Key words: central place theory, downtown shopping areas, hierarchy of shopping ... behaviour of retail activities, retailing, upward hierarchical movement.
GeoJournal 45: 327–336, 1998. © 1998 Kluwer Academic Publishers. Printed in the Netherlands.
327
Spatial dynamics of retail structure and the venerable retail hierarchy Johan G. Borchert Faculty of Geographical Sciences, Utrecht University, PO Box 80115, 3508 TC Utrecht, The Netherlands Received 28 May 1998; accepted in revised form 5 October 1998
Key words: central place theory, downtown shopping areas, hierarchy of shopping centers, large-scale retailing, locational behaviour of retail activities, retailing, upward hierarchical movement.
Abstract The retail system is continuously adapting itself to changing demographic conditions, consumer behavior and economic conditions. Existing retail units, getting larger in order to achieve economics of scale, and new forms of large-scale retailing do not fit easily in the traditional spatial pattern of retail concentrations. Neverteless the spatial distribution of retailing in urban areas, as schematized by Berry thirty-five years ago, still has its value. Through an upward hierarchical movement the retail system has adapted itself to an inflexible spatial system, in which downtown has still a dominant position. Future developments, however, may weaken the position of central shopping areas in favour of peripheral locations.
Introduction Thirty-five years ago Berry (1963) published his well known classification of shopping centers and business configurations (Figure 1). The classification distinguished three types of concentrations: a set of nested hierarchical centers based on the principles of Christaller’s central place theory, commercial ribbons mainly based on traffic flows, and specialized areas based on mutual functional connections, requirements of comparative shopping or specific locational demands of types of business not easily accommodated in the traditional hierarchy of centers. The model resulted from studies in Chicago, but was soon accepted as a universally applicable general model. Although present day textbooks still discuss the model at great length, many consider it as outdated (‘Isn’t it time [. . . ] that Berry’s Chicago study was given a decent burial?’ Brown, 1992, p. 461). With greater consumer mobility and new forms of retailing emerging continuously, the concept of a hierarchy of shopping centers in particular is considered obsolete. The purpose of this contribution is to demonstrate that in spite of the overwhelming dynamism in modern retailing, the spatial pattern in urban areas can still adequately be described by Berry’s model. It will be demonstrated that this also applies to the nowadays controversial hierarchical principles underlying part of the model. The spatial configuration of retailing in The Netherlands will be used as an example. Special attention will be paid to recent developments and to urban policy and legislation concerning inner cities and large scale peripheral retailing. The situation in The Netherlands may represent a special case, as the dense settlement system leaves less opportunities for peripheral retailing, and also because urban planning may be more stringent than in other countries. Nevertheless,
there are strong indications that also in other (West)european countries there is a revival of the position of city centers in the retail structure. Dawson (1995, p. 25) has pointed out that future developments in retailing may well cause a shift back to increased retail investments in town centres and central cities (see also Knee, 1988, Howard et al., 1990, Department of Environment, 1992): ‘As consumer demands fragments, so central locations become the most efficient place to satisfy this varied, often specialist, demand’. Although the retail system in every country has its own peculiarities, it is assumed that many of the trends discussed can be generalized, and are applicable in the modern Westeuropean urban environment.
Retail dynamism The retail system is changing continuously. Demand and supply driven factors are responsible for this dynamism. On the demand side, consumer preferences and shopping behavior are changing rapidly, and in addition demographic transitions and changing living conditions are taking place within the service areas of shops and shopping centers. On the supply side, retail units have to respond to changing economic conditions and increasing competition. These two sets of forces not only have their effects on existing retail businesses, but also give life to new elements enriching the complexity of the retail structure. In a highly planned urban environment, the dynamic retail sector is, however, faced with a spatial structure of retail facilities and a planning system that lack the flexibility to accommodate the rapidly changing requirements. In other words, there is a conflict between the needs of a dynamic retail system and the existing spatial conditions. By nature, adaptations of spatial conditions and planning regulations
328 to meet the needs of the retail system are only gradual and partial, as there are many other interests besides retailing to take care of. Nevertheless the spatial structure of retailing is changing gradually, although sometimes with some dramatic short term changes. Does this imply that the generalizations about the spatial structure of retail activities formulated in the past are outdated? Do we need a new model to describe the spatial pattern of retailing in the 21st century?
