t1r-!/) z9(2 :hvn - keio universitytotal convenience store customers in 2013 (kato, 2016). in recent...

Contents lists available at ScienceDirect

Journal of Urban Managementjournal homepage: www.elsevier.com/locate/jum

The package redelivery problem, convenience store solution, andthe delivery desert: Case study in Aoba Ward, YokohamaShun Nakayamaa,∗, Wanglin Yanba Keio University, Graduate School of Media and Governance, 252-0882, E509 5322 Endo, Fujisawa City, Kanagawa, Japanb Keio University, Faculty of Environmental and Information Studies, 252-0882, E502 5322 Endo, Fujisawa City, Kanagawa, Japan

A R T I C L E I N F O

Keywords:Delivery desertConvenience storeGISAccessibility

A B S T R A C T

The redelivery problem occurs when a delivery service cannot deliver an item to the recipient onthe first attempt, requiring one or more additional attempts. There are impacts on profit, effi-ciency, convenience, traffic, and the environment. In recent years, the redelivery problem inJapan has been aggravated by the growth of e-commerce, which increases delivery volumes. As asolution, many convenience stores offer courier package pickup services, but the actual netbenefits are uncertain. In this industry, store locations are chosen based on profitability andsubject to laws and regulations. This study developed a model to assess the accessibility ofconvenience stores and their possible contribution to solve the redelivery problem. We defined a“delivery desert” as area where a resident cannot access the nearest convenience store packagepickup service within walking distance, and developed a model that considers physical loadrelative to customer age and topographic slope. We then conducted a case study in a suburbanneighborhood in the Tokyo Metropolitan Area and showed that (1) about 65% of residents in thestudy area live in a delivery desert, (2) regulations that restrict the location of convenience storeshave a very small impact on our results, and (3) the percentage of people living in a deliverydesert is low for the age groups targeted by convenience stores. These findings could serve as areference in policy discussions for solving the redelivery problem.

1. Introduction

In recent years, smartphones, the Internet and social media have helped e-commerce grow rapidly (Lin, 2018). Japan is noexception, and it has been reported that e-commerce is spreading nationally (METI, 2018).

Along with the increase of e-commerce is growth in the delivery of goods (METI & MLIT, 2018). This has been viewed as aproblem worldwide, in particular for the “last mile” of delivery (Morganti, Seidel, Blanquart, Dablanc, & Lenz, 2014). In Japan, it isreported that about 20% of courier items have to be redelivered (Cabinet Office Public Relations Office, 2017). Originally, e-com-merce was expected to have a positive effect on greenhouse gas emission reductions (Siikavirta, Punakivi, Kärkkäinen, & Linnanen,2008). However, later studies revealed that redelivery causes annual social costs in Japan of about 420,000 tons of CO2 emissions andconsumes about 180 million hours of driver's time (MLIT, 2015a).

In response, the government has promoted three policies: (1) strengthening communication between consumers and deliverycompanies, (2) improving the environment for active consumer participation in deliveries, and (3) improving delivery receivingmethods to achieve greater convenience (MLIT, 2015b). Accordingly, the industry is working to enable consumers to designate the

https://doi.org/10.1016/j.jum.2019.08.001Received 30 May 2019; Received in revised form 3 August 2019; Accepted 8 August 2019

∗ Corresponding author.E-mail addresses: [email protected] (S. Nakayama), [email protected] (W. Yan).

http://www.sciencedirect.com/science/journal/22265856https://www.elsevier.com/locate/jumhttps://doi.org/10.1016/j.jum.2019.08.001https://doi.org/10.1016/j.jum.2019.08.001mailto:[email protected]:[email protected]://doi.org/10.1016/j.jum.2019.08.001http://crossmark.crossref.org/dialog/?doi=10.1016/j.jum.2019.08.001&domain=pdf

delivery date and time using websites and apps provided by courier and mail-order companies. Active participation encouragesconsumer cooperation through credits for feedback. The government is also encouraging condominiums and convenience stores toinstall delivery boxes to receive packages from couriers. One survey found that among all these efforts, enabling convenience stores totake deliveries was the most popular option for residents (MLIT, 2015b). However, courier services offered at convenience stores maynot provide all the anticipated benefits.

The scale of the redelivery problem is evident from the results of government sample surveys of the courier redelivery rate, asshown in Table 1. The redelivery rate generally falls between 10% and 20%; hence on average, one of five packages must beredelivered because the recipient is not home the first time.

Due to location-related restrictions and biases, the ratio of the population that can go on foot to a convenience store may belimited. In this study, we define a “delivery desert” as an area where a resident cannot access the nearest convenience store packagepickup service within walking distance. This is an analog to the “food desert,” a term that describes areas where people have difficultyaccessing grocery shops (Suzuki, Kimura, Hino, & Kaneko, 2014; Walker, Keane, & Burke, 2010). To date no published study hasgeographically visualized the areas that have difficulty receiving shipments in the context of e-commerce. In this study, we evaluatethe population in the delivery desert in our study area and analyze the potential contribution convenience stores can make to solvethe redelivery problem.

2. Features and roles of convenience stores in Japan

The retail market share of convenience stores is rising every year around the world (Gahinet & Cliquet, 2018). In Japan, con-venience stores are known as essential facilities to support daily life. Since the first convenience store, 7-Eleven, landed on theJapanese islands in 1975 the brand's growth has been conspicuous. Every store was carefully sited and services were strategicallydesigned to reduce costs (Yuasa & Ikegame, 2001). Doing business 24 h a day, 7 days a week, increased business opportunities, andthe franchise business model allowed the stores to spread rapidly across the country. Despite being designed to have limited storespace, the functions of convenience stores in Japan are extremely diverse, including copying and printing, banking, ticketing, postingmail, and delivering official documents, in addition to selling the regular commodities of food, beverages, and stationery, etc. Manyof the services are an extension of the Internet. Consequently, convenience stores have become an integral part of daily life, with theadvantage of providing good accessibility to products and services (Gahinet & Cliquet, 2018; Nogimura, 2015). Some of the servicesnow being provided were a response to social need, while some arose because of fierce competition in the industry, though they mayhave increased running costs and staff workloads. Beyond the needs of consumers for convenience stores, the development of greaterassortment and functions convenient for daily life brought new opportunities and advantages to convenience stores, as shown inTable 2 (Sudo & Masuda, 2014). When it comes to introducing new concepts, 7-Eleven has typically been the market leader.

One of the functions added in this process was the service of receiving courier packages for customers to pick up in person, basedon agreements between delivery companies and convenience store chains. To some extent, the convenience store chains try to

Table 1Sample surveys of courier redelivery rates (MLIT).

