Tourism development and happiness: International evidence

Abstract

The existing tourism- and happiness-related literature commonly investigates whether a happy destination does in fact render tourists happy and how tourist arrivals affect residents’ happiness. This research thus first explores whether the host country and Twitter happiness indices influence tourism development in an international framework (i.e. tourist arrivals, tourism revenues, and travel and leisure sector returns). To account for intricate correlations among variables, the study employs a quantile regression approach on panel data from 119 countries spanning 2006–2017. Evidences find that most host country happiness indices and macroeconomic factors show salient, nonlinear, and asymmetric impacts across both tourist arrival and tourism revenue quantiles in concurrent and subsequent periods, except for European country results. Moreover, Twitter happiness index strongly affects travel and leisure sector returns, but has no impact on tourist arrivals as well as tourism revenues, implying the importance of social media happiness on said returns.

Keywords

host country happiness quantile regression tourism development tourist arrivals travel and leisure sector Twitter happiness

Introduction

A key subject in tourism studies concerns tourism development and happiness (Kwon and Lee, 2020), even as the literature on happiness in the tourism arena did not appear until the start of the 2000s (Filep and Deery, 2010). For instance, Bimonte and Faralla (2016) and Rivera et al. (2016) probe the influences of tourism development on the happiness of residents, while Ivlevs (2017) suggests that tourist arrivals reduce residents’ happiness. However, academic debates still neglect the relationships among social media happiness, host country happiness, and tourism development. Pearce (1994) and Snaith and Haley (1994) pinpoint very early that a happy host is more likely to welcome a tourist and in so doing creates an atmosphere that is favorable for both increased return visits and positive word-of-mouth marketing. To the best of our knowledge, until now few studies report on whether happiness improves tourism development together with the exceptional development of the global travel and leisure sector (Pearce, 1994; Snaith and Haley, 1994).

According to Snaith and Haley (1999), a “happy host” is necessary for the prosperous growth of tourism, but the existing literature sparsely examines how happiness impacts tourism development (Pearce, 1994; Seresinhe et al., 2019; Snaith and Haley, 1994). Additionally, happiness variables are difficult to measure, and thus researchers in the happiness field have had to contend with substantial limitations in gauging happiness stages (Seresinhe et al., 2019). Employing social media (Twitter) and global country happiness indices, our research thus complements previous analyses of the tourism–happiness nexus. Our purpose of this study is to creatively probe the diverse perceptions of happiness and enhance the tourism literature about various practices as well as some perspectives that are neglected. Based on the happy host theory, this research empirically explores whether the host country and Twitter happiness indices attract tourist arrivals, increase tourism revenues, and thus improve travel and leisure sector returns, whether tourist arrivals increase revenues and sector returns, as well as whether tourism revenues enhance sector returns.

According to the World Travel and Tourism Council (2019), one of the world’s largest financial sectors is travel and leisure, which in 2018 contributed US$8.8 trillion to the global economy, created 319 million jobs, and increased global GDP by 10.4%, or 10% of total employment. Travel and leisure, the second fastest-growing sector in 2018 only marginally behind manufacturing, has had an impressive impact on the world economy. In addition, the sector is presently an agent of economic growth and development that has been widely adopted and officially sanctioned internationally (Lee and Chang, 2008; Lee and Chien, 2008). Furthermore, the World Travel and Tourism Council (2019) pinpoints that factors influencing the flow of global travelers, such as destinations’ attractiveness, will continue to impact traveler behavior in the future. With nearly all countries looking to attract international visitors in the global tourism sector, it is therefore vital to apprehend the determinants of the travel and leisure sector’s revenues, arrivals, and returns.

Human beings continually pursue happiness throughout their whole life. While lately there has been an enormous upsurge in scholarly and policy interest on happiness in the tourism field, to the best of our knowledge, scant tourism literature has explored the impacts of tourism host happiness and social media happiness on tourism development. Gradually reproducing and swaying the performance of many other composite systems (Ranco et al., 2015), social media constitute a main device in preparing vacations and can aid to progress demand predicting for tourism products. For example, Twitter has quickly expanded its popularity since its establishment in March 2006 (Wei et al., 2016) with more than 321 million active users as of June 2019. Social media disseminate information to a wide audience around the world. Edelmann et al. (2012) specifies that the user-generated content on Twitter from hundreds of millions of its users explains their collective behavior in a formerly unbelievable fashion. Twitter happiness statistics have found their way into the field of financial research by providing free, reliable, and various data without delay. For example, You et al. (2017) reveal that Twitter has predictive power for stock returns, but the existing literature (e.g. Ivlevs, 2017; Pearce, 1994; Seresinhe et al., 2019; Snaith and Haley, 1994) has sparsely employed a social media (Twitter) happiness index to explore its relationship with tourism. This present study thus initially uses the Twitter happiness index to gauge its impact on tourism development.

The World Happiness Report is a momentous investigation of the condition of worldwide happiness that ranks 156 nations by how happy their residents recognize themselves to be. It presents the available global data on national happiness, presenting that the quality of people’s lives can be coherently, reliably, and validly assessed by a variety of subjective well-being measures. Most tourism studies examine the travel and happiness nexus by using questionnaire data, whereas scant tourism-related literature employs country-level happiness indices to explore this relationship. For example, Bimonte and Faralla (2016) conduct an examination at a resort afore and during the tourist period to test for variances in local dwellers’ perceptions of tourism’s influence and happiness. Kim et al. (2013) employ a survey to test the links of tourism’s impact on dwellers’ happiness. Ivlevs (2017) uses a social investigation accompanied in 32 European nations to explore the influence of tourist arrivals on the happiness of residents. However, until now, there is no evidence regarding whether host country happiness and social media (Twitter) happiness influence tourism development.

Research shows the existence of a nonlinear or asymmetric relation among macroeconomic variables, happiness, and tourism. For example, Sprott (2005) proposes that mathematical happiness models should include a nonlinear effect. When conceptualizing the topic of happiness, Warr (2007) recommends looking for nonlinear patterns. Using threshold regression models, Po and Huang (2008) find a nonlinear association exists between tourism development and economic growth. Zhang et al. (2018) present that Twitter happiness and stock market index returns exhibit strong nonlinear relationships in the United States. Meo et al. (2018) discover the long-term asymmetric impacts of exchange rate, inflation, and oil prices on tourism demand. De Neve et al. (2018) reveal that measures of happiness as well as positive and negative effects are more than double as sensitive to economic recessions as compared to comparable upswings. Kuo et al. (2018) propose that a tour guide’s capability can lead to salient asymmetric negative influences on tourist satisfaction. Most tourism-related studies using conventional ordinary least squares (OLS) regression offer a partial depiction of a conditional distribution (Mosteller and Tukey, 1977) and do not attain the coefficients of the independent variables for the whole regression as a function of the variation in tourism variables. For study of tourism, quantile regression (QR) offers a more elastic and comprehensive feature of its factors at the higher and lower tails of the distributions (Hung et al., 2010; Masiero et al., 2015). As such, Wang et al. (2019) use QR to find that hotels with greater prices are less sensitive to seasonality. Therefore, our research utilizes the QR approach to estimate the happiness–tourism nexus in an international framework.

This study contributes to the related works in several aspects. First, the existing literature (e.g. Pearce, 1994; Seresinhe et al., 2019; Snaith and Haley, 1994) does not consider social media (Twitter) happiness and country happiness together in the tourism theme with few studies ever exploring whether happiness spurs tourism development. Our study initially explores whether the two happiness indices benefit the tourism sector. Second, this study comprehensively discusses three tourism development proxies, offering a policy suggestion that when nations have less international tourists and less tourism revenues, then trying to cultivate a happy host country environment will help tourism growth. However, if managers and/or stakeholders in the travel and leisure sector are targeting to raise stock returns, then they can take into account the increases in Twitter happiness. Third, Ivlevs (2017) finds a negative relationship in which tourist arrivals diminish residents’ life happiness, which is more noticeable in nations with higher tourism intensity. Lee and Jang (2011) reveal nonlinear, asymmetric, and lagging effects of exchange rate exposure for tourism-associated businesses. Therefore, the OLS approach should be evaded in these kinds of study as it offers a fragmented depiction of the tourism–happiness nexus. A quantile analysis regarding Twitter happiness and host country happiness has important implications across different tourism development quantile distributions as it helps to improve management strategies for the tourism sector’s downside and upside conditions. Fourth and finally, most tourism studies focus on single countries or single tourism spots, thus providing only local or limited evidence. Our research employs 12-year panel data from 119 countries that can be generalized into international evidence concerning this topic.

This study analyzes how Twitter and host country happiness indices shape conditional tourist arrivals, tourism revenues, and travel and leisure sector returns for 119 countries for the period 2006–2017. After controlling for the effect of macroeconomic variables, our findings present salient evidence that some happy host countries asymmetrically attract tourist arrivals, and that increases in international tourism revenues differ across quantiles, but no evidence appears on travel and leisure sector returns, supporting that the happy host country theory explains tourist arrivals and tourism revenues. However, Twitter happiness relates neither to tourist arrivals nor to tourism revenue, whereas Twitter happiness does impact travel and leisure sector returns, signifying the importance of social media happiness. Further test results reveal that the relationship mentioned above is robust when considering non-global financial crisis subperiods, endogeneity problems, and country economic development conditions, whereas the effects are insignificant for the European country subsample, meaning geographic regional differences exist in the happiness–tourism nexus. In light of these findings, the article concludes that host country happiness (Twitter) can serve as a determinant of tourism (travel and leisure sector returns).

Our research is organized as follows. The second section offers a brief review of the literature and states the hypotheses. The third section illustrates the research methodology. The fourth section analyzes and discusses the empirical findings obtained. The fifth section concludes the study.

Literature review

Many factors influence tourism development, with researchers attempting to set up an equation of the request for tourism in order to examine the diverse elements that impact tourism development (Surugiu et al., 2011). Existing studies discuss how macroeconomic factors influence tourism (Chaabouni, 2019; Lee and Brahmasrene, 2013; Shahzad et al., 2017; Surugiu et al., 2011), and therefore this study includes several such variables like consumer price, exchange rate, oil price, income, and so on. On the other hand, there is also an increasing number of studies focusing on noneconomic determinants of tourism, such as happiness, which has been the theme of empirical study in social sciences ever since the 1960s (Nawijn, 2011; Nawijn and Veenhoven, 2013; Seresinhe et al., 2019). It is a highly valued topic, as with a few exceptions all humans generally want to be happy, and many people strive to be happier than they are (Nawijn, 2011).

While 100% happiness seems unrealistic, findings on this subject have gradually been taken into account in standard economics (Frey, 2008). Today, tourism is especially seen more and more as a health and well-being activity (Filep and Deery, 2010). Linking two fields of research, unconventional forms of tourism and the economics of happiness, Bimonte and Faralla (2012) test the association between happiness and the kind of tourist a person is. Rivera et al. (2016) investigate the empirical association between tourism development and happiness from the viewpoint of locals in a small island destination. Wu et al. (2017) examine the interrelationships among happiness, rural image, and behavioral targets for China’s rural tourism sector. Chen and Li (2018) examine whether a happy destination brings tourist happiness in Switzerland. Carneiro and Eusébio (2019) analyze the elements swaying the influence of tourism trips on young tourists’ happiness using the Oxford Happiness Inventory. Kwon and Lee (2020) explore how traveling affects the duration of happiness. These studies show how research on topics such as happiness and tourist happiness has flourished in recent years (Filep, 2014).

Whether residents’ happiness and/or social media happiness influence tourism development has been neglected by the related literature. With the quick growth of information technology, social media have developed an enormous repository of rich user-generated content (Thelwall et al., 2011). Gunter (2018) uses Google trends and Facebook pages data to predict tourist arrivals. With its large user base and high engagement, Twitter has increasingly been used by financial field research. You et al. (2017) find that a causal link from stock returns to Twitter happiness occurs only at high quantiles. Zhang et al. (2018) show that Twitter happiness can Granger-cause index returns in a linear causality test, whereas the opposite direction is more noticeable in a nonlinear causality examination. Additionally, Twitter happiness and index returns show robust nonlinear associations in the United States, whereas Twitter happiness does not Granger-cause index returns in Middle East and North Africa (Zhang et al., 2018). Empirical studies employing Twitter happiness data have been conducted for financial markets as well. Based on the abovementioned studies, we hypothesize that Twitter happiness increases travel and leisure sector stock returns, tourism revenues, and tourist arrivals.

