The effects of tobacco taxation in Spain: a study of the importance of the substitution effect in the 2008-2012 period

Ángel Martínez Jorge, Francisco de la Torre

2 Dec, 2023

The effects of taxes on the consumption of goods with negative externalities have, for more than a century now, received considerable attention from economic research and also from national policymakers. However, studying the effect of taxes on these products poses a considerable methodological challenge for two reasons: i) the limited availability of public data on household consumption and ii) the absence of suitable case studies offering a quasi-experimental setting in which to analyse their effect.

In this Policy Brief we address the specific case of tobacco taxes in Spain by studying the period between 2008 and 2012, when the largest tax increase on cigarettes so far this century took place and when the prices of tobacco products (mainly cigarettes) rose by around 33%, with an increase in the average price per pack of 56%. These tax increases, which took place continuously over that period, were followed by years in which the price of tobacco products remained practically constant in real terms. The key to this work lies in the fact that, while mainland Spain and the Balearic Islands saw a large tax increase between January 2009 and January 2013, the evolution of tobacco product prices over the same period in the Canary Islands was much more contained: between January 2009 and 2013, tobacco product prices rose by only 11% in the Canary Islands.

During the same period, cigarette consumption in mainland Spain and the Balearic Islands fell very considerably, from an average annual consumption per household of 164 packs in 2008 to 81 packs in 2012, a 40% drop in just five years. However, we cannot know a priori what part of this drop in consumption can be attributed to the tax policy implemented in this period. It is therefore necessary to estimate econometric models that use the most valid control group possible in order to estimate what part of that observed drop can be causally attributed to this tax increase.

Using microdata from the Household Budget Survey (EPF) for the years 2006-2012, we use a difference-in-differences strategy to estimate the effect that this differential tax policy had on the consumption of tobacco products in mainland Spain and the Balearic Islands compared with households living in the Canary Islands.

Specifically, we focus on studying the impact on cigarette consumption per household per year and, additionally, we try to estimate the extent to which households in the treated regions substituted their cigarette consumption with the substitute product par excellence, rolling tobacco. To do so, we estimate our main specifications on household spending on rolling tobacco, and then convert the results obtained into kilograms of rolling tobacco, using price data from the Commissioner for the Tobacco Market.

Taking the results of our main specification, we conclude that this tax policy implemented in mainland Spain and the Balearic Islands resulted in a drop in cigarette consumption of 51 packs per household per year between 2009 and 2012, which accounts for 73% of the drop in consumption observed in that period in the treated territories. However, this did not imply a fall in the percentage of households consuming tobacco products, partly because of the growth in average spending and the increase in the percentage of households consuming rolling tobacco during the same period.

On this last point, we find that the tax increase also had effects on spending on rolling tobacco, which were particularly relevant among the third of households with the lowest economic capacity. Specifically, we estimate that the tax increase on cigarettes between 2009 and 2012 (a period during which the specific tax on rolling tobacco barely changed) indirectly led to an increase in spending on rolling tobacco of 24 euros per household per year, which meant that around 1 in 4 cigarettes that stopped being smoked as a result of the policy were instead smoked as hand-rolled cigarettes. This had significant fiscal consequences, since the shift of part of the demand (260 cigarettes per household per year) to rolling tobacco reduced excise tax revenue by almost 350 million euros, as rolling tobacco is taxed at substantially lower rates. From a health perspective, this shift in consumption also reduced the potential of the excise tax to achieve health improvements, since the evidence shows that both products are harmful to public health.