Impact of crisis events on Chinese outbound tourist flow: A framework for post-events growth.

International tourism is highly susceptible to external political, economic and environmental crisis events. This paper consists of two studies. Study 1 uses time-series monthly data and the TRAMO/SEATS model to detect and estimate the impact of a range of political, economic and environmental crisis events on the tourist flows from China to Japan and South Korea during the period of 2005-2017. Study 2 uses in-depth interviews to investigate the factors intensifying or alleviating the negative impacts of these crisis events on tourism as well as factors contributing to post-events tourism recovery and growth. Results identify the varying levels of impacts caused by event type and other causative factors for negative impacts. The paper discusses the influence of these factors to provide references to relevant stakeholders for strategic planning, policy making and recovery and growth scheme development.

Introduction

China has become the world's largest source market for international tourists since 2012 (UNWTO, 2018), attracting global industry and academic research attention. Studies on Chinese outbound tourism concerning their motivations, expectations, travel restrictions, and behavioral issues increased significantly in the last decade (Jin & Wang, 2016), however few studies have explored the political dimensions that affect the market development and tourism demand at a regional level.

In the econometric models that aim to predict tourism demand, major events are regarded as an important determinant for tourism demand and tourist flows in addition to determinants such as tourists’ income, tourism prices in a destination relative to those in the source country, prices in the competing destinations, and exchange rates (Song & Li, 2008). Both events that generate tourism flows (such as Olympic Games) and crisis events that hamper tourist arrivals (such as wide-spread epidemic disease) have been investigated so that relevant stakeholders could better understand their positive and negative impacts for strategic planning and destination adaptation purposes. Events that affect negatively on destinations are always termed as crisis events, and publications discussing the impact of crisis events on tourism demand and general destination economy significantly increased over the years (e.g., Goh & Law, 2002; Huang & Min, 2002). The tourism industry could be vulnerable to external environmental changes and crisis events, and recovery could be fast or slow pertinent to a sophisticated interaction among source country, destination and event characteristics.

Literature concerning the impact of crisis events on tourism mainly investigate the following types of crisis events, namely natural disasters, pandemic disease, terrorist activities. This body of literature examines factors diminishing tourism caused by these crises in specific cases (e.g., Eugenio-martin, Sinclair, & Yeoman, 2005) and investigates management problems from the perspective of the extent to which crisis events affect the tourism economy as well as recovery strategies and schemes (e.g., Huang & Min, 2002). The existing body of research mostly explores the impact degree and impact time lag for single events (such as SARS, financial crisis), lack of comparison of the impacts by duration (e.g., long vs short term) and different-types of crisis events (e.g., political vs economic crisis).

Prior literature on the impact of natural disasters on tourism provided evidence that destination tourism industry could be significantly affected by the disasters immediately after the crisis event; however, these destinations could always resume or exceed its former production values within a period of just one to two years (Wang, 2009). Such research indicates the resilient nature of tourism activities and would help the tourism industry learn from past problems and develop prevention and improvement measures (Wang, 2009). Since the tourism industry could be struck at different levels by different types and scales of crisis events, an advanced understanding of the impact of the crisis events would aid the tourism industry and relevant authorities for policy-making and strategy planning purposes.

A wide range of political dimensions affects tourism demand and tourist flows, including international politics, macroeconomics, diplomatic relations and national policies (Wang, 2009). Yet, the tourism literature in general has long neglected the political dimensions in tourism development (Hall, 2010). Social and political events may affect the development of tourism. Discussions on the impact of political events are scarce in literature, although decision-makers in tourism should be more adept in understanding the impact of political dimensions of tourism development and be more sophisticated in their approach to crisis management caused by the political dimensions (Hall, 1994). These decision-makers not only include key decision-makers in the destination country (or regions) but also those in the source country (i.e., tourist generating regions). The increasing number of multinational tourism enterprises that manifest in operation of the international tourism industry is also a key stakeholder in this space. Currently, most of the political risk studies explore actions or events within the destination country (Poirier, 1997). Rarely literature examines actions from the source country. In addition, few studies have examined the impact of crisis events on the flow of Chinese international tourists to the destination countries and regions. Given the present scale and impact of Chinese outbound tourism on global tourism barometer, it is a timely and important topic for tourism scholars and professionals to study how political dimensions divert Chinese tourist flows.

This paper examines the impact of crisis events on Chinese outflows to Japan and South Korea in order to establish a better understanding of the impact and post-events recovery and growth in a mixed political, social and economic context. Japan and South Korea (Republic of Korea) are China's main outbound tourism markets, ranking among the top three in Chinese outbound tourism market in 2016 (Tavel China Guide, 2019). As the close neighbors of China, Japan and South Korea not only have close cultural, economic and trade ties with China, but also have historical disputes over national sovereignty and competitive relations driven by conflicting economic interests. In recent years, China-Japan and China-ROK relations demonstrate a multifaceted and volatile pattern, and major events affecting bilateral relations have regularly occurred. Thus, the following research questions are raised and being explored in this paper:

What types of crisis events, and to what extent, affect Chinese outbound flow to Japan and South Korea from 2005 to 2017?

What factors generate a higher degree of adverse impact on tourism demand?

What factors help the tourism industry of both the destination and the source market recover and grow after the crisis events?

