Model 1† | Model 2† | ||||||||
Mean price model | Median price model | Mean price model | Median price model | ||||||
Choice between tobacco products‡ | MC price | 1.341** | (0.607) | 1.084* | (0.582) | 1.643*** | (0.597) | 1.339** | (0.571) |
RYO price | −0.322 | (0.591) | 0.058 | (0.351) | −0.236 | (0.549) | 0.100 | (0.314) | |
Probability of using RYO§ (conditional on non-smoking MC) | MC price | 0.596 | (0.662) | 0.435 | (0.656) | 0.746 | (0.581) | 0.435 | (0.656) |
RYO price | 0.031 | (0.503) | 0.158 | (0.424) | 0.102 | (0.521) | 0.158 | (0.424) | |
Sociodemographic covariates¶ | x | x | x | x | |||||
Anti-smoking sentiment | x | x | x | x | |||||
Smoking restrictions | x | x |
Source: authors’ calculation based on the 2019 STC-SEE data.
*** p<0.01, ** p<0.05, * p<0.1.
†Model 1 includes control only for s-regional level for anti-smoking sentiment, while model 2 includes controls for both s-regional level for anti-smoking sentiment and restriction levels.
‡Dependent variable in the choice model is a dichotomous variable which takes the value 1 if person is smoking RYO and value 0 if person is smoking MC. Coefficients these rows indicate the per cent change in probability of smoking only RYO rather than smoking only MC resulting from the per cent change in MC and RYO prices.
§Dependent variable in the conditional RYO model is a dichotomous variable which takes the value 1 if person is smoking RYO and value 0 if person is a non-smoker. Coefficients these rows indicate the per cent change in probability of smoking only RYO, compared with being a non-smoker, resulting from the per cent change in MC and RYO prices.
¶Sociodemographic controls include the following characteristics: age (and its square), gender, education. labour market status, household size and personal income category.
MC, manufactured cigarette; RYO, roll-your-own; STC-SEE, Survey on Tobacco Consumption in Southeastern European countries.