Policy commentary

Is Alcohol Consumption Undertaxed?

Dec 18, 2009 | Ian W.H. Parry
This Week's Commentary  Previous Commentaries  Future Commentaries  Objectives

Taxes on beer, wine, and spirits help to deter long-term alcohol abuse and drunk driving, as well as raise revenues for the government. Are prevailing alcohol taxes in the United States (which are low by international and historical standards) anywhere close to their economically efficient level?


Is Alcohol Consumption Undertaxed?
Ian W.H. Parry
December 18, 2009

Because nominal tax rates have not increased in line with inflation, federal and state alcohol taxes have fallen from over 20 percent of pre-tax alcohol prices in 1980 to about 10 percent at present. Is it time to reverse this trend and substantially raise alcohol taxes? There are several factors to consider.

Health and Productivity

Alcohol taxation is warranted to the extent that its consumption leads to broader societal costs—externalities—that are not taken into account by individual drinkers.

One possible externality is the burden of medical treatments, borne largely by the government and insurance companies for alcohol-induced illnesses. Some studies suggest that the resulting annual medical burden easily justifies what federal and state governments collect in alcohol tax revenues (about $15 billion). However, these estimates substantially overstate the external cost because heavy drinkers tend to die younger, lowering the burden of medical costs over their lifecycle.

Heavy drinkers may also suffer difficulties finding and retaining employment as well as concentrating on the job. Although individuals themselves bear much of the cost of reduced workplace productivity and employment, in terms of less take-home pay, the government also shoulders a substantial portion through reduced income and payroll tax revenues. Disentangling, statistically, the productivity effect of alcohol consumption has proved difficult because, for example, a negative association between productivity and alcohol could reflect poor work motivation rather than the actual impairing effects of drinking. Based on the wide range of empirical estimates in the literature, the appropriate tax appears to range from almost zero to as much as 40 percent of pre-tax alcohol prices. 

Drunk Driving

Alcohol-related crashes account for around 40 percent of the roughly 40,000 or so people killed each year on U.S. highways. However, the bulk of fatalities occur in single-vehicle crashes where fatality risks should be taken into account by individuals when they decide whether or not to drink and drive. Studies suggest that only about 20 percent of fatalities in drunk driver accidents represent external risks. Nonetheless, if a statistical life is valued at $6 million (which is approximately the value assumed by the U.S. Department of Transportation), the social costs of these deaths is still over $15 billion a year. Accounting also for broader costs—such as nonfatal injury risks to other vehicle occupants and pedestrians, third-party medical burdens, and property damages—drunk-driver risks appear to warrant an alcohol tax of roughly 30 percent of pre-tax prices.

A more direct response to these risks would be to penalize drunk drivers themselves, rather than taxing all people who consume alcohol. Ramanan Laxminarayan, Sarah West, and I calculated that the average (14-mile) trip by a drunk driver should be taxed at about $20 to account for the full range of external risks.

Unfortunately, only about one in 1,500 trips by drunk driver result in a police-reported accident and subsequent court conviction for the driver. This low detection rate implies that, on average, the optimal fine for convicted drunk drivers—that is, the fine which, when multiplied by the probability of actually paying it, results in an expected penalty of $20 per trip—would be about $30,000. (Ideally, the fine would vary according to accident severity, blood alcohol content, past driving offences, and other related factors.) But this level of fine would be an enormous burden for most people; the average fine for convicted drunk drivers is only about $300 at present.

Another possibility is to impose nonpecuniary penalties such as license suspensions and jail terms, or community service in lieu of jail. However, when valued in monetary terms, the expected cost of these penalties (averaged across states and first-time and repeat offenders) amounts to about $3 per drunk driver trip.



   Ian W.H. Parry

Moreover, unlike fines, nonpecuniary penalties may impose a substantial extra deadweight cost on society because the loss of utility to the individual from a driving ban or jail term is not offset by a corresponding gain in revenue to the government. Yet another option is to mandate that convicted drunk drivers install interlocks requiring them to pass a breathalyzer test in order to start and operate their cars. Experience in New Mexico suggests this policy is highly effective in reducing recidivism.