October 2017 April 2018 October 2018

A B C A B C A B C

Urban 844,935 139,486 16.50% 812,984 132,979 16.40% 883,584 151,386 17.10%Suburban 1,436,175 209,040 14.60% 1,346,059 192,796 14.30% 1,354,016 198,572 14.70%Rural 126,629 16,372 12.90% 116,576 14,721 12.60% 118,947 16,009 13.50%All 2,407,739 364,898 15.20% 2,275,619 340,496 15.00% 2,356,547 365,967 15.50%

A=Courier deliveries B= Redeliveries C= Redelivery rate.

Table 2Year of introduction of major convenience store functions.Function Start Year

7-Eleven Lawson FamilyMart

Open 24 h 1975 1977 1978Copiers 1996 1983 1998Ability to pay utility bills 1987 1989 1990Receive courier deliveries Yamato 1989 1982(∼2004) 1985

Sagawa – 2015 –Japan Post – 2004 2002

Ticketing services 2000 1981 1988Food delivery 2000 2014ATM (banking services) 2000 2001 1999Issuance of official resident's card 2010 2016 2016

S. Nakayama and W. Yan

differentiate themselves by signing contracts with different delivery companies (MLIT, 2015b; Yamato Transport Co., 2004).The main target market of convenience stores was originally the younger generation. However, as store functions have diversified,

the customer base has been changing. In the 2000s, the proportion of customers aged over 50 has rapidly increased, reaching 30% oftotal convenience store customers in 2013 (Kato, 2016). In recent years, convenience stores have also enhanced their support systemsto provide services for seniors. This is another factor driving the increase in senior customers (Nakamura, Matsumoto, Yamamoto-Mitani, Suzuki, & Igarashi, 2018).

With the expansion of the customer base in addition to their inherent advantages of time, distance and opportunity, conveniencestores are also beginning to serve as a kind of social infrastructure provided by the private sector. During the Tohoku earthquake andtsunami disaster that struck on March 11, 2011, convenience stores responded in a timely way and demonstrated resilience byproviding convenient access to food, water, and many emergency needs with their own distributed logistics networks. In fact, in2009, the Japan Franchise Association had already issued the “Convenience Store Declaration as Social Infrastructure,” supported bymajor convenience store chains (Japan Franchise Association, 2009).

This was also a reason for the government to look to convenience stores as a solution to the redelivery problem. Generallyspeaking, the delivery system in Japan is already very well done. For example, a customer can go online to specify the delivery timeand location, and can also track packages being delivered in real time. These options provide convenience and increase the likelihoodthat a customer can receive the package at home. Meanwhile, one government survey reported that 90% of respondents would preferto have courier deliveries go straight to a convenience store on the first attempt, in order to eliminate the need for redeliveries (MLIT,2015b).

On the other hand, we must not forget that convenience store chains target specific customer segments to maximize profits, andoperate under laws and regulations that constrain their location choices. The real potential for convenience stores to eliminate theredelivery problem is still an open question in Japan.

Therefore, in this study, we use geographical information systems (GIS) to estimate the number people living in the deliverydesert in the study area, and define and measure the area where courier services cannot be practically provided via conveniencestores. Our study recognizes the fact that picking up a package at a convenience store requires physical mobility within the deliverydesert, and this could be a source of resistance for vulnerable persons.

3. Delivery desert: definition and quantitative evaluation

3.1. Definition

Before considering the delivery desert, let's look back on the definition of “food desert.” There is no clear definition of this termdespite many studies on the topic. Food desert originally meant an area where people cannot access a grocery store near where theylive. However, studies of food deserts more commonly assess differential access to healthy and affordable food between socio-economically advantaged and disadvantaged areas (Beaulac, Kristjansson, & Cummins, 2009). (Walker et al., 2010) proposed fiveindicators to identify a food desert: (1) distance to the nearest grocery store, (2) number of supermarkets in the area, (3) economiccosts, (4) types of food provided, and (5) nutritional value of the food provided.

In contrast to the concept of food desert, when using e-commerce, the delivery company typically delivers an order to the frontdoor of the consumer, meaning the consumer hardly moves physically. However, in the context of convenience stores being con-sidered as a solution for the redelivery problem, the “last mile” requires the physical movement of the consumer. Thus, just as withthe food desert, in our examination of using convenience stores for package pickup, it is possible for areas to exist where it is difficultto gain access to shipments. In the simplest terms, a delivery desert could be assessed simply by using a distance indicator, i.e., thedistance from home to the nearest convenience store that provides a courier package pickup service.

Previous research targeting Tokyo shows that about 65% of the population has access to a convenience store by walking, and lessthan 80% by walking or cycling (Hwang & Takada, 1995). The use of bicycles decreases as access by walking increases in areas withslopes though the coverage is assumed to become smaller. For this reason, we adopted walkable distance as an indicator for access inthis study. Some studies define a convenience store's walkable radius as 400m or 500m (Takemoto, 2015), equivalent to a 5-minwalk. Thus, the accessible zone is set as a 5-min walking distance for this study. To summarize the above points, the definition ofdelivery desert in this study is “an area where a resident cannot access the nearest convenience store package pickup service withinwalking distance” (a 5-min walk, adjusted for metabolic equivalent).

3.2. Quantitative evaluation

As a method of evaluating walking access while also considering physical load, many studies have focused on energy expenditure(Hall, Figueroa, Fernhall, Kanaley, & Kanaley, 2004; van der Walt & Wyndham, 1973). Thus, in order to consider physical load, weused the metabolic conversion distance instead of actual distance to calculate the area of a delivery desert, with reference to previousstudies (Hara, Ishizaka, & Ohashi, 2009; Satoh, Yoshikawa, & Yamada, 2006). The metabolic conversion distance is the actualdistance calibrated with consideration of the physical load when walking. Parameters to calculate physical load include, walkingspeed, body weight, and basal metabolic rate by age as the attributes of pedestrians, and road slope as an attribute of roads. This isbased on the assumption that the walking speed differs based on the pedestrian's age, basal metabolic rate, and the amount of energyrequired per day per unit of body weight. The population accessible to convenience stores was obtained by conducting a service areaanalysis using metabolic conversion distance and estimating the population size living in the reachable area (a 5-min walkable area,


as mentioned above). The calculation of the metabolic conversion distance is shown below. The variables used in the followingformulas are summarized in Table 3.

The metabolic conversion distance (L) is obtained by multiplying the actual distance (l) by the walking load (R) which takes intoaccount elements such as the slope, then divided by walking load when the road is horizontal (R0) (Satoh et al., 2006).

