Environment and Planning B: Planning and Design 2009, volume 36, pages 279 ^ 299

doi:10.1068/b34011t

The implications of e-shopping for in-store shopping at various shopping locations in the Netherlands Jesse W J Weltevreden ô

Netherlands Institute for Spatial Research (RPB), PO Box 30314, 2500 GH The Hague, The Netherlands; e-mail: [email protected]

Ton van Rietbergen

Section of Economic Geography, Faculty of Geosciences, Utrecht University, PO Box 80115, 3508 TC Utrecht, The Netherlands; e-mail: [email protected] Received 6 February 2007; in revised form 2 September 2008

Abstract. Thus far, the empirical literature on the impact of e-shopping on in-store shopping has paid scant attention to the implications of e-shopping for shopping centres. Using a nationwide sample of 3000 Dutch e-shoppers we provide more insight into this topic. Results indicate that city centres are most likely to face the substitution of e-shopping for in-store shopping, followed by city district centres. Surprisingly, village centres are less affected by e-shopping than city centres. Moreover, for neighbourhood and convenience centres the adverse effects of e-shopping are small. The probability of substituting e-shopping for in-store shopping at particular shopping locations is largely influenced by the extent to which people shop online, as well as personal and geographical factors.

Introduction Since the 1990s, transport and retail geographers have been interested in the implications of e-shopping for in-store shopping. E-shopping can be defined as searching for and/or buying consumer goods and services via the Internet (Mokhtarian, 2004). And with the growth of e-shopping has come an increasing number of empirical studies concerning this topic (Weltevreden, 2007). The main question that many of these studies has tried to answer is: To what extent does e-shopping complement or replace in-store shopping? So far, scholars have found evidence for complementarity as well as substitution, which to a large extent can be attributed to differences in research methodology (Weltevreden, 2007). And while the substitution ^ complementarity debate continues, there remains another important question that so far has not been frequently addressed: Which shopping locations are most influenced by e-shopping? Although many authors highlight that the impact of e-shopping on in-store shopping largely varies among shopping locations, so far there have been only a few empirical studies that address this question (eg Dixon and Marston, 2002; Weltevreden and Van Rietbergen, 2007). Not only academics, but also policy makers, shopping centre owners, and retailers, are interested in the consequences of e-shopping for the viability of retail locations (Foresight, 2000; Kolpron Consultants, 2001; Stec Groep, 2000). Since the potential that e-shopping holds for various products largely differs, it is also important to investigate which retail categories at which locations are most likely to be affected by e-shopping (Weltevreden, 2007). This is also hardly addressed in the literature. Using a sample of 3000 e-shoppers we investigate the extent to which e-shopping is substituted for in-store shopping among different shopping locations in the Netherlands. In addition, we scrutinise the impact of e-shopping on in-store shopping for twenty-seven retail categories. Furthermore, we distinguish between online searching and online buying behaviour, as both activities can have a different impact ô Current address: BOVAG, PO Box 1100, 3980 DC Bunnik, The Netherlands.

280

J W J Weltevreden, T van Rietbergen

on in-store shopping (Farag et al, 2007; Weltevreden, 2007). The objective of this research is to explore the degree to which online searching and online buying are substituted for the number of trips to and purchases at various shopping centres in the Netherlands. The implications of e-shopping for in-store shopping in a shopping centre context: a literature review When studying the impacts of e-shopping on in-store shopping, transport and retail geographers usually distinguish four effects: (1) substitution, (2) complementarity, (3) modification, and (4) neutrality (Weltevreden, 2007). This framework of potential impacts was initially developed by Salomon (1985; 1986) and was further extended by Mokhtarian (1990; 2002; 2004). In this paper we will focus on substitution of e-shopping for in-store shopping, which may adversely affect the viability of shopping centres. Whereas American researchers are particularly concerned with the implications of e-shopping for shopping malls (eg Bean, 2000; Hendershott et al, 2000; Hernandez et al, 2001; Miller, 2000), European scholars are more anxious about the implications of e-shopping for retailing at city centres (eg Burt and Sparks, 2003; Dixon and Marston, 2002; Dixon et al, 2005; Gillespie et al, 2001). This difference in research focus may be explained by variation in the retail context between the US and many European countries (Aoyama and Schwartz, 2004). In the proceeding discussion we will give a brief review of the main assumptions in the literature, as well as a review of the empirical studies. Expectations

Scholars contend that the implications of e-shopping differ among shopping locations. In the Netherlands, a report by the Stec Groep (2000) has been influential. Despite a decline in some high-impact sectors, such as books, CDs, and travel agencies, the Stec Groep expect Dutch city centres to benefit from e-commerce. The rise of multichannel retail formats is expected to improve the attractiveness of city centres as places for recreational shopping. Neighbourhood and convenience centres are also assumed to profit from e-shopping. Their location in residential areas makes them excellent locations for the establishment of collection points for online orders. Furthermore, large-scale retail locations (eg furniture districts), are expected to experience few adverse effects of e-shopping. However, they may use the Internet as a marketing tool to reach potential customers. Like city centres, city district centres are locations for comparison shopping. However, it is assumed that these locations are more likely to be adversely affected by e-shopping than are city centres. Contrary to city centres, city district centres lack the ambience, the size, and the variety of stores to become attractive locations for recreational shopping, which may offset any negative impacts of e-shopping (Stec Groep, 2000). Various Dutch policy reports have adopted these assumptions (eg TNO Inro, 2002; Van Oort et al, 2003). UK scholars have also speculated about the implications of e-shopping for shopping centres. Gillespie et al (2001) argued that smaller (rural) town centres in particular, which have already been detrimentally affected by out-of-town retailing, are most likely to suffer from e-shopping. Large(r) city centres, on the contrary, can more easily shift their emphasis from retail to service and leisure functions. Other scholars developed assumptions by questioning retailers, real estate investors and developers, and their advisers about their expectations concerning the implications of e-commerce for shopping centres in the UK (Dixon and Marston, 2002; Dixon et al, 2005). According to these surveysöconducted in 2000, 2001, and 2003öpractitioners expect that large city centres and out-of-town regional shopping centres are least at risk, followed by

The implications of e-shopping for in-store shopping

281

neighbourhood or city district centres. They further assume that comparison shopping is most at risk, and large cities will be more immune to e-shopping than rural towns. A similar kind of research was conducted in Germany by Wengler (2005). German practitioners expect retail locations in the city centre to benefit most from e-commerce, while for neighbourhood and city district centres a neutral impact is assumed. Shopping centres at the edge of cities, and centres in rural locations in particular, are expected to be the most likely to suffer from e-commerce. For the US, Hendershott et al (2000) expect no decline in the demand for mall space, as the Internet is likely to draw more heavily from catalogue sales. Empirical evidence

