Jeffrey A. Miron, Boston University
The prohibition of alcohol, 1920-1933, is one of the most interesting policy experiments in U.S. history. Temperance movements waxed and waned in the U.S. from early in the nineteenth century, and these movements produced numerous state prohibitions. Many of these prohibitions were subsequently repealed, however, and those that persisted were widely regarded as ineffective. Amid the atmosphere created by World War I, support for national prohibition reached critical mass, and the country ratified the 18th Amendment to the Constitution in January, 1919.1 Under this amendment and the Volstead Act, which provided for the enforcement of Prohibition, the manufacture, transportation, and sale of alcohol were prohibited by federal law.2 The Amendment was popular for many years, but beginning in the late 1920s support began to erode.3 In 1933 the 21st Amendment repealed the 18th, ending Prohibition.
This article provides a brief economic history of Alcohol Prohibition. The first section discusses the major effects one should expect from policies like Prohibition and mentions evidence consistent with these effects. The second and third sections then consider more detailed evidence on two keys issues: Prohibition’s effect on the quantity and price of alcohol, and Prohibition’s effect on violent crime.
The Effects of Prohibitions
The most direct effects of prohibitions are on the supply and demand for the prohibited commodity.4 Prohibitions raise supply costs because black market suppliers face legal punishments for manufacturing, distributing, and selling. Conditional on operating in secret, however, black market suppliers face low marginal costs of evading government regulations and taxes (Miron 2001), which provides a partial offset to the increased costs due to prohibition.5 Prohibitions reduce demand by creating legal penalties for possession and by increasing uncertainty about product quality.6 Prohibitions also reduce demand if consumers exhibit “respect for the law.” At the same time, prohibitions can increase demand through a “forbidden fruit” effect, meaning a tendency for consumers to desire that which has been forbidden to them. Thus, the effect of prohibitions on price and even quantity are ambiguous a priori and must be determined empirically.
In addition to affecting price and quantity, prohibitions potentially increase violent and non-violent crime. Participants in an illegal trade cannot use the legal and judicial system to resolve disputes, so they seek other methods such as violence. Enforcement of prohibitions means reduced resources for enforcement of non-prohibition laws, which implies reduced deterrence of crime generally. Prohibitions can increase income-generating crime, such as theft or prostitution, by raising prices if consumers finance consumption of the prohibited commodity from such crime. And prohibitions give black market suppliers an incentive to corrupt law enforcement officials and politicians. Despite these tendencies to increase crime, the net effect of prohibitions on crime can be negative if prohibitions discourage consumption of the prohibited good and if such consumption encourages criminal activity. Thus, the net effect of prohibitions on crime can only be determined empirically.
Two other effects of prohibitions are the effects on overdoses and accidental poisonings. Because suppliers in a prohibited market must hide their activities from the authorities, they have a strong incentive to produce and ship the good in the most concentrated and hence most easily concealed form (Thornton 1998). This implies that prohibitions help make the potent forms of a good more readily available or even help create more potent forms of a prohibited substance. By itself this effect does not necessarily change the manner in which consumption takes place; consumers can potentially redilute the commodity in question to achieve their desired degree of potency. But in practice such redilution is imperfect, suggesting increased overdoses under prohibitions.7
Consumers in a prohibited market cannot sue the manufacturers of faulty goods or complain to government agencies without incriminating themselves. In addition, the costs of advertising are high in a prohibited market, so producers cannot easily develop a reputation for quality in order to generate repeat business. Thus, uncertainty about quality is likely to be greater in a prohibited market. Combined with the greater existence of high potency products, this further suggests the possibility of increased overdoses, as well as accidental poisonings, under prohibition.8
Alcohol Consumption and Prices under Prohibition
The evidence on alcohol consumption during Prohibition is incomplete, since standard data sources are not available for the Prohibition period. Thus, most analyses of Prohibition’s effect use the cirrhosis death rate as a proxy. Figures 1 and 2 present data on alcohol consumption and cirrhosis, respectively.9 The figures suggests a strong correlation between alcohol consumption and cirrhosis. Both series decline noticeably just before the onset of Prohibition and increase gradually for the first three decades after repeal of Prohibition. Both series then increase more rapidly from the mid-1960’s to the mid-1970’s and decline from 1980 to the present. The correlation is not perfect; alcohol consumption exhibits a noticeable spike relative to cirrhosis in the 1940’s, and cirrhosis starts declining several years earlier than alcohol consumption during the 1970’s. But the figure suggests that cirrhosis is a reasonable proxy for alcohol consumption, and the evidence summarized in Dills and Miron (2001) confirms this impression.
