Keyword

1 Purpose

Many city administrations and managements of shopping streets often utilize various tourist attractions and events to revitalize towns. However, as pointed out by Saito [2], the net number of incoming visitors to the town, which is obviously necessary and indispensable information when viewing the town as one entity for attaining the goal of its revitalization, has hardly been grasped. If the net number of incoming visitors to the town is unknown, it is difficult to judge whether tourist attractions and events have really been effective for revitalizing the town because we cannot accurately evaluate how these attractions and events have contributed to the increase in the net number of incoming visitors to the town.

Furthermore, it can be said that there are no studies that directly address the question such as how many people participated, how they walked around the town, and to what extent and on which part of the town the effect of the tourist attractions and events was spread.

The aim of this study is to address this issue directly. More specifically, on the day of the event, we carry out an on-site survey of Kaiyu behavior, in which a question item is devised to distinguish the visitors who came with the purpose to attend the event from the ordinary visitors. Then we apply the consistent estimation method of Kaiyu patterns developed by Saito et al. [3,4,5] and propose a method to measure how many people participated in the event, how those people shopped around the town, and to what part of the town the effect of the event was spread. Additionally, we also attempt to show the effectiveness of our method with a concrete application example.

While many policies advocating revitalizing towns have been implemented so far, it is surprising that most towns do not know the net number of incoming visitors to their own towns. Some subsidized policies such as vitalization of city center are required to present post-evaluation reports after their implementations, but the situation that policy makers for towns have not known their net number of incoming visitors to their own towns has yet to be improved.

Most of the large retailers such as the department stores and shopping malls have grasped the number of incoming customers to their shops daily. The information is indispensable for their marketing and promoting their shops. The number of incoming customers is usually measured by some people counting system, which automatically counts the number of customers entering and exiting their building in a time interval such as 15 min. As for towns, since it is almost impossible to encircle the whole area of a town, the net number of incoming visitors has not been known. Some areas to implement city center revitalization program employ pedestrian transport volume surveys as a policy evaluation tool, where the sum of the pedestrians counted is used for the substitute for accounting for the number of visitors to their city center.

However, a pedestrian counted at some pedestrian survey point might be counted at another different survey point so that we cannot obtain the net number of incoming visitors to the city center because of this double counting possibility.

Recently, Saito et al. [3] and Saito and Nakashima [4] have developed a unique theoretical method to estimate the net number of incoming visitors to a city center district, which is based on micro-behavioral Kaiyu data obtained from the on-site survey of consumer shop-around behaviors. They call their method a consistent estimation method of consumer shop-around patterns for on-site survey data. While Saito et al. [6] also tried to estimate the number of incoming visitors to the city center by using the weighted Poisson model, this consistent estimation method has quite a different theoretical background from their previous efforts.

The aim of this study is to apply this method to the event to estimate the net number of incoming visitors attracted by the event and to evaluate the effect of the event which spreads over other areas than the event venue through tracing Kaiyu behaviors of the event participants, which has never been addressed to be analyzed in details.

Thus the purpose of this study is to take up the case of the Kumamoto Castle Festival held at the city center district of Kumamoto City on May 31, 2009, conduct an on-site survey of shop-around behaviors on the day of the event, estimate the net number of incoming visitors to the event as well as to the city center district of Kumamoto City, and evaluate to what extent the Kumamoto Castle Festival has influenced on which part of the city center district of Kumamoto CityFootnote 1 through the shop-around behaviors of the event participants.Footnote 2

The significant feature of this study is that to estimate the effects of the event we make a comparison between the visitors who visited the city center with the purpose attending the festival and those visitors with other purposes. The respondents are classified into these two groups by their visit purposes obtained from the questionnaire. This is a sharp contrast to previous studies in which the comparison is based on the difference in respondents’ behaviors such as use or non-use of one-dollar bus and city center café [8, 9].

2 Framework of the Analysis

2.1 Outline of the Kumamoto Castle Festival

The Kumamoto Castle Festival was an event held in accordance with the opening of the Kumamoto Castle Honmaru palace to the public on April 20, 2008. It was held during the period of the local expo “Festival of the 400 years since the construction of the Kumamoto Castle,” held from December 2006 to May 2008 to commemorate the 400th year of the construction of the Kumamoto Castle in 2007.

The data used in this study was collected from the 2nd survey of shop-around behaviors at the city center of Kumamoto City, which was conducted on May 31, 2009 (Sunday). This survey date was the day when the second Kumamoto Castle Festival was held to celebrate the achievement that Kumamoto Castle became the castle with the highest number of visitors in Japan.

