Testing the efficacy of the economic policy uncertainty index on tourism demand in USMCA: Theory and evidence

Abstract

The global economic outlook is more uncertain than ever before and sensitive to uncertainties related to a variety of economic policies decisions of all stakeholders and governments. These perceived uncertainties may be the culprit in shrinking the size of overall economic activity. Under increasing uncertainties, travel and vacation plans of consumers can be canceled or postponed. Therefore, policy-related economic uncertainties are expected to affect tourism demand beyond well-established economic and noneconomic factors. In this study, we explore the efficacy and the impact of the economic policy uncertainty (EPU) index in predicting the tourism demand on international tourist arrivals (a measure of tourism demand) to the United States from Mexico and Canada over the period of January 1996–September 2017. The findings of the study reveal that EPU is a significant predictor as increases in the EPU index lead to decreases in tourism demand to the United States. Canadian tourists seem to be more sensitive to EPUs. Increases in the EPU index cause them to reduce Canadians’ vacations to the United States proportionally more than the Mexicans. To enhance the explanatory power of current models, the uncertainty can be a theoretically significant construct thus needs to be included when calibrating demand models.

Keywords

economic policy uncertainty (EPU) index real exchange rate tourism demand USMCA

Introduction

It is a stylistic fact that nations’ overall economic activity shrink during financial and economic crises as people reduce their consumption of goods and services. When purchase decisions of leisure services and products that depend on discretionary income are delayed or completely eliminated, consumer tends to focus on their basic needs such as food, water, shelter, and medical care as eloquently postulated by Maslow’s Hierarchy of Needs Theory (1987). Sheldon and Dwyer (2010) and Song and Lin (2010) reveal that people reduce their travel budgets in lieu of essential goods and services during the economic crises. Is tourism an essential or luxury good/service? Harrod (1958), a pioneer in economic growth theory, posits the question whether travel is oligarchic wealth or a redundant luxury. Theuns’ (2014) recent argument on this issue (tourism being necessity or a sheer luxury) is also based on Harrod’s (1958) work. According to a report done by the United Nations World Tourism Organization, international tourist arrivals and international tourism receipts declined by 4% and 6%, respectively, during the global financial–economic crisis of 2008 (UNWTO, 2013). The total demand of all tourism goods and services in the United States decreased by 9.3% (BEA, 2011). Similar downturns in tourism demand and investment were also reported during the Asian financial crisis of 1997–1998 (Henderson, 2006). Accordingly, economic crises affect the worldwide tourism demand negatively (see, e.g. Papatheodorou et al., 2010; Ritchie et al., 2010; Sheldon and Dwyer, 2010) and supply (Alonso and Bremser, 2013; Formica and Uysal, 2006; Sheng and Lo, 2010; Uysal, 1998) like in other sector of the economy.

Keynes (1936), the founder of modern macroeconomic theory, had introduced the construct of economic uncertainty following the economic crises of 1930s. Keynes argued that certainty or uncertainty determine much of the economic behavior and prices of stocks. However, the literature on tourism demand remained mute about this concept for more than seven decades until recent works by McKercher and Hui (2004), Bloom (2009), Chen and Chiou-Wei (2009), and more recently Stylidis and Terzidou (2014), Singh et al. (2019) appeared in respected tourism journals. To date, tourism economists tend to focus on measurements that are readily available because of government mandate or ease of access. Since uncertainty requires the construction of an index mainly consisting of qualitative inputs, this reasoning is understandable but not desirable. The quality and speed of data collection allows researchers to incorporate such constructs with ease when modeling demand. Thus, to increase the explanatory powers of existing models and fill a void in the relevant literature, our study centers around the economic policy uncertainty (EPU) as a new proxy construct of economic downturns that was also recommended by Okumus et al. more than a decade ago (2005).

Many scholars applying a variety of methodologies empirically tested the impacts of economic downturns on tourism demand for different countries. For instance, Smeral (2009) in his seminal piece applied the general autoregressive distributed lag model (ARDL) for Australia, Canada, United States, Japan, and the EU-15 countries and found negative impacts of 2008 global economic crisis on outbound tourism demand for these countries. Similarly, Song and Lin (2010) used the autoregressive distributed lag model (ARDL) for inbound and outbound tourism in Asia between 1980 and 2008. The authors reported that economic crisis negatively affected the inbound and outbound tourism in Asia. Song et al. (2011) used the same methodology for Hong Kong between 2009 and 2015 and found that 2008 global crisis had negative impacts for demand of hotel rooms. Page et al. (2012) and Meng et al. (2015) in two independent studies found that 2008 global crisis has negative impact for inbound tourism demand and country’s overall economy. Moreover, in a recent study, Perles-Ribes et al. (2016) reported that the economic crises between 1958 and 2010 reduced the competitiveness tourism of Spain. Applying Bai and Perron’s (1998) method, Cro and Martins (2017) examine multiple structural breaks for 25 countries and Madeira Island between 1995 and 2004 and found that financial crises do indeed decrease tourism demand for all countries reported in their study.

Extant econometric studies traditionally model variables like the changes in consumers’ income, decrease in general prices-levels at a destination, exchange rates, and marketing by destination management organizations when estimating tourism demand (Bloom, 2009; Chen and Chiou-Wei, 2009; Ghosh, 2018; Gozgor and Ongan, 2017; McKercher and Hui, 2004; Singh et al., 2019; Stylidis and Terzidou, 2014; Tsui et al., 2018). The emergence of regional wars, battle for regional hegemony, rising economic nationalism, and the use and abuse of social media in international politics direct economist to incorporate proxy measures of economic crises as a dummy variable to explain additional error variance in demand modeling. Because of its strong correlation with economic crisis, the “economic uncertainty” (index) as another proxy measure has the potential of enhancing the explanatory power of demand models (Fountas et al., 2018; Istiak and Serletis, 2018; Papatheodorou et al., 2010; Sahinoz and Cosar, 2018; Song et al., 2011).

