- Email: [email protected]
Japan's tourism-led foreign direct investment inflows: An empirical study
ECMODE-03808; No of Pages 7 Economic Modelling xxx (2015) xxx–xxx
Contents lists available at ScienceDirect
Economic Modelling journal homepage: www.elsevier.com/locate/ecmod
Japan's tourism-led foreign direct investment inflows: An empirical study☆ Akinori Tomohara 1 University of California, Los Angeles, Anderson Forecast, CA 90095-1481, USA Department of International Economics, Aoyama Gakuin University, Shibuya-ku, Tokyo 150-8366, Japan
a r t i c l e
i n f o
Article history: Accepted 16 September 2015 Available online xxxx Keywords: Immigration International tourism FDI Local economy
a b s t r a c t This study analyzes international tourism–inward foreign direct investment (FDI) relationships by applying dynamic panel models to Japanese data for the 1996–2011 period. A growing number of studies have examined this topic following the publication of the UNCTAD (2007) report; however, the current literature is inadequate. Our analysis is not restricted to tourism-related FDI such as hotels and restaurants, as in the literature, but covers general FDI in the framework of FDI dynamics, because tourism–FDI interactions are not in reality restricted to typical tourism-related FDI. The analysis results reveal that enhanced inbound foreign tourism exhibits spillovers of inward FDI beyond tourism-related sectors. Hence, although policies related to FDI promotion and tourism enhancement are often planned, executed, and evaluated independently under different government jurisdictions, coordination is recommended, because these policies interact with each other in efforts to attain economic development goals. © 2015 Elsevier B.V. All rights reserved.
1. Introduction International organizations have begun promoting tourism–FDI relationships as a vehicle for development (United Nations Conference on Trade and Development (UNCTAD), 2007). Tourism facilitates not only export revenue generation but also service sector job creation and local economic revival through tourism-related FDI. Since the publication of the United Nations Conference on Trade and Development (UNCTAD) (2007) report, a growing number of studies have begun to analyze tourism–FDI relationships. The primary focus in the literature is the relationship between tourism (for which the number of arrivals and revenues are used as a proxy) and tourism-related FDI (i.e., FDI inflows into setting up tourism infrastructure such as accommodation, restaurants, and transportation in less-developed countries; Katircioglu,
☆ I appreciate Jin Ke for her research assistance and acknowledge the financial support from Aoyama Gakuin University. The funding source was not involved in the conduct of the research (study design, the collection, analysis, and interpretation of data) and/or preparation of the article (the writing of the report), and the decision to submit the article for publication. I would like to thank the editor and two anonymous referees for their helpful comments. The usual disclaimer applies. E-mail address: [email protected]. 1 Present address: Department of International Economics, Aoyama Gakuin University, Building No. 8-421, 4-4-25 Shibuya, Shibuya-ku, Tokyo 150-8366, Japan. Tel.: +81 3 3409 8562; fax: +81 3 5485 0782.
2011 for Turkey; Samimi et al., 2013 for 20 developing countries; Selvanathan et al., 2012 for India; Tang et al., 2007 for China). Enhanced tourism is expected to contribute to the development of an economy through increased FDI inflows into tourism-related industries. Thus, tourism-related FDI is assumed to be a primary channel for economic development. However, the existing research in this field is inadequate, because tourism–FDI interactions are not in reality restricted to typical tourism-related FDI such as in hotels, airlines, and restaurants. Tourism development may induce FDI in other sectors as well. For example, China's Suning Commerce Group Co. Ltd. formed a capital business alliance with Japan's Laox Appliance (a retail store for home appliances) and the latter became a subsidiary of the former in 2009.2 One motivation for this alliance is the business opportunity of serving the growing number of Chinese tourists to Japan by offering services such as language support, duty-free arrangements, and acceptance of Chinese credit cards. This is because rice cookers, air purifiers, and digital cameras manufactured in Japan are principal purchase items for Chinese tourists. The example illustrates the possibility that enhanced tourism may induce FDI in sectors beyond hotels, etc. Hence, this study analyzes international tourism–inward FDI relationships. Our analysis is not restricted to typical tourism-related FDI
2
www.laox.co.jp/company/outline/ (last accessed on July 23, 2015).
http://dx.doi.org/10.1016/j.econmod.2015.09.024 0264-9993/© 2015 Elsevier B.V. All rights reserved.
Please cite this article as: Tomohara, A., Japan's tourism-led foreign direct investment inflows: An empirical study, Econ. Model. (2015), http:// dx.doi.org/10.1016/j.econmod.2015.09.024
2
A. Tomohara / Economic Modelling xxx (2015) xxx–xxx
in the literature, but covers general FDI as well. Motivated by anecdotal evidence that tourism influences FDI in non-tourism sectors in Japan, we aim to fill the existing gap in the literature by estimating the effects of inbound tourism on inward FDI beyond tourism-related sectors. We set up our empirical model by referring to the analytical framework of FDI determination. Our model extends the framework to a dynamic context after introducing international tourism. In order to investigate these dynamics, the analysis uses a sample of 29 countries/ areas, each of which is a major FDI donor to Japan, over the period 1996–2011. Our analysis is distinct from other analyses in the literature, which examine the causality between tourism development and tourism-related FDI using time-series analyses including only these two variables. We focus on the spillover effects of tourism development on FDI, and thus, we need to control for other factors determining FDI. Such analysis requires the use of panel data on bilateral FDI and tourism between donor and recipient countries, which are not necessarily required for the time-series analyses presented in the literature. Bilateral FDI data, together with the number of foreign tourists and other control variables, are compiled for the analysis. Then, the tourism–FDI interactions are estimated by using a two-step system generalized method of moments (Arellano and Bover, 1995; Blundell and Bond, 1998; Roodman, 2009; Windmeijer, 2005). The analysis results reveal that an increased number of foreign tourists visiting Japan helps promote FDI into Japan, and that the effects of tourism on FDI in non-tourism industries are observed to be significant. The effects of tourism on total FDI are estimated to be 5.6 times as large as those on tourism-related FDI. The results of a favorable interaction between tourism and FDI imply that the two policies of inward FDI promotion and inbound foreign tourism enhancement are mutually inclusive. There exist positive spillovers from promoting tourism to encouraging FDI in various sectors. A favorable interaction between international tourism and inward FDI is confirmed under different specifications. The results obtained using a benchmark model are robust when we apply a different FDI modeling of the knowledge capital model (Carr et al., 2001, 2003; Markusen, 1997; Markusen et al., 1996) and a different empirical approach using the bias-corrected least-squares dummy variable (LSDVC) estimator of Bruno (2005a,b) to our context. The results of our analysis suggest that the effect of international tourism should have been extended to general FDI dynamics in previous works. The literature focusing on tourism-related FDI has the risk of underestimating the effects of tourism enhancement on FDI in reality. Our results have important policy implications that are not covered in the literature. Tourism enhancement helps attain economic growth in not only less-developed tourism-oriented countries but also others, because international tourism positively influences FDI beyond traditional tourism sectors. Additionally, policy coordination is recommended between FDI promotion and tourism enhancement, although the two policies are often planned, executed, and evaluated independently under different government jurisdictions. These policies interact with each other to enable the attainment of economic development goals. The remainder of this paper is organized as follows. Section 2 presents a description of data used for the analysis and introduces the model. Section 3 discusses the results of the analysis. Section 4 concludes the paper with future suggested lines of research. 2. Model and data Our analysis uses Japanese data to study tourism–FDI interactions. Japan is relevant for analysis because it allows investigating the interactive relationships between inward FDI promotion and inbound foreign tourism enhancement. The Japanese government has aggressively targeted both inward FDI and tourism as priority policy areas for
enhanced growth performance,3 and thus, from a policy perspective, it is critical to know whether there are significant spillover effects among policies. Specifically, the situation in Japan amplifies the importance of examining spillover effects of increased tourism beyond tourism-related industries. Such industries are not major recipients of inward FDI in Japan; rather, Japan receives most of its FDI in other sectors such as finance and insurance, communications, wholesale and retail, chemicals and pharmaceuticals, and electric machinery. The two policies of inward FDI promotion and tourism enhancement may not help in attaining economic development goals if tourism development affects FDI into tourism-related industries only. Moreover, the Japanese situation allows us to examine whether enhanced tourism affects inward FDI in other industry sectors, because such an examination is not feasible if other sectors are not mature enough to receive inward FDI. We use major FDI donor countries to Japan as our sample. The sample comprises 29 countries/areas from Asia (China, Hong Kong, Taiwan, South Korea, Singapore, Thailand, Indonesia, Malaysia, Philippines, India), North America (the United States, Canada), Central America (Mexico, Brazil), Pacific (Australia, New Zealand), Europe (Germany, United Kingdom, France, Netherlands, Italy, Belgium, Luxembourg, Switzerland, Sweden, Spain, Russia), and the Middle East (Saudi Arabia, United Arab Emirates). The Cayman Islands, a major FDI donor, is not included in our sample because the Japanese government does not provide any records of tourists from the Cayman Islands. Our empirical model extends the framework of FDI determination to a dynamic context after introducing international tourism and controlling for other factors determining FDI. Whereas previous works focused on the causality between tourism and tourism-related FDI (Katircioglu, 2011; Samimi et al., 2013; Selvanathan et al., 2012; Tang et al., 2007), our approach is expressed in the following conceptual framework. FDI is a function of Tour and control variables, X: FDI = f(Tour, X). Since FDI is composed of FDItourism − related industry + FDIother indutries, the aforementioned literature with missing effects of tourism on FDIother indutries might mislead the policy effectiveness of tourism enhancement. In order to examine the possibility, FDI dynamics are examined using a system generalized method of moments (GMM), in which current FDI relies on past FDI realizations and the number of foreign tourists. The system GMM accounts for simultaneous bias between FDI inflows and foreign tourist arrivals. FDI it ¼ α þ β1 FDI it−１ ＋β2 Tour it þ X it ρ þ ε it ; where ε it ¼ δi þ μ it
ð1Þ
FDIit represents the net inflow of FDI from country i to Japan at time t, Tourit denotes the number of foreign tourists visiting Japan, Xit is a set of control variables, and εit is an error term, which comprises δi, an 3 The Japanese government has been promoting international tourism ever since the “Basic Policies for Economic and Fiscal Management and Structural Reform” endorsed by the Cabinet in 2002 suggested increasing the number of foreign tourists visiting Japan. Correspondingly, the Ministry of Land, Infrastructure, Transport and Tourism (2003) proposed strategies to promote foreign tourists visiting Japan by providing a report entitled “Global tourism strategies”. The report recognizes international tourism as a promising sector because the size of international tourism is expanding globally according to the World Tourism Organization (UNWTO), and expects expanded international tourism to vitalize the Japanese economy, specifically the declining local economy. The Japanese government has also been trying to promote inward FDI for similar policy ends. During the 1990s, FDI promotion was proposed to revitalize the Japanese economy when Japan slipped into recession after the burst of the bubble economy. After the Economic and Fiscal Policy Management and Structural Reform in 2002, regional governments initiated inward FDI to revitalize the local economy and to acquire greater tax revenues. This is because the reform required regional governments to be less dependent on fiscal support from the central government. Regional governments were expected to raise tax revenues from the activity of multinational companies that are invited to their jurisdictions. The topics of tourism development and inward FDI promotion have attracted academic attention as well. Tourism-led economic growth has been discussed by a few recent papers published in this journal (Antonakakis et al., 2015; Chou, 2013; Jalil et al., 2013). Moreover, Ahmed (2012) has examined the relationship between FDI inflows and economic growth.
