Does healthcare sector development affect inbound tourism? Evidence from ASEAN countries

Abstract

Using the concepts of medical tourism and travel medicine, this research note proposes healthcare sector development has a positive impact on inbound tourism. This hypothesis is tested with the data of ASEAN countries from 2000 to 2018. Because of the presence of cross-section dependence and heterogeneity, appropriate panel cointegration test, panel coefficient estimation method and panel Granger causality test are used. The findings indicate that (1) evidence of cointegration between inbound tourism and healthcare sector development is found when inbound tourism is the dependent variable and (2) there is a Granger causality from healthcare sector development to inbound tourism. The findings suggest that there are temporal relationships from healthcare sector development to inbound tourism in both short-run and long-run. In summary, healthcare sector development has an impact on inbound tourism.

Keywords

healthcare inbound tourism travel medicine medical tourism panel cointegration panel Granger causality

Introduction

Medical tourism is a subset of inbound tourism where travellers move across borders to seek medical care (Ormond and Sulianti, 2017). Adequate medical infrastructure and facilities and large stock of highly trained doctors and nurses are two of the main reasons a country becomes the choice of destination for medical travellers (Ormond and Sulianti, 2017).

Occasionally, travellers are dealing with illness or injury during and/or post travel (Page, 2009). The increasing recognition given to travellers’ health and safety issues leads to the development of travel medicine (Page, 2009). From travel medicine perspective, the risk assessment and prevention at the pre-travel phase, and the travellers’ experience in managing and handling injury or illness during travel phase are dependent on the availability of medical infrastructure and facilities, nurses and doctors at destinations. The travellers, who are risk averse, adopt risk reduction strategies, which lead to the avoidance of destinations that are lack of medical infrastructure and facilities, and highly trained doctors and nurses (Page, 2009; Wang et al., 2019).

In summary, uncompromised medical services have a positive impact on inbound tourism from the perspective of medical tourism and travel medicine. Quality medical services depend on the development of healthcare sector. Positive healthcare sector development increases the number of highly trained doctors and nurses, and the adequacy and quality of medical infrastructure and facilities. These linkages are summarised in Figure 1. Therefore, this research note hypothesises that

Figure 1.

Influence of healthcare sector development on inbound tourism.

H1. Healthcare sector development has a positive impact on inbound tourism.

This hypothesis is tested with the data of the Association of Southeast Asian Nations (ASEANs). ASEAN has 10 countries: Brunei, Cambodia, Indonesia, Lao PDR, Malaysia, Myanmar, Philippines, Singapore, Thailand and Vietnam. The distribution of visitor arrivals is very uneven across ASEAN countries as shown by Table 1 and Table 2. In 2018, Thailand welcomed slightly more than 38 million inbound visitors, whereas Brunei had less than 0.28 million inbound visitors. Also, from 2015 to 2018, the contribution of Brunei to the total visitor arrivals to ASEAN was never higher than 1%, whereas Thailand contributed to almost 30% of the total visitor arrivals to this region. ASEAN is selected because (1) Malaysia, Singapore and Thailand are the main medical tourism destinations in Asia (Ormond and Sulianti, 2017) and (2) Malaysia, Singapore, Thailand and Vietnam were among the top 30 tourism destinations in the world in 2019 (World Tourism Organisation, 2020).

Table 1.

Visitor arrivals to ASEAN member states from 2015 to 2018.

Country	Year
Country	2015	2016	2017	2018
Brunei	218,213	218,809	258,955	278,136
Cambodia	4,775,231	5,011,712	5,602,157	6,201,077
Indonesia	10,406,759	11,519,275	14,039,799	15,810,305
Lao PDR	4,684,429	4,239,047	3,868,838	4,186,432
Malaysia	25,721,251	26,757,392	25,948,459	25,832,354
Myanmar	4,681,020	2,907,207	3,443,133	3,549,428
Philippines	5,360,682	5,967,005	6,620,908	7,127,678
Singapore	15,231,469	16,403,595	17,424,611	18,508,302
Thailand	29,881,091	32,529,588	35,591,978	38,178,194
Vietnam	7,943,651	10,012,735	12,922,151	15,497,791
ASEAN	108,903,796	115,566,365	125,720,989	135,169,697

Source: https://data.aseanstats.org/visitors

Table 2.