In addition, at the second Kumamoto Castle Festival, trams between Kumamotojo-Shiyakusho-Mae Station and Suidocho Station were stopped from 10:30 to 16:00, the entry of cars and motorbikes was prohibited, and the area of Torichosuji became a pedestrian paradise. The trams in the section other than the pedestrian paradise area were operating halfway, and the fare was free all day. Within the pedestrian paradise, a main stage and substages were installed, where various events were held. The opening time of the Kumamoto Castle Festival was from 11:30 to 15:00.

2.2 Data Used

In this study, we used data from the 2nd survey of consumer shop-around behaviors at the city center of Kumamoto City, which was jointly conducted on May 31, 2009 (Sunday), by the Fukuoka University Institute of Quantitative Behavioral Informatics for City and Space Economy (FQBIC) and the Downtown Atelier of the Faculty of Engineering of Kumamoto University. The 2nd survey of consumer shop-around behaviors at the city center of Kumamoto City was carried out as an on-site interview survey with questionnaire sheets, which takes about 15 min. We set up several survey points. The respondents are sampled at random from the visitors at the sampling points. The respondents are asked about their shop-around behaviors at the city center district of Kumamoto City; the frequency of visits to the city center district of Kumamoto City, the Kamitori Arcade, and the Shimotori Arcade; the transport means; etc. The survey was carried out for 6 h between 12:00 and 18:00, with survey points located at the Kamitori Arcade, the Shimotori Arcade, and the Shinshigai Arcade. A total of 122 samples were collected. The outline of the survey is shown in Table 15.1.

Table 15.1 Outline of the 2nd survey of consumer shop-around behaviors at the city center of Kumamoto City
Full size table

2.3 Method for the Analysis

In this survey, to measure the effect of the Kumamoto Castle Festival, we asked the respondents whether or not they came to the city center district of Kumamoto City with the purpose to attend the festival. Figure 15.1 shows the results of the questionnaire item on whether or not the respondents came to the city center district of Kumamoto City with the festival purpose. We see that 21.4% of the people visited the city center of Kumamoto City with the purpose to attend the Kumamoto Castle Festival.

Fig. 15.1
figure 1

Visitors with the festival purpose and with non-festival purposes

Full size image

In this study, we measure the effect of the Kumamoto Castle Festival by the following steps:

  1. 1.

    To estimate the net number of incoming visitors to the city center district of Kumamoto City on the same day of the Kumamoto Castle Festival, first we apply the consistent estimation method of consumer shop-around patterns for on-site survey data to the on-site consumer shop-around data obtained from the 2nd survey of consumer shop-around behaviors at the city center of Kumamoto City and get the consistent estimate of the density of consumer shop-around patterns at the city center district of Kumamoto City, which is defined as the probability distribution over all the observed shop-around routes (paths or trip chains). It should be noted that since we have obtained the probability distribution over all shop-around routes, we can also estimate the density of the origin-destination table by the standard aggregation procedure. Thus we can also obtain the density of the total number of visitors to every commercial site or zone division within the city center district.

  2. 2.

    Next we use the actual number of visitors to a department store on the survey day, which was kindly provided by the department store, divide it by the estimated density of the total number of visitors to that department store obtained in the previous step (1), and get the estimate of the net number of incoming visitors into the city center district of Kumamoto City. Since this is the method of expansion from the density to the actual number, it should also be noticed that the estimates of actual numbers of visitors to various commercial sites or divided zones within the city center district of Kumamoto City can be obtained at the same time.

  3. 3.

    We assume that the percentage of visitors with the festival purpose among the total number of visitors to the city center is equal to the percentage obtained from the shop-around survey data, that is, 21.4%. With this assumption we estimate the respective total numbers of visitors with festival purpose and with non-festival purposes.Footnote 3

  4. 4.

    From the shop-around survey data, we calculate the visit ratios to zones, i.e., the choice probabilities to visit zones, respectively, by samples with the festival purpose and by those with non-festival purposes. From multiplying these respective visit ratios to zones by the total numbers of visitors with the festival purpose and with non-festival purposes estimated in the step (3), we obtain the estimates of the total numbers of visitors to each zone with the festival purpose and with non-festival purposes.

  5. 5.