Please cite this article as: Tomohara, A., Japan's tourism-led foreign direct investment inflows: An empirical study, Econ. Model. (2015), http:// dx.doi.org/10.1016/j.econmod.2015.09.024
A. Tomohara / Economic Modelling xxx (2015) xxx–xxx
3
Table 1 Description of variables with data sources. Variables
Sources
Description
Inward FDI
Japan External Trade Organization
Tourists
“Annual Report of Statistics on Legal Migrants” by the Ministry of Justice, Japan
GDP Immigration Distances
World Development Indicators (WDI) “Statistics on the Foreigners Registered in Japan” by the Japan Immigration Association GeoDist: the CEPII's database on distances
Corruption
The Worldwide Governance Indicators
EPA dummy
The Ministry of Foreign Affairs, Japan
Fair competition
The IMD World Competitiveness Yearbook
Corporate tax rates
The OECD Tax Database
Inward FDI by country and industry
The balance of payments data from the Bank of Japan website
Per capita GDP Taiwanese information
World Development Indicators (WDI) The Taiwanese government website
The FDI data in nominal terms are transformed to real terms using GDP deflators in Japan (100 in 2005) taken from World Development Indicators and U.S.–Japan exchange rates (interbank rates) taken from Principal Global Indicators. The number of foreign tourists with a short-term length of stay (i.e., the Ministry of Justice of Japan defines those who enter Japan and stay for less than or equal to 90 days as foreign entrants with a short-term length of stay). Real GDP (constant 2005 US$) The number of foreign registrations. The “alien registration system” in Japan requires foreigners whose stay exceeds 90 days to register. Geographical distances between Japan and a country of origin of FDI and migration. We use a distance variable, which is coded as distcap (distances) in dist_cepii.xls file. Mayer and Zignago (2011) provide detailed information related to the dataset. An evaluation of governance performance for a host country during 1996–2011 except for the years 1997, 1999, and 2001, and indicators range from −2.5 (weak) to 2.5 (strong). It “reflects perceptions of the extent to which public power is exercised for private gain, including both petty and grand forms of corruption, as well as ‘capture’ of the state by elites and private interests.” Details related to the data can be found in Kaufmann, Kraay, and Mastruzzi (2010). EPA takes the value of 1 if it is effective for more than five months in the year under consideration and 0 otherwise. The dummy is adjusted to zero for Thailand (effective November 1, 2007), Philippines (effective December 11, 2008), Singapore (effective November 30, 2002), and Switzerland (effective September 1, 2009) in the year the EPAs concluded, since the EPAs with these countries were effective for less than five months in the year under consideration. An evaluation of whether “competition legislation is efficient on preventing unfair competition” and indicators range from 0 to 10. “The basic combined central and sub-central statutory corporate tax rate given by the adjusted central government rate plus the sub-central rate.” The FDI data in nominal terms are converted to real terms in US dollars using GDP deflators in Japan (100 in 2005) taken from World Development Indicators and U.S.–Japan exchange rates (i.e., inter-bank rates). Real GDP (constant 2005 US$) Since WDI does not provide Taiwanese data, Taiwanese values are calculated using data on GDP, GDP deflator, population, and exchange rates (NT$ per US$) obtained from the Taiwanese government website.
unobservable time-invariant country-specific fixed effect, and μit, representing idiosyncratic shocks. Table 1 summarizes the description of variables used for the analysis, together with their data sources. The analysis uses net FDI inflow data during 1996–2011. The period corresponds to the time when both FDI promotion and tourism enhancement began to attract attention. Specifically, FDI was expected to complement the decreasing labor force in Japan, because the working-age population began to decrease after reaching its peak in 1995. Technically, the sample period is chosen because of data accessibility. The FDI data before 1995 use different FDI definitions and are not prepared in a style consistent with the FDI data after 1996. Several factors are controlled for in the analysis. Real GDP is a proxy for the market size of a host country. Geographical distances may either discourage FDI because they represent obstacles for international capital movements, or encourage it by replacing trade with countries with greater distances. Higher corporate tax rates can discourage FDI, making host countries less attractive. An Economic Partnership Agreement (EPA) is expected to facilitate cross-border factor movements because it helps countries integrate with the global economy beyond traditional free-trade agreements on goods and services. In addition to institutional factors such as corporate tax rates and EPAs, the analysis controls for quality factors. The corruption and fair competition terms capture attractiveness for investment in a host country. These factors are chosen by referring to Blonigen and Piger (2011) and Eicher et al. (2012), which examined the determinants of FDI. Table 2 presents the sample's summary statistics. Our analysis includes the number of immigrants to control for diaspora effects (or ethnic network externalities). This allows us to
distinguish the effects of tourists from those of immigrants on inward FDI, because recent empirical studies have revealed interactions between migration and FDI (Buch et al., 2006; Foad, 2012; Gheasi et al., 2013; Javorcik et al., 2011; Kugler and Rapoport, 2007; Simone and Manchin, 2012). In the due course, the analysis devises the treatment of Chinese and Korean network externalities. Although the tourist data are different among mainland Chinese, Taiwanese, and Hong Kong citizens, and between North and South Koreans, the immigrant data simply classify them as Chinese or Korean. To fill the gap, the analysis uses the number of Chinese registrations for all of mainland China, Taiwan, and Hong Kong. The treatment assumes that the Chinese share a similar culture to a certain degree and that Chinese network externalities apply to the three countries/areas, so that the total number of Chinese immigrants in Japan affects the number of Chinese tourists to Japan. Similarly, the number of Korean registrations is used for the category of South Korea. This treatment requires a less strict assumption compared with the Chinese case and seems reasonable. The “Annual Report of Statistics on Legal Migrants” shows that the number of North Korean entrants to Japan is negligible compared with those from South Korea, and the most recent entrants from North Korea to Japan are “special permanent residents” (the category is designed for Koreans, Taiwanese, and their offspring who had been living in Japan before September 2, 1945, when Japan signed the Instrument of Surrender) with reentry permits. The distinction of either North or South is not so crucial for registered Koreans who came to Japan before World War II, because North Korea was only established in 1948. We begin to estimate tourism–FDI relationships using the system GMM estimator, where FDI, tourism, immigration, and GDP are treated as endogenous. The treatment accounts not only for simultaneous
Please cite this article as: Tomohara, A., Japan's tourism-led foreign direct investment inflows: An empirical study, Econ. Model. (2015), http:// dx.doi.org/10.1016/j.econmod.2015.09.024
4
A. Tomohara / Economic Modelling xxx (2015) xxx–xxx
Table 2 Summary statistics. Variables
Units of measurement
Mean
Inward FDI Tourists GDP Immigration: type 1 Immigration: type 2 Distances Corruption EPA
Constant 2005 million US$ Logarithm of the number of foreign tourists Constant 2005 million US$ Logarithm of the number of immigrants Logarithm of the number of immigrants Kilometers Indicators ranging from −2.5 to 2.5 1 for EPA taking effect and 0 otherwise
Fair competition Tax rates GDP sum GDP diff squared Factor diff GDP diff ∗ Factor diff
Indicators ranging from 0 to 10 Rates ranging from 0 to 1 Constant 2005 million US$ Constant 2005 million US$ Constant 2005 million US$ Constant 2005 million US$
208.30 1371.51 10.07 2.07 4,502,692 175,722 9.04 2.62 8.94 2.51 7604.48 3398.57 1.25 0.22 Frequency for each EPA dummy 0: 90.6% 5.89 0.58 0.41 0.03 5,732,204 2,319,807 2.E + 13 1.E + 13 −10,102.78 18,203.02 4.E + 10 7.E + 10
Standard deviation
Maximum
Minimum
14,473.11 14.44 4,751,194 13.51 13.43 17,693.20 1.61
−8091.72 4.16 4,221,408 2.48 2.48 1156.67 0.86
1: 6.75 0.50 18,200,000 8.E + 13 50,531.43 2.E + 11
9.5% 4.84 0.40 4,248,272 4.E + 10 −36,321.84 −2.E + 11
Note: Immigration (type 1): The total number of Chinese immigrants is used to control for ethnic network externalities. Immigration (type 2): Taiwanese immigrants are distinguished from other Chinese immigrants. GDP diff stands for differences in real GDP between a source country and a host country. Factor diff is used as a proxy for differences in relative factor endowments in the two countries.
decisions on factor mobility regarding FDI and international tourism, but also for their interaction with immigration and national output. Although a few recent studies have examined the interaction between FDI and immigration, the topic of tourism–FDI relationships given immigration has not received attention in the literature and is thus one of the primary contributions of our study. Another advantage of the approach relates to the selection of instrumental variables. It is often not easy to find relevant instrumental variables. The system GMM enables us to use the information within a dataset without searching for instruments that may not be independent variables in the model. In our analysis, the endogenous variables lagged by two or three periods are used as instruments for difference equations, and their single-lag first differences are
used for level equations. We apply a two-step estimator with robust standard errors, which corrects for finite sample biases (Windmeijer, 2005). Roodman (2009) describes the modest superiority of the twostep estimation with corrected errors to cluster-robust one-step estimation by referring to Windmeijer's simulation.
