The Elasticity of Tobacco Products
In order to estimate tobacco price elasticity, we apply two models, one using individual data from HBS and the other using macro data at country level. Both models contain controls for income and we report income elasticity for both models.
Time-series data model
The model is based on the macro-level time series data, i.e. one measure of price and consumption (and other control variables) per observation period and price elasticity estimated via regression (cointegration) analysis. Comparable time-series estimates are available for the period between 2002 and 2016.
| Price elasticity | Income elasticity | |
| Albania | -0.78 | 0.03 |
| B&H | -0.83 | 0.25 |
| Croatia | -0.44 | -0.92 |
| Kosovo | -0.68 | 0.17 |
| Macedonia | -0.47 | -0.16 |
| Montenegro | -0.68 | -1.37 |
| Serbia | -0.76 | 0.33 |
Individual data model
This model is based on the use of micro-level data and has its theoretical foundation in consumer theory. For this estimation we use the HES data for years between 2012 and 2016, and the model proposed by Deaton (1988).
| Price elasticity | Income elasticity | |
| Albania | -0.57 | 0.24 |
| B&H | -0.65 | 0.50 |
| Croatia | -1.07 | 0.91 |
| Kosovo | -0.29 | 0.24 |
| Macedonia | 0 | 0 |
| Montenegro | -0.57 | 0.27 |
| Serbia | -0.45 | 0.53 |