Despite the growing number of empirical studies concerning the implications of e-shopping for in-store shopping, there are few that take into account the context of shopping centres. Borgers et al (1991) and Timmermans et al (1991) modelled the potential adoption of teleshopping and its consequences for retail locations, using a sample of 186 residents in two neighbourhoods in Eindhoven, the Netherlands. Borgers et al (1991) showed that, depending on diffusion pattern (uniform versus top town), teleshopping strengthens the position of higher order to lower order centres in the retail hierarchy. Timmermans et al (1991) found that, if the attraction of shopping centres deteriorates, the probability of adopting teleshopping increases. They also found that the lower the accessibility to shops in a neighbourhood, the higher the probability of adoption. Hernandez et al (2001) studied multichannel shopping activities of Canadian mall shoppers (N ˆ 1937). Results indicated that, even for products with high online sales contributions (eg travel and computer hardware), viewing online and then buying in-store generated on average three times more sales than direct online purchases. Dixon and Marston (2002) interviewed 450 shoppers in a town in southeast England. They found that almost two thirds of the e-shoppers (28% of all respondents) indicated that some or all of their online purchases had replaced a purchase which otherwise would have been made in the town centre. Using a sample of 957 German shoppers from Hannover and Leipzig, Hassenpflug and Tegeder (2004) examined the implications of e-shopping for in-store shopping in the city. Results showed that e-shoppers shop more frequently in the city centre than non-e-shoppers. For city district and neighbourhood centres they did not find large differences in the frequency of shopping trips between the two types of shopper. Schellenberg (2005) interviewed German high school students (N ˆ 1174) and their parents (N ˆ 881) in the Heidelberg and Neckar ^ Odenwald ^ Kreis region, and 625 German Internet users about their e-shopping activities. The results showed that approximately 25% and 18% of the e-shoppers in his samples made fewer purchases in the city centre and in rural shopping centres, respectively, owing to e-shopping. Using a sample of 3218 Internet users that shop at eight city centres in the Netherlands, Weltevreden (2007) scrutinised the implications of e-shopping for city centre shopping. Results showed that in 2004 no more than 22% of the e-shoppers indicated that e-shopping had caused them to make fewer trips to and purchases in the city centre. However, many Internet users already stated that they used the Internet in a complementary manneröthat is, searching for product information online before making a city centre purchase, and vice versa. Furthermore, Weltevreden concluded that these effects largely varied among retail categories and types of e-shoppers. In an earlier paper we (Weltevreden and Van Rietbergen, 2007) investigated how consumers' perceived city centre attractiveness influenced the relationship between e-shopping and city centre shopping. We identified three factors that relate to city

282

J W J Weltevreden, T van Rietbergen

centre attractiveness: shopping attractiveness, leisure attractiveness, and accessibility attractiveness. The results indicated that, with sociodemographic and other variables accounted for, perceived shopping attractiveness and accessibility influenced the relationship between e-shopping and city centre shopping. The importance of those factors, however, differed among transport mode users. With regard to car users, the perceived car accessibility of the city centre influenced the extent to which they shop online, and the extent to which they purchase less in city centres owing to e-shopping. For other transportation mode users, perceived shopping attractiveness was more important. The higher the perceived shopping attractiveness the lower the probability that non-car users shopped online and substituted e-shopping for purchases at city centre shops. Summary

Most scholars assume that city centres are least at risk from e-shopping. Small rural towns, on the other hand, are perceived to be more at risk from e-shopping. Which shopping locations are most likely to benefit or suffer from e-shopping is a question that can hardly be answered by the few empirical studies conducted thus far. This is because most empirical studies that investigated the relation between e-shopping and in-store shopping within a shopping centre context only looked at one type of shopping centre. Furthermore, most empirical studies are descriptive. Thus, more empirical research is necessary. The Dutch retail context For a better interpretation of the results, some remarks need to be made on the Dutch retail context. Compared with other developed countries, such as France and the United States, the Netherlands has an `old-fashioned' retail structure, characterised by a large number of small-scale shops per capita concentrated in urban areas, and only few large-scale hypermarkets and shopping malls at the edges of major cities (Evers, 2002). As in the UK, fully enclosed shopping malls have emerged mainly in city centres, but to a lesser extent and at a smaller scale (Guy, 1994). The retail hierarchy in the Netherlands consists of eight types of shopping location (Locatus, 2003) (see also table 1): (1) City centre: the largest and most central shopping location in a city (a hundred stores or more). (2) Village centre: the largest and most central shopping location in a village (between five and a hundred stores). (3) City district centre: a shopping centre with fifty stores or more operating next to a large city centre. (4) Neighbourhood centre: a shopping centre with ten to fifty stores, or with five to ten stores and two or more supermarkets, operating next to a city or village centre. (5) Convenience centre: a shopping centre with five to ten stores and one or no supermarket operating next to a city or village centre. (6) Large-scale retail location: a shopping centre with five or more stores with a mean floor space of 500 m2 or more per shop; the retail categories `pets, flowers, and plants', `electrical appliances', `bicycles and car accessories', `do-it-yourself ', and `furniture and home furnishing' must make up at least 50% of the total floor space. (7) Special shopping centre: a shopping centre that does not belong to one of the other categories (eg factory outlet centres, shopping centres at airports). (8) Other retail location: a shopping location with fewer than five stores. It should also be remembered that the Netherlands is a small and highly urbanised country, where even in rural areas consumers have good shop accessibility. As a result,

The implications of e-shopping for in-store shopping

283

Table 1. General characteristics of shopping centres in the Netherlands (2006) (source: Locatus, 2006). Number of locations

Number of outlets

N

%

N

City centre Village centre City district centre Neighbourhood centre Convenience centre Large-scale retail location Special shopping centre Other retail location

138 785 90 601 400 129 14 na

6 36 4 28 19 6 1 na

37 539 24 906 8 860 12 234 3 046 2 409 287 27 571

32 21 8 10 3 2 0 24

6 758 524 4 362 739 1 271 033 2 093 370 511 822 3 192 472 59 214 9 671 737

24 16 5 7 2 11 0 35

Total

2 157

100

116 852

100

27 920 911

100

Retail location

%

Amount of floor space (in m 2 ) N

%

Notes. na: not available. Figures are based on retail categories only, and thus exclude service and handicraft categories.

the Netherlands differs from other West European countries and the USA in terms of the share of total distance that is covered by slow transportation modes. Of all shopping trips in the Netherlands in 2003, approximately 50% were made on foot or by bicycle or moped (AVV, 2004). Methodology Data collection