The fact that cirrhosis was substantially lower on average during Prohibition than before or after might suggest that Prohibition played a substantial role in reducing cirrhosis, but further examination suggests this conclusion is premature. First, there have been substantial fluctuations in cirrhosis outside the Prohibition period, indicating that other factors are important determinants and must be accounted for in analyzing whether Prohibition caused the low level of cirrhosis during Prohibition. Second, there is no obvious jump in cirrhosis upon repeal. This fact does not prove that Prohibition had no effect, since the lags between consumption and cirrhosis mean the effect of increased consumption might not have shown up immediately. Nevertheless, the behavior of cirrhosis after repeal fails to suggest a large effect of Prohibition. Third, cirrhosis began declining from its pre-1920 peak by as early as 1908, and it had already attained its lowest level over the sample in 1920, the year in which constitutional prohibition took effect.
This last fact is the most problematic for the claim that Prohibition reduced alcohol consumption. One possible explanation for the large pre-1920 decline in cirrhosis is that state prohibition laws were becoming increasingly widespread during the 1910-1920 period. Dills and Miron (2001) use state-level data, however, to show the declines in cirrhosis during this period were typically as large or larger in wet states as in states that adopted prohibition laws. More formally, they estimate a fixed-effects regression using state-level cirrhosis data to show that, once aggregate effects are accounted for, there is little effect of state prohibitions on cirrhosis.
A different possible explanation for the large decline in cirrhosis is pre-1920 federal anti-alcohol policies. In February 1913, Congress adopted the Webb-Kenyon Law, which prohibited shipments of liquor from wet states into dry states if such shipments were in violation of the dry state law. This did not prohibit all shipments into dry states, since some dry states allowed importation (Merz 1930, p. 14). In February 1917, Congress passed the Reed bone-dry amendment, which forbade interstate shipment of liquor into states that prohibited manufacture and sale, even if the state allowed importation. (Merz 1930, p. 20). In August 1917, Congress adopted the Food Control Law, which forbade the manufacture of distilled spirits from any form of foodstuff and closed the distilleries (Merz 1930, pp. 26-27, 40-41). In September 1918, it closed the breweries as well (Merz 1930, p. 41). Also in September 1918, Congress approved wartime prohibition, although this did not take effect until July 1, 1919 (Merz 1930, p. 41). Wartime prohibition contained the first general restriction on sale, providing that after June 30, 1919, no liquor could be sold for beverage purposes except for export (Schmeckebier 1929, pp. 4-5).
There are a number of reasons to doubt that these policies were major factors in causing the pre-1920 declines in cirrhosis. First, cirrhosis had been declining since 1908, well before any of these policies took effect. Second, all these policies except war-time prohibition (which did not take effect until July, 1919) were weak; they did not restrict production until August 1917, and none outlawed importation or consumption of existing stocks. Moreover, Congress made no appropriation for the enforcement of any of these laws. In addition, there are other factors that potentially explain a decline in alcohol consumption or cirrhosis. Patriotism might have encouraged temperance, since food was considered vital to the War effort and beer production was associated with Germany. And the high morality rate in Word War I combined with the flu epidemic of 1918 might have removed many persons from the population at risk who would otherwise have died from cirrhosis.
Beyond the results presented here, additional results in Dills and Miron (2001)—which account for the effects of state prohibitions, pre-1920 federal anti-alcohol policies, alcoholic beverage taxes, income and other factors—demonstrate consistently that Prohibition had a small, statistically insignificant, and possibly even a positive effect on cirrhosis. Given the evidence that cirrhosis is a reasonable proxy for alcohol consumption, this implies Prohibition had little impact on the path of alcohol consumption.
The question raised by this result is why consumption did not fall more significantly, since conventional accounts suggest that alcohol prices rose by several hundred percent on average (Warburton (1932), Fisher (1928)). One possibility is that the conventional view is overstated.
