Table 3

Regression results with individual incomes

Variables	(1)	(2)	(3)	(4)	(5)	(6)
Variables	All	All	Most popular brand	Most popular brand	Other brands	Other brands
Smoker income	−0.026***		−0.073**		−0.017*
	(0.006)		(0.013)		(0.008)
Income to price ratio (× 1000)		−0.500**		−0.653**		−0.378**
Income to price ratio (× 1000)		(0.141)		(0.134)		(0.129)
Dummy for tertiary education	0.053	0.059	−0.377*	−0.380*	0.470	0.464
Dummy for tertiary education	(0.379)	(0.404)	(0.120)	(0.127)	(0.630)	(0.669)
Age	0.062*	0.064*	0.101*	0.103**	0.041	0.045
	(0.028)	(0.028)	(0.032)	(0.032)	(0.038)	(0.038)
Squared age	−0.001	−0.001	−0.001**	−0.001**	−0.000	−0.000
	(0.000)	(0.000)	(0.000)	(0.000)	(0.000)	(0.000)
Dummy for illicit cigarette	1.714***	1.566***			1.679***	1.575***
	(0.290)	(0.302)			(0.298)	(0.290)
Dummy for urban commune	0.439	0.413	−0.192	−0.211	0.669*	0.632*
	(0.225)	(0.218)	(0.171)	(0.173)	(0.271)	(0.279)
Dummy for North	−0.556**	−0.564**			−0.804	−0.803
	(0.154)	(0.143)			(0.503)	(0.467)
Dummy for Central	−0.283	−0.304			−0.330	−0.339
	(0.264)	(0.259)			(0.239)	(0.240)
Constant	0.775	0.911	0.056	−0.054	1.077	1.142
	(0.656)	(0.612)	(1.156)	(1.138)	(0.935)	(0.919)
Observations	2094	2094	443	443	1635	1635
R²	0.075	0.079	0.035	0.034	0.090	0.092

Data come from the Tobacco Consumption Survey conducted in Vietnam in 2017–2018 by DEPOCEN. All regressions are weighted by using sample weight. SEs clustered in accordance with the survey design are reported in parentheses. *, ** and *** denote 90%, 95% and 99% significance levels, respectively. Since all packs of the most popular brand collected are licit, and almost all of them are found in the North, dummies for illicit cigarettes and regions are excluded from the second regression. The dependent variable is our estimated price elasticity at the individual level.
DEPOCEN, Development and Policies Research Center.