L= l× R / R0Next, walking load (R) is “the energy consumption when walking from a certain point to a destination point divided by the energy

necessary for one day” (Hara et al., 2009). Energy consumption by walking (E) is expressed by the following equation:E = (RMR + 1.2) × BMR × W × T

Relative metabolic rate (RMR) represents the influence of topography such as road slope. Relative loads are indicated by the valueof (RMR + 1.2), where the constant 1.2 is the resting metabolic rate during daytime life. Basal metabolic rate (BMR) x body weight(W) is basal metabolism and represents the minimum amount of energy required to maintain human life for one day. Because thesediffer depending on age and body weight, the equation is as shown above (Kanamori & Yan, 2018). In addition, the influence of age isreflected in the movement time (T) as a decrease in walking speed (v).= ◊= + ◊ ◊ ◊ ◊ ◊= + ◊ ◊ ◊

R E / (a W)(RMR 1.2) BMR W T {1 / (a W)}(RMR 1.2) BMR (l/v) (1/a)

The rate of relative metabolic rate (RMR) varies with the slope (s) and the walking speed (v). Research in the past used thewalking speed for each age group to substitute for the above moving speed (T), while in the calculation of the energy metabolic rate,many studies substituted 80m/min, indicating that the accuracy was not necessarily high. Therefore, based on previous studies thatfully consider the continuity of the function, this study calculates using the following formula (Hara et al., 2009; Satoh et al., 2006):

when v=80mRMR(s,80)= 10.0 (s≦ - 0.25)RMR(s,80)= -58.07s - 4.52 (0.25≦ s≦ -0.11)RMR(s,80)= 3.113e4.614s (-0.11≦ s≦ 0.25)RMR(s,80)= 10.0 (0.25≦ s)when s= 0RMR(0,v)= 0.4366e0.0246v

This is assigned to the following equation:RMR = RMR(s, 80) + RMR(0. v) – RMR(0, 80)

The metabolic conversion distance can be calculated from the above equation.

4. Case study

4.1. Features of study area

The Japanese population made a significant transformation from rural areas to large cities in the 1960s and 1970s due to labordemand in urban areas. To meet the enormous demand for housing, a policy of ownership with a focus on own-construction was

Table 3Symbol legend.Variable Description Units

R Walking load –E Energy consumption kcalL Metabolic conversion distance mRMR Relative metabolic rate –BMR Basal metabolic rate kcalW Weight kgT Walking time mina Estimated energy requirement of day per weight kcal/kg/dayL Length mv Walking speed m/minS Slope %


promoted so that suburban areas were rapidly converted to residential areas by the private sector (Ishibashi & Taniguchi, 2005).Examples in Tokyo include Tama New Town and Tama Garden City, but for this study, we focused on Aoba Ward, Yokohama City,which is one of these housing complexes. Our targeted area corresponds to the second phase of Tama Garden City. The Radburn streetstructure with a view to the car society was adopted at the center of the area (Abe, Koshizawa, & Sakai, 2007). Thus, it is known as apopular residential area in the metropolitan area. The development has received high reviews, having been awarded the 1987Architectural Institute of Japan Award, the 1989 Green City Award (Prime Minister's Prize), and the 2002 Japan City PlanningSociety Award (Ishikawa Prize).

On the other hand, problems in car-centered urban development and residential land formation remote from the city center inanticipation of land price escalation are becoming apparent with the aging of the town (Matsubara, 1982; Toura, 2018). The AobaWard is located on the Tama Hills, with steep slopes as shown in Fig. 1. The slopes are a challenge for elderly persons to walk if theyhave to surrender their driving licenses. In response to this, Tokyu Corporation, the developer of this area, and the municipality ofYokohama initiated a citizen participatory project, entitled Community Development for Next Generation of Suburban Town.

4.2. Convenience stores in study area

There are 86 convenience stores in Aoba Ward, basically concentrated in front of train stations but sparse in residential areasalong main roads. As shaded pink in Fig. 1, about 47% of the area is zoned under Japan's City Planning Act as a “Category 1 low-riseexclusive residential district,” where convenience stores are not permitted.

4.3. Managing data

In calculating the metabolic conversion distance, we used the coordinate data of the stores, the population data apportioned foreach house, and road data for considering the physical load.

To estimate population for each house, we used 2015 census data and a 2013 City Planning Basic Survey by Yokohama City. First,buildings for housing were extracted from the City Planning Basic Survey, and population data from the census was distributedaccording to the total floor area of each building.

Finally, road section data considering physical load was created from three types of data: section length, slope, and physical load

Fig. 1. Convenience store locations in Aoba ward.


by walking. The road section data originates to the road center line data of the digital National Base Map. In addition, in order tocalculate the slope of each road section, first, the 5m grid Digital Terrain Model by the Geographical Survey Institute was set to theend points of each section of the road center line. Then the slope was calculated for each section. In order to reflect the averageJapanese physical ability, the parameters shown in Table 4 were adopted from the report updated by the Ministry of Health, Laborand Welfare every five years (MHLW, 2015).

Based on these, we calculated the metabolic conversion distance for each road section and estimated the area of the deliverydesert by using “Service Area Analysis,” an ArcMap tool for network analysis.

4.4. Results of analysis

Below we calculate the “delivery desert occupancy rate,” which we define as the ratio of adult residents in Aoba Ward who live ina delivery desert. We used the service area analysis function of ArcMap to estimate, by age group, the population living in a deliverydesert. We found that the overall delivery desert occupancy rate was 64.82% for all adults in the study area. This is ratio of all adultswithout walkable access to any convenience store.

Using the same methodology, we also calculated delivery deserts for adults relying specifically on any one of the three majordelivery companies, which together boast a 94.4% market share for package deliveries in Japan: Yamato Transport Co., Ltd., SagawaExpress, and Japan Post (MLIT, 2018). The results are shown in Fig. 2. Our study revealed that the company-specific delivery desert

Table 4Parameters for metabolic conversion calculations.Age v (m/min) BMR (kcal/kg/min) a (kcal/day)

20∼29 80.24 0.01601 41.9330∼39 80.76 0.01528 38.6940∼49 79.1250∼59 70.19 0.01465 37.5260∼69 63.1870∼ 56.22 36.67

Fig. 2. Map of delivery deserts in Aoba Ward.


occupancy rates were 73.46% for Yamato Transport, 83.63% for Sagawa Express, and 74.72% for Japan Post.The area not classified as delivery desert in the Category 1 low-rise exclusive residential districts was 26.11%, and 27.23% for

Aoba Ward as a whole.Our study also clarified the average percent of population per square kilometer living within walking distance of a convenience

store. Fig. 3 shows that convenience stores have lowest coverage for residents in their thirties and forties. This figure also revealsdifferences in the apparent target segment of each company: Relatively speaking, it appears Lawson tends to attract customers in theirtwenties, 7-Eleven senior customers, and FamilyMart the middle age range.