Contribution of visitor arrivals of each ASEAN member state to the region from 2015 to 2018.

Country	Year
Country	2015 (%)	2016 (%)	2017 (%)	2018 (%)
Brunei	0.20	0.19	0.21	0.21
Cambodia	4.38	4.34	4.46	4.59
Indonesia	9.56	9.97	11.17	11.70
Lao PDR	4.30	3.67	3.08	3.10
Malaysia	23.62	23.15	20.64	19.11
Myanmar	4.30	2.52	2.74	2.63
Philippines	4.92	5.16	5.27	5.27
Singapore	13.99	14.19	13.86	13.69
Thailand	27.44	28.15	28.31	28.24
Vietnam	7.29	8.66	10.28	11.47

Source: Calculated by the authors using the data from https://data.aseanstats.org/visitors

Brief literature review

Many conceptual papers in medical tourism have been written. They focus mainly on identifying the factors that lead to the development of the medical tourism industry (see Enderwick and Nagar, 2011; Kamassi et al., 2020; Vovk et al., 2021; Zarei and Maleki, 2019). Li et al. (2022) indicate some of these papers are marketing-oriented. Empirical studies focussing on the impact of healthcare sector development on inbound tourism are scarce. Limited to our knowledge, there are only two studies, Lee (2010) and Lee and Hung (2010), that use time series econometric techniques to analyse the dynamic interactions between healthcare sector development and inbound tourism. However, these studies produce mixed results.

Lee (2010) and Lee and Hung (2010) study such linkages using the data of Singapore. Lee (2010) uses number of doctors per 10,000 population, whereas Lee and Hung (2010) uses government expenditure on health per capita as the proxy for healthcare sector development. Both use total tourist arrivals as the proxy for inbound tourism. Lee (2010) finds healthcare sector development and inbound tourism are cointegrated, and healthcare sector development has a positive effect on inbound tourism. Also, evidence of short-run Granger causality cannot be found between these two variables. Lee and Hung (2010) are unable to find a cointegrating relationship between healthcare sector development, inbound tourism and economic growth when inbound tourism is the dependent variable, but shows a short-run Granger causality from healthcare sector development to inbound tourism.

Data, methodology and results

Annual data on per capita tourism revenue (TR) and per capita health expenditure (HE) are collected from World Tourism Organization (Yearbook of Tourism Statistics, Compendium of Tourism Statistics and Data Files) and World Health Organization Global Health Expenditure database (http://apps.who.int/nha/database), respectively, over the period 2000–2018 for ASEAN countries.

HE is the proxy for healthcare sector development because it is prepared by the World Health Organisation under the framework of the System Health Accounts 2011, which traces all health spending of a given country that can be used for evidence-based policy making. Furthermore, the data are comparable across countries and time. Different measures of inbound tourism have been used, where the number of tourist arrivals has been used by most of existing studies (Wang and Davidson, 2010). Wang and Davidson (2010) state that economic impacts are TR driven and TR is monitored by policymakers and businesses. Therefore, this study uses per capita TR as the proxy for inbound tourism so that normalization is carried out for the comparability of the proxies for healthcare sector development and inbound tourism. Both are measured in current USD.

To examine the impact of healthcare sector development on inbound tourism in the long run, the equation below is estimated:

{lnTR}_{it} = α_{0} + β_{1} {lnHE}_{it} + ε_{it}

(1)

where ln is natural logarithm. Natural logarithm is applied to both variables to reduce the variability of data. First, to estimate equation (1) with a panel cointegration technique, the order of integration of each variable is checked with the Pesaran (2007) panel unit root test, a second-generation panel unit root test, which allows for the presence of cross-section dependence and heterogeneity. The Pesaran (2007) panel unit root test reveals that both variables at the level have a unit root at the 1% level.