The paper consists of two studies. To find answers to research question 1, Study One employs a quantitative approach and uses the monthly data of Chinese travelers to Japan and South Korea to analyze and explore the impact of various crisis events (e.g., political events, natural disasters and other crisis events) on Chinese outbound flow to Japan and South Korea. To suggest answers to research questions 2 & 3, Study Two uses interviews with key informants (i.e., Chinese outbound tour operators) in light of the factors causative to the rise and fall of the outflows. By corroborating the findings from the two studies, this paper discusses the impacts of these crisis events against the many other factors that affect tourism demand to a destination. A framework is developed via this mixed approach to evaluate crisis events’ impacts to aid strategic planning and post-event recovery. Although crisis events may not be predicted, their occurrence and post-event recovery schemes may be associated with some level of identifiable pattern. The effects of such events on tourism demand and strategies for recovery are to some extent predictable and manageable based on appropriate scenario analysis. Thus, it is of great value to study the impact of crisis events and post-events market recovery and growth. Such understandings will assist the source and destination countries adjust their policy and schemes and recover from the negative impact within a reasonable period. This will contribute to a long-term equilibrium between demand and economic-political fundamentals.

Literature review

Crisis events and tourism

Crisis events in the tourism context refer to events that affect the confidence of tourists travelling to a destination and thus disrupt the continual normal operation of the tourism industry. Crisis events that affect tourism include economic recession/financial crisis, oil/energy issues, political issues, health issues, natural/environmental disasters and increasingly terrorism (Hall, 2010). Crisis events differ by their type, scale, duration, the time and space that they occur (Ren, 2000). Their impacts on tourism could be short or long term pertinent to their characteristics. There is a level of subjectivity in the identification of crisis events. Some events may have significantly reduced visitations to a destination but not identified as crisis events, whereas some perceived crisis events may not have caused serious impacts on the tourism development.

A crisis event may be of a limited duration, but it may receive significant attention in the conventional and new (social) media that enhance concerns travelling to a particular destination where the crisis occurs. An event becomes a crisis event for a particular group of stakeholders when a) the group is aware of the event elements and/or b) the event occurs in a specific political-economic system (Hall, 2002). Pertinent to their socio-demographic and economic profile as well as cultural background, international tourists perceive the influence of natural disasters and travel risks on their travel differently (Park & Reisinger, 2010). Destinations must develop strategies in marketing and service to reduce target groups’ concerns.

Tourism is relatively resilient to economic downturn, but is subject to pressures in its economic, political and natural environment. A substantial literature has discussed how the tourism industry could prepare and respond to crisis, with the focus on crisis and risk management (e.g., Anderson, 2006), impact estimation (Dwyer, Forsyth, Spurr, & VanHo, 2006), forecasting (Lean, Smyth, Barros, Matias, & Santos, 2008), recovery strategies (Scott, Laws, & Prideaux, 2007) and security (Hall, Timothy, & Duval, 2004). The impacts on tourism caused by these crisis events are a result of a system where the world's economies, transport systems, and media networks integrated to reverberate the impact (Hall, 2010). The current body of literature on crisis events focuses on economic and financial crisis, natural disasters, anthropogenic environmental problems, pandemics and disease, and terrorism (Hall, 2010). Political crisis is relatively under-researched in tourism literature. To provide policy-making suggestions, tourism industry and academics should research how policy arenas actually operate as tourism and policy problems are highly interrelated.

Estimating the impact of crisis events on tourist demand

Tourist demand models indicate that macroeconomic variables such as income, relative prices, transportation cost, and exchange rates significantly influence tourist demand (e.g., Dritsakis, 2004). Intervention factors related to political and economic events are incorporated to these demand models, such as terrorist attacks, economic crisis, and natural disasters. Literature concerning the impact of crisis events examines factors diminishing tourism in specific cases (e.g., Huang & Min, 2002), the extent to which crisis events affect the tourism economy (e.g., Goh & Law, 2002) and recovery strategies and schemes (e.g., Campiranon, Scott, Laws, Prideaux, & Chon, 2007; Huang & Min, 2002). Some scholars use qualitative analysis methods, for example, Goodrich (2002) analyzed the short-term effects of the “9.11” incident on US tourism; Okumus, Altinay & Arasli (2005) used the panel interviews to analyze the impact of the Turkish financial crisis on tourism in Northern Cyprus. More scholars use econometric models for the analyses of crisis event impact. Time-series data are typically used in these econometric models, as time-series data typically deal with tasks like clustering, similarity search, forecasting, prediction, classification, outlier detection and noise removal (Eilat & Einav, 2004). Autoregressive Integrated Moving Average time-series model (ARIMA) have been widely used in analyzing time-series data for their accuracy and mathematical soundness (Song & Li, 2008). Most these econometric studies are post-event analyses, which use data prior to, during, and after the event to compare a predicted and actual demand to estimate the effects of the event (Song & Li, 2008). In achieving the estimation, ARIMA and variations of the technique, such as Seasonal Autoregressive Integrated Moving Average (SARIMA), are widely used. For example, Huang and Min (2002) assessed the earthquake impact in tourism in Taiwan, and Min (2005) investigated the effect of Severe Acute Respiratory Syndrome (SARS) on tourism demand in Taiwan. Both used ARIMA models. Chang and Liao (2010) used SARIMA model to predict the scale of Taiwan's travel to Hong Kong, Japan and the United State. Lim and McAleer (2005) used the ARIMA model to analyze the seasonal characteristics of Japanese tourists to Australia. Koc and Altinay (2007) used the X-12-ARIMA model and the TRAMO/SEATS model to analyze the seasonal variation of per capita tourism expenditures in Turkey's inbound tourism from the perspective of market segmentations.

The limitation of such impact analysis may lie in the fact that it is impossible to single out the effects of several crises if one takes place soon after another (Song & Li, 2008). Also, the existing body of research mostly explores the impact degree and impact time lag for single events (such as SARS), lack of comparison of the impact of long-term, different-types of tourism crisis events.

