Malleability of alcohol consumption
Alcohol is one of the leading killers among substances. The World Health Organization estimates that alcohol is responsible for about 5% of all deaths (WHO 2018). Beyond direct health consequences, excessive alcohol consumption also generates broader social and economic burdens (Cook and Moore 2000, Cawley and Ruhm 2011).
Various factors affect alcohol consumption, including taxes and other regulations, peers, and social norms. To identify each factor’s direct impact, studies have used changes in regulation (e.g. Marcus and Siedler 2015) or random assignment of peers (e.g. Eisenberg et al. 2014). These studies give us a well-identified local average treatment effect in the short term. But in the long term the various factors tend to interact, either magnifying or decreasing each one’s direct effect.
For example, peers who do not consume alcohol could vote for stricter alcohol regulation; strict alcohol regulation could lead to norms of consuming less alcohol and affect how children grow up viewing alcohol, which again change the norm and lead to a different local environment. Quantifying the overall impact of the environment is important, as it tells us how malleable alcohol consumption is.
Measuring the combined effect of these factors is challenging because, typically, it requires us to look at a long time horizon. But over the long term, economic conditions change in ways that can also affect alcohol consumption. Therefore, little is known about the magnitude of the combined effect. A notable exception is Yakovlev (2018), who uses a structural model and data on alcohol consumption and peers to estimate the impact of an increase in vodka price in Russia. He finds that peer effects play a large role in magnifying the impact of the price increase.
The importance of the environment
In a recent paper (Hinnosaar and Liu 2020), we measure how much the current environment drives alcohol purchases by analysing changes in consumer behaviour after a move from one state to another. Using data from a panel of individuals who moved in the Nielsen scanner data of alcohol purchases in the US, we observe their alcohol purchases in years before and after the move. The magnitude of the change in the movers’ alcohol purchases allows us to measure the relative importance of the current environment.
Our empirical strategy relies on the fact that the environment – including supply conditions, alcohol regulation, taxes, and movers’ peers – changes substantially when consumers move. If the current environment mainly drives alcohol consumption, we would expect a jump in the mover’s purchases to a level similar to that of other consumers in the destination state. On the other hand, if alcohol consumption is only driven by individual characteristics such as personal preferences and past experiences, we would not expect a change in the mover’s alcohol purchases.
Movers sharply adjust their purchases
Right after a move, consumers’ alcohol purchases converge sharply toward the average level in their destination state, implying that the current environment explains a large share of the differences in alcohol purchases. Figure 1 presents the coefficients from the event study regression. About two-thirds of the gap in alcohol purchases between the origin and destination states close immediately when a consumer moves. No sizeable further convergence is seen after the immediate jump.
Figure 1 Convergence of movers’ alcohol purchases towards the average of the destination (event study estimates)
Adjustment on both extensive and intensive margin
The adjustment takes place both on the extensive and intensive margins. On the extensive margin, movers are more (less) likely to buy any alcohol when moving to a state with a larger (smaller) share of consumers purchasing alcohol. On the intensive margin, movers who bought alcohol before the move adjust the quantity in the direction of the average purchases in the destination state. This is illustrated in Figure 2, which presents similar estimates to those of the event study but pools all time periods before the move and all time periods after the move.
Figure 2 Change in alcohol purchases after the move (difference-in-differences estimates)
Source: Hinnosaar and Liu (2020), Table 2.
Asymmetries and heterogeneity
The adjustment is asymmetric, but mainly on the extensive margin. Consumers adjust more when they move to states with larger average alcohol purchases and adjust less when moving to states with smaller average alcohol purchases. Perhaps it is easier to start consuming alcohol (or consume more) than to stop or sharply decrease consumption. The adjustment is also heterogeneous across product types: consumers adjust their wine purchases more and their liquor purchases less.
Not driven by life events that affect both moving and alcohol purchases
A possible concern with our identification strategy is that a move could occur as a response to a life event that could itself affect alcohol purchases. To alleviate this concern, we provide two pieces of evidence. First, we restrict the sample to movers whose observable characteristics like household size, employment, and marital status don’t change and find that our results remain similar.