Shifting retail locations in a fixed hierarchical system One of the main characteristics of Berry’s model is a set of hierarchical shopping nucleations based on the principles of the central place theory. Berry used a fivefold classification from small convenience centers to metropolitan CBD’s (Figure 1). Many consider this hierarchical system as obsolete and tend to reject all ideas of a hierarchy in a modern urban environment. However, even if we admit that a rigorous hierarchy of shopping centers is a phenomenon of the past, we still find such hierarchies in many cities as a legacy from former principles of urban planning. The real question then is how the retail business adapted itself to this inherited spatial system of centers, which because of inertia is not easily adapted to changing circumstances. The retail system of Dutch cities still has a traditional hierarchical structure, based on postwar planning principles. Before World War II there was almost no planning of shopping facilities; shopping developed spontaneously in the central parts of inner cities and along main streets connecting the inner city with outlying districts. In urban extensions built in the 1930s some of the main street corners were designed to accommodate a few shops. Overall this prewar shopping system was hierarchical in nature. After World War II in new housing estates purpose built shopping centers were planned strictly according to the rules of a hierarchical system. This was especially the case with many larger urban extensions realized from the 1960s onwards. In 1963 a committee was appointed to investigate common planning practices and to report on necessary planning procedures in order to guarantee a well balanced approach to retail planning. As a result guidelines were published by the Netherlands Institute for Physical Planning and Housing (NIROV, 1971) to assist urban planners who had no special knowledge of retail planning. The schematic guidelines made it simple to decide how many hierarchical strata were needed for different sizes of urban extensions. For example, for planning areas comprising 25 000 to 125 000 dwellings a three-tier system was advised with a total of 0.8 m2 of sales floor area per inhabitant. Although it was emphasized that these rules-of-thumb should not be applied as a blue print, nevertheless in many cases these rules were followed blindly. As a consequence the spatial structure of shopping centers in urban areas in the Netherlands follows more closely a hierarchical pattern than in most other countries. At the same time planning regulations prevented retailing from developing outside centers zoned for retail activities.
The spatial pattern of retail concentrations cannot easily be adapted to the changing conditions confronting the retail trade. Especially at the lower end of the hierarchy, two developments jointly create problems. In the first place retail units generally try to achieve an increase in size in order to be able to compete in efficiency and operating costs with larger retail units. An important stimulus for expansion is the wider choice consumers are looking for when their spending power is increasing. Retailers not able to offer a wider variety very soon find themselves in serious problems. The effect of this has been that the average shop at present is two to three times larger than thirty years ago (Table 1). At any given location expansion in size and turnover of a retail unit is only possible when the market area is increasing in size as measured by number of consumers or spending power. Unfortunately a second trend runs counter to the requirements of expanding retail units: even if consumers enjoy improving budgets, the amount spent in retailing at best increases marginally in absolute terms, and normally decreases in relative terms. Still worse, decreasing fertility rates and changing household composition combine to produce a continuing decrease in the number of people per dwelling. Whereas retail units need an increasing market size, in effect at any given location the market is continuously diminishing, both in number of consumers and in retail spending. The combination of both trends has enormous effects. A market area of a given size as expressed by number of dwellings can support in 1997 only half the number of shops supported thirty years ago (Table 2). For food retailing the situation is even more dramatic: the number of shops supported at present is less than a third of that thirty years ago. Under these circumstances it is evident that the spatial pattern created in the past no longer meets the present situation. What happened to the hierarchies of shopping centers? Moving upward in fixed hierarchies The spatial system