Abstract
Empirical evidence gives strong support to a close association between liver cirrhosis mortality and the intake of alcohol and most often a log-linear relationship is assumed in the econometric modeling. The present analysis investigates for unit roots in a panel data set for sixteen European countries - covering the period 1970-2006 - where both alcohol consumption and liver cirrhosis seem best described as trend-stationary variables. Therefore a fixed effects model including individual trends is applied in the analysis but also a more flexible non-linear functional form with fewer restrictions on the relationship between liver cirrhosis mortality and alcohol consumption is included. The conclusion is that the total level of alcohol consumption as well as the specific beverages - beer, wine and spirits - contributes to liver cirrhosis mortality, but the present study also reveals that directly addressing the question of panel unit roots and in this case subsequently applying a trend-stationary modeling methodology reduces the estimates of the impacts from alcohol consumption to liver cirrhosis. Finally, more restrictive alcohol policies seem to have positively influenced the country-specific development in cirrhosis mortality.
Original language | English |
---|---|
Place of publication | Aarhus |
Publisher | Aarhus School of Business |
Number of pages | 28 |
ISBN (Print) | 9788778824547 |
ISBN (Electronic) | 9788778824554 |
Publication status | Published - 2010 |
Keywords
- Alcohol consumption .
- Liver cirrhosis mortality
- Trend-stationary panel data
- Non-linear modelling