5. Discussion

5.1. Discrepancies between delivery desert in study area and actual situation

Our study showed that 64.82% of the population in the study area lived in a delivery desert. In addition, we found that thedelivery desert occupancy rate ranged from 73% to 84% depending on the delivery company, as shown in Table 5. However, inapparent contradiction, the aforementioned survey of home delivery service users found that about 90% of respondents answeredthat they preferred a convenience store as a method of receiving package delivery services without the need for redelivery (MLIT,2015b). In other words, respondents have the perception that they can easily access convenience stores. Why does this gap occur?

A visual interpretation of satellite images revealed that more than half of the convenience stores in Aoba Ward had parking lots.Perhaps customers hesitate to pick up courier items by walking to the store if the items are heavy. This observation suggests thatwhen discussing convenience stores in suburban areas it is also important to consider the means of transportation and the en-vironmental costs that accompany them. This study underestimated the potential contribution of convenience stores to eliminate thedelivery desert in suburban areas where driving a vehicle for errands is common practice. Although it has been suggested that e-commerce would help reduce greenhouse gases emissions (Siikavirta et al., 2008), at least in our study area, the last mile still largelydepends on the customer driving a vehicle.

Fig. 3. Average percent of population per square kilometer living within walking distance of a convenience store (adjusted by age group andconvenience store chain).

Table 5Delivery desert occupancy rates for adults in study area.Delivery company Yamato Sagawa Japan Post

Contracting convenience store chain 7-ElevenFamilyMart

Lawson FamilyMartLawson

Service area (residents in 20s) 18.08 km2 10.17 km2 16.16 km2Delivery desert rate (residents in 20s) 48.60% 71.10% 54.06%Service area (residents in 70s) 9.60 km2 5.04 km2 8.56 km2Delivery desert rate (residents in 70s) 72.71% 85.67% 75.66%Delivery desert occupancy rate 73.46% 83.46% 74.72%


5.2. Delivery deserts and regulations affecting store locations

Laws and regulations have been important factors restricting the location and service areas of convenience stores in Japan(Ishikawa & Asami, 2013). In suburban areas, the location of convenience stores is strictly regulated in land use planning for thepurpose of maintaining the living environment (Mikado & Nakade, 1995). Such regulations have affected the distribution of thestores in the study area and influenced the results of our assessment of the delivery desert. The ratio of land area not classified as adelivery desert was 26.11% in the Category 1 low-rise exclusive residential districts and 27.23% in the zoning district. Certainly, thecoverage of convenience stores tends to be somewhat low within Category 1 low-rise exclusive residential districts. However, the mapin Fig. 1 shows that convenience stores are mostly located not within but between the Category 1 low-rise exclusive residentialdistricts. This reflects the urban planning zoning system. Roadsides and intersections are normally excluded from Category 1 low-riseexclusive residential districts in order to protect a good living environment. Interviews conducted by the author confirmed that thisperception is present among residents in community, based on concerns that a good living environment will be damaged by noise anddisruptions from living close to a convenience store. This perception appears to be stronger in areas that have a community en-vironmental code. At the core of the debate about locations of convenience stores is a kind of tradeoff in convenience versus quality ofliving conditions.

5.3. Private sector constraints in solving the delivery desert problem

The delivery desert and a convenience store's pursuit of profits are inseparable issues. Convenience store chains use area mar-keting techniques to open shops in areas with many potential customers. However, as shown in Fig. 3, in this study area, cohorts intheir thirties and forties tend not to live in areas where it is easy to access convenience stores, although these age ranges are mostlydouble-income households, and they are increasing in number in recent years. Those households are presumably in the busy years ofraising families and would have higher needs for convenient courier delivery services, compared seniors who might tend to staylonger at home during the day. As a business strategy, while convenience stores certainly provide many social infrastructure func-tions, they also need to consider the changing population trends and consumer needs in the areas they serve.

6. Conclusion

Using a study area in Aoba Ward, Yokohama City, part of the Tokyo Metropolitan Area, this study examined the potential forcourier package reception services at convenience stores to be a solution to the redelivery problem. The results showed that thedelivery desert by our definition covers up to about 65% of the total land area in the study area. On the other hand, a survey showedthat 90% of convenience store users preferred to use a convenience store to pick up their courier deliveries and avoid the need forredelivery (MLIT, 2015b). This gap arises in suburban areas where there are limited options to get to a convenience store only bywalking.

Meanwhile, the percentage of land area not classified as a delivery desert was about 26% for Category 1 low-rise exclusiveresidential districts (where in principle, convenience stores are not permitted), and 27% for Aoba Ward as a whole. There is certainlya tendency for coverage to be rather low in the Category 1 low-rise exclusive residential districts, but our study showed that reg-ulations on the location of convenience stores are a very minor factor in the delivery desert.

Generally, convenience stores, which are privately owned, are deployed in Japan to maximize profits by targeting a particularcustomer segment (Kato, 2016). For example, in the past, convenience stores were most popular for customers under age 30, but thetarget market has shifted to higher age segments (Kato, 2016). Fig. 3 shows that 7-Eleven has an advantage in popularity amongseniors, Lawson among young customers, and FamilyMart among customers in their middle years. Our study described differencesbetween the target segments for each convenience store chain. Furthermore, the accessibility of convenience stores for people in theirthirties and forties was low in the study area, indicating a bias in the age structure of the beneficiaries. Our study has described thedelivery desert in the study area and suggests that there are limits to the contributions convenience stores can make to solve theredelivery problem.

However, some limitations should be noted regarding this study. First, this method does not accurately estimate how many peopleare having trouble in receiving packages. For example, elderly persons who cannot drive a car tend to be at home and able to receivea delivery, or a family member may be at home and receive a package on behalf of the official addressee. Such cases were notconsidered in this paper, so our calculations may overestimate the number of people who live in the delivery desert. Moreover, ourstudy area this time was a hilly neighborhood which is heavily car-dependent. Although such conditions are quite common in Japan,the results may vary in neighborhoods with different urban designs, densities, and topographies. Such variations remain for futurestudy.

Acknowledgments

This study was done as part of the M-NEX (no.11314551), a Belmont Forum granted project of Collaborative Research Area onSustainable Urbanization Global Initiative (SUGI)/Food-Water-Energy Nexus.” We would like to thank the City of Yokohama forproviding its basic survey on urban planning, and the Jisedai Kogai Machidukuri (Next Generation Suburban Planning) WISE LivingLab for their cooperation as a research base.