of hierarchical shopping centers is highly inflexible: planned shopping centers normally cannot easily be enlarged. Even if they could, they would be in a wrong location, as a larger shopping center needs a larger market size and consequently a more central location to serve this larger market. Unfortunately, shopping centers cannot be moved around. However, shops can. By moving upward to the next hierarchical level retail branches gain access to a market size no longer available at the level where they used to be. The upward hierarchical movement (Figure 2) can be accomplished by a retail unit actually moving to a higher level center, however the process is better thought of as a gradual development resulting from startups of new units and the death of existing shops. The consequences of this process are fourfold. In the first place there is no reason to abolish the concept of a hierarchy of retail concentrations. As a matter of fact a large part of retailing is still accommodated in a hierarchy of centers largely created in the past. A recent handbook
329
Figure 1. Berry’s model. Table 1. Average size and turnover per annum of retail units
Food - Sales area in m2 - Turnover per m2 sales area (Dfl) - Turnover per retail unit (1000 Dfl) Non-food - Sales area in m2 - Turnover per m2 sales area (Dfl) - Turnover per retail unit (1000 Dfl)
on retailing published by the authoritative Institute of Small and Medium Business (IMK, 1997) presents its data using a traditional hierarchy of shopping centers almost identical with the one proposed for Dutch retailing by Bak twenty five years earlier (Bak, 1971). Of the total sales floor area 73% is to be found in these shopping centers (Figure 3). It is true, however, that the functional composition of each level in the hierarchy has changed to reflect present market conditions. Many retail branches present at the lowest level in the past have moved to intermediate levels, and those of the intermediate levels have moved upward as well. Secondly, at the lower end of the hierarchy problems are to be expected (Tetteroo 1989). As a matter of fact many of the smallest centers have already lost their retail functions completely. A study in the city of Utrecht published in 1988 came to the conclusion that only seven out of the thirtyone neighborhood shopping centers in that city could be maintained in the future (Gemeente Utrecht, 1988). Similar surveys conducted in other cities show comparable results (CIMK, 1985; Kok, 1986). For smaller settlements and vil-
1997
1987
1978
1968
116 14 372 1667
91 12 680 1154
66 12 000 792
35 7000 245
164 5633 924
154 4750 732
119 4200 500
85 2350 200
lages one has been aware of these problems ever since the 1960s. The centers of the lowest hierarchical level present a serious problem indeed: in the Netherlands 76% of all shopping centers belong to this category (Table 3). Thirdly, although with a time lag, planning practice has responded to the changing circumstances by creating larger centers and a smaller number of hierarchical levels, or indeed only one level. In the fourth place, the upward hierarchical movement has consequences for the highest level in the shopping center hierarchy, the main centers. In the main centers of cities and towns there is a growing concentration of retailing. Even in a relative short time span this trend is obvious in all size classes of urban places (Figure 3). The main centers not only improved their share in total sales area, but even witnessed an increasing share of the sales area for food retailing. In most cases the result is an expansion of the main centers. As a consequence, rents are rising, thus causing branches with a lower rent paying capacity to move out to cheaper locations at the edge of the city. Furniture is a good example of a
330 Table 2. Calculation of number of shops supported by a housing area of 10,000 dwellings 1997
1987
1978
1968
10 000 2.44 24 400
10 000 2.61 26 100
10 000 3.04 30 400
10 000 3.64 36 400
Food - Sales floor per 1000 inhabitants (m2 ) - Sales floor area total (m2 ) - Average size retail unit (m2 ) - Number of food shops
221 5392 116 46
220 5742 91 63
198 6019 66 91
150 5460 35 156
Non-food - Sales floor per 1000 inhabitants (m2 ) - Sales floor area total (m2) - Average size retail unit (m2 ) - Number of non-food shops
816 19 910 164 121
810 21 141 154 137
721 21 918 119 184
490 17 836 85 210
Food + non-food - Sales floor per 1000 inhabitants (m2 ) - Sales floor area total (m2 ) - Average size retail unit (m2 ) - Total number of shops
1037 25 302 151 167
1030 26 883 134 200
919 27 937 102 275
640 23 296 63 366
Number of dwellings Inhabitants per dwelling Inhabitants per 10 000 dwellings
Figure 2. The upward hierarchical movement.
