1 Introduction

The purpose of this study is to predict the demand for local public goods, and to examine a method to promote the diversification of the expense burden of supply.

Nowadays public transport is non-existent in many areas of Japan. For person who is unable to drive cars including elderly people and children this is obviously detrimental in terms of mobility restrictions. However, there are cases where community bus operations have been attempted, and succeeded.

If the community bus is a local public good, and that bus service is run by affected residents, this amounts to the private supply of a local public good. According to Shiozu et al. [3] if a reciprocity motive works vis-à-vis the personal supply of local public goods, theoretically under an unsociable network, all members of the network incur a flat expense; more than the expense that social advantage should be in the situation.

However, efforts to instigate community bus services, are not guaranteed to succeed; for example, a consensus might not be established in its favor, or demand might be less than the expense.

It is known that some people find value in knowing that community bus services will exist in the future when their mobility might be restricted. Oguma et al. calls this option value and suggests using the Contingent Valuation Method (CVM) for its measurement.

CVM is technique to measure the economic value of non-market goods and services; it has been applied widely in the academic literature, and moderately in actual decision-making contexts. Because subjects are asked to directly reveal their economic wants and desires, CVM is commonly referred to as an expressed preference technique.

Oguma et al. [2] points out that expressing value in this case-study context raises challenges because the operational cost of the bus service needs to be financed via fares. Assuming that travel expenses are covered by fare income, it is necessary to set fares accordingly or compensate to some degree with subsidies. Because many local governments are already financially challenged, the latter method is not realistic. Additionally, fairness of the cost burden is an important consideration.

In this study, we operationalized a questionnaire which contained questions regarding usage of mobile phone wallets (payment functionality on smart devices) and options for covering operational costs associated with maintaining a community bus service. We found that there was little use of mobile phone wallets. We show that they prefer to bear the cost by fare rather than annual fee and some of them find out the option value. In addition, the quantity of demand did not decrease substantively following fare hikes; but a non-marginal increase in demand is predicted where fare reductions occur. In consideration of the result from other survey which shows that the use of non-contact plastic IC card has been promoted, we suggest to introduce a plural fare system for collecting option value with the non-contact plastic IC card.

The remainder of the paper is organized as follows. In Sect. 2 we describe the survey district. Next, in Sect. 3 we survey the use situation of the wallet phone. In Sect. 4 we estimate community bus demand using the CVM. Finally, Sect. 5 concludes.

2 Survey District

The case-study area is Makishima, Uji-city, Kyoto prefecture. This area is located across suburbs of 3 big Japanese cities. The population of the district is 15,225 people in October 2016. It increases a little than last year. The elderly constitute 21.9% of the population, which is lower than the 26.7% average in Japan. Figure 1 is a map of Uji-city; the Makishima district is located in the northwestern part of the city. Two private railroads run in the district, and a bus operates in north and south directions. There is currently no east-west bus service; as such, the mobility of some residents is restricted.

Fig. 1.
figure 1

Uji-city Map

Full size image

3 Data

We gathered data, via a questionnaire survey, about usage of cellphone payment functionality, and preferences for bus operation. We denote the former as questionnaire test 1 and the latter as test 2.

3.1 Test 1: The Cellphone Questionnaire

We distributed this questionnaire via mail to 8,000 inhabitants in the survey area in February, 2012. The number of effective responses was 1956. The focal item concerned mobile phone electronic payment functionalities. The distribution of respondent ages in this sample is depicted in Fig. 2; the modal age interval therein is respondents in their 60s.

Fig. 2.
figure 2

Distribution of age (test 1)

Full size image

Table 1 expresses two-way frequency data pertaining to this sample in terms of age and the use of mobile phone wallet/QR code functionalities.

Table 1. The use of mobile phone wallet/QR code functionalities by respondent’s age
Full size table

From Table 1, excepting people in their 80s, the choice “I have not used the mobile phone wallet or QR code” is the modal response for all age categories. There is a statistically significant difference in terms of usage of these functionalities if we compare people before 70s and after 70s.

3.2 Test 2: Bus Operation Survey

In the areas that we investigated vis-à-vis cellphone usage, we carried out a further survey about bus services on 1,400 inhabitants. The investigation period is from 15th October 2016 to November 15th. Survey items are about annual contribution payment amount and motivation when introducing a membership fee system when driving a bus, about fare payment amount and motivation when traveling without introducing membership system. The number of effective responses was 438. The distribution of respondent ages in this sample is depicted in Fig. 3.

Fig. 3.
figure 3

Distribution of age for community bus survey

Full size image

Comparing Fig. 3 with Fig. 2, there are more responses from elderly people. Because the focal item is a community bus service, people who have mobility restrictions could have higher propensities to complete the questionnaire because it relates to their well-being to a greater extent than those people without mobility restrictions.

Some areas have no bus services running to the station, city hall, or the hospital. Topographically, the area is flat, thus people without mobility problems can travel by bicycle and on foot. However, most inhabitants travel by privately owned cars.

