Portal de Periódicos da CAPES

Linking Spatial Economics and Sequencing Economics for the Osaka Tourism Agglomeration

2021; Elsevier BV; Volume: 14; Issue: 3 Linguagem: Inglês

10.1111/rsp3.12476

1757-7802

Akifumi Kuchiki,

Regional Economic and Spatial Analysis

An agglomeration is an organization composed of its segments such as infrastructure and institutions. Sequencing economics discusses the sequential process analysis of building the segments of an agglomeration. The concept of 'economies of sequence' can be defined as the selection of any two segments from among the set of segments of an agglomeration and the sequencing of the segments toward the efficient building of an agglomeration. A central place theory in spatial economics gives the initial conditions of sequencing the segments of the tourism industry agglomeration as the economies of sequence. Using Proposition 1 in Henkel et al. (2000) and analyzing the corollary of the proposition additionally, we interpret our results obtained from Granger causality testing. The corollary is that a symmetric equilibrium is more likely to become unstable and break when substitutability between differentiated goods and transport costs becomes sufficiently low. The first priority in the order of the sequencing of segments in a tourism industry agglomeration is thereby given to those of low transport costs and substitutability between differentiated goods. We find that the first priority segment is Universal Studio Japan to lower the degree of substitutability between the differentiated goods and that the second is Kansai International Airport to reduce transport costs. Una aglomeración es una organización compuesta por sus segmentos, como la infraestructura y las instituciones. La economía secuencial aborda el análisis del proceso secuencial de desarrollo de los segmentos de una aglomeración. El concepto de 'economías de secuencia' puede definirse como la selección de dos segmentos cualesquiera de entre el conjunto de segmentos de una aglomeración y la secuencia de esos segmentos en el desarrollo eficiente de una aglomeración. La teoría del lugar central en la economía espacial marca las pautas iniciales de la secuenciación de los segmentos de la aglomeración de la industria turística como economías de secuencia. Mediante la utilización de la Proposición 1 de Henkel et al. (2000) y el análisis adicional del corolario de la proposición, se interpretaron los resultados obtenidos mediante la prueba de causalidad de Granger. El corolario dice que es más probable que un equilibrio simétrico se vuelva inestable y se rompa cuando la sustituibilidad entre los bienes diferenciados y los costos de transporte es lo suficientemente baja. Por tanto, la primera prioridad en el orden de la secuencia de los segmentos en una aglomeración de la industria turística se otorga a aquellos con un bajo costo de transporte y sustituibilidad entre bienes diferenciados. Se encontró que el segmento con la primera prioridad es que Universal Studios Japan reduzca el grado de sustituibilidad entre los bienes diferenciados y que el segundo es que el Aeropuerto Internacional de Kansai reduzca los costos de transporte. 集積とは、インフラや機関などで構成される組織体である。Sequencing economics (シーケンシング経済学)では、集積のセグメント構築の順序のプロセスの分析法を議論している。economies of sequenceの概念は、集積の一連のセグメントから任意のセグメントを2つ選択し、効率的な集積の構築に向けてその順序付けすることと定義することができる。空間経済学の中心の理論は、観光産業集積のセグメントをeconomies of sequenceとして順序付けする初期条件を与えるものである。Henkelら(2000)のProposition 1を用いてその推論をさらに解析し、Granger因果関係検定から得られた結果を解釈する。当然の結果として、対称均衡は、差別化された財の代替可能性と輸送コストが十分に低くなると、不安定になり崩壊しやすくなる。すなわち、観光産業集積におけるセグメントの第一の優先順位は、輸送コストと差別化された財の代替可能性が低いセグメントに与えられる。我々は、ユニバーサル・スタジオ・ジャパンが差別化された財の代替可能性を低下させる第一の優先セグメントであり、関西国際空港が輸送コストを抑制する第二の優先セグメントであることを見出した。 A tourism industry agglomeration provides a base upon which the activation of a regional economy can be achieved. According to the United Nations (2010), tourism can be viewed as a social, cultural, and economic phenomenon related to the movement of people to destinations outside their usual place of residence, usually motivated by pleasure. Agglomeration can be defined as the spatial concentration of economic activity. Thus, policy-makers are keen to contribute to developing a successful agglomeration policy as one means of activating the economy of a region. Many regions around the globe are currently engaged in the construction of agglomerations, but these are yet to be completed (see Ministry of Economy, Trade and Industry, Japan, 2020). Krugman (1991) developed a spatial version of the Dixit–Stiglitz model (1977) with which to examine where economic activity takes place and why. Fujita et al. (1999) provide an explanation of the prototype model of Krugman (1991) in Chapter 4 of their work, and review the established spatial economics, including central place theory, in Chapter 3. Based on a Hotelling-type framework in central place theory, Henkel et al. (2000) determined the inequality conditions of an agglomeration equilibrium related to spatial allocation