Repercussion effects of consumption by domestic tourists in Tokyo and Kyoto estimated using a regional waste input–output approach

The importance of tourism around the world is increasing. In 2014, direct, indirect, and induced economic effects of tourism amounted to 9% of global GDP (United Nations World Tourism Organization (UNWTO), Tourism Highlights 2015). Moreover, the contribution of newly developing countries in the tourism market to this figure increased from 30 to 45% over the last 25 years, which means that tourism has been effective in improving economic well-being in these countries (UNWTO 2015). Hardy et al. (2002) reported that the three key factors in the development of sustainable tourism are environment vision, economic development, and community involvement. More recent studies have begun to examine the environmental loads induced by tourism (Ko 2005; Kytzia et al. 2011; Cernat and Gourdon 2012), although most early tourism studies evaluated only the economic benefits associated with tourism (Archer and Fletcher 1996; Fletcher 1989; Dwyer et al. 2000, 2004; Cai et al. 2006).

Investigating the economic and environmental effects of tourism is important, particularly in certain regions, because the priority associated with reducing environmental loads differs between regions. Economic repercussion effects are not usually limited to the target tourist site but also bring about economic benefits to other areas. However, an increase in tourism is also accompanied by an increase in waste, which is inevitably treated within the target tourist site (Lenzen 2008). Most Japanese tourist sites need to treat increased waste due to tourism because each municipality in Japan must basically treat general waste within the municipality. Moreover, it is extremely difficult to find new landfill sites in Japanese metropolitan cities, and one of the top priorities in these cities is to reduce landfill waste (Nakamura and Kondo 2002b; Sasao 2004; Tsukui et al. 2015). As such, waste generation and landfilling induced by tourists are crucial issues for sustainable tourism.

In a tourism study that considered the environment, Cernat and Gourdon (2012) developed the sustainable tourism-benchmarking tool (STBT), which they used to assess the sustainability of tourism using quantitative indicators initially proposed by Ko (2005) and Fernandez and Rivero (2009). Cernat and Gourdon (2012) conducted an input–output analysis to clarify economic linkages and leakages, which they used to investigate a variety of indicators, such as natural and cultural assets, economic activities, infrastructure, and environmental loads, including GHG emissions. Sun (2014) evaluated GHG emissions in Taiwan using the production and consumption accounting principles applied to the tourism satellite account (TSA) and environmentally extended input–output (EEIO) models. However, these authors did not quantitatively clarify the repercussion effects associated with waste generation or landfilling due to tourism. Other regional-level tourist studies have also examined the effect of tourism on waste generation. Johnson and Moore (1993) evaluated the impact of residents’ and visitors’ whitewater recreation activities on the Klamath River in Oregon, USA. Frechtling and Horvath (1999) estimated the direct, indirect, and induced impacts of tourism consumption on production and value added around Washington DC using the Regional Input–Output Modeling System (RIMSII). Specifically, they estimated the multiplier effects associated with tourist expenditure on the local regional economy. However, neither of these studies investigated the environmental loads induced by tourism.

Our study examines the impact of tourist consumption on tourist attractions in urban areas within the context of waste treatment using a regional waste input–output (WIO) approach. As mentioned previously, the scarcity of landfill sites in large cities is one of the most serious environmental problems affecting these areas. We therefore selected two popular metropolitan tourist destinations in Japan, Tokyo and Kyoto, as target regions for our case studies. We considered the effects of domestic tourism in these target regions, focusing on tourists from other parts of Japan. In the case of Tokyo, we used the interregional WIO (IR-WIO) table for Tokyo compiled in 2000 by Tsukui et al. (2015) to analyze the economic and environmental effects of domestic tourists from other regions. We clarified the tourism consumption in Kyoto and Tokyo as direct effects and estimated the resulting indirect effects in both regions. In our study, we define the combined direct and indirect effects as the repercussion effects. The WIO analysis (WIOA) developed by Nakamura and Kondo (2002a) is one of the most effective approaches for evaluating the interdependence between the flow of goods and waste. Case studies employing the WIOA have been conducted in a variety of countries and regions, including Japan (Nakamura and Kondo 2009), Australia (Reynolds et al. 2014; Fry et al. 2015), Taiwan (Liao et al. 2015), France (Beylot et al. 2016), and the UK (Salemdeeb et al. 2016). However, none of these studies evaluated the repercussion effects of tourism. For the Kyoto case study, we used the regional WIO table for Kyoto compiled in 2000 by Ichikawa et al. (2011). We then compared estimates of the direct and indirect effects of domestic tourism in these two regions on production value, value added, waste generation, waste transportation, landfilling, and GHG emissions.

