on the role of the management decision-making process in locational adaption probably went furthest in the effectivedisplacement of normative theories as the basis of industrial geography, revealingsuboptimal location to be the consequence of adaption.

d Behaviour and structure: The early work developed at the University ofWashington attempted to place the enterprise within a broad framework applicableto regional industrial change generally. This framework included much more thansimply the behaviour of the enterprise as the explanatory base. For example,Steed’s declared aim was to ’... reveal the interrelationships between factorsdiscerned to account for structural change’ (Steed, 1974, 380) by linking structure,behaviour and performance. These were broad aims and it is evident that behaviourwas acknowledged as one aspect out of several which needed to be consideredto explain regional industrial change. The behaviour of the enterprise gave a newinsight, but only when linked with sectoral structure and economic performance.The dominant role of enterprise behaviour alone came in later studies. It must benoted, however, that structure, behaviour and performance cannot be acceptedas variables providing a complete explanation of regional industrial change.

The exemplification of the above aims was presented in work on the NorthernIreland linen complex (Steed, 1971a). It was argued that spatial retrenchment ofthe linen industry involved continuing organizational integration with concentrationand centralization processes allowing rationalization. Changes in the ownershipand production structures within a ’turbulent’ economic environment providedthe framework in which retrenchment took place and therefore its form. The

importance of retrenchment processes and the means by which retrenchmentoccurred - merger; acquisition and closure; independent firm closure variedaccording to the sectoral ownership and production structure. Consequently,behaviour and structure can be considered to be the two main variables to emerge.

Steed’s (1971a) work on the Northern Ireland linen complex appeared toinvolve a wider range of processes than those recognized in industrial locationtheory as it stood. The analysis advocated more than simple internal decision-making in that behaviour was analysed along with structure. However, Steed’s

at PENNSYLVANIA STATE UNIV on May 12, 2016phg.sagepub.comDownloaded from

396

(1971a) practical work suggests to this author the much greater significance ofstructure than of behaviour in influencing forms of change by sector. This conflictswith theoretical papers which emphasize the behavioural element. Therefore, itis all very well advocating the behaviour of enterprise as the causative element inindustrial change, but this does not recognize the permissive or limiting influencesof structure and changes in structure independent of behaviour. The importance ofsectoral structure is particularly well shown up in oligopolistic situations where thefeasibility of decision-making in the formation of, say, a cartel is dependent uponthe degree of centralization/concentration in a sector which influences whether ornot that form of organization can be initiated and maintained. Behaviour is, con.sequently, a conditional variable.

,

Therefore, although the analysis of industrial geographic change is placed in awider context Steed allocates cause to the management of the company to an

unjustifiable degree, overemphasizing endogenous corporate adaption. This hasalways been a problem for behavioural theorists. Steed relies heavily upon theworks done in the process school of management studies and although admittingtheir overconcern with management often comes down to the management variablein the planning, organization, staffing, directing, controlling functions as a causativeelement in his own work (e.g. Steed, 1971c, 93). This is exemplified when a firm’sability to survive is deemed dependent partly upon management’s ability to inter-pret a changing environment and implement adaption (Steed,1971b, 324) and wheremanagement inefficiency is seen as a significant contributory factor in the declineof an industry (e.g. Steed and Thomas, 1971, 345).

The behaviour -of managerinent is perhaps a variable worth noting, but in theview of this author it is not likely to be attributable as the cause of growth ordecline of an industry when questions of competition, labour costs, productivity,etc. are likely to be more influential. The tendency to focus on managementvariables also shows a lack of recognition of the crucial factors of liquidity,solvency, profitability and capital availability. These are not necessarily dependentupon management’s ability, being more relevant to other variables, for example,company ownership. These factors are particularly inhibiting to change in a declin-ing sector requiring rationalization with problems of illiquidity, insolvency, lowprofitability and limited capital backing. One consequence of the overemphasison behavioural variables is that industrial change is now seen as a direct functionof the decision-making choices of managers and entrepreneurs (e.g. Leigh andNorth, 1978) and the understariding of their choices has become an objective ofresearch. The choices of management have consequently become the causes ofindustrial change. Dicken (1977, 434) has extended this to view the whole economiclandscape as being made up of countless decisions each made within that segmentof the environment on which information is received, resulting in purposive be-haviour. Environmental and structural influences pall into insignificance as a

