ALCOHOL PRICE POLICIES: CONNECTING SCIENCE TO PRACTICE
Article first published online: 5 FEB 2010
© 2010 The Authors. Journal compilation © 2010 Society for the Study of Addiction
Volume 105, Issue 3, pages 398–399, March 2010
How to Cite
WAGENAAR, A. C. (2010), ALCOHOL PRICE POLICIES: CONNECTING SCIENCE TO PRACTICE. Addiction, 105: 398–399. doi: 10.1111/j.1360-0443.2009.02862.x
- Issue published online: 5 FEB 2010
- Article first published online: 5 FEB 2010
The alcohol price policy modeling work of Meier and colleagues  reported in this issue is creative, innovative and highly significant. It substantially advances both science and policy. They make effective use of two large national surveys in the United Kingdom to provide a substantive contribution to the state of knowledge on alcohol price elasticities stratified by important population subgroups based on age, sex and drinking level. While the literature on alcohol price elasticities is extensive , few studies provide results for important population subgroups to the degree of specificity reported here; but the contribution of this paper is much larger. It advances the translation of the large evidence base from the basic science of alcohol elasticities into practice—specific, feasible policy options for substantially reducing morbidity, mortality and economic costs due to risky drinking.
Much of what we know in terms of effective measures for the prevention of alcohol-related damage is not put into practice. This problem transcends fields. Ringwalt et al. find only 10% of school districts in the United States implement evidence-based drug prevention programs. Baker et al. report that most psychotherapists base treatments on their personal experience, largely ignoring available science on treatment efficacy, an ‘unconscionable embarrassment’ according to an accompanying editorial . Primary care physicians in the United States and Europe apparently implement only half of the established clinical guidelines [6,7]. The disconnect between science and practice also often holds true for alcohol policy. In the United States, where alcohol taxes are not indexed to inflation, real tax rates today are only a fraction of what they were in the 1950s and 1960s, despite decades of growing research documenting the role of alcohol in leading causes of death and injury, and establishing the effects of alcohol prices on drinking. The work of Meier's team builds directly on the extant evidence base to develop specific practicable policy proposals and model the expected health and fiscal outcomes from each policy alternative. This is exactly the type of translational science needed by policy makers, bridging basic research results to policy practice.
Reader reaction to the paper may cluster in two areas: scientific methods and policy implications. First, in terms of the methods, one might always note methodological details that are less than perfect. Are all their elasticity estimates exactly right? Are their population estimates of drinking and alcohol expenditures precisely correct? Is the survey sample optimally representative? Were the measurement items used in the surveys the highest possible in terms of reliability and validity? Are every category cut-point and every assumption implied in the scenario models exactly the values you or I would have chosen? Most scientists would probably respond ‘no’ to at least one of these. Plus, the project would surely have been strengthened further if longitudinal data over many years had been available. However, it is critically important to note that the pattern of results across a complex matrix of policy options and population subgroups is consistent with well-established theory, and consistent with a very large body of previous empirical research, providing a strong indicator of validity of the models used here.
Meier's report raises numerous interesting and important policy issues. First, there is a clear tension regarding the most appropriate level of aggregation for examining differential policy effects. Epidemiologists and public health professionals are typically most interested in measuring total population-level health effects. We know from a long tradition of experience across many health outcomes that focusing solely upon the high-risk segment of the population produces only modest gains, compared to population-wide interventions . Moreover, health-related behaviors diffuse quite widely through societies in a web of social influences, and efforts to shape the behavior of the few are difficult without also shaping the behavior of the many. In a population, how much the average person drinks affects how much the heavy drinkers drink . Consequently, those with a public health orientation recommend increasing the overall alcohol tax rate. Nevertheless, for many others, the intuitive appeal remains of policy proposals that target higher-risk segments of the population (e.g. young adults, those drinking in bars or heavy drinkers).
Price policy effects are reported for subgroups based on age, sex and drinking pattern, but many other dimensions for such disaggregated effect estimates are also relevant. In many countries some policy options are limited to, or more feasible at, specific levels of government (local, regional or national). Policy makers are most interested in effects of policy options at their disposal, and most interested in effects on their particular constituents; Meier's modeling framework can be adapted easily for such alternative policy matrices in future replications across diverse jurisdictions.
Meier's results show that the average price paid declines the heavier one drinks. This probably reflects a combination of selection and causal effects. Heavier drinkers seek out lower-priced drinks, and drinkers are induced to drink more in the presence of cheap alcohol compared to situations where it is more expensive. These effects are probably quite complex in practice, and causally interactive effects exacerbate the deleterious effects of low alcohol prices. Heavy drinkers and those vulnerable to heavy drinking because of genetic factors and previous environments are attracted to low-cost drinking locations. This situational concentration of heavier drinkers compounds the risk, because any given drinker drinks more in the presence of heavier drinkers than in the presence of lighter drinkers . An obvious policy implication is to structure price policy along with a broader set of alcohol regulations to avoid creating environments that attract or encourage heavy drinking, and designed to maintain consistently a population of patrons where the supermajority exhibits light drinking.
Relatively few previous studies have examined price elasticity specifically among heavy drinkers, finding typically that they are less price-elastic than lighter drinkers. Results reported here advance our knowledge substantially with estimates of heavy drinker cross-elasticities by beverage and by location. Heavy drinkers are found to be more own-price responsive and more cross-price responsive. Thus, when the price of their chosen alcohol product increases, they are more likely to reduce consumption of that product than lighter drinkers, but also more likely to switch to another lower-cost brand, another type of alcoholic beverage or another lower-cost location for drinking. This argues for broader-based price policies (such as overall alcohol tax rates), rather than focused policies limited to affecting prices of specific products in specific venues.
Reported policy effects might be overstated because, as the authors note, they do not account for the potential response of alcohol producers, distributors and retailers to changes in price policies. We know from experience with tobacco where tax and other policies that substantially increased tobacco retail prices generated numerous product design, marketing and pricing innovations (e.g. couponing) designed to minimize sales declines .
Reported policy effects are also underestimated by Meier. Alcohol, tobacco and illegal drugs are complements, not substitutes . Thus, a price policy that reduces drinking will also reduce tobacco and drug use in the population, producing additional health benefits and cost savings. Other effects may emerge from structural changes in local alcohol markets. Minimum-price policies may reduce price competition across establishments, increase profitability and lead to increased competitive attention to non-drinking elements of the experience marketed. The often poorly managed, highest-risk outlets with little to offer except cheap drinks might be more likely to close.
The options evaluated by Meier are not mutually exclusive, and optimum alcohol price policies will combine increases in base tax rates, along with minimum pricing standards. As diverse jurisdictions implement such policy changes, continued evaluation studies using strong multiple time–series designs with high levels of internal validity will advance the science and thus, in turn with patience, the practice of alcohol prevention policy.
Declaration of interest
Current related research projects are funded by the US Robert Wood Johnson Foundation
- 4Current status and future prospects of clinical psychology: toward a scientifically principled approach to mental and behavioral health care. Pyschol Sci Public Interest 2009; 9: 67–103., ,
- 5Connecting clinical practice to scientific progress. Pyschol Sci Public Interest 2009; 9: i–ii.
- 8The Strategy of Preventive Medicine. Oxford: Oxford University Press; 1992.
- 12Paying the Tab: The Economics of Alcohol Policy. Princeton: Princeton University Press; 2007.