decisions when all consumers are active in a marketplace. Herein, we reinterpret the model for the emergence of retailing markets from the perspective of a tourism industry agglomeration. An agglomeration is an organization composed of a set of segments that facilitate physical infrastructure, including ports and highways, institution building, human resource development, and satisfaction with living conditions, cultural aspects, and so on (see Appendix Table). Kanai and Ishida (2000) coined the term 'process analysis' to refer to research on the dynamic process of agglomeration building. Naturally, the construction of an agglomeration cannot be completed in an instant, but requires the sequential construction of segments. For example, it has taken more than 10 years to construct manufacturing agglomerations (see Kuchiki, 2020a). The sequence of segment construction is crucial to the completion of an agglomeration, and any failure in the efficient sequencing of them, or diseconomies of sequence, leads to an interruption in the process. Thus, the construction of an agglomeration relies on a sequential process of segment construction within an agglomeration according to economies of sequence, and this paper sets out to determine these economies of sequence for the tourism industry. Sequencing economics in relation to architecture theory for agglomeration construction is applied to sequencing the segments in terms of 'economies of sequence.' The concept of 'economies of sequence' is defined as the selection and sequencing of any two segments from among the set of segments that make up an industrial agglomeration toward the efficient building of that agglomeration, as defined in Kuchiki (2021). We define the 'economies of sequence' based on Granger causality testing. The Granger causality test allows us to examine whether the facts regarding the 'economies of sequence' are significant. For the Granger causality relationship between segment x and segment y, a null hypothesis that segment x does not Granger-cause segment y is rejected. This paper focuses on the relationship between spatial economics and sequencing economics. Fujita and Kuchiki (2006) analyzed agglomerations from both a sequencing and spatial economics perspective. Based on an examination of case studies in Malaysia, the United States, and China, Kuchiki (2020b) found that the number of foreign tourists Granger-causes revenues in the tourism industry. This fact means that the segments related to reductions in transport costs, such as the construction or expansion of airports, in the transportation infrastructure segment increased tourist numbers, resulting in an increase in revenues across the tourism industry. However, sequencing economics on tourism industry agglomerations is yet to be analyzed from the perspective of central place theory in spatial economics in order to identify the initial conditions for sequencing economics. Calero and Turner (2020) comprehensively reviewed studies of economic geography and tourism, and Dibeh et al. (2020) investigated the tourism-led growth nexus based on a new economic geography (NEG) model in spatial economics. However, Calero and Turner (2020) concluded that 'regional tourism research remains in its infancy.' The flowchart approach described in Kuchiki and Tsuji (2008) and Kuchiki (2006) is a form of process analysis for sequencing economics that does not take into account spatial economics. The relationships between (i) spatial economics and (ii) sequencing economics are demonstrated in Figure 1. In terms of a tourist industry agglomeration in Japan, the number of passengers at JR Kyoto Station was found to Granger-cause the number of tourists to Kyoto City. This fact means that segments related to the reduction of transport costs in the Tourism column in Figure 1 led to an increase in tourist numbers, according to the spatial economics of Henkel et al. (2000) (see Kuchiki, 2020b). With regard to a manufacturing agglomeration, the first priority is given to the construction of traffic infrastructure, including ports and access roads, to reduce transport costs in the manufacturing column in Figure 1, according to the spatial economics of Krugman (1991) (see Kuchiki, 2020a). Further, regarding an information communication technology (ICT) industry agglomeration, the initial condition obtained by Fujita and Thisse (2003) was the development and invitation of researchers, or labor mobility, as shown in the ICT column in Figure 1 (see Kuchiki, 2021). Spatial economics in relation to a railway-led agglomeration in Japan revealed that the first priority must be related to the reduction in transport costs, as obtained in Fujita and Krugman (1995) and Gokan (2016). A case demonstrating the economies of sequence is the opening of Tsukuba Railway and the development of residential towns shown in the Railway-led growth column in Figure 1 (see Kuchiki et al., 2017 and Kuchiki, 2019). However, no studies to date have applied propositions