3. Results The results of the analysis are presented in Table 3. Columns 1–2 present the benchmark results. For each variable, the first row shows the estimated coefficients and the second row in parentheses shows
Table 3 Results of tourism effects. Benchmark Independent variables
L1. Inward FDI Tourists GDP Immigration Fair competition
Knowledge capital
Inward FDI
Inward FDI
(1)
(2)
(3)
(4)
0.376 (0.003)⁎⁎⁎ 109.523 (32.243)⁎⁎⁎ 2.42E−04 (0.067E−03)⁎⁎⁎ −54.421 (27.947)⁎⁎
0.377 (0.020)⁎⁎⁎ 121.672 (22.206)⁎⁎⁎ 0.178E−03 (0.086E−03)⁎⁎ −52.869 (11.507)⁎⁎⁎
0.382 (0.034)⁎⁎⁎ 107.556 (29.637)⁎⁎⁎ 0.235E−03 (0.063E −03)⁎⁎⁎ −56.019 (8.305)⁎⁎⁎
0.373 (0.022)⁎⁎⁎
L1. Inward FDI
117.827 (25.557)⁎⁎⁎ 0.221E−03 (0.070E−03)⁎⁎⁎ −53.463⁎⁎⁎ (12.399)⁎⁎⁎
Tourists
304.348 (35.008)⁎⁎⁎
259.772 (57.603)⁎⁎⁎
289.142 (43.040)⁎⁎⁎
304.098 (29.613)⁎⁎⁎
GDP sum GDP diff squared Factor diff GDP diff ∗ Factor diff
Distances Corruption Tax rates EPA
AR(1) AR(2) No. of observation
0.02 (0.007)⁎⁎⁎ −1306.66 (114.834)⁎⁎⁎
0.026 (0.013)⁎⁎ −1168.075 (224.746)⁎⁎⁎
0.020 (0.007)⁎⁎⁎ −1255.479 (125.185)⁎⁎⁎
0.027 (0.013)⁎⁎ −1334.669 (120.722)⁎⁎⁎
−4542.254 (528.98)⁎⁎⁎ −109.061 229.988
−4160.822 (675.589)⁎⁎⁎
−4303.582 (635.867)⁎⁎⁎ −128.486 206.659
−4632.742 (444.092)⁎⁎⁎
0.072 0.161 319
0.072 0.161 319
0.068 0.162 319
0.072 0.160 319
Distances Corruption Tax rates EPA
AR(1) AR(2) No. of observation
(5)
(6)
0.276 (0.009)⁎⁎⁎ 121.325 (37.074)⁎⁎ −0.11E−03 (2.72E−05)⁎⁎⁎
0.278 (0.010)⁎⁎⁎ 120.987 (37.687)⁎⁎⁎ −0.110E−03 (2.74E−05)⁎⁎⁎
8.17E−11 (6.01E−12)⁎⁎⁎ −0.064 (0.014)⁎⁎⁎ −1.41E−08 (3.04E−09)⁎⁎⁎ 0.119 (0.035)⁎⁎⁎
8.06E−11 (6.17E−12)⁎⁎⁎ −0.062 (0.013)⁎⁎⁎ −1.43E−08 (2.15E−09)⁎⁎⁎ 0.108 (0.037)⁎⁎⁎
−867.473 (110.954)⁎⁎⁎ −3912.430 (508.908)⁎⁎⁎ 119.383 115.908
−807.251 (122.168)⁎⁎⁎ −3786.713 (636.529)⁎⁎⁎
0.077 0.142 319
0.076 0.144 319
Note: Figures in parentheses indicate standard error. The Arellano–Bond test was conducted to assess autocorrelation. The test for second-order correlation in differences, AR (2), shows no serial correlation in the first-difference disturbances. ⁎⁎⁎ Denotes coefficient significance at 1% level. ⁎⁎ Denotes coefficient significance at 5% level. ⁎ Denotes coefficient significance at 10% level.
Please cite this article as: Tomohara, A., Japan's tourism-led foreign direct investment inflows: An empirical study, Econ. Model. (2015), http:// dx.doi.org/10.1016/j.econmod.2015.09.024
A. Tomohara / Economic Modelling xxx (2015) xxx–xxx
standard errors. Column 1 presents the results obtained with all control variables. Since EPA is estimated to be statistically insignificant, we attempt estimation without the EPA variable (Column 2). We take the logarithm of tourism and immigration variables, which have distributions with a positive skew, so that the distributions of these variables will be more normal. Our results show that international tourism has synergy with inward FDI. The coefficient on inbound foreign tourists is estimated as positive and statistically significant. For example, a one percent increase in foreign tourists into Japan encourages inward FDI into Japan of about $1.22 million (Column 2).4 A favorable interaction between inward FDI and foreign tourism is observed when they share similar policy targets. The enhancement of foreign tourism amplifies the effectiveness of FDI promotion. The results remain valid under different specifications of Chinese network externalities when the sensitivity of network externalities is examined by different treatments of Chinese immigrants (Columns 3 and 4). We try disaggregated data, where Taiwanese immigrants are distinguished from other Chinese immigrants.5 In fact, it does not seem crucial to distinguish Taiwanese immigrants from other Chinese immigrants in capturing diaspora dynamics. The magnitude of coefficient estimates on tourists is similar in Columns 4 and 2 even after adjusting for some differences in Taiwan and mainland China, including Hong Kong. It would be of interest to know whether expanded international tourism mainly increases inward FDI in tourism-related industries, and thus, whether most effects of international tourism on inward FDI are explained by the increased inward FDI in the industry. However, this seems less likely, as the level of FDI inflows into tourism-related industry sectors such as transportation and services is not large compared to the total inward FDI in Japan. Rather, major recipients of inward FDI in Japan are industry sectors such as finance and insurance, communications, wholesale and retail (non-manufacturing sectors), chemicals and pharmaceuticals, and electric machinery (manufacturing sectors).6 We conduct further examination using recent data on inward FDI by country and industry to support our assertion. We define sectors labeled as transportation and services as tourism-related industries and merge the data on tourism-related inward FDI to our original dataset sorted by country and year. The data are restricted, as they are available only from the year 2005 and information is concealed to protect privacy if the inward FDI cell for each country and industry contains less than three reports. There are three cases where tourism-related inward FDI data by country and year are missing because of privacy protection. When inward FDI into the transportation (or services) sector is positive but data on the other sector is concealed, we treat the positive values in one sector as inward FDI into tourism-related industries by assuming that the concealed values are negligible as data are concealed only if the industry contains less than three reports. There are 24 such cases.
4 Princeton University Library, Data and Statistical Services http://dss.princeton.edu/ online_help/stats_packages/stata/log.html (last accessed on July 23, 2015). 5 The government statistics began to distinguish the Taiwanese from the Chinese after the “alien registration system” was abolished and the “new residency management system” was introduced in 2012. Since the data on Taiwanese immigrants is available only for 2012, we apply a ratio of the Taiwanese among all the Chinese in 2012 to split the total Chinese immigrants into Taiwanese immigrants and other Chinese immigrants during the sample period. 6 Because net FDI inflows take negative values, simply taking the share of tourismrelated inward FDI to total inward FDI does not provide the relative importance of tourism-related industry sectors among all industry sectors. Therefore, we take absolute values of inward FDI in all sectors and sum them. Then, we calculate the share of tourism-related inward FDI to total inward FDI for each sample year. The values vary within the range of 0.08%–27.6% during the period 2005–2011. These values are mostly less than 13%, except for 2005, when the transportation sector received large FDI inflows. Data used for the analysis are available at http://www.jetro.go.jp/world/japan/stats/fdi/ (last accessed on July 23, 2015).
5
Table 4 Spillover effects.
Independent variables
L1. Dependent variable Tourists
System GMM
LSDVC
Tourism-related Total FDI FDI
Tourism-related Total FDI FDI
(1)
(2)
(3)
(4)
0.356 (0.020)⁎⁎⁎ 292.418 (68.493)⁎⁎⁎
0.281 (0.155)⁎
0.341 (0.145)⁎⁎ 387.835 (1164.849) 0.002 (0.003) −322.502 (5973.756) −346.176 (1603.951) 302.564 (945.773)
0.047 (0.175) 49.505 (27.608)⁎ GDP −5.E−05 (3.E−05) Immigration −37.809 (19.719)⁎ EPA 22.018 (47.099) Fair competition 14.396 (12.016) AR(1) 0.484 AR(2) 0.415 No. of 106 observation
11.248 (89.771) −7.E−04 −2.E−04 (0.8E−04)⁎⁎⁎ (2.E−04) −149.664 440.201 (54.091)⁎⁎⁎ (525.120) −738.501 −10.819 (238.024)⁎⁎⁎ (106.631) 313.244 −42.301 (67.724)⁎⁎⁎ (93.228) 0.080 0.156 141 105
141
Note: Figures in parentheses indicate standard errors. Since bias-corrected LSDV estimators do not report standard errors, bootstrap standard errors are reported. The Arellano– Bond test was conducted to assess autocorrelation. The test for second-order correlation in differences, AR(2), shows no serial correlation in the first-difference disturbances. ⁎⁎⁎ Denotes coefficient significance at 1% level. ⁎⁎ Denotes coefficient significance at 5% level. ⁎ Denotes coefficient significance at 10% level.
When inward FDI into the transportation (or services) sector equals zero and data on the other sector is concealed, we treat the case as missing data. There are 31 such cases. The results of the analysis are shown in Table 4. The column labeled as tourism-related FDI shows the results, when tourism-related inward FDI replaces total inward FDI in Eq. (1). The column labeled as total FDI replicates the analysis of total FDI under our benchmark model during the period 2005–2011. The coefficient of tourists is estimated to be positive and statistically significant in both columns, but the coefficient in Column 1 is smaller than that in Column 2. This indicates that tourism enhancement may have positive spillovers on inward FDI beyond tourism-related industries. The coefficient in Column 2 is larger because it includes not only the direct effects in tourism-related sectors but also indirect spillover effects in other sectors. Other sectors also enjoyed increased FDI inflows from the same country of origin of the tourists when tourists from the country increased. The results suggest that the two policies of inward FDI promotion and inbound foreign tourism enhancement are mutually inclusive. The robustness of the results is examined by using the bias-corrected least-squares dummy variable (LSDVC) estimator developed by Bruno (2005a,b). The LSDVC estimator is an alternate approach to solve dynamic panel bias.7 Bruno extends the discussions of LSDVC in Kiviet (1995, 1999), and Bun and Kiviet (2003), which is applicable only for balanced panels, to unbalanced panels. The results of the analysis using the LSDVC estimator of Bruno are summarized in the right-hand side of Table 4. They are similar to those obtained using the system GMM: the degree of effects of tourism expansion is estimated to be larger for total FDI than for tourism-related FDI. Although the two
7 In the analysis of dynamic panels for finite time dimension and large number of crosssectional units, the lagged dependent variable's coefficient estimated by least squares dummy variable estimators is biased because the lagged dependent variable is correlated with the fixed effects (Nickell, 1981). GMM estimators are designed to handle the bias (and endogeneity issues) in analyzing panel data with short time dimension and large number of cross-sectional units.
