The determinants of household vacation travel: some empirical evidence
The determinants of household vacation travel: some empirical evidence
复制标题
家庭度假旅行的决定因素：一些经验证据
DOI：
10.1080/00036848100000027
复制
发表时间：
1981
影响因子：
2.2
通讯作者：
R. Hagemann
中科院分区：
文献类型：
--
作者：
R. Hagemann
令人惊讶的是，对家庭度假旅行的决定因素的估计相对较少。Houthakker（1952）使用英国的数据，得出了“假日旅行”需求的总支出弹性的估计值为3.52* 0.23。此外，Prais和Houthakker（1970）利用英国的数据估计，工人阶级的可比弹性为2.16- 0.17，中产阶级为1.10 - 0.23。在最近的一项研究中，美国旅游数据中心使用1972-73年消费者支出调查（CES）的数据，得出了旅游需求收入弹性的估计值为1.88。在每一项研究中，都指定了某种形式的对数线性或半对数回归方程，并用普通最小二乘估计参数。这意味着不旅行的家庭，其旅行支出为零。排除在样本之外。这在很大程度上代表了样本选择偏差，回归系数可能存在偏差。隐含的假设是，不旅行的家庭对旅行的需求为零。但这一点并不完全是先验的。例如，有些家庭可能在调查前一年已经旅行，并计划在第二年旅行。一个更合适的假设是，不旅行的家庭选择不旅行也是在表现市场行为。在这种假设下，在估计旅行费用函数的参数时，重要的是要考虑非出行者和出行者。在某种程度上，某些因素对旅行可能性的影响可能比对支出的影响更大，因此区分旅行支出的收入弹性和旅行可能性的收入弹性是至关重要的。
Surprisingly, there have been relatively few estimates of the determinants of household vacation travel. Houthakker (1952), using British data, obtained an estimate of 3.52* 0.23 for the total expenditure elasticity of demand for'Holiday Travel'. Also, using British data, Prais and Houthakker (1970) estimated the comparable elasticity to be 2.16-c 0.17 for the working class and 1.10 2 0.23 for the middle class. In a more recent study, the US Travel Data Centre, using data from the 1972-73 Consumer Expenditure Survey (CES), obtained an estimate of 1.88 for the income elasticity of demand for travel.In each of these studies, some form of log-linear or semi-logarithmic regression equation was specified, and the parameters estimated by ordinary least squares. The implication is that non-travelling households, for which travel expenditures are zero, were. excluded from the sample.'This represents to a large extent a sample selec'tion bias, and the regression coefficients are likely to be bia~ ed.~ The implicit assumption is that non-travelling households have a zero demand for travel. But this is not at all clear a priori. For instance, some households may have travelled in the year prior to the survey, and are planning a trip in the following year. A more appropriate assumption is that non-travelling households are also displaying market behaviour by choosing not to travel. Under such an assumption, it is important to in~ lude non-travellers as well as travellers in estimating the parameters of the travel expenditure function. To the extent that some factors may have a proportionately greater effect on the probability of travel than on expenditures, it is crucial to differentiate the income elasticity of expenditures on travel from the income elasticity of the probability of travel.