The Breusch and Pagan (1980) test (CD_LM1), the Pesaran (2004) test (equation (6) of Pesaran (2004)) (CD_LM2), the Pesaran (2004) test (equation (7) of Pesaran (2004)) (CD_LM3), and the Pesaran et al. (2008) test (LM_adj) are used to test for the presence of cross-section dependence of the disturbance terms of equation (1). The Pesaran and Yamagata (2008) test is performed to check for the presence of slope homogeneity in equation (1). The results of cross-section dependence and slope homogeneity tests are reported in Table 3. The null hypothesis of each test is rejected at the 1% level. The results suggest that cross-section dependence and heterogeneity are present in equation (1).

Table 3.

Results of cross-section dependence and slope homogeneity tests.

Cross-section dependence test
Test	Statistic	p-value
CD_LM1	210.996	0.000***
CD_LM2	17.497	0.000***
CD_LM3	12.672	0.000***
LM_adj	41.694	0.000***
Slope homogeneity test
$\hat{Δ}$	9.459	0.000***
${\hat{Δ}}_{adj}$	10.494	0.000***

*, **, *** indicate the rejection of the null hypothesis at the 10%, 5%, and 1% level, respectively.

Since equation (1) has cross-section dependence and heterogeneity, two second-generation panel cointegration tests, the Westerlund (2007) test, and the Westerlund and Edgerton (2007) test, are used. Also, both tests have good small-sample properties. The null hypothesis of the Westerlund (2007) test is there is no cointegration, whereas the null hypothesis of the Westerlund and Edgerton (2007) test is there is a cointegration. The results of both tests are reported in Table 4. Both G_T and Gα of the Westerlund (2007) test rejects the null at least at the 5%. The Westerlund and Edgerton (2007) test accepts the null at even the 10% level. Therefore, equation (1) is a cointegrating regression.

Table 4.

Results of panel cointegration tests.

The Westerlund (2007) test
Statistic	Statistic	Bootstrap p-value
G_T	−6.121**	0.026
G_α	−10.908***	0.004
The Westerlund and Edgerton (2007) test
LM statistic	Asymptotic p-value	Bootstrap p-value
0.811	0.209	0.591

Only G_T and Gα of the Westerlund (2007) test are produced because of the presence of heterogeneity

*, **, *** indicate the rejection of the null hypothesis at the 10%, 5%, and 1% level, respectively.

Since there is a presence of a cointegrating relationship, the estimation of heterogeneous long-run parameters is carried out with the Eberhardt and Teal (2010) augmented mean group (AMG) estimator. The results are reported in Table 5. From the panel, the AMG estimation indicates that healthcare sector development has a positive impact on inbound tourism because the estimated slope coefficient is equal to 0.267 and the slope coefficient is statistically different from zero at the 10% level. When country-specific results are evaluated, healthcare sector development positively influences inbound tourism for only five ASEAN countries: Cambodia, Lao PDR, Myanmar, Thailand and Vietnam.

Table 5.

Results of AMG long-run cointegrating coefficients.

Country	Intercept	Slope coefficient
Brunei	57.258*** [0.005]	−7.998*** [0.010]
Cambodia	1.892*** [0.001]	0.563*** [0.000]
Indonesia	2.878*** [0.000]	−0.013 [0.915]
Lao PDR	1.401** [0.035]	0.535*** [0.005]
Malaysia	6.967*** [0.000]	−0.227 [0.384]
Myanmar	−2.340*** [0.000]	1.117*** [0.000]
Philippines	3.567*** [0.000]	−0.098 [0.410]
Singapore	6.988*** [0.000]	0.012 [0.953]
Thailand	3.311*** [0.000]	0.388*** [0.004]
Vietnam	1.953*** [0.000]	0.299*** [0.006]
Panel	3.008*** [0.004]	0.267* [0.071]

Values in parentheses are p-values

*, **, *** indicate the rejection of the null hypothesis at the 10%, 5%, and 1% level, respectively.

