본 연구는 칼만필터 알고리즘의 관광수요모형을 이용하여 1995년부터 2012년까지 인바운드 외래관광객의 관광수요의 탄력성, 구조적 변화 및 변동성을 조사하고자 하였다. 첫째, 미국, 일본 및 중국의 자체가격탄력성이 모두 탄력적이고 미국은 보완재, 일본과 중국은 대체재의 관계, 미국의 소득탄력성만이 탄력적으로 사치재로 취급되고 교차가격탄력성은 모두 비탄력적이었다. 둘째, Chow 결과에서 미국은 글로벌 금융위기기간인 2007년(13)에 구조적 변화가 존재하고 1-Step의 결과에서 2003년과 2012년은 계수값이 안정적이나, N-Step의 결과에서 2000년 초부터 2006년 말까지의 계수값이 불안정하였다. 또한 CUSUM 결과에서는 모든 계수값이 안정적이나, CUSUMSQ 결과에서 2003년 초부터 2006년 말까지 계수값이 불안정하였다. 일본은 1999(5)년부터 2008(14)년까지 구조적 변화가 존재하고 1-Step의 결과에서 2008년은 계수값이 불안정하고, N-Step의 결과에서 1999년, 2005년, 2006년, 2007년은 계수값이 불안정하였다. 또한 CUSUM 결과에서는 2008년 이후부터 2012년 말까지 계수값이 불안정하고, CUSUMSQ 결과에서 2006년 초부터 2007년 말까지 계수값이 불안정하였다. 중국은 1999년(5)부터 2007년(13)까지의 기간에 구조적 변화가 존재하고 1-Step의 결과에서 2006년과 2007년은 계수값이 불안정하고, N-Step의 결과에서 2000년, 2001년, 2004년, 2005년, 2006년은 계수값이 불안정하고, CUSUM 결과에서는 2007년 이후부터 2012년 말까지 계수값이 불안정하고, CUSUMSQ 결과에서 모든 기간의 계수값이 안정하게 나타났다. 셋째, 시간가변 수요탄력성의 결과, 2007년과 2008년에 미국과 일본의 자체가격탄력성은 2000년대 초보다 대폭 높고, 2010년 2011년의 자체가격탄력성에서 미국과 중국은 대폭 증가하였다. 2007년과 2008년의 소득탄력성에서 미국과 일본은 약간 높고, 반면에 2010년 2011년의 소득탄력성에서는 미국과 일본이 증가되었다. 넷째, 다변량 비대칭 BEKK 모형의 조건부 분산식에서 대부분의 일방향 또는 양방향으로 강한 조건부 변동성전이효과가 존재하고 자체 또는 변수간에서 비대칭효과가 존재하였다.
따라서 본 연구는 관광객들의 행동변화가 관광수요의 탄력성, 구조적 변화 및 변동성에 영향을 미친다는 것을 확인하고 상태공간모형은 전통적인 관광수요모형보다 더욱 우수하고 소비자의 취향, 기대감, 관광정책과 관광체제전환과 같은 비관측요인들에 의해서 자주 변경되고 있는 관광수요의 탄력성, 구조적 변화 및 변동성을 파악하고자 할 때 유용하였다.
The purpose of this study was to examine the elasticity and structural changes and the volatility in the tourism demand of inbound foreign tourists using the model of tourism demand of Kalman filter algorithm from 1995 to 2012.
The results of this study can be summarized as follows: First, the own price elasticities of the US, Japan, and China were elastic in all periods, indicating there were relations with complements in the US and substitutes in Japan and China. And the income elasticity of the US was treated as luxuries and the cross price elasticities of three countries were inelastic and insensitive.
Second, the coefficients of the US showed structural changes in Chow tests in 2007(13) including the global financial crisis and the coefficients were stable in the 1-Step prediction error tests of 2003 and 2012, but unstable in N-Step prediction error tests from 2000 to 2006. The coefficients also were stable in CUSUM tests and unstable in CUSUMSQ tests from 2003 to 2006. And the coefficients of Japan presented structural changes in Chow tests from 1999(5) to 2008(14) and the coefficients were unstable in the 1-Step prediction error tests of 2008 and unstable in N-Step prediction error tests in 1999, 2005, 2006 and 2007 and unstable in CUSUM tests from 2008 to 2012 and unstable in CUSUMSQ tests from 2006 to 2007. Additionally the coefficients of China showed structural changes in Chow tests from 1999(5) to 2007(13) and the coefficients were unstable in the 1-Step prediction error tests of 2006 and 2007 and unstable in N-Step prediction error tests in 2000, 2001, 2004, 2005 and 2006 and unstable in CUSUM tests from 2007 to 2012 and stable in CUSUMSQ tests in all periods.
Third, in the results of the time-varying demand elasticity, the own price elasticities of the US in 2007 and Japan in 2008 were significantly higher than in the early 2000s, and the US and China had greatly increased in the own price elasticities of 2010 and 2011. The income elasticities of the US and Japan were slightly higher in 2007 and 2008, whereas those of the US and Japan were increased in 2010 and 2011.
Fourth, there were one-way or two-way strong volatility spillover effects and asymmetric effects in the conditional variance equations of the multivariate BEKK model.
As implications, first, the state equation includes the assumption that follows the probability of Random Walking process, and can be compared to the predicted future performance, including tourism products and tourism services. Second, the geographical or seasonal nature of Korea needs to take advantage of the environment according to the season. Also, if you correctly forecast the Korea Tourism demand, it will be able to improve the economic efficiency of the various policies and actions of tourism policy and economic subjects. Third, tourism strategy and measures to overcome the seasonality of acceptance for enlargement can see it necessary. Fourthly, there is a state space model can be used to analyze the elasticity related to the determinant of the long-term demand for tourism in Korea. If the forecast performance of the state-space model can be compared with alternative tourism demand models, tourism practitioners can help even more. Fifth, the test procedure is theoretically due to the structural instability of the model to be developed in the future. Six, aggressive regulations to attract tourists should be relaxed. Seven, as a prerequisite for attracting tourists, it is needed to be redevelopment of tourism infrastructure. Eight, it is necessary to enforce foreign tourists in tourist promotion policies in order to expand revenue from foreign tourists. Nine, shopping tourism as part of personalized marketing to attract tourists needs to be activated. Ten, it is considered necessary the development of domestic tourism products competitive with aggressive measures to attract tourists. Eleven, Korea’s tourism industry should actively respond and participate in the competitive pricing or non-price competition in order to preempt inbound foreign tourists.
Thus, the findings provided new insights into the understanding of the dynamics of tourists’ consumption behaviour regarding a group of tourism goods and services and to evaluate the competitiveness of Korea as an international tourist destination, and had some useful implications that might assist the relevant stakeholders in both the public and private sectors in Korea to make strategic decisions regarding tourism planning and management.
In conclusion, we confirmed that inbound tourist behavior changes had affected on the elasticity and structural changes and the volatility using the state-space model with Kalman filter algorithm. Moreover the tourism demand model was superior to the traditional model and it was useful when you want to identify the elasticity and structural changes and the volatility in tourism demand that were changed by the unobservable factors such as consumer tastes, expectations, tourism policy and tourism regulation transition.