References

Abe, Y., Koshizawa, A., & Sakai, A. (2007). Planning of pedestrian road and change of land-use in utsukushigaoka, Yokohama. AIJ Journal of Technology and Design,13(26), 751–755. https://doi.org/10.3130/aijt.13.751.

Beaulac, J., Kristjansson, E., & Cummins, S. (2009). A systematic review of food deserts, 1966-2007. Preventing Chronic Disease, 6(3), A105. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/19527577.

Cabinet Office Public Relations Office (2017). Overview of “polls on redelivery issues.”.Gahinet, M.-C., & Cliquet, G. (2018). Proximity and time in convenience store patronage: Kaïros more than chronos. Journal of Retailing and Consumer Services, 43, 1–9.

https://doi.org/10.1016/J.JRETCONSER.2018.02.008.Hall, C., Figueroa, A., Fernhall, B. O., Kanaley, J. A., & Kanaley, J. (2004). Energy expenditure of walking and running: Comparison with prediction equations.

Prediction Equations. Med. Sci. Sports Exerc, 36(12), 2128–2134. https://doi.org/10.1249/01.MSS.0000147584.87788.0E.Hara, T., Ishizaka, K., & Ohashi, Y. (2009). Evaluation of residential area from a point of elderly people's walking accessibility in local center city. Journal of

Architecture and Planning (Transactions of AIJ), 74(635), 129–135. https://doi.org/10.3130/aija.74.129.Hwang, I.-C., & Takada, K. (1995). Resolving parking problems near convenience stores. Journal of the City Planning Institute of Japan, 30, 661–666. https://doi.org/10.

11361/journalcpij.30.661.Ishibashi, N., & Taniguchi, H. (2005). Study on the Planning Process for the Development of “Tama Garden City” : Study on the planning of suburban residential area

developed through a combination of land readjustment projects (1). Journal of Architecture and Planning (Transactions of AIJ), 70(598), 129–136. https://doi.org/10.3130/aija.70.129_3.

Ishikawa, T., & Asami, Y. (2013). Residents' perception of the quality of urban living and attitudes toward mixed land use. Urban Housing Sciences, 2013(81), 98–107.https://doi.org/10.11531/uhs.2013.81_98.

Japan Franchise Association (2009). About declaration of convenience store as social infrastructure. Retrieved from http://www.jfa-fc.or.jp/misc/static/pdf/090528.pdf.Kanamori, T., & Yan, W. (2018). Walking accessibility assessment of elderly people in collective relocation and their effects on lifestyle after the great east Japan

earthquake —a case study in kesennuma city, Japan. Journal of Japan Society of Civil Engineers, 74(4), 261–274. Ser. D3 (Infrastructure Planning and Management)https://doi.org/10.2208/jscejipm.74.261.

Kato, N. (2016). Direction of food development in the aging society. Journal of the Japanese Society of Taste Technology, 14(2), 36–42. https://doi.org/10.11274/bimi.14.2_36.

Lin, Y. (2018). E-urbanism: E-Commerce, migration, and the transformation of taobao villages in urban China. Cities. https://doi.org/10.1016/J.CITIES.2018.11.020.Matsubara, H. (1982). A study of large-scale residential development by private railway enterprises -the case of Tama garden city-. Geographical Review of Japan, 55(3),

165–183. https://doi.org/10.4157/grj.55.165.METI (2018). 2017: Development of infrastructure for data-driven society in Japan.METI, & MLIT (2018). Productivity improvement liaison conference for home delivery business and EC business.MHLW (2015). Dietary reference intakes for Japanese (2015).Mikado, I., & Nakade, B. (1995). Study on the effects for residential environments of CVS in residential districts. Journal of the City Planning Institute of Japan, 30,

163–168. https://doi.org/10.11361/journalcpij.30.163.MLIT (2015a). About estimate of social loss by outbreak of redelivery of home delivery.MLIT (2015b). Study group on promotion of diversification of delivery methods to reduce redelivery.MLIT (2018). Survey and count method of the number of delivery services (2017). Retrieved from http://www.mlit.go.jp/common/001252227.pdf.Morganti, E., Seidel, S., Blanquart, C., Dablanc, L., & Lenz, B. (2014). The impact of E-commerce on final deliveries: Alternative parcel delivery services in France and

Germany. Transportation Research Procedia, 4, 178–190. https://doi.org/10.1016/J.TRPRO.2014.11.014.Nakamura, Y., Matsumoto, H., Yamamoto-Mitani, N., Suzuki, M., & Igarashi, A. (2018). Impact of support agreement between municipalities and convenience store

chain companies on store staff's support activities for older adults. Health Policy, 122(12), 1377–1383. https://doi.org/10.1016/j.healthpol.2018.09.015.Nogimura, T. (2015). A review of controversy on convenience store industry regulation in Japan. Journal of Japan Management Diagnosis Association, 15, 93–98.

https://doi.org/10.11287/jmda.15.93.Satoh, E., Yoshikawa, T., & Yamada, A. (2006). Investigation of converted walking distance considering resistance of topographical features and changes in physical

strength by Age : Model for location planning of regional facilities considering topographical condition and aging society Part 1. Journal of Architecture andPlanning (Transactions of AIJ), 71(610), 133–139. https://doi.org/10.3130/aija.71.133_2.

Siikavirta, H., Punakivi, M., Kärkkäinen, M., & Linnanen, L. (2008). Effects of E-commerce on greenhouse gas emissions: A case study of grocery home delivery inFinland. Journal of Industrial Ecology, 6(2), 83–97. https://doi.org/10.1162/108819802763471807.

Sudo, M., & Masuda, S. (2014). A convenience store's evolution in the retail business and its new challenge as the core business player in the society. TechnologicalInnovation and Social Change, 7(1), 14–32.

Suzuki, Y., Kimura, K., Hino, S., & Kaneko, Y. (2014). Shopping status of elderly and support measures from the view point of various of shopping. Journal of JapanSociety of Civil Engineers, 70(5), I_371–I_382. Ser. D3 (Infrastructure Planning and Management) https://doi.org/10.2208/jscejipm.70.I_371.