consequence. -

at PENNSYLVANIA STATE UNIV on May 12, 2016phg.sagepub.comDownloaded from

397

II Conclusions

The Washington School of behavioural industrial location theorists proved to be avery important element in the initiation and development of a theoretical basewhich was critical of traditional methods of approach to the analysis of industrialgeography in the late 1960s and early 1970s. Quite justifiably, they realized seriousproblems in the ability of classical and neoclassical theories to explain change inindustrial geography. Traditional theories had not recognized the crucial role ofbehavioural variables, such as decision-making, operating within the’ dynamicorganization structure of the enterprise. As a consequence, the direction of sub-sequent industrial geographical research shifted towards attempting to understandthe function and interaction of behavioural variables within the enterprise.

However, it is the impression of this author that behavioural industrial locationtheorists have lost sight of the wider goals of industrial geographical research inthat internal decision-making and the enterprise have become very much morethan the objects of analysis. The problem appears to be one of emphasis in excessof that attributed to decision-making and the enterprise by the early researchersof the Washington School. Although the important contribution to the under-standing of industrial change that decision-making and the enterprise has hadcannot be denied, behavioural researchers have tended to overemphasize theirrole. Steed’s goal of revealing the interrelationships between structure, behaviourand performance has been achieved, if it has been achieved at all, with an excessiveemphasis on the behavioural variable. This emphasis has increased in more recentstudies, particularly in the UK. ~

,

One consequence of this emphasis has been the relegation of the cause of indus-trial change to ’decision-making’ which is as unacceptable as ’economic man’ isin traditional theory. It is one thing to analyse decision-making to improve theunderstanding of industrial change but quite another to argue that decision.makingis the cause of change. The latter is in fact what behavioural theory has led to.

Although behavioural industrial location theory has provided important insightsinto certain processes of change; the overemphasis of behavioural variables hasproved a serious limitation to the further advancement of understanding in indus-trial geography. This is because behavioural industrial location theory itself doesnot incorporate concepts which relate the enterprise, and decision-making withinit, to change in the economic environment. The economic environment had beenrecognized in the early stages of behavioural industrial geography; however, inlater studies it was not recognized analytically at all. The fact that decision-makingitself may be understood more fully if analysed as an aspect of a theoretical frame-work which incorporates change in the economic environment has been over-looked.

Decision-making is now viewed, unacceptably, as the initiating cause of industrialchange and this has led directly to the study of locational &dquo;adaption’. The processesrelevant to ’adoption’ involving pressures exogenous to the enterprise have

largely gone unrecognized. The development of research on aspects of industrial

at PENNSYLVANIA STATE UNIV on May 12, 2016phg.sagepub.comDownloaded from

398

geographical change where behavioural variables, particularly decision-making, areinsignificant and variables relevant to the pressures of the economic environmentare dominant is essential if industrial geography is to reestablish its credibility.Relevant hypotheses in this direction are more likely to be defined within thesphere of industrial decline, particularly in periods of recession. This will encourageindustrial geographers to consider other theories and move away from their currentblinkered stance.

The inclusion of variables relevant to the economic environment as well asbehaviour is a development which a political economic approach is likely to allow.A political economic approach has been discussed in behavioural terms by Thorelli(1965) in an attempt to incorporate internal and external interest groups within asingle - framework. This has been touched on in industrial geography with the

recognition of the need to study ’interfirm relations’ and their empirical appli-cation (Watts, 1978; 1980). However, in the view of this author, a behaviouralpolitical economic approach to industrial geography will always be unacceptable.The point remains that a behavioural approach of whatever nature will be un-suitable where behavioural variables are insignificant in the explanation of indus-trial geographical change.

