The Road Less Traveled: States That More Tightly Control the Sale and Distribution of Alcohol Have Lower Alcohol-Related Fatalities

Stephen Herzenberg
Publication Date: 
May 15, 2012

The Road Less Traveled: Download the ReportDownload the Full Policy Brief (PDF)

Read the Press Release

Get the Raw Data: Data File | Do File 

Listen to a Media Conference Call on the Policy Brief


What impact would the privatization of Pennsylvania’s retail and wholesale alcohol operations have on the number of alcohol-related traffic fatalities each year? This should be a critical question for state policymakers to answer as they consider proposals to privatize Pennsylvania’s state-operated wine and spirits system. 

Following a comprehensive review of state-level data, we find that states with tighter control over the sale and distribution of alcohol have lower rates of alcohol-related traffic deaths than states that take a more hands’ off approach.

Our findings are consistent with those of a panel of 21 public health experts[1] who concluded that the privatization of retail alcohol sales and distribution leads to increases in per capita alcohol consumption, creating a greater risk of alcohol abuse and its associated social costs.[2] Those findings were published by The Task Force on Community Preventive Services (Task Force), a panel appointed by the director of the Centers for Disease Control and Prevention (CDC), in the April edition of the peer-reviewed American Journal of Preventive Medicine (Hahn et al, 2012).[3]

Economists John Pulito and Antony Davies, in research released by two self-described “free market” think tanks, reached a very different conclusion. Their research found that states with tighter control of the sale and distribution of alcohol, in some instances, have higher rates of alcohol-related fatalities.[4]

This policy brief replicates Pulito and Davies’ analysis and then demonstrates that their results are reversed once you account for two variables excluded from the Pulito-Davies analysis but key to understanding the differences in alcohol-related motor vehicle fatality rates among the states — average vehicle miles traveled and average per capita income. Once these variables are included, states that more heavily control the sale and distribution of alcohol do have lower alcohol-related fatality rates for adults than either states that do not regulate or only lightly regulate alcohol sales.  

All else equal, a heavy control state with the characteristics of Pennsylvania sees 58 fewer adult deaths each year from alcohol-related traffic accidents than a comparable state that has no such controls. (Heavy control states are defined as maintaining control over the sale of at least two types of alcohol at the retail level and at least one type of alcohol at the wholesale level.)

We find no difference in fatality rates for youth ages 15 to 19 according to the degree of state control over the distribution and sale of alcohol.  Among youth under the age of 15, a group Pulito and Davies do not analyze, we find lower fatality rates for alcohol-related car accidents in states that exercise heavy control over the distribution and sale of alcohol. 

Although the full technical explanation is complex, there are straightforward reasons for including vehicle miles traveled and per capita income in this analysis. First, both variables have an impact on alcohol-related fatalities and, therefore, belong in a comprehensive analysis of variations in fatality rates across states. Second, control states tend to be ones in which people drive further and are lower income. Therefore, when the two variables are excluded, some of the higher fatalities that should be attributed to driving further and having lower incomes are wrongly attributed to state alcohol controls. 

This policy brief is not intended for the benefit of researchers on alcohol distribution and consumption, such as the public health experts on the Task Force. They do not need this policy brief to know that they can confidently ignore the Pulito and Davies research. Rather, this brief is intended to set the record straight in the public policy debate on liquor privatization in Pennsylvania. The Pulito-Davies research has been used repeatedly by policymakers and others—including in testimony before the legislature—to muddy the water regarding the negative social impacts of privatizing alcohol distribution. 

As our findings and those of the Task Force show, the water remains clear on this topic: privatization brings with it potential negative social impacts, including increased alcohol-related traffic fatalities.  These negative social impacts should be taken into account by lawmakers as they consider proposals to privatize Pennsylvania’s wine and spirits stores.

Read the Full Policy Brief (PDF)


[1] The names and affiliations of the Task Force are available online at


[3] Available online  Also published in April in the same journal was a commentary by Cook (2012) and available online  Cook’s data analysis is broken into two parts with the first part focusing on per capita wine sales in seven states that privatized the sale of wine between 1970 and 1985 and the second part on what is known about the effects of privatizing liquor sales in the U.S. which is the case of Iowa in 1987.  With respect to the privatization of wine, Cook concludes there is evidence that the privatization leads to a “notable increase in consumption.”  With respect to Iowa, Cook found again evidence of rising wine consumption but not of liquor. 

[4] John Pulito and Antony Davies, Does State Monopolization of Alcohol Markets Save Lives? Dr. Davies provided us with an updated version of this paper on July 21, 2011. This version is available at As of January 2012, a previous version (State Control of Alcohol Sales As A Means of Reducing Traffic Fatalities: A Panel Analysis) of the paper could be found on line at The findings were also summarized in a Mercatus Center Working paper: Antony Davies and John Pulito, Binge Thinking: A Look at the Social Impact of State Liquor Controls, Mercatus Center Working Paper No. 10-70, November 2010 available online at