331
Figure 3. Share of main centers, supplementary centers and other locations in nett retail sales area.
332 Table 3. The hierarchy of shopping centers in The Netherlands, 1997 Shopping center hierarchya
Number of centers
Average number of stores
Sales area (nett) Average % natioin m2 nal total
food/ non-food
Market size (population, thousands)
Main centers 1 Village 2 Town 3 Small cities 4 Medium cities 5 Cities >70 000 inh
1010 164 130 79 39
9 39 76 144 497
900 5400 10 000 24 500 81 500
5.9 5.7 8.4 12.4 20.4
34/66 26/74 22/78 19/81 8/92
100
Total
1422
39
5767
52.8
17/83
Supplementary centers 1 Convenience 2 Neighborhood 3 Community 4 Regional
1077 164 64 20
12 41 80 151
1250 5500 11 000 23 000
8.6 5.8 4.6 3.0
53/47 33/67 23/77 16/84
Total
1325
21
2580
22.0
36/64
All retail centers Peripheral locationsb Dispersed retailing
2747 25
30 19
4230 27 200
74.8 4.4 20.8
23/77 0/100 16/84
100.0
21/79
Total a 1 = Nett floor area < 3750 m2
< 10 10–20 20–40 40–100
b Locations specifically zoned for large scale
2 = Nett floor area 3750–7500 m2
furniture retailing and some other branches 3 = Nett floor area 7500–15000 m2 (so called pdv locations, that is locations for 4 = Nett floor area 15 000–45 000 m2 perifere detailhandelsvestigingen). 5 = Nett floor area > 45 000 m2 Source: Author’s calculations based on data from IMK/REA, Handboek Detailhandel (n.d.).
branch which has shifted from the main centers to peripheral locations. The improved position of main centers in the Netherlands is somewhat exceptional when compared with other urban areas in Europe. It may be explained by the urban strategy developed after 1980, when the new towns policy came to an end. Hence-forth attention shifted to urban areas, introducing the ‘compact city’ concept. New housing was directed to urban centers, resulting after 1985 in a population increase in many cities that had seen the number of inhabitants declining in the previous period. This was accomplished by adopting higher building densities and occupying empty spaces within the built up area or at the edge of the city. This trend of urban concentration will continue: future urban developments will be confined to locations adjacent to larger cities as laid down in the Extension to the Fourth Report on Physical Planning (in Dutch: Vierde Nota Ruimtelijke Ordenig Extra, for short VINEX). These new building locations are called VINEX-locations, and will accommodate almost 500 000 new dwellings to be built by 2015. The largest VINEXlocation for 30 000 dwellings is Leidsche Rijn, southwest of the city of Utrecht (234 000 inhabitants in 1997). This policy of concentrated urban development certainly will stimulate the position of the main centers. The city of Utrecht is anticipating future developments with the Utrecht City Project, a large scale expansion and renovation of its central area.
Nevertheless the improving position of downtown centers may be only temporarily, as will be explained further on. The conclusion that the concept of a hierarchy of shopping centers is still of value today may come as a surprise for those impressed by new retail configurations which do not seem to follow the development path predicted by the central place theory (Klop and Haringsma, 1997). However, the basic principle formulated by Christaller is applicable as before: goods and services with a high frequency of consumption can be distributed from a denser network of locations than those responding to less frequent needs. Of course threshold values and the range of goods are changing continuously, but this in no way injures the value of the theory. Actually, Christaller was well aware of the changeability of these variables, which makes the central place theory suitable in dynamic situations. There is no reason why a concept such as threshold should be limited to traditional types of retailing and not extended to innovative forms of retailing. Some consumers may feel compelled to visit occasionally an imaginative center offering a wide variety of shopping and leisure facilities, but shopping trips to a straightforward supermarket in order to buy necessities of daily life will be more frequent. Phrased in terms of the theory, the threshold values for these forms of retailing are different, and therefore they are predestined for different hierarchical levels with their corresponding market sizes.
