Summary: | 碩士 === 國立臺灣海洋大學 === 應用經濟研究所 === 100 === According the statistic data from Department of Health, most of Taiwanese have cancer or heart disease that cause their death. The lung cancer and the liver cancer are the most ones among all types of cancers. Excessive smoking and drinking are possibly causing heart disease and many kinds of cancer. Thus, if we want to reduce the negative impact on health because of tobacco and alcohol, we have to understand what factors that causes the household has difference tobacco and alcohol expenditures. On recently ten year, the basic household consumer expenditure on food, clothing and living had decrease even though the household income increases. Replace those basic household consumer expenditure are medical expenditure, education expenditure and entertainment expenditure. The main purpose of this study is to observe what kinds of family characteristic cause household had differently tobacco and alcohol expenditure. The other purpose is to know if different tobacco and alcohol expenditure had different affects on household medical expenditure. The sources of information used in this study was the DGBAS “Household Income and Expenditure Survey Report” in 2010. This study uses the quantile regression analysis to analyze the relationship between dependent variable and independent variable.
The results of quantile regression on tobacco and alcohol expenditure equation displaying that the age and education of household owner, the members more than 65 years old in house are negative relationship with tobacco and alcohol expenditures. The owner’s gender as male, number of employed person in house, single-person households, income of household are positive relationship with tobacco and alcohol expenditures. Owner’s work profession also had effect on tobacco and alcohol expenditures. If owner’s work related to education and medical will cause tobacco and alcohol expenditures decrease.
The results of quantile regression on medical expenditures equation displaying that the size of household, the members more than 65 years old in house, household income are positive relationship with medical expenditures in each quantile. The other family characteristics had different effect extent in each quantile of medical expenditure, and those family characteristics will be change the direction of influence on medical expenditure. In lower and highest quantile of medical expenditure that how much tobacco and alcohol expenditures was not significantly effect on medical expenditure, that’s means tobacco and alcohol expenditures was not the main factor cause household had different medical expenditure in lower and highest quantile. Even so, if household had higher tobacco and alcohol expenditures will pay more on medical.
The result show that using quantile regression analysis can observe those factor cause different effect on each different quantile of expenditure, not only on average level. The results also show that different household characteristics cause different tobacco and alcohol expenditures, also medical expenditure.
|