Though online questionnaire surveys are quite popular, they suffer the disadvantages of specialized populations, self-selection bias, and gathered samples from virtual groups (Wright, 2005). The World Happiness Report is a momentous study of the condition of worldwide happiness that ranks 156 nations by how happy their residents recognize themselves to be and presents global data on national happiness, revealing that the quality of people’s lives can be consistently, reliably, and genuinely measured by a diversity of subjective well-being measures. The World Happiness Index has become a hot topic of discussion, because it allows for comprehension about the quality of life (Hariyanto, 2017). Tourism sector development can help reduce unemployment and remove poverty, thereby augmenting the position of India in terms of the World Happiness Index and further contributing to global peace (Mishra and Verma, 2017). Until now, scant research employs this index to investigate the tourism field. The World Happiness Index (2019) includes 12 factors, for which we use 10 of them: life ladder, GDP, social support, healthy life expectancy, freedom to make life choices, generosity, perceptions of corruption, country negative affect, confidence in national government, and democratic quality. We remove 2 of the 12 factors due to a high correlation among the factors.

Demir et al. (2017) indicate that international tourists Granger-cause tourism stock returns in Turkish tourism firms. However, Chen (2011) reveals that growth in total tourists has a more straight impact on hotel sales and profitability than it does on hotel stock performance. He attributes the lack of a robust link between tourist arrivals and stock returns to the time-varying discount rate caused by investors’ fluctuating expectations over the outlook of upcoming cash flows. Chen (2010) shows that changes in GDP and international tourists are notable explanatory elements of hotel industry performance, but these two elements do not have any salient impact on hotel stock returns. Chen (2007) presents that foreign tourist arrivals show a positive but insignificant effect on Chinese hotel stock returns, and general macro-level variables are more sensitive to these returns. Thus, the relationships among the abovementioned variables are inconsistent.

Po and Huang (2008) are the first to point out that a nonlinear association exists between tourism growth and economic development and also find that tourism does not necessarily bring about economic growth under certain conditions. Sprott (2005) states that the simple linear model supposes that, over a sufficiently long time, people tend to adjust to their circumstances and thus experience equivalent extents of happiness and unhappiness. Consequently, persistent happiness is an impractical and unattainable objective. Zhang et al. (2018) reveal that Twitter happiness and index returns exhibit strong nonlinear relationships in the United States. Meo et al. (2018) discover long-run asymmetric impacts of exchange rate, inflation, and oil prices on tourism demand. Kuo et al. (2018) denote that a tour guide’s ability leads to salient asymmetric negative effects on tourist satisfaction. Thus, our study utilizes the QR approach to estimate whether the relationship between happiness and tourism development varies at different quantiles of the tourism distributions. Based on international evidence, this study develops the following hypotheses to generalize the relationships of tourist arrivals, tourism revenues, and travel and leisure sector returns internationally and depicts our research concept in Figure A1 in the Online Appendix.

H1a: Twitter happiness has significant effects on international tourist arrivals.

H1b: Host country happiness has significant effects on international tourist arrivals.

H2a: International tourist arrivals have significant effects on international tourism revenues.

H2b: Twitter happiness has significant effects on international tourism revenues.

H2c: Host country happiness has significant effects on international tourism revenues.

H3a: International tourist arrivals have significant effects on travel and leisure sector returns.

H3b: International tourism revenues have significant effects on travel and leisure sector returns.

H3c: Twitter happiness has significant effects on travel and leisure sector returns.

H3d: Host country happiness has significant effects on travel and leisure sector returns.

H4: The relationship between happiness and tourism development varies at different quantiles of the tourism distribution.

Economic globalization and developments in transportation and communications technology in the 21st century have spurred governments to trace and endorse fruitful sectors to overcome macroeconomic difficulties such as inflation, unemployment, and stagnant growth (Chaabouni, 2019). Tourism is regarded as one such essential industry, because it can improve the balance of payments and create income, taxes, hard currency, and jobs (Lee and Brahmasrene, 2013). Tourism is one of the fastest developing industries in the world and correlates closely with economic growth and socioeconomic improvement for many developing nations as well as for some developed nations (Shahzad et al., 2017). The travel and leisure industry, as a leader of economic growth, can inspire GDP growth through jobs and enterprise formation and offer substantial foreign exchange revenues (Chang et al., 2013). Chiu and Yeh (2017) find robust evidence of a nonlinear link between tourism growth and economic growth, signifying that it is not unceasing and constant. Demir et al. (2017) note that growth in consumer price index (CPI), imports, exchange rate, oil prices, and foreign tourist arrivals is related. Therefore, we include CPI, exchange rate, foreign direct investment, industrial production, inflation, imports, oil prices, and unemployment as control variables and employ the QR model to explore the impacts of independent variables under different quantiles of tourist arrivals, tourism revenues, and sector returns.

Methodology

Data

This research employs two sources of happiness indices. One covers the daily happiness data obtained from Twitter. This index is a derivative of natural language processing skills by utilizing a random sampling of around 10% of all messages presented on Twitter. We operate Amazon’s Mechanical Turk service to score the level of happiness of particular words appearing on Twitter, such as joy, successful, laughter, winning, and excellent. Because the country happiness indices and tourism-related data are available on a yearly frequency, we use the Twitter happiness data at the end date of the sampled year as yearly data to conduct the analysis. The other source is the yearly country happiness data extracted from the World Happiness Report. The indices are taken as measures of a country’s subjective happiness. We aim to illuminate the differences across countries by means of explanatory variables: life ladder, GDP, social support, health life expectancy, freedom to make life choices, generosity, country perceived absence of corruption, country negative effect, country confidence in the national government, and democratic quality. The year 2017 has the least amount of data on tourist arrivals and tourism revenues. The earliest starting time for data on Twitter happiness is September 9, 2008. Moreover, there are 156 countries with data in the World Happiness Report website. We match the 156 countries with those countries that have travel-related data in the World Bank database and thus use 119 countries in this study. Table A1 in the Online Appendix outlines all variables employed in this article.

Sreekumar and Parayil (2002) employ the ratio of foreign tourism revenues among total exports to study the contention of tourism as a development option. Gokovali (2010) also utilizes the ratio of foreign tourism revenues among total exports to probe the contribution of tourism to economic development. Eugenio-Martin et al. (2004) study the nexus between the amount of tourist arrivals and economic development. Following the abovementioned studies, we obtain three dependent variables: log of international tourists’ arrivals and international tourism revenues (percentage of total exports) from the World Bank database and the travel and leisure sector’s yearly prices from the DataStream database. We gauge sector returns by $(P_{t} - P_{t}_{- 1}) / P_{t}_{- 1}$ , where P_t is the adjusted closing price at time t in US dollars. Because the latest data of tourist arrivals and tourism revenues are from 2017, the study period is 2006–2017. Table A2 in the Online Appendix lists the sample countries. For robustness check, we take the log of international tourism receipts in US dollars to proxy tourism revenues.

Chen (2007) investigates the association between macro-level and non-macro-level explanatory variables and Chinese hotel stock returns by using industrial production, imports, CPI, and international tourist arrivals. Chiu and Yeh (2017) find that tourism growth has a significant link with inflation and exchange rate change. Demir et al. (2017) reveal that consumer confidence index, exchange rate, imports, oil prices, and international tourist arrivals affect tourism stock returns. Santana-Gallego et al. (2010) reveal that the exchange rate is a main determinant of tourist demand. Zapata and Rambaldi (1997) indicate that a causal link exists between foreign direct investment and tourism. Perles-Ribes et al. (2016) present an unemployment effect of financial crises on hotel and tourism destinations. Tang (2011) finds bilateral causality between unemployment and tourist arrivals.

We therefore account for the influence of economic factors by considering CPI, log of official currency exchange rate per US$ (EXG), net inflow of foreign direct investment (FDI), log of industrial production in constant US$ (IND), inflation (INF), log of imported merchandise (IMP), crude oil price ratio (OIL), and unemployment (UMP). All data are gathered in US dollars. Among the eight macro-level factors, seven of them are from the World Bank database, while OIL is from http://oilpirce.com. Moreover, the hospitality and tourism sector has been experiencing numerous challenges following the global financial crisis (Kapiki, 2012; Smeral, 2009). Following Hill et al. (2015), we thus set the global financial crisis period as 2008–2009. Balli et al. (2019) forecast visitor arrivals by controlling regional structural changes. Eugenio-Martin et al. (2004) find that a nexus in the amount of tourist arrivals and economic growth exists in emerging nations, but not in developed nations. To conduct a comprehensive analysis, we divide the sampled data into three subgroups: developing country, non-global financial crisis, and European countries.

Table 1 affords summary statistics of the key variables. We understand from Table 1 that sector returns differ between −0.84% and 5.85% during the sample period, with the median being 0.02% and the mean being 0.12%. The log of average tourism revenue is 11.30%, and the range varies between 0.35% and 79.18% with an 11.24% standard deviation during the sample period, signifying huge differences among countries’ tourism revenues. The log of average tourist arrival is 14.75%. Moreover, the travel and leisure sector returns are positively skewed, signifying that the tail on the right-hand side of the probability density function is heavier than that on the left-hand side. The kurtosis coefficients are larger than 33.02 for sector returns, representing that the series have fatter tails. Moreover, the distributions are not symmetric.

Table 1.

Descriptive statistics.

Variable	Mean	Median	Maximum	Minimum	SD	Skew	Kurtosis	Jarque–Bera	Obs.
SR	0.12	0.02	5.85	−0.84	0.62	4.29	33.02	24,651.04	607
REV	11.30	7.39	79.18	0.35	11.24	2.23	9.42	3374.08	1326
ARR	14.75	14.72	18.28	10.09	1.70	−0.23	2.62	19.58	1359
LL	5.53	5.44	8.02	2.66	1.13	0.11	2.16	44.54	1411
GDP	9.32	9.53	11.47	6.65	1.16	−0.38	2.08	83.91	1411
SS	0.82	0.85	0.99	0.37	0.11	−1.11	3.95	341.83	1410
HLE	63.63	65.50	76.50	32.30	7.61	−0.88	3.21	183.39	1411
FLC	0.74	0.75	0.99	0.26	0.15	−0.55	2.69	74.97	1403
GEN	0.00	−0.02	0.55	−0.32	0.16	0.54	2.85	68.91	1383
PER	0.74	0.81	0.98	0.04	0.20	−1.39	4.29	545.42	1387
NEG	0.26	0.24	0.59	0.09	0.08	0.75	3.96	187.24	1409
CON	0.48	0.46	0.99	0.07	0.20	0.32	2.42	43.36	1370
DEM	−0.02	−0.08	1.58	−2.33	0.86	−0.05	2.07	51.60	1411
TH	6.14	6.13	6.19	6.10	0.03	0.34	1.84	90.25	1190
CPI	110.63	104.89	2740.27	54.36	77.97	28.70	953.19	51,537,562.00	1365
EXG	3.49	3.16	22.63	−1.54	2.84	0.58	3.64	88.71	1224
FDI	5.20	3.04	198.07	−43.46	9.51	7.91	133.25	1,013,471.00	1413
IND	24.77	25.01	29.30	20.49	1.78	−0.12	2.61	7.75	921
INF	5.32	3.55	254.95	−8.97	9.48	15.47	370.79	7,787,425.00	1372
IMP	10.22	10.44	14.67	1.51	2.30	−1.03	5.05	442.21	1258
OIL	35.70	36.26	37.73	30.24	2.05	−1.33	4.49	552.86	1427
UMP	7.33	6.13	31.11	0.29	5.11	1.56	5.94	1083.90	1416

Note: The yearly data in this study span from January 1, 2006 to December 31, 2017. The 10 country happiness indices are LL, GDP, SS, HLE, FLC, GEN, PER, NEG, CON, and DEM. The Jarque–Bera statistics of all variables state departures from normality and present the existence of nonlinear components in the data-generating process. SD: standard deviations; SR: country travel and leisure sector returns; REV: international tourism revenues, % of total exports; ARR: log number of international inbound tourists; LL: life ladder; GDP: log GDP per capita of country; SS: country social support; HLE: country healthy life expectancy; FLC: country freedom to make life choices; GEN: country generosity; PER: country perceived absence of corruption; NEG: country negative effect; CON: country confidence in the national government; DEM: democratic quality; TH: Twitter happiness index; CPI: consumer price index (2010 = 100); EXG: log of official currency exchange rate per US$; FDI: net inflow of foreign direct investment; IND: log of industrial production in constant US$; INF: inflation, consumer prices of annual %; IMP: log of imported merchandise; OIL: crude oil price ratio; UMP: unemployment, total % of total labor force.