obtained from spatial economics to obtain initial conditions as 'economies of sequence' for the sequencing the segments of an agglomeration. This paper directly applies the central place theory model of Henkel et al. (2000) in spatial economics to obtaining the initial conditions for tourism agglomeration policy in sequencing economics. This paper, therefore, aims to present the facts of the 'economies of sequence' for agglomerations in Osaka Prefecture, Japan, by applying the proposition obtained from the model of Henkel et al. (2000) to the Granger causality test of segment sequencing for the tourism agglomerations. The proposition is that, when the degree of substitutability between differentiated goods as well as the transport costs are lower, symmetric equilibrium is more likely to become unstable. Figure 2 shows the workflow of the paper, summarizing the results obtained from Tables 1, 2, and 3. By lowering the degree of substitutability between the differentiated goods, the opening of Universal Studio Japan (USJ) in 2000 increased the number of foreign passengers arriving at Kansai International Airport (KIA) within 2 years, as presented in Table 2. A structural change in the trend regarding the number of foreign passengers at Kansai International Airport then occurred in 2003, as presented in Table 3. An increase in the number of foreign passengers simultaneously led to increases in both the number of flights at KIA and the foreign tourists to Osaka Prefecture, as presented in Table 1. This paper demonstrates that the opening of USJ to lower the degree of substitutability between differentiated goods is the first priority in the sequencing of segments in a tourism agglomeration as an initial condition, and that the second priority is the development of Kansai International Airport to reduce transport costs. The following presents an outline of Osaka Prefecture, Japan. The gross regional product (GRP) of Osaka Prefecture is ranked third, at 30.0 trillion yen, after those of the Tokyo Metropolitan area and Aichi Prefecture, which are ranked first and second at 103 and 39.5 trillion yen, respectively. The main economic blocks in Japan consist of the two regions of Kanto, the area surrounding Greater Tokyo, and Kansai, the area surrounding Osaka Prefecture. Kuchiki et al. (2017) found that the Japanese model Ministry of Economy, Trade and Industry, Japan of a regional growth, as represented by the Hankyu Railway established in 1907 in Kansai area, was diffused to allow the development of Japan from the Kansai area to the Kanto area. The commercial center of Japan was located in the Kansai area, and the first international exposition in Asia was held in Osaka in 1970. Greater Tokyo in the Kanto area developed railway networks by utilizing private and public companies, resulting in centralization in Greater Tokyo and slowly widening the economic disparity between the Tokyo and Osaka areas. The Osaka Convention and Tourism Bureau was opened in 2003 with the aim of promoting the tourism industry in Osaka Prefecture, and was later reorganized as the Osaka Tourism Bureau in 2015 for the development of the tourism industry across the entire Kansai area, which includes Osaka Prefecture and Kyoto Prefecture. The tourism strategy developed by Osaka Prefecture in 2005 prioritized measures in terms of shopping, restaurants, and cultural entertainment at night through the selection of priority areas, such as Osaka Castle, and the creation of a city known for its international entertainment. As the economy in the Osaka area failed to maintain its upward trend, the policy of Osaka Prefecture changed its focus to develop the tourism industry through policy measures such as the opening of KIA in 1994. However, the airport was unsuccessful in contributing to the tourism industry agglomeration in the Osaka area, partly because the user costs at the airport were high and direct access from the airport to the Osaka city center was relatively expensive. Information related to the segments including infrastructure is as follows (see Appendix Table). The population in Osaka Prefecture and Greater Tokyo on January 12, 2017 was 8.83 and 13.51 million, respectively. With regard to the respective airports, the number of passengers at KIA in 2019 was 24.8 million, while the combined number at Narita International Airport and Tokyo International Airport in the Kanto Area was 53.2 million. The length of railway track in operation in the Osaka area in 2010 was 1,504 km, while that in the Tokyo area was 2,459 km, with the number of the passengers in the Osaka area in 2010 being 4,647 million, and that in Tokyo being 14,329 million. With regard to ports, the quantity of goods handled at Osaka Port and Kobe Port in the Kansai area in 2019 was 5.328 million twenty-foot equivalent units (TEU), which ranks 27th in the world, while that at Tokyo Port, Yokohama Port, and Kawasaki Port in the Kanto area was 8.161 million TEU, which ranks 19th in the world. In the