Please cite this article as: Tomohara, A., Japan's tourism-led foreign direct investment inflows: An empirical study, Econ. Model. (2015), http:// dx.doi.org/10.1016/j.econmod.2015.09.024
6
A. Tomohara / Economic Modelling xxx (2015) xxx–xxx
estimation approaches are not comparable exactly, it is interesting to show similar trends regarding tourism effects on FDI.8 The robustness of the analysis is further examined using different specifications of economic theories. We investigate the effects of international tourism in the context of the knowledge capital model, where both vertical and horizontal FDI arise (Markusen, 1997; Markusen et al., 1996). We introduce FDI dynamics, together with tourism effects, into Carr et al. (2001, 2003), which studies the knowledge capital model empirically: 2 FDI it ¼ α þ β1 FDI it−１ ＋β2 GDP it þ GDP jt þ β3 GDP it−GDP jt þ β4 Factor it −Factor jt þ β 5 Factor it −Factor jt GDP it −GDP jt þ β6 Tourit þ X it ρ þ εit where εit ¼ δi þ μ it : GDPit (or GDPjt) represents the real GDP of a source country i (or a host country j) at time t, Factorit is the per capita real GDP, and Xit is a set of control variables related to investment and trade costs. Horizontal FDI increases with increasing similarity in income and relative factor endowments between the source and host countries. Vertical FDI, on the other hand, increases with the dissimilarity of factor endowments between the two countries. Differences in per capita GDP are used as a proxy for differences in relative factor endowments. We take the logarithm of tourism variable, which has a positively skewed distribution. Results support a favorable interaction between inward FDI and international tourism, as obtained from the previous analysis. Columns 5–6 in Table 3 show the results of the analysis using a system GMM. A two-step estimator with robust standard errors is applied by treating FDI, GDP-related terms, and foreign tourists as endogenous. The endogenous variables lagged by two or three periods are used as instruments for difference equations, and their once-lagged first differences are used for level equations. Since EPA is estimated to be statistically insignificant again, we attempt estimation without the EPA variable. The coefficient on foreign tourists is estimated to be positive and statistically significant. Interestingly, the degree of impacts of international tourism on inward FDI in Columns 5–6 is similar to those in Columns 1–4 despite different model specifications. A one percent increase in foreign tourists into Japan encourages inward FDI into Japan of about $1.21 million (Column 6). It is notable to report that a favorable interaction between international tourism and inward FDI is confirmed under different model specifications. Spillover effects of tourism enhancement on FDI are observed.
enhanced tourism exhibits spillovers of FDI beyond tourism-related sectors. This study is intended to investigate a specific policy implication: Two policy instruments—FDI promotion and foreign tourism enhancement—may entail an unexpected virtuous circle, as in the case of Japan. Our analysis reveals a favorable interaction between inward FDI and foreign tourists visiting Japan. Specifically, tourism enhancement has positive spillover effects on inward FDI beyond tourism-related industries. Additionally, the robustness of favorable interactions between tourism and FDI is confirmed under different model specifications. The results suggest that FDI in other industries needs to be included when discussing the effects of tourism enhancement. The inclusion can alter the degree of effects of tourism enhancement from earlier studies. Considering that the literature focuses on only tourism-related FDI and does not include FDI beyond tourism-related industries, previous studies might have underestimated the real effects of tourism enhancement. Our analysis implies that the two policies of inward FDI promotion and inbound foreign tourism enhancement are mutually inclusive, and that one policy amplifies the effectiveness of the other. The results have implications for policy designs and evaluation. Since these two policies are often planned, executed, and evaluated independently under different government jurisdictions, policy coordination between tourism and FDI should be encouraged to attain goals such as revitalization of declining local economies. Additionally, the analysis suggests practical implications to develop an index for evaluating the tourism– FDI policy. The degree of interaction can be used as a proxy when evaluating the effectiveness of policy coordination between tourism and FDI. The current analysis has several potential avenues of extension. One possible extension is to examine if similar implications might apply to situations in other countries. Although the analysis uses Japanese data, a favorable interaction between tourism and FDI might be observed in other countries. If this is the case, another more interesting extension includes examining the factors that influence such interactions, using panel data. The degree of interaction might vary depending on the fundamentals in each country. The usage of panel data would allow us to compare cross-country differences regarding the degree of policy interactions, and to clarify the conditions under which stronger tourism–FDI interactions occur. Such an analysis requires additional data on bilateral FDI and tourism in several countries, and is beyond the scope of the current paper. All of these topics represent potential future lines of research.
4. Conclusions
References
This study analyzes tourism–FDI interactions in the framework of FDI determination. Although the literature focuses on the relationships between tourism and tourism-related FDI, we examine whether
Ahmed, E.M., 2012. Are the FDI inflow spillover effects on Malaysia's economic growth input driven? Econ. Model. 29 (4), 1498–1504. Antonakakis, N., Dragouni, M., Filis, G., 2015. How strong is the linkage between tourism and economic growth in Europe? Econ. Model. 44, 142–155. Arellano, M., Bover, O., 1995. Another look at the instrumental variable estimation of error-components models. J. Econom. 68, 29–52. Blonigen, B.A., Piger, J., 2011. Determinants of foreign direct investment. NBER Working Paper No. 16704. Blundell, R., Bond, S., 1998. Initial conditions and moment restrictions in dynamic panel data models. J. Econom. 87, 115–143. Bruno, G.S.F., 2005a. Approximating the bias of the LSDV estimator for dynamic unbalanced panel data models. Econ. Lett. 87 (3), 361–366. Bruno, G.S.F., 2005b. Estimation and inference in dynamic unbalanced panel-data models with a small number of individuals. Stata J. 5 (4), 473–500. Buch, C.M., Kleinert, J., Toubal, F., 2006. Where enterprises lead, people follow? Links between migration and FDI in Germany. Eur. Econ. Rev. 50, 2017–2036. Bun, M.J.G., Kiviet, J.F., 2003. On the diminishing returns of higher order terms in asymptotic expansions of bias. Econ. Lett. 79 (2), 145–152. Carr, D.L., Markusen, J.R., Maskus, K.E., 2001. Estimating the knowledge-capital model of the multinational enterprise. Am. Econ. Rev. 91 (3), 693–708. Carr, D.L., Markusen, J.R., Maskus, K.E., 2003. Estimating the knowledge-capital model of the multinational enterprise: reply. Am. Econ. Rev. 93 (3), 995–1001. Chou, M.C., 2013. Does tourism development promote economic growth in transition countries? A panel data analysis. Econ. Model. 33, 226–232. Eicher, T.S., Helfman, L., Lenkoski, A., 2012. Robust FDI determinants: Bayesian model averaging in the presence of selection bias. J. Macroecon. 34, 637–651. Foad, H., 2012. FDI and immigration: a regional analysis. Ann. Reg. Sci. 49, 237–259.
8 Both approaches have advantages and limitations. LSDVC estimators have the advantage of being suitable for panel data with smaller number of cross-sectional units. Additionally, Monte Carlo experiments reveal that LSDVC estimators outperform GMM estimators in several cases (Bruno, 2005b; Judson and Owen, 1999). However, LSDVC estimators of Bruno assume strictly exogenous explanatory variables and their analytical standard errors often break down, requiring the use of bootstrap (Bruno, 2005a, 2005b). On the other hand, GMM estimators allow for endogeneity that arise in our model. While GMM estimators assume the absence of second-order serial correlation, our GMM analysis is shown to satisfy the assumption. It is difficult to conclude that one estimator outperforms others in all cases, since previous Monte Carlo experiments examined only a limited number of cases. Our baseline analysis follows that of Judson and Owen (1999), who recommend GMM rather than LSDVC estimators in the case of unbalanced panels with a time dimension of 20 years or less. Moreover, our analysis fits closer to the experimental setting of Judson and Owen (1999) than to that of Bruno (2005a, b). For example, the number of cross-sectional units (29) is slightly larger than 10 and 20, which Bruno (2005a, b) considers to be small in a Monte Carlo analysis.
Please cite this article as: Tomohara, A., Japan's tourism-led foreign direct investment inflows: An empirical study, Econ. Model. (2015), http:// dx.doi.org/10.1016/j.econmod.2015.09.024
A. Tomohara / Economic Modelling xxx (2015) xxx–xxx Gheasi, M., Nijkamp, P., Rietveld, P., 2013. Migration and foreign direct investment: education matters. Ann. Reg. Sci. 51, 73–87. Jalil, A., Mahmood, T., Idrees, M., 2013. Tourism–growth nexus in Pakistan: evidence from ARDL bounds tests. Econ. Model. 35, 185–191. Javorcik, B.S., Özden, Ç., Spatareanu, M., Neagu, C., 2011. Migrant networks and foreign direct investment. J. Dev. Econ. 94 (2), 231–241. Judson, R.A., Owen, A.L., 1999. Estimating dynamic panel data models: a guide for macroeconomists. Econ. Lett. 65 (1), 9–15. Katircioglu, S., 2011. The bounds test to the level relationship and causality between foreign direct investment and international tourism: the case of Turkey. E + M Econ. Manag. 1, 6–13 (E + M Ekonomie A Management). Kaufmann, D., Kraay, A., Mastruzzi, M., 2010. The worldwide governance indicators: methodology and analytical issues. Policy Research Working Paper Series 5430. Kiviet, J.F., 1995. On bias, inconsistency, and efficiency of various estimators in dynamic panel data models. J. Econom. 68 (1), 53–78. Kiviet, J.F., 1999. Expectation of expansions for estimators in a dynamic panel data model: some results for weakly exogenous regressors. In: Hsiao, C., Lahiri, K., Lee, L.F., Pesaran, M.H. (Eds.), Analysis of Panels and Limited Dependent Variable Models. Cambridge University Press, Cambridge, pp. 199–225. Kugler, M., Rapoport, H., 2007. International labor and capital flows: complements or substitutes? Econ. Lett. 94, 155–162. Markusen, J.R., 1997. Trade versus investment liberalization. NBER Working Paper No. 6231. Markusen, J.R., Venables, A.J., Konan, D.E., Zhang, K.H., 1996. A united treatment of horizontal direct investment, vertical direct investment, and the pattern of trade in goods and services. NBER Working Paper No. 5696. Mayer, T., Zignago, S., 2011. Notes on CEPII's distances measures: the GeoDist database. CEPII Working Paper No. 2011-25. Ministry of Land, Infrastructure, Transport and Tourism, Japan, 2003. Global tourism strategies, mimeo. http://www.mlit.go.jp/kisha/kisha02/01/011224_3/011224_3.pdf (last accessed on July 23, 2015). Nickell, S.J., 1981. Biases in dynamic models with fixed effects. Econometrica 49 (6), 1417–1426.
7
Roodman, D., 2009. How to do xtabond2: an introduction to difference and system GMM in Stata. Stata J. 9 (1), 86–136. Samimi, A.J., Sadeghi, S., Sadeghi, S., 2013. The relationship between foreign direct investment and tourism development: evidence from developing countries. Inst. Econ. 5 (2), 59–68. Selvanathan, S., Selvanathan, E.A., Viswanathan, B., 2012. Causality between foreign direct investment and tourism: empirical evidence from India. Tour. Anal. 17 (1), 91–98. Simone, G.D., Manchin, M., 2012. Outward migration and inward FDI: factor mobility between Eastern and Western Europe. Rev. Intern. Econ. 20 (3), 600–615. Tang, S., Selvanathan, E.A., Selvanathan, S., 2007. The relationship between foreign direct investment and tourism: empirical evidence from China. Tour. Econ. 13 (1), 25–39. United Nations Conference on Trade and Development (UNCTAD), 2007. FDI in Tourism: The Development Dimension (New York and Geneva). Windmeijer, F., 2005. A finite sample correction for the variance of linear efficient twostep GMM estimators. J. Econom. 126 (1), 25–51. Akinori Tomohara received his Ph.D. in Economics from Johns Hopkins University. Prior to graduate study, Dr. Tomohara worked for the Government of Japan. His experience includes economic research at Columbia University and faculty positions at Queens College and Graduate Center, City University of New York and at University of Pittsburgh. He was an economist at University of California, Los Angeles. A broad range of research skills, work experience, and contacts earned Dr. Tomohara the distinction of being an economic consultant at the Urban Institute, the World Bank, and the Inter-American Development Bank. Currently, he is a Professor in Aoyama Gakuin University.
