THE IMPACT OF AZERBAIJAN'S NATIONAL ECONOMIC DEVELOPMENT ON TOURISM SECTOR

This article the influence of GDP on travel vouchers in a longand short-term perspective, travel vouchers sold to Azerbaijani citizens for long-term internal journeys, the price of travel vouchers sold to Azerbaijani citizens, the number of beds per person in a day, the number of beds per all citizens in a day, and the number of residential areas per citizens. In addition, it analyzes the impact of GNI on those same items. Unlike other research about the influence of tourism development on GDP, this article examines the reverse impact the impact of GDP on tourism development. In Azerbaijan, the primary source of GDP growth is the export of energy resources. Tourism is not a major contributor to economic development. The problem, therefore, is to find alternative ways to develop tourism based on the economic growth that comes from the export of energy resources. To explain this further, tourism development should rely on supply rather than the demand generated by GDP and GNI growth. As a result, GDP positively affects tourism development while GNI has a negative impact. We recommend considering the development of a stable supply for tourism that is not dependent on the revenues derived by energy exports.


INTRODUCTION
Tourism is thought to be one of the most dynamically developing sectors for the economy of any country in the modern world. The tourism sector is allocated significant investment in Azerbaijan, and sustainable development can add to the country's socio-economic progress. Investment in tourism is reflected in opening new workplaces and establishing tourism entities and infrastructures. The tourism sector also plays an important role in diversifying the regions within Azerbaijan. Competition and rivalry among tourism producers to provide profitable opportunities are high (Mikayilov et al., 2019). At the same time, this competition is a mechanism for regulating a social product (Nazaraliyev, 2019).
Since gaining its independence, Azerbaijan has been developing an open market economy and has extended economic ties with different countries, integrating itself into the international sphere. And thanks somewhat to its ages-old link to the Great Silk Way, Azerbaijan is full of natural, cultural, national, and ethnographic monuments. This, plus its suitable geographical position -nine climate zones, incredible nature, location on the coast of the Caspian Sea, cuisine, and hospitality -are the main factors that influence tourism. Besides, Azerbaijan has announced that tourism is crucial to developing the non-oil sector.
There is a wide range of opportunities to develop tourism facilities (village, health, ecology, culture, social, commercial sport, religion, etc.). In that regard, Azerbaijan has a lot of aims to develop tourism: research for tourism resources; surveys to determine the demand for tourism; the establishment of necessary services; the extension of health centers; building more tourism facilities such as hotels; and increasing types of tours. The main issue is to turn the country's economy into a sustainable, effective, competitive, and developed one. There is a rising tendency in this direction.
Natural resources, many types of climate zones, and rich cultural and historical heritance provide an opportunity to create a tourism product. In the last 10 years, reforms, state programs, and legislation prove that tourism is one of the main sectors. As a result of these measures, Azerbaijan tourism has penetrated into a new phase and the country is going to be recognized as a tourism-oriented country; integration into the global tourism market is developing every year.
In recent years, Azerbaijan's tourism potential, including information about infrastructure, public transport, tourism products, companies and hotels, were improved for internal and external tourists, providing information about and a database for tourism. Much of the information was targeted at German and Russian tourist-oriented websites.
Many actions have been taken for the development of the Azerbaijan tourism sector. New hotels, relaxation centers were built, and visa terms were simplified. Focusing on its beautiful nature, mineral water resources, the only health unit in the world with oil (Naftalan oil), ancient monuments, culture, national cuisine and other factors, the goal is to provide a good and safe rest in Azerbaijan that will appeal to all types of tourism, including business, sport, health, hunting, exotic, beach and winter tourism. And, due to COVID-19, online work is expected to continue to be popular so that people will have more time for leisure activities, further boosting tourism. It is noteworthy that, unlike other foreign countries, reforms in Azerbaijan tourism caused an increase in quality with a price decrease.
This article includes the following sections. The Introduction presents the urgency and importance of the topic related to the tourism sector, which suffers from the wave of COVID-19.

