Dyadic hostility, tourism, and regional trade agreements

ABSTRACT

This study uses a gravity approach to investigate the effect of dyadic hostility levels on bilateral tourism flows and determine whether regional trade agreements (RTAs) attenuate the link between dyadic hostility levels and tourism flows. We used a balanced panel dataset of 189 destination countries, 200 origin countries, and 16,059 country-pairs from 2005 through 2015. The results show that while low-level dyadic hostility suppresses international tourism flows, its effect is often weak or statistically insignificant. However, the analysis shows strong evidence that high-level dyadic hostility dampens tourism flows across a range of model specifications. Furthermore, the results indicate that RTA membership helps to reverse the negative effect of both low and high levels of dyadic hostility on tourism flows.

KEYWORDS:

• Dyadic hostility
• international tourism flows
• regional trade agreement
• gravity framework
• conflict resolution

Disclosure statement

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

Notes

1 For example, tourist inflows to Ukraine dropped significantly during the armed conflict that followed Russia’s annexation of Crimea in 2014 (Tomczewska-Popowycz & Quirini-Popławski, 2021).

2 The term ‘countries’ is used throughout this study for the sake of convenience and without judgement as to the legal status or sovereignty of any region, territory, or jurisdiction.

3 Here, we set levels of hostility, three and four, to missing values to ensure that the reference category is no hostility – which equals zero. We used a similar approach to generate other indicators of hostility.

4 To calculate the precise change in bilateral tourist flows due to high-level hostility, we used the following formula: $\left(e^{\text{β}}-1\right)\times 100.$ For instance, in Column 2, the coefficient for high-level hostility from origin to destination is −0.246; hence the precise value is $\left(e^{-.246}-1\right)\times 100=-22\text{\%}.$ Therefore, tourist flows would be 22% lower for a country-pair having high-level hostility from origin to destination compared with a country-pair with no hostility.

5 For detailed definitions of control variables, please refer to the supplementary material A.

6 The percentage difference is based on the coefficient of the interaction term and calculated as follows: $\left(e^{0.747}-1\right)\times 100=110\text{\%}.$ The coefficient on low-level hostility from the origin to destination countries is −0.349, indicating that international tourist flows are 29% lower for non-RTA member countries experiencing low-level hostility from origin to destination (calculated as follows: $\left(e^{-0.349}-1\right)\times 100=-29\text{\%}$). The difference between tourist flows for RTA and non-RTA member countries experiencing low-level hostility is given by the coefficient of the interaction term of RTA with low-level hostility from the origin to the destination, which is 0.747, indicating that the international tourist flows for RTA member countries experiencing low-level hostility from the origin to destination countries is 110% higher vis-à-vis a country-pair that is not part of an RTA.

Additional information

Funding

This work was supported by United Arab Emirates University [grant number UPAR grant # 12B001].