333 What has to be abandoned from the classical central place theory is the spatial component. Increased mobility releases the necessity of a regular spatial pattern. For various reasons less ‘central’ sites may be chosen for centers that in the past would have been confined to downtown locations. Eventually a spatial pattern of urban retail nucleations may arise that does deviate completely from the theoretical regular pattern. It should be stressed that the freer interpretation of the theory by Berry & Garrison (1958) emphasized the hierarchical structure of centers rather than their spatial distribution and the pattern of trade areas.
Berry’s hierarchy revisited From the discussion so far, it is clear that the traditional shopping center hierarchy has collapsed in the sense that the lowest levels have lost a good deal of their position, whereas the highest level takes an increasing share of the total shopping space. In the case of the Netherlands, the loss in the share of the supplementary centers was partly compensated by the increasing share of the main centers. Available data suggest that the main centers benefited by one third from the shift away from the supplementary centers, whereas two thirds contributed to non-center locations. Between 1993 and 1996 the share of the non-center locations almost doubled from 14.6 to 27% at the expense of the shopping center hierarchy (Figure 3). Taking these shifts into consideration the traditional model by Berry has been adjusted (Figure 4). The fivefold classification of hierarchical centers has been replaced by a threefold system. Rather than putting the hierarchy of centers in a neatly nested pattern, a freer form was deliberately chosen to indicate the greater freedom of location of these centers. The ‘ribbon’ component of Berry’s model has been maintained, although highway oriented ribbons seem to be a typical American phenomenon. Finally, the specialized areas component has been extended in order to accommodate emerging types of retailing and retail nucleations.
Future directions The evidence presented so far suggests a strong position of the main centers in Dutch cities. The situation in the Netherlands cannot be considered as representative in all respects, as the Dutch settlement system is characterized by a large number of small places: only eighteen cities have more than 100 000 inhabitants. The size of the largest city (Amsterdam 750 000 inhabitants) is modest even when compared with medium-sized cities in surrounding countries. Obviously, in a settlement system dominated by smaller cities a larger part of retailing is concentrated in inner cities, just because the number of inner cities is larger than in a settlement system dominated by a few large cities. There are however two trends which may cause a diminishing share of downtown retailing in the future. Both are unintentionally effects of government policy, and therefore specific for the Dutch situation.