Table 2 displays that the correlations between sector returns and tourist arrivals as well as tourist arrivals and tourism revenues are positive, while the link between sector returns and tourism revenues is negative. Among the happiness indices, sector returns positively relate with life ladder, social support, freedom to make life choices, generosity, confidence in government, democratic quality, and Twitter happiness, while they negatively relate with GDP, healthy life expectancy, perceived absence of corruption, and negative effect. Tourism revenues negatively relate to almost all happiness variables except perceived absence of corruption, negative effect, and confidence in government. Nevertheless, tourist arrivals are positively linked with life ladder, GDP, healthy life expectancy, freedom to make life choices, generosity, and democratic quality, while negatively linked with social support, perceived absence of corruption, negative effect, confidence in government, and Twitter happiness. A high correlation (0.79) exists between GDP and healthy life expectancy, signifying different aspects of happiness effects. Twitter happiness has a positive effect on sector returns. OIL is positively related with sector returns, tourism revenues, and tourist arrivals, while EXG is negatively related with them. Previous research indicates a negative influence of oil prices on the tourism sector (e.g. Becken, 2011; Becken and Lennox, 2012; Yeoman et al., 2007). However, Chatziantoniou et al. (2013) pinpoint that these previous results provide a partial depiction, as they do not deliberate the cause of oil price variations; that is, aggregate demand shocks have a meaningfully positive effect on tourism revenues and the economy; conversely, oil-specific demand tremors exercise a significantly negative influence on tourism sector returns. Our positive EXG and tourism results are consistent with Forsyth et al. (2014) in that a greater exchange rate significantly impacts tourism. The panel unit-root test findings display a uniform conclusion that the null of the unit root can be rejected for the levels of the variables, meaning the variables are stationary.

Table 2.

Unconditional correlation.

	SR	REV	ARR	LL	GDP	SS	HLE	FLC	GEN	PER	NEG	CON	DEM	TH	CPI	EXG	FDI	IND	INF	IMP	OIL	UMP
SR	1.00	−0.04	0.00	0.02	−0.02	0.00	−0.03	0.10	0.03	−0.01	−0.01	0.03	0.00	0.05	−0.07	−0.03	−0.02	0.02	0.04	0.06	0.21	−0.05
REV		1.00	0.03	−0.38	−0.34	−0.23	−0.16	−0.24	−0.16	0.32	0.37	0.05	−0.26	−0.02	0.11	−0.29	−0.09	−0.51	0.01	−0.06	0.00	0.28
ARR			1.00	0.17	0.14	−0.01	0.13	0.12	0.14	−0.08	−0.11	−0.10	0.00	−0.06	0.12	−0.20	0.05	0.55	−0.04	0.68	0.02	0.05
LL				1.00	0.65	0.60	0.59	0.50	0.50	−0.59	−0.33	−0.01	0.50	0.02	−0.26	−0.08	0.08	0.36	−0.36	0.07	0.01	−0.15
GDP					1.00	0.61	0.79	0.25	0.26	−0.60	−0.44	−0.08	0.65	−0.01	−0.24	0.10	0.27	0.20	−0.38	−0.03	0.01	−0.05
SS						1.00	0.53	0.39	0.28	−0.33	−0.38	−0.16	0.51	0.04	−0.28	−0.20	0.09	0.03	−0.42	−0.13	−0.05	−0.24
HLE							1.00	0.35	0.36	−0.56	−0.35	−0.16	0.70	−0.03	−0.26	0.04	0.30	0.05	−0.57	0.13	0.03	0.05
FLC								1.00	0.62	−0.59	−0.37	0.38	0.34	−0.05	−0.07	0.02	0.13	0.23	−0.31	0.14	0.06	−0.49
GEN									1.00	−0.45	−0.26	0.26	0.26	0.05	−0.15	0.01	0.17	0.16	−0.27	0.11	0.01	−0.35
PER										1.00	0.38	−0.44	−0.46	−0.01	0.11	−0.10	−0.33	−0.14	0.18	−0.10	−0.02	0.20
NEG											1.00	−0.16	−0.27	−0.07	0.28	−0.27	−0.16	−0.20	0.25	−0.08	0.01	0.23
CON												1.00	−0.34	0.04	0.11	0.27	0.16	−0.08	0.20	−0.08	−0.02	−0.32
DEM													1.00	0.03	−0.39	−0.03	0.21	0.07	−0.54	−0.04	−0.01	−0.06
TH														1.00	−0.22	−0.01	−0.06	−0.02	−0.07	−0.03	−0.21	0.03
CPI															1.00	0.00	−0.08	0.07	0.42	0.07	0.15	0.06
EXG																1.00	0.25	−0.03	0.03	−0.11	−0.03	−0.25
FDI																	1.00	−0.16	−0.08	−0.13	0.01	−0.20
IND																		1.00	0.09	0.52	0.01	−0.12
INF																			1.00	0.04	0.08	0.04
IMP																				1.00	0.04	0.16
OIL																					1.00	0.00
UMP																						1.00

Note: Yearly data for the period 2006–2017. SR: country travel and leisure sector returns; REV: international tourism revenues, % of total exports; ARR: log number of international inbound tourists; LL: life ladder; GDP: log GDP per capita of country; SS: country social support; HLE: country healthy life expectancy; FLC: country freedom to make life choices; GEN: country generosity; PER: country perceived absence of corruption; NEG: country negative effect; CON: country confidence in the national government; DEM: democratic quality; TH: Twitter happiness index; CPI: consumer price index (2010 = 100); EXG: log of official currency exchange rate per US$; FDI: net inflow of foreign direct investment; IND: log of industrial production in constant US$; INF: inflation, consumer prices of annual %; IMP: log of imported merchandise; OIL: crude oil price ratio; UMP: unemployment, total % of total labor force.

Models

Chiu and Yeh (2017) identify that if one disregards the possibility that the tourism–growth nexus could be nonlinear, then the results of a linear model frequently cause bias due to a false estimate method. Baggio and Sainaghi (2011) denote that tourism system is nonlinear and complex. Sainaghi and Baggio (2019) confirm the tourism system is nonlinear. Meo et al. (2018) also suggest a long-term asymmetric association among inflation, exchange rate, oil prices, and tourism demand. Conventional OLS offers summary point estimations for the average outcome of the explanatory factors (Binder and Coad, 2011). Concentrating on the average effects may under- or overestimate the related coefficient estimations or may even fail to identify imperative relations (Binder and Coad, 2011). Lew and Ng (2012) use a QR measure on tourist spending in diverse kinds against a fixed range of dependent variables and show that it is less susceptible to impact by outlier values and is better able to aim finer tourist spending market parts. Gunderson and Ng (2005) analyze the influences of quality of life attributes and tourism on regional economic development via QR.

Gunderson and Ng (2005) present that OLS provides nothing more than an assessment of the average of the dependent variable conditioned on the independent variables. Due to the innate heterogeneity in financial markets, the links between market returns and independent variables might differ across their condition distributions (Badshah, 2013; Lee and Chen, 2020). Heterogeneity is generally greater under unstable market situations. OLS, which captures the relationships at the average, therefore might lead to a misspecification, and information around the tails of a distribution is ignored. To account for these concerns, we usage QR as proposed by Koenker and Bassett (1978), which is more robust and consequently offers more efficient assessments, since it permits us to comprise a full range of conditional quantile functions (Chiang et al., 2010; Lee and Chen, 2020). Moreover, it is strong to heteroscedasticity, skewness, and leptokurtosis, which are general features of financial data (Baur et al., 2012). Thus, the QR model primarily helps to estimate whether country and Twitter happiness variables influence tourist arrivals, tourism revenues, and sector returns by intensifying upon the descriptive statistics in Table 1 and testing equations (1 ) to (3) by using tourist arrivals, tourism revenues, and sector returns as dependent variables, respectively.

Before identifying the QR model for the assessments, the benchmark OLS, which is accustomed for heteroscedasticity (White cross-section standard errors), has the following form

{ARR}_{i t} = α_{0} + β_{1} {CH}_{i, t} + β_{2} {TH}_{i, t} + β_{3} {CV}_{i, t} + ε_{i t}

{REV}_{i t} = α_{0} + β_{1} {ARR}_{i, t} + β_{2} {CH}_{i, t} + β_{3} {TH}_{i, t} + β_{4} {CV}_{i, t} + ε_{i t}

{SR}_{i t} = α_{0} + β_{1} {REV}_{i, t} + β_{2} {ARR}_{i, t} + β_{3} {CH}_{i, t} + β_{4} {TH}_{i, t} + β_{5} {CV}_{i, t} + ε_{i t}

where ARR _it , REV _it , and SR _it respectively denote country i’s tourist arrivals, tourism revenues, and travel and leisure sector returns in time t. CH represents 10 host country happiness variables: life ladder, GDP per capita, social support, healthy life expectancy, freedom to make life choices, generosity, corruption perception, country negative effect, confidence in the national government, and democratic quality. TH represents the Twitter happiness index extracted from Twitter. The vector of CV covers eight control variables that might influence tourist arrivals, tourism revenues, and sector returns: CPI, official currency exchange rate, foreign direct investment, industrial production, inflation, import, crude oil price, and unemployment.

We next conduct equations (1 ) to (3) by the QR models. We consider the impacts of host country and Twitter happiness variables on tourism variables’ dynamics by allowing for that the conditional ∂ quantile of Ys’ (tourism) variation distribution (y_t), $y (t), Q_{y} (\partial | x)$ , is influenced by the effects from happiness variables and other control variables. Therefore, we identify the ∂th conditional quantile function of tourism y as follows

Q_{y} (\partial | x) = inf {γ | F_{y} (γ | x) \geq \partial} = \sum_{h} α_{h} (\partial) x_{h} = x α (\partial)

where $F_{y} (γ | x)$ is the conditional distribution function of y given x, and the QR coefficient $α (\partial)$ governs the dependence association between vector x and the ∂th conditional quantile of y. The values of $α (\partial)$ for $\partial \in [0, 1]$ govern the complete dependence structure of y. The coefficients $α (\partial)$ for a given ∂ are assessed by diminishing the weighted absolute deviations between y and x

\hat{α} (\partial) = arg min \sum_{t = 1}^{T} (\partial - 1_{{y_{t} < x_{t}' α (\partial)}}) | y_{t} - x_{t}' α (\partial) |

where $1_{{y_{t} < x_{t}^{'} α (\partial)}}$ is the typical indicator function. The resolution to this difficult is attained via the linear programming algorithm proposed by Koenker and D’Orey (1987). The standard errors for the appraised coefficients can be acquired from the pair bootstrapping procedure suggested by Buchinsky (1998), because it offers standard errors that are asymptotically valid under heteroscedasticity and misspecifications of the QR model.

In the situation of linear dependence on a vector of exogenous variables (X), we can inscribe the linear conditional quantile function as

Q_{y} (\partial | x) = β (\partial) + \sum_{h} γ_{h} (\partial) x_{h}

We evaluate tourism variables across five quantiles $q = {0.1; 0.25; 0.5; 0.75; 0.9}$ . The values of $γ (\partial)$ describe the variable effects. It is worth highlighting that this is the initial time that an analysis of tourism, Twitter, and country happiness indices is executed within the QR framework for international evidence. Prior studies based their findings as to the presence of varying effects on questionnaire survey happiness data within a single country or a limited number of countries.

Results and discussion

Impacts of the happiness indices on tourist arrivals

As a benchmark to our quantile analysis in equation (1), we first display estimates from OLS regressions of international tourist arrivals of time t on the independent variables. To clearly identify the influences of host country and Twitter happiness indices, we separate the results of the two variables in Table 3, which shows estimates of the slope coefficients β for each predictor factor. The standard errors for the quantile assessments are accomplished by the bootstrapping approach. The left-hand side of Table 3 displays that Twitter happiness has no significant influence on tourist arrivals, and thus we reject hypothesis 1a that Twitter happiness has significant effects on international tourist arrivals. However, we uncover findings from the right-hand side that host country happiness saliently affects tourist arrivals. Life ladder negatively and significantly predicts returns at the 75th and lower quantiles, implying that the host country’s citizens in the lower life ladder affect more international tourist arrivals, but life ladder does not affect the 90th quantile’s tourist arrivals. We consider this situation to be quite reasonable under the tourism–economic growth nexus in that people with a low life ladder are willing to appeal more tourist arrivals and vice versa.

Table 3.

Estimates of the QR-based international tourist arrival models.