Takemoto, R. (2015). Estimated by municipality of convenience store refugees. Retrieved from https://www.smtri.jp/report_column/report/pdf/report_20150810.pdf.Toura, R. (2018). Shitetsu 3.0. Wanibooks.Walker, R. E., Keane, C. R., & Burke, J. G. (2010). Disparities and access to healthy food in the United States: A review of food deserts literature. Health & Place, 16(5),

876–884. https://doi.org/10.1016/j.healthplace.2010.04.013.van der Walt, W. H., & Wyndham, C. H. (1973). An equation for prediction of energy expenditure of walking and running. Journal of Applied Physiology, 34(5),

559–563. https://doi.org/10.1152/jappl.1973.34.5.559.Yamato Transport Co (2004). Kuroneko Yamato does not change. Retrieved April 4, 2019, from http://www.yamato-hd.co.jp/news/pdf1/iken040825.pdf.Yuasa, Y., & Ikegame, T. (2001). The proposition of the conceptual model on the coming style of convenience store : The basic viewpoint for the composition of the

method of SDS(Sustainable Design System)(Technical Session Contents). Proceedings of the Annual Conference of Jssd, 48, 290–291. https://doi.org/10.11247/jssd.48.0_290.


https://doi.org/10.3130/aijt.13.751http://www.ncbi.nlm.nih.gov/pubmed/19527577http://www.ncbi.nlm.nih.gov/pubmed/19527577http://refhub.elsevier.com/S2226-5856(19)30130-X/sref3https://doi.org/10.1016/J.JRETCONSER.2018.02.008https://doi.org/10.1249/01.MSS.0000147584.87788.0Ehttps://doi.org/10.3130/aija.74.129https://doi.org/10.11361/journalcpij.30.661https://doi.org/10.11361/journalcpij.30.661https://doi.org/10.3130/aija.70.129_3https://doi.org/10.3130/aija.70.129_3https://doi.org/10.11531/uhs.2013.81_98http://www.jfa-fc.or.jp/misc/static/pdf/090528.pdfhttps://doi.org/10.2208/jscejipm.74.261https://doi.org/10.11274/bimi.14.2_36https://doi.org/10.11274/bimi.14.2_36https://doi.org/10.1016/J.CITIES.2018.11.020https://doi.org/10.4157/grj.55.165http://refhub.elsevier.com/S2226-5856(19)30130-X/sref15http://refhub.elsevier.com/S2226-5856(19)30130-X/sref16http://refhub.elsevier.com/S2226-5856(19)30130-X/sref17https://doi.org/10.11361/journalcpij.30.163http://refhub.elsevier.com/S2226-5856(19)30130-X/sref19http://refhub.elsevier.com/S2226-5856(19)30130-X/sref20http://www.mlit.go.jp/common/001252227.pdfhttps://doi.org/10.1016/J.TRPRO.2014.11.014https://doi.org/10.1016/j.healthpol.2018.09.015https://doi.org/10.11287/jmda.15.93https://doi.org/10.3130/aija.71.133_2https://doi.org/10.1162/108819802763471807http://refhub.elsevier.com/S2226-5856(19)30130-X/sref27http://refhub.elsevier.com/S2226-5856(19)30130-X/sref27https://doi.org/10.2208/jscejipm.70.I_371https://www.smtri.jp/report_column/report/pdf/report_20150810.pdfhttp://refhub.elsevier.com/S2226-5856(19)30130-X/sref30https://doi.org/10.1016/j.healthplace.2010.04.013https://doi.org/10.1152/jappl.1973.34.5.559http://www.yamato-hd.co.jp/news/pdf1/iken040825.pdfhttps://doi.org/10.11247/jssd.48.0_290https://doi.org/10.11247/jssd.48.0_290

The Inter-University Symposium on Asian Megacities, Khabarovsk, Russia, September 4–7, 2019

Analysis on Food Environmental Gaps among Residential Areas

in Metropolitan Suburbs

Shun Nakayama1, Wanglin Yan2

1 Keio University Graduate School of Media and Governance, Fujisawa city, Japan, [email protected] 2 Keio University Faculty of Environment and Information Studies, Fujisawa city, Japan, [email protected]

ABSTRACT

This study analyzed the food environment in metropolitan suburban Tokyo, Japan, to clarify the spatial mismatch between supply and demand for food as a kind of gaps. In recent years, Japanese society is shrinking because of aging and population decreasing. This phenomenon increases the number of elderly people without mobility in the car-dependent society. And at the same time, as a result of pursuing economic rationality, stores has being consolidated to populous areas so that food environment, in particular the accessibility to grocery shops has being deteriorated. This study developed a model to assess the accessibility of grocery stores, and food environmental gaps by analyzing the spatiotemporal change of accessibility in suburban town of Tokyo Metropolitan. A foodshed is defined as the area where a resident can access the grocery store within 5 minutes by walk, and a model that considers physical load relative to customer age and topographic slope is developed to calculate the foodshed. The result shows that (1) the choice of grocery store is related to its location, and there is a gap in the area of 1000 to 1500m from a railway station, (2) there is a difference of the foodshed with type of urban development. These findings could serve as a reference in policy discussions for solving the food accessibility problem in shrinking society.

Keywords: foodshed, gap analysis, GIS, shrinking society

1. INTRODUCTION Japan's population peaked in 2008 and turned to decline, becoming a shrinking society. At the same time, the declining birthrate and aging population are progressing. The concern with these issues is the decline in the quality of life. One example is the issue of access to grocery stores. Population decline reduces the absolute amount of demand and increases the ratio of service maintenance costs to store profits. Also, the declining birthrate and aging population shift the center of demand from young people to elderly people. Due to the decline in physical ability, the daily living area of the elderly is narrow, which reduces the demand for shops and services in the conventional daily living area. As a result, there is a possibility that the local small supermarkets and shops will be closed. Furthermore, when there are no more stores in the daily life area, use of large stores by private cars will increase. Furthermore, the demand for shops and services in daily life areas is reduced. After several decades, the number of elderly people is higher than now, and the number of people who can not drive a car for reasons such as return of licenses increases. However, the convenience would be very bad at that time because small supermarkets and shops were not in the daily life area. Therefore, the problem of access is the problem of spatial mismatch between the elderly without

Shun NAKAYAMA, Wanglin YAN

2

mobility and the stores that can not stand the low demand and withdraw in a car-dependent society. This problem is a problem that produces a negative spiral. In this research, we analyzed the change over time and space about the decrease of the store as a foothold to consider this problem. In this analysis, we used data from five periods since 1998 to visualize the number of stores that can be accessed on foot. This clarified where food access problems occur spatially. The grocery store is the store that sells fresh food.

