本文使用世界252個國家1960-2013年之飲酒量追蹤資料，分別為總飲酒量、烈酒、葡萄酒及啤酒，估計總體變數對飲酒需求的影響。研究方法分為兩個部分，第一部分使用分量迴歸，估計不同飲酒分量下，總經變數對各類飲酒需求的影響。結果顯示在特定分量下失業率、消費者物價成長率、國民平均受教程度、每人平均GDP、每周平均工作時數對飲酒需求影響為正相關，經濟成長及貪汙指數對飲酒需求為負相關，根據酒類的不同，上述變數顯著分量亦不相同。第二部分使用動態追蹤資料系統GMM估計，觀察不同酒類飲酒量落後期對當期飲酒量之影響，發現均為顯著正相關，顯示酒精飲料確實存在成癮性，依照落後期之邊際效果由大至小排列，分別為啤酒、烈酒及葡萄酒。 This paper uses data on the consumption of spirits, wine, and beer from a panel of 252 countries from the period 1960-2013 to study the relationship between macroeconomic variables and alcohol consumption. The results obtained using quantile regression indicate that positively correlated with unemployment rate, inflation rate, corruption, average years of schooling, the level of GDP per capita, and working time are positively correlated with alcohol consumption at certain levels, whereas economic growth is negatively correlated with alcohol consumption. The quantile levels at which these variables are significant differs across types of alcoholic beverages. Results obtained using dynamic panel data models indicate that alcohol consumption is positively correlated with lagged level of alcohol consumption, confirming the presence of alcohol addition.
