Economic Policy Uncertainty, COVID-19, and Tourist Stays in Croatia: Evidence from a Fourier Toda-Yamamoto Modeling Approach

ABSTRACT

This study extends the literature with respect to economic policy uncertainty measures and tourism flows to Croatia through the use of the Toda and Yamamoto modeling approach with a Fourier approximation to capture structural breaks. The results show that domestic economic policy uncertainty does not have a significant impact on tourist overnight stays. However, an increase in European economic policy uncertainty reduces total and domestic tourist overnight stays. An increase in COVID-19 cases has a negative and significant impact on total, domestic, and foreign tourist overnight stays, and contributes to increases in both Croatian and European economic policy uncertainty.

KEYWORDS:

• Tourist overnight stays
• economic policy uncertainty
• tourism demand
• Fourier Toda-Yamamoto
• Croatia

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

¹ Croatian Bureau of Statistics (2019) and Institute for Tourism (2018).

² Payne, Gil-Alana, and Mervar (2022) outline the government policy actions to support the tourism sector and the Croatian economy with the onset of the COVID-19 pandemic.

³ In addition to tourism flows, the impact of economic policy uncertainty on the financial and operating performance of tourism-related firms has also been explored, see Demir and Ersan (2018); Ersan, Akron, and Demir (2019); Madanoglu and Ozdemir (2019); Akron et al. (2020); Demir, Gozgor, and Paramati (2020); Hadood and Irani (2020); Garcia-Gomez et al. (2022); Irani, Athari, and Hadood (2021); Kumar (2021); Ozdemir, Han, and Dalbor (2022); Bashir and Kumar (2022); and Jiang et al. (2022).

⁴ The period of the analysis corresponds with the start date for the Croatian economic policy uncertainty index.

⁵ https://podaci.dzs.hr/en/

⁶ www.policyuncertainty.com

⁷ https://fred.stlouisfed.org/series/EUEPUINDXM

⁸ https://www.hnb.hr/en/statistics/statistical-data/rest-of-the-world/competitiveness

An increase in the real effective exchange rate represents a depreciation.

⁹ https://www.hzjz.hr/aktualnosti/covid-19-izvjesce-hzjz-a/

¹⁰ Under the null hypothesis, the KPSS statistic does not follow the asymptotic distribution with the critical values provided by Kwiatkowski et al. (1992).

¹¹ Under the null hypothesis, the asymptotic distribution of the $KPSS\left(k\right)$ statistic depends on the Fourier frequency $k$ with the critical values reported in Becker, Enders, and Lee (2006).

¹² The stationarity tests were conducted with the GAUSS library for Time Series and Panel Data Methods (TSPDLIB Version 2.0) developed by Nazlioglu (2021)

¹³ The bootstrap distribution is used to increase the power of the test statistic in small samples and is robust to the unit root and co-integration properties of data (see Mantalos (2000); Hatemi-J (2002); and Hacker and Hatemi-J (2006)). The bootstrap distribution of the Wald statistic is obtained by employing the residual sampling bootstrap approach originally proposed by Efron (1979).

¹⁴ The Toda and Yamamoto tests with Fourier approximation was conducted with the GAUSS library for Time Series and Panel Data Methods (TSPDLIB Version 2.0) developed by Nazlioglu (2021).