Variables	OLS	Q10	Q25	Q50	Q75	Q90	OLS	Q10	Q25	Q50	Q75	Q90
C	4.794	7.620	4.253	3.224	4.864	13.787	1.029	−2.441	1.162	1.636	0.824	3.905**
LL							−0.257***	−0.242**	−0.208**	−0.245***	−0.213***	−0.017
GDP							0.276***	0.425	0.557**	0.261*	0.124	0.653**
SS							−0.059	0.170	−0.480	1.730**	0.351	−1.685
HLE							−0.003	0.001	−0.043	−0.001	0.010	−0.048
FLC							1.783***	3.103***	2.050***	0.723	1.102*	−0.146
GEN							0.528***	1.224*	0.626	0.205	−0.096	−0.241
PER							0.352**	0.114	0.858**	0.259	0.312	0.926***
NEG							−0.373	2.323**	0.787	−1.159	−3.195***	−3.580***
CON							0.110	−0.390	−0.194	0.246	0.382	0.413
DEM							−0.041	−0.348*	0.024	−0.001	0.014	−0.032
TH	−0.553	−1.127	−0.636	−0.236	−0.263	−1.598
CPI	0.007***	0.009*	0.009***	0.008***	0.005*	0.004**	0.005***	0.004	0.005*	0.007***	0.008***	0.006***
EXG	−0.099***	−0.154**	−0.061**	−0.076***	−0.085***	−0.114***	−0.117***	−0.097**	−0.075***	−0.081***	−0.072***	−0.099***
FDI	0.038***	0.037***	0.046***	0.042***	0.028***	0.027***	0.028***	0.029***	0.027***	0.018*	0.016**	0.020**
IND	0.509***	0.541***	0.521***	0.476***	0.455***	0.445***	0.432***	0.377***	0.318***	0.353***	0.493***	0.376***
INF	−0.023***	−0.078***	−0.071***	−0.016	−0.006	0.004	−0.012	−0.044***	−0.039**	−0.008	0.000	0.004
IMP	0.126***	0.143**	0.131*	0.151***	0.114***	0.092***	0.133***	0.225**	0.275***	0.180***	0.066**	0.076***
OIL	−0.020	−0.050	−0.014	−0.017	−0.005	0.001	−0.022**	−0.028	−0.019	−0.016	−0.004	0.005
UMP	0.013***	0.002	−0.001	0.013	−0.004	0.009	0.024***	0.021*	0.005	0.007	0.034**	0.018
Adj./Pseudo R²	0.633	0.390	0.430	0.456	0.437	0.408	0.631	0.453	0.440	0.467	0.455	0.433

Note: This table reports the estimation results of host country happiness indices and Twitter happiness index impacts on the log of international tourist arrivals according to equation (1). Yearly data are used for the period 2006–2017. Standard errors are bootstrapped replications. The 10 host country happiness variables are LL, GDP, SS, HLE, FLC, GEN, PER, NEG, CON, and DEM. TH represents the Twitter happiness index extracted from Twitter. Control variables are the eight control variables that might influence ARR, REV, and SR: CPI, EXG, FDI, IND, INF, IMP, OIL, and UMP. LL: life ladder; GDP: GDP per capita; SS: social support; HLE: healtdy life expectancy; FLC: freedom to make life choices; GEN: generosity; PER: corruption perception; NEG: negative effect; CON: confidence in tde national government; DEM: democratic quality; CPI: consumer price index (2010 = 100); EXG: log of official currency exchange rate per US$; FDI: net inflow of foreign direct investment; IND: log of industrial production in constant US$; INF: inflation at consumer prices in annual %; IMP: log of imported merchandise; OIL: crude oil price ratio; UMP: unemployment as % of total labor force; SR: country travel and leisure sector returns; REV: international tourism revenues, % of total exports; ARR: log number of international inbound tourists; OLS: ordinary least squares; QR: quantile regression.

* Significance at the 10% level.

** Significance at the 5% level.

*** Significance at the 1% level.

GDP positively influences the 25th, 50th, and 90th quantiles, signifying that greater GDP affects higher quantile tourist arrivals. Freedom to make life choices, generosity, and perceived absence of corruption also have positive impacts on tourist arrivals. Negative effect has a positive impact at the 10th quantile and a saliently negative impact at 75th and above quantiles, representing that a negative tourist arrival effect in a host country with greater tourist arrivals decreases that country’s well-being. This finding is in line with Ivlevs (2017) that tourist arrivals diminish citizens’ life satisfaction and with Bimonte and Faralla (2016) that tourism has an unseen cost in terms of perceived life satisfaction. The findings partly support hypothesis 1b that host country happiness has significant effects on international tourist arrivals. Specifically, 8 out of 10 country happiness indices saliently impact tourist arrivals (excluding healthy life expectancy and confidence in government). The empirical results indicate strong evidence of asymmetric relations between tourist arrivals and host country happiness indices. The findings do not confirm the assessment of a one-size-fits-all and thus support that the relationship between happiness and tourist arrivals varies at different arrival distributions.

Regarding the control variables, CPI, FDI, industrial production, import, and unemployment are positively and significantly related with tourist arrivals, EXG and inflation are negatively related with tourist arrivals, and OIL has no significant relation with tourist arrivals. Specifically, FDI, industrial production, and imports show a positive impact on tourist arrivals, while EXG shows a negative impact at all quantiles, meaning that a linear model fits the economic growth and tourist arrival nexus. Moreover, FDI, industrial production, and imports influence tourist arrivals, which is consistent with Chatziantoniou et al. (2013) that using industrial production as an economic growth proxy shows that tourism is closely related with economic growth. Habibi (2015) presents that tourism, FDI, and economic growth are cointegrated. Our result is in line with Hanafiah et al. (2011) that tourism demand negatively correlates with the exchange rate. In addition, Forsyth et al. (2014) find that a higher exchange rate poses significant problems for tourism. Consistent with Hanafiah and Harun (2012) that the inflation rate reduces the number of tourist arrivals, we further find a notable and negative relation under the 25th quantile for tourist arrivals. Thus, tourist arrivals are sensitive to macroeconomic variables.

Figure 1 displays the marginal effect of happiness variables for all quantiles within the (0,1) range of tourist arrivals. Figure 1 also gives us a consistent result for Twitter happiness whereby no noteworthy coefficients in any of the QRs are stated. Comparatively, country happiness variables are more related to tourist arrivals. Generally, we find that host country happiness affects tourist arrivals more at lower quantiles than at higher ones. Therefore, increasing GDP, freedom to make life choices, generosity, perceived absence of corruption, and negative effect should be considered as a major policy objective in countries with low tourist arrivals so as to improve their incoming tourists.

Figure 1.

Quantile regression estimates with 95% confidence intervals for the impacts of happiness on tourist arrivals. Note: The blue line shows the quantile regression estimates for the quantile ranging from 0.1 to 0.9. The red line depicts 95% confidence intervals for the quantile regression estimates. LL: life ladder; GDP: GDP per capita; SS: social support; HLE: healthy life expectancy; FLC: freedom to make life choices; GEN: generosity; PER: corruption perception; NEG: negative effect; CON: confidence in the national government; TH: Twitter happiness index.

Impacts of tourist arrivals and happiness indices on tourism revenues

Table 4 indicates quantile estimates from equation (2). Figure 2 states QR parameter assessments along with the 95% confidence intervals (red lines) for the predictive power of Twitter and country happiness variables’ effect on tourism revenues. Figure 2 exhibits that the estimated confidence intervals of perceived absence of corruption and confidence in government are reduced, indicating that these variables can function as determinants in clarifying tourism revenues. Moreover, most of the host country happiness indices show positive impacts on tourism revenues. The results in Table 4 show that tourist arrivals saliently and positively influence tourism revenues, supporting hypothesis 2a that international tourist arrivals have significant and positive effects on international tourism revenues. This result matches with Akal (2004) that tourism revenues can be illuminated by present tourist arrivals. However, Twitter happiness does not significantly affect tourism revenues, thus rejecting hypothesis 2b that Twitter happiness has a significant effect on tourism revenues.

Table 4.

Estimates of the QR-based international tourism revenue models.

Variables	OLS	Q10	Q25	Q50	Q75	Q90	OLS	Q10	Q25	Q50	Q75	Q90
C	57.534**	19.720	6.972	51.502	70.837	117.954	−8.227	4.497	−0.091	−10.488	−15.865*	−24.802**
ARR	3.293***	0.642***	0.994***	1.793***	3.418***	4.488***	3.150***	1.286***	1.390***	1.893***	2.412***	2.395***
LL							0.043	0.338	−0.551*	−0.296	0.165	0.202
GDP							0.053	−0.808*	−0.477	−0.973	−0.064	2.963*
SS							−6.403**	3.600	5.981*	−2.361	−17.607**	−18.435**
HLE							0.624***	0.228***	0.270***	0.645***	1.003***	0.882***
FLC							5.273	1.775	6.720***	4.288	2.809	−1.026
GEN							−0.327	−0.654	1.748	2.128	4.109**	6.692**
PER							14.498***	4.551***	6.009***	11.663***	20.891***	27.464***
NEG							27.686***	6.736***	8.988***	19.038***	20.055**	36.575***
CON							12.417***	−0.163	1.025	8.176***	14.770***	20.649***
DEM							−0.137	−0.200	−0.776*	−0.103	−0.502	0.914
TH	−1.337	−0.084	3.726	−2.541	−3.588	−9.159
CPI	−0.002	0.001	0.001	0.017	0.002	−0.016	−0.020***	−0.017*	−0.012	−0.018*	−0.015	−0.054***
EXG	−0.155***	−0.226***	−0.248***	−0.075	0.001	−0.452**	−0.312***	−0.036	−0.009	−0.340***	−0.583***	−0.836***
FDI	−0.130***	0.014	−0.033**	−0.112***	−0.194***	−0.272***	−0.092***	0.001	−0.046*	−0.078**	−0.070**	−0.162***
IND	−3.510***	−1.012***	−1.539***	−2.250***	−3.639***	−4.408***	−3.548***	−1.471***	−1.612***	−2.174***	−3.345***	−3.747***
INF	−0.057	−0.085***	−0.076**	−0.031	−0.082	−0.043	0.051	−0.037	−0.039	0.077	−0.035	0.017
IMP	−0.050	0.269***	0.281***	0.004	−0.095	−0.316	−0.097	0.237***	0.160*	−0.107	−0.152	−0.430
OIL	−0.134	−0.078	−0.124**	−0.075	−0.068	−0.084	−0.145	−0.088**	−0.112**	−0.102	−0.191	−0.160
UMP	0.420***	0.005	0.018	0.147	0.933***	0.963***	0.263***	−0.019	0.049	0.082	0.694***	0.669***
Adj./Pseudo R²	0.389	0.161	0.152	0.184	0.291	0.430	0.512	0.225	0.217	0.262	0.367	0.503

Note: This table reports the estimation results of host country happiness indices and Twitter happiness index impacts on international tourism revenues according to equation (2). Monthly data are used for the period 2006–2017. The same note as Table 3.

Figure 2.

Quantile regression estimates with 95% confidence intervals for the impacts of tourist arrivals and happiness on tourism revenues. Note: The same as Figure 1. ARR: log number of international inbound tourists.

Regarding country happiness indices, healthy life expectancy, perceived absence of corruption, and negative effect remarkably and positively impact tourism revenues at all quantiles. GDP has a considerably negative influence on tourism revenues at the 10th quantile, but a considerably positive influence at the 90th quantile, suggesting that GDP leads high quantile tourism revenues to a better performance, while GDP leads lower quantile tourism revenues worsen tourism revenues. Çağlayan et al. (2012) denote that the association between tourism revenue and GDP varies with geographic areas. Ekanayake and Long (2012) indicate that the elasticity of tourism revenue with respect to real GDP is not statistically substantial for all counties; its positive sign reveals that tourism receipt positively supports economic growth in emerging nations. Using Spain (developed nation) as a sample nation, Albaladejo et al. (2014) find that tourism growth and real GDP have a bi-causal relationship. Therefore, the relation between GDP and tourism revenues is inconsistent and asymmetric. Social support has a negatively significant impact at 75th and above quantiles, but has positively significant impacts at the 0.25 quantile. This suggests that social support leads a higher tourism revenue country to worse tourism revenues, while it leads a lower quantile tourism revenue country to more tourism revenues. Healthy life expectancy (perceived absence of corruption) is constantly, significantly, and positively related with tourism revenues at all quantiles, indicating that a higher healthy life expectancy (perceived absence of corruption) is an incentive for more tourism revenues. Generosity (confidence in government) positively affects tourism revenues at the (intermediate and) higher quantiles. Negative effect is significantly positive at all quantiles, signifying that a country with more unhappiness tends to promote tourism revenues. Therefore, our findings support hypothesis 2c that host country happiness has significant effects on tourism revenues.

We draw some interesting findings from Table 4. First, Twitter happiness has no impact on tourism revenues. Second, healthy life expectancy, perceived absence of corruption, and negative effect positively impact all tourism revenues quantiles. Third, the influences of life ladder, GDP, social support, freedom to make life choices, generosity, confidence in government, and democratic quality on tourism revenues are asymmetric across quantiles, supporting hypothesis 4 that the relationship between happiness and tourism revenues varies at different revenue quantile distributions. Compared to the influences of host country happiness variables in Tables 3 and 4, the country happiness indices show more salient impacts on tourism revenues than on tourist arrivals. Therefore, one possible reason for the importance of country happiness on tourism revenues than on tourist arrivals is that tourism revenues are a more direct variable to estimate tourism development.