The EPU index, developed by Baker et al. (2016), contains three main components: The first component extracts newspaper coverage in terms of the volume of news articles related to policy-related economic uncertainty. The second component of the index contains extracts from the reports by the Congressional Budget Office about current and future tax code provisions. The third component includes forecasts by the Federal Reserve Bank of Philadelphia’s about the expected level of inflation, federal expenditures, and state and local expenditures. The EPU index provides wide-range time-series data covering daily as well as yearly expenditures dating back to 1900s for the United States.

Accordingly, the purpose of our study is to examine the efficacy and the impacts of the EPU index on US tourism demand, measured as the number of international tourist arrivals from Mexico and Canada over the period of January 1996–September 2017. We apply unit root- and cointegration-tests with structural breaks in order to examine the efficacy of policy-related economic uncertainties (EPU index) when predicting the tourism demand to the United States. Mexico and Canada are the two largest tourist-generating countries for the US tourism market. About 51.1% of all international tourists who visited the United States in 2017 came from these two neighbors of United States (NTTO, 2018). Moreover, because Canada and Mexico are the two tourist-generating countries with much different economic structures and developmental levels, we expect that travelers from these countries will respond to the changes in the EPU index much differently. Because the literature on this topic is virtually nonexistent, we cannot postulate any directional hypotheses in terms of the expected magnitude of tourism demand to the United States.

Nonetheless, our study is expected to enhance the current knowledge in the area mainly in two major ways: The first is its contribution to the practice and management of tourism by drawing attention of policy makers in the United States to the nature of the US Mexican and Canadian tourists in the wake of economic uncertainty. The two main questions that guide this study: How do potential tourists react to economic uncertainty? Is the tourist market sensitive to economic uncertainty, if yes how do they respond? The second contribution is the use of an innovative index (EPU) that was specifically constructed based on policy-related economic uncertainties. The efficacy of EPU index in this tourism demand study lies in its breadth and depth of the coverage of all current political discussions (especially from the perspective of the United States) about the future of the United States–Mexico–Canada Agreement (USMCA). Therefore, the impacts of the changes in the EPU index on US tourism demand will show the US policy makers how their economic–political discussions on the USCMA affect the US tourism market.

Literature review

The EPU index is widely used to reveal the impacts of uncertainties on some economic and macroeconomic variables such as stock market returns (Antonakakis et al., 2013; Brogaard and Detzel, 2015; Karnizova and Li, 2014), unemployment (Colombo, 2013; Mumtaz and Surico, 2016), housing prices (Aye, 2018; Chow et al., 2017), exchange rates (Dai et al., 2017; Dogru et al., 2019), and oil prices (Arouri et al. 2014; Bekiros et al., 2015). Many scholars used the concept of uncertainty in relation to the economic crises in tourism models as well (see Bloom, 2009; Chen and Chiou-Wei, 2009; Dogru and Sirakaya-Turk, 2018; Ghosh, 2018; Isik and Radulescu, 2017; Işık, Radulescu and Fedajev, 2019; Isik et al., 2018; Isik et al., 2017b; McKercher and Hui, 2004; Stylidis and Terzidou, 2014; Wang, 2009). Furthermore, the EPU index began to use as a new independent variable in tourism demand models. For instance, Gozgor and Ongan (2017) applied the unit root and a cointegration test and found that uncertainty has negative impacts on US tourists’ domestic spending. Similarly, Ongan and Gozgor (2017) applied the same methodology and found that EPU leads to decreases in tourism demand by Japanese tourists to the United States. Madanoglu and Ozdemir (2019) used the generalized estimating equations method for the United States and found that uncertainty has negative impacts on the hotels’ operating performances. Tsui et al. (2018) applied panel data gravity model for New Zealand and found that EPU has statistically significant impact on the country’s business tourism flows. Demir and Ersan (2018) applied multiple regression approach and found that the EPU index in Europe and Turkey has significant negative impacts on the stock returns of Turkish tourism companies. The rising uncertainties worldwide make the uncertainty concept necessary to be considered as an additional determinant besides the most used below traditional determinants in tourism models. Gross national product or gross domestic product (GDP) or in their per capita form and nominal or real exchange rate adjusted by consumer prices index (CPI) of both destination and generator countries are the most used independent variables by many scholars including Lim (1999), Garín-Muñoz and Teodosio (2000), Kulendran and Kenneth (2000), Li et al. (2005), Thompson and Thompson (2010), and Isik et al. (2017a).