    Furthermore, in order to measure the effect on the amount of expenditure, we use the following method. We assume that all the amount of expenditure spent in the city center district by the visitors who came to the city center district of Kumamoto City with the festival purpose is included in the effect of the Kumamoto Castle Festival on the expenditure. We estimate this effect by each zone from multiplying the total number of visitors to this zone with the festival purpose estimated in the step (4) by the average amount of expenditure per visitor to this zone obtained from the shop-around survey data. On the other hand, for those visitors who had not come with the festival purpose, we assume that their expenditure in the pedestrian paradise, the venue of the Kumamoto Castle Festival, would not have been spent if it were not for the Kumamoto Castle Festival. Thus we regard this as a part of the effect of the Kumamoto Castle Festival on expenditure and estimate it from multiplying the number of visitors to the venue with non-festival purposes by the amount of expenditure per visitor in the pedestrian paradise. By summing up these two, we measure the effect of the Kumamoto Castle Festival on the amount of expenditure at the city center district of Kumamoto City.

Here we would like to point out two points. First, it should be noted that the effect of this study is limited to the city center district of Kumamoto City, as it does not measure the effect on other areas and it differs from the concept of the net effect on society as a whole. Second, in this study we employed steps (3) and (4), but if we had much more survey samples, we could have applied the consistent estimation method to the shop-around data separately for the samples with the festival purpose and those with non-festival purposes. This is left for a future task.

2.4 Consistent Estimation Method

In this study, we apply the consistent estimation method of consumer shop-around pattern for on-site survey data to the on-site data obtained from the 2nd survey of consumer shop-around behavior at the city center of Kumamoto City. The consistent estimation method was developed by Saito et al. [3]. The method makes it possible to consistently estimate the density of shop-around patterns from the data obtained by the on-site sampled shop-around data while removing the choice-based sampling bias that inherently occurs in the on-site sampled data. The density of shop-around patterns is the probability distribution over all shop-around routes or paths. In other words, it is the distribution over all types of trip chains. The most innovative consequence of the method is that it gives a theoretical foundation to estimate the net number of incoming visitors to the area such as the city center district or the downtown. The reason is that since the method gives us the accurate probability over all shop-around routes, it also gives the density of the amount of visitors to a particular commercial facility when the density of the net number of incoming visitors equals 1. Usually the large retailers such as department stores are equipped with a people counting system, which automatically counts the actual number of people entering and exiting their buildings. Thus the consistent estimation method provides us with a theoretical method to estimate the net number of incoming visitors to the city center if we had the actual number of visitors to at least one commercial facility. In fact, the net number of incoming visitors to the city center can be obtained by using the usual expansion method which divides the actual number of visitors to the facility by the density of its visitors.

In this study, we apply the consistent estimation method to the data obtained from the 2nd survey of consumer shop-around behaviors at the city center of Kumamoto City to estimate the density of shop-around patterns within the city center. We also have the data of the actual number of visitors to one commercial facility. With these data, following the above procedure, we can obtain the net number of incoming visitors to the city center district of Kumamoto City.

The consistent estimation method is formalized as the following formula. This formula shows the algorithm of how to calculate the consistent estimate for the distribution over shop-around routes, fc(r) from the on-site shop-around behavior history data obtained where the subscript c indicates the distribution f is defined at on-site.Footnote 4

$$ {\widehat{f}}_c(r)\propto \frac{\sum \limits_{v>0}\sum \limits_{s\in S}\sum \limits_{t\in T}{\delta}_t^c\left(r,v,s\right)\frac{1}{l_S\left({r}_t\right)}\frac{{\tilde{f}}_c\left({s}_t|{v}_t\right){\tilde{\tilde{f}}}_c\left({v}_t\right)}{H\left({s}_t\right){\tilde{\tilde{f}}}_c\left({v}_t|{s}_t\right)}}{\sum \limits_{r\in R}\sum \limits_{v>0}\sum \limits_{s\in S}\sum \limits_{t\in T}{\delta}_t^c\left(r,v,s\right)\frac{1}{l_S\left({r}_t\right)}\frac{{\tilde{f}}_c\left({s}_t|{v}_t\right){\tilde{\tilde{f}}}_c\left({v}_t\right)}{H\left({s}_t\right){\tilde{\tilde{f}}}_c\left({v}_t|{s}_t\right)}} $$

Here, fc(r) denotes the distribution over shop-around routes r∈R, r denotes a shop-around route, v denotes the visit frequency, s∈S denotes the survey point, t∈T is the sample, and ls(r) denotes the length of the shop-around route r over the set of survey points S. \( {\delta}_t^c\left(r,v,s\right) \) is the random variable that assumes 1 when sample t takes the route r and the visit frequency v, and is sampled at the survey point s, and 0 otherwise. H(s) is the sampling ratio of each sampling point s.

