-----------------------------------------------------------------------------------------------

Preface

A dissertation submitted to the University of Manchester for the degree of MA (Econ) in the Faculty of Economic and Social Studies, department of Economic Studies, 1997.

I declare that no portion of the work referred to in the dissertation has been submitted in support of an application for another degree or qualification of this or any other university or other institution of learning.

---------------------------------------------------------------------------------------------------

---------------------------------------------------------------------------------------------------

- Contents -

Acknowledgements

Chapter 1. Introduction

Chapter 2. History of the growth pole concept

The 1st wave : Perroux's world and the birth of "growth poles"

The 2nd wave: geographical applications

The 3rd wave : growth poles as a cause of regional inequality

The 4th wave ? : revisionism

General characteristics of the Growth Pole Strategy

National Economic Growth

Regional Economic Growth

Inter-regional Equality

Rural Development

Urban system and Migration control

Variety of the Growth Pole Strategy

Difficulties of Evaluation

Developed countries

Developing countries

----------------------------------------------------------------------------------------------------

Acknowledgements

I am very thankful to the great supervising of Prof. Nixson for this dissertation. I was inspired on deciding the subject of my dissertation since I have attended at his interesting lecture about regional development in the University of Manchester, and his useful comments and encouragement helped me to complete this work. My gratitude also goes to Prof. Parr in the University of Glasgow. His latest writing on the issue of growth pole strategies has been a base of this dissertation. I also have to mention Paul Lancaster in checking my English. He even gave some helpful comments on my draft. I am also grateful to Mr. John Morley for his lasting encouragement to improve my English writing skill. I met a lot of good people in Manchester and they certainly deserved to be given my special thanks: Tomiharu, Yuichi, Satoshi, Naoto, Maimuna, Sophie, Alison, Sardo, Carmen, Amparo, Annalisa, Richard, Fong, Ann-Pei, Brian, Lewis, Hajime, Yoko, Chiho, Midori, Pei-in, Stella-Maria, Ms. Jane Elliot, Dr. Dale, Dr. Joseph, Dr. Walter, Andrew (Yan-san), Alex, Junko and all students in the presessional Englsih school in 1997.

--------------------------------------------------------------------------------------------------

Chapter 1: Introduction

The concept of a growth pole was developed by Perroux [1955] and it indicated that development has to be brought about by a certain concentration (agglomeration) of economic activities in an abstract space. Although Perroux had considered it as an economic phenomenon and it had a complex system, the concept of a growth pole had become "an idea in good currency" and it enjoyed the privilege of "all mythic catchwords" in regional development planning in the 1960s [Lasuen, 1969, p20]. A lot of articles had been published about the growth pole. For instance, in 1969, Darwent had already noted 155 references relating to the growth pole and there are 296 references listed by Moseley [1974]. Indeed, by the early 1970s, "reliance on growth pole analysis was a dominant characteristic of operational regional planning in both developed and developing countries" [Richardson and Richardson, 1974, p163].

While it raised academic interests, practical strategies based on the growth pole concepts, named "growth pole strategies", had been intensely considered and implemented in developed and developing countries in the 1960s. By the late 1970s, the growth pole strategies had been "implemented, provided for, or seriously discussed" in at least 28 developed and developing countries. They were Austria, Belgium, Bolivia, Brazil, Bulgaria, Canada, Chile, Colombia, Cuba, France, Ghana, Great Britain, India, Ireland, Italy, Kenya, Libya, Malaysia, Nigeria, Peru, Poland, Russia, Spain, Sweden, Tanzania, the United States, Venezuela, and Yugoslavia [Gaile, 1978, in Polenske, 1988, p101].

However, most such strategies have afterwards been thought to have "failed". Parr [1996, p2] argued "the recent history of regional economic planning in many parts of the world is littered with examples of growth pole strategies having failed or having been prematurely abandoned".

The governments turned their back on the growth pole strategies [Conroy, 1973], and "they pass away, undestroyed, but tarnished by their inefficiency" [Lasuen, 1969, p21]. Some regional economists, such as Gore [1984] and Lipton [1977], have argued that growth pole strategies depend heavily on several favourable assumptions, and that they have been suspicious of the validity of a situation where the growth pole strategies have worked properly. Moreover, the strategies even "have aroused a suspicion, ...., on ideological grounds" [Parr, 1996, p2].

The growth pole strategies seem to be abandoned and ignored by current academic interests, too. There is no vivid discussion relating to growth poles even as an economic phenomenon. It seems that the growth pole concept is useless or inappropriate in reality. However, most of the previous studies and reviews of growth poles have some problems as follows.

Firstly, the definitions of growth poles in those studies were quite diverse such as;

"A spatial agglomeration of related industries" : McCrone [1969]

"A spatial agglomeration of related industries which contains a growing propulsive industry" : Lasuen [1974]

"A spatial agglomeration of related industries, located in an urban centre, which, through their expansion, induce growth in its surrounding hinterland" : Boudeville [1966]

"A growing urban centre inducing growth in its surrounding hinterland" : Nichols [1969]

"A growing urban centre" : Parr [1973]

Gore argued that such a transformation of the concept of growth poles results in "what is described as 'growth pole theory' is characterized by considerable definitional confusion.... This confusion is reproduced in planning practice..." [Gore, 1984, p89]. With such a confused definition, it seems to be impossible to judge the validity of the growth pole theory.

Secondly, while the discussion of growth poles has been confused by the various definitions, the argument has been quite far from the original characteristics of Perroux's intention. In comments on the existing reviews and studies of growth poles, Polenske argued "few analysts provide the historical context for Perroux's work" [Polenske, 1988, p91]. A small number of studies dealt with a contradiction of abstract space and geographical space, and most of the discussions have been made only with the geographical space. Therefore the theoretical base of all growth pole strategies is "ill-specified" [Gore, 1984, p92] and it has prevented constructive discussion.

Thirdly, there has not been enough discussion on the objectives of the strategies, their presumed effects and their needed programme set. The growth pole strategies cannot be generalised as only one programme. For example, if it is targeted to provoke the development of a depressed region, the growth pole must be located in the depressed region rather than other wealthy regions, although it raises a further question how to set the growth poles (investment or infrastructure), whether the growth poles must be set in a depressed area or wealthy area in the depressed region, and so on. In this case, the effect should not be judged by only one measurement, such as household income. Stohr and Todtling [1979] argued that the measurement of spatial equity in the existing studies is too materialistic. They ignored some social aspects such as family and local solidarity, self-realization, access to the central decision making power, and so on. Unless making clear policy objectives and programme settings for each case, it does not seem that the growth pole strategies have been fairly judged in practice.

Has the growth pole concept died? No, the growth pole has existed despite the academic neglect. Perroux [1988] replied to the criticisms of the validity of the concept of growth poles with a "straightforward" question : "Do you know of an example, anywhere on earth, of growth and development where these processes have taken place without the presence and effects of development centers, whether territorialized or not?" [Perroux, 1988, p66]. While Richardson and Richardson admitted some negative effects of growth pole strategies, they argued that "it is difficult to devise a strategy which did not include an element of polarization" [Richardson and Richardson, 1974, p175]. Regional inequality exists all over the world, and no development policy has the freedom from problems of spatial selectivity. If the reality has been to face the regional problems such as regional inequality and spatial selectivity, the growth pole concept seems to have been abandoned in a quite premature condition. Indeed, as Parr admitted, "there has never existed an adequate framework or general theory of regional economic policy within which the growth pole strategy could be evaluated, either conceptually or operationally" [Parr, 1996, p3].

This dissertation will consider the validity of the growth pole strategies from a comprehensive point of view. First, it examines the historical change of the concept from Perroux to its critics and it will make a clear grasp of the concept. Second, the practical policy issues will be discussed in terms of various objectives, and it will show widely varying (or sometimes contradictory) characteristics of the strategy. Third, it will review written empirical studies considered with the strategies and it will examine the conditions of the successes and failures of the strategies for each objective.

Chapter 2: History of the growth pole concept

The 1st wave: Perroux's world and the birth of "growth poles"

The concept of growth poles has been developed by Perroux. Even though a lot of studies about growth pole strategies began their analysis from Perroux's work on growth poles, few actually attempted to fully understand this theory, since "the theory was too complex, too abstract, and too non operational" [Higgins, 1988, p44]. It seems to be necessary to understand the concept more comprehensively by reviewing Perroux's philosophical framework.

Perroux criticised traditional analytical frameworks of contemporary economics in his early work [Perroux, 1950]. Keynesian and neo-classical economists treat the economy as a set of aggregate variables such as capital, savings, investment, consumption and so on. In the 1950s, the main-stream of growth theory was this functional analysis as represented by the Harrod-Domar model and it had assumed the equilibrium path of growth. Perroux criticised such an analytical framework - "treat the nation as a local complex of factors of production, of which the contents are determined by the relative supplies of these factors contributed by various nations" and in this framework all individuals are "contained in a container" as a "big individual" [Perroux 1950, p33]. Rather, Perroux presumed there to be a more dynamic mechanism in economic growth. Higgins [1988] provided an interesting analysis of Perroux's social philosophical background by referring to Perroux's recent article "Au-Delà du Welfare State" (Beyond the welfare state - Perroux, 1984). In this article, according to Higgins, Perroux pointed out that all humankind was born as a member of a society or a group, and "the Classical school, from Adam Smith to John Stuart Mill, understood very well that society is composed of groups with conflicting interests" [Higgins, 1988, p35]. Human beings are not "robots of price" as neo-classical economics assumed and Perroux rejected any concept of a basic harmony of interests in society. As Perroux argued in his recent writing, "each individual carries energy for change, usually in the form of expansion energy" [Perroux, 1988, p51]. This human tendency constructs some oligopolistic "Leading Actors". Thus, in Perroux's thought, "economies and societies are pulled and pushed in various directions by ever-changing constellations of Leading Actors, with constant groupings and regroupings through (irreversible) time, so that the distribution of power is constantly changing too" [Higgins, 1988, p36-37]. In his article in 1965, Perroux argued that economics is no longer a study of the network of 'exchange', rather it is a network of 'force' or power, and he called it "the theory of economic domination" [Perroux, 1955]. This dynamism and an un-harmonized system of conflicts among various groups in the economy seems to be the philosophical basis of Perroux's analytical framework.

Perroux argued that 'observable' economic growth is not 'smooth nor regular', and he associated it with structural change - "the appearance and disappearance of industries... the varying proportion of various industries in total output in the course of successive periods.... the different rates of growth for diffusion of the growth of an industry" [Perroux, 1955, p56 translated in Livingstone, 1981]. This image of a divergent (or unbalanced) nature of growth leads to his famous concept of 'growth poles'. Actually, Perroux had already mentioned the term "pole" and "centre" in his early work in 1950. Perroux had already written that establishments of each firm are "geographically dispersed", and "bounds of organization of varying strength are formed among them" [Perroux, 1950, p26]. Perroux argued that "as a field of forces, economic space consists of centres (or poles or foci) from which centrifugal forces emanate and to which centripetal forces are attracted" [Perroux, 1950, p27]. As a reason for such agglomeration, Perroux argued that dominant (leading) firms are comparatively efficient, and they can achieve an effective use of innovations and thus increase output more greatly than others. The extent of this effect would spread and would be received by everybody in society through a multiplier effect. Therefore, a certain level of polarisation must be a necessary bait to benefit the masses [Perroux, 1950].

This point of view is quite similar to Schumpeter's view. Schumpeter wrote : "Progress - in industrial as well as any other sector of social and cultural life - not only proceeds by jerks and rushes but also by one-sided rushes productive of consequences other than those which ensue in the case of co-ordinated rushes" [Schumpeter, 1939, p79]. In his view, capitalism is a routine of "creative destruction". A new industry (sector) which has brought a new technology with a new innovation breaks all old industries (sectors) and would form a certain domination. Schumpeter argued that such expansion of a sector could create multiplier effects to other sectors through Leontieff type input-output relationships. Thus, Schumpeter also admitted that monopoly was a necessary bait to attract new innovation. Monopoly is even good since "sensitive" and "delicate" innovative entrepreneurs need to be nurtured. Innovation gives a certain benefit to "a cluster of followers", so that others can reach up with an expansion of the technology.

Higgins [1988] argued "Perroux took over the whole of the Schumpeterian system and put it into space." Both of them believed that progress was due to innovation, and also Perroux's "propulsive industries", "polarization" and "propulsive effects" can be attributed directly to Schumpeter's "entrepreneurs", "monopoly", and "a cluster of followers".

From this context, Perroux [1955, p56 translated in Livingstone, 1981] has written the following famous sentence : "the fact, rough but solid, is this : growth does not appear everywhere at the same time; it manifests itself in points or 'poles' of growth, with variable intensities; it spreads by different channels with variable terminal effects for the economy as a whole." However, as previously discussed, the growth poles are not individual phenomenon, rather Perroux focused on the 'relationship' of propulsive units, and stated that the growth pole is "a propulsive unit coupled with the surrounding environment" [Perroux, 1968, p247, in Gore, 1984]. Since this concept has been frequently misunderstood as simple agglomeration effects in an economy, he had to restate it as "I went from the concept of 'growth pole' to that of 'pole of development'..... the growth pole is a set that has the capacity to induce the growth of another set ("growth" being defined as a lasting increase in the dimensional indicator); the pole of development is a set that has the capacity to engender a dialectic of economic and social structures whose effect is to increase the complexity of the whole and to expand its multidimensional return" [Perroux, 1988, p49]. As seen in this writing, it seemed that Perroux presumed a highly complex system with an interrelation of various factors in his analytical framework. In this sense, the growth pole concept should have been treated more carefully. At least, it is not suitable to be examined only as a geographical agglomeration effect on the economy within regional planning. As discussed in the next section, however, the academic interest with the growth pole has been raised only in this limited geographical meaning, and it has resulted in a wrong direction of the development of this strategy.