Extant literature in tourism demand is replete with studies that illustrate causal relationships between international trade and international tourist arrivals (Ibrahim, 2013) and business travels (Tsui and Fung, 2016). Population at the tourist originating country (Leitao, 2010) or as a proxy for per capita demand models is used in some studies. However, different population segments based on sociodemographics such as age have also been modeled in tourism demand (Glover and Prideaux, 2009; Kapiki, 2012). Ledesma-Rodríguez et al. (2001) included tourism promoting expenditures which may increase the awareness of the destinations. Song et al. (2003) and Ayeh and Lin (2011) find that repeat visits and Word-of-Mouth highly influence tourism demand (Song et al., 2009). While political instability and violence are used as explanatory variables in some demand models (Neumayer, 2004; Seddighi et al., 2001), tourist arrivals, tourism expenditures, and tourism receipts are the most used dependent variables in tourism demand models (Akal, 2004; Au and Law, 2002; Chu, 2011; Coshall, 2005). Other studies use the length of stay (Gokovali et al., 2007; Kim and Moosa, 2005) and tourist nights spent (Gouveia and Rodrigues, 2005; Guizzardi and Mazzocchi, 2010). Furthermore, many different empirical methodologies were applied in tourism demand models over time such as linear regression in early tourism models (Uysal and Crompton, 1984). Ordinary least squares seem to have provided substantial advantages in modeling demand (Payne and Mervar, 2002; Wooldridge, 2010; when data were nonstationary). Isik et al. (2018) used nonlinear ARDL (autoregressive distributed lag) model under the assumption that rising uncertainties create asymmetric (nonlinear) relationships between the variables. Greenidge (2001) used structural time series modeling including time varying components in the regression equation. Song et al. (2003) and Lim and Mcaleer (2001) used error correction models (ECMs) considering short-run dynamic characteristics of tourism demand. Narayan (2004) used ARDL (autoregressive distributed lag) model needing no prior knowledge about the integration properties of the variables. Han et al. (2006) used almost ideal demand system model, derived economic theory, providing a strong theoretical underpinning over single equation models. Kim et al. (2011) used (ARIMA) model explaining a variable’s past and a random disturbance terms statistically. De Vita and Kyaw (2013) used generalized autoregressive conditionally heteroscedastic model. This model provides detailed analysis of volatilities in tourism demand models. Studies of Greenidge (2001), Lim and Mcaleer (2001), Song et al. (2003), Narayan (2004), Han et al. (2006), Kim et al. (2011), De Vita and Kyaw (2013), and Isik et al. (2018) found that tourism leads to an increase in the economy.

Theoretical framework and empirical methodology

In order to determine the impact of the economic policy-related uncertainties (EPU index) on international tourist arrivals to the United States from its neighboring countries of Mexico and Canada, we apply the following tourists’ demand behavior function based on the Marshallian theory of consumer demand. According to this theory, demand for a product (tourist product) is the function of price of the product and income of the individual (tourist). In this function, besides income and price effects which are the most used variables in tourism demand models, EPU effect as an index is also added. This function will transform to the demand model of this study in the following equations in regression and ECM forms (Gómez et al., 2018):

{COT}_{t} = f ({RGDPC}_{t}^{\propto 2}, {REER}_{t}^{\propto 3}, {EPU}_{t}^{\propto 4})

where COT denotes arrivals from Canada and Mexico to the United States, RGDPC is real GDP per capita of these two countries, REER refers to the real exchange rates between Canadian Dollar (CAD)–Mexican Peso (MXN) and the USD, and EPU is the EPU indices of Canada and Mexico. Due to monthly data availability, we use Industrial Production Index as the proxy of GDP since it provides monthly data for the empirical studies. Recently, scholars recommended the use of this index in tourism demand models (Balli et al., 2018; Demir and Gozgor, 2018; Dogru et al., 2017; González and Paz, 1995; Ongan et al., 2017; Seo et al., 2009). Furthermore, the REER is calculated as:

R E E R = N E R \times (P_{d} / P_{f})

where NER is the nominal exchange rate (USD per MXN and USD per CAD), P_d denotes the CPI of Canada and Mexico, and P_f is the CPI of the United States. The reference year of CPI is 2005 (2005 = 100). The model in equation (1) can be rewritten in the following regression form as in equation (3):

{ln COT}_{t} = \propto_{0} + \propto_{1} ln {COT}_{t - k} + \propto_{2} ln {RGDPC}_{t - k} + \propto_{3} ln {REER}_{t - k} + \propto_{4} ln {EPU}_{t - k} + ε_{t}

In order to examine the short-run and long-run impacts of the EPU indices on the number of arrivals from Canada and Mexico to the United States, we apply ECM given in the following equation:

Δ ln {COT}_{t} = \propto_{0} + \sum_{i = 1}^{n} \propto_{1} Δ ln {COT}_{t - k} + \sum_{i = 0}^{n} \propto_{2} Δ ln {RGDPC}_{t - k} + \sum_{i = 0}^{n} \propto_{3} Δ ln {REER}_{t - k} + \sum_{i = 0}^{n} \propto_{4} Δ ln {EPU}_{t - k} + \propto_{5} ε_{t - 1} + μ_{t}

where Δ represents changes in variables; $ε_{t - 1}$ is the lagged error correction term (ECT).

Expected effects (signs) of the independent variables and data set

In this equation, we expect the sign of $\propto$ ₂ is to be positive since rises in RGDPC of Canada and Mexico will lead to more trips to the United States from these two countries. The sign of $\propto$ ₃ is expected to be positive as well since appreciations in the values of CAD and MXN against to the USD will lead to more vacations in the United States. However, the sign of $\propto$ ₄ is to be negative because rising uncertainties in the economies through the EPU indices in Canada and Mexico will lead to less vacation trips to the United States from these countries. ε_t is the error term. The data of monthly NERs, the CPI of Canada, Mexico, and the United States, and RGDPC and EPU indices were obtained from the database of Federal Reserve Bank of St. Louis (FED) (2015). The monthly arrivals were obtained from the National Travel and Tourism Office.

