What is the impact on alcohol purchases of increasing the price of sugary drinks?

With the global prevalence of obesity rising, fiscal measures to reduce the consumption of energy-dense foods, and particularly sugar-sweetened beverages (SSB), are being advocated and implemented in many countries. In the UK, the government has proposed an industry levy on SSBs to be implemented in 2018. However, even if the levy were to be passed through fully and result in higher prices for SSBs, the overall health impact depends on the complement and substitution relationships across the whole diet. An aspect that receives less attention is the contribution of consumption of alcoholic beverages on the energy intake of households, with beer and wine providing on average 43 and 85 kilocalories per 100 ml, respectively; in comparison to an average of 40 kcal in 100 ml of Cola drink. Limited references in the literature are available considering alcohol consumption from the energy intake perspective, but it is not clear how public health policies targeting either the price of non-alcoholic or alcoholic beverages affect the consumption of alcoholic drinks vis-à-vis SSBs. In this paper we draw on a comprehensive dataset of individual purchases of food and drinks for 30,000 households in the UK (2012-2013) which allows us to estimate the own- and cross-price elasticities for SSBs, considering alcoholic and non-alcoholic drinks. Overall, households in the sample on average spent £17.5 and £10.3 per month on alcoholic and non-alcoholic drinks respectively. However, the share of households who never purchased a particular drink category ranged from 16% for fruit juices to more than 50% for certain alcoholic drinks such as cider or sparkling wine. The estimated demand model will account for (i) censoring of expenditure at zero, (ii) endogeneity of prices and expenditure, as well as (iii) unobserved heterogeneity at the household level.