Kansai area, the Genroku Culture prospered among the townspeople around 1688–1704 in the Edo Period of Japan. Osaka is popular in culture terms. Segments related to cultural factors in Osaka Prefecture are octopus dumplings as food, Kawachi folk songs as music, Osaka Castle as history, Senshu toweling as a textile, Sakai City cutlery as handiwork, Kyoto and Kobe as resorts, and the local sake, Akishika, as an alcoholic beverage (see Appendix Table). KIA developed to become a hub airport for low-cost carriers (LCCs) in 2012. To make KIA more of a base for tourists, Osaka Prefecture began operating the second terminal exclusively for LCCs on October 28, 2012. The number of LCC flights per week subsequently increased from 43 in the summer of 2011 to 236 in the summer of 2015. 11 New Kansai International Airport Company. 2020. http://www.nkiac.co.jp/news/2014/2144/2015summerschedule.pdf. (Accessed on 16 August 2020). Three companies started running airport shuttle buses from the airport to Osaka Station on November 6, 2012. A railway company also discounted its passenger tickets from the airport to Namba in the center of Osaka. 22 Kansai Airport Research Institute. 2020. http://www.kar.or.jp/wpcms/wp-content/uploads/2012/01/review1212.pdf. (Accessed on 16 August 2020). Further, USJ introduced the Wizarding World of Harry Potter in 2014. 33 Kadokawa Corporation. 2020. https://www.walkerplus.com/trend/matome/article/168696/. (Accessed on 16 August 2020). The paper is structured as follows: Section 2 provides a definition of the concept of 'economies of sequence' and derives a proposition regarding 'Universal Studio Japan' and 'low-cost carriers' as the initial conditions for sequencing the segments of a tourism agglomeration. Section 3 looks at Granger causality testing to identify the facts of 'economies of sequence' and linear regression and the dummy-variable regression analyses. Finally, the last section offers a summary and conclusions, integrating the results discussed in Sections 2 and 3. The economies of sequence can be mathematically defined as follows. Suppose that there are three periods: the first, second, and third. Let us examine two examples of segment formation sequencing in an agglomeration. Policy measures, for which there are two candidate sequences (A and B), form a segment. Suppose that an agglomeration consists of three segments {s1, s2, s3}, say {international airport, Universal Studio Japan, and the quality enhancement of cultural products} in the tourist industry model. The difference between A and B lies in the ordering of s2 and s3. In A, the sequence of segment formation is assumed to be the sequence s1, s2, and s3; thus, it is assumed that the sequence of policy measures to form a production function in B is the sequence s1, s3, and s2. Accordingly, A and B can be notated as follows: A = {s1, s2, s3} = {international airport, Universal Studio Japan, and the quality enhancement of cultural products}, and: B = {s1, s3, s2} = {international airport, the quality enhancement of cultural products, and Universal Studio Japan}. The productivity for sequence A and sequence B can be given as YA and YB, respectively. The productivity after the implementation of the successful policy measures for {s1, s2, s3} is YA = f {s1, s2, s3}). Now, the productivity of A and B in the third period can be compared. Suppose that there exist economies of sequence between s2 and s3; that is, YB is very small and nearly zero because of the diseconomies of sequence between s3 and s2, while YA is large. The sequence of the segment formation from s3 to s2 is inefficient in comparison with that of the segment formation from s2 to s3. Accordingly, there exist 'economies of sequence' between s2 and s3 in the case such that YA < YB in the case that {s1, s3, s2} is more efficient than {s1, s2, s3}. We derived a proposition for the interpretation of results obtained from Granger causality testing, based on the model of Henkel et al. (2000) regarding the economies of scale. The proposition obtained implies that the first priority in the order of the sequencing of segments in a tourism industry agglomeration should be given to the low degree of substitutability between differentiated goods as s1 and low transport costs as s2. The Granger causality test allows us to examine whether the facts regarding the 'economies of sequence' are significant. Analyses in a time series assume stationary stochastic processes, while drift is an intercept component. On the other hand, ordinary least square estimations with a drift term (constant term) are nonstationary stochastic processes. It is generally accepted that equations without a drift term be used for stationary stochastic processes. Thus, our models adopt Model 1 and Model 2 without a drift term once it was confirmed that equations with a drift term provided results for nonstationary stochastic processes. H0: all c2i = 0, i = 1, 2,…, n. The null hypothesis holds if H1: at least one ci ≠ 0, i = 1, 