3.2.1 Preferences for Contributing Toward Bus Operation Costs

To operationalize the CVM in this context, we asked respondents to express their fare payment intentions in monetary amounts; this was done in two phases based on a binary choice principle.

CVM is a method by which a researcher directly asks the subject about economic value. In particular, at first, the investigator presents the scenario to the subject, then the subject replies his/her willingness to pay. There are two ways to represent one’s willingness to pay. One is the subject directly presents the amount of payment. The other is the subject just declares approval or disapproval of the amount of payment which the investigator shows. Because both methods do not ask the subject to pay the willingness to pay, it is well known that the former method overestimates the willingness to pay. Recently, the researchers usually adopt the latter method. To use the latter method, the case where the investigator asks once is called single bound, and the case of twice is called double bound.

In this case, to operate bus needs initial cost, for example, vehicle cost or installation cost of bus stop, and operation cost. Suppose a case where these fixed costs are procured by annual membership fee and a case where they are procured from fare. We asked subjects by survey questionnaire of the binary choice double bound method for intention to pay for annual fee and intention to pay for freight.

3.2.1.1. Preferences for Annual Payment

First of all, we asked respondents if they would agree to pay 1,500 yen annually per household with a part of bus travel expenses as local burden (First bid price). Next, would they agree to pay 2,000 yen per year (Second bid price 1). Where respondents did not agree to paying the initial amount, they were asked to pay an annual payment of 1,000 yen (Second bid price 2). The results are shown in Table 2. Respondents were also asked to explain the reasons why the burden was acceptable.

Table 2. Preferences for annual payment
Full size table

We asked the person who said that they would pay annual payment even a little for the reason. About half of the 279 people who responded were concerned for the mobility and well-being of others (Fig. 4).

Fig. 4.
figure 4

Reasons why people agreed to a hypothetical annual payment

Full size image

Where applicable, we also asked people to explain why they did not want to pay the burden of bus operation. As a result, 61.8% of respondents said that the passengers on the bus should pay. Moreover, 10.8% (17 people) answered that they cannot afford to pay the contribution See Fig. 5.

Fig. 5.
figure 5

The reason why you may not pay the annual payment

Full size image

3.2.1.2. Bus Fare Preferences

Respondents who agree with the initial offering amount of 200 yen account for about 70% of the total effective responses. More than half agree to the fare of 250 yen. However, there are also a certain number of respondents who do not want to pay even if the fare is 150 yen (Table 3).

Table 3. Bus fare preferences
Full size table

We asked the respondent who would pay the bus fare even a bit for the reason. 172 people answered that they wanted to support if the service was principally designed to meet the travel needs of children and elderly people (Fig. 6).

Fig. 6.
figure 6

The reasons why you may pay the bus fare

Full size image

We also ascertained the reasons why 47 respondents said they did not want to pay the bus fare at all; the most frequent answers were that myself and family members did utilize a bus service (Fig. 7).

Fig. 7.
figure 7

The reasons why you may not pair the bus fare

Full size image

In the free description column, there was an opinion that when you and your family cannot drive a car now or in the near future, thinking of cooperating with bus operation might be advantageous. This suggests that option value exists in this case-study context.

4 Estimation of Willingness to Pay for Community Bus

In the case of this paper, when a community bus service is in operation, each household may bear the annual contribution of T yen, and may pay the t yen fare by getting on board. In both cases, if the community bus does not operate, the cost will be 0 yen. There are four answers. First one is the subject agrees both first bid price and second price bid, second one is the subject agrees first bid price but disagrees second bid price, third one is the subject disagrees first bid price but agrees second bid price, and fourth one is the subject disagrees neither first bid price nor second bid price. These are denoted as [YY], [YN], [NY], [NN] respectively. At this time, the probability of obtaining each answer is as follows.

$$ \begin{array}{*{20}c} {{ \Pr }\left[ {\text{YY}} \right] = 1- {\text{G}}\left( {\text{TU}} \right) = 1- {\text{G}}\left( {\text{tU}} \right)} \\ {{ \Pr }\left[ {\text{YN}} \right] = {\text{GU}}\left( {\text{TU}} \right) - {\text{G}}\left( {\text{Tf}} \right) = {\text{GU}}\left( {\text{tU}} \right) - {\text{G}}\left( {\text{tf}} \right)} \\ {{ \Pr }\left[ {\text{NY}} \right] = {\text{G}}\left( {\text{Tf}} \right) - {\text{G}}\left( {\text{Ts}} \right) = {\text{GU}}\left( {\text{tf}} \right) - {\text{G}}\left( {\text{ts}} \right)} \\ {{ \Pr }\left[ {\text{NN}} \right] = {\text{G}}\left( {\text{Ts}} \right) = {\text{G}}\left( {\text{ts}} \right)} \\ \end{array} $$

Where f is the initial presentation amount, s is the second presentation amount, and G (T) and G (t) are distribution functions when the presented amount is T or t yen. Herein, we assume the following logit type distribution function.

