3.1. Econometric Model and Estimation Method
In this study, the baseline model is as follows:
where
i is the home country index,
t is the time index,
α is the unknown parameter to be estimated,
FDI_T/GDP is the ratio of
FDI inflows in tourism to
GDP,
FDI_T/GDP (−1) is the lagged ratio of
FDI inflows in tourism to
GDP,
TERR is the measure of international terrorism,
GDPgrowth is
GDP growth rate,
INTARR is the number of international tourism arrivals,
ηi is the unobserved country-specific effect term and
εi,t is the usual error term.
The FDI share of tourism in GDP (FDI_T/GDP) was included as a dependent variable of the model. As independent model variables, the former level of FDI contribution to tourism in GDP (FDI_T/GDP (−1)), International Terrorism (INTTERR), GDP growth rate and the number of international tourist arrivals (INTARRs) were included. All of the variables are in logarithm forms.
In order to obtain more reliable research results, the model subsequently included the specific control variables:
where
POLSTAB is the index of Political Stability and Absence of Violence,
CORR is the index of Control of Corruption and
DOINGBUS is the Starting a business index, part of a Doing business index.
This research employed the System-Generalized Method of Moments (SYS-GMM) state-of-the-art econometric estimation method [
45]. Early research of similar models used standard OLS techniques that are susceptible to the well-known spurious regression problem [
46]. According to [
47], “the pooled OLS estimator does not deal with either country-specific effects across the panel or endogeneity bias”. Ref. [
48] in 1982 introduced GMM. GMM is commonly used to study the dynamics of adjustment in samples with relatively large cross-sections and short time periods. The standard GMM estimator controls for measurement errors and endogeneity. On the other hand, it does not account for unobservable country-specific effects and can be vulnerable to inaccuracy due to small-sample bias.
The SYS-GMM estimator is developed by [
49,
50]. It produces more efficient and precise estimates compared to dynamic GMM by improving precision and reducing the finite sample bias [
51] by allowing for more instruments [
52]. This estimator resolves some of the small-sample biases of the standard GMM estimator without enforcing particularly strong assumptions [
53]. This estimator creates a system of two equations; the first equation is differenced while the second one remains in levels [
54]. Also, in addition to the corrections for serial correlations, measurement error and endogeneity also accounted for the underlying data dynamics [
55]. The consistency of the SYS-GMM estimator depends on the validity of the instruments. To address this issue, two diagnostic tests were used to test the validity of the instruments, the Sargan test and the Hansen test.
3.2. Empirical Data and Sample Selection
The research was carried out based on annual time series for the period 2000 to 2016. The panel of research countries was made up of the following 50 countries: Australia, Austria, Belgium, Bosnia and Herzegovina, Bulgaria, Chile, China, Croatia, Czech Republic, Denmark, Estonia, Finland, France, Germany, Greece, Hong Kong, Hungary, Iceland, India, Ireland, Israel, Italy, Kazakhstan, Korea, Kosovo, Latvia, Lithuania, Luxembourg, Macedonia, Mauritius, Mexico, Morocco, Mozambique, Netherland, Norway, Poland, Portugal, Russia, Serbia, Slovak Republic, Slovenia, Spain, Sweden, Switzerland, Thailand, Tunisia, Turkey, United Kingdom, United States and Vietnam. The selection of this sample was made due to data availability of a dependent variable
FDI inflow in tourism. However, this sample is sufficiently representative because the bulk of
FDI in tourism is geared towards developed countries, i.e., 85–90 percent of TNC hotels are located in developed countries [
56]. Out of a total of 145 destination countries of
FDI in tourism, the top five account for 30.2% of greenfield
FDI in tourism projects, and the top 10 account for 45% of the project. This panel includes almost all of the top 10 host countries (except UAE). Additionally, although the risk of terrorism is by no means absent in developing countries, it appears to be primarily associated with industrialized countries [
57]. The representativity of the sample is supported by the fact that it includes the top 10 world destinations as far as international tourist arrivals and international tourism receipts [
2] are concerned. All 50 countries in the panel make up 72% of total international tourism receipts (see
Appendix A). Summary statistics can be found in
Appendix B.
The variable
FDI in tourism (
FDI_T) has been obtained from [
58,
59,
60]. The variable is employed in millions of US dollars.
Data for terrorism derives from The Global Terrorism Database (GTD) [
61]. This research used international total casualties as a terrorism variable. Total casualties include both injuries and fatalities (killed). The GTD database does not offer per se a column distinguishing domestic and international terrorist incidents. Decomposition was done following established methodology by [
62,
63]. Decomposition started with [
62] five-step procedure. After this procedure had been exhausted, the process continued with the known perpetrator group identity parameter used by [
63].
Starting a Business indicator was obtained from the Doing Business database [
64]. All the other variables were obtained from [
65].
Hypothesis 1. The level of FDI in tourism with a time lag of one period significantly affects the future FDI inflows in tourism.
The movement of one company may initiate a chain reaction of countermeasures at the domestic and international level by rivals who want to protect their positions [
66]. In oligopolistic industries, companies will often imitate interaction because alternatives to imitation following the strategy of differentiation may prove to be costly and dangerous [
67]. The related concept is so called “herding” [
68]. “Herding” is essentially unscrupulous behavior based on the security of numbers; as long as everyone else behaves unconsciously, the probability of serious consequences for a particular company is low. Such behavior does not necessarily have to be regarded as irrational, i.e., “it is not that they are blind—this is simply the logical result of competitive processes in an oligopolistic industry” [
69]. In the context of
FDI, the idea of this perspective is that transnational corporations are largely doing what other companies are doing in the organization field when there is a high level of uncertainty [
70].
Hypothesis 2. Terrorism has no significant impact on FDI in tourism.