The 2nd wave: geographical applications

In the 1960s, a lot of articles about growth poles were published, and all of them treated this concept with enthusiasm and aggressively applied it in regional planning. Most notably we can find such intensive application of growth pole strategies from Boudeville's and Hirschman's work. Indeed, Gore [1984] mentioned their names with a challenging title "two false starts". This section will examine growth pole strategies developed by Boudeville, Hirschman, and other economists, and consider their contribution and faults.

Boudeville

Boudeville is one of the most eminent Francophone economists who applied the growth pole concept in regional planning. According to his definition, the growth pole is "a set of expanding industries located in an urban area and inducing further development of economic activity throughout its zone of influence" [Boudeville, 1966, p11]. He reached this concept from his typology of regions: homogeneous regions, polarised regions, and planned regions. About the second definition, he noted "the region can be studied from the point of view of its more or less pronounced degree of coherence, that is to say, according to the interdependence of its diverse parts; it is more or less polarized" [Boudeville, 1961, quoted in Hansen, 1971, p14]. Later, he developed the growth pole concept into an 'operational planning model" which explained a condition in which growth would be created in 'polarized regions'. He defined three types of polarisation : national, regional and local, but polarised centres at all levels must be the same because of the goods circulation throughout the country. As the reason for such polarisation, he pointed out that exchange (trade) would be done more efficiently in "the regional metropolis" than other places in the nation. Although Boudeville noted that it is an endogenous economic phenomenon, economic growth must be brought about also by exogenous policies which "plan with the greatest possible efficiency the development of growth poles through the mechanism of their propulsive industries" [Boudeville, 1966, p112]. In this theoretical progress by Boudeville, the growth pole theory was turned into a practical concept for regional planning.

Boudeville expected there to be "Leontieff-type multiplier effects" (upstream and downstream effects) and "polarization effects" in propulsive industries, he also argued that it encouraged further "efficient" development and it resulted in benefits for all, with the same logic as Perroux. Although Boudeville's theory explained the impact of the growth pole in geographical space, "it is not by itself a theory of location explaining where the functional growth poles are or where they in the future will be localized in geographical space" [Hermansen, 1972, p179].

Obviously, the propulsive industries and the growth pole had been brought from Perroux's previous work, however his new attempt was to assume that this natural economic phenomenon could be used within geographical regional planning. Coragio [1974] called this "the pure strategy of polarized development" since it tried to pursue the building up of growth poles, which had been presumed as a natural phenomenon by Perroux, in a geographical space through practical regional planning and be developed as a "growth pole strategy". In order to construct a practical regional planning model with the growth pole concept, Boudeville has to limit space into a purely geographical space, while Perroux's first attempt considered it in an abstract topological space. As Lasuen [1969, p21] noted, Perroux conceived that "geography had been cast as a passive rigid container that conditions the dynamic evolution of the economic forces, and that this was only a partial, limiting, and dangerous perspective."

Hirschman

Hirschman has also attempted to conceptualise the growth pole theory as a regional planning strategy. In his eminent book "The Strategy of Economic Development" [Hirschman, 1958], he has advocated an "unbalanced growth strategy". Like Perroux and Boudeville, Hirschman saw development as "a chain of disequlibria". He mentioned Scitovsky's [1954] notion of inter-action, and argued that "development is a lengthy process during which interaction of the kind described by Scitovsky takes place not only between two industries, but up and down and across the whole of an economy's input-output matrix" [Hirschman, 1958, p66]. In practice, Hirschman argued that the key problem for LDCs economies was a series of "interlocking vicious circles" where the factors of production and abilities were "hidden, scattered and badly utilized" [Hirschman, 1958, p5], thus they need "pressures" or "tensions" to mobilise the largest amount of resources. Therefore, a certain level of polarisation will be expected to stimulate the development at the depressed area. As with Boudeville, Hirschman was also influenced by the previous work of Perroux, as he wrote "economic progress does not appear everywhere at the same time" and " that once it has appeared powerful forces make for the spatial concentration of economic growth around the initial starting-point" [Hirschman, 1958, p183].

Although he presumed the Perrouxian type of polarisation and he argued that such concentration (i.e., growth poles) should be located in large cities, he focused on intermediate and basic industries, rather than high industrial complexes, since they would induce greater backward and forward linkage effects. He assumed a simple model for this discussion, and he set the "North" as the growing region and the "South" which indicated underdeveloped regions. Within this model, Hirschman expected "trickling down" and "polarization" effects: "the growth of the North will have a number of direct economic repercussions on the South, some favourable, others adverse" [Hirschman, 1958, p187]. In long run, the trickle down effects have been expected to prevail more so than the polarisation effects, thus reducing the regional disparities. He explained that this trend is inevitable because of agglomeration diseconomies. Actually, one year after Myrdal's [1957] critique, Hirschman showed that trade and factor mobility would automatically reduce regional income disparities. Moreover, Hirschman attributed this compensation mechanism to a certain level of state intervention, which Gore pointed out as "an integral part of Hirschman's unbalanced strategy" [Gore, 1984, p96]. In his growth pole strategy, state intervention cannot concentrate only on making agglomeration economies, but it should be attached with a supplement policy such as building residential satellite cities around the growth pole, investments in service infrastructure in rural areas, establishment of appropriate transport system in a whole country, etc., thus reducing the tensions or pressures for polarisation. As he noted : "if the market forces that express themselves through the trickling down and polarisation effects result in a temporary victory of the latter, deliberate economic policy will come into play to correct the situation" [Hirschman, 1958, p190]. Although he noticed negative aspects of polarisation and agglomeration diseconomies, it seems to be the point that the growth pole strategy could be still the best strategy with this supplementary state intervention.

Other regional economists

In the 1960s, the trend of development economics had a general characteristic. This was well represented by Rostow's stages of development. In his influential thesis "The Stages of Economic Growth: A non-communist manifesto" in 1960, Rostow argued that all countries could be located in one of a hierarchy of development stages, such as the traditional society, the transitional stage, the take off, the drive to maturity, and the stage of high mass consumption. In this hierarchy, developing countries are in the traditional or the transitional stage. For such underdeveloped countries, it is necessary to accelerate economic growth to shift to the take off by "a leading sector" [Rostow, 1960]. This notion of a leading sector has a similarity to Perroux's "propulsive units" [Higgins, 1988]. This assumption is that all countries must follow the same path of development which industrialised countries have experienced. Thus, the 1960s' discussion of growth poles had been influenced by this prevailing attitude, and some regional economists applied the "established" development pattern to the regional development of developing countries.

Williamson [1965] has investigated the history of some developed country's economies and concluded that income disparities among regions are rising at the early stage of national economic growth, and then the regional dualism or divergence disappears at the mature stage [Williamson, 1965]. Thus, Williamson drew an inverted-U curve between a horizontal axis of income per capita and a vertical axis of inter-regional income disparities. Williamson's famous axiom is "rising income inequalities and increasing North-South dualism is typical of early development stages, while regional convergence and the disappearance of severe North-South problems are typical of more mature stages of national growth and development" [Williamson, 1965, p179]. This hypothesis of Williamson had been considered as a pattern of regional development in the 1960s, and it was believed that this pattern might also occur in LDCs in the 1970s.

Friedmann [1966] contributed to this view. In his 1966 publication, Friedmann argued that the development pattern of regions of the USA would be applied to all developing countries. He derived this conclusion from a synthesis of some empirical studies concerned with regional development of the USA, such as Schultz [1951], North [1955] and Perloff [1960]. Schultz [1951] provided a "retardation hypothesis" where economic development would occur in industrial-urban growth centres with manufacturing activities and it required the absorption of agricultural activities in the periphery. Friedmann argued this hypothesis was supported by an empirical study by Nicholls [1961] in South Carolina, Georgia and Tennessee from 1860 to 1950. North [1955] studied the history of economic growth of North America and pointed out that "the success of the export-base has been the determining factor in the rate of growth of regions.... The importance of the export base is a result of its primary role in determining the level of absolute and per capita income in a region, and therefore in determining the amount of residentiary secondary and tertiary activity that will develop" [North, 1955, p346], and he argued that the expansion of the export sector is the key for mass development. Perloff [1960] also thought that manufacturing industry was identified as a key factor for regional economic growth, and businessmen decided the location of economic activities by the regional advantages of the most needed natural resources. If one region started developing, it may continue because of 'cumulative advantages' and the first developing region will be the "heart land", and regional disparity would occur. However, this disparity will be diminished by a "filling-in process" of regional growth. Perloff argued that this pattern had happened in 1920s USA. Although these studies were about the development patterns of the USA, Friedmann [1966] synthesised these findings and applied them as the development pattern which developing countries must follow. Friedmann argued in his "proposition of the spatial incidence of economic growth" that regional economies must be opened up to the outside to accelerate the export sector growth, that economic growth tends to occur in urban centres where more firms are more likely to locate, and that flows of labour may help to reduce regional disparities. These propositions are obviously drawn from the empirical findings of the previously mentioned studies, but the critical point is that Friedmann applied this pattern to the regional development of developing countries.

Lewis [1954] and other "dualists" saw that development had to be brought about by the expansion of the urban-industrial sector which produces more efficiently and more economically with the higher technologies compared to the rural-agricultural sector. For example, in his famous article "Economic development with unlimited supplies of labour", Lewis [1954] assumed there to be two contrasting sectors; one is a traditional, overpopulated rural subsistence sector whose marginal labour productivity is zero or near zeroand where there is abundant surplus labour, another is a high-productivity modern urban industrial capitalist sector which has enjoyed higher income and technology and where profits are saved and reinvested.

With this situation in mind, the Lewis model constructed the structural transformation. Since the traditional sector is overpopulated and is characterised by low productivity while the modern sector has ensured a higher wage, then surplus labour in the traditional sector will shift to the modern sector where high productivity and more employment is enjoyed. This growth of the modern sector does not stop here, because there will be a reinvestment from profits, which enables this sector to receive more labour from the traditional sector. Since there is an unlimited supply of labour from the traditional sector, "capital-saving technology" must be chosen for employing more labour. In this cycle, the modern industrial sector will overtake the traditional agricultural sector. Finally, the Lewis model has implied self-sustaining growth of the modern sector which seems to be the key factor in the development process. This process of development seems to have some similarities with the growth pole concept. Lewis' model conceived some disparities in an economy, and this model explained that the winner, who has more efficiency in its production with a modern technology, absorbs (or destroys) the loser, who has a relative backwardness. The development has been considered as an expansion of the winner's influence. Thus, the rationale of a derived strategy from this model is to make the winner win, i.e., more investment in a growing sector (or region).

By regarding these growth theories in the 1960s, Gore [1984] concluded that these 1960s regional economists have a significant feature in common. Firstly, their arguments were based on a Rostovian "stages-of-growth" conception and Rosenstein-Rodan's "big-push" conception. Secondly, they believed that growth pole strategies can achieve various regional policy objectives because of the relationship between urban-industrial growth and regional-rural development. Thirdly, they believed, and also provided some proof that such relationships have been empirically verified.

General problems

Before starting to review the critiques of the 1970s which mainly asserted that the growth pole just brought about regional inequality, it is possible to mention some general problems of those 1960s arguments from a methodological point of view.

From the early 1960s, attention was given to the planned growth pole, and thus to pursuing a growth pole strategy. Parr [1996] pointed out that two strands of thinking lay behind this. Firstly, as seen in Friedmann's work, there was some evidence that investment at a limited number of economic centres would provide a satisfactory condition for development. Secondly, there had been established Perroux's growth pole notion and a practical application of this notion into geographical space by Boudeville. Parr [1996], however, argued that they had two sources of confusion. A confusion is "the assumption that the growth pole in geographic space was simply a spatial manifestation of the growth pole in economic space". Another is "the assumption that the natural growth pole in geographic space could be replicated in the form of a planned growth pole, by the implanting of propulsive industries at particular urban centres" [Parr, 1996, p6-7]. These assumptions are illogical, as can be seen easily just by imagining the desired size of migration to "make" a growth pole, e.g. the natural growth pole has to have a substantial magnitude (at least 250,000 people) in order to have a certain influence within and outside of the region, but of course most of the planned growth poles cannot have such an influential size of population mainly due to the lack of enough financial support in reality.

Gore [1984] criticised Boudeville's and Hirschman's views since they conceived only urban-industrial growth pole strategies, and they conceived only a geographical space as its analytical framework . He also criticised the other economists that are quite far behind the first attempt of Perroux's study, due to their "inadequate" explanation of how regional and rural growth occurs and their conclusion that the important policy is still just to "manipulate general economic relationships rather than through policies designed to change the spatial distribution of industry and urban population directly" [Gore 1984, p117]. He pointed out that whether those 1970s growth pole strategies would be successful, ultimately depends on the assumptions about "how development occurs". The 1960s discussion of growth poles had not mentioned that once one of the assumptions is denied in reality, the growth pole strategies result in a different outcome.