Regarding control variables, CPI, EXG, FDI, and OIL show negative impacts on tourism revenues at some of the quantiles, implying a worsening economy might stimulate better tourism revenues. Industrial production has constant, salient, and negative impacts on all tourism revenues quantiles, and thus no matter how tourism revenue performs, worse industrial production correlates to increasing tourism revenues. However, unemployment has a positive impact on 75th and above quantiles, indicating that the unemployment problem decreases the costs of travel and leisure sector operations and then increases tourism revenues, especially for higher tourism revenues quantiles. Likewise, increasing oil prices raises operation costs in the travel and leisure sector, and thus OIL has a negative impact on tourism revenues.

Impacts of tourist arrivals, tourism revenues, and happiness indices on travel and leisure sector returns

Table 5 displays the assessments of the QR model introduced by equation (3). Figure 3 reveals a graphical demonstration of the point estimates of the model parameters. Demir et al. (2017) reveal that foreign tourist arrivals enhance tourism stock returns. However, Chen (2010) finds that foreign tourist arrivals do not have a substantial effect on hotel stock returns. Our findings show that tourist arrivals significantly and positively impact at the 10th quantile. In other words, tourist arrivals positively influence sector returns only for a country’s travel and leisure sector returns in the lowest quantile. The explanatory power of tourist arrivals decreases for higher quantiles, meaning that for higher sector returns the same rise in tourist arrivals results in a larger decrease in sector returns. Our finding is more consistent with Chen (2010) that tourist arrivals do not have a salient influence on travel and leisure sector returns. Hence, our result merely supports hypothesis 3a that international tourist arrivals have significant effects on travel and leisure sector returns in the lowest sector return quantiles. Further tests are displayed in robustness checks.

Table 5.

Estimates of the QR-based travel and leisure sector return models.

	SR						SR
Variables	OLS	Q10	Q25	Q50	Q75	Q90	OLS	Q10	Q25	Q50	Q75	Q90
C	−11.253	−1.010	−6.156	−9.747**	−12.038	−30.337*	−1.539*	−5.098***	−2.313	−0.421	0.431	2.942
REV	−0.003	0.008	0.004	0.002	0.001	0.005	−0.005*	−0.009	0.001	−0.003	−0.004	0.000
ARR	−0.012	−0.004	0.019	0.003	0.042	−0.015	−0.023	0.094*	0.009	0.009	−0.007	−0.057
LL							0.024	−0.006	−0.031	−0.022	−0.011	0.154
GDP							0.118	0.051	0.080	0.053	0.054	−0.089
SS							−0.413	−0.190	−0.406	−0.874*	−0.916	−0.921
HLE							−0.011	0.016	−0.003	−0.002	−0.007	−0.003
FLC							0.842	0.451	0.596	0.579	0.539	0.732
GEN							−0.068	−0.188	0.063	0.012	0.005	0.037
PER							0.374	0.134	0.030	0.152	0.393*	0.407
NEG							0.287	1.301*	0.278	−0.041	−0.511	−1.211
CON							0.097	0.257	−0.129	0.042	0.145	−0.842
DEM							0.012	−0.017	−0.030	0.006	0.022	−0.046
TH	1.612	−0.254	0.516	1.293*	1.785	4.659*
CPI	−0.005	0.001	−0.002	−0.006**	−0.008**	−0.002	−0.005	−0.003	−0.003	−0.006**	−0.006*	−0.005
EXG	−0.009	0.015	0.006	−0.010	0.002	0.011	−0.016	−0.004	−0.009	−0.024*	−0.021	0.005
FDI	−0.002	0.003	0.001	0.003	−0.003	−0.009	0.001	0.000	0.002	0.001	0.002	0.004
IND	−0.017	0.046	0.026	0.023	−0.023	−0.034	−0.044	0.016	0.019	−0.003	−0.025	−0.119
INF	0.018	−0.031***	−0.022*	0.009	0.036**	0.058**	0.014	−0.020	−0.020	0.003	0.017	0.069
IMP	0.027	−0.032**	−0.010	0.005	0.013	0.028	0.038**	−0.039*	−0.014	0.008	0.039**	0.067**
OIL	0.068***	0.036**	0.061***	0.049***	0.056***	0.092***	0.055**	0.040***	0.043****	0.028**	0.027*	0.057*
UMP	−0.011**	0.001	−0.004	−0.009*	−0.008	−0.015	−0.001	0.002	0.000	−0.004	−0.003	−0.014
Adj./Pseudo R²	0.053	0.140	0.099	0.059	0.091	0.132	0.021	0.156	0.096	0.064	0.095	0.117

Note: This table reports the estimation results of host country happiness indices and Twitter happiness index impacts on travel and leisure sector returns of t and $t + 1$ periods according to equation (3). Yearly data are used for the period 2006–2017. The same note as Table 3.

Figure 3.

Quantile regression estimates with 95% confidence intervals for the impacts of tourist arrivals, tourism revenue, and happiness on travel and leisure sector returns. Note: The same as Figure 1. REV: international tourism revenues; ARR: log number of international inbound tourists.

Chen (2010) indicates that neither real GDP growth nor the growth rate of international tourists have a substantial effect on hotel stock performance. He attributes the immaterial effect of tourism growth on hotel stock returns to a time-varying discount rate initiated by investors’ changing opinion about the riskiness of cash flows. Additionally, Chen (2010) finds that both economic and industry issues are positive and substantial explanatory factors of tourist hotels’ whole financial performance.

As shown by Table 5, tourism revenues have an important and negative influence on sector returns only in the OLS results. Thus, our QR results reject hypothesis 3b that international tourism revenues have significant effects on travel and leisure sector returns. Being diverse from Chen (2010), our finding can be explained due to the different tourism revenue variables and different coverage of sectors. Twitter happiness, one of our central explanatory variables, is estimated to have a positive effect on sector returns at the 50th and 90th quantiles, albeit statistically indistinguishable at other quantiles. Thus, our finding partly supports hypothesis 3c that Twitter happiness has significant effects on travel and leisure sector returns.

Figure 3 displays that most country happiness indices show insignificant impacts on sector returns. Regarding these country happiness indices, social support is negative at all quantiles, while it remarkably and negatively affects sector returns at the 50th quantile. This means that social support has a negative impact on most of the quantiles, especially at intermediate sector return quantiles. Perceived absence of corruption has a considerably positive effect on sector returns at the 75th quantile, suggesting that perceived absence of corruption leads the 75th tourism revenues quantile to better sector returns. Negative effect is considerably and positively related with sector returns at the 10th quantile, representing that a higher negative effect seems to denote a spur for more sector returns. Therefore, our findings tend to reject hypothesis 3d that host country happiness has significant effects on travel and leisure sector returns.

We draw some interesting findings from Table 5. First, tourism revenues have little impact on sector returns. Second, tourist arrivals positively affect sector returns at the lowest sector return quantile. Third, Twitter happiness shows a significantly positive effect at some higher quantiles, signifying the influences of Twitter happiness on sector returns are asymmetric across quantiles. Fourth, host country happiness indices show less salient impacts on sector returns, compared to the influences of Twitter happiness.

Findings for the control variables comprised in the model are also informative. OIL shows constantly positive impacts on sector returns at all quantiles, implying higher OIL might positively stimulate increases in sector returns. Mohanty et al. (2014) discover that fluctuations in oil prices have a substantial effect on the US travel and leisure sector returns. Our finding is consistent with Kristjanpoller and Concha (2016) that a rise in oil prices is trailed by a rise in airline stock prices. CPI has a salient and negative impact on intermediate and higher (50th–75th) sector return quantiles, which is consistent with Chen (2007) that CPI has a negative influence on hotel stock returns. Inflation and imports show negative and substantial effects at the lower quantiles, but are positive and substantial at the upper quantiles, implying the asymmetric effects of inflation and imports on sector returns.

The negative impact of unemployment in the OLS result is similar with Chen et al. (2005), who explore the relationship between macroeconomic factors (i.e. expected inflation, growth rate of industrial production, and change in the unemployment rate) and hotel stock returns, finding only that the unemployment rate considerably (and negatively) explains the movement of hotel stock returns. Additionally, Demir et al. (2017) find that CPI as well as imports Granger-cause a tourism firm’s stock returns in Turkey; after allowing for a structural break in 2007, the pre-break findings reveal that CPI, EXG, and tourist arrivals could Granger-cause sector returns, while the findings in the post-structural break period show only increases in OIL and import. We also find salient impacts of CPI, imports, EXG, and tourist arrivals on sector returns. In the following section of robustness checks, we further test the impact of geographic area, non-global crisis subperiod, different variable proxies, and country development condition. In sum, one possible reason for the insignificance of country happiness on sector returns is due to the time-varying discount rate triggered by investors’ changing outlooks over the prospect of future cash flows from holding travel and leisure stocks (Chen, 2011).

Robustness checks

Effect of Twitter and host country happiness indices on present and subsequent arrivals

The results in Table 3 do not address issues regarding endogeneity. To avoid the endogeneity problem and incorporate the main variables into the models, we advance tourist arrivals by one period to avoid simultaneity in Table 6 and Figure 4. This advance allows for the effect of any change in happiness indices to show up in tourist arrivals. Figure 4 demonstrates the effects of happiness indices on tourist arrivals, whereby most happiness variables are positively related to tourist arrivals. Among the country happiness indices, life ladder is the only index that constantly negatively correlates with tourist arrivals. We notice that the impacts of negative effect, no matter in tourist arrivals for concurrent or subsequent periods of Table 6 as well as in Table 3, are negative at upper quantiles and positive at lower quantiles.

Table 6.

Estimates of the QR-based international tourist arrivals of t and $t + 1$ periods’ models.

	ARR _t						${ARR}_{t + 1}$
Variables	OLS	Q10	Q25	Q50	Q75	Q90	OLS	Q10	Q25	Q50	Q75	Q90
C	2.218	3.478	0.397	1.346	2.103	9.514	0.707	−0.051	−3.597	1.281	6.294	9.186
LL	−0.288***	−0.341***	−0.305***	−0.289***	−0.204**	−0.062	−0.303***	−0.347**	−0.339***	−0.290***	−0.244***	−0.105
GDP	0.322***	0.320	0.640***	0.442**	0.142	0.724**	0.293***	0.174	0.645***	0.332**	0.058	0.353
SS	−0.123	0.138	−1.289	1.329	0.315	−1.502	−0.175	0.317	−1.175	0.750	0.310	−1.110
HLE	−0.002	0.021	−0.017	−0.013	0.022	−0.043	−0.003	0.022	−0.022	−0.022	0.009	−0.015
FLC	1.861***	3.214***	2.449***	1.198*	1.016	0.028	1.797***	2.963***	2.397***	1.238**	1.017	−0.491
GEN	0.483***	1.614**	0.352	0.400	−0.207	0.041	0.530***	2.195***	0.553	0.244	−0.019	−0.005
PER	0.260	0.122	0.344	0.487	0.378	0.941**	0.175	−0.637	0.319	0.023	0.311	0.869**
NEG	0.023	3.183***	2.188**	−0.404	−3.176***	−3.277***	−0.163	3.287**	2.339**	−0.375	−2.451**	−3.982***
CON	0.043	−0.894	−0.436	0.214	0.368	0.125	0.070	−0.999	−0.548	0.164	0.632	0.584
DEM	−0.048	−0.402**	−0.027	−0.060	−0.020	−0.142	−0.038	−0.367	−0.011	0.006	0.146	0.082
TH	−0.274	−1.066	−0.398	−0.094	−0.369	−0.940	0.024	−0.365	0.378	0.204	−0.823	−1.011
CPI	0.005***	0.003	0.003	0.007**	0.007**	0.003	0.006***	0.002	0.003	0.007*	0.008**	0.007**
EXG	−0.116***	−0.105**	−0.075***	−0.093***	−0.076***	−0.091***	−0.109***	−0.107***	−0.069***	−0.079***	−0.061***	−0.098***
FDI	0.032***	0.032***	0.029**	0.030**	0.017*	0.019*	0.031***	0.024**	0.021*	0.032***	0.020*	0.020
IND	0.445***	0.444***	0.457***	0.347***	0.491***	0.362***	0.455***	0.505***	0.444***	0.392***	0.507***	0.460***
INF	−0.016*	−0.043**	−0.038**	−0.018	0.004	0.004	−0.020**	−0.035	−0.044***	−0.024	−0.008	−0.003
IMP	0.126***	0.179	0.132	0.172***	0.066*	0.064**	0.115***	0.109	0.135	0.146***	0.072**	0.072***
OIL	−0.024**	−0.033	−0.019	−0.014	0.000	0.003	−0.020*	−0.017	−0.014	−0.017	−0.013	−0.001
UMP	0.025***	0.015	0.003	0.009	0.030	0.018	0.027***	0.020	0.002	0.017	0.042**	0.031
Adj./Pseudo R²	0.634	0.463	0.454	0.468	0.452	0.435	0.632	0.457	0.453	0.470	0.459	0.447

Note: This table reports the estimation results of host country happiness indices and Twitter happiness index impacts on international tourist arrivals of t and $t + 1$ periods according to equation (1). Yearly data are used for the period 2006–2017. The same note as Table 3.