Please cite this article as: Tomohara, A., Japan's tourism-led foreign direct investment inflows: An empirical study, Econ. Model. (2015), http:// dx.doi.org/10.1016/j.econmod.2015.09.024
Contents lists available at ScienceDirect
Economic Modelling journal homepage: www.elsevier.com/locate/ecmod
Japan's tourism-led foreign direct investment inflows: An empirical study☆ Akinori Tomohara 1 University of California, Los Angeles, Anderson Forecast, CA 90095-1481, USA Department of International Economics, Aoyama Gakuin University, Shibuya-ku, Tokyo 150-8366, Japan
a r t i c l e
i n f o
Article history: Accepted 16 September 2015 Available online xxxx Keywords: Immigration International tourism FDI Local economy
a b s t r a c t This study analyzes international tourism–inward foreign direct investment (FDI) relationships by applying dynamic panel models to Japanese data for the 1996–2011 period. A growing number of studies have examined this topic following the publication of the UNCTAD (2007) report; however, the current literature is inadequate. Our analysis is not restricted to tourism-related FDI such as hotels and restaurants, as in the literature, but covers general FDI in the framework of FDI dynamics, because tourism–FDI interactions are not in reality restricted to typical tourism-related FDI. The analysis results reveal that enhanced inbound foreign tourism exhibits spillovers of inward FDI beyond tourism-related sectors. Hence, although policies related to FDI promotion and tourism enhancement are often planned, executed, and evaluated independently under different government jurisdictions, coordination is recommended, because these policies interact with each other in efforts to attain economic development goals. © 2015 Elsevier B.V. All rights reserved.
1. Introduction International organizations have begun promoting tourism–FDI relationships as a vehicle for development (United Nations Conference on Trade and Development (UNCTAD), 2007). Tourism facilitates not only export revenue generation but also service sector job creation and local economic revival through tourism-related FDI. Since the publication of the United Nations Conference on Trade and Development (UNCTAD) (2007) report, a growing number of studies have begun to analyze tourism–FDI relationships. The primary focus in the literature is the relationship between tourism (for which the number of arrivals and revenues are used as a proxy) and tourism-related FDI (i.e., FDI inflows into setting up tourism infrastructure such as accommodation, restaurants, and transportation in less-developed countries; Katircioglu,
☆ I appreciate Jin Ke for her research assistance and acknowledge the financial support from Aoyama Gakuin University. The funding source was not involved in the conduct of the research (study design, the collection, analysis, and interpretation of data) and/or preparation of the article (the writing of the report), and the decision to submit the article for publication. I would like to thank the editor and two anonymous referees for their helpful comments. The usual disclaimer applies. E-mail address: [email protected]. 1 Present address: Department of International Economics, Aoyama Gakuin University, Building No. 8-421, 4-4-25 Shibuya, Shibuya-ku, Tokyo 150-8366, Japan. Tel.: +81 3 3409 8562; fax: +81 3 5485 0782.
2011 for Turkey; Samimi et al., 2013 for 20 developing countries; Selvanathan et al., 2012 for India; Tang et al., 2007 for China). Enhanced tourism is expected to contribute to the development of an economy through increased FDI inflows into tourism-related industries. Thus, tourism-related FDI is assumed to be a primary channel for economic development. However, the existing research in this field is inadequate, because tourism–FDI interactions are not in reality restricted to typical tourism-related FDI such as in hotels, airlines, and restaurants. Tourism development may induce FDI in other sectors as well. For example, China's Suning Commerce Group Co. Ltd. formed a capital business alliance with Japan's Laox Appliance (a retail store for home appliances) and the latter became a subsidiary of the former in 2009.2 One motivation for this alliance is the business opportunity of serving the growing number of Chinese tourists to Japan by offering services such as language support, duty-free arrangements, and acceptance of Chinese credit cards. This is because rice cookers, air purifiers, and digital cameras manufactured in Japan are principal purchase items for Chinese tourists. The example illustrates the possibility that enhanced tourism may induce FDI in sectors beyond hotels, etc. Hence, this study analyzes international tourism–inward FDI relationships. Our analysis is not restricted to typical tourism-related FDI
2
www.laox.co.jp/company/outline/ (last accessed on July 23, 2015).
http://dx.doi.org/10.1016/j.econmod.2015.09.024 0264-9993/© 2015 Elsevier B.V. All rights reserved.
Please cite this article as: Tomohara, A., Japan's tourism-led foreign direct investment inflows: An empirical study, Econ. Model. (2015), http:// dx.doi.org/10.1016/j.econmod.2015.09.024
2
A. Tomohara / Economic Modelling xxx (2015) xxx–xxx
in the literature, but covers general FDI as well. Motivated by anecdotal evidence that tourism influences FDI in non-tourism sectors in Japan, we aim to fill the existing gap in the literature by estimating the effects of inbound tourism on inward FDI beyond tourism-related sectors. We set up our empirical model by referring to the analytical framework of FDI determination. Our model extends the framework to a dynamic context after introducing international tourism. In order to investigate these dynamics, the analysis uses a sample of 29 countries/ areas, each of which is a major FDI donor to Japan, over the period 1996–2011. Our analysis is distinct from other analyses in the literature, which examine the causality between tourism development and tourism-related FDI using time-series analyses including only these two variables. We focus on the spillover effects of tourism development on FDI, and thus, we need to control for other factors determining FDI. Such analysis requires the use of panel data on bilateral FDI and tourism between donor and recipient countries, which are not necessarily required for the time-series analyses presented in the literature. Bilateral FDI data, together with the number of foreign tourists and other control variables, are compiled for the analysis. Then, the tourism–FDI interactions are estimated by using a two-step system generalized method of moments (Arellano and Bover, 1995; Blundell and Bond, 1998; Roodman, 2009; Windmeijer, 2005). The analysis results reveal that an increased number of foreign tourists visiting Japan helps promote FDI into Japan, and that the effects of tourism on FDI in non-tourism industries are observed to be significant. The effects of tourism on total FDI are estimated to be 5.6 times as large as those on tourism-related FDI. The results of a favorable interaction between tourism and FDI imply that the two policies of inward FDI promotion and inbound foreign tourism enhancement are mutually inclusive. There exist positive spillovers from promoting tourism to encouraging FDI in various sectors. A favorable interaction between international tourism and inward FDI is confirmed under different specifications. The results obtained using a benchmark model are robust when we apply a different FDI modeling of the knowledge capital model (Carr et al., 2001, 2003; Markusen, 1997; Markusen et al., 1996) and a different empirical approach using the bias-corrected least-squares dummy variable (LSDVC) estimator of Bruno (2005a,b) to our context. The results of our analysis suggest that the effect of international tourism should have been extended to general FDI dynamics in previous works. The literature focusing on tourism-related FDI has the risk of underestimating the effects of tourism enhancement on FDI in reality. Our results have important policy implications that are not covered in the literature. Tourism enhancement helps attain economic growth in not only less-developed tourism-oriented countries but also others, because international tourism positively influences FDI beyond traditional tourism sectors. Additionally, policy coordination is recommended between FDI promotion and tourism enhancement, although the two policies are often planned, executed, and evaluated independently under different government jurisdictions. These policies interact with each other to enable the attainment of economic development goals. The remainder of this paper is organized as follows. Section 2 presents a description of data used for the analysis and introduces the model. Section 3 discusses the results of the analysis. Section 4 concludes the paper with future suggested lines of research. 2. Model and data Our analysis uses Japanese data to study tourism–FDI interactions. Japan is relevant for analysis because it allows investigating the interactive relationships between inward FDI promotion and inbound foreign tourism enhancement. The Japanese government has aggressively targeted both inward FDI and tourism as priority policy areas for
enhanced growth performance,3 and thus, from a policy perspective, it is critical to know whether there are significant spillover effects among policies. Specifically, the situation in Japan amplifies the importance of examining spillover effects of increased tourism beyond tourism-related industries. Such industries are not major recipients of inward FDI in Japan; rather, Japan receives most of its FDI in other sectors such as finance and insurance, communications, wholesale and retail, chemicals and pharmaceuticals, and electric machinery. The two policies of inward FDI promotion and tourism enhancement may not help in attaining economic development goals if tourism development affects FDI into tourism-related industries only. Moreover, the Japanese situation allows us to examine whether enhanced tourism affects inward FDI in other industry sectors, because such an examination is not feasible if other sectors are not mature enough to receive inward FDI. We use major FDI donor countries to Japan as our sample. The sample comprises 29 countries/areas from Asia (China, Hong Kong, Taiwan, South Korea, Singapore, Thailand, Indonesia, Malaysia, Philippines, India), North America (the United States, Canada), Central America (Mexico, Brazil), Pacific (Australia, New Zealand), Europe (Germany, United Kingdom, France, Netherlands, Italy, Belgium, Luxembourg, Switzerland, Sweden, Spain, Russia), and the Middle East (Saudi Arabia, United Arab Emirates). The Cayman Islands, a major FDI donor, is not included in our sample because the Japanese government does not provide any records of tourists from the Cayman Islands. Our empirical model extends the framework of FDI determination to a dynamic context after introducing international tourism and controlling for other factors determining FDI. Whereas previous works focused on the causality between tourism and tourism-related FDI (Katircioglu, 2011; Samimi et al., 2013; Selvanathan et al., 2012; Tang et al., 2007), our approach is expressed in the following conceptual framework. FDI is a function of Tour and control variables, X: FDI = f(Tour, X). Since FDI is composed of FDItourism − related industry + FDIother indutries, the aforementioned literature with missing effects of tourism on FDIother indutries might mislead the policy effectiveness of tourism enhancement. In order to examine the possibility, FDI dynamics are examined using a system generalized method of moments (GMM), in which current FDI relies on past FDI realizations and the number of foreign tourists. The system GMM accounts for simultaneous bias between FDI inflows and foreign tourist arrivals. FDI it ¼ α þ β1 FDI it−１ ＋β2 Tour it þ X it ρ þ ε it ; where ε it ¼ δi þ μ it
ð1Þ
FDIit represents the net inflow of FDI from country i to Japan at time t, Tourit denotes the number of foreign tourists visiting Japan, Xit is a set of control variables, and εit is an error term, which comprises δi, an 3 The Japanese government has been promoting international tourism ever since the “Basic Policies for Economic and Fiscal Management and Structural Reform” endorsed by the Cabinet in 2002 suggested increasing the number of foreign tourists visiting Japan. Correspondingly, the Ministry of Land, Infrastructure, Transport and Tourism (2003) proposed strategies to promote foreign tourists visiting Japan by providing a report entitled “Global tourism strategies”. The report recognizes international tourism as a promising sector because the size of international tourism is expanding globally according to the World Tourism Organization (UNWTO), and expects expanded international tourism to vitalize the Japanese economy, specifically the declining local economy. The Japanese government has also been trying to promote inward FDI for similar policy ends. During the 1990s, FDI promotion was proposed to revitalize the Japanese economy when Japan slipped into recession after the burst of the bubble economy. After the Economic and Fiscal Policy Management and Structural Reform in 2002, regional governments initiated inward FDI to revitalize the local economy and to acquire greater tax revenues. This is because the reform required regional governments to be less dependent on fiscal support from the central government. Regional governments were expected to raise tax revenues from the activity of multinational companies that are invited to their jurisdictions. The topics of tourism development and inward FDI promotion have attracted academic attention as well. Tourism-led economic growth has been discussed by a few recent papers published in this journal (Antonakakis et al., 2015; Chou, 2013; Jalil et al., 2013). Moreover, Ahmed (2012) has examined the relationship between FDI inflows and economic growth.
