收入与出境旅游率的非线性关系——基于门槛面板模型的实证证据

Non-linear Relationships between Income and Outbound Tourism Rate:Empirical Evidence Based on Threshold Panel Models

在各种有关出境旅游市场影响因素的研究中，收入水平与出境旅游的关系存在着显著的争议。文章判断二者之间存在着非线性关系，且因国家而异。该研究引入门槛回归模型，以世界95个国家和地区为研究总样本，并按照面积大小划分为3组．利用其1995--2011年的人均国民收入、出境旅游率、年龄结构、教育水平和就业水平数据，对收入和出境旅游率之间的相关关系进行了深入分析。研究发现：（1）各样本中，收入与出境旅游率均呈显著非线性相关。（2）样本不同，收入门槛值的数量和位置不同，促使出境旅游快速发展的收入门槛值也不同。说明国土面积对收入与出境旅游的关系具有潜在影响。（3）各控制变量总体上对出境旅游率具有正面影响。（4）中国出境旅游实际表现超过理论基准，处于快速超前发展的阶段。以上研究结果可以为一个国家或地区判定出境旅游的发展阶段及制定合适的旅游政策提供依据。

Research on factors affecting outbound tourism markets has revealed significant controversies over the relationship between income levels and outbound tourism. This paper compared studies on the relationship between income and outbound tourism rate and found that there was a nonlinear relationship between these two variables, and that the characteristic of this non-linear relationship was different in different countries. To confirm this judgment, we developed a threshold panel regression model based on data obtained from 95 countries and regions across the world. The total sample was divided into three size groups. Information relating to annual income per capita, outbound tourism rates, age structures, education levels and employment levels was retrieved for each country and region for the period 1995-2011. The regression analysis showed that： （a） There was a significant positive non-linear correlation between the income and outbound tourism rates. This means that there are specific income thresholds during the development of outbound tourism. That is, income levels reaching a certain point will further promote or weaken the rapid development of outbound tourism; （b） The relationship between income levels and outbound tourism presented dual threshold characteristics in the total sample and large country sample, and triple threshold characteristics in the medium and small country sample. Thus, the number and position of the thresholds present different levels in different study samples. Correspondingly, the position of the income threshold generally deciding the rapid development of outbound tourism is different. It is the highest in the large country sample, followed by the medium-country sample, and smallest in the small country sample. This illustrates the potential impact of land area as a factor affecting the relationship between income and outbound tourism; （c） In general, age structure, employment levels and higher education levels had positive effects on the development of outbound tourism. An increase in any of these will induce an increase in the rate of outbound tourism; and （d） The analysis also verifies the fact that China＇ s outbound tourism is at an advanced stage of development. In theory, China＇ s outbound tourism should not be in a period of explosive growth while the actual performance is ahead of this theoretical assumption. All the factors considered in this paper might play a role in this situation. However, in the light of China＇ s more recent special development conditions, there will be diverse factors to be further discussed in future studies. For example, the influence of the paid leave system, of outbound tourism policy, of traveling abroad on public funds, of adventure psychology and sports psychology, of luxury consumption, of the intention to study abroad, and of currency appreciation must all play a significant role in this relationship. The findings of this study provide some effective support for a country or region to determine the development stage of its outbound tourism and to make appropriate policies. The findings provide a basis for China to make and coordinate outbound tourism policy in the future. This study is therefore of prime importance not only theoretically, but practically as well.