After the terrorist attack in the US in 2001, terrorism became a source of concern for international investors and has entered the scene as a form of political risk [
71,
72]. The impact of political risk varies depending on which industry
FDI is focused on. The research focused on sectoral differences is very modest and points to the specificity of particular industries [
73,
74]. Ref. [
75], in one of the early reviews of research on the effect of political risk, concluded that empirical evidence is inconsistent and has mixed results related to the influence of political instability on
FDI stocks or flows. Secondly, back to 1983, terrorism generally did not significantly affect
FDI, although it had significant localized impacts in places such as the Basque region in Spain or Northern Ireland [
31]. Finally, all further research leaves a shadow on the existence of such a postulate [
33,
76]. Ref. [
41] explored the link between terrorism and
FDI with the example of Egypt and amply warned on exaggeration connected with the negative impacts of terrorism on
FDI in tourism. Finally, ref. [
44] also points to a lack of existing research related to the
FDI in tourism and terrorism and concluded that the issues are very complex and that the impact of terrorism on
FDI in tourism cannot be generalized. Since tourism demand and tourism supply are closely linked, it is logical to assume that if terrorism does not significantly affect the decline in tourist arrivals, there will be no downturn in tourism supply, i.e.,
FDI in tourism will not decline. The arguments in favour of the set hypothesis are as follows: according to the latest research, tourists are not too concerned about terrorism [
12,
77,
78], after the terrorist attack, tourism had already been recovering for 13 months [
23], the latest UNWTO data show a continuous growth rate of international tourist arrivals [
2], less significant and limited terrorist attacks have little impact on the expected returns of an investment project [
33], out of 39 sectors, tourism is ranked 10th in terms of capital investment [
79] and, above all,
FDI in tourism continues to enter the countries affected by terrorism [
80].
Hypothesis 3. The size of the market is an important determinant of FDI in tourism.
Market size is generally the most important determinant of
FDI inflows. It is also a significant determinant of
FDI in the services market [
81]. The larger the market, the more likely it is that the investor will regain its fixed costs [
82]. The size of the market is considered the most important location factor the investor considers when deciding on
FDI [
83]. The size of the overall economy market is measured by
GDP. When the market size is small compared to other competitors in the country, such a market fails to attract
FDI due to difficulty to achieve the economies of scale [
84]. It should also be noted that the market size does not only apply to the domestic market, however also to the regional market in which the country is located. Corporations locate their
FDI considering the regional context and context of the country as the country’s attractiveness is limited by regional development characteristics [
85]. Consequently, the main objective of regional political development has become attracting
FDI [
86]. The high economic growth affects the
FDI inflows due to increased revenue and the effects of consumption [
87,
88].
Hypothesis 4. The number of tourist arrivals significantly influences FDI in tourism.
Although indicators such as
GDP or
GDP per capita determine the market size, the more relevant measures for tourism would be the propensity to travel within the economy [
89]. The level and the degree of tourism products and tourism development are important because
FDI in tourism is under their influence [
56]. The country will attract foreign investors to tourism if it has an effective tourism marketing strategy and promotional programs that are significantly funded [
90]. Tourist arrivals and
FDI are interconnected. Tourist arrivals are considered the main cause for
FDI in tourism [
91]. A large number of tourist arrivals in the country also indirectly complement the existing market, thus affecting the attraction of
FDI in the hotel industry [
56,
92]. The primary driver of service companies to invest abroad is based on tracking citizens and clients [
93]. A significant number of studies have demonstrated the existence of a causal link between international tourist arrivals and
FDI in tourism [
94,
95,
96,
97,
98,
99].
Hypothesis 5. Political stability and the absence of violence positively influence the FDI in tourism.
Political stability, along with macroeconomic stability, are key factors influencing the location decision of foreign investors. For every foreign investor, each country is the potential destination of its capital. However, given that every investor is a rational investor, one of the most important criteria when selecting a country in which to invest their capital is the investment risk. Generally, as long as the foreign investor believes it can operate profitably without excessive risk for its capital and staff, it will continue to invest. A host country with a high political risk will discourage
FDI inflows into its market since the political volatility harms the profitability of
FDI. The three major forms of political risk discourage
FDI because of damage to their profitability and survival [
100]: nationalization or expropriation of foreign assets (which is rare) and breach of contract (which is much more common) endanger foreign investment; political instability and arbitrary regulation in policies related to
FDI create uncertain investment environments and undermine the profitability of
FDI; and political violence, including terrorist activities, can immediately damage foreign property and discourage productivity in the country for a long time.
Hypothesis 6. The higher level of corruption in the country negatively affects the FDI in tourism.
As corruption is widespread, less investor capital will flow to the country. Corrupt states are less likely to attract
FDI in order to get assistance in the long-term economic development of the state. The amount of corruption in the country that foreign investors want to invest in sis as important as the cost of labor and the tax rate [
101]. Corruption is occurring in countries where government transparency is low. Investors in these countries are either pulling or not investing at all, precisely because of the unstable political environment and inefficient bureaucracy and corruptive actions that ultimately damage the reputation of the investor and his profits. Lower levels of corruption leads to higher productivity of the sector [
102,
103].
Hypothesis 7. The ease of starting a business has a positive impact on FDI in tourism.
Starting a Business measures the paid-in minimum capital requirement, number of procedures, time and cost for a small- to medium-sized limited liability company to start up and formally operate in the economy’s largest business city. The ranking of the Doing Business list indicates the attractiveness of the investment environment, where a higher position on the list means a more attractive investment environment. According to [
104], a better ranking on Doing Business is significantly associated with higher
FDI inflow. In addition, countries with more effective regulations for starting a business have greater benefits from
FDI inflows. According to [
105], one step higher on the Doing Business scale can bring an additional
$44 million in
FDI to the government.