The relationship between tourism and equality in income distribution in developed and developing countries: An application of Panel VAR model in income in An of

The purpose of this study was to examine the relationship between the tourism and equitable distribution of income for the developed and developing countries using Panel VAR. The results indicate that economic growth and the first lag (one-year lag) of international tourism income negatively affect the poverty index in developed and developing countries. Impulses due to economic growth and poverty index have the greatest effect on international tourism income so that its effect remains after 10 periods permanently. The relationship between tourism and equality countries: application Panel


Introduction
Poverty is a multidimensional phenomenon that influences the quality of life in several ways rather than other phenomena. It damages the human growth, limits the human development, and ultimately affects the well-being of families (Hooshmand et al, 2015). One of the appropri-ate strategies for reducing poverty and unemployment is to develop tourism. Sharpley and Naidoo (2010), Jiang, DeLacy, Mkiramweni, and Harrison (2011), Croes (2014), Li, Chen, Li, and Goh (2016), showed that tourism has a significant positive impact on poverty reduction and income equality. On the other hand, Seckelmann (2002), Tosun, Timothy and Ozturk (2003), Marcouiller, Kim, and Deller (2004), Lee and O'Leary (2008), Lee (2009), Alam and Reddy Paramati (2016) found negative impact of tourism development on income equality. Manyara and Jones (2007), Scheyvens and Momsen (2008) showed no significant relationship between tourism and the fair distribution of incomes. Considering the variety of the results, this research seeks to assess tourism's impact on the fair distribution of incomes in the two groups of developed and developing countries. The main objective of the paper is to examine the impact of international tourism revenues on the poverty indicators in developing and developed countries by Panel VAR methods. Guris, Akay, Zeytinoglu, Sacilidi and Sadic (2016), Góes (2016), Lin and Zhu (2017), Attinasi and Metelli (2017), Ouyang and Li (2018) and Jouida (2018), used the Panel VAR model in various research. The paper also evaluates the impact of trade, inflation, economic growth, and unemployment on poverty in these two groups.

Methodology
Panel VAR is used as a multivariate model to estimate the relationship between tourism and the equitable distribution of income in developing (Brazil, Algeria, Egypt, Indonesia, India, Iran), and developed countries (Austria, Canada, Germany, Spanish, France, U.K. and United States) in 1999-2015. The selection criteria for countries were based on the data access in both groups.

HPI=F(INF,GDP,UN,TR,TOUR)
HPI represents the poverty index, which is a function of INF, GDP, UN, TR, and TOUR. INF, GDP, UN, TR, and TOUR are inflation, economic growth, unemployment, trade index, and international tourism revenue to GDP ratio, respectively.
The Panel VAR was used to estimate coefficients and extract Impulse Response Functions. Panel VAR model is as follows: Where, Xit is the dependent variable vector. ΓL is a polynomial matrix with lag as a dependent variable defined as The optimum lag is a one-year lag. Ui and εit are fixed effects and error vectors, respectively.

Results and Discussion
Panel VAR method includes 5 stages. First, panel unit root testing was used. Im, Pesaran and Shin (IPS) Test was applied to examine the stationary of variables (Table 1). Second, data were evaluated by panel cointegration. Pedroni Panel cointegration test was used to examine the long-term cointegration of variables (Table 2). Regarding the results of Pedroni panel cointegration, the variables were significant at 1% for the groups.   Third, the optimal lag was selected to explain the model. Table 3 illustrates the duration of optimal lag for the groups. According to Table  3, Schwartz statistic was used to protect the degree of freedom and data. The number of optimal lags equals 1 for the groups.
Fourth, the relationship between variables was considered. Finally, the long-term relationship for convergent panels was extracted through vector autoregression (VAR). Table 4 shows the results of the developed countries estimation using the Panel Var model.
The economic growth with one lag, the poverty index with one lag and international tourism revenue to GDP with one lag affect the poverty index and the coefficients of these variables are -0.274, 0.977, -0.4 respectively. The unemployment rate is the last variable that affects the poverty index as much as 0.105. Table 5 shows the estimation results of the developing countries using the Panel VAR model.
The economic growth with one lag and the international tourism revenue to GDP with one lag decrease poverty index as much as 0.48, 0.03 respectively. The poverty index with one lag and the unemployment rate had a significant and positive effect on the poverty index as much as 0.904, 0.309 respectively in developing countries.
The dynamic interactions among variables and Variance analysis were conducted. Impulse Response functions show the response of endogenous variable to the effects of another variable (Figures 1, 2).

Figure 1. Tourism income response to explanatory variables for developed countries
The result showed that the response of tourism income to economic growth, poverty index, and inflation are incremental, and the effect is stable after 10 periods in developed countries. Considering the Variance analysis for two groups of countries, Impulse Response functions show the response of endogenous variable to the effects of another variable while variance analysis measures each effect on endogenous variable variance.
In Table 6, 88.371% of tourism incomes are explained by the same variable and 10.789% by economic growth in period 1. 83.80% of Figure 2. Tourism income response to the explanatory variable for developing countries error variance for tourism incomes is explained by the same variable, 15.006% by economic growth and 1% by other variables in period 2. According to Table 7, 95.734% of tourism incomes are explained by the same variable, followed by economic growth with 3.724%.

Conclusion, limits and perspectives
This paper examined the relationship between tourism and equality in income distribution in developed and developing countries using PanelVar. The estimation results showed that the economic growth rate and international income of tourism had a significant and positive  (2015)).
Further studies could be focused on collected data of poverty index and other indicators that effect on tourism. Given the tourism policy implications, we suggest to adopt appropriate policies to improve the business environment, the income distribution system to support the poor, the general livelihoods, and reduce income inequality in both groups. It is also proposed to expand and strengthen tourism infrastructure to increase the number of tourists. Therefore, making benefit for indigenous people is required to deal with expanding poverty, given the undeniable role of tourism in poverty alleviation.