3 The Effect of the Kumamoto Castle Festival from the Viewpoint of Kaiyu Behaviors

3.1 Estimating the Net Total Number of Incoming Visitors to the Whole City Center District of Kumamoto City

Here, we estimate the net number of incoming visitors to the city center of Kumamoto City by applying the consistent estimation method of shop-around patterns to the on-site shop-around data obtained from the 2nd survey of consumer shop-around behavior at the Kumamoto city center.

Specifically, first we estimate the density of shop-around patterns in the Kumamoto city center by applying the consistent estimation method of shop-around patterns. The Tsuruya Department Store, which is the long-established local department store, kindly provided the data of the actual number of its visitors. Based on the estimated result of probability distribution over shop-around routes and the actual number of visitors at one commercial facility, we can estimate the net number of incoming visitors to the whole city center district of Kumamoto City.

According to the estimated result of the consistent estimation,Footnote 5 the density of the amount of total visitors to the Tsuruya Department Store turns out to be 0.153 when the total number of visitors to the city center of Kumamoto City including net incoming visitors and shop-around visitors is set to 1. Also, the number of visitors to Tsuruya, according to an interview with people involved with Tsuruya, was 26,250 (in this study, we use the data on the number of visitors to Tsuruya on September 28, 2008, the previous year when the first Kumamoto Castle Festival was held).

Furthermore, the average number of steps in the shop-around, that is, the average number of all shops visitors visited while they shopped around the city center, which is obtained from the data collected in the 2nd survey of consumer shop-around behaviors at the city center of Kumamoto City, turns out to be 2.78.

Based on the above, the net number of incoming visitors to the whole city center district of Kumamoto City is calculated as follows.Footnote 6

The net number of incoming visitors to the whole city center district = (the actual number of visitors to Tsuruya Department Store ÷ the density of the amount of visitors to Tsuruya) ÷ the average number of all shops visitors visited.

It becomes as follows:

$$ \mathrm{61,675}\ \left(\mathrm{people}\ \mathrm{per}\ \mathrm{day}\right)=\left(\mathrm{26,250}\ \left(\mathrm{people}\ \mathrm{per}\ \mathrm{day}\right)\div 0.153\right)\div 2.78 $$

Also, the following holds.

The total number of visitors to the city center district of Kumamoto City (the net number of incoming visitors + the number of visitors visited by their shop-around) = the net number of incoming visitors × the average number of all shops visitors visited.

It becomes as follows:

$$ \mathrm{171,456}\ \left(\mathrm{people}\ \mathrm{per}\ \mathrm{day}\right)=\mathrm{61,675}\ \left(\mathrm{people}\ \mathrm{per}\ \mathrm{day}\right)\times 2.78 $$

3.2 Comparing Visit Ratios to Zones by Visit Purposes of Visitors

Figure 15.2 is a map of the city center district of Kumamoto City. The area depicted in the map is used as the area defined as the city center district for this study. Based on the zone divisions shown in the figure, we look at the visit ratios to these zones in the city center district by visitors with the festival purpose and by those with non-festival purposes from the data collected by the 2nd survey of consumer shop-around behaviors. Table 15.2 shows the result of these visit ratios to zones in the city center district for visitors by visit purposes of visitors, that is, by the festival purpose and non-festival purposes. Looking at the top five zones, it can be seen that when comparing visitors who came for the festival purpose with those who did not, the former spreads along the pedestrian paradise and arcades (Zones 3, 5, 8, 10, and 28). In contrast, the latter were concentrated on the pedestrian paradise and around the Tsuruya Department Store (Zones 5, 6, and 7).

Fig. 15.2
figure 2

Map of the city center district of Kumamoto City

Full size image
Table 15.2 Visit ratios to zones for visitors with the festival purpose and with non-festival purposes
Full size table

3.3 Comparing the Number of Visitors to Zones by Visit Purposes of Visitors

Figure 15.3 shows the result of the number of visitors to zones in the city center district of Kumamoto City by visit purposes of visitors. This was calculated by multiplying the total numbers of visitors of respective visit purposes by the corresponding visit ratios to zones within the city center. From the figure, we can compare the differences in shop-around behaviors by visit purposes on a real number basis.

Fig. 15.3
figure 3

Comparison of Kaiyu visits at zones between visitors with festival purposes and with non-festival purposes

Full size image

Among visitors who came for the festival purpose, 6900 people stopped by the pedestrian paradise zone, which was the highest number, followed by Tsuruya (Zone 5) with 4185 people. Also, Ginza Street (Zone 28) had high numbers.

