Today, it is becoming increasingly important to correctly plan tourism development strategy, as the global tourism market is recovering from the shock of the COVID-19 pandemic. Correct forecasting and rational planning of tourism development can be a crucial factor in providing a decisive advantage in the global tourism market. Taking into account the impact of major factors is a fundamental condition in planning future tourism development.

There are many fascinating historical sites and attractive exotic places to visit in the heart of Central Asia. Thus, in theory, Uzbekistan has great advantages with regard to becoming a new mainstream tourism destination. However, the tourism infrastructure is not developed in many parts of the country, and there may be many other factors that prevent Uzbekistan from becoming major world tourism attraction [1]. These problems include lack of qualified personnel, lack of tourism infrastructure, high transportation costs and others. Tourism services in Uzbekistan are mainly focused around historical sites. Most tourists visit Uzbekistan for historical sightseeing, and unfortunately, the Republic still should do more to hold tourists for longer time by offering high quality accomodation services, entertaining shows, festivals, qualified medicine, inclusive transportation and other crucial tourism services. Nevertheless, government is taking measures to support investors in building infrastructure, thereby contributing to the development of domestic tourism, which is crucial in the progress of overall tourism. In our view, if the country develops a reasonable plan of tourism development then it has a chance to change the balance of power in the competition for global tourist inflows.

With the emergence of COVID-19 in 2020, tourism and the travel sector faced unprecedented challenges. Uzbekistan’s tourism sector was hit by the effect of strict quarantines, when the whole tourism sphere was essentially frozen. The measures taken by government to support the economy included creating a USD 1 billion Anti-Crisis Fund, which made it possible to lengthen the tax-free period for travel companies and tourism infrastructure. Moreover, the state expanded funding healthcare, covered the cost of quarantines and salary supplements for healthcare workers, assisted affected businesses via subsidies, created additional public works, and extended a moratorium on tax audits as well as delaying tax declarations [2]. Taking these measures enabled the republic to maintain its key tourism infrastructure facilities in working condition.

According to Dwyer, L. and Forsyth, P., the domestic output (level of consumption) is a major factor that drives inbound tourism [9]. Researchers tried to assess the short- and long-term effect of real GDP per capita, which can represent the level of consumption and welfare, on the inbound tourism using the ARDL model. In addition, we forecast inbound tourism demand using the ARIMA model, which showed that the COVID-19 pandemic will continue to exert a negative effect for more than five years. Our study stands out from similar researches on inbound tourism demand by focusing on assessing the impact of factors which shape destination competitiveness on inbound tourism, with extensive empirical analysis using the example of Republic of Uzbekistan. Previous studies rarely used macroeconomic data to assess the inbound tourism demand of Uzbekistan; rather, the priority was on survey analysis [10]. Allaberganov and Preko [10], in their research, found a positive correlation between travel motives and the frequency of visitation of international tourists to Uzbekistan. They used survey data from 563 international tourists; in other words, it was static data taken only for one period. However, it is crucial to estimate the impact of various factors on inbound tourism using dynamic data. Our research partially fills this gap by using data from 21 time series as well as econometric models to assess the impact of four crucial factors on inbound tourism. The purpose of this research was to explore macro-level data that was related to constructing a tourism development strategy able to boost the country’s competitiveness as a tourism destination. At the end of the day, we are able to state that all of the reported results obtained in this research were statistically significant and can be applied in practice. The statistical data used for analysis were taken from the official website of Department of Statistics of Uzbekistan.

2. Literature Review

The use of economic and mathematical methods makes it possible to conduct a qualitative and quantitative analysis of economic phenomena in order to provide a quantitative assessment of the significance of risk and market uncertainty and choose an effective solution. According to [11], three main methods based on empirical studies of tourism forecasting can be distinguished, namely, causal models (econometric or spatial models), time series models, and various qualitative models. Goh and Law [12] identify three types of quantitative forecasting methods: time series models, econometric models, and AI-based models. Time series models require only one data series, where past data is extrapolated into the future patterns. Even though time series models are widely used, these models are hard to interpret as they are not based on any economic theory. Time series models are divided into simple (simple moving average, single exponential smoothing) and advanced (double exponential smoothing, autoregressive moving average, simple structural time series) subcategories. Peng et al. [13] emphasize the use of econometric models for determining the causality structure and evaluate the influence of different variables on future tourism demand. To put it simply, the aim of econometric methods is not extrapolation, it is identifying the group of explanatory factors [14]. Simple regression, gravity models, vector autoregression, error correction models, cointegration, and autoregressive distributed lag models are common types of econometric models [12]. From the beginning of the 2000s models based on artificial intelligence have been used for forecasting purposes in various fields [15]. AI-based models such as artificial neural networks, support vector machines, fuzzy time series, genetic algorithms, and expert systems have proven to be more effective than traditional forecasting methods [16]. Even though the forecasted values seem to be more accurate for AI-based models, it is hard to identify the path taken by the learning process, which is based on adjusting weights of respective neurons (nodes) via synapses [15]. Therefore, AI-based models are usually used with big data, where classification or identification of clusters has greater priority. Identifying clusters can be important in forming hypotheses, which can then be verified by econometric methods. Nowadays many researchers use various forecasting methods in combination in order to reach research goals [17]. The research of Dwyer and Forsyth on assessing inbound tourism is the most notorious and extensive work on the analysis of inbound tourism demand. In their research, they point out domestic consumption as the strongest factor driving tourism demand upwards [9]. Milenkovski et al. [18] analyzed the impact of traffic infrastructure on the inbound tourism in the Republic of Macedonia. They assert the critical impact of security and environment of inbound tourism. In their research, Breda et al. [19] studied the impact of safety and security measures on inbound tourism in China. Their research results show that political and social stability, security, and fashion trends significantly affected inbound tourism demand. Moreover, they point out the impact of outbreaks on structural changes in Chinese tourist agencies. However, the research conducted by Biagi et al. [20] showed that crime could be positively correlated with tourism. According to Sunlu [21], tourism might negatively impact the environment when the number of visitors is greater than the capacity of the tourism destination. Among the potential negative impacts he lists air, water, and land pollution as well as other physical impacts on the ecosystem of a destination.