Empirical results and discussion

Before examining the long-run relationship between the variables, researchers should understand whether the series are stationary. Economic activities (series) exhibit structural changes (structural breaks) over time. In a similar manner, tourism activities contain various seasonalities. Hence, in this time series model, we apply Lee and Strazicich’s (2003) unit root test among many other alternative and interacting tests following each other (Lee et al., 2012; Meng et al., 2017; Narayan and Popp, 2010; Zivot and Andrews, 1992). Lee and Strazicich’s (2003) test takes into account and allows two endogenous structural breaks in the null and alternative hypothesis. Furthermore, this test also avoids the problems of bias and spurious rejections. The results of this test are reported in Table 1.

Table 1.

Results of the unit root test of Lee and Strazicich (2013).

	LM	Lag	Break date	LM	Lag	Break date
Variables (Canada)
ln EPU index	−3.114	2	2008(3)	−7.651***	2	2008(2)
ln RGDPC (2005 = 100)	−3.069	2	2008(4)	−5.978***	2	2008(2)
ln REER (2005 = 100)	−3.095	3	2001(2)	−6.434***	3	2009(2)
ln number of arrivals from Canada (2005 = 100)	−3.101	3	2010(4)	−5.457***	3	1998(3)
Variables (Mexico)
ln EPU index	−2.782	2	2008(2)	−8.458***	2	2008(1)
ln RGDPC (2005 = 100)	−2.483	3	2008(4)	−5.546***	3	2008(2)
ln REER (2005 = 100)	−2.496	2	2001(4)	−7.456***	2	2010(4)
ln number of arrivals from Mexico (2005 = 100)	−2.538	4	2010(4)	−5.429***	4	1998(3)
Critical value
%1	−4.239			−5.11
%5	−3.566			−4.50
%10	−3.211			−4.21

Note: Null hypothesis: the series has a unit root. The detected structural break dates in the table correspond to the pre–post effects of 2008 Global Financial Crisis and Mexican Currency Crisis in 1994.

*** The rejection of the null hypothesis at the 1% level.

According to Table 1, all variables are stationary I(1). Hence, we can run cointegration test. For this purpose, we apply Maki (2012) cointegration test considering structural break dates up to five. The null hypothesis of Maki (2012) test is “no cointegration among the series.” For the cointegration methodology by Maki (2012), the model in equation (4) can be represented in the following form:

y_{t} = μ + \sum_{i = 1}^{k} μ_{i} K_{i, t} + Ɣ χ + β χ_{t} + \sum_{i = 1}^{k} β_{i} χ_{i} K_{i, t} + ν_{t}

where $K_{i, t}$ is the dummy variable, $μ_{i}$ is the constant (intercept) term, Ɣ is the coefficient term, $β$ is the time trend term, and v_t is the error term. The results of this test are reported in Table 2.

Table 2.

Results of the cointegration test of Maki (2012).

ln number of arrivals from Canada and Mexico–ln RGDPC–ln REER–ln EPU
Variables for Canada	(Model 0) Test statistics/(break date)	(Model 1) Test statistics/(break date)	(Model 2) Test statistics/(break date)	(Model 3) Test statistics/(break date)
Variables for Canada	−9.23*** 2001(3), 2008(1), 2012(4)	−9.31*** 2001(3), 2008(1), 2011(4)	−10.27*** 2001(3), 2005(1), 2011(4)	−10.45*** 2001(3), 2005(1), 2011(4)
Variables for Mexico	(Model 0) Test statistics/(break date)	(Model 1) Test statistics/(break date)	(Model 2) Test statistics/(break date)	(Model 3) Test statistics/(break date)
Variables for Mexico	−17.89*** 1998(3), 2009(3), 2011(2)	−17.94*** 1998(3), 2009(3), 2011(2)	−19.56*** 1998(3), 2005(3), 2009(2)	−19.73*** 1998(3), 2005(3), 2009(2)
Critical value		%1	%5	%10
Critical value		−6.229	−5.704	−5.427

Note: The detected structural break dates in the table correspond to the pre–post effects of 2008 Global Financial Crisis and Mexican Currency Crisis in 1994.

*** The rejection of the null hypothesis at the 1% significance level.

The results in Table 2 indicate that there are statistically long-run relationships between the number of arrivals from Canada and Mexico to the United States and RGDPC-REER-EPU of these two countries. The next step is to estimate the long-run coefficients of the model in equation (4) with the dynamic ordinary least squares (DOLS) estimations of Stock and Watson (1993).

The test results of the DOLS estimations are reported in Table 3.

Table 3.

The estimated DOLS results.

Dependent variable	ln number of arrivals from Canada	ln number of arrivals from Mexico
Results of the DOLS for the long-run coefficients
ln RGDPC	1.823*** (0.321)	3.657*** (0.424)
ln REER	0.217*** (0.146)	0.468*** (0.224)
ln EPU index	−0.378*** (0.059)	−0.194*** (0.088)
D1998(3)	−0.297*** (0.021)	−0.329*** (0.019)
D2008(1)	−0.016 (0.021)	−0.018 (0.018)
D2011(2)	0.139 (0.014)	0.140 (0.012)
Constant term	−2.942*** (0.434)	−1.241*** (0.232)
Adjusted R ²	0.951	0.838
Standard error of regression	0.113	0.124
Durbin–Watson statistics	1.958	1.998
Long-run variance	0.043	0.039
Sum squared residuals	0.941	0.886
Results of the DOLS for the short-run coefficients
Error correction term	−0.058** (0.031)	−0.086** (0.029)
ΔLagged ln number of arrivals from Canada and Mexico	−0.485*** (0.077)	−0.510*** (0.085)
ΔLagged ln RGDPC	2.861** (1.622)	4.128** (1.521)
ΔLagged ln REER	−0.125 (0.112)	−0.392 (0.128)
ΔLagged ln EPU index	−0.354* (0.063)	−0.122* (0.058)
Adjusted R ²	0.928	0.813

Note: DOLS: dynamic ordinary least squares. The optimal number of the lag length is selected by the Schwarz Information Criteria. The impacts of the break dates of 2008(1) and 2011(2) were found to be statistically insignificant. The break date of 1998(3) was to be statistically significant at the 1% level.