To investigate the implications of e-shopping for in-store shopping of various products at different shopping locations in the Netherlands, a large dataset of approximately 3000 e-shoppers was required. We used the online panel of Multiscope, which contains more than 100 000 Internet users: 30 484 users were randomly selected to participate in the research,(1) of whom 4327 filled in the selection questionnaire; 993 did not fit our criteria, because they were (1) not buying via the Internet (628 respondents, 63%), (2) not responsible for (nondaily) shopping (300 respondents, 30%), or (3) unable to indicate at which locations they shop (65 respondents, 7%); 334 respondents did not want to fill in the main questionnaire, leaving a net number of exactly 3000 respondents. The data were collected from 30 August until 19 September 2006. To investigate the impact of e-shopping on shopping locations, respondents could choose among the 2157 shopping centres that are present in the retail location database of Locatus. This dataset includes every shopping centre in the Netherlands. Several important characteristics of these localities can be obtained from this dataset, such as the mean x, y coordinate (2) of the shopping centre, the type of shopping centre, the total number of outlets, the total amount of floor space, the total number of retail categories (variety), and the total number of vacant shops. For each town or city we also added an `other shopping location', to represent retail locations that are not classified as shopping centres, such as solitary shops. In addition, for respondents at (1) Multiscope estimated that one third of their panel members should receive an invitation to obtain 3000 net responses. (2) In order to calculate the distance between a respondent's place of residence and the shopping centre, we needed a single x, y coordinate for the whole shopping centre. As such, we took the mean of all x, y coordinates of the shops in the shopping centre as a proxy for the x, y coordinate of the whole shopping centre.

284

J W J Weltevreden, T van Rietbergen

the border regions we added the option of shopping in Belgium or Germany. This brings the number of shopping locations which respondents could choose from to 4352 locations. One should take into account that the implications of e-shopping also largely vary among products (Weltevreden, 2007). This is especially important as retail categories are not equally distributed across shopping locations. For example, some shopping locations are more specialised in maintenance (grocery) shopping (eg convenience and neighbourhood centres), while others are more specialised in nonmaintenance, or recreational shopping (eg city centres). As such, we need to differentiate among shopping locations, according to the main type of shopping activity that is conducted at each location. In the online questionnaire respondents could select two shopping locations (3) for each of the following types of goods: daily goods (ie groceries and health and personal care items), nondaily goods (eg clothing, shoes, books, CDs, toys, telecom, tickets), and speciality goods (furniture, home furnishing, kitchens, cars, electrical appliances). This distinction closely resembles Copeland's (1923) goods classification of convenience goods, shopping goods, and speciality goods, which is also used in other empirical studies (eg Hassenpflug and Tegeder, 2004. Response analysis

To investigate whether our sample is representative of the Dutch e-shopper we compared it with a nationwide sample of (non)e-shoppers from Statistics Netherlands (table 2). We also compared our respondents with two nonresponse groups: the none-shoppers and the nonshoppers. Our sample closely resembles the e-shopper profile from Statistics Netherlands (table 2). On the urbanisation variable, however, our sample differs from the Statistics Netherlands sample (w 2 ˆ 61:24 with p < 0:05 and degrees of freedom 4). E-shoppers in the highly and moderately urbanised areas are overrepresented in our sample, while e-shoppers in the strongly and weakly urbanised areas are underrepresented. Nevertheless, for the multivariate analyses there are still sufficient numbers of cases in each category. Comparing our sample with the nonresponse samples shows that we have less females than the non-e-shopper sample. Nonshoppers, on the contrary, are more often males and people between 15 and 35 years of age, as compared with the response sample. This seems to reflect the traditional labour division in the household, as well as the fact that a lot of youngsters are not, or do not feel, responsible for shopping. Dependent variables

Several questions in the survey dealt with actual shopping behaviour, as well as the perceived impact of e-shopping on in-store shopping. We could only ask about respondents' perceptions, as the data are cross sectional. Asking for information about two shopping centres per purchase category could provide data about 6000 shopping centres per category (3000 respondents  2 shopping centres). The actual number of shopping centres was somewhat smaller, however, as some respondents only shopped at one shopping centre, or did not shop for daily items themselves. Moreover, some people did not shop at a shopping centre, but at retail locations which are not classified as shopping centres (eg solitary shops). We do not have information about the characteristics of these locations.

(3) Although

we are aware that most consumers shop at more than two shopping locations for particular types of goods, we only asked respondents to select two locations per product category because otherwise the online questionnaire would have become too lengthy.

The implications of e-shopping for in-store shopping

285

Table 2. Sample characteristics. Features

Response (%) e-shopper

Nonresponse (%)

Total population (2006) (%) a, b

nonnonshopper e-shopper

e-shopper

none-shopper

Gender Male Female

56.6 43.4

53.3 46.7

65.0 35.0

55.5 44.5

45.9 54.1

Age 15 to 35 years 35 to 55 years 55‡ years

43.2 45.2 11.6

39.3 38.4 22.3

78.4 12.8 8.8

43.2 45.5 11.3

30.6 40.0 29.4

Education Low Middle High

24.1 41.5 34.4

24.3 48.7 27.0

29.1 58.4 12.5

24.1 40.4 35.5

47.3 35.0 17.7

Urbanisation Highly urbanised Strongly urbanised Moderately urbanised Weakly urbanised Not urbanised (rural)

22.3 25.7 22.8 17.3 11.8

21.3 30.6 19.9 15.3 12.9

13.7 23.7 23.7 18.7 20.1

18.6 28.5 20.4 21.2 11.3

18.1 26.6 21.8 21.5 12.0

N

3000

628

300

5 848 000

4 759 000

a Source:

Statistics Netherlands (2006). b The original data have been transformed to reflect the total Dutch (non)e-shopper population.

We investigate the impact of e-shopping on two characteristics of respondents' shopping activity at shopping centres: the frequency of a shopping trip and the number of purchases made on a shopping trip. First, we asked respondents whether they visit a shopping centre more often, as often, or less often because of e-shopping [table 6(a)]. Second, we asked whether people make more, as many, or less purchases at a certain shopping location because of e-shopping [table 6(b)]. Table 3 shows the mean frequency and duration of shopping at various shopping centres, which is helpful for the interpretation of the dependent variables in the multivariate analyses (tables 7 ^ 10). These outcomes are largely consistent with the findings of Dutch studies that used other data collection methods, such as travel diaries (eg AVV, 2006) and time-use diaries (eg SCP, 2006). Independent variables

In the regression analyses we include sociodemographic, behavioural, e-shopping, household, and spatial variables. The sociodemographic and behavioural variables are: gender (0 ˆ female, 1 ˆ male), age in years (continuous), age squared (continuous, included to measure any nonlinear effects of increasing age), education (low, medium, high), and degree of enjoyment of shopping in the concerned shopping centre [not at all/hardly, to some extent, (very) much]. Since the frequency of trips to a shopping centre is influenced by the transport mode consumers use to get there (Monheim, 1998), we also included a transport mode variable per shopping centre (on foot, bicycle or moped, and car or public transport). For city centres we further distinguished between car and public transport, while for the other centres we used the combined variable.