University of the Witwatersrand, Johannesburg, South Africa

A ckno wledgem ents

The author would like to .thank Professor R.J. Johnston and Dr H.D. Watts, De-partment of Geography, University of Sheffield for their advice and commentsduring the preparation of this paper.

The research on which this paper is based was supported by the Social ScienceResearch Council.

III References

Alchian, A.A. 1950: Uncertainty, evolution and economic theory. Journal ofPolitical Economy 58, 211-21.

Bater, J.H. and Walker, D.B: 1972: Further comments on industrial location andlinkage. Area 2, 59-63.

Chapman, K. 1974: Corporate systems in the United Kingdom petrochemicalindustry. Annals of the Association of American Geographers 64, 126-37.

Cooper, M.J.M. 1975: The industrial location decision-making process. Centre forUrban and Regional Studies, University of Birmingham Occasional PaperNo.34.

Cyert, R.M. and March, J.G. 1963: Behavioural theory of the firm. EngelwoodCliffs, NJ: Prentice-Hall.

Dicken, P. 1971: Some aspects of decision-making behaviour of business organiz-ations. Economic Geography 47, 426-37.

at PENNSYLVANIA STATE UNIV on May 12, 2016phg.sagepub.comDownloaded from

399

1976: The multiplant business enterprise and geographical space: some issues onthe study of external control and regional development. Regional Studies 10,401-42.

1977: A note on location theory and the large business enterprise. Area 9,138-45.

Erickson, R.A. 1972: The lead firm’ concept. Tijdschrift voor Ecotioinische enSociale Geographie 63,426-37.1974. The regional impact of growth firms: the case of Boeing, 1963-68.Land Economics 50, 127-36.

Hamilton, F.E.I. 1974: A view of spatial behaviour, industrial organizations anddecision-making. In Hamilton, F.E.I., editor, Spatial perspectives on industrialorganization and decision-making, Chichester: John Wiley.

Krumme, G. 1969a: Towards a geography of enterprise. Economic Geography 45,30-40.

1969b: Notes on locational adjustment patterns in industrial geography. Geo-grafiska Annaler 51B, 15-19.1970: The interregional corporation and the region: a case study of Siemensgrowth characteristics and the response pattern in Munich, West Germany.Tijdschrift voor Economische en Sociale Geographie 61, 318-33.

Le Heron, R.B. 1973: Best-practice technology, technical leadership and regionaleconomic development. Environment and Planning 5, 735-49.

1976: Best-practice firms and productivity changes in the Pacific Northwestplywood and veneer industry, 1960-72: some regional implications. Environ-ment and Planning 8, 163-72.

Leigh, R. and North, D.J. 1978: Regional aspects of acquisition activity in Britishmanufacturing industry. Regional Studies 12, 227-45.

Machlup, F. 1967: Theories of the firm: marginalist, behavioural, managerial.American Economic Review 57, 1-3 3.

Massey, D. 1974a: Towards a critique of industrial location theory. Centre forEnvironmental Studies Research Paper 5, 1-27.

1974b: Is the behavioural approach really an alternative? In Massey, D. andMorrison, W.I., editors, Industrial location: alterrtative frameworks, Centrefor Environmental Studies, Conference Paper 15, 79-86.

1975a: Approaches to industrial location theory: a possible spatial framework.In Cripps, E.L., editor, Regional science - new concepts and old problems,London: Pion.

1975b: Behavioural research. Area 7, 201-203.1979: A critical evaluation of industrial location theory. In Hamilton, F.E.I.and Linge, G.D.R., editors, Spatial analysis, industry and the industrial environ-ment. Volume 1 Industrial systems, Chichester : John Wiley, 57-72.