A few studies indicated approaches to reduce negative impact of crisis events on tourism. The public sector plays an anchor role in emergency response system (Morakabati, Page, & Fletcher, 2017). Government travel warnings had a significant impact on the tourism market's travel decisions relating to the threat of terrorism (Walters, Wallin, & Hartley, 2019). The role of media can be crucial in the impact of crisis on tourism (Faulkner, 2001). The impacts of disasters on the market are often out of proportion with their actual disruptive effects because of exaggeration by the media, particularly if it is dramatized and distorted through rumors and the media (Cassedy, 1991). Yet, there is a lack of empirical research investigating the impact of tourism crisis events on international tourism and causative factors for recovery in recent years. Also, few studies have examined the impact of crisis events on the outflow of Chinese international tourists and corresponding reasons.

Apart from econometric factors, a complicated system involving a number of factors may influence tourist flows from one country to another country. The system covers factors from different perspectives, e.g., the tourists' perspective, destination perspective, and the integration of the tourism industry development of both the source market and the destinations. From tourists’ perspective, destination risk and fashion are quite important for destination choice (Eilat & Einav, 2004). Fluctuation in bilateral trade over time for a given pair of countries are uncorrelated with fluctuations in international tourism (Eilat & Einav, 2004). Yet, political, economic, security, military, and cultural forces – all of which function to promote or inhibit tourist flows from China to the destination countries and regions. In the context of Chinese outflow to Japan and South Korea, cultural and human exchanges (e.g., media, pop culture, entertainment industry) between the countries has evolved in multidimensional and multichannel forms. Japanese and Korean pop culture, in particular, are popular in China. These exchanges in trade and culture, and power balance in the negotiation of the exchanges, have not been included in the analyses of the impact of crisis events on tourism or post events recovery and growth.

Chinese travelling to Japan and South Korea from January 2005 to December 2017

Chinese travelling to Japan from January 2005 to December 2017

As Fig. 1 indicates, there is a phenomenal and steadfast growth of the total number of Chinese travellers to Japan from 2005 to 2017, although the curve in both growth rate and total visitor numbers reveal market volatility within the period. From 2005 to 2009, the average annual growth rate was relatively high (e.g., 15.8%). There was a slight decrease in 2008 and 2009. Year 2010, however, saw the highest growth rate (i.e., 52.9%). The first half of 2011 saw a sharp reduce of Chinese tourists; however, the latter half of 2011 once again experienced skyrocketed growth, which continued to midyear 2012. The period from September 2012 to mid-year 2013 witnessed a sharp drop in Chinese to Japan in both growth rate and the total visitor numbers. From September 2013, Chinese travelling to Japan experienced another explosive growth. After 2016, the growth rate slowed down, but the total visitation number reached to a new height, and Chinese tourists became the biggest spenders of all international tourists to Japan (Liu, 2017)

Fig. 1

Plot of monthly visitations by Chinese tourists to Japan and year-on-year growth rates.

The last quarter of a year is generally a low season for travel to Japan (Przystup, 2013), and thus the downward curves in that quarter could be the results of seasonality. Apart from the effect of seasonality, four events could allegedly have affected Chinese travelling to Japan since 2005: 1) the global financial crisis in 2009; 2) Fukushima earthquake and nuclear accident, 3) disputes over Diaoyu/Senkaku Island in 2012/2013; and 4) Chinese boycotts Japan hotel chain APA over Nanjing Massacre denials in 2017. Year 2008 and 2009 experienced the global financial crisis. In March 2011, a magnitude 9 earthquake shook Fukushima, unleashing a savage tsunami and causing a nuclear accident (i.e., release of radioactive materials particularly in contaminated water). In September 2012, the Noda government nationalized the Diaoyu/Senkaku Island, which prompted tensions between the two countries, resulted in anti-Chinese sentiment in Japan and anti-Japan demonstrations in China (Katsuyuki, 2013). In January 2017, two Chinese tourists found books in a chain hotel room (i.e., APA chain hotel) denying Nanjing Massacre, which caused an uproar on Chinese social media, with many calling for a boycott of the hotel chain (Griffith & Ogura, 2017).

Chinese travelling to South Korea from January2005 to December 2017

Chinese outbound travel to South Korea had been growing at a constant, high rate from 2005 to 2015, lending China to become the largest source market for South Korea. There were volatility in the growth in both the growth rate and the total number of arrivals, especially for Year 2015 and 2017. In August 2015, the year-on-year decrease rate was 76%. The numbers bounced back in late 2015 and 2016. The growth rate of Chinese travels to South Korea slowed down in 2016, the total number of visitors was in the highest in June–July of 2016. However, Year 2017 saw a cliff-drop from the beginning to the end of the year. The number of tourists at the end of 2017 went back to the same level as that in 2013 (See Fig. 2 ).

Fig.2

The number of Chinese outbound tourists to Korea and Year-on-year growth rates.

Possible events that could have generated either positive or negative impact on Chinese outflows to South Korea include the Middle East Respiratory Syndrome (MERS) in 2015, the 2016 ‘Visit Korea Year’, and disputes over The Terminal High Altitude Area Defense (THAAD) in 2017. The MERS in Korea from May to July in 2015 involved 186 cases and claimed 38 deaths, with over 16,000 individuals quarantined for 14 days for the control of the outbreak (Oh et al., 2018). The 2016 ‘Visit Korea Year’ initiated policies and resources to boost Chinese travels to South Korea, including visa fee waivers for group travelers, new air routes, specialized tour packages, social media promotion, incentive trips, and local partnerships (Liu & Lian, 2016). THAAD (2017) ignited Chinese concerns over the implications for its strategic political and defense interests. China-ROK frictions over THAAD resulted in Chinese call on social media to boycott travels to South Korea.