The practical difficulties of imposing stiff penalties on convicted drunk drivers, and the resource costs involved in apprehending, convicting, and penalizing them, suggest that alcohol taxes still have a role to play as part of a broader package of measures to deter drunk driving. In the absence of more aggressive policies, most of the external costs of alcohol-related crashes can be included in an assessment of optimal alcohol taxes. An alcohol tax of roughly three times the current level of taxes seems to be warranted on efficiency grounds. Higher taxes might also be appropriate if people underestimate the future costs of becoming addicted to alcohol, though economists disagree on whether people do in fact misperceive addiction risks. 

Fiscal Considerations

Is the ability to provide revenue enough reason to set higher levels of alcohol taxation than warranted on externality grounds? Externalities aside, the desirability of partly financing government through alcohol taxes depends on their economic costs compared with other taxes, such as income and payroll taxes.

Taxes on labor income distort the overall level of employment in the economy; for example, by reducing take-home pay, income taxes reduce labor force participation rates. Taxes that fall on specific goods also cause economic costs, by changing household behavior and inducing people to consume less of the taxed product and more of other products than they would otherwise prefer. And, by raising the general level of product prices and depressing the amount of goods people can buy with their earnings, product taxes tend to reduce (albeit very slightly) labor supply across the economy.

Economists usually find that it is less costly to raise revenue from taxes with very broad bases, such as income and payroll taxes, than narrowly focused taxes on specific products. One important exception, however, is products that can be taxed with less effect on economywide employment than the employment effects of raising the same amount of extra revenue through higher income or payroll taxes. Some of my preliminary work with Sarah West suggests that alcohol may indeed be one of these exceptions—implying that fiscal considerations may greatly reinforce the case for higher alcohol taxes. This assumes that alcohol tax revenues are used to cut distortionary income taxes. If, instead, revenues are not used to enhance economic efficiency, the fiscal argument for alcohol taxes is undermined and perhaps reversed.

Effects of Higher Taxes

Suppose that alcohol taxes were increased to 30 percent of pre-tax prices. What effect would this have? Ideally, the tax would rise in proportion to the alcohol content of a beverage, as alcohol content is what matters for external costs. Tripling alcohol taxes would, for example, add roughly $1.20 to the price of both a six-pack of beer and a bottle of wine, and would reduce total alcohol consumption by roughly 10 percent. We estimate that the annual net economic efficiency benefits of the tax increase could easily exceed $10 billion if the revenue displaced other distorting taxes.

This tax increase would be regressive, as lower-income households tend to spend a greater share of their income on alcohol than higher-income households. Distributional concerns, however, are best addressed through the broader tax and benefit system. It is not clear that households with strong preferences for alcohol consumption deserve any special government compensation, even if they do have low income. Alcohol taxes should be set so that retail alcohol prices reflect not only production costs but also the external costs of alcohol abuse, and perhaps should also be considered for raising revenue.

Ian W.H. Parry is Allen V. Kneese Chair and Senior Fellow, Resources for the Future. He specializes in quantifying the costs and benefits of environmental, transportation, and energy policies.

Further Readings:

Donald S. Kenkel. 1996. New Estimates of the Optimal Tax on Alcohol. Economic Inquiry 34.

Parry, Ian W.H., Ramanan Laxminarayan, and Sarah E. West, 2009. Fiscal and Externality Rationales for Alcohol Taxes. B.E. Journal of Economic Analysis & Policy 9, Article 29: 1-45.

Philip J. Cook and Michael J. Moore. 2009. Alcohol. In Handbook of Health Economics, edited by Anthony J. Culyer and Joseph P. Newhouse. Elsevier.

West, Sarah E. and Ian W.H. Parry. 2009. Alcohol/Leisure Complementarity: Empirical Estimates and Implications for Tax Policy. National Tax Journal LXII: 611-634.