Shopping centre

City centre Village centre City district centre Neighbourhood centre Convenience centre Large-scale retail location Special shopping centre Other shopping locations in the Netherlands Shopping locations in Belgium or Germany Mean value for all shopping centres N

Daily goods

Nondaily goods

286

Table 3. Average frequency of shopping, duration of shopping, and distance travelled, according to shopping centre and product type. Speciality goods

frequency duration (per month) (minutes)

distance (km)

frequency duration (per month) (minutes)

distance (km)

frequency (per month)

duration (minutes)

distance (km)

8.53 10.22 9.28 9.46 8.34 6.15

46 38 42 34 31 51

4.91 3.78 3.60 2.54 2.50 4.30

2.63 3.49 2.66 2.56 2.64 1.43

111 65 81 48 62 86

13.08 7.62 7.70 4.67 8.12 9.98

3.61 2.75 3.38 4.14 2.74 3.24

112 76 92 82 75 115

12.19 10.00 8.98 10.91 10.61 14.05

3.38 8.42

62 41

7.44 na

1.32 2.61

109 85

25.50 na

3.55 2.63

94 111

17.10 na

4.29

89

na

1.23

147

na

2.87

169

na

9.07

39

4.91

2.64

99

11.66

3.24

109

12.32

5247

4287

5726

5717

5193

5529

5521

4109

5256

J W J Weltevreden, T van Rietbergen

Notes. na: not available; figures are based on respondents that actually visited a certain shopping centre for purchasing daily, nondaily, and/or speciality items.

N:/psfiles/epb3602w/

The implications of e-shopping for in-store shopping

287

The e-shopping variables are: frequency of online searching (continuous logarithm), frequency of online buying (continuous logarithm), online buying experience in years (continuous). The household variables include: children (0 ˆ no children, 1 ˆ with children), household situation (0 ˆ single or other, 1 ˆ living together or married), and the total number of hours (paid) work in the household (continuous). We also developed five spatial variables. The first measures the size or attractiveness of a shopping centre. Principal component analysis was used to construct this variable. The size/attractiveness component contains the following four highly correlated variables: (1) the total number of outlets, (2) the total amount of floor space, (3) the total number of retail categories, and (4) the entropy of the retail category distribution which measures the skewness of the distribution. The second spatial variable measures the share of vacant shops in a shopping centre (continuous logarithm). The third variable is the urban density of the respondents' place of residence, which is defined as the number of street addresses in 2004 per four-digit zip code. The fourth variable is the number of shopping opportunities in the vicinity of the consumers' place of residence. Using Flowmap version 7.2 we developed shop accessibility measures by combining retail data at the zip code level (destination) with the respondent's zip code (origin) and a roadmap of the Netherlands (street-network-based travel distances). We used a regular proximity count, which calculates the total number of shops respondents can reach by car from their place of residence in a certain time distance, ranging from five to forty-five minutes. Since the Netherlands is a small and highly urbanised country, we did not include the total number of shops a respondent could reach by car in more than forty-five minutes. We use the logarithm of these variables, as the original measures are skewed. Empirical studies have shown that urban density and shop accessibility influence the probability to adopt e-shopping (Farag et al, 2006; 2007). Finally, the probability to substitute e-shopping for in-store shopping is likely to be influenced by the distance separation between the consumer's place of residence and the shopping centre (Borgers et al, 1991). Using Flowmap we developed a variable which measures the actual travel time from the respondent's place of residence to the shopping centre he or she visits, according to the mode of transportation. We use the logarithm of this variable, as the original measure is skewed. Results

Potential substitution of in-store shopping, according to retail category and the type of shopping centres

Before examining the potential substitution of e-shopping for in-store shopping at various locations, it is necessary to give an overview of the products most frequently purchased online (table 4). With this ranking in mind one can interpret better the potential substitution of in-store shopping as presented in figure 1. Secondhand items, books, CDs/DVDs/videos, travel, and outer clothing are the most popular products purchased online by Dutch e-shoppers. The popularity of secondhand items is especially remarkable. In 2004 only 6% of all online purchases in the Netherlands were secondhand goods (Weltevreden, 2007). Let us first explore to what extent shopping centres are hypothetically influenced by e-shopping, according to retail category. For the most popular online purchases we investigated the distribution of those retail categories across the various shopping centres, to get an idea which shopping locations are most likely to face the substitution of e-shopping. Retail outlets that hypothetically are most at risk from e-shopping (ie the five most popular online purchases in table 4) are largely concentrated in city centres (50%) and village centres (23%). High-risk retail categories are considerably less often

288

J W J Weltevreden, T van Rietbergen

Table 4. Products purchased online in the Netherlands, on the basis of the respondents' last three online purchases. Retail category

N

Second-hand products Books CDs, DVDs, and videos Travel Outer clothing Tickets Computers and accessories Underwear Photographic goods Software Telecom Toys Office supplies Brown goods (eg CD players, TVs) Health and personal care goods Erotica Sporting goods Shoes Financial products White goods (eg fridges, microwaves) Household goods Food and drinks Jewellery Furniture and home furnishing Art and antiquities Bedroom furniture and accessories Other products

1090 1002 907 852 796 460 448 240 231 215 192 138 135 133 115 113 103 95 90 80 75 65 60 53 40 40 443

Total online purchases Total respondents

8211 3000

% 13.3 12.2 11.0 10.4 9.7 5.6 5.5 2.9 2.8 2.6 2.3 1.7 1.6 1.6 1.4 1.4 1.3 1.2 1.1 1.0 0.9 0.8 0.7 0.6 0.5 0.5 5.4 100 100

located in city district centres (9%), neighbourhood centres (8%), convenience centres (1%), and large-scale retail locations (0%). However, the fact that e-commerce-sensitive retail categories are largely concentrated in city centres and village centres is not surprising given the fact that more than 50% of all retail outlets in the Netherlands are located in those centres (table 1). As such, one also needs to look at e-commerce-sensitive retail categories as a share of all retail categories within a certain shopping centre (table 5). As shown in this table, the share of e-commerce-sensitive retail categories is highest in city centres, city district centres, and village centres. After identifying the hypothetical risk of e-shopping for shopping centres, we move to the potential risk. For their last three online purchases respondents had to indicate how they would have bought the product if the Internet could not be used for the transaction (figure 1). Respondents were able to select the shopping centres they mainly patronise. A comparison of the outcomes shown in figure 1 with the results presented in the preceding tables shows that city centres face more substitution of in-store shopping than one would expect on the basis of the distribution of high-risk retail categories across all shopping locations in the Netherlands. Village centres, on the other hand, face considerably less substitution of in-store shopping with e-shopping. These outcomes, however, again largely vary among retail categories.