Please cite this article as: Tomohara, A., Japan's tourism-led foreign direct investment inflows: An empirical study, Econ. Model. (2015), http:// dx.doi.org/10.1016/j.econmod.2015.09.024
A. Tomohara / Economic Modelling xxx (2015) xxx–xxx
3
Table 1 Description of variables with data sources. Variables
Sources
Description
Inward FDI
Japan External Trade Organization
Tourists
“Annual Report of Statistics on Legal Migrants” by the Ministry of Justice, Japan
GDP Immigration Distances
World Development Indicators (WDI) “Statistics on the Foreigners Registered in Japan” by the Japan Immigration Association GeoDist: the CEPII's database on distances
Corruption
The Worldwide Governance Indicators
EPA dummy
The Ministry of Foreign Affairs, Japan
Fair competition
The IMD World Competitiveness Yearbook
Corporate tax rates
The OECD Tax Database
Inward FDI by country and industry
The balance of payments data from the Bank of Japan website
Per capita GDP Taiwanese information
World Development Indicators (WDI) The Taiwanese government website
The FDI data in nominal terms are transformed to real terms using GDP deflators in Japan (100 in 2005) taken from World Development Indicators and U.S.–Japan exchange rates (interbank rates) taken from Principal Global Indicators. The number of foreign tourists with a short-term length of stay (i.e., the Ministry of Justice of Japan defines those who enter Japan and stay for less than or equal to 90 days as foreign entrants with a short-term length of stay). Real GDP (constant 2005 US$) The number of foreign registrations. The “alien registration system” in Japan requires foreigners whose stay exceeds 90 days to register. Geographical distances between Japan and a country of origin of FDI and migration. We use a distance variable, which is coded as distcap (distances) in dist_cepii.xls file. Mayer and Zignago (2011) provide detailed information related to the dataset. An evaluation of governance performance for a host country during 1996–2011 except for the years 1997, 1999, and 2001, and indicators range from −2.5 (weak) to 2.5 (strong). It “reflects perceptions of the extent to which public power is exercised for private gain, including both petty and grand forms of corruption, as well as ‘capture’ of the state by elites and private interests.” Details related to the data can be found in Kaufmann, Kraay, and Mastruzzi (2010). EPA takes the value of 1 if it is effective for more than five months in the year under consideration and 0 otherwise. The dummy is adjusted to zero for Thailand (effective November 1, 2007), Philippines (effective December 11, 2008), Singapore (effective November 30, 2002), and Switzerland (effective September 1, 2009) in the year the EPAs concluded, since the EPAs with these countries were effective for less than five months in the year under consideration. An evaluation of whether “competition legislation is efficient on preventing unfair competition” and indicators range from 0 to 10. “The basic combined central and sub-central statutory corporate tax rate given by the adjusted central government rate plus the sub-central rate.” The FDI data in nominal terms are converted to real terms in US dollars using GDP deflators in Japan (100 in 2005) taken from World Development Indicators and U.S.–Japan exchange rates (i.e., inter-bank rates). Real GDP (constant 2005 US$) Since WDI does not provide Taiwanese data, Taiwanese values are calculated using data on GDP, GDP deflator, population, and exchange rates (NT$ per US$) obtained from the Taiwanese government website.
unobservable time-invariant country-specific fixed effect, and μit, representing idiosyncratic shocks. Table 1 summarizes the description of variables used for the analysis, together with their data sources. The analysis uses net FDI inflow data during 1996–2011. The period corresponds to the time when both FDI promotion and tourism enhancement began to attract attention. Specifically, FDI was expected to complement the decreasing labor force in Japan, because the working-age population began to decrease after reaching its peak in 1995. Technically, the sample period is chosen because of data accessibility. The FDI data before 1995 use different FDI definitions and are not prepared in a style consistent with the FDI data after 1996. Several factors are controlled for in the analysis. Real GDP is a proxy for the market size of a host country. Geographical distances may either discourage FDI because they represent obstacles for international capital movements, or encourage it by replacing trade with countries with greater distances. Higher corporate tax rates can discourage FDI, making host countries less attractive. An Economic Partnership Agreement (EPA) is expected to facilitate cross-border factor movements because it helps countries integrate with the global economy beyond traditional free-trade agreements on goods and services. In addition to institutional factors such as corporate tax rates and EPAs, the analysis controls for quality factors. The corruption and fair competition terms capture attractiveness for investment in a host country. These factors are chosen by referring to Blonigen and Piger (2011) and Eicher et al. (2012), which examined the determinants of FDI. Table 2 presents the sample's summary statistics. Our analysis includes the number of immigrants to control for diaspora effects (or ethnic network externalities). This allows us to
distinguish the effects of tourists from those of immigrants on inward FDI, because recent empirical studies have revealed interactions between migration and FDI (Buch et al., 2006; Foad, 2012; Gheasi et al., 2013; Javorcik et al., 2011; Kugler and Rapoport, 2007; Simone and Manchin, 2012). In the due course, the analysis devises the treatment of Chinese and Korean network externalities. Although the tourist data are different among mainland Chinese, Taiwanese, and Hong Kong citizens, and between North and South Koreans, the immigrant data simply classify them as Chinese or Korean. To fill the gap, the analysis uses the number of Chinese registrations for all of mainland China, Taiwan, and Hong Kong. The treatment assumes that the Chinese share a similar culture to a certain degree and that Chinese network externalities apply to the three countries/areas, so that the total number of Chinese immigrants in Japan affects the number of Chinese tourists to Japan. Similarly, the number of Korean registrations is used for the category of South Korea. This treatment requires a less strict assumption compared with the Chinese case and seems reasonable. The “Annual Report of Statistics on Legal Migrants” shows that the number of North Korean entrants to Japan is negligible compared with those from South Korea, and the most recent entrants from North Korea to Japan are “special permanent residents” (the category is designed for Koreans, Taiwanese, and their offspring who had been living in Japan before September 2, 1945, when Japan signed the Instrument of Surrender) with reentry permits. The distinction of either North or South is not so crucial for registered Koreans who came to Japan before World War II, because North Korea was only established in 1948. We begin to estimate tourism–FDI relationships using the system GMM estimator, where FDI, tourism, immigration, and GDP are treated as endogenous. The treatment accounts not only for simultaneous
Please cite this article as: Tomohara, A., Japan's tourism-led foreign direct investment inflows: An empirical study, Econ. Model. (2015), http:// dx.doi.org/10.1016/j.econmod.2015.09.024
4
A. Tomohara / Economic Modelling xxx (2015) xxx–xxx
Table 2 Summary statistics. Variables
Units of measurement
Mean
Inward FDI Tourists GDP Immigration: type 1 Immigration: type 2 Distances Corruption EPA
Constant 2005 million US$ Logarithm of the number of foreign tourists Constant 2005 million US$ Logarithm of the number of immigrants Logarithm of the number of immigrants Kilometers Indicators ranging from −2.5 to 2.5 1 for EPA taking effect and 0 otherwise
Fair competition Tax rates GDP sum GDP diff squared Factor diff GDP diff ∗ Factor diff
Indicators ranging from 0 to 10 Rates ranging from 0 to 1 Constant 2005 million US$ Constant 2005 million US$ Constant 2005 million US$ Constant 2005 million US$
208.30 1371.51 10.07 2.07 4,502,692 175,722 9.04 2.62 8.94 2.51 7604.48 3398.57 1.25 0.22 Frequency for each EPA dummy 0: 90.6% 5.89 0.58 0.41 0.03 5,732,204 2,319,807 2.E + 13 1.E + 13 −10,102.78 18,203.02 4.E + 10 7.E + 10
Standard deviation
Maximum
Minimum
14,473.11 14.44 4,751,194 13.51 13.43 17,693.20 1.61
−8091.72 4.16 4,221,408 2.48 2.48 1156.67 0.86
1: 6.75 0.50 18,200,000 8.E + 13 50,531.43 2.E + 11
9.5% 4.84 0.40 4,248,272 4.E + 10 −36,321.84 −2.E + 11
Note: Immigration (type 1): The total number of Chinese immigrants is used to control for ethnic network externalities. Immigration (type 2): Taiwanese immigrants are distinguished from other Chinese immigrants. GDP diff stands for differences in real GDP between a source country and a host country. Factor diff is used as a proxy for differences in relative factor endowments in the two countries.
decisions on factor mobility regarding FDI and international tourism, but also for their interaction with immigration and national output. Although a few recent studies have examined the interaction between FDI and immigration, the topic of tourism–FDI relationships given immigration has not received attention in the literature and is thus one of the primary contributions of our study. Another advantage of the approach relates to the selection of instrumental variables. It is often not easy to find relevant instrumental variables. The system GMM enables us to use the information within a dataset without searching for instruments that may not be independent variables in the model. In our analysis, the endogenous variables lagged by two or three periods are used as instruments for difference equations, and their single-lag first differences are
used for level equations. We apply a two-step estimator with robust standard errors, which corrects for finite sample biases (Windmeijer, 2005). Roodman (2009) describes the modest superiority of the twostep estimation with corrected errors to cluster-robust one-step estimation by referring to Windmeijer's simulation.