Large-scale peripheral retailing Dutch retail planning policy has been very restrictive as far as peripheral large scale retailing is concerned (Borchert, 1995). Soon after large outlying superstores came into being in the early 1970s, measures were taken to prevent this form of commercial suburbanisation. Large-scale retailing had to be located in or adjacent to existing retail concentrations. As an exception to this general rule only a few branches were allowed to locate outside existing shopping areas on locations specifically designated, so called pdv-locations (pdv stands for perifere detailhandelsvestigingen). In one of the first bills on peripheral retailing in 1973 outlets dealing with goods presenting a fire or explosive hazard were mentioned along with the sale of cars, boats and caravans because of the large spaces required for these goods. In 1984 firms selling furniture and do-it-yourself and building materials were added, provided it could be shown that existing shopping areas could not accommodate them. A few years later furniture was extended to include home fabrics and home furnishings. In 1996 the sales floor area of all pdv-locations made up 4.4% of the total floor area (Handboek Detailhandel). Apart from the pdv-branches large scale peripheral retailing has been limited, although not completely lacking. The total sales area of large scale peripheral retailing outside pdv-locations can be estimated at six to eight percent of the total sales area in the Netherlands. Quite unexpectedly the policy concerning large scale peripheral retailing was relaxed in 1993 by the introduction of so called gdv-locations (gdv = grootschalige detailhandelsvestigingen) for stores exceeding 1500 square meters gross floor area. Permission to create this new type of retail location has only been granted to thirteen larger cities, which have been awarded a special status as an urban growth pole (stedelijk knooppunt) in the Fourth Physical Planning Report Extra (VINEX) . An additional requisite for zoning a gdv-location is that in consultation with neighboring local authorities a regional retail structure vision will be made. As expected, the new opportunities for large-scale peripheral retailing were met with an enthusiastic response from developers and some retail firms. One year after publication of the new guidelines an annual inventory of all new retail developments indicated that 40% (770 000 square meter) of retail space under construction or projected concerned gdv-locations. Interestingly enough the new gdv-projects were not limited to the thirteen cities mentioned in the guidelines. The municipalities concerned have been more reluctant to embrace the new possibilities. All have commissioned commercial research institutes to investigate the possibilities of gdv-developments. The methodology applied in these studies falls into two categories. In most cases total future retail spending is confronted with anticipated turnover of the existing and planned retail provisions. Only if there is a surplus large enough to warrant an additional floor space of a few tens of thousands of square meters, would a provisional decision in favour of a gdv-development be made and the effects on the existing retail network calculated. The other
334
Figure 4. Berry’s model adjusted.
approach starts from the opposite end with the supposition that a gdv with a certain mix of superstores be located at a specific location. Taking the typical superstore size in each branch the turnover is estimated. The next step is to calculate the origin of the turnover by distance zone, and therefore the effect the gdv might have in different zones of its tributary area. The decision for or against a gdv-development then depends on a political judgment, whether or not the anticipated effect on the existing retail system is acceptable. So far one gdv has been realized in Rotterdam (Alexandrium II, 16 superstores, 25 000 m2), The Hague, Amsterdam and Maastricht will follow, and the other cities decided not to develop gdv-sites for the time being. Potential gdvdevelopment can have an important effect on the spatial distribution of retail trade. As these policies have yet to come to fruition, it is difficult to assess their impact. There are indications that the widespread interest mentioned earlier has been tempered somewhat, probably because most local governments have no intention for realization in the short term. This has created a climate of uncertainty surrounding these new developments. Nevertheless some experts assume that gdv-retailing could eventually be a threat to the position of downtown shopping, although in the present state of affairs the coming of centers of the type CentrO or Metrocentre is not likely as they would violate the intentions (if not the letter) of the gdv-guidelines. An important factor to bear in mind is that in a small country like the Netherlands the number of important players in retailing is limited, and
most of them have invested heavily in downtown shopping locations in the recent past. On the other hand the new generation of gdv-centers is creating new opportunities for foreign retailers to enter the Dutch market. At present the future development of gdv-locations is uncertain. If the downtown areas become less attractive, the hidden opportunities of the gdv-locations may become more important.