The implications of e-shopping for in-store shopping

289

Second-hand products Books CDs, DVDs, videos Travel Outer clothing Tickets Computers and accessories Underwear Photographic goods Software Telecom Toys Office supplies Brown goods Health and personal care goods Erotica Sporting goods Shoes Financial products White goods Household goods Food and drink Jewellery Furniture and home furnishing Art and antiquities Bedroom furniture and accessories Other products Total

0

10

20

30

40

50 60 Share (%)

70

80

90

100

Village centre Neighbourhood centre Large-scale retail location By mail/telephone Other

City centre City district centre Convenience centre Other retail locations Product would not have been bought then

Figure 1. Where would a product have been purchased if it was not possible to buy it online (N ˆ 8207 purchases)? Table 5. Cumulative share of e-commerce-sensitive retail categories within each shopping centre. All values are percentages. Online purchases a

City centre

Village centre

City Neigh± district bourhood centre centre

Conve- Large-scale nience retail centre location

Other shopping locations

Total

Top 5 Top 10 Top 15 Top 20 Top 25 All products

20 24 31 39 48 100

14 18 24 33 44 100

15 19 25 32 43 100

5 6 12 16 30 100

3 3 39 50 58 100

11 14 19 26 35 100

a Based

9 12 18 23 36 100

2 3 6 9 36 100

on table 4. The `top 5' are the five most popular products online, etc.

An important outcome is that, on average, 19% of the respondents would not have bought the product at all if the Internet could not be used for the transaction (figure 1), which indicates that e-shopping creates additional demand. For retailers this is good news, because it means that e-commerce is not a zero-sum game, though it has to be said that the additional demand is especially large for secondhand items, which in fact is consumer-to-consumer (c2c) e-commerce. The rise of online auctions

290

J W J Weltevreden, T van Rietbergen

and advertisement sites such as eBay and Marktplaats has indeed created a complete new market apart from traditional (e-)retailing. One should note, however, that the results in figure 1 only represent people's intentions. The reality can of course be different. Nevertheless, we believe that they provide valuable insights into the implications of e-shopping with regard to in-store shopping for different products at various shopping centres. Table 6. Impact of e-shopping on: (a) the frequency of visiting a shopping centre, according to the type of good; (b) the number of in-store purchases in a shopping centre. Shopping centre

Daily goods

Nondaily goods

as often/ less N more often often (%) (%)

as often/ less more often often (%) (%)

N

as often/ less N more often often (%) (%)

(a) City centre 93 Village centre 97 City district 96 centre Neighbourhood 97 centre Convenience 97 centre Large-scale retail 94 location Special shopping 100 centre Other shopping 95 location Shopping locations 88 in Belgium or Germany Total

96

(b) City centre 95 Village centre 97 City district 97 centre Neighbourhood 97 centre Convenience 99 centre Large-scale retail 97 location Special shopping 100 centre Other shopping 97 location Shopping locations 88 in Belgium or Germany Total

96

Speciality goods

7 3 4

1013 1017 552

80 88 82

20 12 18

3828 469 457

79 82 78

21 18 22

2280 304 258

3

1287

85

15

224

90

10

115

3

415

82

18

50

97

3

35

6

78

81

19

198

84

16

1169

0

8

88

12

25

82

18

11

5

870

76

24

436

79

21

1305

13

16

92

8

39

77

23

52

4

5256

81

19

5726

81

19

5529

5 3 3

1013 1017 552

75 83 79

25 17 21

3828 469 457

77 82 78

23 18 22

2280 304 258

3

1287

83

17

224

87

13

115

1

415

80

20

50

97

3

35

3

78

82

18

198

85

15

1169

0

8

96

4

25

73

27

11

3

870

76

24

436

78

22

1305

13

16

74

26

39

79

21

52

4

5256

77

23

5726

80

20

5529

The implications of e-shopping for in-store shopping

291

Actual substitution of in-store shopping, according to the type of shopping centre

In this section we present the estimation results for the probability of substituting e-shopping for in-store shopping, while accounting for sociodemographic and behavioural variables. We look at the substitution of shopping trips (tables 7 and 9), as well as the substitution of in-store purchases (tables 8 and 10). The reason for looking at both dependent variables is that the substitution of in-store purchases does not necessarily lead to fewer shopping trips, as consumers often buy more than one product during a shopping trip. Before turning to the multivariate analyses we first look at some descriptive results (table 6). Table 6 shows the impact that e-shopping has on the frequency of visits to the various shopping centres and the number of in-store purchases at these locations. Logically, the effect of e-shopping on daily goods is limited. The share of shopping centres for daily purchases that experience a decrease in the number of purchases at, and shopping trips to, these locations is approximately 4%. However, for nondaily and speciality goods, the adverse effects are already becoming significant, as about 20% of the shopping centres experience a decrease in the number of purchases and visits. These adverse effects are largest for city centres and city district centres. For neighbourhood, convenience, and village centres these impacts are smaller, but still significant. However, the results also show that neighbourhood and convenience centres are not important locations for purchasing nondaily and speciality goods. As such, e-shopping does not seem to have large detrimental consequences for these localities. Multivariate results

Using logistic regression we attempted to find the characteristics of people who, owing to e-shopping, visit shopping centres less, and/or buy in them less often, for the purchase of nondaily items and speciality goods. The dependent variable in these models is, thus, whether people substitute (1) or do not substitute (0). We did not estimate regression for shopping centres in which people buy their daily items, as these localities are hardly affected by e-shopping. The choice whether or not to conduct an analysis on a specific shopping centre is based on the popularity of a centre for the purchase of the three types of goods. The cut off is 225 respondents per shopping centre and per type of shopping activity (see table 6). The final models were constructed after log-likelihood tests were carried out to check whether the statistical significance of the model deteriorated when insignificant variables were left out. For this reason, some nonsignificant variables have been included in the final models. Before estimation we also checked the data for outliers and multicollinearity. It should also be noted that, as shown by the R 2, the models presented in tables 7 ^ 10 perform rather poorly. Consequently only part of the variation in substituting e-shopping for in-store shopping can be attributed to differences in the personal, household, residential environment, and behavioural variables included in our models. Other important explanatory variables, such as attitudes (eg Farag et al, 2007) and perceived shopping centre attractiveness (Weltevreden and Van Rietbergen, 2007), were not included in our dataset. With regard to nondaily shopping, results indicated that the more frequently people buy online, the higher the probability that they will substitute e-shopping for shopping trips and in-store purchases (see tables 7 and 8). Not surprisingly, this variable is the most important determinant in the models. Furthermore, people with a low accessibility to shops seem especially to shop less frequently for nondaily items at city centres, village centres, and city district centres. In half of the models, age (significant when age squared is deleted) has a negative coefficient, which implies that, for nondaily items, older people have a lower likelihood than do young people to substitute e-shopping for in-store shopping.