3. Results The results of the analysis are presented in Table 3. Columns 1–2 present the benchmark results. For each variable, the first row shows the estimated coefficients and the second row in parentheses shows
Table 3 Results of tourism effects. Benchmark Independent variables
L1. Inward FDI Tourists GDP Immigration Fair competition
Knowledge capital
Inward FDI
Inward FDI
(1)
(2)
(3)
(4)
0.376 (0.003)⁎⁎⁎ 109.523 (32.243)⁎⁎⁎ 2.42E−04 (0.067E−03)⁎⁎⁎ −54.421 (27.947)⁎⁎
0.377 (0.020)⁎⁎⁎ 121.672 (22.206)⁎⁎⁎ 0.178E−03 (0.086E−03)⁎⁎ −52.869 (11.507)⁎⁎⁎
0.382 (0.034)⁎⁎⁎ 107.556 (29.637)⁎⁎⁎ 0.235E−03 (0.063E −03)⁎⁎⁎ −56.019 (8.305)⁎⁎⁎
0.373 (0.022)⁎⁎⁎
L1. Inward FDI
117.827 (25.557)⁎⁎⁎ 0.221E−03 (0.070E−03)⁎⁎⁎ −53.463⁎⁎⁎ (12.399)⁎⁎⁎
Tourists
304.348 (35.008)⁎⁎⁎
259.772 (57.603)⁎⁎⁎
289.142 (43.040)⁎⁎⁎
304.098 (29.613)⁎⁎⁎
GDP sum GDP diff squared Factor diff GDP diff ∗ Factor diff
Distances Corruption Tax rates EPA
AR(1) AR(2) No. of observation
0.02 (0.007)⁎⁎⁎ −1306.66 (114.834)⁎⁎⁎
0.026 (0.013)⁎⁎ −1168.075 (224.746)⁎⁎⁎
0.020 (0.007)⁎⁎⁎ −1255.479 (125.185)⁎⁎⁎
0.027 (0.013)⁎⁎ −1334.669 (120.722)⁎⁎⁎
−4542.254 (528.98)⁎⁎⁎ −109.061 229.988
−4160.822 (675.589)⁎⁎⁎
−4303.582 (635.867)⁎⁎⁎ −128.486 206.659
−4632.742 (444.092)⁎⁎⁎
0.072 0.161 319
0.072 0.161 319
0.068 0.162 319
0.072 0.160 319
Distances Corruption Tax rates EPA
AR(1) AR(2) No. of observation
(5)
(6)
0.276 (0.009)⁎⁎⁎ 121.325 (37.074)⁎⁎ −0.11E−03 (2.72E−05)⁎⁎⁎
0.278 (0.010)⁎⁎⁎ 120.987 (37.687)⁎⁎⁎ −0.110E−03 (2.74E−05)⁎⁎⁎
8.17E−11 (6.01E−12)⁎⁎⁎ −0.064 (0.014)⁎⁎⁎ −1.41E−08 (3.04E−09)⁎⁎⁎ 0.119 (0.035)⁎⁎⁎
8.06E−11 (6.17E−12)⁎⁎⁎ −0.062 (0.013)⁎⁎⁎ −1.43E−08 (2.15E−09)⁎⁎⁎ 0.108 (0.037)⁎⁎⁎
−867.473 (110.954)⁎⁎⁎ −3912.430 (508.908)⁎⁎⁎ 119.383 115.908
−807.251 (122.168)⁎⁎⁎ −3786.713 (636.529)⁎⁎⁎
0.077 0.142 319
0.076 0.144 319
Note: Figures in parentheses indicate standard error. The Arellano–Bond test was conducted to assess autocorrelation. The test for second-order correlation in differences, AR (2), shows no serial correlation in the first-difference disturbances. ⁎⁎⁎ Denotes coefficient significance at 1% level. ⁎⁎ Denotes coefficient significance at 5% level. ⁎ Denotes coefficient significance at 10% level.
Please cite this article as: Tomohara, A., Japan's tourism-led foreign direct investment inflows: An empirical study, Econ. Model. (2015), http:// dx.doi.org/10.1016/j.econmod.2015.09.024
A. Tomohara / Economic Modelling xxx (2015) xxx–xxx
standard errors. Column 1 presents the results obtained with all control variables. Since EPA is estimated to be statistically insignificant, we attempt estimation without the EPA variable (Column 2). We take the logarithm of tourism and immigration variables, which have distributions with a positive skew, so that the distributions of these variables will be more normal. Our results show that international tourism has synergy with inward FDI. The coefficient on inbound foreign tourists is estimated as positive and statistically significant. For example, a one percent increase in foreign tourists into Japan encourages inward FDI into Japan of about $1.22 million (Column 2).4 A favorable interaction between inward FDI and foreign tourism is observed when they share similar policy targets. The enhancement of foreign tourism amplifies the effectiveness of FDI promotion. The results remain valid under different specifications of Chinese network externalities when the sensitivity of network externalities is examined by different treatments of Chinese immigrants (Columns 3 and 4). We try disaggregated data, where Taiwanese immigrants are distinguished from other Chinese immigrants.5 In fact, it does not seem crucial to distinguish Taiwanese immigrants from other Chinese immigrants in capturing diaspora dynamics. The magnitude of coefficient estimates on tourists is similar in Columns 4 and 2 even after adjusting for some differences in Taiwan and mainland China, including Hong Kong. It would be of interest to know whether expanded international tourism mainly increases inward FDI in tourism-related industries, and thus, whether most effects of international tourism on inward FDI are explained by the increased inward FDI in the industry. However, this seems less likely, as the level of FDI inflows into tourism-related industry sectors such as transportation and services is not large compared to the total inward FDI in Japan. Rather, major recipients of inward FDI in Japan are industry sectors such as finance and insurance, communications, wholesale and retail (non-manufacturing sectors), chemicals and pharmaceuticals, and electric machinery (manufacturing sectors).6 We conduct further examination using recent data on inward FDI by country and industry to support our assertion. We define sectors labeled as transportation and services as tourism-related industries and merge the data on tourism-related inward FDI to our original dataset sorted by country and year. The data are restricted, as they are available only from the year 2005 and information is concealed to protect privacy if the inward FDI cell for each country and industry contains less than three reports. There are three cases where tourism-related inward FDI data by country and year are missing because of privacy protection. When inward FDI into the transportation (or services) sector is positive but data on the other sector is concealed, we treat the positive values in one sector as inward FDI into tourism-related industries by assuming that the concealed values are negligible as data are concealed only if the industry contains less than three reports. There are 24 such cases.
4 Princeton University Library, Data and Statistical Services http://dss.princeton.edu/ online_help/stats_packages/stata/log.html (last accessed on July 23, 2015). 5 The government statistics began to distinguish the Taiwanese from the Chinese after the “alien registration system” was abolished and the “new residency management system” was introduced in 2012. Since the data on Taiwanese immigrants is available only for 2012, we apply a ratio of the Taiwanese among all the Chinese in 2012 to split the total Chinese immigrants into Taiwanese immigrants and other Chinese immigrants during the sample period. 6 Because net FDI inflows take negative values, simply taking the share of tourismrelated inward FDI to total inward FDI does not provide the relative importance of tourism-related industry sectors among all industry sectors. Therefore, we take absolute values of inward FDI in all sectors and sum them. Then, we calculate the share of tourism-related inward FDI to total inward FDI for each sample year. The values vary within the range of 0.08%–27.6% during the period 2005–2011. These values are mostly less than 13%, except for 2005, when the transportation sector received large FDI inflows. Data used for the analysis are available at http://www.jetro.go.jp/world/japan/stats/fdi/ (last accessed on July 23, 2015).
5
Table 4 Spillover effects.
Independent variables
L1. Dependent variable Tourists
System GMM
LSDVC
Tourism-related Total FDI FDI
Tourism-related Total FDI FDI
(1)
(2)
(3)
(4)
0.356 (0.020)⁎⁎⁎ 292.418 (68.493)⁎⁎⁎
0.281 (0.155)⁎
0.341 (0.145)⁎⁎ 387.835 (1164.849) 0.002 (0.003) −322.502 (5973.756) −346.176 (1603.951) 302.564 (945.773)
0.047 (0.175) 49.505 (27.608)⁎ GDP −5.E−05 (3.E−05) Immigration −37.809 (19.719)⁎ EPA 22.018 (47.099) Fair competition 14.396 (12.016) AR(1) 0.484 AR(2) 0.415 No. of 106 observation
11.248 (89.771) −7.E−04 −2.E−04 (0.8E−04)⁎⁎⁎ (2.E−04) −149.664 440.201 (54.091)⁎⁎⁎ (525.120) −738.501 −10.819 (238.024)⁎⁎⁎ (106.631) 313.244 −42.301 (67.724)⁎⁎⁎ (93.228) 0.080 0.156 141 105
141
Note: Figures in parentheses indicate standard errors. Since bias-corrected LSDV estimators do not report standard errors, bootstrap standard errors are reported. The Arellano– Bond test was conducted to assess autocorrelation. The test for second-order correlation in differences, AR(2), shows no serial correlation in the first-difference disturbances. ⁎⁎⁎ Denotes coefficient significance at 1% level. ⁎⁎ Denotes coefficient significance at 5% level. ⁎ Denotes coefficient significance at 10% level.
When inward FDI into the transportation (or services) sector equals zero and data on the other sector is concealed, we treat the case as missing data. There are 31 such cases. The results of the analysis are shown in Table 4. The column labeled as tourism-related FDI shows the results, when tourism-related inward FDI replaces total inward FDI in Eq. (1). The column labeled as total FDI replicates the analysis of total FDI under our benchmark model during the period 2005–2011. The coefficient of tourists is estimated to be positive and statistically significant in both columns, but the coefficient in Column 1 is smaller than that in Column 2. This indicates that tourism enhancement may have positive spillovers on inward FDI beyond tourism-related industries. The coefficient in Column 2 is larger because it includes not only the direct effects in tourism-related sectors but also indirect spillover effects in other sectors. Other sectors also enjoyed increased FDI inflows from the same country of origin of the tourists when tourists from the country increased. The results suggest that the two policies of inward FDI promotion and inbound foreign tourism enhancement are mutually inclusive. The robustness of the results is examined by using the bias-corrected least-squares dummy variable (LSDVC) estimator developed by Bruno (2005a,b). The LSDVC estimator is an alternate approach to solve dynamic panel bias.7 Bruno extends the discussions of LSDVC in Kiviet (1995, 1999), and Bun and Kiviet (2003), which is applicable only for balanced panels, to unbalanced panels. The results of the analysis using the LSDVC estimator of Bruno are summarized in the right-hand side of Table 4. They are similar to those obtained using the system GMM: the degree of effects of tourism expansion is estimated to be larger for total FDI than for tourism-related FDI. Although the two
7 In the analysis of dynamic panels for finite time dimension and large number of crosssectional units, the lagged dependent variable's coefficient estimated by least squares dummy variable estimators is biased because the lagged dependent variable is correlated with the fixed effects (Nickell, 1981). GMM estimators are designed to handle the bias (and endogeneity issues) in analyzing panel data with short time dimension and large number of cross-sectional units.