*, **, and *** denote the statistical significance at the 1%, 5%, and 10% level, respectively.

The results in Table 3 indicate that the long-run coefficients of the income measures (RGDPC) of Canada and Mexico are positive (as expected) and elastic (coefficients 1.82 and 3.65, respectively). Similarly, the long-run coefficients of the price measures (the REER) of Canada and Mexico are positive (as expected) and inelastic (Canada: 0.21 and Mexico: 0.46). These negative relationships reveal that rises in the EPU index of Canada and Mexico lead to falls in the arrivals from these two countries to the United States in the long run. With respect to comparative elasticity coefficients of the EPU index of Canada and Mexico, Canadian tourists respond to the increases in the EPU index in their country proportionately more than the Mexican tourists do. While 1% rise in the EPU index in Canada leads to 0.37% decrease in the number of arrivals from this country to the United States, same percent rise in the EPU index in Mexico leads to only 0.19% decrease in the number of arrivals from this country. Canadian tourists are more sensitive to the increases in the EPU index in their country and thus reduce their vacations to the United States proportionately more than the Mexicans. Both results clearly reveal that uncertainties in the economies have negative impacts on vacation plans of the people. These results also affirm the findings of two recent studies by Ongan and Gozgor (2017) and Gozgor and Ongan (2017).

On the other hand, rises in real RGDPC of both countries lead to increases in the number of arrivals from these two countries to the United States. Similarly, appreciations in MXN and CAD against to the USD lead to more vacations to the United States from these two countries in the long run, a finding that is theoretically expected and significant at the 1% level.

The ECT for the ECM regression variables is −0.058 and −0.086 and are statistically significant at the 5% significance level. The negative signs of ECT indicate that the numbers of arrivals from Canada and Mexico to the United States converges with their long-run equilibrium paths at a 5.8% and 8.6% speed of adjustment which shows the speed with which the countries close the gap between each other.

Conclusion and future directions

Rising uncertainty is one of the prominent phenomenon of today’s modern economic systems. It not only affects some macroeconomic parameters but also changes the behavior decisions of the consumers in the selection of the goods and services according to their essential priorities. Thus, it can be assumed that travel plans or vacation decisions in this priority order can be easily postponed or completely cancelled during economic crises. In this study, we examined the question whether rising uncertainties in the economies affect potential travelers’ vacation plans. To this aim, we introduce and use a recently created EPU index and applied unit root- and cointegration tests with structural breaks. Empirical results of this study reveal that rises in the EPU index (uncertainties) of Canada and Mexico decrease the number of arrivals from these countries to the United States. This means that the Mexican and Canadian tourists are sensitive to economic uncertainties when they make travel plans to the United States. This finding would help guide US policy makers when they design tourism policies for these two tourist markets. Another significant finding of this study is the differential response behavior of Canadian and Mexican tourists to uncertainties in their economies. While Mexican tourists respond proportionally more to the increases in their income levels (income elastic) and appreciations in their currency, Canadian tourists are more sensitive to the uncertainty levels in their economy when they decide to travel to the United States. This outcome may be due to the fact that Mexican tourists have been exposed to uncertainties in their economy for a long time, and thus are familiar what it means living under economic uncertainty, and rising uncertainty levels may not affect them as much as it does the Canadians. It should be noted that the Mexicans are defined high uncertainty avoidance people in uncertainty avoidance index created by Hofstede (1980). This index ranks the societies according to their tolerance for uncertainty and ambiguity.

The final outcome of this study is consistent with the results of the forenamed previous empirical studies using the EPU index in tourism demand models. This study also finds that policy-related economic uncertainty plays determining role on tourists’ vacation plans. In order to reaffirm this expected result and also to compare the response levels of outbound tourists coming from the countries having different level uncertainties to the uncertainties, this study uses Mexican and Canadian tourists to the United States as the sample tourist groups. Furthermore, the empirical findings of this study lie in its breadth and depth of the coverage of all current political discussions (especially from the perspective of the United States) about the future of the USMCA. Because, there has been an increasing amount of discussion and criticism in the US side about USMCA since the large and persistent negative bilateral trade of the United States balances with Canada and Mexico.

In conclusion, the findings of this study suggest that the EPU index, as an additional independent variable, should be included in tourism demand models besides the traditional variables related to the economic factors. Today’s complex and uncertain world economy more than ever makes this consideration necessary. Further empirical studies using different methodologies and data sets involving other countries may be needed.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Cem Işık

References

Akal

(2004) Forecasting Turkey’s tourism revenues by ARMAX model. Tourism Management 25(5): 565–580.

Alonso

Bremser

(2013) Strategic responses of the Spanish hospitality sector to the financial crisis. International Journal of Hospitality Management 32: 141–148.

Antonakakis

Chatziantoniou

Filis

(2013) Dynamic co-movements of stock market returns, implied volatility and policy uncertainty. Economics Letters 120(1): 87–92.

Arouri

Rault

Teulon

(2014) Economic policy uncertainty, oil price shocks and GCC stock markets. Economics Bulletin 34(3): 1822–1834.

Law

(2002) Categorical classification of tourism dining. Annals of Tourism Research 29(3): 819–833.