Study One

Research method

Study One uses time-series monthly data to try to detect and estimate the impact of a range of crisis events on the tourist flows from China to Japan and South Korea respectively in the period of 2005–2017. The monthly data on the number of Chinese tourists to Japan were obtained from the official website of the Japan Government Tourism Bureau. The monthly data on the number of Chinese travelers to ROK were obtained from the official website of the Korea Tourism Agency. Due to the unavailability of data prior to Year 2005, the research period of this paper is set from 2005 to 2017.

Visitor arrival data are used to detect (discriminate) crisis events that have statistically significant impact on tourist demand and estimate the duration and severity of these events. This will assist accurate clustering and efficient indexing of ARIMA time series. The specific research method is the Time Series Regression with ARIMA Noise, Missing values and outliers (TRAMO) and Signal Extraction in ARIMA Time Series (SEATS). The methods were developed by Caporello and Maravall in 1996 and have been widely used by institutions such as US Census Bureau and Eurostat (Koc & Altinay, 2007). The general estimation follows three steps. The first analysis stage is to follow the conventional approach to pre-process and decompose the time series (Koc & Altinay, 2007). Specifically, TRAMO preprocesses the time series data, then passes the result to SEATS, and uses SEATS to decompose the time series into the following two models: Where: Y represents a time series; TC represents a long-term trend-cycle element of a time series, which is a long-term feature of a period of several years. S represents the seasonal element, which is a cyclical change that occurs repeatedly every year. The period is 12 months or 4 quarters, and I represents irregular elements.

The second stage is the identification of outliers. Among them, the LS (Level Shift) outlier indicates that a crisis event has the greatest influence on the time series with the longest impact duration, which often affects the trend - cyclic component, and has a fundamental and lasting impact on the business activities of the tourist destination. However, it does not affect the irregular factors and seasonal factors in the sequence decomposition. The TC (Temporary Change) outliers represent the impact of a crisis event inferior to the LS outliers. The TC outliers do not affect the trend-cycle components, often affecting only the irregular components. In addition, the AO (Additive Outlier) outliers caused by a crisis event often only affect the irregular factors, but the impact period is very short. It often appears as a single jump point in the time series, because its influence level will not disturb the normal operation of the tourist destination. Study One aims to determine the crisis events affecting Chinese outbound travels based on the types of outliers extracted from the seasonal adjustment model. In the estimation, Level Shift (LS) and temporary outlier (TC) are analyzed but AO outliers were excluded.

In the third analysis stage, point estimation and interval estimation are performed. The estimated value is added or multiplied to the corresponding unaffected component to obtain an estimated range of the natural trend curve. The loss estimate is performed accordingly. When an outlier is TC, the impact of a crisis is attributable to the irregular factors and has no effect on the trend and the seasonality. Therefore, an irregularity factor is checked to determine if it has a normal characteristic, and the interval is estimated for each month of the year in which the crisis has not occurred. This interval is the normal variation range of the irregular factor without the crisis event. Add or multiply this interval to the trend and seasonal data to get the interval estimate of the background line, If an outlier is LS, the impact comes down to the trend factor, and has no effect on the seasonal and irregular factors. Therefore, there is a need to re-estimate trends that are not affected by crisis events. The pre-crisis data are usually used to predict the trend at the time of crisis, and then the background is determined according to the seasonal adjustment model and irregular seasonal factors. During the computation, data with an average loss rate of less than 5% are excluded from the impact period. The “nbdemetra” software developed by the European Union Statistics Center was used for the computation.

Results

Estimation of the impact of crisis events on Chinese travelling to Japan

The pre-adjustment results show that the time series data is suitable for the multiplication model, that is, the magnitude of the time series fluctuation is significantly correlated with the level of the time series, and the logarithmic transformation is required for the computation. The TC, Seasonal and Irregular patterns of Chinese travelling to Japan from 2005 to 2017 are shown in Fig. 3 .

Fig. 3

The sequence decomposition results of Chinese tourists to Japan from 2005 to 2017.

The model identifies five outliers as shown in Table 1 . Because the AO values are outside the scope of this paper, this article will identify events and effects for the other three abnormalities. Three crisis events could be determined, namely the 2009 financial crisis, the 2011 Fukushima earthquake and nuclear accident, and the 2012 Diaoyu Islands incident. Some other small-scale events are not detected in the estimation. For example, Kumamoto earthquake in April 2016 or APA hotel boycott in 2017 did not lead to a decrease in total tourists from China to Japan. The calculated degree of impact and impact period of each of the events on Chinese travelling to Japan are reported in Table 2 .

Table 1

Outlier test of Chinese travelling pattern to Japan (2005–2017).

Outliers	Coefficients	T-Stat	P[|T| > t]
AO (2–2009)	−0.4485	−4.47	0.0000
TC (6–2009)	−0.6244	−5.40	0.0000
LS (3–2011)	−0.5550	−4.67	0.0000
AO (2–2012)	−0.4570	−4.56	0.0000
LS (10–2012)	−0.5830	−4.94	0.0000

Table 2

The impact of crisis events on Chinese travelling to Japan (2005–2017).