The remainder of the present paper is structured as follows. Section 2 describes the regional WIO models used in the analysis. Section 3 presents the expenditure results estimated using a non-survey method for domestic tourists in Tokyo and Kyoto. Section 4 presents a comparison between the two regions with estimates of the repercussion effects for each region. Finally, conclusions and issues for future consideration are presented in Sect. 5.

In our study, we compared the repercussion effects for Tokyo and Kyoto in terms of domestic tourist expenditure using an IR-WIO approach. As mentioned in Sect. 1, the advantage of using an IR-WIO analysis to analyze the repercussion effects of consumption in a specific region is that economic and environmental effects, such as waste generation, transportation, and landfilling, can be considered simultaneously.

In our study, we applied the same two-region IR-WIO model that we applied in Tsukui et al. (2015). Briefly, the model is based on Kagawa and Kondo (2007) and Kagawa et al. (2007), which is an expansion of the conventional regional IO model of Isard (1951) and the WIO model of Nakamura and Kondo (2002a). A detailed description of this model has been provided in Tsukui et al. (2015). We estimated the repercussion effects of visitors’ consumption on industrial sectors, waste generation, waste treatment sectors, and the environmental loads of each sector in Tokyo and Kyoto, respectively. Environmental loads, such as GHG emissions, can be estimated by multiplying generation coefficient matrices.

For the case study of Tokyo, we used an IR-WIO table for Tokyo for the year 2000, which we compiled in a previous study (Tsukui and Nakamura 2009; Tsukui et al. 2010, 2011a, b, c,  2015) and which was based on “The Interregional Input–Output Table in the Year 2000 in Tokyo” published by Tokyo Metropolitan General Affairs (TMGA 2007). In the Kyoto case study, we used a regional WIO table for the year 2000 in Kyoto (KWIOT2000), which was compiled by Ichikawa et al. (2011). To compile this table, Ichikawa et al. used the “Year 2000 Kyoto Prefecture Input–Output Table,” published by Kyoto Prefecture (2004). We also referred to the industrial waste data published by the Ministry of the Environment Secretariat, Waste/Recycling Measures Department (2002) and NEDO (2005) for the estimation of waste generation from industrial sectors in both Tokyo and Kyoto. In order to apply the IR-WIO model described in Tsukui et al. (2015) to our study, we extended the KWIOT2000 to produce a two-region IR-WIO table (KIRWIOT2000) using Nakamura (2010). In the WIOTs for both Tokyo and Kyoto, estimating the activities of waste treatment sectors was based on Matsuto (2005). Estimates of the GHG emissions for each sector were based on the Embodied Energy and Emission Intensity Data for Japan Using Input–Output Tables (3EID) database (Nansai et al. 2008).

Since we had already developed a multi-regional WIO table for 47 prefectures of Japan (47WIOT) (Tsukui et al. 2014), we herein compare the advantages of each table before adopting a two-region IO framework in the case study. Using the 47WIOT, we can confirm the consistency of the data sets for investigating both Tokyo and Kyoto. However, the 47WIOT has only 46 industrial sectors and the two important sectors for investigating tourism, “Hotels and other lodging places” and “Eating and drinking places,” are treated as belonging to the same sector, “Other personal services.” Upon careful consideration of the advantages and disadvantages of the 47WIOT and WIO tables for Tokyo and Kyoto, we decided to adopt the WIO tables for use in our study.

Before presenting the results for the repercussion effects associated with tourist consumption in Tokyo and Kyoto, in this section, we clarify the differences in tourist expenditure between Tokyo and Kyoto.

We defined “tourists” as visitors from regions in Japan outside the target region who stayed overnight. While the primary objective of overnight visitors is tourism and shopping, the main purpose of day visitors (i.e., shopping or tourism) is more difficult to clarify.

For the analysis of Tokyo, we used the tourist expenditure data specified in the interregional input–output table (IR-IOT) for Tokyo 2000 (TMGA 2007), which we used to estimate the goods and services sector of the IR-WIO table for Tokyo in 2000 (Tsukui et al. 2015).

However, in the case of Kyoto, there was no category in the regional IO table for Kyoto 2000 showing consumption by visitors from other regions staying overnight, which was the basis for the WIO table for Kyoto compiled in our study, even in the most detailed version of the table, which included 104 sectors. We considered that the expenditure of visitors was incorporated into the “Household consumption expenditure” of the final consumption sector in the regional IOT of Kyoto. Thus, tourist consumption in Kyoto by residents of other regions was estimated by non-survey-based methods, as described in Sect. 3.2.