Aye

(2018) Causality between economic policy uncertainty and real housing returns in emerging economies: a cross-sample validation approach. Cogent Economics and Finance 6(1): 1–16.

Ayeh

Lin

(2011) Estimating tomorrow’s tourist arrivals’: forecasting the demand for China’s tourism using the general-to-specific approach. Tourism and Hospitality Research 11(3): 197–206.

Bai

Perron

(1998) Estimating and testing linear models with multiple structural changes. Econometrica 66(1): 47–78.

Baker

Bloom

Davis

(2016) Measuring economic policy uncertainty. Quarterly Journal of Economics 131(4): 1593–1636.

10.

Balli

Shahzad

SJH

Uddin

(2018) A tale of two shocks: What do we learn from the impacts of economic policy uncertainties on tourism? Tourism Management 68: 470–475.

11.

Bureau of Economic Analysis (BEA) (2011) Bureau of Economic Analysis, U.S. Department of Commerce. Available at: https://www.bea.gov/ (accessed January 2018).

12.

Bekiros

Gupta

Paccagnini

(2015) Oil price forecastability and economic uncertainty. Economics Letters 132: 125–128.

13.

Bloom

(2009) The impact of uncertainty shocks. Econometrica 77(3): 623–685.

14.

Brogaard

Detzel

(2015) The asset-pricing implications of government economic policy uncertainty. Management Science 61(1): 3–18.

15.

Chen

Chiou-Wei

(2009) Tourism expansion, tourism uncertainty and economic growth: new evidence from Taiwan and Korea. Tourism Management 30(6): 812–818.

16.

Chow

Cunado

Gupta

, et al. (2017) Causal relationships between economic policy uncertainty and housing market returns in China and India: evidence from linear and nonlinear panel and time series models. Studies in Nonlinear Dynamics and Econometrics 22(2): 1–15.

17.

Chu

(2011) A piecewise linear approach to modeling and forecasting demand for Macau tourism. Tourism Management 32(6): 1414–1420.

18.

Colombo

(2013) Economic policy uncertainty in the US: Does it matter for the Euro area? Economics Letters 121(1): 39–42.

19.

Coshall

(2005) A selection strategy for modelling UK tourism flows by air to European destinations. Tourism Economics 11(2): 141–158.

20.

Cro

Martins

(2017) Structural breaks in international tourism demand: Are they caused by crises or disasters? Tourism Management 63: 3–9.

21.

Dai

Zhang

, et al. (2017) Causality between economic policy uncertainty and exchange rate in China with considering quantile differences. Theoretical and Applied Economics XXIV (3(612)): 29–38.

22.

Demir

Ersan

(2018) The impact of economic policy uncertainty on stock returns of Turkish tourism companies. Current Issues in Tourism 21(8): 847–855.

23.

Demir

Gozgor

(2018) Does economic policy uncertainty affect tourism? Annals of Tourism Research 69: 15–17.

24.

De Vita

Kyaw

(2013) Role of the exchange rate in tourism demand. Annals of Tourism Research 43: 624–627.

25.

Dogru

Sirakaya-Turk

(2018) Modeling Turkish outbound tourism demand using a dynamic panel data approach. Tourism and Hospitality Research 18(4): 411–414.

26.

Dogru

Isik

Sirakaya-Turk

(2019) The balance of trade and exchange rates: theory and contemporary evidence from tourism. Tourism Management 74: 12–23.

27.

Dogru

Sirakaya-Turk

Crouch

(2017) Remodeling international tourism demand: old theory and new evidence. Tourism Management 60: 47–55.

28.

Federal Reserve Bank of St. Louis (FED) (2015) Economic Research. Available at: https://fred.stlouisfed.org/categories/32263 (accessed January 2018).

29.

Formica

Uysal

(2006) Destination attractiveness based on supply and demand evaluations: an analytical framework. Journal of Travel Research 44(4): 418–430.

30.

Fountas

Karatasi

Tizika

(2018) Economic policy uncertainty in Greece: measuring uncertainty for the Greek macroeconomy. South-Eastern Europe Journal of Economics 1: 79–92.

31.

Garín-Muñoz

Teodosio

(2000) An econometric model for international tourism flows to Spain. Applied Economics 7: 525–529.

32.

Ghosh

(2018) Uncertainty, economic growth its impact on tourism, some country experiences. Asia Pacific Journal of Tourism Research 24(3): 1–25.

33.

Glover

Prideaux

(2009) Implications of population ageing for the development of tourism products and destinations. Journal of Vacation Marketing 15(1): 25–37.

34.

Gokovali

Bahar

Kozak

(2007) Determinants of length of stay: a practical use of survival analysis. Tourism Management 28(3): 736–746.

35.

Gómez

Ciarreta

Zarraga

(2018) Linear and nonlinear causality between energy consumption and economic growth: the case of Mexico 1965–2014. Energies 11: 784.

36.

González

Paz

(1995) An analysis of the international tourism demand in Spain. International Journal of Forecasting 11: 233–251.

37.

Gouveia

Rodrigues

PMM

(2005) Dating and synchronizing tourism growth cycles. Tourism Economics 11(4): 501–515.

38.

Gozgor

Ongan

(2017) Economic policy uncertainty and tourism demand: empirical evidence from the USA. International Journal of Tourism Research 19(1): 99–106.

39.

Greenidge

(2001) Forecasting tourism demand: an STM approach. Annals of Tourism Research 28(1): 98–112.

40.

Guizzardi

Mazzocchi

(2010) Tourism demand for Italy and the business cycle. Tourism Management 31(3): 367–377.