Event	The affected period	Actual value	Background value	Loss rate
Financial crisisTC(6–2009)	5–2009	60530	80719	25.0%
	6–2009	36597	72060	49.2%
	7–2009	67944	113896	40.3%
	8–2009	109017	121096	10.0%
	9–2009	98697	118664	16.8%
	10–2009	108301	111237	2.6%
	11–2009	81462	90963	10.4%
	12–2009	62527	72104	13.3%
Fukushima earthquake and nuclear accidentLS(3–2011)	3–2011	62450	107167	41.7%
	4–2011	76164	117384	35.1%
	5–2011	58608	95624	38.7%
	6–2011	61419	82043	25.1%
	7–2011	86963	127793	31.9%
	8–2011	102640	136161	24.6%
	9–2011	112493	121602	7.5%
	10–2011	106174	123130	13.8%
Diaoyu Island "nationalization"LS(10–2012)	10–2012	69713	154904	55.0%
	11–2012	51993	121323	57.1%
	12–2012	52336	97506	46.3%
	1–2013	72301	153241	52.8%
	2–2013	80903	140038	42.2%
	3–2013	102265	167421	38.9%
	4–2013	100160	181003	44.7%
	5–2013	81571	136851	40.4%
	6–2013	98996	127598	22.4%
	7–2013	139905	195976	28.6%
	8–2013	162288	205517	21.0%
	9–2013	156201	183219	14.7%

The results show that the impact cycle of the 2009 financial crisis on the number of Chinese tourists was about 7 months, of which the impact was the strongest in June 2009 with a loss rate of 49.2%. The impact period of the Fukushima earthquake and nuclear accident was about 8 months, of which the impact was the strongest in March 2011with a loss rate of 41.7% and a loss rate of over 20% in the next five months. The impact period of the Diaoyu Islands incident in 2012 was as long as 12 months, and the loss rate over 40% in a consecutive eight-month period right after the incident.

Estimation of the impact of crisis events on Chinese travel to South Korea

Pre-inspection shows that the change in the number of Chinese tourists entering Korea also applies to the multiplication model. The passenger traffic characteristics of China to South Korea in 2005–2017 are reported in Fig. 4 .

Fig. 4

The sequence decomposition results of Chinese tourists to Korea from 2005 to 2017.

The model identifies four outliers, of which AO (7–2015) is not within the scope of this paper (see Table 3). The effect of the 2017 THAAD event is to superimpose the effects of trend-circulation and irregularity. Into the calculations, the impact of the MERS epidemic in 2015 and the 2017 THAAD incident on Chinese travelling to Korea are reported in Table 4 . The results show that the impact cycle of the MERS epidemic in 2015 was only about four months, of which the loss rate was the highest in July 2015, reaching 68.7%. The impact period of the THAAD incident was from March to December 2017. The loss rate from April to November 2017 was above 50% for consecutively nine months. Since the data end at Dec 2017, the visual impact may have continued to 2018 or beyond.

Table 3

Outlier test of Chinese travel pattern to Korea (2005–2017).

Outliers	Coefficients	T-Stat	P[|T| > t]
TC (6–2015)	−0.7346	−7.76	0.0000
AO (7–2015)	−0.5488	−7.16	0.0000
TC (3–2017)	−0.6062	−5.61	0.0000
LS (4–2017)	−0.8520	−8.63	0.0000

Table 4

The impact of crisis events on Chinese travelling to Korea (2005–2017).

Event	The affected period	Actual value	Background value	Loss rate
MERS epidemicTC(6–2015)	6–2015	315095	658802	52.2%
	7–2015	255632	815710	68.7%
	8–2015	513275	901104	43.0%
	9–2015	591242	738964	20.0%
THAAD incidentTC(3–2017)LS(4–2017)	3–2017	360,782	662793	45.6%
	4–2017	227,811	794070	71.3%
	5–2017	253,359	789452	67.9%
	6–2017	254,930	883059	71.1%
	7–2017	281,263	1085508	74.1%
	8–2017	339,388	1092556	68.9%
	9–2017	318,682	933795	65.9%
	10–2017	345,384	886487	61.0%
	11–2017	299,247	679521	56.0%
	12–2017	332,474	656491	49.4%

Study Two

In order to investigate what factors generate a higher degree of crisis events' adverse impact on tourism demand and what factors help the tourism industry recover and grow after the crisis events, a qualitative method is deemed appropriate. Study 2 utilizes a qualitative method (individual and small group interviews) to report insiders’ (in this case Chinese tour operator and tourism experts) views on factors intensifying or alleviating negative impacts and factors for recovery and post-events growth. Individual and small group interviews are appropriate in such a context as this gathers relevant information and provides insights and suggests solutions. Tour operators and tour leaders are interviewees, as they play a key role as tourism product developers, destination marketers, and on-site managers in the context of Chinese outbound tourism. They understand and direct the pulse of the market.

Interviews were semi-structured with an interview guide to ensure that important topical areas were covered in the discussions. The main topical areas include: 1) Chinese tourists flows to Japan and South Korea and their own business operations prior, during and post various crisis events during the time period, 2) the destination market and product characteristics (i.e., Japan and South Korea), 3) tourism enterprise operating model in the tourism generating region (i.e., China), 4) other factors influencing the impact of the crisis events on the tourism businesses in generating and destination regions.

Interviewees were recruited using a personal referral system. Altogether nine operators from six leading tour companies based in Guangzhou participated in the interviews. These six tour companies are among the top 15 tour enterprises nationwide in terms of business turnover, involving both wholesale and retail of tour products via online and offline mechanism. The informants were either directors of Japan and South Korea markets or the general managers of the tour companies. All interviews were conducted in the meeting rooms of the tour companies. Group interviews lasted about one and a half hour each and individual interviews lasted about 1 h each. The researcher conducted all interviews. Interview materials were transcribed and analyzed using content analysis method. Analyses were conducted using the original Chinese dialogue and quotations were translated into English for reporting purposes.

Informants generally agree that natural disasters and public health crisis have different level of impacts on tourism as compared to that of political events. In addition, factors contributing to the degree of impact in the types of crisis events are different. Thus, this section is divided into three sub-sections, namely, the context of natural disasters and public health crisis, the context of political crisis events, and other factors causative of the degree of impact and conducive for post-event recovery and growth.