The Emirmahmutoglu and Kose (2011) panel Granger causality test is also carried out to investigate the presence of Granger causality from lnHE to lnTR. The results are reported in Table 6. The obtained Fisher test statistic is equal to 28.534, suggesting that lnHE Granger causes lnTR at the 10% level for the whole panel. However, among the 10 countries, only Indonesia and Malaysia provide evidence of Granger causality from lnHE to lnTR at the 10% level and 5% level, respectively. In summary, evidence of Granger causality from healthcare sector development to inbound tourism is found.

Table 6.

Results of Emirmahmutoglu and Kose (2011) panel Granger causality test.

Country	H₀: lnHE does not Granger cause lnTR
Country	Lag	Wald statistic	p-value	Decision
Brunei	1	0.000	0.985	Accept
Cambodia	1	0.161	0.721	Accept
Indonesia	1	4.641*	0.061	Reject
Lao PDR	1	3.432	0.108	Accept
Malaysia	1	7.556**	0.030	Reject
Myanmar	1	0.641	0.461	Accept
Philippines	1	2.717	0.151	Accept
Singapore	1	1.338	0.274	Accept
Thailand	1	0.880	0.390	Accept
Vietnam	1	0.320	0.597	Accept
Panel Fisher	1	28.534*	0.097	Reject

*, **, *** indicate the rejection of the null hypothesis at the 10%, 5%, and 1% level, respectively. The lag lengths are determined according to Akaike information criteria as 1.

Conclusion

This research note aims to understand whether healthcare sector development has a positive impact on inbound tourism in the context of ASEAN. From panel perspective, there exists a long-run (cointegrating) relationship between healthcare sector development and inbound tourism when inbound tourism is the dependent variable, and a Granger causality from healthcare sector development to inbound tourism. Based on the empirical findings of this research note, policymakers should allocate sufficient financial resources to enhance medical services if destinations intend to expand their tourism industry. The findings of this study will still be relevant when COVID-19 becomes an endemic because the availability of quality medical services influences the image of countries as tourism destinations.

This study is with a limitation because it uses a bivariate model to deal with small sample size constraint. Such bivariate model may lead to the omission of important variables (Brida et al., 2016). There is a risk of getting biased results. However, Lutkepohl (2007) states that a simplified model, which is justified by the theoretical foundation, is more likely to provide interesting findings than a more complex model under small sample size constraint. Furthermore, Lee (2021) emphasises that in practice it is difficult to decide which other variables to be added into the model. Therefore, the addition of new variables may affect the findings.

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.

Data availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.

ORCID iD

Chew Ging Lee

Author biographies

Salih Türedi is an Associate Professor at Rize Recep Tayyip Erdoğan University. His research interests include economic growth, environmental economics, and applied econometrics. He has published more than 20 articles in different international journals.

Mustafa Şit is an Associate Professor at Harran University. He worked as Vice Manager of many units in this university. Also, he is a head of tourism business department. His research interests are in macroeconomics and tourism economics. He has published more than 40 articles in different national and international journals.

Haydar Karadağ is an Associate Professor at Rize Recep Tayyip Erdoğan University. He received his PhD from Istanbul University and his M.A. from Kocaeli University. Dr. Karadag’s current work focuses on macroeconomics, monetary politics, real estate, and tourism economics. He often examines these fields by using the data of different governmental institutions with convincing evidence for public policy decisions. He has published articles in different national and international journals. He is also a vice dean at the Faculty of Economics and Administrative Sciences.