2,…, n. Then x Granger-causes y. In other words, in the case that the variance of Model 2 is equal to that of Model 1, x does not Granger-cause y. In the case that the variance of Model 2 is greater than that of Model 1, x Granger-causes y. We apply F-test to find the facts of 'economies of sequence.' Here we use the same model of Henkel et al. (2000) in central place theory to explain the tourism industry agglomeration, or the selection of two tourist regions by specialized one product tourism firms. The model is a partial equilibrium model to focus on the tourism industry. We suppose that shops, consumers, and firms in the model of Henkel et al. (2000) are tourism spots, tourists, and tourism firms in this section, respectively. We use Proposition 1 in Henkel et al. (2000) and consider the symmetric fragmentation equilibrium in which the same number of tourists visit at both ends of the line, or the two locations. The proposition states that there exists a unique equilibrium involving an identical number of tourists at the two locations. We apply the corollary to interpreting the results of the Granger causality testing in the next section. The development strategy of each region is to implement policy measures to satisfy the inequality conditions. The two branches of the Dixit and Stiglitz (1977) framework for a monopolistic competition model and the Hotelling (1929) framework are adopted, with firms restricted to locating at the line's end point. This supposes that many tourism firms produce a variety of differentiated goods and services, such as entertainment, food, shopping, etc., in line with 'the key factors of the cultural aspects' of living conditions (see cultural aspects in Appendix Table). Upon patronizing one of these points, called shops, or tourism spots, at a cost purely dependent on distance, tourists buy the utility-maximizing varieties of tourist goods and services. Let us consider the case in which a tourist visiting to buy a specific differentiated product or service can freely choose its location each time. Corollary of 'Proposition 2 in Henkel et al. (2000)': The lower the degree of substitutability between differentiated goods, and the lower the transport costs, the more likely it is for the symmetric equilibrium to become unstable. The corollary gives an interpretation for the results obtained by the empirical analysis: first, Universal Studio Japan in Osaka and its new program for the Wizarding World of Harry Potter made the degree of substitutability between differentiated goods, σ, smaller than the one associated with the agglomeration threshold; second, low-cost carriers (LCC) and new schedules of Brett trains are to make transport costs, t, lower. This section shows the results for 'economies of sequence' from Table 1 to Table 4 and explains an interception dummy-variable method, in which the criterion year is important. Regarding Universal Studio Japan of the differentiated goods and services and Kansai International Airport of transportation infrastructure, Tables 1–4 obtain the cases of 'economies of sequence,' as summarized in column 1 of Figure 2. The Granger causality test and dummy-variable method are applied to the relationships between the number of foreign tourists to Osaka Prefecture and the number of foreign passengers at Kansai International Airport. Universal Studio Japan in Osaka opened in March 2001, LCC flights started to operate in October 2012, and the Tokaido Shinkansen Nozomi (bullet train) schedules were revised significantly in March 2012. The results of the regression analyses presented in Table 1 reveal that the positive relationship between the number of foreign passengers and the number of foreign flights at Kansai International Airport is significant, with a p-value of 9.4 × 10−11 since the coefficient of the flights is positive, and the positive relationship between the number of foreign tourists visiting Osaka Prefecture and the number of foreign passengers at Kansai International Airport is significant, with a p-value of 9.7 × 10−9 since the coefficient of foreign passengers is positive. To summarize Table 1, the positive relationships between the number of foreign flights, the number of foreign passengers at Kansai International Airport, and the number of foreign tourists visiting Osaka Prefecture are significant, as shown in column 3 of Figure 2. The results allow us to conclude that any increase in the number of foreign flights induces an increase in the number of foreign passengers and, subsequently, an increase in the number of tourists visiting Osaka. As shown in column 1 of Figure 2, the results of the Granger causality test in Table 2 prove the economies of sequence of the number of visitors to Universal Studio Japan, the number of foreign passengers at Kansai International Airport, and the number of foreign tourists. The paper finds that the two following cases related to the economies of sequence are statistically significant. First, the null hypothesis of