Information and media influence in natural disaster and public health crisis

Informants confirm that public health and natural disasters at the destinations negatively impact on inbound tourism, but the scale and characteristics of such events determine the severity and consequences of the aftermath. Generally speaking, negative impact of natural disasters and health hazards on tourism is short-lived - usually no more than three months. General earthquake impact is often limited to the earthquake-struck area and may not often lead to a significant reduction in the overall number of tourists to a country, provided other tourist resources/destinations within the country are not affected.

One month after the volcanic eruption in Indonesia [in 2018] group tours went back to the normal scale and operation (ID1, female, 26–35).

Informants posit that various factors contribute to the short-lived impact of natural disasters or public health crisis, including the desire for international travel in Chinese, confirmation of safety from media, and general human characteristics.

We sometimes joked that the Chinese are very forgetful. I estimate that the negative impacts of [epidemic disease] lasts no more than three months. For example, three months after MERS in South Korea, as media ceased to report it, travels resume as normal (ID2, female 36–45).

Media (information adequacy, transparency) definitely play an important role in the context. Fukushima earthquake and nuclear incident had a much stronger detrimental impact than other natural disasters and health crisis on tourist flows due to public panic over the severity of the nuclear leakage, as information was not adequate, or deemed as being provided with a high level of transparency.

There were various rumors about the severity of the nuclear leakage which led to panic and sharp drop of travel. I felt the impact of Fukushima earthquake on travel to China lasted two years (ID7, male, 36–45).

Informants indicated that information provided by home country's tourism authority is significant in recovery from a crisis blow. For example, less than three months after the Fukushima crisis, the Chinese National Tourism Administration issued a last travel notice -‘Japan has entered the stage of post-disaster reconstruction. Except for Fukushima and other hard-hit areas, social order and public life in most other areas have returned to normal’, which could be regarded as a release of the travel warnings.

To sum up, information on the impact of a non-political crisis event, provided by relevant authorities, deemed as transparent, scrupulous, and symmetric, is important in tourists’ travel decision-making right after a crisis event. From the source market perspective, information (travel warnings given by the authoritative tourism bodies) are important for tourists to make their travel or non-travel decisions.

Government policy and media influence in political crisis events

Informants posited that the influence of the government is pertinent and probably dominant in the case of political crisis events on tourism. Government policies during the crisis period directly impact on the marketing and sales of the tour products. The level of government involvement may defer, resulting in varied levels of impacts on outbound tourism. For example, during THAAD, China's tourism administration authority banned travel agencies from organizing group tours to South Korea. Chinese OTA (online travel agency) websites "consciously" removed all relevant products for travelling to South Korea; these products included travel information, travel routes, accommodation, and entry tickets to various attractions. Tour operators ceased operation in that market. Informant indicated:

The Chinese government was involved in the THAAD incident. We received a notification from the government that we shall not organize group tours to South Korea. If the government has spoken it out, of course we dare not act against it (ID5, female, 26–35).

Even now (March 2019) relevant tour products to South Korea are not listed on the websites of OTAs (online tour operators) such as Ctrip. Offline tour operators are allowed to organize group tours to South Korea but these operators would not invest much in marketing and promotion (ID8, male, 36–45).

During Diaoyu Island dispute, Chinese tourism authority did not prohibit travel agencies from operating tours to Japan but adopted a tacit discouragement approach, resulting in tour operators' cancellations of organized tours to Japan. A shrinking of tourist routes between China and Japan due to the political conflicts resulted in the reduction in the number of tourists. Although Japan took remedial measures, China had imposed a “blockade” on certain preferential policies. The market was replenished for a long time, making the impact cycle of this incident longer than that of the Japanese nuclear power plant. However, informants posited that the varying level of Chinese authority's involvement during the two major political crisis events (THAAD vs. Diaoyu Island dispute) may explain why THAAD had a longer and more severe negative impact on Chinese travels to South Korea than Diaoyu Island dispute on Chinese travels to Japan.

But, during the Diaoyu Island incident, we did not receive any explicit notice from the government. During the Diaoyu Island dispute, tour operators in Beijing has received some oral notifications that prohibit organizing group tours to Japan. But operators in Guangzhou did not receive any such notifications from the government (ID9, male, over 45).

Informants also confirmed the influence of conventional and social media on Chinese outflow. One example is Chinese social media outcry for boycott on LOTTE in the THAAD incidents. That LOTTE duty free shop leased land for the deployment of THAAD on February 27, 2017 had become a significant marker of Chinese boycott of tourism to South Korea, resulting in a slump in the total visiting numbers. LOTTE and a few tourism related businesses became clear targets of the boycott on the Chinese social media. One informant indicated:

Shopping in South Korea has been a significant element of Chinese tourists’ activities in the country. Chinese consumers used to constitute 70 percent of the customers of Lotte Group. The Lotte Group became a major target of the boycott on tours to South Korea and purchase of Korean products, which had a huge adverse effect on the business of the Group (ID8, male, 36–45).

Interestingly, Chinese social media's call for boycotting Japanese hotel chain APA over Nanjing Massacre denials in 2017 did not make much influence on the visiting numbers to Japan. This further confirms the influence of governing body over social media in maneuvering tourist flows.

All informants agreed that the political crisis over Diaoyu Islands created a less severe impact on Chinese overall tourism to Japan than THAAD to South Korea. Apart from government influence, another partial explanation could be that the dispute has been a long-standing historical issue rather than an acute incidence. During the period of 2005–2012, diplomatic and political tensions frequently occurred between China and Japan over the Diaoyu Island issue, so that the public's sensitivity to the crisis is reducing over time. Correspondingly, the blow on tourism progressed with a slow pace.

At the beginning of the crisis, strong nationalist sentiment in China had not yet erupted, and we did not immediately cancel the tourism products already sold to the consumers. Although potential tourists began to decrease after the ‘nationalization’ crisis, the change was relatively slow and insignificant (ID2, female, 36–45).