Chew Ging Lee is a Professor and Head of Business School at the University of Southampton Malaysia. He was the Associate Vice Chancellor and a Professor at Wawasan Open University. He is a Fellow of the Higher Education Academy and a Fellow of Chartered Management Institute. His research interests include economics of sports and economics of tourism. He has published more than 50 articles in different international journals, such as Tourism Economics, Tourism Management, and International Journal of Hospitality Management.

References

Breusch

Pagan

(1980) The LM test and its application to model specification in econometrics. The Review of Economic Studies 47(1): 239–254.

Brida

Cortes-Jimenez

Pulina

(2016) Has the tourism-led growth hypothesis been validated? A literature review. Current Issues in Tourism 19(5): 394–430.

Eberhardt

Teal

(2010) Productivity analysis in global manufacturing production. Economics Series Working Papers 515. University of Oxford, Department of Economics.

Emirmahmutoglu

Kose

(2011) Testing for Granger causality in heterogeneous mixed panels. Economic Modelling 28(3): 870–876.

Enderwick

Nagar

(2011) The competitive challenge of emerging markets: the case of medical tourism. International Journal of Emerging Markets 6(4): 329–350.

Kamassi

Abd Manaf

Omar

(2020) The identity and role of stakeholders in the medical tourism industry: state of the art. Tourism Review 75(3): 559–574.

Lee

(2010) Health care and tourism: evidence from Singapore. Tourism Management 31(4): 486–488.

Lee

(2021) Tourism-led growth hypothesis: international tourism versus domestic tourism-Evidence from China. International Journal of Tourism Research 23(5): 881–890.

Lee

Hung

(2010) Tourism, health and income in Singapore. International Journal of Tourism Research 12(4): 355–359.

10.

Xue

, et al. (2022) Chinese medical tourism: review and research agenda. Journal of Quality Assurance in Hospitality & Tourism: 1–25. Published online ahead of print 18 May 2022. DOI: 10.1080/1528008X.2022.2073939.

11.

Lutkepohl

(2007) General-to-specific or specific-to-general modelling? An opinion on current econometric terminology. Journal of Econometrics 136(1): 319–324.

12.

Ormond

Sulianti

(2017) More than medical tourism: lessons from Indonesia and Malaysia on South–South intra-regional medical travel. Current Issues in Tourism 20(1): 94–110.

13.

Page

(2009) Current issue in tourism: the evolution of travel medicine research: a new research agenda for tourism? Tourism Management 30(2): 149–157.

14.

Pesaran

(2004) General diagnostic tests for cross section dependence in panels. Cambridge Working Papers in Economics No. 435, University of Cambridge, and CESifo Working Paper Series No. 1229.

15.

Pesaran

(2007) A simple panel unit root test in the presence of cross-section dependence. Journal of Applied Econometrics 22(2): 265–312.

(2008) A bias-adjusted LM test of error cross-section independence. The Econometrics Journal 11(1): 105–127.

17.

Pesaran

Yamagata

(2008) Testing slope homogeneity in large panels. Journal of Econometrics 142(1): 50–93.

(2021) Identification of factors for the development of medical tourism in the world. International Journal of Environmental Research and Public Health 18: 11205.

(2010) A review of micro-analyses of tourist expenditure. Current Issues in Tourism 13(6): 507–524.

, et al. (2019) Travellers’ self-protections against health risks: an application of the full Protection Motivation Theory. Annals of Tourism Research 78: 102743.

21.

Westerlund

(2007) Testing for error correction in panel data. Oxford Bulletin of Economics & Statistics 69(6): 709–748.

22.

Westerlund

Edgerton

(2007) A panel bootstrap cointegration test. Economics Letters 97(3): 185–190.

23.

World Tourism Organisation (2020) World Tourism Barometer and Statistical Annex December 2020. https://www-e--unwto-org-443.web.bisu.edu.cn/doi/epdf/10.18111/wtobarometereng.2020.18.1.7.

24.

Zarei

Maleki

(2019) Asian medical marketing, a review of factors affecting Asian medical tourism development. Journal of Quality Assurance in Hospitality & Tourism 20(1): 1–15.