In this section, we compare the repercussion effects in Tokyo and Kyoto of the domestic tourist expenditures described in Sect. 3. As shown above, consumption by domestic tourists staying overnight in Kyoto Prefecture is less than half that for Tokyo. We considered the disparity in magnitude of the economic activities of the two prefectures when we investigated and compared the repercussion effects induced by tourist consumption in Tokyo and Kyoto.

First, the impacts were compared from an economic perspective. The repercussion effects on the production value attributed to consumption by domestic tourists were approximately 434 billion yen in Tokyo and 260 billion yen in Kyoto. Considering that the direct effects of tourism in Tokyo were twice as large as those in Kyoto, the repercussion effects on the production value in both regions were relatively similar. The value-added repercussion effects amounted to approximately 276 billion yen in Tokyo and approximately 108 billion yen in Kyoto.

The trends in repercussion effects on production and value-added effects were similar between Tokyo and Kyoto. The repercussion effects on civil services and the service industry were 62.7% in Tokyo and 68.0% in Kyoto, on traffic, telecom, and broadcasting were 17.2% in Tokyo and 13.0% in Kyoto, on commerce were 10.8% in Tokyo and 8.0% in Kyoto, and on secondary manufacturing industries were 6.3% in Tokyo and 8.5% in Kyoto. However, the impacts on certain industries differed between Tokyo and Kyoto (Table 4). The repercussion effects of the production value on “Hotels and other lodging places,” “Eating and drinking places,” and “Amusement and recreational services” accounted for more than 50% in both regions. The repercussion effect on “Hotels and other lodging places” was larger in Kyoto than in Tokyo. Moreover, the repercussion effect on “Amusement and recreational services” in Kyoto was half that in Tokyo. As mentioned in Sect. 3.2, regarding the direct effects of tourists’ consumption of goods and services, the repercussion effect on “Amusement and recreational services” in Kyoto is smaller than that in Tokyo because, unlike in Tokyo, most of the tourists visiting Kyoto go to historical attractions, such as shrines and temples. The repercussion effect on “Financial and insurance services” was also larger in Tokyo than in Kyoto. Since most insurance companies have their headquarters in Tokyo, the repercussion effects of travel insurance are larger in Tokyo than in Kyoto. In Kyoto, the repercussion effect on “Petroleum refinery products” accounted for 4634 million yen, which is considerably larger than that in Tokyo. This disparity in repercussion effects associated with petroleum products exists because there is no petroleum industry in Tokyo. Indeed, all petroleum products are imported to Tokyo from other parts of Japan. A production value of 8032 million yen for petroleum products was induced in areas outside Tokyo. As shown in Fig. 2, the difference in the repercussion effects on the production value is caused by differences in the economic structures of Tokyo and Kyoto. Whereas the main industry in Tokyo is the service industry, in Kyoto, the main industries are manufacturing industries, including the automotive industry, the food industry, industries producing semiconductor devices and integrated circuits, and the publishing and printing industries. The difference in the economic structures of Tokyo and Kyoto also affects the amounts and types of waste generated.

Table 5 shows the repercussion effects of waste generation (considering recycling) in Tokyo and Kyoto. Although the direct and indirect effects on the production value in Kyoto are half that in Tokyo, the induced waste generation for Kyoto is as same as that for Tokyo. Business waste is generated by the service industries, such as “Hotels and other lodging places” and “Eating and drinking places,” and waste generation by these industries is larger in Kyoto than in Tokyo. As shown in Fig. 2, the difference in industrial waste generation is mainly attributable to differences in the economic structure of Tokyo and Kyoto. The main industries in Kyoto are manufacturing industries, which have higher waste generation coefficients than the tertiary industries that dominate Tokyo. Kyoto is also famous for its livestock industry, and “Livestock excreta and livestock corpses” accounts for a total of 3345 tonnes of industrial waste, which is considerably larger than that generated in Tokyo.