Gore [1984, p89] warned that "the growth pole was divorced from its analytical basis in Perrouxian growth". Indeed, "Once the growth pole is defined in purely spatial terms, it is possible to bring any body of theory... the theoretical basis of all growth pole strategies.... is ill-specified" [p92]. A limitation of Perroux's growth pole conception into only a geographical meaning "robs the concept of much of its usefulness in regional economic analysis" [Parr 1996, p5], since it can not be applied in a functional or sectoral analysis, e.g. a credit card company exists with a concentration of power under its name even while its offices are dispersed around the world. Perroux himself wrote "it would be untenable to reduce the theory of development poles to a mere instrument of regional policy" [Perroux, 1988, p54]. A methodological fallacy of 1960s discussions of growth poles was that they seemed to have followed the academic trend of the 1960s without any careful treatment of Perroux's growth pole theory, and they were too quick to apply it as a regional planning tool.

The 3rd wave : growth poles as a cause of regional inequality

In the 1970s, the academic trend surrounding the growth pole strategies was completely changed. As mentioned in the introduction, most of the regional economists started criticising the growth pole strategies and governments turned their back on the growth pole strategies as a policy option of regional planning. This change was caused by the fact that a lot of regional policies known as "the growth pole strategies" resulted in unsatisfactory and sometimes unfavourable outcomes. Stohr and Todtling [1977] synthesised the case studies and found that growth pole strategies could not bring development to the hinter land. They concluded that the growth pole strategies may have been successful in reducing interregional disparities, but as local spread effects were weak, moreover, they had at the same time caused intra-regional, in particular rural-urban, disparities.

In this context, some economists explained the mechanism of disparities "caused" by the growth poles.

Myrdal

In his 1957 publication, Myrdal had already noticed that development is undertaken in a highly unequal manner. With a quotation from the Bible, he noted "that there is a tendency inherent in the free play of market forces to create regional inequalities, and that this tendency becomes more dominant the poorer a country is, are two of the most important laws of economic development and underdevelopment under laissez-faire" [Myrdal, 1957, p34]. This argument has a similarity with previously mentioned views by Boudeville and Hirschman, however Myrdal did not expect this disparity to be reduced in the long-run. Rather, it would be strengthened by a process of circular and cumulative causation. From the case study of the African-American population, he pointed out that once a particular centre in a region starts growing, the region gains its economic power through the process of cumulative causation, which is an expansive train of economic activities, and this region will continue its development further by absorbing resources of other regions; in his words "at the expense of other locations and regions" [1957, p27]. Thus, he argued that trade and factor mobility has backwash effects on other underdeveloped regions, so that poorer regions become poorer in the development process. Besides the backwash effects, he also mentioned "spread effects" which diffuse the technology and innovation to other lagging regions. However, Myrdal strongly rejected the optimistic assumption of Hirschman, and he wrote "in no circumstances do spread effects establish the assumption for an equilibrium analysis" [1957, p32]. Rather, he had a pessimistic view; he wrote "It has never occurred in recorded history that a privileged group,.......,has climbed down from its privileges and opened its monopolies to the unprivileged. The unprivileged have to become conscious of their demands of greater equality and fight for their realisation" [Myrdal, 1971, p88-89]. The backward regions in the developing countries, where most of the population are underprivileged, have less political power and can not catch up with the privileged region.

From this pessimistic point of view, the growth pole strategies (here the writer uses this name purely in terms of 1960s industrial-urban centre type growth pole) help the growth of already wealthy regions, but it has no effect or even unfavourable effects on the regional equality. In order to reduce such a political and economic gap, Myrdal [1957] asserted that development planning must focus on the improvement of education, transportation, and other means of communication in the depressed regions. In his view, growth poles are not a tool to help egalitarian policy, rather they are the cause of regional inequality which must be attacked by a proper development policy.

Friedmann

Friedmann, who used to represent one of the pro-growth pole strategies, also changed his views and admitted that regional inequality had been accelerated by the development process.

In his 1972 publication, he reviewed the academic trend of regional development in the 1960s and wrote "none of them can be accepted as an adequate framework for regional development planning in its more general connotation" [Friedmann, 1972, p41]. Rather, he provided a "general theory" of polarised development from a socio-political point of view. In his theory, innovation happens in a limited place, the people and the place of innovation have power, and the core-region has six effects such as a "dominance effect", "information effect", "psychological effect", "modernization effect", "linkage effect", and "production effect". All of these effects are working to weaken the periphery region and it causes a conflict between core region and peripheral region elites. The outcome of this conflict is indeterminate and depends on whether peripheral region elites are repressed, neutralised or co-opted, or succeed in replacing central elites. As with Myrdal, Friedmann explained that developing countries have more disparities in political power between core region elites and peripheral region elites, so it is difficult to reach egalitarian objectives. Therefore, this leads to a conclusion that the decentralisation of decision making functions is the crucial point. This "general theory" of Friedmann's was important since it introduced a political dimension more clearly into regional development analysis.

In this framework, the growth pole strategies which emphasise an artificial concentration of economic activities, have no ability to reduce the regional inequality for which Friedmann saw more political causation. Rather the growth pole strategies just help the richer remain rich so it worsens the backward situation of the peripheral regions.

Dependency theory

Dependency had emerged as a challenging theory in development economics during the 1970s. There were two main arguments: one is related to unequal exchange, and the other is based upon the concept that the world economy is divided into a core of dominant nations and a periphery of dependent ones. The latter argued that all countries in the world have been forced to be involved in a world capitalist system since the sixteenth century, which benefits only the western developed capitalist countries .

Baran [1957, p196] pointed out "the continued foreign exploitation of underdeveloped countries and their domination by the imperialist powers.", and he argued that economic development in underdeveloped countries is "profoundly inimical" to the dominant interests in the advanced capitalist countries and powerful foreign corporations.

In this context, Frank [1967] considered "metropolitan - satellite relationships" which were similar to Friedmann's "core-prephery model", and in these relationships only the metropolitan area received the benefit of growth at the expense of the satellites. Thus, such benefits of urban-industrial growth are never spread nor 'trickle-down' to the surrounding backward area. This is due to "lack of access to their own surplus" of satellites and "the same polarization and exploitative contradictions which the metropolis introduces and maintains..." [Frank 1967, p9] Indeed, he noted that "economic development and underdevelopment are opposite faces of the same coin" [Frank, 1967, p9].

Slater [1975, p137] called colonialism "an integration that produced underdevelopment" since such colonialism broke down induced spatial relationships in developing countries and integrated only to the world economy, so it made developing countries "internally disintegrated". [Slater, 1975, p137]. He mentioned the Tanzanian case where all product functions were linked only with export, and he argued that the development of LDCs has been exploited by "the metropolitan powers" [p138].

Santos [1979] provided a dualistic model with "two circuits": the "upper circuit" is associated with modern urban types of economic activities such as banking, export trade and industry, services, and wholesaling, etc., and it tends to have less internal linkages; and the "lower circuit" which consists of induced forms of manufacturing, non-modern services and it has mainly internal linkages. He concluded that "It is clear that this situation will only benefit the upper circuit: the lower circuit will become progressively 'lower'. In absolute terms the upper and lower circuit will develop, but the relative gap between them will become pronounced" [Santos, 1979 p199]. In developing countries, this upper circuit tends to be monopolised by foreign corporations and, since reinvestments are made in totally unconnected industries or foreign countries, the growth is unlikely to have the desired effects from backward and forward linkages.

What kind of conclusion can be derived from this pessimistic point of view? One clear devastating conclusion seems to be that no regional policy can solve the regional problems in developing countries in a situation where the developing countries have been forced to depend on the world capitalist economy.

The growth pole strategy is not an exception. Polenske [1988] wrote "while the growth pole theorists maintain that the domination of certain firms is a positive factor in the development process required to help the mass of the population, the dependency theorists argue that domination leads to expropriation of the surplus product not for use by the masses, but for use by the capitalists" [Polenske, 1988, p96]. Therefore, the growth pole strategies again help the regional inequality rather than equality in this context.

These critiques of the growth pole strategies were powerful enough to make people feel that the growth pole strategies may lead to unfavourable results, even though they only argued about a simple type of growth pole strategy which was just about a concentration of investment in a limited industrialised area. Whereas it also seems that the critiques are too powerful since they succeeded in making people abandon even the growth pole concept itself.

The 4th wave ? : revisionism

Most of the criticisms against the growth pole strategies seem to have a general characteristic. They see only one aspect of the strategy namely in terms of industrial-urban development strategies. This image had been established through the 1960s applications of growth pole concepts, but it does not mean that the concept of growth poles and its application are useless or ineffective.

Richardson and Richardson [1974, p169] wrote that "the disenchantment with growth centre policies in many countries is not evidence that the principle of polarization is wrong. On the contrary, it reflects the over-optimism and short-run time horizon of regional policy-makers, the failure of sustained political will, the use of deficient investment criteria, bad locational choices and lack of imagination in devising appropriate policy instruments." Higgins [1988, p3] argued that "the strategy applied was seldom the one that Perroux favoured. Far from straightening and encouraging existing growth poles, the strategy has been one of trying to create growth poles in retarded, disadvantaged regions, hoping for spread effects from the chosen pole to its own geographical regions... In my view, it is very important to make it clear that "the failure of the growth-pole strategy" was not the failure of the Perroux theory, but failure of a distorted version applied by his disciples, mainly Boudeville". Savoie [1988] asserted the validity of the growth pole concept, however he warned "it is very different from the one that has usually been applied by governments that have announced a growth pole strategy" [p378]. Even Perroux [1988] admitted the growth pole strategy does not work in certain situations, and "the establishment of foreign-tied mining or raw material producing enterprises inadequately integrated into the domestic economy is a common feature of the less-developed countries. These countries furnish striking examples of markedly polarized urban areas and the same time of vast empty or very backward regions" [p69] and "a growth or development pole is a growth-inducing unit coupled with its surrounding environment. In an environment that is already well knit and vigorous, thanks to active individuals, the establishment of a new enterprise generates wide and varied agglomeration and complementary effects" [p70].

Gore [1984] discussed the difficulty of measuring the impact of the growth pole strategy. He argued that geographical mapping can be used as a "development surface" with growth poles but that we do not know what the 'expected' effects would be for the comparison, so it is not logically possible to judge its effects. It may be substituted by a "before-after test", even though it is not suitable for cross country analysis. From a different point of view, Stohr and Todtling [1979] argued that the measurement of spatial equity is too "materialistic" and it ignores other important values such as family and local solidarity, self-realisation, and access to the central decision making power.

From the previous discussion, one clear conclusion is that the growth pole strategies have various aspects and it is not possible to argue the validity of this complex strategy from only one aspect. Various growth pole strategies can be considered in practice depending on their objectives such as national growth, regional growth, regional equality, rural development, and so on. The methodological fallacy of previous discussions of the growth pole strategies is that they discussed the various strategies as a generalised strategy, and sometimes they criticised the strategies in terms of a particular objective which the strategies did not target.

Therefore, it seems to be necessary and fair to distinguish all types of growth pole strategies in terms of development objectives and discuss their validity one by one. Thus, the next chapter will discuss anticipated growth pole strategies by each objective, and consider their required conditions and expected effects.

Chapter 3. Policy issues of the Growth Pole Strategies

General characteristics of the Growth Pole Strategies

This chapter will discuss practical policy issues for various problems to which the growth pole strategies can be applied. Although the growth pole strategies have been proposed and implemented in widely diverse ways for various problem settings, they have a set of general characteristics among them. According to Parr [1996, p28-29], the growth pole strategy :

"involves encouraging the growth of employment and population within a region at particular locations or planned poles over some specified period."

"requires a limitation on the number of locations or centres which are designated as planned poles"

"necessarily requires spatial discrimination or selectivity among locations"

"inevitably involves a modification of the spatial structure of employment and population within a region."

While the general characteristics of growth pole strategies in practice can be stated as above, they show various styles in the implementation process. Sometimes, the growth pole strategies for different objectives are contradictory to each other. The following sections will discuss derived growth pole strategies for various objectives, and this will make it clear that an extension of the growth pole strategy can have a broad range in its application.

National Economic Growth

A growth pole strategy for national economic growth is a strategy which a country pursues unconsciously in practice. All capital cities have some form of leading sector, not only industries but also tertiary sectors. Such naturally growing capital cities have been located their governmental bodies and most of the authorities, which attract people and firms to settle in the capital. There is no example where another city outside the capital (except of planned capitals like Brasilia) grew first. In the early stage of economic growth, the government tends to invest firstly in its capital city. This rationale can be quite right from an economic point of view, since the capital has more of the population and potential to bring about a rapid growth, and it is easier to invest in infrastructure in the growing capital city. In addition to investment for the capital city, the industrialised region which has already grown by itself must be attractive for the government to invest in infrastructure or a new project, since this is financially easier and more economical if the government is focusing only on national economic growth. Actually, this case can be observed in most countries at the early stage of development. In this case where the government concentrates its investment only on the growing area, it can be argued that the government has implemented a growth pole policy for national economic growth, whether or not they conceived an idea for growth poles. In order to achieve this objective, the growth pole strategy must involve: investment in the infrastructure in the growing area (housing, hospitals, water supply, transportation within the area, etc.); surrounding the area with some power stations; enabling an adequate supply of food to the growth poles by setting up some regions of agriculture or considering importing agricultural products; transport construction to link the growth poles, and so on.