Variables

City centre

Village centre

City district centre

standard error

exp (B )

B

standard error

exp (B )

B

standard error

exp (B )

Constant

ÿ0.748

0.661

0.473

ÿ0.041

1.338

0.960

ÿ0.066

1.174

0.936

Sociodemographic and behavioural variables Age Education: high Enjoyment of in-store shopping: neutral Enjoyment of in-store shopping: high

ÿ0.025*** 0.186* ÿ0.339* ÿ0.895***

0.005 0.111 0.177 0.179

0.975 1.205 0.713 0.408

ÿ0.037**

0.016

0.964 ÿ0.609**

0.289

0.544

1.530*** 0.040*

0.136 0.020

4.616 1.040

0.898** 0.093

0.411 0.059

2.455 1.098

1.474*** 0.330

4.367

0.795* ÿ0.015**

0.463 0.007

2.214 0.985

ÿ0.695**

0.293

0.499

ÿ0.658*

0.388

0.518

ÿ0.861**

0.346

0.423

ÿ0.185 1.047*

0.176 0.555

0.831 2.850

Household variables Living together/married Number of hours (paid) work in household Spatial variables Log10 number of shops within 10 minutes by car from place of residence Log10 travel time in minutes to shopping centre Factor attractiveness/size shopping centre Log10 share of vacant outlets in the shopping centre w2 ÿ2log likelihood Nagelkerke R 2 N * p < 0:1; ** p < 0:05; *** p < 0:01.

ÿ0.323**

0.149

0.724

0.154

0.228

1.166

0.109*

0.064

1.115

230.827*** 2 290.203 0.138 2 556

19.837*** 287.540 0.089 423

36.831*** 347.256 0.140 418

J W J Weltevreden, T van Rietbergen

B

E-shopping variables Log10 frequency of online buying Log10 number of years buying online

292

Table 7. Binary logistic regression models concerning the substitution of e-shopping for trips to a shopping centre, for the purchase of nondaily items.

Variables

City centre

Village centre

City district centre

B

standard error

exp (B )

B

standard error

exp (B )

B

Constant

ÿ0.778

0.615

0.459

ÿ3.892

0.591

0.020

ÿ2.421*** 0.336

Sociodemographic and behavioural variables Gender: male Age Education: medium Education: high Transport mode: public transport Enjoyment of in-store shopping: high

ÿ0.017*** 0.259** 0.401*** ÿ0.275* ÿ0.365***

0.005 0.132 0.137 0.150 0.099

0.983 1.296 1.493 0.760 0.694

1.785***

0.126

5.962

ÿ0.284***

0.108

0.753

ÿ0.276**

0.141

0.759

ÿ0.186

0.212

0.830

0.062

1.300

E-shopping variables Log10 frequency of online buying Number of years buying online Household variables Living together/married Spatial variables Log10 number of shops within 10 minutes by car from place of residence Log10 travel time in minutes to shopping centre Factor attractiveness/size shopping centre w2 ÿ2log likelihood Nagelkerke R 2 N

0.262*** 294.496*** 2 588.917 0.161 2556

1.318*** 0.363 0.112** 0.053

3.735 1.119

0.551

1.736

26.804*** 348.347 0.105 411

0.379

standard error

exp (B )

0.089

0.431*

0.255

1.539

ÿ0.681**

0.269

0.506

1.380*** 0.300

3.976

The implications of e-shopping for in-store shopping

Table 8. Binary logistic regression models concerning the number of purchases of nondaily items for which e-shopping is substituted.

33.374*** 404.678 0.118 418

* p < 0:1; ** p < 0:05; *** p < 0:01. 293

Table 9. Binary logistic regression models concerning the substitution of e-shopping for trips to a shopping centre, for the purchase of speciality goods. Variables

Constant Sociodemographic and behavioural variables Gender: male Age Age squared Education: high Transport mode: public transport Transport mode: car/public transport Enjoyment of in-store shopping: neutral Enjoyment of in-store shopping: high E-shopping variables Log10 frequency of online searching Log10 frequency of online buying Number of years buying online Household variables Household with children Number of hours (paid) work in household Spatial variables Log10 number of shops within 10 minutes by car from place of residence Log10 travel time in minutes to shopping centre 2

w ÿ2log likelihood Nagelkerke R 2 N * p < 0:1; ** p < 0:05; *** p < 0:01.

City centre

Village centre

City district centre

Large-scale retail location

B

standard exp (B ) error

B

standard exp (B ) error

B

standard exp (B ) error

B

standard exp (B ) error

ÿ1.838***

0.672

0.159

ÿ1.318**

0.193

ÿ3.554***

0.752

ÿ4.714***

1.160

0.009

0.264** ÿ0.024***

0.127 0.006

1.302 0.977

0.124* ÿ0.002**

0.064 0.001

1.132 0.998

0.325*

0.195

1.384

ÿ0.381** ÿ0.371*

0.193 0.193

0.683 0.690

ÿ0.490***

0.186

0.613

0.353** 0.773*** 0.073***

0.147 0.187 0.024

1.423 2.165 1.076

0.497** 0.628**

0.210 0.250

1.644 1.874

0.255**

0.127

1.290

0.191 0.004

1.444 0.994

0.155

0.768

0.230

1.661

ÿ0.264* 0.507** 110.384*** 1663.420 0.096 1763

ÿ0.619*

0.330

0.268

0.029

0.654*

0.334

1.923

0.972**

0.475

2.643

0.538

0.367* ÿ0.006

1.165** 3.681* 262.181 0.021 288

13.107*** 233.244 0.083 236

0.520

3.205 51.090*** 841.500 0.084 1019

Table 10. Binary logistic regression models concerning the number of purchases of speciality goods for which e-shopping is substituted. Variables

Constant Sociodemographic and behavioural variables Gender: male Age Age squared Education: high Enjoyment of in-store shopping: neutral Enjoyment of in-store shopping: high E-shopping variables Log10 frequency of online searching Log10 frequency of online buying

City centre

Village centre

City district centre

Large-scale retail location

B

standard exp (B ) error

B

standard exp (B ) error

B

standard exp (B ) error

B

standard exp (B ) error

ÿ4.057***

0.722

0.017

ÿ7.189***

2.120

0.001

ÿ3.188***

0.678

0.041

ÿ2.662***

0.345

0.070

0.541*** 0.083** ÿ0.001***

0.123 0.033 0.000

1.718 1.087 0.999

0.839** 0.256** ÿ0.003**

0.350 0.106 0.001

2.315 1.292 0.997 0.659*

0.338

1.933

ÿ0.514*** ÿ0.474***

0.183 0.184

0.598 0.623

ÿ1.262*** ÿ0.968**

0.422 0.428

0.283 0.380

ÿ0.492** ÿ0.803***

0.247 0.258

0.612 0.448

1.026***

0.146

2.790

1.354***

0.499

3.874

0.565*** 0.815***

0.213 0.251

1.760 2.260

Household variables Household with children Spatial variables Log10 travel time in minutes to shopping centre Factor attractiveness/size shopping centre w2 ÿ2log likelihood Nagelkerke R 2 N * p < 0:1; ** p < 0:05; *** p < 0:01.