As indicated by the results shown in Table 5, the waste generation induced in Kyoto by “Waste oil,” “Waste acid,” “Waste alkali,” and “Waste plastics” accounted for −1616, 22, 15, and −2778 t, respectively (total: −4356 t). The reason why some repercussion effects are negative is that the amount of waste that is recycled is larger than that generated in these waste categories. For example, in “Road transport” consumption by tourists induced production activity in the petroleum industry that recycled 1655 t of waste oil and 3140 t of waste plastics. In KIRWIOT2000, “Waste oil” and “Waste plastics” are generated by other chemical industries, but these industries are not stimulated by tourist consumption. The relative amount of “Cinder, dust, and molten slag” from Kyoto, which accounted for 2680 t, is also much larger than that from Tokyo. In investigating which industrial sectors generate “Cinders, dust, molten slag” waste, the repercussion effect of “Dust” from the “Electricity” industry in Kyoto is 2286 t. Kyoto Prefecture has a thermal power plant at Maizuru, which uses primarily coal and crude oil as fuel (KEPCO 2011). In contrast, in 2000, most thermal power plants in Tokyo were off-line or used natural gas, which produces only small amounts of “Dust.” Therefore, the quantity of “Dust” produced in Tokyo and Kyoto appears to differ markedly between the two prefectures. Since there has been a change in the usage of power plants and fuels since the Great East Japan Earthquake in March 2011, the results may differ if the analysis is performed using more recent data.

The large amounts of waste generated led to a large amount of landfill waste. Regarding the repercussion effect of landfill consumption area, Kyoto consumed 1479 m² of landfill compared to 7369 m² for Tokyo. This appears to contradict the results indicating that waste generation, which is induced by visitors’ consumption in Kyoto, was the same as that in Tokyo. The reason for the observed disparity in landfill consumption area is because the waste categories associated with waste generation in Kyoto, which was larger than that in Tokyo, were mainly treated in incineration facilities. Since incineration effectively reduces the amount of residue that needs to be landfilled, Kyoto consumed less landfill than Tokyo. For example, in Kyoto, 15% of the induced waste generation belonged to the “Livestock excreta, Livestock corpses” category, which was incinerated. In contrast, in Tokyo, the induced waste generation of “Scrap metal,” considering recycling, accounted for approximately 1359 tonnes, which is 3.5 times larger than that in Kyoto. Scrap metal is difficult to reduce in size and is directly landfilled (Fig. 3).

For the repercussion effects of GHG emissions, a noticeable difference was observed between Tokyo and Kyoto. The repercussion effect of GHG emissions in Tokyo was 269,945 t-CO₂ (eq.), whereas that in Kyoto was 141,103 t-CO₂ (eq.). However, as shown in Fig. 4, the proportion of GHG emissions by the industries of Tokyo and Kyoto was quite different. As stated in Sect. 3.2, the direct effect of transportation and traffic-related activities is high for Tokyo, and “Road transport” and “Air transport,” which have high GHG emission coefficients, also have high relative importance in Tokyo.

Our study analyzed the direct and indirect effects of consumption by domestic tourists using a regional WIO analysis, which enabled us to investigate the economic and environmental impacts of final consumption induced by tourism. As indicators of environmental loads, we focused on waste generation, landfilling, and GHG emissions. We conducted case studies in Tokyo and Kyoto and compared the direct and indirect effects induced by visitors from other parts of Japan.

The findings revealed that the economic impacts of tourism in Tokyo and Kyoto were generally similar, but the environmental impacts of domestic tourism were quite different. Although the magnitude of the direct and indirect effects was much larger in Tokyo, tourism was more important in Kyoto than in Tokyo. Similarly, although the amount of waste induced in both regions was similar, the amount of landfill generated in Tokyo was approximately five times larger than that in Kyoto. The GHG emissions in both regions were approximately proportional to the direct effects of domestic visitors’ consumption, but the industries that emitted GHGs were quite different in each area, reflecting differences in the economic structures of Tokyo and Kyoto. The results of our study also demonstrated the need for considering the regional characteristics of the target region when promoting tourism in a way that is environmentally sound.

Future research should examine the importance of the interregional relationships between different areas of a country. When a site attracts tourists, it sometimes induces large environmental loads in other parts of the country. For example, economic activities that are stimulated by tourism, such as shopping, eating in restaurants, and lodging, which generate relatively small environmental loads, tend to stimulate production activities in industries such as agriculture, the food industry, and manufacturing, which have larger environmental loads. Hasegawa et al. (2015) constructed a multi-regional IO table for 47 prefectures in Japan (47IOT) in order to evaluate the carbon footprint of each region. Expanding the 47IOT to include environmental loads other than GHG emissions would allow the direct and indirect effects of tourism to be clarified within the context of the interregional economic and environmental relationships between regions. It is also important to consider the effect of the Great East Japan Earthquake in 2011, particularly insofar as the shift in power generation methods in Japan is concerned. The model and data obtained in our study provide a powerful tool for investigating the economic and environmental impacts of tourism.