After reaching a peak, however, such strategies will be faced with a physical limitation of city size. This is due to the diseconomies of a large city, such as expensive public services, under-utilised infrastructure, over population, traffic congestion, pollution, etc. Moreover, the growth pole strategies for a purely national economic growth objective will cause new regional problems, such as urban-rural inequality, intra- and inter- regional inequality, a vast immigration to the urban area, and so on. Therefore, an objective of national economic growth should not be a priority of the objectives of growth pole strategies for regional planning, although the "national" government naturally tends to do it in reality.

Regional Economic Growth

If a region is to be developed with a regional policy, the targeted region must have a leading sector or a propulsive industry to boost the economic growth in this region. Basically, the strategy to develop the region seems to be similar to the strategy for national economic growth as mentioned above, since the proposed policy is to choose the leading sector to be intensively invested in. It is, however, a critical point as to whether this development has a purely economic objective or contains some egalitarian objectives within the region. If the strategy attempts to pursue egalitarian objectives for the whole population within the region, an investment concentrated on the growth pole will lead to the reverse result. Even with the construction of a transport system linking the growth poles and other hinterlands, it is common that the migration tends to be from rural to urban, rather than producing a spread effect of the population, as a response to the establishment of new transport linked to an urban growth centre (for example, see Gilbert and Goodman, 1976). Thus, in order to achieve the intra-regional equality, it is necessary to consider a comprehensive regional planning which concerns the multiple growth poles in the backward areas co-operating well with the leading sector in the growth centre. For instance, if the growth pole has been a highly industrialised urban city, it is possible to set up agricultural centres around the city and link them with transportation, or perhaps recreation centres might be possible to be considered for such hinterlands.

It may be possible to argue that an economic growth objective and an egalitarian objective in regional planning can be complementary rather than competitive, however, in most of the cases, these two objectives do not move in the same direction. For an extension of the above example, the regional planning which considers the agricultural development in the lagging areas will be faced with a severe choice between equality and efficiency, when imported agricultural products are relatively cheap in price for that region.

Inter-regional Equality

Inter-regional inequality has been a common phenomenon in the process of national economic growth, as Williamson [1965] observed. However, it also can be said that such inter-regional inequality has been 'created' by the misallocation of investment and resources, sometimes by a growth strategy based on growth pole concepts targeting national economic growth as mentioned in the previous section. Theoretically, this issue is the same as that of intra-regional equality discussed in the last section. For example, if it is assumed that the South is lagging behind the Northern region, it is necessary to set up some growth poles in the South expected to act as 'countermagnets' against the Northern growth poles. In this case, again it is necessary to consider the existing economic system in which the previously existing growth poles will be activated. A simple policy such as the establishment of growth poles in the South, or creating links to the North, cannot be a solution, rather it may end up causing a deterioration of the Southern economy (because of a misallocation of resources in a planned economy) or may cause new pressures of out-migration to the North via the new transportation. Again it is an irony that the new growth poles are in danger of causing new intra-regional inequality within the South. Even if the Southern region has developed according to the economic indices, it does not necessarily mean that the "whole" of the region has developed, rather it tends to end up causing a new intra-regional inequality in the South. It may be possible to consider a highly dispersed distribution of investment as an alternative, and the next chapter will mention some cases based on this policy. However, if it is necessary to distribute investment in as dispersed away as possible, then that one unit of investment will be quite limited in being able to achieve development objectives in the whole region. Also, the government may not have a large enough budget to pursue this kind of policy. Here again, regional planners must face a severe choice between equality and efficiency.

Rural Development

Rural areas have tended to be forgotten in development strategies in the past, however the current situation is that they have now become a main objective. Stohr and Taylor [1981] discussed the different types of development strategies, and they grouped them into the "from above" type and the "from below" type: the "from above" is top-down, centre-outward development; the "from below" is bottom-up, periphery-inward development. They concluded that "from below" was especially appropriate for LDCs, where most of the poorer population lived in the periphery areas and migrated to the urban areas, but were still in poverty. In this context, there has been an argument about the 'agropolitan' development in the rural areas. Friedmann and Douglas [1978] studied the development processes of Asian countries and argued that the traditional top-down policy based on the industrialisation paradigm was "inoperative as a guide line to the future" [Friedmann and Douglas, 1978, p181] because of the situation which LDCs have been facing, such as rising import prices, declining export markets, and a deterioration in the terms of trade. For an appropriate development strategy, they made the following concluding remarks [Friedmann and Douglas, 1978, p181-182] :

"Limited and specific human needs should replace unlimited generalized wants as the fundamental criterion of successful national development..."

"Agriculture should be regarded as a leading or 'propulsive' sector of the economy."

"Attaining self-sufficiency in domestic food production should be regarded as a high priority objective."

"Existing inequalities in income and living conditions between social classes and between urban and rural classes should be reduced."

"Facilitative measures to increase production of wage goods for domestic consumption should be given high priority."

"A policy of planned industrial dualism should be adopted whereby small-scale production for the domestic market is protected against competition form large-scale capital-intensive enterprise."

Besides the agropolitan development, it is possible to establish small market towns in rural areas, as Johnson [1970] advocated. Johnson argued that "the countryside is inadequately provided with accessible market centres" [Johnson, 1970, p171], and argued for the establishment of more small-scale industries and less concentrated urban strategies. Although he was against the urban-biased growth pole strategies based on the development of large industrial complexes, he still followed the growth pole concepts, since he wrote "a well-dispersed network of promising 'growth points' should be selected in different regions of a country" to establish "agro-urban community" [Johnson, 1970, p219]. Moreover, Johnson acknowledged the value of industrial growth poles in a limited sense, in that they should have some positive linkages with small towns and rural growth centres.

Urban system and Migration Control

According to Townroe [1979], LDCs have been "suffering from a condition of 'primacy' in their urban hierarchy: the largest city in the country is considerably larger than any other city and is growing as fast or faster than other cities" [Townroe, 1979, p93]. Moreover, most of these developing countries have been experiencing a rapid increase of national population. These high growth rates of population are reducing the agricultural land area available per capita in many countries, thereby maintaining or increasing rural-urban migration flows to cities which are themselves experiencing large rates of natural increase [Townroe, 1979, p93]. Therefore, it is necessary to consider a proper urban system that can reduce such migratory pressures on the limited number of urban areas. The growth pole strategies have been considered explicitly and implicitly as regional planning tools to control the migration.

Even in developed countries, the urban system is important since there has been the assumption that the "long-run performance of the national economy is strongly influenced by the nature and form of the urban system" [Parr, 1996, p16].

In this case, the strategy must be considered with the influence of the existing urban cities. Some planned growth poles tend to be located in intermediate urban centres as they can be 'countermagnets' against larger urban cities. Of course, it is possible in theory to establish new growth poles in some of the rural areas, however it is more difficult since it is expensive to produce a new competitive city, and it is difficult to find a new appropriate place. A problem is how to choose the location of such influential intermediate urban cities. In theory, they are the cities which have more backward and forward linkages and have more potential to compete with the existing large cities, however in most of the cases, the selection is limited in terms of numbers and quality. Even if the planner could choose the growth poles in practice, the chosen growth poles could not compete with an existing powerful urban city since the economic activities in the chosen growth poles are limited, and they would not be influential enough to attract immigrants.

Variety of the Growth Pole Strategy

As seen in the above discussion, the growth pole strategy can never be generalised as a set of programmes. The confusing nature of this strategy, in terms of the various objectives and programmes, has caused a misunderstanding and misuse in its application. Indeed, it is not possible to generalise various growth pole strategies as a fixed set of programmes, although some empirical evidence has shown that many objectives were included within a growth pole strategy. The failures of strategies in practice can be attributed partly to the confusing utilisation of this strategy. The next chapter will review some empirical case studies.

Chapter 4. Empirical Evidence

Difficulties of Evaluation

This chapter will review various case studies of growth pole strategies. It is, however, problematic to evaluate the effects of a regional policy since there are some difficulties as follows.

Limited Availability of Case Studies

There are a lot of articles in journals and published books with titles such as "a case study of a growth pole", however, most of them are categorised into three types : explanations of published regional plans applying growth pole concepts; a regional analysis using growth pole concepts; and policy recommendations of growth pole strategies (for example, see Hansen, 1969, Kuklinski 1972, also 1981, and Lo, 1978). Few of them have actually done an evaluation of a real case where a growth pole strategy was implemented. This is because most of the articles were written in the period when the growth pole strategy was still a "new" strategy and so perhaps some of the writers applied it too quickly rather than reviewing its effects retrospectively. Thus, the number of case studies which this dissertation refers to has had to be limited, despite the fact that a lot of references were theoretically available.

Time Issues of Evaluation

Unfortunately, most of the case studies which this dissertation refers to were published in the 1970s, so most of them were evaluated too soon to analyse the effects adequately. Theoretically, the effects of a regional policy must be assessed in the long run, more than 10 years, since a regional policy must be targeted to effect not only the local geographical points but also the whole of the region and the effects must be assessed in a comprehensive manner, e.g. an analysis of structural (spatial) changes of the economy. Of course, it does not mean that all assessments must be done only in the long-run. Some certain urgent objectives, such as an eradication of urban (or rural) poverty, should be assessed over a series of shorter periods.

Problems of Measurement

It is difficult to choose a measure to evaluate the effects. The possible measures are limitless, such as regional gross output, gross output per capita, income, population growth, city size, the number of industries (firms) or houses and so on. Frameworks for measurement are also various, such as "in a growth pole", "in the zone of influence", "urban-rural disparity", "interregional disparity", "intra-regional disparity", "national economy", and so on. Of course, a functional approach by considering an inter-sector analytical framework can be a possible measurement. A more dynamic analysis is possible by analysing the spatial change of the economy of a country, and also the urban system in a country. Moreover, Stohr and Todtling suggested that an assessment of regional policy should consider more non-material indicators such as "loving" or of "being" [Stohr and Todtling, 1979]. There is no theory for choosing the correct measure, and no complete measurement system exists. Thus, it is necessary to use various measures to see a comprehensive effect of a policy according to its policy objectives, even though most of the case studies were based on mainly economic measurements.

Mixed Effects Problem

Allen and Maclennan [1970] pointed out that any evaluation of a regional policy cannot avoid the problem of the changes in the objectives of a regional policy and the problem of external elements. For example, most of the countries had implemented other forms of regional policy before they introduced a new regional policy. In this case, it is very difficult to distinguish the effects of the old policies from the new policy. On the latter problem, for instance, if a new oil field was discovered and the economic status of the targeted region was increased, it is impossible to know how much the regional policy influenced the economic growth of this region. Stohr and Todtling [1979] also noted this problem and admitted that "most authors found it difficult to distinguish clearly between autonomous and policy induced trends of spatial development" [Stohr and Todtling, 1979, p43].

Even though there are some difficulties as mentioned above, it is still worthwhile to learn from the past by looking at case studies, one by one. The cases of successes and failures of the growth pole strategies certainly showed general characteristics, which will be summarised in the final chapter.

Developed Countries

United States of America

In 1965, two important Acts were passed in the USA: the Appalachian Regional Development Act and the Public Works and Economic Development Act, and obviously "their outlines have utilised the growth centre concept in their operation" [Hansen, 1972b, p266]. The Appalachian Region extends from north-eastern Mississippi to southern New York. Even though this region was between two growth areas, the Atlantic megalopolis and the industrial Midwest, this region was lagging and underdeveloped: low wages, relatively high unemployment rate, very poor quality of housing, low levels of retail sales and savings, a poverty of local services, a heavy dependence on public assistance, low urbanisation, etc. Out-migration was more than two million between 1950-60 [Hansen, 1972b]. Therefore, the Appalachian Regional Commission which was set up by the Act, considered that migration was related to a good job market, so that the establishment of growth centres was designated to boost the economic development of this region. In the Act, it was stated that "the public investments made in the region under this Act shall be concentrated in areas where there is the greatest potential for future growth, and where the expected return on public dollars will be greatest" [Appalachian Region Development Act, 1960, quoted in Hansen, 1972b]. Sixty development districts were selected and a distinction was made between a growth centre, a growth area, and the surrounding hinterland in terms of its size and economic growth potential. To these districts, various investments were made : vocational (educational) projects, health projects, community development, child development, local development district support (research), environmental and natural resource projects, and other (chiefly non-highway transport). The Developmental Highway Programme planned the construction of 2300 miles of 'developmental highway' and about 1000 miles of 'local access road' to link those development districts. It is worth to mentioning that, during the process of these projects, the Appalachian Regional Commission had been working as well as an intermediate to construct co-operation schemes among related organisations. The result of this comprehensive regional policy seemed to be quite positive. Estall concluded that an "undoubtedly considerable advance has been made in Appalachia since 1965" [Estall, 1982, p49]. He showed the evidence that, in the late 1960s, a lot of new job opportunities were provided in Appalachia and the unemployment rate of Appalachia was just 5.4% in 1970, while the average in the USA was 4.9%. Income backwardness of the region was also improved from 74% in 1970 to 81% in 1980 compared to the average income of the whole USA. However, he put in a caution that "the achievements of the sixties must, be assessed in the context of a period of strong national economic growth.... Appalachia was bound to benefit from such influences" [Estall, 1982, p49]. He also pointed out a possibility that the development pattern of the Appalachian region since the end of the 1960s merely followed the nation-wide pattern of regional development in the USA. In other words, the regional policy might not have had any influence over the changes in Appalachia. For example, "the revival of numerous small centres is associated with the reversal of the old-established move of population to metropolitan areas, and with the vigorous growth of the 'inter-metropolitan peripheries' that is taking place over much of the U.S.A." [Estall, 1982, p51]. It seems that the effectiveness of the regional development policy may have been a reflection of the national development performance.