With Chinese domestic nationalist sentiments becoming increasingly conspicuous on social media and anti-Japanese demonstrations taking place across China, travel to Japan became a risk and travel agencies had to cancel their travel products to Japan. Following the sentiment, many tourists cancelled their travel because they were worried about safety. Some tourists cancelled their travels due to social pressure (Lin, Qiu Zhang, Gu, & Peng, 2017). However, group tours to Japan did not come to a complete halt. One informant said:

Even during the peak of the crisis, we had a group tour [to Japan]. Although I told these tourists that our company would like to cancel the tour and they would receive a full refund, a few tourists insisted to go as scheduled and they went (ID7, male, 36–45).

In May 2018, Chinese Premier Li Keqiang publicly remarked that China-Japan relations had returned to a path of normality (Nakazawa & Nikkei, 2018). This official remark has immediate and direct impact on media sentiment and tour operations to Japan.

After Premier Li's official visit to Japan, the current media really has not much negative coverage on Japan which is very good for our business (ID9, male, over 45).

Other factors associated with detrimental impact or recovery

Informants commented on various factors that accelerate or alleviate the negative impact caused by crisis events as well as factors that contribute to a fast recovery aftermath. It seems that, specific to the Chinese outbound market, government-led promotional tourism campaigns at the national level and favorable visa policy have been effective to increase tourist arrivals, conducive to recovery. Right after the Fukushima 2011 crisis, Japan launched a large number of preferential policies in response to the tourism crisis to stimulate the market demand, including new visa policies for Chinese tourists (Japanese Embassy in China, 2012). After MERS 2015 crisis, South Korea collaborated with Chinese tourism industry and launched ‘2016 Visit Korea Year’, which initiated policies, travel resources and facilities to boost Chinese travels to South Korea.

Right after MERS, July 2015, South Korea exempted the Chinese group visa fee. The visa policy stimulated the demand for travel to South Korea to some extent. China was still Korea's largest source of inbound tourists in 2015. Although the exchange rate at the time declined, the number of Chinese visitors to Korea grew rapidly since the end of the epidemic in 2015 (ID3, female, over 45).

In the initiatives of the ‘Visit Korea Year’, more relaxed visa policies, visa fee waivers for group travelers, new air routes, specialized tour packages, social media promotion, incentive trips, and local partnerships (Lean et al., 2008) are prominent. This view has already been discussed in literature. However, informants commented that South Korea government and tourism enterprises did not initiate similar or new incentive schemes after THAAD, which intensified the negative impact on Chinese outflows to South Korea.

Travel and transportation mode could play a role in during-event impact and post-event recovery and growth. China's transportation to South Korea was restricted during the crisis and the situation was not improved post the peak of the crisis. One informant commented on tours to South Korea:

Part of the increase of Chinese tourists to South Korea [prior to THAAD] was brought by cruises, which was about 2 million in the Year of 2015. Due to THAAD, these cruises cancelled South Korea as the port of call and instead travel to Japan, Vietnam and Philippines. It is easy for the cruise companies to change destinations, as tourists choose cruise liners rather than destinations for cruise tours. They see on-board activities as the main attraction of the cruise tour (ID9, male, over 45).

This view was also supported by industry literature. For example, it was reported that cruise companies operating cruises from China to South Korea cancelled more than 300 cruises to South Korean ports, involving more than 1 million tourists (World Maritime, 2017). Since most incentive tours are business travel per se, informants seem to hold the opinion that individual pleasure/leisure travels are less affected by crisis events compared to business travelers.

The ease access to alternative destinations and tourism products may play a key role in the recovery of tourist flows after a crisis event. One informant indicated:

South Korea has been a destination for incentive travels, and lots of Chinese companies organized incentives for their employees to South Korea via cruise trips. But this incentive market has quickly been replaced by similar products (such as incentives to Thailand or cruises to other Southeast Asian destinations) after THAAD (ID2, female, 36–45).

Destination attractiveness significantly direct tourist flows in recovery. Destinations with abundant resources and heterogeneous tour products recover more quickly than destinations with more homogenous products. There was general agreement among the informants that Japan is more appealing than South Korea to the Chinese market, with more diversified products and destinations.

Japan has tourism resources all over the country, not only in Honshu, but also in Kyushu, Shikoku and Hokkaido. But South Korea seems to only have three main routes – Seoul, Busan and Jeju Island, a bit monotonous in product development (ID2, female, 36–45).

Japan has more variety of tour products than Korea. For example, Disneyland, family tours, spa tours, food tours. The products in South Korea are more or less the same (ID9, male, over 45).

Facilities and services in Japan are of world class. A lot of Chinese business tours go to Japan for learning purposes. Leisure tourists are usually impressed by the products and services in Japan. The satisfaction level of Chinese tourists to Japan has been very high. South Korea is different in this regard (ID1, female, 26–35).

Destination market structure could alleviate or intensify the negative impact from one market. From a social network perspective, in the case of the impact of THAAD on South Korea's tourism, a slump from the Chinese market may lead to the boost of inbound tourists from other markets. Yet, tourists from the other markets were not able to offset the loss from the Chinese market due to its scale and dominance. One informant commented that Chinese consumers used to constitute 70 percent of the customers of Lotte Group. It was a market too large to be replaced easily. Perhaps the only group of Chinese tourists continued to visit South Korea was Daigou (i.e., individual or a syndicated group located outside China purchase commodities for customers in China(ID3, female, over 45).