Granger causality is that the number of visitors at Universal Studio Japan does not Granger-cause the number of foreign passengers at Kansai International Airport, with a p-value of 1.324 × 10−7 considered statistically significant at n = 2. However, secondly, the null hypothesis of Granger causality is that the number of foreign passengers at Kansai International Airport does not Granger-cause the number of foreign tourists, with a p-value of 0.02015 considered statistically significant at n = 3. Table 2 presents the Granger causality tests of the data of the first order difference to examine autocorrelation ergodic. First, the number of visitors at Universal Studio Japan does not Granger-cause the number of foreign passengers at Kansai International Airport, with a p-value of 2.98 × 10−5 considered statistically significant at n = 2. Second, the null hypothesis of Granger causality is that the number of foreign passengers at Kansai International Airport does not Granger-cause the number of foreign tourists, with a p-value of 0.1977 considered not statistically significant at n = 3. Regarding Kansai International Airport, Table 3 presents the results for the dummy-variable method with the p-values for the number of foreign passengers being significant in cases in which the starting years were 2003 and 2015. First, the p-values for the number of foreign passengers at Kansai International Airport are significant in cases in which the starting year was 2003 and 2015. The null hypothesis is that the dummy variable of 2003 is not significant, with a p-value of 0.007468 considered statistically significant. Also, the null hypothesis is that the dummy variable of 2015 is not significant, with the p-value of 2.72 × 10−7 considered statistically significant. Second, the p-values of the number of foreign flights at Kansai International Airport are significant in cases in which the starting year was 2002 and 2015. The null hypothesis is that the dummy variable of 2002 is not significant, with a p-value of 0.0106 considered statistically significant. The null hypothesis is that the dummy variable of 2015 is also not significant, with a p-value of 6.52 × 10−9 considered statistically significant. Accordingly, column 2 of Figure 2 supports the results of Granger causality tests of Table 2. Regarding Universal Studio Japan, Table 4 presents structural changes in the number of visitors that occurred in 2007 and 2015 (p-values of 0.01194 and 0.00129, respectively), as shown in column 2 of Figure 2. We can see structural changes in the role of Universal Studio Japan around 2003 and 2015. The following two facts support the above result on structural changes: the GRP of Osaka Prefecture occurred in 2008 and 2015 (p-values of 0.00026 and 0.0567, respectively); the Tokaido bullet trains in 2008 and 2015 (p-values of 0.00034 and 0.00916, respectively). Provided that the autocorrelation ergodicity of tests is satisfied, the results of the Granger causality tests presented in Table 2 reveal that the number of visitors to USJ Granger-causes the number of the foreign passengers on international airlines arriving at KIA, and that the number of the foreign passengers arriving at KIA Granger-causes the number of foreign tourists visiting Osaka Prefecture. We tested the Granger causality on the more stationary data for each variable of the first order difference to avoid autocorrelation ergodicity, revealing that the number of visitors at USJ Granger-causes the number of foreign passengers arriving at KIA. Although we cannot reconfirm that the number of foreign passengers arriving at KIA Granger-causes the number of foreign tourists visiting Osaka Prefecture, Table 1 shows that the number of passengers arriving at KIA is positively correlated with the number of foreign tourists visiting Osaka Prefecture, indicating that number of visitors at USJ Granger-causes both the number of passengers arriving at KIA and the number of foreign tourists visiting to Osaka Prefecture. Figure 2 clearly demonstrates that the conditions of an agglomeration equilibrium derived from the model of Henkel et al. (2000) hold when the differentiated goods and services of USJ play a role in lowering the degree of substitutability between the goods. We find that an increase in the number of visitors to Universal Studio Japan Granger-causes an increase in the number of foreign passengers arriving at Kansai International Airport and, subsequently, an increase in the number of foreign flights arriving at Kansai International Airport Granger-causes an increase in the number of foreign passengers, simultaneously resulting in an increase in the number of foreign tourists. The results of linear regression and dummy-variable regression analyses also support the notion that the facts of the 'economies of sequence' are necessary conditions for an agglomeration. Hence, the economies of sequence for the agglomeration discussed herein show the sequence of airport construction followed by the