Cohort, age and time effects in alcohol consumption by Italian households: a double-hurdle approach

Abstract

Abstract

A double-hurdle approach to modelling tobacco consumption in Italy

“…These outcomes are in line with the findings of Garcia and Labeaga (1996) and imply that tobacco expenditure rises as household income increases, but at a decreasing rate. The estimated age effects are negative and significant in both participation and consumption equations, indicating the existence of a strong lifecycle pattern (Kerr et al, 2004;Aristei et al, 2005). As in Jones (1996) andYen (2005b), the probability of smoking and tobacco expenditure levels decrease with the age; moreover, the estimated age squared parameter shows a positive non-linear relationship between consumption levels and age.…”

Türkiye’de Ev Dışı Gıda Tüketiminde Kuşak Etkisi: Double Hurdle Yaklaşımı

“…As we mentioned above the main aim of this paper is to investigate cohort, time and age effects on FAFH expenditure in Turkey based on the pioneer studying of Aristei et al (2008) and Deaton and Paxson (1994). There are the linear relationship between age, birth and the time.…”

“…For this reason it is not possible to identify the effects of age, birth's year and time separately and estimating the time effect on FAFH expenditure requires a special specification. According to Deaton and Paxson (1994), one way of solving this problem is to assume that the time effects capture cyclical fluctuations that average to zero over the long run and excluding first two year dummies in the model estimation and specifying the remaining ones as (Aristei et al 2008):…”

“…Each household budget survey is conducted by random sampling in that particular year (2002, 2005, etc.) Therefore, the panel data which were built upon are not a fixed set of households sampled across the years. Five-year age bands cohorts were used, depending on the age of the household head, following the studies Browning et al (1985) and Aristei et al (2008). The analysis is restricted to those households headed by individuals aged between 20 and 82.…”

“…Any trends in the data can be arbitrarily attributed to either the year effect or to a combination of age and cohort effects. This leads us to drop one column from both the age and cohort matrices of dummy variables, and the first and second year dummy variables (see Aristei et al, 2008 for details). Moreover, first two dummies which created for years 2002 and 2003 were dropped from this study.…”