On the other hand, a development programme based on the Public Works and Economic Development Act was not successful [Hansen, 1972b]. It was targeting the development of some underdeveloped regions, such as the Ozarks, New England, the Four Corners, the Coastal Plains, and the Upper Great Lakes. In these regions, development centres were selected by anticipating that the development centre would provide job opportunities for depressed area residents and rural area residents. However, this programme could not be implemented because of several reasons. Firstly, an established development commission simply did not have an adequate budget. Secondly, most of the selected centres were too small (less than 50,000 population) to boost the economy and to create enough jobs in the community. Thirdly, the development commission (Economic Development Administration) did not make enough effort to establish links among the related organisations. Therefore, projects under this Act ended up just as a point-distribution investment policy in the backward region, and they could not achieve any agglomeration effects with the growth poles as the Appalachian Regional Development Act did [Hansen, 1972b]. This contrast between two regional policies in the same country and in the same period shows that a regional policy is still important, and a comprehensive approach is more favourable when implementing a growth pole strategy.

Hansen [1972a] presented an interesting case where movement of an industry does not always bring about an increasing effect on employment. In 1956, Kaiser Aluminium and Chemical Corporation set up a new aluminium reduction and rolling mill at Ravenswood, West Virginia, fifty miles from Charleston. The main objective was to provide job opportunities to the local area, which suffered from high unemployment. About 300 jobs were created for the local community by this project, however most of them were primarily low-skill jobs in retail trade, and high salary jobs which required a profession went to persons from outside the community. Hansen [1972a] pointed out a problem that "directly applicable skills and even adequate basic schooling were lacking in the local manpower pool" in that local area [Hansen, 1972a, p112]. He also presented another case of a plant built by American Standard in 1970 in eastern Kentucky. In this case, most of newly created jobs were filled by the returned migrants who had learned a skill in the North. These cases show that the movement of industries is not a sufficient condition to solve an unemployment problem in a backward region.

France

France had implemented growth pole strategies intensely in the 1960s and 1970s. In 1960, The Fourth Plan had been activated which selected "development areas" in order to bring about industrial development in backward regions (the West, South-West, parts of the North and Centre of France), which were seriously behind Paris [Allen and Maclennan,1970].

The development centres such as Nantes-St Nazaire, Limoges, Montpellier and Bordeaux were selected, while certain rural areas were scheduled in Brittany and the Loire basin on the consideration with the problem of agricultural development. The public investments were then concentrated on the infrastructure in those development areas, and especially the educational investments (schools and vocational training) were emphasised to encourage farmers to work in industrial firms. It is important to mention that the Fourth Plan put a priority on the infrastructure rather than financial aid to private firms. Following this was the Fifth Plan (1966-1970) which extended the growth areas. Eight provincial cities such as Lyons-St Etienne, Marseilles-Aix-Delta du Rhone, Bordeaux, Lille-Roubaix-Tourcoing, Toulouse, Strasbourg, Nates-St Nazaire and Nancy-Metz-Thionville were selected as 'countermagnets' against Paris. Allen and Maclennan [1970] evaluated the early policy and argued that nearly half of the decentralisation effects (e.g. building a new firm) during the period 1950-63 by this policy were made within a radius of two hundreds kilometres around the city of Paris, thus "There is little evidence to suggest that this concentration in the Paris basin is being reduced" [Allen and Maclennan, 1970, p243]. Rather, they pointed out that the direct settlement control on new firms was more effective to halt the urbanisation of Paris. They also mentioned that the effects were variable among regions. The West region experienced 15.1 per cent rise in the number of firms and 19.7 per cent rise of the created jobs during the period 1950-1962, however, other parts for instance the South-West and the Central regions did not enjoy the benefits of decentralisation: only 4.7 % of new firms and 4.5% of new jobs created in the South-West, and 3.8% of new firms and 2.7% of new jobs created in the Central regions [Data from Allen and Maclennan, 1970]. Thus, the result is not straight forward, but this seems to be attributed to "the nature of regional policy" [Allen and Maclennan, 1970, p224], i.e., it is difficult to judge the effectiveness of policy since this kind of policy must be implemented for backward regions which have already been behind and have fewer potential for growth.

Italy

Italy has been facing a problem of serious regional disparity. Southern regions have been lagging and underdeveloped. In 1957, the Italian authorities considered a growth centre was needed to stimulate the Southern economy and to reduce the disparity by a development policy named 'Law 634' [Allen and Maclennan, 1970, p67]. There were two types of growth centres considered in this scheme: "The Areas of Industrial Development" (Aree di Sviluppo Industriale), which must have a population of at least 200,000 and must contain a rapidly developing industry with basic infrastructure; the other is "The Nuclei of Industrialisation" (Nuclei di Industrializzazinone), whose industries must be relatively small and of mainly labour intensive technology. With this categorisation, 12 Areas and 39 Nucleis, whose population sizes were the same as the Northern region, were selected by each local authority, and not by the central planning office. As a result, in 1965, most of the governmental enterprises located in Areas and Nuclei, and 85% of grants and 50% of infrastructure within the development scheme, were available only to Areas and Nucleis. It prevented the out-migration from the South but it made migrants head for the Areas in the South. However, it seems that there was no spread effects in the zone of influence. Indeed, the South has still remained backward compared to the North, although the final judgements can not be made until the future.

Spain

The first attempt of a growth pole strategy in Spain was expressed in the development policy in 1964 [Richardson, 1975]. In that plan, two out of seven of the development poles, (Burgos and Huelva) were designated as industrial promotion to boost the regional development while the reminder were industrial development poles to spread the effects of growth of the industrial promotion poles. In 1969, another four poles were included (Granada, Cordoba, Oviedo and Logrono) and operated in 1970-72, while Villagarcia de Arosa was included and operated in 1972. The selection of poles was based on various regions: for example, Vigo and La Coruna development was expected to strengthen the north-west coast economy of Galicia and radiate growth diffusion to its hinterland. Both have good natural ports and have urban centres, while Huelva, Cadiz and Seville became a development triangle along the River Guadalquivir and contributed to the economy of Southern Spain. Richardson argued that "many of the choices were determined by clear-cut evidence of development potential" [Richardson, 1975, p116]. For the inducement of industries to those poles, there were some incentives, such as a 10 per cent subsidy to capital investment, preference in obtaining official credit, and fiscal reduction and tax exemptions. Even though the preferences were put on small and labour intensive projects, most of the applicants to this scheme were capital intensive projects. Actually, only 16.7% of the realised investment at the pole was made via the official credit. The result was that industrial development at the poles has been dominated by the chemicals and metal industry sectors, which are highly capital intensive. For employment, it was not so successful since most of the projects were middle size and capital intensive. Spread effects were varied amongst the poles. The Burgos-Valladolid link had little spread effects, while Seville and Zaragoza showed very strong hinterland effects. Richardson pointed out the reason as being that "the intensity of hinterland effects depends on the size of the central pole itself" [Richardson, 1975, p125]. He also argued that the level of infrastructure was a decisive constraint on development in some poles, e.g., a serious problem of water supply in Huelva, Burgos, and Vigo. In addition, an adequate technical infrastructure was needed in those areas. Direct effects (activities of firms themselves), indirect effects (inter-industry links) , and induced effects (income) with growth poles have shown variation among the growth poles. 141 (36%) out of 396 firms were established for the new growth poles. The population increase was much faster in the growth pole areas, but decreased in other places in the provinces. Even though the results were mixed and vague, Richardson concluded that "from the point of view of industrial promotion the initial seven poles have been an undoubted success" [Richardson, 1975, p132]. However, Buttler [1975] was suspicious of the effects of growth poles and argued that "except for those of Huelva and Valladolid, the new activities were not attracted by new complex formations.... The degree of inter-industrial regional linkage of the poles is thus low on the whole" [Buttler, 1975, p181-183]. Whereas, Lasuen [1974] evaluated the Spanish regional policy over the previous two decades, and pointed out the reduction in inter-urban income differentials and an increase in urban-rural income differentials. But he concluded that there was a certain improvement in regional disparity of income per capita as a net effect. In his view, the Spanish growth pole strategy was fairly successful in terms of inter-regional equality with the payment of causing intra-regional and urban-rural disparity.

United Kingdom

The UK had also attempted to pursue the growth pole policy. In 1945, the Distribution of Industry Act set 'Development Areas' which had relatively high unemployment rates. Thus, the primary objective of the Act was to provide job opportunities through the internal distribution of industries. In 1960, the Local Employment Act replaced 'Development Areas' with smaller 'Development Districts' and it made clear the concern about unemployment problems. This scheme provided a 10 per cent grant to new investment in the Development Districts. In 1966, the new Labour government set forth the Industrial Development Act, which again set new enlarged Development Areas. Financial support was strengthened, so that 40% and 25% of grants were provided for new investment in equipment and buildings respectively. Haynes and Dignan [1988] synthesised the empirical studies about the effects of these policies from some empirical studies. They pointed out that percentage changes in manufacturing employment within the British sub-regions were insignificant before 1958, but significantly positive for the period 1958-66, because of the increased subsidy. This increase was brought about by immigrant firms. A regression study found that unemployment was a negative factor preventing a firm to move into Development Areas. The IDC controls were the most important factors in generating employment in backward regions, not only manufacturing jobs, but also spin-off jobs in the service sector provoked by multiplier effects. The kind of jobs created in the Development Areas were mainly unskilled and semi-skilled with a greater dependence on cheaper female labour than in the core areas.

Ireland

In the 1950s, while Ireland's economy was experiencing a recovery from the last world war, "the western half of the country, characterised by a profusion of small farms, an inhospitable environment, and heavy out-migration, was scheduled as a block of 'Underdeveloped' Areas" [Breathnach, 1982, p36]. In the late 1950s, industrial grants were introduced to encourage industrialisation in the Western region, and the growth pole concept emerged. In 1969, the so-called Buchanan report, which was obviously influenced by the growth pole concept, proposed that the new industrial employment policy should be concentrated in nine urban centres, which consisted of two national growth centres and seven regional centres. However, this argument was abandoned by the government whose party needed votes in the coming election from depressed areas. Instead, an alternative plan, presented by the Industrial Development Authority, focused rather on a balanced development policy, whose objective was the dispersal of industries. In this scheme, industrial plants were built in 271 separate locations, and half of the new plants were located in areas of less than 3000 population [Breathnach, 1982, p39]. With the national economic policy which opened Ireland's economy to the world during the 1960s, this policy was also supported by investment from overseas and it worked well to boost the economic growth of the Western region. The population has spread throughout the country, pulled by the spreading industrialisation. The West and the Northwest received a benefit from this scheme, although it was accompanied by a decline in the relative position of the Midlands region.

Japan

Mizushima, an integral part of Kurashiki City in Okayama Prefecture, in Japan used to be an underdeveloped area before the Korean war, but the Korean war in the 1950s accelerated its industrialisation [Lo, 1978]. Mizushima had been considered as an industrial centre since the first attempt made by "the Mizushima Coastal Industrial District Development Programme" (1952). The Law for the Promotion of New Industrial City Construction in 1962 intensively promoted a growth pole approach to the development of the Kurashiki - Okayama city area and Mizushima was put into the core industrial area as the growth pole. Within this scheme, infrastructure for industrialisation, such as dredging of waterways, redevelopment of harbour areas, and railroads were emphasised to support the industrialisation. It encouraged the development of petroleum and steel industries and other import-export oriented heavy industries from the very beginning. Thus, the main objective was obviously to boost national exports rather than regional development. In the implementation process, the Okayama Prefecture Government took "a leading role in seeking assistance from corresponding ministries of the central government and were also deeply devoted to the promotion and establishment in the Mizushima area of branch factories by the leading industrial firms" [Lo, 1978, p29]. Lo [1978] argued that the development of Mizushima was a typical case, in that private manufacturing investment followed the public infrastructure investment and it brought about increasing job opportunities. The result was quite successful during the 1960s: in Mizushima, manufacturing output increased 10.7 times, while manufacturing employment increased 4 times; in Okayama Prefecture, manufacturing output increased 3.6 times while employment increased 1.5 times. The basic industries in the Mizushima complex created more output impacts on the labour intensive sectors than on the capital intensive sectors. In 1970, Mizushima reached the peak of its population growth and then decreased, and "this might suggest that industries in Okayama Prefecture outside Mizushima had started their own growth process" [Lo, 1978, p36]. However this rapid growth of the Mizushima complex was accompanied by the corresponding decrease in growth rates of the Northern urban centres. Nevertheless, Lo [1978] concluded that "the spread effect of the manufacturing sector began to dominate the polarization effects by 1968-9 and provided a growth impetus to the other urban centres" [Lo, 1978, p43].