On the other hand, among visitors who came for other purposes, 21,850 people stopped by the Tsuruya Department Store which was the most visited, followed by the pedestrian paradise zone with 14,845 people, and PARCO (Zone 6) with 13,579 people. In addition, a lot of visitors stopped by the surroundings of Shinshigai Arcade (Zone 12).

4 Measuring the Effect of the Kumamoto Castle Festival on Expenditure

4.1 Average Expenditure per Visit for Each Zone

In this section we estimate the effect of the Kumamoto Castle Festival from the viewpoint of expenditure. Table 15.3 shows the result of calculating the average expenditure per visit for each zone within the city center district of Kumamoto City based on the data obtained from the 2nd survey of consumer shop-around behaviors. Note that the notion of the average expenditure per visit is different from the concept of customer-based expenditure. For example, suppose that one customer visited shopping sites, A and B, located in the same zone X and spent 2000 yen and 3000 yen, respectively, at A and B. The total expenditure for this customer is 5000 yen. But the average expenditure per visit is 2500 yen per visit since this customer drops two visits to zone X. Here for the calculation of the average expenditure per visit, the visit with no expenditure is included.

Table 15.3 Average expenditure by zone per step
Full size table

In the next section, using the average expenditure per visit for each zone, we estimate the effect of the Kumamoto Castle Festival on expenditure.

4.2 The Effect of the Kumamoto Castle Festival on Expenditure

First, we measure the effect of the Kumamoto Castle Festival from the viewpoint of the amount of money spent by visitors who came for the festival purpose. Table 15.4 shows the result of the measurement. The effect of the Kumamoto Castle Festival on expenditure contributed by visitors with the festival purpose is the sum of the amount of money visitors with the festival purpose spent within the Kumamoto city center. The expenditures in each zone are calculated by multiplying the number of visitors with the festival purpose to that zone by the average expenditure per visit obtained in the previous section. As shown in the table, the total of the expenditure visitors with the festival purpose spent in the city center district of Kumamoto City turns out to be 51,320,000 yen.

Table 15.4 Expenditure by visitors with festival purpose spent at the city center district of Kumamoto City
Full size table

On the other hand, the amount of money visitors with non-festival purposes spent at the pedestrian paradise, the venue of the Kumamoto Castle Festival, which should be thought of as a part of the monetary effect by the festival according to the above discussion, was calculated in the following way.

The amount of money spent by visitors with non-festival purposes at the festival venue = the number of visitors with non-festival purposes who visited to the pedestrian paradise × the average expenditure per visit at the pedestrian paradise.

Therefore, the effect contributed by visitors with non-festival purposes turns out as follows:

$$ \mathrm{1,350,000}\ \mathrm{yen}=\mathrm{14,845}\ \mathrm{visit}\mathrm{s}\times 91\ \mathrm{yen}\ \mathrm{per}\ \mathrm{visit} $$

From the above, the total effect of the Kumamoto Castle Festival on expenditure turns out to be 52,670,000 yen (or 526,700 dollars).

$$ \mathrm{52,670,000}\ \mathrm{yen}=\mathrm{51,320,000}\ \mathrm{yen}+\mathrm{1,350,000}\ \mathrm{yen} $$

5 Conclusion and Future Challenges

The most significant contribution of this study is to provide a methodology of how we theoretically estimate the net number of incoming visitors attracted by some event and how we measure the effect of the event which is possibly spread over other areas than that of the event venue, both of which are based on the micro-behavioral data obtained from the on-site survey of consumer shop-around behaviors. Thus we have presented a coherent methodology to evaluate the effects of attractions and events often adopted for revitalizing towns from the viewpoint of consumer Kaiyu behaviors with a typical example of the Kumamoto Castle Festival, in which the effects spread over other spots within the city center are shown to be tracked from the Kaiyu behaviors of the event participants both on the number of visitors basis and on monetary basis.

While this study might be said as a pilot study because of its small size of the samples, we see that visitors who came for the festival purpose mainly moved around the arcades near the festival venue and the effect of the Kumamoto Castle Festival on expenditure is roughly estimated as 52,670,000 yen.

Our future work would be for anything to find another event opportunity where we can expand the number of samples and perform a separate consistent estimation for the Kaiyu behaviors of event participants. Another future research topic would be to compare the cost and effect of holding the event such as the Kumamoto Castle Festival.