0.639***

0.212

1.895 0.377*

101.471*** 1764.322 0.086 1763

33.516*** 229.201 0.184 288

0.210

1.458

0.795**

0.361

2.214

1.047** ÿ0.434**

0.530 0.189

2.848 0.648

16.629*** 229.722 0.105 236

53.198*** 812.262 0.089 1019

296

J W J Weltevreden, T van Rietbergen

The same holds for enjoyment of in-store shopping: the more people enjoy shopping at city centres and city district centres, the less likely they will substitute e-shopping for in-store shopping. Furthermore, gender and education do not seem to be important predictors. A surprising outcome of tables 7 and 8 is that larger/more attractive city centres have a higher probability of facing substitution than smaller/less attractive city centres. Note that this variable reflects the characteristics of a shopping centre, rather than the opinion of the visitors. Consumer perception of the attractiveness does not necessarily correspond with these characteristics and is most important for explaining (changes in) shopping behaviour (Weltevreden and Van Rietbergen, 2007). Unfortunately, we do not have information in our dataset on the consumers' perception of the attractiveness of shopping centres. The results for substitution of in-store shopping regarding speciality goods are quite similar to the outcomes for nondaily items. Again, in most models the frequency of online buying and enjoyment of in-store shopping are important predictors. However, contrary to the models for nondaily items, the impact of age in the speciality goods models is often nonlinear. The probability of making fewer trips to large-scale retail locations increases up to the age of 37 years; after that age it decreases (table 9). In addition, the likelihood of buying less speciality goods at city and village centres, owing to e-shopping, increases up to the age of 36 years and 41 years, respectively, and then decreases (table 10). Furthermore, males are more likely to substitute e-shopping for in-store shopping, for speciality goods, at city and village centres than are females. The same holds for people that live relatively far from their city centres and city district centres. A final interesting outcome is that, especially at large-scale retail locations, online searching positively influences the substitution of in-store shopping regarding speciality goods (tables 9 and 10). This means that online searching is not necessarily complementary to in-store shopping, as claimed by some scholars (eg Farag et al, 2007). Conclusions and directions for further research Despite the frequently made assumption in the literature that (large) city centres are most likely to benefit from e-shopping, our results seem to indicate otherwise. City centres are most likely to face substitution of e-shopping for in-store shopping, followed by city district centres. For the latter, projections of the impact of e-shopping were not very optimistic and our results indeed substantiate this. Surprisingly, village centres are less affected by e-shopping than are city centres. Moreover, for neighbourhood and convenience centres the adverse effect of e-shopping are small, as these localities are mainly used for daily shopping. This also has much to do with the fact that online grocery shopping is not (yet) very popular in the Netherlands. However, the extent to which shopping centres are influenced by e-shopping not only depends on the behaviour of consumers, as presented in this paper, but also on how retailers respond to changes in consumers' shopping patterns as a result of e-shopping (Burt and Sparks, 2003; Wrigley et al, 2002). Weltevreden et al (2008) recently showed that, while accounting for retail category and firm size, Dutch retailers at city and village centres are more likely to have a website and to sell online than are their counterparts at city district, neighbourhood, and convenience centres. The (relatively) proactive reaction by the former may (partly) offset the adverse effects of e-shopping that mainly take place at city centres. Retailers at city district centres should take a more proactive stance towards e-commerce, as these shopping centres are also relatively vulnerable to e-shopping öthe more so, since city district centres are less suited for recreational shopping than are city centres.

The implications of e-shopping for in-store shopping

297

Besides showing evidence that the impact of e-shopping largely varies among shopping centres, the results also show that shop accessibility and travel time to shopping centres influence the decision whether or not to substitute e-shopping for in-store shopping. Consumers with few shopping opportunities in their surroundings and consumers that have to travel a large distance to a specific shopping centre are more likely to substitute e-shopping for in-store shopping. These results further support the importance of geography for explaining the impacts of e-shopping on society, which has already been stressed elsewhere (eg Farag et al, 2006; 2007; Weltevreden and Van Rietbergen, 2007). For retailers another important outcome is that more than 19% of all online purchases would not have been made if the Internet was not used for the transaction. Thus, e-shopping is not always substituted for in-store purchases, but also creates additional demand, though it has to be said that the additional demand is especially large for secondhand shopping, which in fact is c2c e-commerce. Moreover, people that highly enjoy in-store shopping are less inclined to substitute e-shopping for in-store shopping at a particular shopping centre. On the other hand, the frequency of online buying significantly influences the decision whether or not to substitute e-shopping for in-store shopping. As such, people that are buying a lot online are less likely to go to shopping centres. Our results also show that online searching is not always complementary to in-store shopping, particularly in the case of purchasing speciality goods. As growing numbers of people become e-shoppers, and current e-shoppers are expected to buy online more frequently in the future (Thuiswinkel.org, 2007; Verdict, 2005), the adverse effects will only grow bigger. However, the extent to which e-shopping replaces in-store shopping also largely depends on the type of product. Secondhand goods stores, telecom shops, electrical appliances stores, lingerie stores, bookstores, and CD shops currently seem to be the principal victims of e-shopping. Regarding the implications of e-shopping for shopping centres, progress in future research lies in two areas. First, future studies should include a more complete set of explanatory variables, such as attitudes, lifestyle variables, and perceived shopping centre attractiveness. Our models performed relatively poorly, which indicates that other explanatory variables should be considered in future research. Future studies should also investigate the extent to which the frequency of visiting shopping centres or purchasing goods decreases because of e-shopping. Second, like many other e-shopping studies we used cross-sectional data. As such, we could only ask about peoples' perceptions with regard to substitution of e-shopping for in-store shopping, which need not necessarily reflect the real impact of e-shopping on in-store shopping at various shopping centres. Therefore, future research should try to collect longitudinal data that make it possible to compare people's online and in-store shopping behaviour over a certain period of time. These kinds of data could be collected by using software that registers people's online (purchase) behaviour in combination with travel diaries and/or GPS devices that register people's offline behaviour. This type of research may further enhance our understanding of the implications of e-shopping for retailing and shopping centres. Acknowledgements. We would like to thank Hans van Amsterdam and Stephaan Declerck for the preparation of the retail data, and Tom de Jong for his help on constructing the accessibility measures and the calculation of the travel time distances. Finally, we wish to thank the two anonymous reviewers for their constructive comments on earlier versions of this paper.