41.

Han

Durbarry

Sinclair

(2006) Modelling US tourism demand for European destinations. Tourism Management 27(1): 1–10.

42.

Harrod

(1958) The possibility of economic satiety – use of economic growth for improving the quality of education and leisure. In: Harrod

(ed.) Originally published in Problems of United States economic development, vol. I, 2017-13. New York: Committee for Economic Development, pp. 7–12.

43.

Henderson

(2006) Managing Tourism Crises. Kidlington: Elsevier.

44.

Hofstede

(1980) Culture’s Consequences. Beverly Hills: Sage.

45.

Ibrahim

MA MA

(2013) The determinants of international tourism demand for Egypt: panel data evidence. European Journal of Economics, Finance and Administrative Sciences, ISSN, 1450–2275.

46.

Isik

Radulescu

(2017) Investigation of the relationship between renewable energy, tourism receipts and economic growth in Europe. Statistika: Statistics and Economy Journal 97(2): 85–94.

47.

Isik

Doğan

Ongan

(2017) Analyzing the tourism–energy–growth nexus for the top 10 most-visited countries. Economies 5(4): 40.

48.

Isik

Dogru

Sirakaya-Turk

(2018) A nexus of linear and non-linear relationships between tourism demand, renewable energy consumption, and economic growth: theory and evidence. International Journal of Tourism Research 20(1): 38–49.

49.

Isik

Kasımatı

Ongan

(2017b) Analyzing the causalities between economic growth, financial development, international trade, tourism expenditure and/on the CO₂ emissions in Greece. Energy Sources, Part B: Economics, Planning, and Policy 12(7): 665–673.

50.

Işik

Radulescu

Fedajev

(2019) The effects of exchange rate depreciations and appreciations on the tourism trade balance: the case of Spain. Eastern Journal of European Studies 10(1).

51.

Istiak

Serletis

(2018) Economic policy uncertainty and real output: evidence from the G7 countries. Applied Economics 50(39): 4222–4233.

52.

Kapiki

(2012) Current and future trends in tourism and hospitality: the case of Greece. International Journal of Economic Practices and Theories 2(1): e-ISSN 2247–7225. Available at: https://ssrn.com/abstract=2150562 (accessed 18 November 2019).

53.

Karnizova

(2014) Economic policy uncertainty, financial markets and probability of US recessions. Economics Letters 125(2): 261–265.

54.

Keynes

(1936) The General Theory of Employment, Interest and Money. New York: Harcourt Brace and World.

55.

Kim

Moosa

(2005) Forecasting international tourist flows to Australia: a comparison between the direct and indirect methods. Tourism Management 26(1): 69–78.

56.

Kim

Wong

Athanasopoulos

, et al. (2011) Beyond point forecasting: evaluation of alternative prediction intervals for tourist arrivals. International Journal of Forecasting 27: 887–901.

57.

Kulendran

Kenneth

(2000) Modelling business travel. Tourism Economics 6: 47–59.

58.

Ledesma-Rodríguez

Navarro-Ibáñez

Pérez-Rodríguez

(2001) Panel data and tourism: a case study of Tenerife. Tourism Economics 7(1): 75–88.

59.

Lee

Strazicich

(2003) Minimum Lagrange multiplier unit root test with two structural breaks. Review of Economics and Statistics 85(4): 1082–1089.

60.

Lee

Strazicich

Meng

(2012) Two-step LM unit root tests with trend-breaks. Journal of Statistics and Economics Methods 1: 81–107.

61.

Leitao

(2010) Portuguese tourism demand. Journal of Global Business and Technology 5: 63–72.

62.

Song

Witt

(2005) Recent developments in econometric modeling and forecasting. Journal of Travel Research 44: 82–99.

63.

Lim

(1999) A meta-analysis review of ınternational tourism demand. Journal of Travel Research 37: 273–284.

64.

Lim

Mcaleer

(2001) Cointegration analysis of quarterly tourism demand by Hong Kong and Singapore for Australia. Applied Economics 33: 1599–1619.

65.

Madanoglu

Ozdemir

(2019) Economic policy uncertainty and hotel operating performance. Tourism Management 71: 443–452.

66.

Maki

(2012) Tests for cointegration allowing for an unknown number of breaks. Economic Modelling 29(5): 2011–2015.

67.

Maslow

(1987) Motivation and Personality, 3rd ed. New York: Harper Collins.

68.

McKercher

Hui

(2004) Terrorism, economic uncertainty and outbound travel from Hong Kong. Journal of Travel and Tourism Marketing 15(2–3): 99–115.

69.

Meng

Lee

Payne

(2017) RALS-LM unit root test with trend breaks and non-normal errors: application to the Prebisch-Singer hypothesis. Studies in Nonlinear Dynamics & Econometrics 21(1): 31–45.

70.

Meng

Chin

Grant

(2015) Long-run effect of the global financial crisis on Singapore’s tourism and the economy. Asian Economic Journal 29(1): 41–60.

71.

Mumtaz

Surico

(2016) Policy uncertainty and aggregate fluctuations. CEPR Discussion Paper No. DP9694. ARDL.

72.

Narayan

(2004) Fiji’s tourism demand: the ARDL approach to cointegration. Tourism Economics 10(2): 193–206.

73.

Narayan

Popp

(2010) A new unit root test with two structural breaks in level and slope at unknown time. Journal of Applied Statistics 37(9): 1425–1438.

74.

National Travel and Tourism Office (NTTO) (2018) Monthly Tourism Statistics. Işık.

75.