Retail perspectives in inner cities The last twenty years have seen a revival of the inner cities. Urban regeneration not only reversed the decreasing population trend, but also gave rise to a variety of new activities in retailing, leisure and the catering industry. Inner city shopping areas in particular have taken advantage of the more liberal trading hours regulations during weekdays and the possibility for shops to open for a limited number of Sundays. The increased attraction of the inner cities is reflected by growing numbers of pedestrians. There are also developments threatening the position of the inner city as the most attractive retail environment. There are many threats (see Boekema et al., 1996), however, the most important problem to be discussed here is the car. In historic cities with a finely-woven structure of narrow streets car traffic and parking are at odds with the requirements of a pleasant shopping environment. In the 1970s pedes-
335 trianization of the main shopping streets was introduced successfully in many cities. Recently more drastic policies have been introduced in some cities in order to free larger parts of the inner city from traffic. An early example is the city of Groningen, where in 1976 a traffic system for the inner city (vcp = verkeerscirculatieplan) was adopted with a very limited access of cars. A more recent and controversial system was introduced in Utrecht on 1 November 1996. To prevent through traffic, the inner city was divided into six sections which could be entered from a ring road only. Interconnections between the sectors were disrupted, and within the sectors a complicated one-way system was introduced. The amount of parking spaces at street level was further limited. At the same time a park-and-ride system was introduced with a small parking fee which included use of the shoppers shuttle bus to the inner city. The case of Groningen is well documented. The effects on the retail business have been studied by Buit in a beforeand-after comparison. Shortly after the introduction of the new traffic system forty percent of the shopkeepers reported a decrease in number of customers, and a slightly higher percentage also had a lower turnover (Buit, 1979). Two years later two out of ten shopkeepers from the same sample claimed a smaller number of buyers where visiting their shops because of the new traffic system, whereas one out of ten maintained the new situation had brought more customers to their shops (Buit and Walen, 1981). In total only a small proportion of those interviewed considered the new traffic system as a success. The initial effects of the Utrecht case were more dramatic. According to an investigation conducted immediately after the introduction of the new traffic system twenty one percent of the inhabitants of Utrecht and twenty seven percent of those in surrounding municipalities stated to avoid future visits of the inner city. Rather than being the effect of the traffic measures in itself, part of this reaction must be attributed to the negative publicity surrounding its introduction. The official designation ‘car-barren-city’ for the new traffic system certainly frightened many prospective visitors. Some shopkeepers in the inner city claimed a loss of turnover of twenty five to thirty percent, whereas shopping centers in the vicinity enjoyed a remarkable increase in their business. On average the turnover of shops in the inner city during the three months from November 1996 until January 1997 was six percent lower than in a similar period one year earlier, whereas for the Netherlands as a whole there was an increase of three percent. Some lines of trade were more affected than others. Book shops and services did relatively well, food retailing suffered most, with a loss of 12%. A further investigation half a year after the introduction of the ‘car-barren-city’ demonstrated a shift in modal split: before the introduction thirty nine percent of the visitors came by car, afterwards twenty two percent, whereas five percent had shifted to public transport. Using the categorization of shopping activities by Carr (1990) research findings suggested that technical shopping was more affected than either expressive or recreational shopping. Moreover shops
selling voluminous goods or goods bought in quantities awkward to carry, as in the case of wine sellers, experienced negative effects. The two cases reported here are no isolated cases. Many cities pursue a policy to discourage visits by car to their downtown area. Insofar as these policies are successful, the result will be a changing environment for retailing in which some types of shops will prosper and others might gradually shift to other locations. As was demonstrated in the preceding sections this is not an uncommon process. It will only be accelerated and extended to forms of retailing so far still typical for inner city locations. The process will be most pronounced in the largest cities, which therefore will find it difficult to keep their competitive position compared with medium-sized cities. Already the monitoring system ‘consumer and retail trade’ found that consumer orientation to the twenty largest cities from 1995 till 1996 diminished in favour of smaller cities and the maximum distance travelled to larger cities dropped from seventeen to fourteen kilometers. Conclusion Berry’s model of shopping nucleations can still be used to adequately describe the functional differentiation of urban retailing, provided some minor alterations are carried through. The spatial implications of the model have to be relaxed, although inertia in spatial structure have forced retail activities to a partial adaptation of inherited patterns of shopping centers. The hierarchical component of the model has changed: retail activities have moved upward through the hierarchy. The lowest levels of the hierarchy have lost their position, whereas the highest level has gained in importance. For the Netherlands it is to be expected, that the recent reform of retail policy lifting the ban on peripheral large scale retailing, in combination with attempts by larger cities to discourage the use of cars for downtown shopping, may somewhat weaken the position of the highest level of the shopping hierarchy in the future. Nevertheless, the city center will remain the most important retail concentration in urban areas. This will be the case in many Westeuropean countries, as nowadays a major component of planning policy is to support retail activities in town and city centres (Guy, 1994, p. 83). References Bak L., 1971: Hoofdwinkelcentrum. Samson, Alphen aan den Rijn. Berry B.J.L., 1963: Commercial structure and commercial blight. University of Chicago, Department of Geography, Chicago (Research Paper 85). Berry, B.J.L. & Garrison W.L., 1958: A note on central place theory and the range of a good. Econ. Geogr. 34: 304–311. Boekema F., Buursink J. & Van der Wiel J., 1996: Het behoud van de binnenstad als winkelhart. Van Gorcum, Assen. Borchert J.G., 1995: Retail planning policy in the Netherlands. In: Davies R.L. (ed.), Retail planning policies in Western Europe, Routledge, London/NY, pp. 160–181. Brown S., 1992: Book review of Jones & Simmons, The retail environment. Progr. Human Geogr. 16: 461–462.
