3.1 The Empirical Model
The main purpose of this thesis is to study the determinant of the inbound tourism flows to Taiwan. The estimation model of tourism flows can be constructed from the conventional gravity model of bilateral trade. The gravity-type model of trade is based on the gravity theory in physics, the Newton’s law of Universal Gravitation. Normally, the value of bilateral trade flows would increase with the economic sizes of countries and decrease their distance. Thus, the basic gravity equation can be written as
𝑋𝑖𝑗 = 𝛽0(𝑌𝑖)𝛽1(𝑌𝑗)𝛽2(𝐷𝑖𝑗)𝛽3(𝐴𝑖𝑗)𝛽4𝑢𝑖𝑗, (1)
where 𝑋𝑖𝑗 is the value of export from country i to country j, 𝑌𝑖 and 𝑌𝑗 are the economic sizes (GDP) of country i and country j respectively, 𝐷𝑖𝑗 the distance between country i to country j, and 𝐴𝑖𝑗 any other factor which would affect the value of trade between two countries.
In this study, the estimation model of tourism flows can be specified as 𝑇𝑖𝑗𝑡 = 𝛾0+ 𝛽1𝐺𝐷𝑃_𝑃𝐶𝑖𝑡+ 𝛽2𝐷𝐼𝑆𝑇𝑖𝑗 + 𝛽3𝐼𝑁𝐷𝑖𝑡+ 𝛽4𝑃𝑂𝑃𝑖𝑡
+ 𝛽5𝑇𝑅𝐴𝐷𝐸𝑖𝑗𝑡+𝛽6𝑋𝑅𝑖𝑡+𝛽7𝑈𝑅𝐵_𝑃𝑂𝑃𝑖𝑡+ 𝛽8𝐷𝐸𝐹𝐿𝐴𝑇𝑂𝑅𝑖𝑡+ 𝛽9𝑌𝐸𝐴𝑅𝑡 + 𝜀𝑖𝑗𝑡,
(2)
where 𝐺𝐷𝑃_𝑃𝐶𝑖𝑡 is country i’s GDP, 𝐷𝐼𝑆𝑇𝑖𝑗 is the distance between country i and Taiwan (country j), 𝐼𝑁𝐷𝑖𝑡 is the industry value added (percent of GDP) in country i, 𝑃𝑂𝑃𝑖𝑡 is the total population of country i, 𝑇𝑅𝐴𝐷𝐸𝑖𝑗𝑡 is the value of bilateral trade between Taiwan and country i, 𝑋𝑅𝑖𝑡 is the exchange rate to US dollar of country i,
16
𝑈𝑅𝐵_𝑃𝑂𝑃𝑖𝑡 is the ratio of urban population to total, and 𝐷𝐸𝐹𝐿𝐴𝑇𝑂𝑅𝑖𝑡 is the GDP deflator of country i. The values of estimator variables except for the dummy variables are measured in logarithmic value except for the year dummy 𝑌𝐸𝐴𝑅. The signs of predicted estimation coefficients are listed below. The explanations of variables are shown in Table 3.
β1 > 0: β1 is the coefficient of country i’s per capita GDP. Per capita GDP is usually used as the level of the living standard of a country. Holding all other things constant, people with higher per capita GDP tend to have more leisure and be willing to go abroad. Thus, a country with a higher per capita GDP is expected to have more outbound tourist arrivals to a destination country.
β2 < 0: β2 is the coefficient of the distance between two capital cities. Distance between two countries implies the cost of transportation and tourist travel cost. Holding all other things constant, the closer between two countries is, more tourists outbound to a destination country are expected.
β3 > 0: β3 is the coefficient of industry value in percent of GDP. Holding all other things constant, a country with higher percent in industry is expected to have more tourist arrivals to Taiwan.
β4 > 0: β4 is the coefficient of country i’s population. Holding all other things constant, the more population country i has, the more potential outbound tourists are. Thus, more tourist arrivals to Taiwan are expected.
β5 > 0: β5 is the coefficient of trade volume between two countries. Holding all other things constant, the larger trade volume between two countries is expected to have more tourist flow.
β6 < 0: β6 is the coefficient of country i’s exchange rate against US dollar. Low exchange rate would lead strong purchasing power on international goods.
17
Holding other things constant, the lower exchange rate is expected to have more outbound tourists.
β7 > 0: β7 is the coefficient of the urban population ratio. Holding all other things constant, a country with more urban population would have more potential tourist outbound. Thus, more tourist arrivals to Taiwan are expected.
β8 ≧ 0 or ≦ 0: β8 is the coefficient of country i’s GDP deflator. It shows the rate of price change in the economy. The estimation can be positive or negative.
β9 ≧ 0 or ≦ 0: 𝑌𝐸𝐴𝑅 can be the year dummy or time trend variable. The coefficient estimates can be positive or negative.
Table 3. The Definition of Variables
Variables Definition Source
𝑇𝑖𝑗𝑡 Total tourist arrivals from country i to j. Tourism Bureau, M.O.T.C. of Taiwan 𝐺𝐷𝑃_𝑃𝐶𝑖𝑡 GDP per capita is gross domestic product divided by
mid-year population.
World Bank
𝐷𝐼𝑆𝑇𝑖𝑗 The distance between the capital cities of two countries, expressed in miles.
Website：How far is it?
𝐼𝑁𝐷𝑖𝑡 Industry value added, percent of GDP. World Bank
𝑃𝑂𝑃𝑖𝑡 Total population of country i. World Bank
𝑇𝑅𝐴𝐷𝐸𝑖𝑗𝑡 The total trade between country i and Taiwan. The Bureau of Foreign Trade, M.O.T.C. of Taiwan
𝑋𝑅𝑖𝑡 The average exchange, local currency per US dollar. World Bank 𝑈𝑅𝐵_𝑃𝑂𝑃𝑖𝑡 The ratio of urban population. The data of population is
taken from the World Bank WDI indicators.
World Bank
𝐷𝐸𝐹𝐿𝐴𝑇𝑂𝑅𝑖𝑡 The GDP deflator is the ratio of nominal (current price) GDP to real GDP.
World Bank
𝑌𝐸𝐴𝑅𝑡 Dummy variable of specific year t, or time trend variable
18
3.2 The Data
To study the determinant of the tourism flows between tourist-origin countries and Taiwan, I construct a data set which consists of 32 countries, covering year 1993 to year 2013, with 578 observations. Table 4 demonstrates the 32 countries included in this study. The independent variables include the per capita GDP of tourist-origin countries, population of origin country, urban population in percent of total, distance between Taiwan and origin countries, total volume of between Taiwan and origin country, exchange rate of origin country, industry value in percent of GDP, GDP deflator of origin country.
Table 4. The List of Selected 32 Countries and Regions
Asia America Europe Others
Hong Kong, China Argentina Austria Australia
India Brazil Belgium New Zealand
Indonesia Canada France South Africa
Japan Mexico Germany
Korea, Republic United States Greece
China, People Republic* Italy
Malaysia Netherlands
Philippines Russia*
Singapore Spain
Thailand Sweden
Vietnam* Switzerland
United Kingdom
Note *: The data of Hong Kong include Macao, China, from year 2008 to 2013. The data of Russia and Vietnam cover from year 2010 and year 2012to 2013 respectively.
19