One of the most recent arguments about growth pole theory is "Just-in-Time manufacturing". This idea originated from the case of a Japanese automobile manufacturing system. A famous Japanese car manufacturing company located its industry base at a town with population size of 345,000 and actually this town subsequently changed its name to this company's in 1959. About 70% of the citizens in this town work for organisations related to Toyota. The Toyota company subsidises most of the social welfare for the citizens, such as hospitals and recreation activities, and even operates some schools from Kindergarten to technical universities. Citizens and local governments are quite favourable to the existence of this 'big-brother' like enterprise. Although the size is not like so large, Japan has a lot of similar cases, e.g., Hitachi in Hitachi city. This traditional Japanese style was widely applied in North American and European motor industry localities, and it has attracted academic interests. Mair [1995] explained that this strategy is "characterized by emphasis of strong local linkages - in growth pole fashion - among manufacturers, workforce and local governments, and is accompanied by a set of specific policy instruments relating to attraction of new companies, provision of infrastructures and job training, and efforts to restructure local industrial organization" [Mair, 1995, p207]. Although it is possible to argue that this strategy is quite far from the traditional growth pole strategy concept, there is no reason to deny this widely used strategy of concentrated and localised industrial complex in the context of the growth poles concept. Since this strategy is mainly used by private companies who seek only their own profits, one critical question is whether this strategy can be implemented by the government or any public authorities as "a regional development policy".

Korea

The rapid economic growth of post-war Korea has left a problem of heightened interregional inequalities. 25% and 9% of gross output was concentrated in Seoul and the Busan area respectively [Kim, 1978]. The value added disparity was widened by 9.5 times at the mean deviation during the period 1962-1972. In this context, the Korean government introduced a new development plan in 1972, in which the industrialisation was to continue but accompanied with rural development and comprehensive land development. This plan included the construction of an industrial complex in the Southern coastal area and the establishment of a new growth centre, Ulsan Town, as a special industrial area which was connected to the Busan industrial area. This regional development, aimed at decentralising industries away from Seoul, has been successful to a certain extent. In Ulsan, the average annual population growth rate was 7.4% during the period 1962-1974, with stable in-migrants. The amount of products and exports of Ulsan City increased by about 300 times during the same period. Twenty-two petrochemical plants, fifteen non-petrochemical plants, and three power plants have been located in the industrial complex [Kim, 1978, p72]. In terms of economic development, this project was quite successful, however, Kim [1978] pointed out some problems with those projects. Firstly, industrial complex cities in the southern coastal area did not establish expected linkages with rural hinterlands, so there was little spread (diffusion) effects of growth. Secondly, the urbanisation of Seoul did not appear to be halted, and an imbalance has still remained. Thirdly, industrialisation has reduced agricultural and fishery production in the southern coastal area, and environmental disruption has begun. Finally, the economic growth of the local economy has not brought about increased fiscal benefits to the local governments.

Developing countries

Brazil

Brazil had expanded regional disparities in development with its economic growth. According to the empirical study by Gilbert and Goodman [1976], which tested Williamson's hypothesis of a convergence of inter-regional income disparities in Brazil for the period 1939-68, it found that the statistical evidence was not supportive to Williamson's optimistic hypothesis and suggested that "regional income differentials have remained fairly stable" [Gilbert and Goodman, 1976, p129]. Besides this regional disparity, inter-personal income inequality was increased during the 1960s, especially in the urban areas. During this period, the Brazilian economy was operated mainly within its capital city, Sao Paulo, the second largest LDC city in the world after Mexico City. While the South-East regions developed, the North-East regions were left underdeveloped. At the end of the 1960s, the SUDENE program was implemented to promote development in Salvador and in rural development centres, such as Aratu, and to achieve some redistribution of capital out of the wealthy Southeast regions. According to Richardson and Richardson, successful results were "the promotion of small metropolitan poles within Sao Paulo state itself (e.g. , in the Plano de Acao area) and the transfer of the administrative capital to Brasilia" [Richardson and Richardson, 1975, p174]. Richardson and Richardson attributed this partly to some supplementary policies which helped the distribution of the effects, such as "an ambitious road construction program for the interior" and "complementary industrial and agricultural promotion schemes in nearby regions" [Richardson and Richardson, 1975, p174]. This indicates that a growth pole strategy generally works well within a comprehensive approach. On the statistical data, the annual growth rate of city population declined from 6% to 4% while the suburbs population remained steady at around 9%-10% during the period 1950-1980. The whole of metropolitan Sao Paulo's growth rate of population declined from 7% to 5%, while the hinterland was relatively steady at 6% to 5% during the same period. However, Townroe [1983] argued that this trend must be mainly a process of "polarisation reversal". He argued that all polarisation must be reversed at a certain point towards decentralisation, similar to Williamson's hypothesis. He pointed out that a main reason for this process is a diseconomy of over urbanisation (congestion, pollution, expensive goods prices, etc.), rather than the result of a regional policy.

Chile

In the late 1960s, Chile implemented the first comprehensive national plan for regional development based on the growth pole concept in Latin America [Richardson and Richardson, 1975]. In the strategy, three levels of growth poles were created : (1) a national pole centred around Santiago, (2) three multi-regional poles centred on Antofagasta, Valparaiso, and Concepcion, and (3) a series of regional growth poles corresponding to the twelve regions. The actual locations were chosen on the basis of the greatest development potential. Thus, some poles were functional poles, for example Arica for automobiles and electronics, Santiago for electronics, Concepcion for steel and petrochemicals and Punta Arena for petrochemicals. Boisier [in Richardson and Richardson, 1975] evaluated the earlier strategy, and concluded that its results were successful in Antofagasta, Valparaiso, and Talca, slightly good in Concepcion (the principal pole), and negative in Arica and Punta Areas where the policy created national enclaves. According to him, such "negative" growth areas had weak policy instruments. Moreover, they were "implemented too passively and abandoned too soon" [Richardson and Richardson, 1975, p172]. Actually, soon after this policy was implemented, the thought that "such a development strategy might have dire consequences for non-favoured regions ....not to be sustained in the long run" was widely recognised. In the early 1970s, a newly elected government which was more socialist, changed the direction of regional policy to anti-concentrative and integrative objectives.

Peru

Peru suffered from the high economic dominance of Lima-Callao, which dominated 70% of the national industrial activities with 20% of the population, and this situation called for a decentralisation policy. Geographically, Lima is located in the centre of a development axis on the Pacific coast and the other major urban centres developed around this axis (Chiclayo and Truljillo in the north and Arequipa in the south). This geographical set of urban systems was anticipated to be suitable for a growth pole strategy. Some of the growth centres were set at the existing urban centres, and public investments were concentrated in these areas. However, Richardson and Richardson argued that the growth poles strategy failed, and this was because "the industrial aspects of pole development have not been emphasized enough, policy instruments are weak, and no action has been taken to curb the growth of Lima" [Richardson and Richardson, 1975, p173]. Thus, Peru had to turn its policy to more direct decentralisation such as rural development.

Bolivia

Bolivia had planned a long-run growth pole in 1970 [in Richardson and Richardson, 1975]. This plan considered a comprehensive growth strategy with a complicated hierarchy of poles and centres: Cochabamba (the second city) was designated as a service pole, Santa Cruz and Oruno as industrial poles (petro-chemicals, iron and steel, and agro-industries in Santa Cruz and metals in Oruro), and the much smaller centres of Sucre and Tarija as agro-industrial poles, and so on. It is obvious that this strategy was based on functional development poles. However, it ended up just as a plan on paper. This decision seemed to be influenced by the trend of abandonment of the growth pole strategy by LDCs. Richardson and Richardson [1975] considered this choice was right, since the functional growth pole strategy was not suitable to the reality of the Bolivian economy, due to its heavy dependence on agriculture and mining, and because its manufacturing industries are light, small and labour intensive.

Colombia

Colombia also followed the same path. The 1969-72 plan pointed out the problem of "the absence of internalization of economies of specialization and economies of scale as a principal source of low levels of efficiency and competitiveness", so "it called for concentration of economic location in regions of high development potential" [Conroy, 1973, p372]. The selection of growth poles was based on the current urban system, intra- and inter-regional linkages, existing economic activities, and sectoral analysis of functional poles. Thus, the resulting policy consisted of "the superimposed functional hierarchy and the interest in concentration production" [Conroy, 1973, p372]. It stated that the development should be in the form of the 3 large cities (Medellin, Cali, and Barranquilla) through subsidies for propulsive industries and provision of industrial infrastructure. However, when the Borrero administration came to power in 1970, this strategy was abandoned. The new policy intended to improve the standard of living in the regions, i.e., the smaller and intermediate cities in the hierarchy were emphasised.

Venezuela

Venezuela implemented a growth pole strategy in order to achieve a balanced regional development. Using a green field site, the Ciudad Guayana project was to emphasise on heavy industrialisation by a government development authority 'Corporacion Venezolana de Guayana'.

According to Richardson and Richardson [1975], it resulted in Ciudad Guayana being developed as a frontier region into an integral part of the Venezuelan economy, and it also had linkage effects to its hinterlands, although it might have had adverse effects on the secondary pole of Ciudad Bolivar. However, this type of strategy caused a problem, in that the spatial integration of the national economy required an expansion in the number and size of other secondary cities, and for this purpose other supplement policies which encouraged the decentralisation of industries would be necessary. Otherwise, the development process with the growth poles would be halted by limited natural resources and the diseconomies of over urbanisation in the new growth areas.

Argentina

In the 1970-1975 development plan, Argentina warned of the high dependency on its capital, Buenos Aires, and argued that growth poles should be designated to boost the economy outside of Buenos Aires. In reviewing the plan, Richardson and Richardson [1975] argued that " they were poor choices" [Richardson and Richardson, 1975, p174]. Firstly, three of the poles were in Patagonia, but its population size was less than 2% of the national population. Secondly, the selection of poles was made on the basis of an import-substitution policy, so the resources may have been misallocated. Thirdly, in the selection of the poles, political and security motives seemed to dominate over economic and social objectives, so they were ineffective. Finally, they were too small and weak in terms of economic power to have a 'countermagnet' function.

Mexico

Mexico set several growth poles in the 1960s: the establishment of the first industrial complex in the south-east in the state of Yucatan; the promotion of an industrial complex in Tabasco, also in the south east; an ambitious plan of development of an industrial corridor from the Pacific to the Gulf linking these ports with Monterrey, Moncolova, and Durango; and the establishment of a smaller industrial corridor in the state of Jalisco in the West to relieve pressure on Guadalajara [in Richardson and Richardson, 1975]. This project resulted in the provision of a certain amount of improvement at the border zone in the Northern region. However, "the only previous successes had been a development of the border zones in the far north promoted by direct investment by the United States and geared to supplying industrial products to that market" [Richardson and Richardson, 1975, p174]. Actually, it is not clear how much the situation was improved by this policy since the post-1971 plans turned to a much more integrated strategy and dealt with the country as a whole.

China

Before 1949, industries in China were concentrated in the coastal areas of Manchuria. Most of the factories, motive power used and the manpower employed were concerned within the mechanised manufacturing operations. While those coastal areas enjoyed high industrialisation and rapid growth, interior cities outside of these growth areas were left backward, poor, highly dependent on primitive agriculture and with fragmented economies. In order to solve these problems, the First Five-Year Plan proposed 472 out of a total of 694 industrial projects to be placed in the backward interior provinces, but it did not succeed since, physically, China is too large to expect some linkage effects through the limited number of rural industrialisation projects. Thus, a form of the growth poles concept was introduced in order to achieve "a proper balance between intra-regional self-sufficiency and interregional specialization and exchange" [Johnson, 1970, p328]. Public investment was then distributed mainly among medium size cities which contained a population of 300,000 to 1 million. The result was that medium cities grew faster than larger cities, however, the number of small cities had actually decreased by 16% in the 1960s, and the number of large cities had increased by 67% in the same period. Thus, the Chinese regional policy which encouraged the growth of medium size cities was effective at helping to increase the increasing size of the city and cause more urbanisation. If the objective was to encourage urbanisation, that policy was successful, but since the objectives were decentralisation, this medium size policy resulted in adverse effects. Half of China's industrial capacity was still (in 1966) in three provinces (Kiangsu, Hopeh, Liaoning) as it was before 1949.

India

In India, the economy was highly concentrated in a few urbanised cities. According to Johnson [1970], the planners' first policy was to "cluster industrial estates on the outskirts of large cities, thereby unwittingly aggravating an already excessive urban polarisation and exacerbating acute problems of housing and sanitation" [Johnson, 1970, p337-338]. This centralisation left rural villages underdeveloped and poor, thus, the Indian Fourth Plan (1969 -74) implemented an egalitarian policy targeting a balanced regional development. The main scheme was investment subsidies (15%) to backward areas with less than 2,500 persons and attempted to establish a village centred industry. However, the result was "a scandalously wasteful underutilization of factory space, for the reason that a small center with less than 2,500 people simply could not supply the versatile manpower needed for a range of industrial enterprises and was utterly incapable of providing the requisite marketing, transport, repair, and credit facilities which a group of small shops would require" [Johnson, 1970, p338]. Mathur [1978] pointed out that 62% of the total investment subsidy utilised till the end of 1974 ended up in high income states, with most of it being invested in Maharashtra, Tamil Nadu, Karnataka, and Kerala, which contained only 20% of the total national population. There was no decentralisation of factories for the rural areas. Mathur argued that the backwardness of such rural areas was more serious than expected, and that the inflexible investment scheme (fixed 15% subsidy for any project) could not be utilised in poor region where sometimes 100% subsidies were required. Some improvements with this project were reported especially in the area of education and vocational training, but Johnson [1970] argued that such success cases were attributed mostly to people's efforts, and not to the project itself.