Neumayer

(2004) The impact of political violence on tourism. Journal of Conflict Resolution 4882: 259–281.

76.

Okumus

Altinay

Arasli

(2005) The impact of Turkey’s economic crisis of February 2001 on the tourism industry in Northern Cyprus. Tourism Management 26(1): 95–104.

77.

Ongan

Gozgor

(2017) Tourism demand analysis: the impact of the economic policy uncertainty on the arrival of Japanese tourists to the USA. International Journal of Tourism Research 20(3): 308–316.

78.

Ongan

Isik

Ozdemir

(2017) The effects of real exchange rate and income on international tourism demand for the USA from some European Union countries. Economies 5(51): 1–11.

79.

Ongan

Isik

Özdemir

(2018) The asymmetric effects of exchange rates on the US tourism balances in the NAFTA countries: an application of the nonlinear ARDL approach. Journal of Applied Economics & Business Research 8(3).

80.

Page

Song

(2012) Assessing the impacts of the global economic crisis and swine flu on ınbound tourism demand in the United Kingdom. Journal of Travel Research 51(2): 142–153.

81.

Papatheodorou

Rossello

Xiao

(2010) Global economic crisis and tourism: consequences and perspectives. Journal of Travel Research 49(1): 39–45.

82.

Payne

Mervar

(2002) A note on modelling tourism revenues in Croatia. Tourism Economics 8(1): 103–109.

83.

Perles-Ribes

Belén Ramón-Rodríguez

Rubia-Serrano

, et al. (2016) Economic crisis and tourism competitiveness in Spain: permanent effects or transitory shocks? Current Issues in Tourism 19(12): 1210–1234.

84.

Ritchie

Molinar

ACM

Frechtling

(2010) Impact of the economic crisis on global tourism, foundations of the crisis: a North American perspective. Journal of Travel Research 49(1): 5–15.

85.

Sahinoz

Cosar

(2018) Economic policy uncertainty and economic activity in Turkey. Applied Economics Letters 25(21): 1517–1520.

86.

Seddighi

Nuttall

Theocharous

(2001) Does cultural background of tourists influence the destination choice? An empirical study with special reference to political instability. Tourism Management 22(2): 181–191.

87.

Seo

Park

(2009) The analysis of the relationships of Korean outbound tourism demand: Jeju Island and three international destinations. Tourism Management 30(4): 530–543.

88.

Sheldon

Dwyer

(2010) The global financial crisis and tourism: perspectives of the academy. Journal of Travel Research 49(1): 3–4.

89.

Sheng

(2010) Evaluating the tourism potential of public museums in Hangzhou: a supply-side perspective. Journal of Travel and Tourism Marketing 27(3): 287–305.

90.

Singh

Das

Jana

, et al. (2019) A wavelet analysis for exploring the relationship between economic policy uncertainty and tourist footfalls in the USA. Current Issues in Tourism 22(15): 1789–1796.

91.

Smeral

(2009) Impacts of the world recession and economic crisis on tourism: forecasts and potential risks. Journal of Travel Research 49(1): 31–38.

92.

Song

Lin

(2010) Impacts of the financial and economic crisis on tourism in Asia. Journal of Travel Research 49(1): 16–30.

93.

Song

Lin

Witt

, et al. (2011) Impact of financial/economic crisis on demand for hotel rooms in Hong Kong. Tourism Management 32: 172–186.

94.

Song

Witt

(2009) The Advanced Econometrics of Tourism Demand. Milton Park: Taylor and Francis.

95.

Song

Wong

KKF

Chon

KKS

(2003) Modeling and forecasting the tourism demand for Hong Kong Tourism. International Journal of Hospitality Management 22: 435–451.

96.

Stock

Watson

(1993) A simple estimator of cointegrating vectors in higher order integrated systems. Econometrica 61(4): 783–820.

97.

Stylidis

Terzidou

(2014) Tourism and the economic crisis in Kavala, Greece. Annals of Tourism Research 44: 210–226.

98.

Theuns

(2014) Is tourism a luxury: Can we do without? Tourism Recreation Research 39(2): 221–233.

99.

Thompson

(2010) The exchange rate, Euro, switch and tourism revenue in Greece. Tourism Economics 16: 773–778.

100.

Tsui

WHK

Balli

Tan

DTW

(2018) New Zealand business tourism: exploring the impact of economic policy uncertainties. Tourism Economics 24(4): 386–417.

101.

Tsui

WHK

Fung

MKY

(2016) Causality between business travel and trade volumes: empirical evidence from Hong Kong. Tourism Management 52: 395–404.

102.

United Nations World Tourism Organization (UNWTO) (2013) Economic Crisis, International Tourism Decline and its Impact on the Poor. Madrid: World Tourism Organization. eISBN: 978-92-844-1444-4.

103.

Uysal

(1998) The determinants of tourism demand: a supply-side analysis. In: Ioannides

Debbage

(eds) The Economic Geography of the Tourist Industry. London: Routledge, pp. 79–98.

104.

Uysal

Crompton

(1984) An overview of the approaches used to forecast tourism demand. Tourism Management 5(4): 288–297.

105.

Wang

(2009) The impact of crisis events and macroeconomic activity on Taiwan’s international inbound tourism demand. Tourism Management 30(1): 75–82.

106.

Wooldridge

(2010) Econometric Analysis of Cross Section and Panel Data. Cambridge: MIT Press.

107.

Zivot

Andrews

(1992) Further evidence on the great crash, the oil-price shock, and the unit-root hypothesis. Journal of Business and Economic Statistics 10(3): 251–252.