Second, we compare trends in pre-move purchases of movers to higher- versus lower-alcohol-purchasing states. This shows that the movers who chose to go to different states before the move had similar trends in their purchases. We perform several robustness checks using alternative samples, functional forms, controls, and geographic aggregation levels. Throughout, we find that movers’ alcohol purchases converge sharply toward the average level of their destination.
In the context of previous findings
Let us place these findings in the context of the literature that has studied how much purchases are influenced by the current environment. Bronnenberg et al. (2012) found that the current environment heavily influences brand purchases, and consumers change which brands to buy right after a move. They also found, similar to our study, that 60% of the gap between the destination’s and the origin’s average purchases of grocery products is bridged immediately after the move.
On the other hand, the literature on the healthiness of food purchases hasn’t found much evidence that the current environment matters. Specifically, Allcott et al. (2019) and Hut (2020) study how the healthiness of food purchases changes with a move. Both find that the magnitude of the change in the few years after the move is very small.
Our findings on alcohol are in contrast to the evidence of little change in the healthiness of food purchases; instead, they are more in line with larger changes in brand choices. We hypothesise that large regional differences in alcohol regulation (availability and taxes) are the main reason for the large adjustment in alcohol purchases. Large regional differences in supply conditions are absent in the case of food healthiness but do exist for brands.
The recent work of Chetty et al. (2016) and Chetty and Hendren (2018) finds that where one grows up affects long-term outcomes such as intergenerational mobility and earnings. Our findings partly echo this point and could provide an additional mechanism of why the environment matters. According to our findings, the current environment largely determines individuals’ alcohol purchases.
Using a simple back-of-the-envelope calculation would suggest that if a household of two adults and one underage child moves from Utah to New Hampshire, the family’s alcohol consumption would permanently increase by $27 per quarter. This permanent shift in the household’s alcohol consumption could have a direct impact on household asset accumulation as well as other indirect impacts, such as on household earnings, crime, and potential exposure to alcohol abuse. This permanent increase could affect the wellbeing of both the household heads and their child.
Allcott, H, R Diamond, J-P Dubé, J Handbury, I Rahkovsky and M Schnell (2019), “Food deserts and the causes of nutritional inequality”, The Quarterly Journal of Economics 134: 1793–844.
Bronnenberg, B J, J-P H Dube and M Gentzkow (2012), “The evolution of brand preferences: Evidence from consumer migration”, American Economic Review 102: 2472–508.
Cawley, J, and C J Ruhm (2011), “The economics of risky health behaviors”, in M V Pauly, T G Mcguire, and P P Barros (eds.), Handbook of Health Economics, vol. 2, Elsevier, 95–199.
Chetty, R, N Hendren and L F Katz (2016), “The effects of exposure to better neighborhoods on children: New evidence from the moving to opportunity experiment”, American Economic Review 106: 855–902.
Chetty, R, and N Hendren (2018), “The impacts of neighborhoods on intergenerational mobility II: County-level estimates”, The Quarterly Journal of Economics 133: 1163–228.
Cook, P J, and M J Moore (2000), “Alcohol”, in Handbook of Health Economics, vol. 1, Elsevier, 1629–73.
Eisenberg, D, E Golberstein and J L Whitlock (2014), “Peer effects on risky behaviors: New evidence from college roommate assignments”, Journal of Health Economics 33: 126–38.
Hinnosaar, M, and E M Liu (2020), “Persistence in alcohol consumption: Evidence from migrants”, CEPR Discussion Paper 15196.
Hut, S (2020), “Determinants of dietary choice in the US: Evidence from consumer migration”, Journal of Health Economics 102327.
Marcus, J, and T Siedler (2015), “Reducing binge drinking? The effect of a ban on late-night off-premise alcohol sales on alcohol-related hospital stays in Germany”, Journal of Public Economics 123: 55–77.
World Health Organization (2018), Global status report on alcohol and health 2018, Geneva: World Health Organization.
Yakovlev, E (2018), “Demand for alcohol consumption in Russia and its implication for mortality”, American Economic Journal: Applied Economics 10: 106–49.