336 Buit J., 1979: Repercussies van het verkeerscirculatieplan Groningen voor binnenstadsbedrijven en binnenstadsmilieu; resultaten van een verkennend onderzoek omtrent de periode jan 1976 – febr 1978. Geografisch en Planologisch Instituut Vrije Universiteit, afdeling Planologie, Amsterdam. Buit J. & Walen D.J., 1981: Repercussies van het verkeerscirculatieplan Groningen voor binnenstadsbedrijven en binnenstadsmilieu; resultaten van de tweede fase van een verkennend onderzoek omtrent de periode january 1978 – february 1980. Geografisch en Planologisch Instituut Vrije Universiteit, afdeling Planologie, Amsterdam. Carr J., 1990: The social aspects of shopping: pleasure or chore? Roy. Soc. Arts J. 138, 189–197. CIMK, 1985: De levensvatbaarheid van kleine buurtwinkelcentra. Coordinerend Instituut Midden- en Kleinbedrijf, Diemen. Dawson J.A., 1995: Retail change in the European Community. In: Davies R.L. (ed.), Retail planning policies in Western Europe, Routledge, London/NY, pp. 1–30. Department of Environment, 1992: The effects of major out-of-town retail development. HMSO, London. DIS, 1996: Handboek DIS 1995/96; kengetallen voor de detailhandel. D&P Onderzoek en Advies/Vereniging van Kamers van Koophandel, Den Haag/Woerden.
Gemeente Utrecht, 1988: Een winkelhart van buurt tot binnenstad; detailhandelsnota Gemeente Utrecht. Gemeente Utrecht, Utrecht. Guy C., 1994: The retail development process; location, property and planning. Routledge, London. Handboek detailhandel, 1997: Instituut voor Midden- en Kleinbedrijf IMKREA, no place (Hoofddorp). Howard M. et al., 1990: Britain in 2010: Future patterns of shopping. Reg. Sci. Assoc. J 5403: 162–203. Klop P. & Haringsma J., 1997: Het gedroomde winkelcentrum; een collage over het verschijnsel winkelcentrum. Instituut voor Midden- en Kleinbedrijf IMK-REA, No place (Hoofddorp). Knee D., 1988: City centre retailing in continental Europe. Longman, Harlow. Kok R., 1986: Winkelstructuur ’s-Gravenhage Zuidwest. Gemeente ’sGravenhage, Gravenhage. Nirov, 1971: De winkelvoorziening in de ruimtelijke ordening. Samson, Alphen aan den Rijn (Publikatie 76 van het Nederlands Instituut voor Ruimtelijke Ordening en Volkshuisvesting). Tetteroo N.H.M., 1989: Herstructurering van winkelcentra. In: Ruimtelijke Verkenningen 1989; Jaarboek Rijksplanologische Dienst. SDU Uitgeverij, ’s-Gravenhage.