LITERATURE REVIEW
In modern times, the tourism economy and its role in overall economic development were of great interest to economists (Madlberger, 2014). Thus, the income generated from tourism undeniably improves the country's balance of payments and increases the level of employment and budget revenues (Khan et al., 1990;Durbarry, 2002). Although publications on the various economic effects of tourism and the cause-and-effect relationship between tourism and economic growth have long been known, such studies have grown relatively rapidly since 2002. As such, Kim et al. (2006) concluded that theoretical models examining the cause-andeffect relationship between tourism and economic growth are a product of recent years. Lanza and Pigliaru (2000), Balager and Cantavella Jorda (2002) were the first to study the relationship between tourism and economic growth empirically. Since then, publications have increased significantly (Croes and Manuel, 2008;Muhammad, 2013;Khaled, 2020). In addition to the studies mentioned above that analyze the main cause-and-effect relationship between growth and tourism activities, many other aspects of tourism have been studied (Lucas 1988;Lanza et al.,2003;Algieri, 2006;Smykova, 2015;McKenzie, 2014), including the analysis of the characteristics of the tourist flow of independent tourists (Chun-Yan and Hyung-Ho Kim, 2020), econometric analysis of tourist demand (Bayramli and Aliyev, 2020), the sustainable development of the tourism sector (Khalid Mohammad Hasan et al., 2020), ecotourism (Tai-GiAn and Lim-Soo Shin, 2020), empirical research on sports tourism service providers (Vanessa et al., 2020), instead of tourism in households (Dwi et al., 2019), environmental capital investments and research on the relationship between tourism (Mohammad et al., 2019). T→ PA In addition to the research mentioned above, we would claim that although there are several research studies about the impact of tourism on economic development -Garrigos-Simon et al.,
URT Before evaluating regression equations, the stationarity of the variables using URT has to be determined. This is because when estimating the relationship between two or more variables using regression analysis, it is necessary to have stability in the variables over time. For a time series variable to be stationary, its probability distribution must be the same for each time interval taken.
To make the stationary test results more reliable, three different single root tests are performed, both with and without a trend. The single root tests used are: (ADF), (PP), and (KPSS).

ARDLBT
It should be noted that the ARDL method is more valuable than other cointegration approaches and offers reliable results for small samples (Abbasov and Aliyev, 2018).
ARDL boundary test models can be expressed as follows (Equations (1-6)): The Impact of Azerbaijan's National Economic Development… Sugra Ingilab Humbatova et al Here logarithmic functions, 0 , is a constant quantity, -indicates a white noise error.
The ARDL error correction model (ECM) is based on the following short-term relationships between variables (Equations (7-12)): The ARDLBT approach examines whether there is a cointegration relationship between the variables after the is established. The ARDL bounds-testing cointegration method uses the Wald test ( −stat) on i to indicate the existence of long-term cointegration between selected variables, which checks whether there is a cointegration relationship ( 0 : 1 = 2 = 3 = 0; 1 : 1 ≠ 2 = 3 ≠ 0). An alternative hypothesis is the cointegration relationship between variables. Pesaran et al., (2001), − proposed two types of boundaries based on criterion statistics (i.e., upper bound and lower bound). If the estimated value of the −criterion is below the lower limit, it means that there is no significant long-term relationship between the variables. Furthermore, if the predicted value of the − criterion is higher than the upper limit, this indicates the existence of a long-term relationship between the variables. However, if the calculated statistics of the −test are within limits, this means that the results are uncertain.

Engel-Granger cointegration test
The Engel-Granger (EG) (Engle and Granger, 1987) cointegration test allows the verification of the existence of a long-term relationship and investigates the short-term relationship, and determines the direction of the relationship between the related variables. In the first stage of the EG cointegration test, the regression equation is evaluated for variables.
Thus, for the three variables given in our example (Equations (13-18)): Table 3. Equations for the three variables.
Here 0 , 1 , 2 , 3 , 1 , 1 , 3 , 4 , 5 express the coefficients; p is the optimal delay size, and ε is the white noise error of the model and the direction of the relationship between the variables. In the first stage of the EG cointegration test, the regression equation is evaluated for variables. EG shows that if there is cointegration between variables, this dependence should also be evaluated. If the cointegration relationship is stable (the ), −1 coefficients 1 , 1 , 3 , 4 , 5 and 6 should be negative and statistically significant. Is usually rated between -1 and 0.
The following cause-and-effect relationships can be tested using equations 7-12.
The Granger cause-and-effect relationship for the short run is evaluated using −statistical or − square statistical values by checking the statistical significance of the coefficients of all delayed first-order differences (all ∆ − and ∆ − ) together for each free variable (zero hypotheses: 0 : 2 = 3 = 0, = 1 … ). The rejection of the zero hypothesis suggests that GNI and GDP have short-term effects on LCTV, LCTVACAC, LCTVACOC, LBND, and LACCBND.
Using the t-test to check the Granger causeand-effect relationship for the long run, the statistical significance of the coefficient −1 is checked. The zero hypothesis for this ( 0 : 1 = 0, 1 = 0, 3 = 0, 4 = 0, 5 = 0 və 6 = 0 ) needs to be tested. If, as a result, the null hypothesis is rejected, this long-run period shows that deviations from the equilibrium state affect the dependent variable and will return to the equilibrium state over time.
A strong cause-and-effect relationship is, in fact, both a short-term and a long-term cause-andeffect relationship. In other words, using the − statistic or − square statistical values through the Wald test as a zero hypothesis for each variable taken hypotheses are tested.