Algeria

In 1965-71, Algeria promoted a highly concentrated industrialisation growth policy based on the export of oil. Indeed, about 90% of total exports were dependent on oil during this period. With an abundant trade surplus from the oil exports, Algeria concentrated its public investment around the capital city to establish an industrial complex which consisted mainly of hydrocarbon industries and steel-manufacturing industries. However, as Sutton [1975] pointed out, this highly capital intensive industrialisation policy, which was relying on natural resources, brought about little effect on employment. Sutton argued "the policies of the 1965-71 period were thoroughly state capitalist, and the single-minded pursuit of centralism and industrialisation by such means has left a legacy of rural stagnation" [Sutton, 1975, p353]. Indeed, during this period, while Algeria had enjoyed high national economic growth, its regional imbalance was worsened and food shortages prevailed in the country because of a neglect of a proper agricultural policy. It is difficult to judge whether Algeria had a certain conception about regional policy during this period since "national rather than regional policy dominated into the 1970s" [Sutton, 1975, p356]. This is an example in which a growth pole was utilised purely as a national development objective, and it worsened the regional development in terms of balanced growth.

Mali

Mali has suffered regional disparities and a heavy flow of migrants to Bamako, an existing growth centre. In order to reduce the flow of migrants to Bamako and to create favourable grounds for diversified economic development, especially for the backward region, the Mali government set up a new growth pole in Djoliba between the Niger River and the Bamako-to-Guinea road. One of the important reasons is that "it was far enough away from the capital city so that it would not become a 'dormitory' satellite" [Johnson, 1970, p343]. Even though the growth pole was set simply in one area, Djoliba, the investment was made in broad aspects, such as the service sector (establishment of a market yard, a medical centre, and enlargement of schools, etc.) ; construction of new streets, new mills and other agricultural facilities; housing; and an industrial training program. The United States Agency for International Development reported that this project was successful since millet production in the Djoliba area had increased 2.5 times and that groundnut output had expanded over 5 times from 1962 to 1965, while about half of the increased agricultural output seemed to be provided by the project [in Johnson, 1970]. It is not clear whether or not this project brought about intra- and inter-regional disparities.

Kenya

Kenya has experienced a high concentration of population and economic activities in Nairobi and Mombasa, so that it has resulted in high migration and unemployment rates. Thus, Kenyan regional policies have always been aimed at the redistribution of economic activities and urban population. The Second Development Plan (1970-1974) clearly expressed that it has been following the line of a growth pole strategy. It identified seven growth centres (Nakuru, Kisumu, Thika, Eldoret, Kakamega, Nyeri and Embu), which were selected in terms of an expected potential of growth and they were given a high priority in public works. In particular, the first four were to be promoted as industrial centres. In this plan, the urban growth of Nairobi and Mombasa was not intended to be discouraged, however, soon the need for a more intensive plan was recognised. The Third Development Plan (1974-1978) added new growth centres, such as Kitale and Meru, and emphasised the discouragement of the growth of Nairobi, which absorbed more than three-fifths of the rural migrants. However, it does not seem that these policies were effective in reducing the prevailing urban growth. Richardson [1978] argued that "the growth center policies adopted in Kenya during the 1970s hardly merit the name since the designations were insufficiently selective, the centres were not integrated into an overall strategy for the country as a whole, and implementation was ineffective" [Richardson, 1978, p150]. He pointed out that one of the reasons of the failure is that the distribution of investment was too fair because it was distributed purely depending on the proportion of the population. This example shows that the growth pole strategy requires a careful consideration of the economic system rather than just focus on location issues.

Poland

The Konin region in Poland is a relatively new area which was developed during the period 1953-1967 in order to reduce the differences in the spatial structure of the Polish economy. Before this project, industries had been concentrated in the southern parts of the country, the so-called old industrial regions represented by Lodz, Poznan, Warszawa, and in the Baltyk districts. Power industries had been built only in the southern and south-western parts. Wojtasiewicz characterised Konin before the project as "over populated in relation to the amount of work available, had no industry, and agriculture was under-developed due to the large partition of farms, weak network of communication and transport, and a low standard of urbanization" [Wojtasiewicz, 1972, p223]. This area did not have good quality soil to enable the development of agriculture, nor was it suitable for tertiary sector development such as tourism or recreation centres. Therefore, in Konin, industrialisation was the only choice. By the industrialisation of this area, the objective was to encourage the general industrialisation of the central regions of Poland and to increase the employment in these regions. Actually, various industries were located in the Konin area, such as mining activities, the establishment of a large energy base (power plants), aluminium works and metal work factories, chemical industries, etc. Wojtasiewicz [1972, p224] mentioned "two out-standing points" of this project. Firstly, it provided a diversified branch structure of the economic system to the region. Secondly, it enabled the establishment of a large energy centre in the central part of Poland, and changed the general spatial structure of Poland's economy. This development of Konin contributed to the establishment of a new industrial development region of Konin-Keczyca-Inowroclaw. Employment was increased from 5,600 to 26,300 during the period 1948-68, with the most rapid increase occurring in the 1960s. This increase was attributed to a rapid growth of the urban population, and it caused a boom of housing construction in the Konin region. The industrialisation absorbed much agricultural labour, and there was heavy migration from the countryside to the urban area in the Konin region. Indeed, Wojtasiewicz [1972] pointed out that there were some symptoms that the countryside has been lagging since this project.

Chapter 5. Conclusions

This dissertation has discussed the growth pole strategies in a comprehensive manner, such as the theoretical issues, practical policy issues, and empirical evidence.

Regarding the theory, it appears that the concept of growth poles has been transformed and sometimes mistreated outside of the early work of Perroux. Perroux [1955] developed the concept of growth poles purely as a dynamic system explaining dominance processes in an abstract economic space. His first attempt was to construct a general framework of economic analysis which should not have been limited to just regional analysis. However, in the 1960s, some regional economists, represented by Boudeville and Hirschman, applied this concept as an analytical tool in regional planning. Academic discussion of growth poles which has followed, proceeded mainly with this limited concept. Regional planning based on this limited growth pole concept resulted in providing wrong policy instruments in various countries. An unhealthy trend of anti-growth pole sentiment was often formed and prevailed in the 1970s. Moreover, the discussion itself seemed to be abandoned along with this disappointment. Some economists, represented by Richardson and Richardson [1975], have been arguing that growth pole strategies were misconducted and mistreated in their application. In the current situation where regional problems still exist, the need to revise this concept in a more careful manner is recognised.

We then discussed the various objectives for which the growth pole strategies seem to have been applied. This discussion revealed the contradictory nature of the growth pole strategy, that a growth pole strategy cannot avoid causing new spatial inequality, while the strategy may reduce inequality in a spatial frame (inter-regional, intra-regional and urban-rural). For example, the growth pole strategy which targets to reduce inter-regional disparity, usually results in causing a new inequality within the planned region (urban-rural disparity). This discussion also pointed out the fact that the planner always has to face a severe choice between efficiency and equality, although it is possible that both objectives can be attained together in long run by a supplementary policy coping with such newly 'created' disparity.

We then reviewed some empirical studies, which confronted two types of difficulties: limited sources and problems of measurement. However, the empirical review of various growth pole strategies showed some valuable suggestions for the future.

Firstly, most of the growth pole strategies were abandoned or insufficiently implemented because of the ideological change. Since the 1970s, the development purpose was changed to consider more the eradication of poverty and the development of small cities. Therefore, this change should not be seen as an abandonment of the analytical framework of the growth pole theory, rather it abandoned the idea of the growth centred development. Indeed, there can be a growth pole strategy for rural development, e.g., agropolitan development. As discussed in the policy issues, the objectives between national (regional) growth and rural development are against to each other in terms of their anticipated effects. Thus, the need to study the effect of such rural development policy in a framework of growth poles is recognised.

Secondly, the growth pole strategy has been faced with a financial problem. The governments have tended to misunderstand that a growth pole strategy is just investment in a limited area, however in order to boost the agglomeration economies, other conditions, such as a good transport system and skilled labour are necessary. Thus, the growth pole strategies had been more expensive than the governments had expected during the early stages. However, it is not necessarily logical to abandon a growth pole strategy just because it may be too expensive. The government always has to choose the location of investment and at least consider the effects of a policy in a spatial structure. Thus, financial problems certainly provide a good reason to abandon the optimistic view of the growth pole strategy, however they should not be the reason to reject the spatial thinking of the effect of a policy. The growth pole concept still seems to be a useful framework of analysis, rather than constructing a new framework.

Thirdly, the empirical evidence has shown that the effects of regional policies are related to the condition of the national economy. Even if the government has the best regional policy, it will end up as a failure if it is without the backup of growth of the national economy. Indeed, the Appalachian project in the USA may have been impossible without the financial backup of the 1960s national economic growth. This can also be seen in the success cases of Japan and Korea. All of these cases show that the success of regional planning was related strongly to the success of national economic development. This indicates the limitation not only of the growth pole strategy, but also of all regional policies.

Finally, while regional planning, including growth pole strategies, has such a limitation, there is still a reason to adopt good regional planning. This can be seen in the USA case. There were two growth pole strategies, but they showed contrasting results, success and failure: one which had been proceeding carefully in a comprehensive manner resulted in showing a certain improvement, while the other has just suggested some movement of industries and suffered from financial shortages then resulted in achieving nothing in the end. This contrast tells us that no "trick" exists in regional planning, in other words, planning which has been made with careful analysis and supported by sufficient investment may be a success, otherwise, it doesn't bring about any effects, or sometimes it causes negative effects, and is therefore a failure.

This point is important when considering the story of growth pole strategies. The growth pole strategies have been treated in quite an insufficient manner in theory and in practice. The growth pole concept seems to have been utilised as a framework of analysis at first. Only a policy derived from a careful analysis of this framework with growth pole concept must be called a "growth pole strategy". Thus, the growth pole strategy must have various forms from large, industrial, urban perspectives to small, agricultural, rural perspectives and it depends on the characteristics of each regional problem. However, the academic trend in the 1960s seemed to skip the analysis stage and apply a fixed growth pole strategy to all cases. This is the origin of most of the failures. The future analysis of regional planning should avoid this inadequate attitude and return to the original thinking, where Perroux attempted to construct a general framework of economic analysis.

-----------------------------------------------------------------------

- References -

Allen, K. and MacLennan, M. C. (1970) Regional Problems and Policies in Italy and France. London: Allen and Unwin

Alonso, W. (1968) "Urban and Regional Imbalances in Economic Development", Economic Development and Cultural Change, 17, p1-14

Alonso, W. (1971) "The Economics of Urban Size", papers of the Regional Science Association, 26, p67-83

Baran, P. (1957) The Political Economy of Growth, New York, Monthly Review Press.

Berry, B. J. L. (1971) "Hierarchical diffusion: the basis of developmental filtering and spread in a system of growth centres." in N. M. Hansen (Ed.) (1971) Growth Centres in Regional Economic Development. New York, The Free Press.

Berry, B. J. L. (1971a) "City Size and Economic Development: Conceptual Sythesis and Policy Proglems, with Special Reference to South and Southeast Asia" in L. Jakobson and V. Prakash (Eds.), Urbanization and National Development. Beverly Hills, Calif.: Sage Publications

Boisier, S. (1980) "Growth Poles: Are They Dead", Unpublished report, UN Latin American Institute for Economic and Social Planning, Santiago

Boudeville, J. R. (1966) Problems of Regional Economic Planning. Edinburgh: Edinburgh University Press.

Breathnach, P. (1982) "The Demise of Growth-Centre Policy: The Case of the Republic of Ireland" in R. Hudson and J. Lewis (Eds.), Regional Planning in Europe (London Papers in Regional Science, 11), p35-56.

Buttler, F. (1975) Growth Pole Theory and Economic Development. Portsmouth: Lexington Books.

Bylund, E. (1972) "Growth Centre and Administrative Area Problems within the Framework of the Swedish Location Policy" in A. Kuklinski (Ed.), Growth Poles and Growth Centres in Regional Planning. Paris and The Hague: Mouton.

Cameron, G. C. (1970) "Growth Areas, Growth Centres and Regional Conversion", Scottish Journal of Political Economy, 17, p19-38

Conroy, M. E. (1973) "On the Rejection of 'Growth Center' Policy in Latin American Regional Development Planning", Land Economics, 49, p371-380

Coragio, J. L. (1973) "Towards a Revison of Growth Pole Theory", Viertel Jahres Berichte, 53, p289-308.

Darwent, D. F. (1969) "Growth poles and growth centres in regional planning: a review", Environment and Planning, 1 (1), p5-31. Reprinted in Friedmann, J. and Alonso, W. (1975)

Drewe, P. (1971) "Steps toward Action-Oriented Migration Research", Papers of the Regional Science Association, 36, p145-164

Estall, R. C. (1982) "Planning in Appalachia: an Examination of the Appalachian Regional Development Programme and its Implications for the Future of the American Regional Planning Commissions", Transactions of the Institute of British Geographers, 7, p35-58

Estall, R. C. (1985) "Stock Control in Manufacturing: the Just-In-Time System and its Locational Implications", Area, 17, p129-133

Frank, A. G. (1967) Capitalism and Underdevelopment in Latin America, New York, Monthly Review Press.