2. QUANTITATIVE EVALUATION The attributes of people who live in the city are diverse, and the food demand varies depending on their attributes. [1][2].It has become clear that these people need options that can access diverse food needs in order to enjoy the same high-quality diet[3]. In other words, it is one of the conditions of a good food environment to have various food options that correspond to the various needs of the various residents living in the city. Therefore, in this study, the number of options for food stores was evaluated. As a method of evaluating walking access while also considering physical load, many studies have focused on energy expenditure [4], [5]. Thus, in order to consider physical load, we used the metabolic conversion distance instead of actual distance to calculate the area of a delivery desert, with reference to previous studies [6], [7]. The metabolic conversion distance is the actual distance calibrated with consideration of the physical load when walking. Parameters to calculate physical load include, walking speed, body weight, and basal metabolic rate by age as the attributes of pedestrians, and road slope as an attribute of roads. This is based on the assumption that the walking speed differs based on the pedestrian’s age, basal metabolic rate, and the amount of energy required per day per unit of body weight. The population accessible to convenience stores was obtained by conducting a service area analysis using metabolic conversion distance and estimating the population size living in the reachable area (a 5-minute walkable area, as mentioned above). The calculation of the metabolic conversion distance is shown below. The variables used in the following formulas are summarized in Table 1.

Table 1. Symbol legend

Variable Description Units R Walking load - E Energy consumption kcal L Metabolic conversion distance m RMR Relative metabolic rate - BMR Basal metabolic rate kcal W Weight kg T Walking time min a Estimated energy requirement of day per weight kcal/kg day L Length m v Walking speed m/min S Slope %


3

The metabolic conversion distance (L) is obtained by multiplying the actual distance (l) by the walking load (R) which takes into account elements such as the slope, then divided by walking load when the road is horizontal (R0) [7].

L = l × R / R0 Next, walking load (R) is “the energy consumption when walking from a certain point to a destination point divided by the energy necessary for one day” [6]. Energy consumption by walking (E) is expressed by the following equation:

E = (RMR + 1.2) × BMR × W × T Relative metabolic rate (RMR) represents the influence of topography such as road slope. Relative loads are indicated by the value of (RMR + 1.2), where the constant 1.2 is the resting metabolic rate during daytime life. Basal metabolic rate (BMR) x body weight (W) is basal metabolism and represents the minimum amount of energy required to maintain human life for one day. Because these differ depending on age and body weight, the equation is as shown above [8]. In addition, the influence of age is reflected in the movement time (T) as a decrease in walking speed (v).

R = E / (a × W) = (RMR + 1.2) × BMR × W × T × {1 / (a × W)} = (RMR + 1.2) × BMR × (l/v) × (1/a)

The rate of relative metabolic rate (RMR) varies with the slope (s) and the walking speed (v). Research in the past used the walking speed for each age group to substitute for the above moving speed (T), while in the calculation of the energy metabolic rate, many studies substituted 80 m/min, indicating that the accuracy was not necessarily high. Therefore, based on previous studies that fully consider the continuity of the function, this study calculates using the following formula [6], [7]:

when v=80m RMR(s,80) = 10.0 (s ≦ - 0.25) RMR(s,80) = -58.07s - 4.52 (0.25 ≦ s ≦ -0.11) RMR(s,80) = 3.113e4.614s (-0.11 ≦ s ≦ 0.25) RMR(s,80) = 10.0 (0.25 ≦ s)

when s=0 RMR(0,v) = 0.4366e0.0246v

This is assigned to the following equation: RMR = RMR(s, 80) + RMR(0. v) – RMR(0, 80)

The metabolic conversion distance can be calculated from the above equation.


4

3. CASE STUDY 3.1. Features of Study Area The Japanese population made a significant transformation from rural areas to large cities in the 1960s and 1970s due to labor demand in urban areas. To meet the enormous demand for housing, a policy of ownership with a focus on own-construction was promoted so that suburban areas were rapidly converted to residential areas by the private sector [9]. Examples in Tokyo include Tama New Town and Tama Garden City, but for this study, we focused on residential area of Yokohama City. The target area is approximately 3km long x 5.5km wide, including the area developed and sprawled. There is typical Japanese suburban residential area. Especially the development area is known as a popular residential area in the metropolitan area. The development has received high reviews, having been awarded the 1987 Architectural Institute of Japan Award, the 1989 Green City Award (Prime Minister’s Prize), and the 2002 Japan City Planning Society Award (Ishikawa Prize). On the other hand, problems in car-centered urban development and residential land formation remote from the city center in anticipation of land price escalation are becoming apparent with the aging of the town [10], [11]. The Aoba Ward is located on the Tama Hills, with steep slopes as shown in Figure 1. The slopes are a challenge for elderly persons to walk if they have to surrender their driving licenses. In response to this, Tokyu Corporation, the developer of this area, and the municipality of Yokohama initiated a citizen participatory project, entitled Community Development for Next Generation of Suburban Town.

3.2. Target stores There are several studies focusing on living related facilities for suburban residential areas in Japan, and the relationship between the development method and the occurrence of regional facilities has been clarified, the summary of the occurrence of regional facilities around the railway station, the suburbs What analyzed and patterned the secular change of the regional facilities in the residential area[12]–[15]. On the other hand, accessibility has not been sufficiently examined in these studies. Therefore, this study focused on the analysis of the relationship between the change in store location and the accessibility.

Table 2. Parameters for metabolic conversion calculations

Age v (m/min) BMR

(kcal/kg/min)

a (kcal/day)

Over 70 56.22 0.01465 36.67


5

3.3. Managing Data In calculating the metabolic conversion distance, we used the coordinate data of the stores, the store data, and road data for considering the physical load. For store data, We used Zenrin's Zmap TOWN II from CSIS, the University of Tokyo. Among these, there is nameplate data, and after discrimination by Python from here, discrimination was further conducted visually. Road section data considering physical load was created from three types of data: section length, slope, and physical load by walking. The road section data originates to the road center line data of the digital National Base Map. In addition, in order to calculate the slope of each road section, first, the 5 m grid Digital Terrain Model by the Geographical Survey Institute was set to the end points of each section of the road center line. Then the slope was calculated for each section. In order to reflect the average Japanese physical ability, the parameters shown in Table 2 were adopted from the report updated by the Ministry of Health, Labor and Welfare every five years [16]. Based on these, we calculated the metabolic conversion distance for each road section and estimated the number of store choices by using “Service Area Analysis,” an ArcMap tool for network analysis.

4. RESULT The relationship between the distance to the nearest station from each house and the number of grocery store options was analyzed using the K-means method, an unsupervised clustering method, using python. The results shown in Fig. 1 were obtained as a result of setting the

Fig 1. Clustering the number of food choices and the distance (m) from each residential station (From upper left: 1998, 2003, 2008 From lower left: 2013, 2016)


6

number of clusters appropriate for the previous analysis. First of all, the number of options was initially at a maximum of 12 stores, but it is decreasing year by year. Also, initially there were the most grocery options available in the area 2500 m to 3000 m from the station. After that, the number of stores around 1500m has increased slightly. Furthermore, it became clear that there were few food store options from around 1000m to 1500m.