Massey, D.B. and Meegan, R.A. 1976: The inner city and the international com-petitiveness of British industry: the employment implications of the IndustrialReorganization Corporation. Centre for Environmental Studies, Working Note437, 1-35.

1978: Industrial reorganization versus the cities. Urban Studies 15, 273-88.1979: The geography of industrial reorganization: the spatial effects of therestructuring of the electrical engineering sector under the Industrial Reorgan-ization Corporation. Progress in Planning 10, 155-237.

at PENNSYLVANIA STATE UNIV on May 12, 2016phg.sagepub.comDownloaded from

400

Moseley, M.J. and Townroe, P.M. 1973: Linkage adjustment following industrialmovement. Tijdschrift voor Economische en Sociale Geographie 64, 137-44.

Pred, A. 1967: Behaviour and location. Lund Studies in Geography, Series B,Human Geography 27 and 28, Lund: Gleerup.

Rees, J. 1974: Decision-making, the growth of the firm and the business envirori-ment. In Hamilton, F.E.I., editor, Spatial perspectives on industrial organ-ization and decision -making, Chichester: John Wiley, 189-211.

1978: On the spatial spread and oligopolistic behaviour of large rubber com-panies. Geoforum 9, 319-30.

Simon, H.A. 1959: Theories of decision-making in economics and behaviouralscience. American Economic Review 49, 253-83.

1965: Administrative behaviour. New York: Free Press.

Smith, D.M. 1970: On throwing out Weber with the bathwater: a note on industriallocation and linkage. Area 2,15-18.

Smith, I.J. 1979: The effect of external takeovers on manufacturing employmentchange in the northern region between 1963 and 1973. Regional Studies 13,421-37.

Stafford, H.A. 1974: The anatomy of the location decision: content analysis andcase studies. In Hamilton, F.E.I., editor, Spatial perspectives on industrialorganization and decision-making, Chichester: John Wiley, 1969-88.

Steed, G.P.F. 1971a: Internal organization, firm integration and locational change:the Northern Ireland linen complex; 1954-64. Economic Geography 47,371-83.

1971b: Plant adaptation, firm environments and locational analysis. The Pro-fessional Geographer 23, 324-28. 1971c: Forms of corporate-environmental adaptation. Tijdschrift voor

Economische en Sociale Geographie 62, 90-94.1971d: Changing processes of corporate-environmental relations. Area 3,207-211.

1971e: Locational implications of the corporate organization of industry.Canadian Geographer XV, 54-57.

1974: The Northern Ireland linen complex: 1950-70. Annals of the Associationof American Geographers 64, 397-408.

Steed, G.P.F. and Thomas, M.D. 1971: Regional industrial change: NorthernIreland. Annals of the Association of American Geographers 61, 344-60.

Taylor, M.J. 1970: Location decisions of small firms. Area 2, 51-54.Thorelli, H.B. 1965: The political economy of the firm: basis for a new theory of

competition? Schweitz. Zeitschr. Volkwirtschaft und Stat. 101, 248-62.Townroe, P. 1969: Locational choice and the individual firm. Regional Studies

3, 15-24. 1971: Industrial location decisions. Centre for Urban and Regional Studies,University of Birmingham, Occasional Paper 15.

1972: Some behavioural considerations in the industrial location decision.

Regional Studies 6, 261-72.Watts, H.D. 1974: Spatial rationalization in multiplant enterprises. Geoforum

17, 69-76.1978: Interorganizational relations and the location of industry. RegionalStudies 12, 215-25.

at PENNSYLVANIA STATE UNIV on May 12, 2016phg.sagepub.comDownloaded from

401

1980: Conflict and collusion in the British sugar industry, 1924-28. Journalof Historical Geography 6, 291-314.

Wood, P.A. 1969: Industrial location and linkage. Area 1, 32-39.

at PENNSYLVANIA STATE UNIV on May 12, 2016phg.sagepub.comDownloaded from