$$ \begin{aligned} {\text{Logit G}}\left( {\text{T}} \right) = 1/( 1+ { \exp }(\upbeta_{0} + \beta_{\text{T}} \,\ln \,\text{T + }\sum\upbeta_{\text{k}} \text{x}_{\text{k}} )) \hfill \\ {\text{Logit}}\,{\text{ G}}\left( {\text{t}} \right) = 1/( 1+ { \exp }(\upbeta_{0} + \beta_{\text{t}} \,\ln \,\text{t} + \sum\upbeta_{\text{k}} \text{x}_{\text{k}} )) \hfill \\ \end{aligned} $$

We estimate the parameters with the maximum likelihood method so as to maximize these log-likelihood functions. It is known that the log-likelihood function becomes nonlinear. For this reason, the BHHH estimator is used for updating algorithms necessary for parameter estimation (For details of the derivation process, see Tsuge [4]).

The median value of willingness to pay is when the probability of answering Yes to the second presentation is 0.5, and the average value is obtained by integrating the probability of answering Yes to the second presentation amount. In this paper, since we are making maximum adjustments, we have integrated from 0 to the maximum offering amount (For details, see Kuriyama et al. [1]).

4.1 Estimation Results Concerning Annual Payment

From Fig. 8 and Table 4, the intention to pay an annual contribution for bus operation revealed a median value of 1,042 yen; and it was 1,134 yen on average when the maximum bid amount was set.

Fig. 8.
figure 8

Willingness to pay annual payments

Full size image
Table 4. Estimation results for annual payments
Full size table

4.2 Estimation Results Concerning Bus Fares

From Fig. 9 and Table 5, the median willingness to pay bus fare was 226; and it was 209 yen on average when the maximum offering amount was set.

Fig. 9.
figure 9

Willingness to pay bus fare

Full size image
Table 5. Estimation results for bus fares
Full size table

4.3 Price Elasticity of Bus Fares

From the results in Table 2, nearly half of the respondents are opposed to paying annual contributions; this is justified in terms of bus users should pay the cost. Also, assuming the fare is an option from Table 3, even if a fare higher than the initial offering amount is favored, agreement is higher than opposition. This could reflect willingness to cooperate to support others.

Therefore, in this section, we examine how the number of passengers change when the fare changes, by invoking the price elasticity of demand. If the number of passengers does not change even if the fare is raised somewhat, the operation of the community bus can be said to be sustainable if the initial cost and fixed cost for bus operation are collected by freight. Conversely, if the bus fare is set high and the number of passengers is drastically reduced, it is necessary to examine the collection of the initial cost and fixed cost in another way.

The equation for calculating the price elasticity (ε d) of bus demand is as follows. Where p is the fare after change, q is the number of passengers after the change, Δ p is the price difference, and Δq is the variation in the number of passengers.

$$ \upvarepsilon_{\text{d}} = - ({\Delta p/p})/({q/\Delta q}) $$

Calculating by applying the values in Table 3, when the fare falls from 200 yen to 150 yen, the price elasticity becomes 2.07; as the fare for this bus service decreases, it is expected that demand substantively increases. By contrast, when the fare increases from 200 yen to 250 yen, the price elasticity becomes 0.22; as such, demand only decreases marginally.

Therefore, according to these results, it is expected that demand will increase significantly as the fare falls. If the fare rises, however, it can be posited that demand will not decrease so much.

5 Conclusion and Remarks

The result of the cellphone survey obtains the subjects do not have experience using electronic wallet and QR code functionalities. Recently, Felica Network undertook a similar survey in Tokyo based on a sample of residents 15 years or older between May 28–June 8, 2016. Therein, about 20% of the subjects used phone based electronic wallet functionalities. They asked questions pertaining to the usage of electronic money; women, adolescents and men over 60 years old were more likely to use physical, plastic debit/cards. In other words, it is thought that the people who are more likely to use community bus services are more likely to use plastic debit/credit cards.

From the result of the bus operation survey, we can estimate the median of the willingness to pay for annual payment is lower than first bid price. On one hand the median of the willingness to pay for bus fare is higher than first bid price. From this, we infer that subjects prefer the fare which includes whole cost to annual payment.

However, when all the initial expense is financed by the fare, it needs more number of passengers or higher bus fare.

From the price elasticity of bus demand, it can be said that the bus demand does not decrease so much if the bus fare is 50yen higher than first bid price, however, the bus demand increases much more if the bus fare is 50yen lower than first bid price.

Therefore, in order to realize the bus operation, it is supposed to suggest that the operator hands an IC chip card to the person who paid money of annual fee and apply the discount bus fare, or the system that user can pay financial reserve funds on the IC chip card little by little to prepare for future use.

We do not differentiate estimates of bus service payment intention amounts by age categories herein. We would try to estimate the amount of payment intention according to the generation. And we would like to consider the way of the expense burden that a bus service becomes sustainable by using a classification model.

We have collated GPS data from certain survey district residents; we hope to use the data in the future to further explore bus service demand.