298

J W J Weltevreden, T van Rietbergen

References AoyamaY, Schwarz G, 2004,``From mail order to e-commerce: competition, regulation, and politics of non-store retailing in Germany'' Urban Geography 25 503 ^ 527 AVV, 2004 Kernicijfers Personenvervoer [Personal transport key figures], Ministerie van Verkeer en Waterstraat, The Hague, http://www.platformschonevoertuigen.nl/ default.aspx?iti=117&app=news AVV, 2006 Mobiliteitsonderzoek Nederland [Mobility research Netherlands], Ministerie van Verkeer en Waterstraat, The Hague, http://www.rws.nl/dvs/themas/mobiliteit/personenvervoer/mon/ index.jsp Bean J S, 2000,``The effect of technology on retail sales, commercial property values and percentage rates'' Journal of Real Estate Portfolio Management 6 185 ^ 201 Borgers A W J, Gunsing M, Timmermans H J P, 1991, ``Teleshopping and the dynamics of urban retail systems: some numerical simulations'', in Spatial Analysis and Population Dynamics Ed. D Pumain (Congress and Colloquia, Chantilly, France) pp 229 ^ 242 Burt S, Sparks W, 2003, ``E-commerce and the retail process: a review'' Journal of Retailing and Consumer Services 10 275 ^ 286 Copeland M T, 1923, ``Relation of consumers' buying habits marketing methods'' Harvard Business Review 1 282 ^ 289 Dixon T, Marston A D, 2002, ``UK retail real estate and the effects of online shopping'' Journal of Urban Technology 9 19 ^ 47 Dixon T, Thompson B, McAllister P, Marston A D, Snow J, 2005 Real Estate and the New Economy: The Impact of Information and Communications Technology (Blackwell, Oxford) Evers D, 2002, ``The rise (and fall?) of national retail planning'' Tijdschrift voor Economische en Sociale Geografie 93 107 ^ 113 Farag S, Weltevreden J W J, Van Rietbergen T, Dijst M, Van Oort F G, 2006, ``E-shopping in the Netherlands: does geography matter? '' Environment and Planning B: Planning and Design 35 59 ^ 74 Farag S, Schwanen T, Dijst M, Faber J, 2007, ``Shopping online and/or in-store? A structural equation model of the relationships between e-shopping and in-store shopping'' Transportation Research A 41 125 ^ 141 Foresight, 2000 Clicks and Mortar: The New Store Fronts (Department of Trade and Industry, London) Gillespie A, Marvin S, Green N, 2001, ``Bricks versus clicks: planning for the digital economy'', in Digital Futures: Living in a Dot-Com World Ed. J Wilsdon (Earthscan, London) pp 200 ^ 218 Guy C M, 1994 The Retail Development Process: Location, Property, and Planning (Routledge, London) Hassenpflug D, Tegeder G, 2004 Sta«dtischer Einzelhandel in vernetzten Zeiten [Urban retailing in the Internet age] (Bauhaus University, Weimar) Hendershott P, Hendershott R, Hendershott T, 2000, ``Will the Internet reduce the demand for mall space? '' Real Estate Finance 2000 41 ^ 46 Hernandez T, Gomez-Insausti R, Biasiotto M, 2001, ``Non-store retailing and shopping centre vitality'' Journal of Shopping Centre Research 8 58 ^ 81 Kolpron Consultants, 2001, ``De binnenstad: instrumenten voor offensief beleid'' [The city centre: instruments for proactive policy] (HBD/Platform-Detailhandel.nl, The Hague), http://www.platformdetailhandel.nl/index.cfm/28,551,c,html/DeBinnenstad.pdf Locatus, 2003 Locatus Database Beschrijving [Locatus' database description] (Locatus, Woerden) Locatus, 2006 Retail Locatie Database [Retail location database] (Locatus, Woerden) Miller N G, 2000, ``Retailing leasing in a web enabled world'' Journal of Real Estate Portfolio Management 6 167 ^ 184 Mokhtarian P L,1990,``A typology of relationships between telecommunicatons and transportation'' Transportation Research A 24 231 ^ 242 Mokhtarian P L, 2002, ``Telecommunications and travel: the case for complementarity'' Journal of Industrial Ecology 6 43 ^ 57 Mokhtarian P L, 2004, ``A conceptual analysis of the transportation impacts of b2c e-commerce'' Transportation 31 257 ^ 284 Monheim R, 1998, ``Methodological aspects of surveying the volume, structure, activities and perceptions of city centre visitors'' Applied Urban Geography 45 273 ^ 287 Salomon I, 1985, ``Telecommunications and travel: substitution or modified mobility? '' Journal of Transport Economics and Policy 19 219 ^ 235

The implications of e-shopping for in-store shopping

299

Salomon I, 1986,``Telecommuniations and travel relationships: a review'' Transportation Research A 20 223 ^ 238 Schellenberg J, 2005 Endverbraucherbezogener E-Commerce. Auswirkungen auf die Angebotsund Standortstrukture im Handel und Dienstleistungssektor [B2C e-commerce. Impacts on retail structure] (LIS, Passau) SCP, 2006, ``Tijdsbestedingsonderzoek'' [Time spending research], Social en Cultureel Planbureau, The Hague, http://www.tijdsbesteding.nl Statistics Netherlands, 2006, ``Resultaten POLS onderzoek 2005'' [Results of POLS questionnaire 2005] (Statistics Netherlands, Voorburg/Heerlen). Stec Groep, 2000 De Invloed van E-business op de Nederlandse Vastgoedmarkt [The impact of e-business on the Dutch real estate market] (IVBN, Voorburg) Thuiswinkel.org, 2007, ``Thuiswinkelen: onderzoek en cijfers'' [Homeshopping: research and figures], http://www.thuiswinkel.org Timmermans H J P, Borgers A W J, Gunsing M, 1991, ``The potential adoption of teleshopping technologies in a spatial context: a decompositional choice experiment'' The International Review of Retail, Distribution and Consumer Research 1 549 ^ 567 TNO Inro, 2002 Op Weg naar een Omgevingsverkenner ICT, Ruimte en Mobiliteit [Towards an ICT, space and mobility monitor] (Ministry of Transport, Public Works, and Water Management/Ministry of Economic Affairs/Ministry of Housing, Spatial planning, and Environment, The Hague) Van Oort F G, Raspe O, Snellen D, 2003 De Ruimtelijke Effecten van ICT [The spatial implications of ICT] (Nai Uitgevers/Ruimtelijke Planbureau, Rotterdam) Verdict, 2005 Verdict on e-retail 2005 (Verdict Research Limited, London) Weltevreden J W J, 2007, ``Substitution or complementarity? How the Internet changes city centre shopping'' Journal of Retailing and Consumer Services 14 192 ^ 207 Weltevreden J W J, Van Rietbergen T, 2007, ``E-shopping versus city centre shopping: the role of perceived city centre attractiveness'' Tijdschrift voor Economische en Sociale Geografie 98 68 ^ 85 Weltevreden J W J, Atzema O A L C, Frenken K, De Kruif K,Van Oort F G, 2008, ``The geography of Internet adoption by independent retailers in the Netherlands'' Environment and Planning B: Planning and Design 35 443 ^ 460 Wengler T, 2005 Auswirkungen des Internet-Handels auf Shopping-Center [The impacts of B2C e-commerce on shopping centres] (Leipzig University, Leipzig) Wrigley N, Lowe M, Currah A, 2002, ``Retailing and e-tailing'' Urban Geography 23 180 ^ 197

ß 2009 Pion Ltd and its Licensors

Conditions of use. This article may be downloaded from the E&P website for personal research by members of subscribing organisations. This PDF may not be placed on any website (or other online distribution system) without permission of the publisher.