Figure 4.

Quantile regression estimates with 95% confidence intervals for the impacts of tourist arrivals and happiness on tourism revenues of the subsequent period. Note: The same as Figure 1. ARR: log number of international inbound tourists.

Our QR results in Table 6 display the following. (1) The host national happiness variables have significantly asymmetric impacts on tourist arrivals, thus supporting hypothesis 1b. (2) Twitter happiness has no notable impact on tourist arrivals, thus rejecting hypothesis 1a. (3) The constantly negative impact of life ladder demonstrates that when a country’s citizens are at lower life ladder quantiles, the host country might increase tourist arrivals. (4) Higher GDP, freedom to make life choices, generosity, and perceived absence of corruption notably increase tourist arrivals at some quantiles. Again, our findings display the asymmetry of country happiness indices’ effect on tourist arrivals. (5) The QR models of concurrent and subsequent tourist arrivals have quite similar signs and significant results. We also test the problem of endogeneity to incorporate the advanced tourism revenues by one period to avoid simultaneity in Table 7. We obtain similar findings with those of Table 4. Our findings illustrate that host country happiness indices have both concurrent and lagged effects on tourist arrivals and tourism revenues.

Table 7.

Estimates of the QR-based international tourism revenues of t and $t + 1$ periods’ models.

	REV _t						${REV}_{t + 1}$
Variables	OLS	Q10	Q25	Q50	Q75	Q90	OLS	Q10	Q25	Q50	Q75	Q90
C	−29.504	−6.821	−10.852	−24.584	3.299	−0.111	−15.177	10.223	−43.018	−36.414	17.042	49.758
ARR	3.112***	1.067***	1.402***	1.772***	2.293***	2.283***	3.013***	0.822**	1.259***	1.735***	2.295***	2.293***
LL	−0.277	−0.142	−0.703**	−0.777	−0.115	0.053	−0.137	−0.286	−0.428	−0.537	−0.204	0.607
GDP	0.232	−0.916**	−0.547	−0.515	−0.365	3.483**	0.278	−0.498	−0.077	−0.195	−0.515	1.271
SS	−7.062**	6.142**	5.504	−1.235	−16.186*	−21.812**	−7.179**	5.378*	2.729	−3.516	−13.698	−16.891**
HLE	0.627***	0.242***	0.282***	0.562***	1.042***	0.858***	0.693***	0.271***	0.252**	0.533***	1.079***	0.986***
FLC	7.873**	2.446	6.202**	7.195*	4.007	4.305	7.616**	3.098	5.947**	8.961**	6.327	1.436
GEN	0.275	0.107	2.458*	2.114	5.027**	7.665***	0.653	0.748	2.960**	2.465	6.156**	8.350***
PER	15.413***	5.093***	6.340***	9.473***	20.857***	29.481***	17.028***	5.877***	7.030***	10.980***	22.566***	30.607***
NEG	29.996***	8.101***	9.727**	21.694***	26.820***	30.533***	29.586***	8.598***	12.048***	20.962***	28.115***	28.309***
CON	12.626***	1.366	2.188	5.756*	15.262***	18.787***	14.394***	2.140	2.918	5.447*	16.541***	23.794***
DEM	0.047	−0.024	−0.621	−0.414	0.060	0.703	0.085	0.002	−0.456	−0.393	−0.159	1.165
TH	2.719	1.222	1.607	2.533	−4.272	−4.505	−1.034	−1.982	5.850	3.477	−8.703	−11.723
CPI	−0.021***	−0.017*	−0.008	−0.021	−0.017	−0.037*	−0.021**	−0.013	0.000	−0.018	−0.027	−0.025
EXG	−0.271***	−0.057	−0.029	−0.236*	−0.513***	−0.818***	−0.247***	−0.056	−0.079	−0.189	−0.514***	−0.610***
FDI	−0.101***	0.009	−0.060**	−0.084**	−0.070*	−0.158***	−0.103***	0.013	−0.036	−0.083**	−0.069	−0.165***
IND	−3.497***	−1.297***	−1.581***	−2.130***	−3.260***	−4.187***	−3.476***	−1.297***	−1.552***	−2.134***	−3.142***	−3.682***
INF	0.071*	−0.031	−0.058	0.099	0.027	−0.018	0.089*	−0.012	−0.006	0.082	0.021	−0.061
IMP	−0.082	0.280***	0.183	−0.012	−0.092	0.082	−0.047	0.276***	0.225	0.070	−0.077	−0.095
OIL	−0.135	−0.056	−0.113*	−0.087	−0.096	−0.032	−0.073	−0.062	−0.029	−0.010	0.063	−0.402*
UMP	0.318***	−0.001	0.066	0.151	0.716***	0.781***	0.302***	0.015	0.069	0.128	0.626***	0.869***
Pseudo R²	0.522	0.236	0.224	0.268	0.385	0.519	0.508	0.227	0.217	0.266	0.373	0.521

Note: This table reports the estimation results of host country happiness indices and Twitter happiness index impacts on international tourism revenues of t and $t + 1$ periods according to equation (2). Yearly data are used for the period 2006–2017. The same note as Table 3.

Subsample of the non-global financial crisis period

As tourism involves discretionary income, it is vulnerable to economic ambiguity and instability. Because international tourism figures arrived negative territory in 2008, the situation was expected to get worse during 2009 (Papatheodorou et al., 2010). Thus, we remove the effect of the 2008 global financial crisis and set 2006–2007 and 2010–2017 as non-global financial crisis subperiods to run QR models in Table 8. We observe the same notably insufficient life ladder condition as in the front sections; that is, a lower life ladder denotes more tourist arrivals. Excess social support more likely negatively affects the highest tourist arrivals and tourism revenue quantiles. Most country happiness impacts are positive and significant on tourist arrivals and tourism revenues with miniscule impacts on sector returns. Twitter happiness shows significantly positive impacts on sector returns with no salient impact on tourist arrivals and tourism revenues, signifying the importance of social media happiness on sector returns. Tourist arrivals have a notably positive relation with tourism revenues, but no salient relation with sector returns.

Table 8.

Estimates of the QR-based tourism development for non-global financial crisis period models.

	ARR				REV				SR
Variables	Q10	Q25	Q75	Q90	Q10	Q25	Q75	Q90	Q10	Q25	Q75	Q90
C	9.280	2.204	10.957	9.203	−33.607	−38.389	−81.932	−124.299	−14.639**	−13.413**	−18.593**	−31.793*
REV									−0.005	0.000	−0.010*	−0.013
ARR					1.284***	1.450***	2.182***	2.893***	−0.004	0.012	0.065	0.071
LL	−0.262	−0.379***	−0.307***	−0.151*	−0.195	−0.845**	−0.120	−0.233	−0.031	0.044	−0.023	0.088
GDP	−0.189	0.303	0.475**	0.755***	−0.677	−0.174	0.502	3.554*	0.288	0.137	0.069	−0.374
SS	−2.449	−1.105	−1.662	−2.925**	3.866	0.366	−13.855*	−28.519***	−0.632	−1.046	−1.091	−1.236
HLE	0.096***	0.049	0.052	0.004	0.329***	0.357***	0.975***	1.216***	0.006	−0.002	−0.002	0.018
FLC	2.220***	2.086***	0.934	−0.140	3.463	7.845**	6.440	2.586	1.236**	0.800	1.305**	1.675
GEN	2.683***	1.044	0.461	0.517	−1.697	0.382	3.588	2.762	0.174	−0.062	−0.286	−0.204
PER	−0.163	−0.007	0.658	0.745**	5.149***	6.816***	17.583***	30.381***	0.081	0.497*	0.305	0.525
NEG	3.829***	2.562*	−2.538***	−2.944**	13.586***	14.999***	44.363***	35.286***	1.258	−0.876	−0.720	−0.338
CON	−0.520	−0.675	0.378	−0.166	3.675**	2.050	14.098***	18.590***	−0.094	0.313	−0.161	0.250
DEM	0.220	0.061	−0.160	−0.323**	−0.332	−0.725	−1.030	−0.438	−0.053	0.042	−0.093	−0.127
TH	−1.825	−0.855	−1.652	−0.718	1.777	4.131	8.926	16.460	1.603	1.868**	2.701**	5.021**
EXG	−0.135***	−0.063**	−0.096***	−0.063**	−0.035	−0.073	−0.587**	−1.021***	−0.009	−0.028*	−0.033	−0.015
IND	0.612***	0.606***	0.459***	0.392***	−0.963***	−1.402***	−3.228***	−4.867***	0.021	−0.003	−0.032	−0.017
OIL	−0.056	−0.056	−0.044	−0.049	0.039	−0.069	−0.162	−0.347	0.006	0.007	0.050*	0.068
Pseudo R²	0.414	0.401	0.432	0.430	0.201	0.206	0.339	0.480	0.143	0.091	0.109	0.147

Note: This table reports the estimation results of host country happiness indices and Twitter happiness index impacts on tourist arrivals, tourism revenues, and travel and leisure sector returns of the non-global financial crisis period according to equations (1 ) to (3). Yearly data are used for the period 2006–2017. The same note as Table 3.

Subsample of developing countries

Eugenio-Martin et al. (2004) find that the nexus in the amount of tourists and economic growth exists in emerging nations, but not in developed nations. We investigate whether happiness variables have different impacts on dependent variables, use developing countries as a subsample, and report the empirical results in Table 9. Tourist arrivals show a notably positive relation with tourism revenues under lower quantiles, while there is no salient relation with sector returns. Tourist arrivals and tourism revenues report no significant impact on sector returns. Consistent with Table 8, Twitter happiness shows significantly positive impacts on sector returns, while there is no salient impact on tourist arrivals and tourism revenues, denoting the importance of social media happiness on sector returns. We also observe the same notable insufficient life ladder condition as in the front sections; that is, a lower life ladder means more tourist arrivals. The excess negative social support more likely affects the highest tourist arrival quantiles. Moreover, most country happiness impacts are asymmetric and significant on tourist arrivals and tourism revenues with a miniscule impact on sector returns. The results of the developing country subgroup are similar with those of the whole sample.

Table 9.

Estimates of the QR-based tourism development for the developing countries’ models.

	ARR				REV				SR
Variables	Q10	Q25	Q75	Q90	Q10	Q25	Q75	Q90	Q10	Q25	Q75	Q90
C	3.580	2.556	15.211*	13.013	4.322	18.798	−61.404	−178.273	−25.785**	−30.170**	−25.870	−57.608**
REV									−0.003	0.004	−0.014	−0.019
ARR					0.516***	0.349*	0.003	−3.178**	−0.022	0.005	−0.053	−0.001
LL	0.002	−0.078	−0.223**	−0.070	−0.695	−1.717***	−5.874***	−11.075***	−0.112	−0.007	0.105	0.243
GDP	0.265	0.878***	1.234***	1.316***	−1.074***	−1.565***	−1.471	5.317**	0.113	0.076	−0.213	−0.405
SS	−2.007**	−0.775	−0.538	−1.553***	8.036***	7.626**	3.700	14.761	0.452	−0.239	−0.705	−0.267
HLE	0.090***	0.035	0.000	−0.001	0.063	0.289***	1.011***	1.106***	0.017	−0.014	−0.006	0.023
FLC	2.138**	1.989***	2.350***	1.413*	−4.746**	−1.203	−9.712**	−27.434**	0.972	1.419**	−0.729	−1.080
GEN	1.056	0.717	0.116	0.150	0.534	5.263**	15.890***	29.852***	−0.290	0.057	0.236	−0.275
PER	1.204	−0.871	1.404***	1.356***	2.070	−1.348	−13.284***	−21.100	0.272	−0.303	0.784	0.681
NEG	−0.115	0.481	−3.479***	−1.633**	6.901**	4.292	21.382***	31.071**	0.906	−0.428	−1.282	−0.658
CON	−0.722	0.009	0.770**	−0.026	2.751*	−2.576	−3.636	−6.360	−0.032	−0.115	0.502	0.648
DEM	−0.164	−0.115	−0.565***	−0.788***	2.763***	2.155***	3.667***	5.849***	0.053	−0.095	−0.161	−0.033
TH	0.090	0.187	−2.098	−1.604	−0.904	−1.724	11.924	37.919	3.287**	4.296**	4.286	8.880**
OIL	0.005	0.001	0.024	0.017	−0.010	−0.150*	0.003	−0.404	0.070***	0.088***	0.089*	0.149**
Pseudo R²	0.178	0.230	0.291	0.332	0.098	0.101	0.178	0.237	0.258	0.183	0.166	0.308

Note: This table reports the estimation results of host country happiness indices and Twitter happiness index impacts on tourist arrivals, tourism revenues, and travel and leisure sector returns of developing countries according to equations (1 ) to (3). Yearly data are used for the period 2006–2017. The same note as Table 3.