Friedmann, J. (1956) "Locational Aspects of Economic Development", Land Economics, 32, p213-227

Friedmann, J. (1963) "Regional Economic Policy for Developing Areas", Papers and Proceedings of the Regional Science Associatoin, 11, p41-61

Friedmann, J. (1966) "Regional Development in Post-Industrial Society", Journal of the American Institute of Planners, 30, p84-90

Friedmann, J. (1966) Regional Development Policy: A Case study of Venezuela. Cambridge, Mass.: MIT Press.

Friedmann, J. (1969) "The Future of Urbanization in Latin America: some Observations on the Role of the Periphery", Papers of the Regional Science Association, 23, p161-174

Friedmann, J. (1972) "A general theory of polarized development." in N. Hansen (Ed.) (1972) Growth Centres in Regional Economic Development. New York: The Free Press.

Friedmann, J. and Alonso, W. (1964) Regional Development and Planning: a Reader, Cambridge, Mass., MIT Press.

Friedmann, J. and Alonso, W. (1975) Regional Policy: Readings in Theory and Applications, Cambridge, Mass., MIT Press.

Friedmann, J. and Douglass, M. (1978) "Agropolitan Development: Towards a New Strategy for Regional Planning in Asia" , in F. C. Lo and K. Salih (Eds.), Growth Pole Strategy and Regional Development Policy: Asian Experience and Alternative Apporaches. Oxford: Pergamon Press.

Friedmann, J. and Weaver, C. (1979) Territory and Function: the Evolution of Regional Planning. London: Edward Arnold.

Gaile, G. L. (1978) "The concept of growth centres." unpublished report, Storrs, Conn.: Department of Geography, University of Connecticut.

Gaile, G. L. (1980) "The Spread-Backwash Concept", Reginal Studies, 14, p15-25

Gore, C. (1984) Regions in Question: Space, Development Theory and Regional Policy. London: Menthuen

Hansen, N. M. (1971) Intermediate-size cities as growth centers : applications for Kentucky, the Piedmont Crescent, the Ozarks, and Texas. New York: Praeger Publishers

Hansen, N. M. (1972a) "Criteria for a growth centre policy" in Kuklinski (Ed.) (1972) Growth Poles and Growth Centres in Regional Planning. The Hague: Mouton Publishers

Hansen, N. M. (1972b) "Growth Center Policy in the United States" in N M. Hansen (Ed.), Growth Centers in Regional Economic Development. New York: The Free Press.

Hansen, N. M. (1975) "An Evaluation of Growth-Centre Theory and Practice", Environment and Plannning, 7, p821-832

Hansen. N. M. (Ed.) (1969) Growth centers in regional economic development. London: The Free Press

Harper, R. A., Schmudde, T. H. and Thomas, F. H. (1966) "Recreation Based Economic Development and the Growth-Point Concept", Land Economics, 42, p95-101

Harrison, B. (1992) "Industrial Districts: Old Wine in New Bottles", Regional Studies, 26, p469-483.

Haynes, K. E. and Dignan, T. (1988) "Evaluating capital grants for regional development" in B. Higgins and D. Savoie (Eds.), Regional Economic Development: Essays in Honour of Francois Perroux. Boston: Unwin Hyman.

Helmsing, A. H. J. (1988) "Rural Industries and Growth Points Issues in An Ongoing Polity Debate in Zimbabwe, Working Paper Sub-series - Simbabwe Rural and Regional Planning - No.4. Hague: Institute of Social Studies.

Hermansen, T. (1972) "Development poles and related theories: a synoptic approach." in N. M. Hansen (Ed.) (1972) Growth Centres in Regional Economic Development. New York: The Free Press.

Higgins, B. (1972) "Growth Pole Policy in Canada" in N. M. Hansen (Ed.) Growth Centers in Regional Economic Development. New York: The Free Press.

Higgins, B. and Savoie, D. (Eds.) (1988), Regional Economic Development: Essays in Honour of Francois Perroux. Boston: Unwin Hyman.

Hilhorst, J. G. M. (1981) "Peru: regional planning 1968-1977; frustrated bottom-up aspirations in a technocratic military setting" in R. Stohr and D. R. F. (Eds.), Development from Above or Below? Chichester: Jorn Wiley

Hirshman, A. O. (1958) The Strategy of Economic Development. New Haven. Conn.: Yale University Press.

Hoover, E. M. (1969) "Some Old and New Issues in Regional Development" in E. A. G. Robinson (Ed.), Backward Areas in Advanced Countries. London: Macmillan

Hoover, E. M. (1971) An introduction to Regional Economics. New York: Alfred A. Knopf.

James, B. G. S. (1964) "The Incompatibility of Industrial and Trading Cultures: A Critical Appraisal of the Growth Poit Concept", Journal of Industrial Economics, 13, p90-94.

Johnson, E. A. J. (1970) The organization of Space in Developing Countries. Cambridge, Mass.: Harvard University Press.

Kim, An-jae (1978) "Industrialization and Growth-Pole Development in Korea: A Case Study of Ulsan Industrial Complex", in F. C. Lo and K. Salih (Eds.), Growth Pole Strategy and Regional Development Policy: Asian Experience and Alternative Apporaches. Oxford: Pergamon Press.

Krumme, G. (1972) "Development Centers and Central Places in West German Regional Planning Schemes", Review of Regional Studies, 2, p215-234

Kuklinski, A. (Ed.) (1972) Growth Poles and Growth Centres in Regional Planning. The Hague: Mouton Publishers

Kuklinski, A. (Ed.) (1981) Polarized development and regional policies : tribute to Jaques Boudeville. The Hague: Mouton Publishers

Lasuen, J. R. (1969) "On Growth Poles", Urban Studies, 6, p137-161

Lasuen, J. R. (1974) "A generalization of the growth pole notion." in R. S. Thoman (ed.) Proceedings of the Commission on Regional Aspects of Development of the International Geographical Union, Volume I, Methodology and Case studies. Canada

Lewis, W. A. (1954) "Economic development with unlimited supplies of labour", Mancheter School of Economics and Social Sciences, 22, p139-91

Lipton, M. (1977) Why Poor People Stay Poor: Urban Bias in World Development. Cambridge, Mass.: Harvard University Press.

Lo, Fu-Chen and Salih (Eds.) (1975) Growth pole strategy and regional development policy: Asian experience and Alternative approaches. Oxford: Pergamon

Mair, A. "New Growth Poles? Just-In-Time Manufacturing and Local Economic Development Strategy", Regional Studies, 27, p207-221

McCann, P. (1995) "Rethinking the Economics of Location and Agglomeration", Urban Studies, 32, p563-577

McCrone, G. (1969) Regonal Policy in Britain. London: Allen and Unwin.

Misra, R. P., Sundaram, V. K. and Parakasa Rao, V. L. S. (1974) Regional Development Planning in India: a New Strategy. Delhi: Viking.

Moseley, M. J. (1973a) "The Impact of Growth Centres in rural Regions", Regional Studies, 7. p57-75 and p77-94

Moseley, M. J. (1973b) "Growth Centres: A Shibboleth?", Area, 5, p143-150

Moseley, M. J. (1974) Growth Centres in Spatial Planning. Oxford: Pergamon Press.

Myrdal, G. (1957) Economic Theory and Urderdeveloped Regions. London: Duckworth.

Nichols, V. (1969) "Growth poles: an evaluation of the propulsive effects", Environment and Planning, 1, p193-208 (G. Econ)

North, D. C. (1955) "Location theory and regional economic growth", Journal of Political Economy, 63 (3), p 243-58 In Friedmann, J. and Alonso, W. (1975)

Parr, J. B. (1973) "Growth poles, regional development and central place theory", Papers and Proceedings of the regional Science Association, 31, p172-212

Parr, J. B. (1979) "Spatial Structure as a Factor in Economic Adjustment and Regional Policy" in D. Maclennan and J. B. Parr (Eds.), Regional Policy: Past Experience and New Directions. Oxford: Martin Robertson

Parr, J. B. (1981) "The Distribution of Economic Oppotunity in a Central Place System: Dynamic Aspects and Growth Poles" in A. Kuklinski (Ed.), Polarized Development and Regional Policies: Tribute to Jacques Boudeville. The Hague: Mouton

Parr, J. B. (1996) Growth Pole Stategies in Regional Economic Planning : A Retrospective View, draft report, Department of Social and Economic Research, University of Glasgow.

Perloff, H. S. et al. (1969) Regions, Resources and Economic Growth, Baltimore, Johns Hopkins Press for Resources for the Future Inc.

Perroux, F (1950) "Economic Space: Theory and Application", Quarterly Journal of Economics, 64, p89-104, reprinted in J. Friedmann and W. Alonso (1964) Regional Development and Planning: a Reader. Cambridge, Mass.: MIT Press, p21-36

Perroux, F (1955) "Note sur la notion de pole de croissance", Economie Appliquee, 8, p307-320. or in I. Livingstone (1979) Development Economics and Policy: Selected Readings. London: George Allen & Unwin

Perroux, F (1961) "La firme motrice dans une region et la region motrice", Cahiers de l'ISEA, Series L, No.9.

Perroux, F (1968) "Les investissements multinationaux et l'analyse des poles de developpement et des poles d'integration", Revue Tiers Monde, 19, p239-65

Perroux, F (1988) "The Pole of Development's New Place in a General Theory of Economic Activity" in B. Higgins and D. Savoie (Eds.), Regional Economic Development: Essays in Honour of Francois Perroux. Boston: Unwin Hyman.

Reid, N. (1995) "Just-In-Time Inventory Control and the Economic Integration of Japanese-owned Manufacturing Plants with the County, State and National Economies of the United States", Regional Studies, 29, p345-355.

Richardson, H. W. (1975) Regional Development Policy and Planning in Spain. Farnborough: Saxon House.

Richardson, H. W. (1978) "Growth Centers, Rural Development and National Urban Policy: A Defense", International Regional Science Review, 3, p133-152

Richardson, H. W. (1980) "National Urban Strategies in Developing Countries", Urban Studies, 18, p267-283.

Richardson, H. W. and Richardson, M. (1975) "The relevance of Growth Center Strategies to Latin America", Economic Geography, 51, p163-178

Rosenstein-Rodan, P. N. (1943) "Problems of industrialization of eastern and south-eastern Europe", Economic Journal, 52, p202-11. Reprinted in A. N. Argarwala and S. P. Singh (Eds) (1963) The Economics of Underdevelopment. New York: Oxford University Press.

Rostow, W. (1971) The Stages of Economic Growth: A Non-Communist Manifesto. Cambridge: Cambridge University Press.

Salih, K., Pkkasem, P., Prantilla, E. and Soegijoko, S. (1978) "Decentralization Policy, Growth Pole Apporoach, and Resource Frontier Development" in F. C. Lo and K. Salih (Eds.), Growth Pole Strategy and Regional Development Policy: Asian Experience and Alternative Apporaches. Oxford: Pergamon Press.

Santos, M. (1979) The Shared Space. London: Methuen.

Schultz, T. W. (1951) "A framework for land economics - the long view", Journal of Farm Economics, 33, p204-15

Schumpeter, J. A. (1939) Business Cycles: A Theoretical, Historical and Stastical Analysis of the Capitalist Process, abridged, with and introduction by Rendig Fels (1964), New York: McGraw-Hill

Sen, L. K. (Ed.) (1978) Readings on Micro-Level Planning and Rural Growth Centres. Hyderabad: National Institute of Community Development

Slater, D. (1975) "Underdevelopment and spatial inequality: approaches to the problem of regional planning in the Third World", Progress in Planning, 4 (2).

Stohr, W. B. and Todtling, F. (1977) "An evaluation of regional policies experiences in market and mixed economies." in N. M. Hansen (Ed.) (1978) Human Settlement Systems: International Perspectives on Structure, Change and Publick Policy. Cambridge, Mass.: Ballinger.

Stohr, W. B. and Todtling, F. (1979) "Spatial equity - some antitheses to current regional development doctrine", Papers and Proceeding of the Regional Science Association, 38, p33-53

Stohr, W. B. (1981) "Development from below: the bottom-up and periphery-inward development paradigm". in W. B. Stohr and D. R. F. Taylor (Eds.) (1981) Development from Above or Below? The Dialectics of Regional Planning in Developing Countries, Chichester: John Wiley.

Sutton, K. "Algerilia : Centre-Down Development, State Capitalism, and Emergent Decentralization" in F. C. Lo and K. Salih (Eds.), Growth Pole Strategy and Regional Development Policy: Asian Experience and Alternative Apporaches. Oxford: Pergamon Press.

Townroe, P. M. (1979) "Employment Decentralization: Policy Instruments for Large Cities in Less Developed Countries", Progress in Planning, 10, p85-154

Townroe, P. M. and Keen, D. (1984) "Polarization in the State of Sao Paulo, Brazil", Regional Studies, 18, p45-54.

Wekwete, K. (1987) Growth Centre Policy in Zimbabwe - A Focus on District Centres", Working Paper Sub-series - Simbabwe Rural and Regional Planning - No.2. Hague: Institute of Social Studies.

Williamson, J. G. (1965) "Regional Inequality and the Process of National Development: A Description of the Patterns", Economic Development and Cultural Change, 13, p3-45

Wojtasiewicz, L. (1972) "Konin industrial region in the economic and social development of Poland" in A. Kuklinski (Ed.), Growth Poles and Growth Centres in Regional Planning. Paris and The Hague: Mouton.