Please cite this article as: Tomohara, A., Japan's tourism-led foreign direct investment inflows: An empirical study, Econ. Model. (2015), http:// dx.doi.org/10.1016/j.econmod.2015.09.024
6
A. Tomohara / Economic Modelling xxx (2015) xxx–xxx
estimation approaches are not comparable exactly, it is interesting to show similar trends regarding tourism effects on FDI.8 The robustness of the analysis is further examined using different specifications of economic theories. We investigate the effects of international tourism in the context of the knowledge capital model, where both vertical and horizontal FDI arise (Markusen, 1997; Markusen et al., 1996). We introduce FDI dynamics, together with tourism effects, into Carr et al. (2001, 2003), which studies the knowledge capital model empirically: 2 FDI it ¼ α þ β1 FDI it−１ ＋β2 GDP it þ GDP jt þ β3 GDP it−GDP jt þ β4 Factor it −Factor jt þ β 5 Factor it −Factor jt GDP it −GDP jt þ β6 Tourit þ X it ρ þ εit where εit ¼ δi þ μ it : GDPit (or GDPjt) represents the real GDP of a source country i (or a host country j) at time t, Factorit is the per capita real GDP, and Xit is a set of control variables related to investment and trade costs. Horizontal FDI increases with increasing similarity in income and relative factor endowments between the source and host countries. Vertical FDI, on the other hand, increases with the dissimilarity of factor endowments between the two countries. Differences in per capita GDP are used as a proxy for differences in relative factor endowments. We take the logarithm of tourism variable, which has a positively skewed distribution. Results support a favorable interaction between inward FDI and international tourism, as obtained from the previous analysis. Columns 5–6 in Table 3 show the results of the analysis using a system GMM. A two-step estimator with robust standard errors is applied by treating FDI, GDP-related terms, and foreign tourists as endogenous. The endogenous variables lagged by two or three periods are used as instruments for difference equations, and their once-lagged first differences are used for level equations. Since EPA is estimated to be statistically insignificant again, we attempt estimation without the EPA variable. The coefficient on foreign tourists is estimated to be positive and statistically significant. Interestingly, the degree of impacts of international tourism on inward FDI in Columns 5–6 is similar to those in Columns 1–4 despite different model specifications. A one percent increase in foreign tourists into Japan encourages inward FDI into Japan of about $1.21 million (Column 6). It is notable to report that a favorable interaction between international tourism and inward FDI is confirmed under different model specifications. Spillover effects of tourism enhancement on FDI are observed.
enhanced tourism exhibits spillovers of FDI beyond tourism-related sectors. This study is intended to investigate a specific policy implication: Two policy instruments—FDI promotion and foreign tourism enhancement—may entail an unexpected virtuous circle, as in the case of Japan. Our analysis reveals a favorable interaction between inward FDI and foreign tourists visiting Japan. Specifically, tourism enhancement has positive spillover effects on inward FDI beyond tourism-related industries. Additionally, the robustness of favorable interactions between tourism and FDI is confirmed under different model specifications. The results suggest that FDI in other industries needs to be included when discussing the effects of tourism enhancement. The inclusion can alter the degree of effects of tourism enhancement from earlier studies. Considering that the literature focuses on only tourism-related FDI and does not include FDI beyond tourism-related industries, previous studies might have underestimated the real effects of tourism enhancement. Our analysis implies that the two policies of inward FDI promotion and inbound foreign tourism enhancement are mutually inclusive, and that one policy amplifies the effectiveness of the other. The results have implications for policy designs and evaluation. Since these two policies are often planned, executed, and evaluated independently under different government jurisdictions, policy coordination between tourism and FDI should be encouraged to attain goals such as revitalization of declining local economies. Additionally, the analysis suggests practical implications to develop an index for evaluating the tourism– FDI policy. The degree of interaction can be used as a proxy when evaluating the effectiveness of policy coordination between tourism and FDI. The current analysis has several potential avenues of extension. One possible extension is to examine if similar implications might apply to situations in other countries. Although the analysis uses Japanese data, a favorable interaction between tourism and FDI might be observed in other countries. If this is the case, another more interesting extension includes examining the factors that influence such interactions, using panel data. The degree of interaction might vary depending on the fundamentals in each country. The usage of panel data would allow us to compare cross-country differences regarding the degree of policy interactions, and to clarify the conditions under which stronger tourism–FDI interactions occur. Such an analysis requires additional data on bilateral FDI and tourism in several countries, and is beyond the scope of the current paper. All of these topics represent potential future lines of research.
4. Conclusions
References
This study analyzes tourism–FDI interactions in the framework of FDI determination. Although the literature focuses on the relationships between tourism and tourism-related FDI, we examine whether
Ahmed, E.M., 2012. Are the FDI inflow spillover effects on Malaysia's economic growth input driven? Econ. Model. 29 (4), 1498–1504. Antonakakis, N., Dragouni, M., Filis, G., 2015. How strong is the linkage between tourism and economic growth in Europe? Econ. Model. 44, 142–155. Arellano, M., Bover, O., 1995. Another look at the instrumental variable estimation of error-components models. J. Econom. 68, 29–52. Blonigen, B.A., Piger, J., 2011. Determinants of foreign direct investment. NBER Working Paper No. 16704. Blundell, R., Bond, S., 1998. Initial conditions and moment restrictions in dynamic panel data models. J. Econom. 87, 115–143. Bruno, G.S.F., 2005a. Approximating the bias of the LSDV estimator for dynamic unbalanced panel data models. Econ. Lett. 87 (3), 361–366. Bruno, G.S.F., 2005b. Estimation and inference in dynamic unbalanced panel-data models with a small number of individuals. Stata J. 5 (4), 473–500. Buch, C.M., Kleinert, J., Toubal, F., 2006. Where enterprises lead, people follow? Links between migration and FDI in Germany. Eur. Econ. Rev. 50, 2017–2036. Bun, M.J.G., Kiviet, J.F., 2003. On the diminishing returns of higher order terms in asymptotic expansions of bias. Econ. Lett. 79 (2), 145–152. Carr, D.L., Markusen, J.R., Maskus, K.E., 2001. Estimating the knowledge-capital model of the multinational enterprise. Am. Econ. Rev. 91 (3), 693–708. Carr, D.L., Markusen, J.R., Maskus, K.E., 2003. Estimating the knowledge-capital model of the multinational enterprise: reply. Am. Econ. Rev. 93 (3), 995–1001. Chou, M.C., 2013. Does tourism development promote economic growth in transition countries? A panel data analysis. Econ. Model. 33, 226–232. Eicher, T.S., Helfman, L., Lenkoski, A., 2012. Robust FDI determinants: Bayesian model averaging in the presence of selection bias. J. Macroecon. 34, 637–651. Foad, H., 2012. FDI and immigration: a regional analysis. Ann. Reg. Sci. 49, 237–259.
8 Both approaches have advantages and limitations. LSDVC estimators have the advantage of being suitable for panel data with smaller number of cross-sectional units. Additionally, Monte Carlo experiments reveal that LSDVC estimators outperform GMM estimators in several cases (Bruno, 2005b; Judson and Owen, 1999). However, LSDVC estimators of Bruno assume strictly exogenous explanatory variables and their analytical standard errors often break down, requiring the use of bootstrap (Bruno, 2005a, 2005b). On the other hand, GMM estimators allow for endogeneity that arise in our model. While GMM estimators assume the absence of second-order serial correlation, our GMM analysis is shown to satisfy the assumption. It is difficult to conclude that one estimator outperforms others in all cases, since previous Monte Carlo experiments examined only a limited number of cases. Our baseline analysis follows that of Judson and Owen (1999), who recommend GMM rather than LSDVC estimators in the case of unbalanced panels with a time dimension of 20 years or less. Moreover, our analysis fits closer to the experimental setting of Judson and Owen (1999) than to that of Bruno (2005a, b). For example, the number of cross-sectional units (29) is slightly larger than 10 and 20, which Bruno (2005a, b) considers to be small in a Monte Carlo analysis.
Please cite this article as: Tomohara, A., Japan's tourism-led foreign direct investment inflows: An empirical study, Econ. Model. (2015), http:// dx.doi.org/10.1016/j.econmod.2015.09.024
A. Tomohara / Economic Modelling xxx (2015) xxx–xxx Gheasi, M., Nijkamp, P., Rietveld, P., 2013. Migration and foreign direct investment: education matters. Ann. Reg. Sci. 51, 73–87. Jalil, A., Mahmood, T., Idrees, M., 2013. Tourism–growth nexus in Pakistan: evidence from ARDL bounds tests. Econ. Model. 35, 185–191. Javorcik, B.S., Özden, Ç., Spatareanu, M., Neagu, C., 2011. Migrant networks and foreign direct investment. J. Dev. Econ. 94 (2), 231–241. Judson, R.A., Owen, A.L., 1999. Estimating dynamic panel data models: a guide for macroeconomists. Econ. Lett. 65 (1), 9–15. Katircioglu, S., 2011. The bounds test to the level relationship and causality between foreign direct investment and international tourism: the case of Turkey. E + M Econ. Manag. 1, 6–13 (E + M Ekonomie A Management). Kaufmann, D., Kraay, A., Mastruzzi, M., 2010. The worldwide governance indicators: methodology and analytical issues. Policy Research Working Paper Series 5430. Kiviet, J.F., 1995. On bias, inconsistency, and efficiency of various estimators in dynamic panel data models. J. Econom. 68 (1), 53–78. Kiviet, J.F., 1999. Expectation of expansions for estimators in a dynamic panel data model: some results for weakly exogenous regressors. In: Hsiao, C., Lahiri, K., Lee, L.F., Pesaran, M.H. (Eds.), Analysis of Panels and Limited Dependent Variable Models. Cambridge University Press, Cambridge, pp. 199–225. Kugler, M., Rapoport, H., 2007. International labor and capital flows: complements or substitutes? Econ. Lett. 94, 155–162. Markusen, J.R., 1997. Trade versus investment liberalization. NBER Working Paper No. 6231. Markusen, J.R., Venables, A.J., Konan, D.E., Zhang, K.H., 1996. A united treatment of horizontal direct investment, vertical direct investment, and the pattern of trade in goods and services. NBER Working Paper No. 5696. Mayer, T., Zignago, S., 2011. Notes on CEPII's distances measures: the GeoDist database. CEPII Working Paper No. 2011-25. Ministry of Land, Infrastructure, Transport and Tourism, Japan, 2003. Global tourism strategies, mimeo. http://www.mlit.go.jp/kisha/kisha02/01/011224_3/011224_3.pdf (last accessed on July 23, 2015). Nickell, S.J., 1981. Biases in dynamic models with fixed effects. Econometrica 49 (6), 1417–1426.
7
Roodman, D., 2009. How to do xtabond2: an introduction to difference and system GMM in Stata. Stata J. 9 (1), 86–136. Samimi, A.J., Sadeghi, S., Sadeghi, S., 2013. The relationship between foreign direct investment and tourism development: evidence from developing countries. Inst. Econ. 5 (2), 59–68. Selvanathan, S., Selvanathan, E.A., Viswanathan, B., 2012. Causality between foreign direct investment and tourism: empirical evidence from India. Tour. Anal. 17 (1), 91–98. Simone, G.D., Manchin, M., 2012. Outward migration and inward FDI: factor mobility between Eastern and Western Europe. Rev. Intern. Econ. 20 (3), 600–615. Tang, S., Selvanathan, E.A., Selvanathan, S., 2007. The relationship between foreign direct investment and tourism: empirical evidence from China. Tour. Econ. 13 (1), 25–39. United Nations Conference on Trade and Development (UNCTAD), 2007. FDI in Tourism: The Development Dimension (New York and Geneva). Windmeijer, F., 2005. A finite sample correction for the variance of linear efficient twostep GMM estimators. J. Econom. 126 (1), 25–51. Akinori Tomohara received his Ph.D. in Economics from Johns Hopkins University. Prior to graduate study, Dr. Tomohara worked for the Government of Japan. His experience includes economic research at Columbia University and faculty positions at Queens College and Graduate Center, City University of New York and at University of Pittsburgh. He was an economist at University of California, Los Angeles. A broad range of research skills, work experience, and contacts earned Dr. Tomohara the distinction of being an economic consultant at the Urban Institute, the World Bank, and the Inter-American Development Bank. Currently, he is a Professor in Aoyama Gakuin University.
Please cite this article as: Tomohara, A., Japan's tourism-led foreign direct investment inflows: An empirical study, Econ. Model. (2015), http:// dx.doi.org/10.1016/j.econmod.2015.09.024