Subsample of European countries

Because the relationship between tourism revenue and GDP varies with geographic areas (Çağlayan et al., 2012), we use the largest sample area as a subgroup to test the robustness of happiness indices in specific region samples in Table 10. Inconsistent with Table 8, Twitter happiness shows no significant impacts on sector returns, tourist arrivals, and tourism revenues, signifying that the influence of social media happiness on sector returns is minimal in European countries. Tourist arrivals still significantly and positively influence tourism revenues. No Twitter happiness index impact on sector returns exists in European countries. GDP (perceived absence of corruption) shows a constant positive and noteworthy influence on tourist arrivals. The results of the European country subgroup are dissimilar with those of the whole sample; that is, there is no Twitter happiness index influence, and there is a slight impact of host country happiness on tourism variables.

Table 10.

Estimates of the QR-based tourism development of the European countries’ models.

	ARR				REV				SR
Variables	Q10	Q25	Q75	Q90	Q10	Q25	Q75	Q90	Q10	Q25	Q75	Q90
C	−60.880***	−67.083***	−49.847**	−63.658**	−181.901	231.566	380.826	119.383	38.285	20.570	15.908	15.925
REV									0.047	0.039	0.063	0.064
ARR					11.138**	15.264***	17.523***	17.494***	−0.232	0.338	−1.316	−1.316
LL	−0.291	−0.291	−0.124	−0.138	4.792	6.550*	7.236*	10.380**	0.016	−0.130	−0.028	−0.028
GDP	5.834***	6.567***	5.847***	5.669***	3.662	−33.032	−41.632	−37.230	−0.253	1.014	−2.526	−2.526
SS	−1.682	−2.167	−2.441	−4.326	−45.997	−30.349	−2.518	0.378	1.948	2.224	2.270	2.268
HLE	0.124	0.120	0.114	0.207	5.014**	2.819	2.831	3.644	−0.112	−0.325	0.337	0.337
FLC	0.780	0.381	0.018	1.755	−36.895**	−22.378	−13.142	−27.272	1.972	0.076	1.210	1.211
GEN	0.246	0.292	−0.227	0.161	−11.351	−1.986	3.093	−14.124	1.655	1.050	1.039	1.038
PER	6.249***	6.263***	6.377***	7.284***	25.924	−0.260	−0.011	2.644	−3.029	−2.324	−4.988	−4.993
NEG	0.467	1.606	0.895	−0.567	16.504	45.602	58.549**	71.971**	0.836	0.215	0.659	0.654
CON	1.572*	1.213*	1.417	1.905*	−0.789	2.469	7.644	5.111	0.006	−0.312	−1.312	−1.314
DEM	0.766	0.589	1.421*	1.478*	1.849	5.570	7.126	8.918	−0.853	−0.677	−2.569	−2.570
TH	2.105	1.922	0.359	2.425	−7.875	−13.632	−13.748	12.140	−2.278	−2.253	1.785	1.786
CPI	0.035	0.049*	0.041	0.043	−0.786**	−0.592*	−0.869**	−0.885*	0.035	0.031	−0.059	−0.060
EXG	−1.109***	−1.140***	−1.395***	−1.383***	−5.101	0.944	4.014	0.291	−0.200	0.156	−0.381	−0.383
IND	−0.376	−0.393*	−0.375	−0.521**	−8.721**	−7.767**	−12.462***	−12.666**	−0.649	−0.148	0.232	0.230
Pseudo R²	0.918	0.886	0.746	0.720	0.741	0.762	0.862	0.873	0.673	0.467	0.450	0.669

Note: This table reports the estimation results of host country happiness indices and Twitter happiness index impacts on tourist arrivals, tourism revenues, and travel and leisure sector returns of European countries according to equations (1 ) to (3). Yearly data are used for the period 2006–2017. The same note as Table 3.

Alternative tourism revenue variables

To check the consistency of our variables, we apply the log of international tourism revenues in current US dollars to proxy tourism revenues. Due to space limitation, the results of proxy revenues are available upon request. Tourist arrivals also show a notably positive relation with proxy revenues at all quantiles. Consistent with Table 8, host country happiness shows significantly positive impacts on proxy revenues, while the Twitter happiness index shows no salient impact on tourism revenues, signifying the importance of host country happiness spurring tourism revenues. GDP and generosity show a notably positive impact on proxy revenues, while most country happiness impacts are asymmetric and significant on proxy revenues. The results of different tourism revenue variables are similar with those of tourism revenues. Basically, except for the European country results (i.e. slight impact of host country happiness), we discover no other notable variances between the main results and other robustness tests, demonstrating that country happiness determines tourist arrivals and tourism revenues, while the Twitter happiness index has more influence on sector returns. We summarize the findings in Table 3A in the Online Appendix.

Discussion

Sainaghi and Baggio (2017) suggest favoring a more holistic view of the tourism destination. Therefore, this research uses two happiness indices to explore whether happiness nonlinearly or asymmetrically affect tourism development across the conditional distribution of relevant tourism factors. The two happiness indices are Twitter happiness index, which is a social media happiness index and search engine data, and country happiness, which is a milestone study of the condition of global happiness that ranks 156 countries by how happy their populations identify themselves to be.

This study’s results are imperative to tourism development under highly competitive global tourism markets. These findings have practical implications for establishing management implications and trading strategies based on social media and country happiness trading signals and for forming national policy. First, for policymaking, financial analysts, management, stockholders, and academic literature, our study estimates host country happiness and Twitter happiness indices as determinants of tourism development based on a 12-year data set and international evidence that can be generalized to practical tourism and academic fields. Specifically, travel and leisure sector returns have important suggestions for capital investment and tourism management. Clarifying the determinants’ impacts on tourist arrivals, tourism revenues, and sector returns can assist governments, airline management, tourism managers, and investors to plan and conduct effective policies and/or strategies to supply to the needs of foreign tourists.

We next show how the tourism variables in different quantiles can help us assess the social media and host country happiness effects across diverse tourism conditions, thus providing more detailed support to develop tourism. Our results also confirm that the non-global financial period, subsequent period results, and developing countries offer evidence that matches similar findings in the full sample, supporting that our evidence can be generalized to universal samples. However, we find an insignificant impact of Twitter and host country happiness indices on European tourism factors, suggesting that ignored, important, and geographic-specific determinants merit interesting research questions.

Conclusion and implication

This research aims at complementing studies on the relationships between host country and social media (Twitter) happiness indices as well as tourism by paying special attention to the distribution of tourist arrivals, tourism receipts, and travel and leisure sector returns using yearly data of 119 countries for the period 2006–2017. The focal contribution of this research is to explore whether host country and/or social media happiness indices attract tourism development across the conditional distortion of relevant tourism factors. For this purpose, we employ the QR model. From a practical perspective, we provide a richer feature of the relationship between host country happiness, Twitter happiness indices, economic factors, and tourism. Moreover, we consider that the dependence relationships might change by a non-global financial crisis, geographic region, country economic development level, and different variable proxies.

Our findings reveal that the influence of Twitter happiness index on travel and leisure sector returns is mostly significant and positive across a non-global financial crisis period and different country economic development levels, except for the European country subgroup, implying that it is wise to take Twitter happiness, as one social media happiness proxy, into account when structuring portfolios of tourism sector stocks on travel and leisure sector returns. However, the impacts of the Twitter happiness index on both tourist arrivals and tourism revenues are insignificant. Most relationships of country happiness on both tourist arrivals and tourism revenues are significant and asymmetric in current and subsequent years, whereas the relationship is insignificant on travel and leisure sector returns. Among the country happiness indices, GDP, freedom to make life choice, perception on absence of government corruption, and generosity are positive and salient across several quantiles, supporting that a happy host country attracts international tourist arrivals. Specifically, life ladder is negatively and notably related with tourist arrivals, suggesting that a host country with a low life ladder prefers to increase tourist arrivals. However, the tourist arrivals caused by a negative effect are positive and salient at lower tourist arrival quantiles and turn negative at the highest quantiles, supporting the nonlinear association between happiness and tourism development. In sum, the majority of the country happiness indices are saliently and positively related with tourism revenues, supporting that a happy host country increases tourism revenues. Additionally, tourist arrivals remarkably and positively affect tourism revenues. However, neither tourist arrivals nor tourism revenues notably impact travel and leisure sector returns.

We also examine whether the effects of macroeconomic factors are related with tourist arrivals, tourism revenues, and travel and leisure sector returns. We find evidence that CPI, FDI, industrial production, and imports (exchange rate and inflation) are significantly positively (negatively) related with tourist arrivals. CPI, exchange rate, FDI, industrial production, inflation, and oil price (imports and inflation) are notably, asymmetrically, and negatively (positively) related with tourism revenues. Oil prices are constantly positively related with travel and leisure sector returns, while inflation and imports are asymmetric and salient with travel and leisure sector returns. Our findings are consistent with the current studies for the asymmetric relationships between tourism demand, oil prices, exchange rate, and inflation (Meo et al., 2018).

Understanding the happiness and tourism nexus has been a focal theme in the tourism field. The central suggestion of this study is that residents’ happiness and Twitter happiness should be considered when developing a tourism industry. Additionally, Po and Huang (2008) pinpoint a nonlinear relationship for the tourism–economic growth nexus. Our study is the first to point out that there exists a nonlinear association between tourism and happiness, implying that happiness does not necessarily bring about tourism development under all quantiles, and that a nonlinear happiness impact on tourism development is more adequate and complete for delineating the relationship.

Future research can examine whether circumstances support a more negative effect that prefers more tourist arrivals at lower arrival quantiles or if a more negative effect does not welcome more tourists. Additionally, it would be interesting to study the influences of Twitter happiness and host national happiness indices on tourism factors for diverse geographic areas. Admittedly, we only explore the relationship between happiness indices and tourism. Our study has not explored the issue from the diverse parameter values in the estimated 10 host country happiness factors considered herein. If we could recognize the distinct roles played by individual country happiness indices, then this would be useful to realize how to improve tourism development more accurately. We leave this topic for future work.

Supplemental material

Supplemental Material, appendix - Tourism development and happiness: International evidence

Supplemental Material, appendix for Tourism development and happiness: International evidence by Chien-Chiang Lee, Mei-Ping Chen and Yi-Ting Peng in Tourism Economics

Footnotes

Acknowledgements

The authors would like to thank the editors and two anonymous referees for their helpful suggestion.

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) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was financially supported by National Taichung University of Science and Technology in Taiwan.

ORCID iDs

Chien-Chiang Lee

Mei-Ping Chen

Supplemental material

Supplemental material for this article is available online.

References

Akal

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

Albaladejo

González-Martínez

Martínez-García

(2014) Quality and endogenous tourism: an empirical approach. Tourism Management 41: 141–147.

Badshah

(2013) Quantile regression analysis of the asymmetric return-volatility relation. Journal of Futures Markets 33(3): 235–265.

Baggio

Sainaghi

(2011) Complex and chaotic tourism systems: towards a quantitative approach. International Journal of Contemporary Hospitality Management 23(6): 840–861.

Balli

Tsui

WHK

Balli

(2019) Modelling the volatility of international visitor arrivals to New Zealand. Journal of Air Transport Management 75: 204–214.

Baur

Dimpfl

Jung

(2012) Stock return autocorrelations revisited: a quantile regression approach. Journal of Empirical Finance 19(2): 254–265.

Becken

(2011) A critical review of tourism and oil. Annals of Tourism Research 38(2): 359–379.

Becken

Lennox

(2012) Implications of a long-term increase in oil prices for tourism. Tourism Management 33(1): 133–142.

Bimonte

Faralla

(2012) Tourist types and happiness: a comparative study in Maremma, Italy. Annals of Tourism Research 39(4): 1929–1950.

10.

Bimonte

Faralla

(2016) Does residents’ perceived life satisfaction vary with tourist season? A two-step survey in a Mediterranean destination. Tourism Management 55: 199–208.

11.

Binder

Coad

(2011) From Average Joe’s happiness to Miserable Jane and Cheerful John: using quantile regressions to analyze the full subjective well-being distribution. Journal of Economic Behavior & Organization 79(3): 275–290.

12.

Buchinsky

(1998) Recent advances in quantile regression models: a practical guideline for empirical research. Journal of Human Resources 33: 88–126.

13.