5. DISCUSSION From this analysis result, it has become clear that the number of stores that can be accessed by foot is decreasing overall. Looking at the details, first of all, the choice is decreasing rapidly in the area 3000 m from the nearest station. In this area, there are many application areas where commercial facilities can be located, and there are many small-scale private stores such as shops and liquor stores, but it is thought that the decrease due to overconcentration on large stores in front of the station. Then there was an increase in grocery store options in the 1500 m zone. This seems to be related to the opening of a convenience store at the roadside store. Furthermore, the area between 1000m and 1500m consistently has fewer options than originally. There are many low-rise detached residential areas in this area, and many places are designated as restricted areas where the opening of stores is strictly restricted. And as the number of accessible stores increases as the station is approached, it is considered that this is because the station front is defined as a commercial use area. However, some limitations should be noted regarding this study. Because I could not use high-precision data. For the location data of the store, Zenrin's Zmap TOWN II obtained from the University of Tokyo CSIS was used. However, it was found that there were no supermarkets or convenience stores that should be located in the station building. We left a challenge in accurate data maintenance of food stores in large commercial facilities. In addition, in this study, it was possible to go back only to 1998 due to the data improvement. In the future, it is necessary to proceed with work such as digitizing paper maps..

6. CONCLUSION In this study, with the purpose of making it possible to view the change in the number of stores in the residential area on the outskirts of a large city in space and time, we take up part of the Tama rural city and its surroundings The We evaluated the transition of the number of accessible food stores and considered the factors. It has become clear that the number of accessible grocery store options in the area is reduced compared to the past. This is considered to be due to the decrease in small stores. In addition, cluster analysis of the distance from the station and the number of choices of accessible grocery stores showed improvement in the food environment such as an increase in the number of choices in the zone 1500 to 2000 m from the station. On the other hand, it was revealed that the number of food options did not increase in the zone of 1000m to 1500m, and it was the lowest value in the area. There are many detached housing areas formed in the early stages of development in this area, and it is thought that problems with food access will be exposed as the population ages.


7

REFERENCES [1] D. M. Nash, J. A. Gilliland, S. E. Evers, P. Wilk, and M. K. Campbell, “Determinants

of Diet Quality in Pregnancy: Sociodemographic, Pregnancy-specific, and Food Environment Influences,” J. Nutr. Educ. Behav., vol. 45, no. 6, pp. 627–634, Nov. 2013.

[2] H. Patrick and T. A. Nicklas, “A Review of Family and Social Determinants of Children’s Eating Patterns and Diet Quality,” J. Am. Coll. Nutr., vol. 24, no. 2, pp. 83–92, Apr. 2005.

[3] B. A. Laraia, A. M. Siega-Riz, J. S. Kaufman, and S. J. Jones, “Proximity of supermarkets is positively associated with diet quality index for pregnancy,” Prev. Med. (Baltim)., vol. 39, no. 5, pp. 869–875, Nov. 2004.

[4] W. H. van der Walt and C. H. Wyndham, “An equation for prediction of energy expenditure of walking and running.,” J. Appl. Physiol., vol. 34, no. 5, pp. 559–63, May 1973.

[5] C. Hall, A. Figueroa, B. O. Fernhall, J. A. Kanaley, and J. Kanaley, “Energy Expenditure of Walking and Running: Comparison with Prediction Equations,” Predict. Equations. Med. Sci. Sport. Exerc, vol. 36, no. 12, pp. 2128–2134, 2004.

[6] T. Hara, K. Ishizaka, and Y. Ohashi, “Evaluation of Residential Area From a Point of Elderly People’s Walking Accessibility in Local Center City,” J. Archit. Plan. (Transactions AIJ), vol. 74, no. 635, pp. 129–135, 2009.

[7] E. Satoh, T. Yoshikawa, and A. Yamada, “Investigation of Converted Walking Distance Considering Resistance of Topographical Features and Changes in Physical Strength by Age : Model for location planning of regional facilities considering topographical condition and aging society Part 1,” J. Archit. Plan. (Transactions AIJ), vol. 71, no. 610, pp. 133–139, 2006.

[8] T. Kanamori and W. Yan, “Walking Accessibility Assessment of Elderly People in Collective Relocation and Their Effects on Lifestyle after the Great East Japan Earthquake —A Case Study in Kesennuma City, Japan—,” J. Japan Soc. Civ. Eng. Ser. D3 (Infrastructure Plan. Manag., vol. 74, no. 4, pp. 261–274, 2018.

[9] N. Ishibashi and H. Taniguchi, “Study on the Planning Process for the Development of ‘Tama Garden City’ : Study on the planning of suburban residential area developed through a combination of land readjustment projects (1),” J. Archit. Plan. (Transactions AIJ), vol. 70, no. 598, pp. 129–136, 2005.

[10] R. Toura, Shitetsu 3.0. Wanibooks, 2018. [11] H. Matsubara, “A study of Large-Scale Residential Development by Private Railway

Enterprises -The Case of Tama Garden City-,” Geogr. Rev. Jpn., vol. 55, no. 3, pp. 165–183, 1982.

[12] A. Kita, “Changes of Community Facilities and Residents&Apos; Attitude toward Residential Environment in Suma New Town : Maturity process and changes of


8

residents' opinions in new town,” J. Archit. Plan. (Transactions AIJ), vol. 72, no. 616, pp. 99–106, 2007.

[13] M. Lee, S. Kashihara, H. Yoshimura, and T. Yokota, “On The Temporal Change Of Distribution Patterns of Community Facilities around Railway Stations,” J. Archit. Plan. (Transactions AIJ), vol. 59, no. 455, pp. 77–86, 1994.

[14] H.-S. Kim, K. Okada, S. Kashihara, H. Yoshimura, and T. Yokota, “On Method Of Predicting Generation of Community Facilities In New Towns : Study on Supply Planning of Community Facilities in New Towns,” J. Archit. Plan. Environ. Eng. (Transactions AIJ), vol. 407, pp. 97–105, 1990.

[15] S. Kashihara, K. Okada, H. Yoshimura, T. Yokota, and H.-S. Kim, “On Relations Between Development Techniques and Distribution Patterns Of And Generated Quantity of Community Facilities : Study on supply planning of community facilities in new towns,” J. Archit. Plan. Environ. Eng. (Transactions AIJ), vol. 404, pp. 69–78, 1989.

[16] MHLW, “Dietary Reference Intakes for Japanese (2015),” 2015.

t1r-!/) z9(2 :hvn - keio universitytotal convenience store customers in 2013 (kato, 2016). in recent...

Documents