provision of the differentiated goods, such as USJ, to be an initial condition. We analyzed tourism agglomeration segment construction from three perspectives: a case study, a spatial economics model, and Granger causality testing. We examined the tourism industry agglomeration in Osaka Prefecture, Japan, taking into consideration central place theory in the spatial economy to apply economies of sequence to sequencing the segments of the tourism industry agglomeration. The model of Henkel et al. (2000) was used to derive the initial conditions for segment sequencing of the tourism industry agglomeration. We identified the 'economies of sequence' in the tourism industry agglomeration using Granger causality testing with the intention of establishing sequencing economics for a tourism agglomeration, allowing its practical application to a regional growth strategy. The results revealed that, as the initial conditions, the value of the degree of substitutability between the differentiated goods and services in the tourism industry need to be reduced below the threshold value at which an agglomeration equilibrium, with all tourists concentrated in one region, exists. Granger causality testing and regression analyses revealed that an increase in the number of visitors to USJ Granger-causes an increase in the number of foreign passengers visiting Kansai International Airport and, subsequently, an increase in the number of foreign passengers visiting Kansai International Airport causes an increase in the number of foreign flights, resulting in a further increase in the number of foreign tourists. The policy implications of the above are as follows: if we suppose that a region plans to develop a tourism agglomeration, the region must satisfy the conditions derived as the initial conditions for that region through central place theory in spatial economics. Thereafter, the sequencing of the segments sees those related to lowering the degree of substitutability between the differentiated goods and the reduction of transport costs given priority in spatial economics. Hence, the policy measures are as follows: first, the degree of substitutability between the goods and services in the tourism industry needs to be lowered by producing differentiated goods and services, such as the introduction of USJ and the opening of a new attraction, such as the Wizarding World of Harry Potter; second, transport costs need to be reduced by opening KIA and introducing the new LCC flight schedules. To date, no studies have analyzed the sequencing process of segment construction while taking account of the substitutability between goods by linking propositions derived from spatial economics with sequencing economics in a tourism industry agglomeration. However, three main issues remain for future study. First, the number of case studies should be increased to further confirm the identified facts of the 'economies of sequence' of the Osaka tourism agglomeration, not only for the tourism industry but also for other industries such as the ICT industry. Second, segments related to innovations in Porter's diamond model (1990) remain to be analyzed. Third, environmental issues, such as overtourism resulting from an agglomeration, need to be sequenced with reference to sustainable development goals (SDGs), particularly after the coronavirus disease 2019 (COVID-19) pandemic. This work was supported by the Japan Society for the Promotion of Science, grant/award number 17H04549. I wish to thank to Masahisa Fujita, Seiichi Fukui, Hideyoshi Sakai, and three anonymous referees for comments on the drafts of this paper. This appendix represents the same model of Henkel et al. (2000) to interpret the results of the Granger causality testing of the Osaka tourism agglomeration in section 3. y ≡ Y/t and β ≡ 1/(σ − 1). In the symmetric fragmentation equilibrium, setting z* = 1/2, n(2) ≡ βMt [y − (1/4)]/{2F[(β + 1)]} = n0* = n1* firms locate in either tourist spot. We seek to derive the conditions under which the symmetric equilibrium breaks and becomes unstable. First, we suppose that the symmetric equilibrium of two regions is stable. Second, we obtain the sustainability condition for the symmetric equilibrium. Third, we derive the instability conditions under which the symmetric equilibrium breaks, using the sustainability conditions. From the above, a region gives the first priority in the order of the sequencing of segments in a tourism industry agglomeration to those segments that satisfy the instability condition of the symmetric equilibrium. The above inequality means as follows: The symmetric equilibrium does not hold. A larger relative number of tourism industry firms a (= (n0 − n1)/2) makes region 0 more profitable and more attractive. Then, larger π0 makes n0 larger. The processes of the concentration of all firms in a single location start. That is, β > βb implies that there exists an agglomeration equilibrium with all tourists concentrated at one of the line's ends.