Diagnostics
This study uses the Breusch-Godfrey LM test, the heteroscedasticity test, the Breusch-Pagan-Godfrey test, and the Autoregressive Conditional Heteroskedasticity test ARCH and Ramsey RESET Test (statistical) to check the stability of the ARDL model. The J-B Normality test is used to check the normal distribution of the white noise error. The CUSUM and CUSUMSQ tests are also used to investigate the stability of the ARDL model.

Unit Root Tests Results
According to the ADF test in the With Intercept According to the KPSS test in the With Intercept only model, all variables are I(0). In addition, LCTVACAC is also I(1). In the With Intercept & Trend model, all variables of the same year are I(0). In addition, LCTVACAC is also I(1).
The ADF, PP, and KPSS unit root test evaluation results suggest that the ARDL method and the ARDL boundary-test approach can be used to evaluate short-term and long-term associations between variables.   Table 5. Related to ARDL, but in the second equation Restricted Constant and No Trend: -statistic-Model 1 (4.642194**), Model 3 (4.826476**), Model 4 (7.773800***), Model 5 (3.685004), Model 6 (6.915941***) and Unrestricted Constant and Restricted Trend: -statistic-Model 2 (6.709281***). Note: *** ≤ 1%, ** ≤ 5% and * ≤ 10%. Table 7 presents the results of the long-term and short-term approaches of ARDL. The evaluation results of the ARDL model show that in Model 1, GDP has a positive and significant effect on CTV in both long-term and short-term perspectives. Thus, a 1% increase in GDP increases CTV by 12.57% and 8.30%, respectively. In Model 2, GDP has a negative and insignificant impact on the CTVACAC in the long run and a positive and negligible impact in the short term (-7.38% and 0.99%). In Model 3, GDP has a positive and significant impact (10.03% and 6.38%) on CTVACOC both long and short term. In Model 4, 5, and 6, the GDP as BND, ACCBND, and CCNPP has a positive and insignificant effect in the long term and the short term (4.63% and 0.42%; 5.06% and 0.77%, respectively; 31.32% and 0.18%).
The situation is entirely different with GNI, however. In Model 1, GNI has a negative and significant effect on CTV in both the long term and the short term. A 1% increase in GNI reduces the CTV by 11.51% and 7.60%, respectively. In Model 2, GNI based on the CTVACAC has a positive and insignificant impact, in the long run, a negative and negligible effect in the short term (6.84% and -0.82%). In Model 3, GNI has a negative and significant (-8.83%) long-term (-8.83%) and insignificant (-5.61%) impact on CTVACOC. In Model 4, 5, and 6, the GNI as the BND, ACCBND, and CCNPP has a negative and insignificant impact in the long term as well as in the short term (-6.12% and -0.56%; -5.61% and -0.85%, respectively; -46.23% and -0.23%). Note: *** ≤ 1%, ** ≤ 5% and * ≤ 10%.

CONCLUSION AND RECOMMENDATION
To summarize the whole article, we come to these conclusions: GDP affects positively and significantly travel voucher price in a long-and short-term perspective, impacts negatively and little on travel vouchers sold to Azerbaijani citizens for long-term internal journeys, but influences positively in the short term. In addition, GDP has a positive effect either in the long term or short-term perspective on travel voucher's price sold to Azerbaijani citizens, the number of beds per person in a day, the number of beds per all citizens in a day, and the number of people in total.
The situation regarding gross national income is entirely different, however. GNI affects negatively yet significantly travel voucher price in a long-and short-term perspective impacts positively yet little on travel vouchers sold to Azerbaijani citizens for long-term internal journeys, influences negatively yet significantly on travel voucher price sold to Azerbaijani citizens for external journey in a short-term period; oppositely, it affects less in the long term. In addition, gross national income affects negatively and little to the number of beds per person in a day and the number of beds per all citizens in a day both in the long and short term.
We concluded that proper state policy is a must to ensure the sustainable development of tourism, which has a tremendous reciprocal impact on national economic development. From this perspective, Azerbaijan has determined the tourism sector as one of the leading industries besides oil. In fact, state programs approve this trend by adopting relevant laws and acts.
The research does have some limitations: the topic is new to Azerbaijan; data is only given by years; tourism potential is not fully utilized; freed areas in Azerbaijan can play an important role in tourism, and the pandemic is still in progress. All of these make the interpretation of the results difficult.