Cultural export from the destination regions (in this case Japan and South Korea) to the tourist generating regions (in this case China) shapes destination image and attractiveness. Cultural export activities and status during the crisis events period play a role in the formation of impact and post-events recovery. Both Japan and South Korea have strong cultural connections with China. Japanese TV program, animated culture and imagery have had long influence in East Asia (Iwabuchi, 1998). Japanese animation has been marketed with substantial financial support by the Ministry of Foreign Affairs (MOFA) of Japan and private companies and localized by anime and manga fans (Otmazgin, 2012). As a result, Japan has an image of a culturally exciting country – ‘cool Japan’ (Otmazgin, 2012). South Korea has also been active in exporting culture in the recent decade. The Korea Wave (Hanliu – the popularity of Korean pop culture) was perceived as successor to the Western influence dominant in China in the 1980s and the Japanese influence of the 1990s (Yang, 2004). Hanliu resulted in increased interest in Korean tourism resources and positively affected tourism to South Korea (Kim, Kim, Agrusa, & Lee, 2012). Informants pointed out that, during and even after THAAD, on Chinese online streaming platforms, South Korean popular drama and entertainment is removed from the shelf of main entertainment portals of China. This adversely affected tourists' interest in the country.

Jeju Island became a popular destination partially contributing to the Korean dramas, but you don't see much Korean dramas or entertainment programs now. This resulted in no interest in tourism to South Korea. Also, I believe Chinese entertainment giant has bought a number of Korean TV programs and localized these programs. This further lowers Chinese’ interests in Korean culture (ID2, female, 36–45).

I don't think the Japanese anime culture was affected by political issues among young Chinese (ID5, female, 26–35).

Development stage of tourism to the crisis-affected destination country/region in the source market also plays a role in the impact. For example, South Korea had been one of the top destinations for Chinese tourists in the last decade; many Chinese tourists in first, second and even third tier cities had been to the country, pushing the development to the consolidation or stagnation stage on Butler 1980 tourism area life cycle model. This propels challenges post crisis events as the destination already faces rejuvenation issues. One informant described:

After the long halt on travels to South Korea due to THAAD, Chinese tourists have almost lost interests to visit South Korea, as Chinese tourists have transferred to new/trendy destinations (e.g., Myanmar, Laos) as South Korea is no longer a hot destination for Chinese tourists (ID7, male, 36–45).

Source market tourism industry structure further contributes to the changes in tourists’ interests in the destinations. There used to be a hierarchical system in Chinese outbound tourism, in which tourists in the first tier cities lead the trend and tourists in second and third tier cities follow suit. However, with the development of information technology and increased monopoly of organized tours by big tour companies via charter planes and bulk purchasing for new destinations, there is a trend that destination development becomes more synchronized. That is, tourists throughout the country, regardless of their region of residence, could be mobilized to visit emerging trendy destinations. This is especially true for short-haul destinations where charter planes are easily manageable. One informant commented:

If a destination becomes hot in one part of the country, it becomes a hot destination for the whole country. For example, Myanmar is the new hot destination, now you see all operators in China advise Myanmar and airline companies open new flights to Myanmar (ID7, male, 36–45).

Based on abstracting concepts from the interview data and using an inductive approach, Fig. 5 is developed to sums up the key influencing factors on tourism demand during and post crisis events.

Fig. 5

An integrated guide for post crisis events recovery and growth.

Conclusion

Using both quantitative and qualitative approaches, this paper tested the impact of five major crisis events affecting the flow of Chinese tourists to Japan and South Korea from 2005 to 2017 and tried to explore why destinations suffer or recover at a different pace. Political crisis events at the national level had demonstrated a greater impact concerning both the intensity and duration than natural disasters, epidemic disease and financial crisis. South Korean's tourism industry was still reeling from the THAAD turmoil by the end of 2017 and beyond. The findings suggest that long-term political tension may exert the most negative impact on tourism development. The impact of political tension varies based on the level of involvement of the governance bodies in regulating tour business operations and managing social influence. Results show that sound international relations are important for bilateral tourist flows and development of tourism. When international relations deteriorate, the tourism industry is more likely to be adversely impacted than other exports and trades, resulting in the loss of economic gains for the host country as well as the tourism businesses of the source country. Governments need to maintain a sound bilateral political relation for the development of tourism. This paper also indicates the effect of convention and social media on managing social sentiment during and post crisis events, which echoes prior research such as Faulkner (2001) that the role of media can be crucial in the impact of crisis on tourism. The results also indicate that, in various crisis contexts, tourists cancel their travel plans due to fear of health concerns, physical risk or social pressure. With other determinants of tourist demand remain constant, once the fear or pressure is vetoed, the potential source market will recover at a relatively rapid pace, as most tourists only temporarily lose interest in the destination country. In post crisis events recovery and growth, the travel industry infrastructure and structure at both the generating and destination regions, the status of the tourism products, and the attractiveness of the destinations all play important roles. It proves that destination government policy per se can have a significant impact on the directions of tourist outflows. Political, social, and cultural exchanges between the generating and destination regions, especially cultural export from the destination regions to the generating countries may assist in tourism development in the destinations. The paper posits that a severe crisis impact was created by the intertwine of the political, social and tourism systems. At the time of crisis, crisis management capabilities, timely positive publicity, and effective marketing campaigns will help to recover the potential source market in the affected region. At the post crisis stage, status and degrees of political, economic, social and cultural relations between the two regions largely determine the post crisis tourist flow. This raises significant questions as to the reactive nature of tourism versus the development of greater predictive capacity and theory generation.

Author contributions

Mingya Qu undertook the quantitative work for Study One and participated in the interviews for Study Two. She drafted an early version of Study One. Jigang Bao contributed to the conceptualisation of the research, conducted the interviews, supervised Mingya throughout the conduct of the research, and contributed to the writing of the final manuscript. Xin Jin contributed to the conceptualisation and design of the research, conducted the analysis for Study Two, reworked the preliminary draft of Study One, and completed the writing of the paper.