Çağlayan

Sak

Karymshakov

(2012) Relationship between tourism and economic growth: a panel Granger causality approach. Asian Economic and Financial Review 2(5): 591.

14.

Carneiro

Eusébio

(2019) Factors influencing the impact of tourism on happiness. Anatolia 30(4): 475–496.

15.

Chaabouni

(2019) China’s regional tourism efficiency: a two-stage double bootstrap data envelopment analysis. Journal of Destination Marketing & Management 11: 183–191.

16.

Chang

Hsu

McAleer

(2013) Is small beautiful? Size effects of volatility spillovers for firm performance and exchange rates in tourism. The North American Journal of Economics and Finance 26: 519–534.

17.

Chatziantoniou

Filis

Eeckels

, et al. (2013) Oil prices, tourism income and economic growth: a structural VAR approach for European Mediterranean countries. Tourism Management 36: 331–341.

18.

Chen

(2007) Macro and non-macro explanatory factors of Chinese hotel stock returns. International Journal of Hospitality Management 26(4): 991–1004.

19.

Chen

(2010) The economy, tourism growth and corporate performance in the Taiwanese hotel industry. Tourism Management 31(5): 665–675.

20.

Chen

(2011) The response of hotel performance to international tourism development and crisis events. International Journal of Hospitality Management 30(1): 200–212.

21.

Chen

Kim

(2005) The impact of macroeconomic and non-macroeconomic forces on hotel stock returns. International Journal of Hospitality Management 24(2): 243–258.

22.

Chen

(2018) Does a happy destination bring you happiness? Evidence from Swiss inbound tourism. Tourism Management 65: 256–266.

23.

Chiang

Tan

(2010) Empirical investigation of herding behavior in Chinese stock markets: evidence from quantile regression analysis. Global Finance Journal 21(1): 111–124.

24.

Chiu

Yeh

(2017) The threshold effects of the tourism-led growth hypothesis: evidence from a cross-sectional model. Journal of Travel Research 56(5): 625–637.

25.

Demir

Alıcı

Chi Keung Lau

(2017) Macro explanatory factors of Turkish tourism companies’ stock returns. Asia Pacific Journal of Tourism Research 22(4): 370–380.

26.

De Neve

Ward

De Keulenaer

, et al. (2018) The asymmetric experience of positive and negative economic growth: global evidence using subjective well-being data. Review of Economics and Statistics 100(2): 362–375.

27.

Edelmann

Höchtl

Sachs

(2012) Collaboration for open innovation processes in public administrations. In: Charalabidis

Koussouris

(eds) Empowering Open and Collaborative Governance. Berlin: Springer, pp. 21–37.

28.

Ekanayake

Long

(2012) Tourism development and economic growth in developing countries. The International Journal of Business and Finance Research 6(1): 61–63.

29.

Eugenio-Martin

Martín Morales

Scarpa

(2004) Tourism and economic growth in Latin American countries: a panel data approach. FEEM Working Paper No. 26.2004. www.ssrn.com (accessed April 2020).

30.

Filep

(2014) Moving beyond subjective well-being: a tourism critique. Journal of Hospitality & Tourism Research 38(2): 266–274.

31.

Filep

Deery

(2010) Towards a picture of tourists’ happiness. Tourism Analysis 15(4): 399–410.

32.

Forsyth

Dwyer

Spurr

(2014) Is Australian tourism suffering Dutch disease? Annals of Tourism Research 46: 1–15.

33.

Frey

(2008) Happiness: A Revolution in Economics. Cambridge: The MIT Press.

34.

Gokovali

(2010) Contribution of tourism to economic growth in Turkey. Anatolia 21(1): 139–153.

35.

Gunderson

(2005) Analyzing the effects of amenities, quality of life attributes and tourism on regional economic performance using regression quantiles. Journal of Regional Analysis and Policy 35: 1–22.

36.

Gunter

(2018) What makes an Airbnb host a superhost? Empirical evidence from San Francisco and the Bay Area. Tourism Management 66: 26–37.

37.

Habibi

(2015) Foreign direct investment, tourism and economic growth in Malaysia. European Journal of Social Sciences 48(3): 381–390.

38.

Hanafiah

Harun

Jamaluddin

(2011) Trade and tourism demand: a case of Malaysia. In: International conference on business and economic research, Malaysia.

39.

Hariyanto

(2017) Leisure time, leisure activities and society happiness index in Bandung City. International Journal of Scientific & Technology Research 6(10): 352–357.

40.

Hill

Nadig

Hougan

(2015) A Comprehensive Guide to Exchange-Traded Funds (ETFs). Charlottesville: CFA Institute Research Foundation.

41.

Hung

Shang

Wang

(2010) Pricing determinants in the hotel industry: quantile regression analysis. International Journal of Hospitality Management 29(3): 378–384.

42.

Ivlevs

(2017) Happy hosts? International tourist arrivals and residents’ subjective well-being in Europe. Journal of Travel Research 56(5): 599–612.

43.

Kapiki

(2012) The impact of economic crisis on tourism and hospitality: results from a study in Greece. Central European Review of Economics and Finance 2(1): 19–30.

44.

Kim

Uysal

Sirgy

(2013) How does tourism in a community impact the quality of life of community residents? Tourism Management 36: 527–540.

45.

Koenker

Bassett

Jr (1978) Regression quantiles. Econometrica: Journal of the Econometric Society 46: 33–50.

46.

Koenker

D’Orey

(1987) Algorithm AS 229: computing regression quantiles. Journal of the Royal Statistical Society. Series C (Applied Statistics) 36(3): 383–393.

47.

Kristjanpoller

Concha

(2016) Impact of fuel price fluctuations on airline stock returns. Applied Energy 178: 496–504.

48.

Kuo

Cheng

Chang

, et al. (2018) The asymmetric effect of tour guide service quality on tourist satisfaction. Journal of Quality Assurance in Hospitality & Tourism 19(4): 521–542.

49.

Kwon

Lee

(2020) Why travel prolongs happiness: longitudinal analysis using a latent growth model. Tourism Management 76: 103944.

50.

Lee

Chang

(2008) Tourism development and economic growth: a closer look at panels. Tourism Management 29(1): 180–192.

51.

Lee

Chen

(2020) Do natural disasters and geopolitical risks matter for cross-border country exchange-traded fund returns? North American Journal of Economics and Finance 51: 101054.

52.

Lee

Chien

(2008) Structural breaks, tourism development, and economic growth: evidence from Taiwan. Mathematics and Computers in Simulation 77(4): 358–368.

53.

Lee

Brahmasrene

(2013) Investigating the influence of tourism on economic growth and carbon emissions: evidence from panel analysis of the European Union. Tourism Management 38: 69–76.

54.

Lee

Jang

(2011) Foreign exchange exposure of US tourism-related firms. Tourism Management 32(4): 934–948.

55.

Lew

(2012) Using quantile regression to understand visitor spending. Journal of Travel Research 51(3): 278–288.

56.

Masiero

Nicolau

Law

(2015) A demand-driven analysis of tourist accommodation price: a quantile regression of room bookings. International Journal of Hospitality Management 50: 1–8.

57.

Meo

Chowdhury

MAF

Shaikh

, et al. (2018) Asymmetric impact of oil prices, exchange rate, and inflation on tourism demand in Pakistan: new evidence from nonlinear ARDL. Asia Pacific Journal of Tourism Research 23(4): 408–422.

58.

Mishra

Verma

(2017) Tourism and peace in economic development perspective of India. Journal of Environmental Management & Tourism 8(4(20)): 927–934.

59.

Mohanty

Nandha

Habis

, et al. (2014) Oil price risk exposure: the case of the US travel and leisure industry. Energy Economics 41: 117–124.

60.

Mosteller

Tukey

(1977) Data Analysis and Regression: A Second Course in Statistics. Addison-Wesley Series in Behavioral Science: Quantitative Methods. Boston: Addison-Wesley.

61.

Nawijn

(2011) Determinants of daily happiness on vacation. Journal of Travel Research 50(5): 559–566.

62.

Nawijn

Veenhoven

(2013) Happiness through leisure. In: Freire

(ed), Positive Leisure Science. Dordrecht: Springer, pp. 193–209.

63.

Papatheodorou

Rosselló

Xiao

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

64.

Pearce

(1994) Tourism-resident impacts: examples, explanations and emerging solutions. In: Theobald

(ed), Global Tourism: The Next Decade. Abingdon: Routledge, pp. 103–123.

65.

Perles-Ribes

Ramón-Rodríguez

Sevilla-Jiménez

, et al. (2016) Unemployment effects of economic crises on hotel and residential tourism destinations: the case of Spain. Tourism Management 54: 356–368.

66.

Huang

(2008) Tourism development and economic growth—a nonlinear approach. Physica A: Statistical Mechanics and its Applications 387(22): 5535–5542.

67.

Ranco

Aleksovski

Caldarelli

, et al. (2015) The effects of Twitter sentiment on stock price returns. PloS One 10(9): e0138441.

68.

Rivera

Croes

Lee

(2016) Tourism development and happiness: a residents’ perspective. Journal of Destination Marketing & Management 5(1): 5–15.

69.

Sainaghi

Baggio

(2017) Complexity traits and dynamics of tourism destinations. Tourism Management 63: 368–382.

70.

Sainaghi

Baggio

(2019) Destination events, stability, and turning points of development. Journal of Travel Research. DOI: 10.1177/0047287519890927.

71.

Santana-Gallego

Ledesma-Rodríguez

Pérez-Rodríguez

(2010) Exchange rate regimes and tourism. Tourism Economics 16(1): 25–43.

72.

Seresinhe

Preis

MacKerron

, et al. (2019) Happiness is greater in more scenic locations. Scientific Reports 9(1): 4498.

73.

Shahzad

SJH

Shahbaz

Ferrer

, et al. (2017) Tourism-led growth hypothesis in the top ten tourist destinations: new evidence using the quantile-on-quantile approach. Tourism Management 60: 223–232.

74.

Smeral

(2009) The impact of the financial and economic crisis on European tourism. Journal of Travel Research 48(1): 3–13.

75.

Snaith

Haley

(1999) Residents’ opinions of tourism development in the historic city of York, England. Tourism Management 20(5): 595–603.

76.

Snaith

Haley

(1994) Tourism’s impact on host lifestyle realities. In: Seaton

(ed) Tourism: The State of Art. New York: John Wiley, pp. 826–835.

77.

Sprott

(2005) Dynamical models of happiness. Nonlinear Dynamics, Psychology, and Life Sciences 9(1): 23–36.

78.

Sreekumar

Parayil

(2002) Contentions and contradictions of tourism as development option: the case of Kerala, India. Third World Quarterly 23(3): 529–548.

79.

Surugiu

Leitão

Surugiu

(2011) A panel data modelling of international tourism demand: evidences for Romania. Economic Research-Ekonomska Istraživanj 24(1): 134–145.

80.

Tang

(2011) An exploration of dynamic relationship between tourist arrivals, inflation, unemployment and crime rates in Malaysia. International Journal of Social Economics 38(1): 50–69.

81.

Thelwall

Buckley

Paltoglou

(2011) Sentiment in Twitter events. Journal of the American Society for Information Science and Technology 62(2): 406–418.

82.

Wang

Sun

Wen

(2019) Tourism seasonality, online user rating and hotel price: a quantitative approach based on the hedonic price model. International Journal of Hospitality Management 79: 140–147.

83.

Warr

(2007) Searching for happiness at work. The Psychologist 20(12): 726–729.

84.

Wei

Joseph

Liu

, et al. (2016) Exploring characteristics of suspended users and network stability on Twitter. Social Network Analysis and Mining 6(1): 51.

85.

World Travel and Tourism Council (2019) Available at: https://wttc.org/en-gb/

86.

Wright

(2005) Researching Internet-based populations: advantages and disadvantages of online survey research, online questionnaire authoring software packages, and web survey services. Journal of Computer-Mediated Communication 10(3): JCMC1034.

87.

Cheng

(2017) A study of experiential quality, equity, happiness, rural image, experiential satisfaction, and behavioral intentions for the rural tourism industry in China. International Journal of Hospitality & Tourism Administration 18(4): 393–428.

88.

Yeoman

Lennon

Blake

, et al. (2007) Oil depletion: what does this mean for Scottish tourism? Tourism Management 28(5): 1354–1365.

89.

You

Guo

Peng

(2017) Twitter’s daily happiness sentiment and the predictability of stock returns. Finance Research Letters 23: 58–64.

90.

Zapata

Rambaldi

(1997) Monte Carlo evidence on cointegration and causation. Oxford Bulletin of Economics and Statistics 59(2): 285–298.

91.

Zhang

Wang

et al. (2018) Twitter’s daily happiness sentiment and international stock returns: evidence from linear and nonlinear causality tests. Journal of Behavioral and Experimental Finance 18: 50–53.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.25 MB