The first problem with the calculations presented by Warburton or Fisher is that they neglect the behavior of the overall price level. Warburton’s data compare prices between 1911-1915 and 1926-1930, while Fisher’s compare prices between 1916 and 1928. Both authors examine the behavior of nominal prices, yet the price level increased by approximately 75 percent between these two periods (Bureau of the Census (1975), p.211). Thus, at a minimum, the raw data presented by Warburton and Fisher overstate the increase in the relative price of alcohol.
In addition, Warburton presents a broad range of prices for the Prohibition period, and the lowest prices reported suggest that, even ignoring inflation, some alcoholic beverage prices fell relative to the pre-Prohibition period. This does not prove that consumers paid less, on average, for alcohol, but they certainly faced an incentive to buy at the lowest prices and then stockpile the quantities purchased at these prices. The available data do not allow computation of the average price actually paid, and the extremely high prices reported in many cases by both Warburton and Fisher allow for the possibility that the average price paid in fact rose. But the magnitude of this rise is undoubtedly less than they asserted, and it is at least possible prices failed to rise substantially overall. If prices did not increase very much, there is no puzzle in the failure of consumption to fall substantially.
Alcohol Prohibition and Crime
The evidence on Prohibition and crime focuses on the homicide rate, since this is the only type of crime for which data are reported consistently both before, during, and after Prohibition.10 Figure 3 presents the homicide rate in the United States (measured as homicides per 100,000 population) for the period 1900-1995. Starting from around 1906, the homicide rate rises steadily through 1933-1934, when it begins a general decline until approximately 1960, interrupted by a spike during World War II. Beginning in the early 1960’s the homicide rate rises steadily until the early 1970’s—to a level slightly above the previous peak in 1933-1934—and then fluctuates around a relatively high value for the remainder of the sample.
Roughly speaking, therefore, there have been two periods with high homicide rates in U.S. history, the 1920-1934 period and the 1970-1990 period (Friedman 1991). Both before the first episode and between these two episodes, homicide rates were relatively low or clearly declining. Prima facie, this pattern is consistent with the hypothesis that alcohol prohibition increased violent crime: homicide rates are high in the 1920-1933 period, when constitutional prohibition of alcohol was in effect; the homicide rate drops quickly after 1933, when Prohibition was repealed; and the homicide rate remains low for a substantial period thereafter. Further, the homicide rate is low during the 1950s and early 1960s, when drug prohibition was in existence but not vigorously enforced, but high in the 1970-1990 period, when drug prohibition was enforced to a relatively stringent degree (Miron 1999).
To see this more carefully, consider Figure 4, which plots real per capita expenditure by the federal government for enforcement of alcohol and drug prohibition over this same period. As discussed in Miron (1999, 2001), the effect of prohibition on violence depends not just on the existence of a prohibition but on the degree to which it is enforced. Increased enforcement narrows the scope of legal exceptions to the prohibition (e.g., medical uses), thereby increasing the size of the black market, and increased enforcement destroys reputations and implicit property rights within the black market. Both effects increase the use of violence.
Real per Capita Expenditures for Enforcement of Alcohol and Drug Prohibition
Note: Vertical axis is measured in 1992 dollars.
The data in Figure 4 combined with the data in Figure 3 show that expenditure climbs along with the homicide rate during Alcohol Prohibition and then falls at the end of this prohibition, as does the homicide rate. The relation is not perfect; other factors undoubtedly play a role. For example, the homicide rate begins rising about a decade before constitutional Prohibition takes effect, a fact that potentially reflects demographics (the enormous levels of immigration during the early part of this century), the violence-inducing effects of World War I, or perhaps merely changes in the sample of states used to compute homicide rates (Miron 1999). Regression analysis in Miron (1999) considers this more formally and confirms that enforcement of alcohol prohibition played a central role in causing the increasing and decreasing homicide rates during this period.
Prohibition represents one of the most dramatic policy experiments in U.S. history, with implications for a broad range of economic, historical, and political issues. This summary has focused narrowly on the most basic economic effects of Prohibition. The analysis shows that the evidence is consistent with the predictions of standard economic theory about the effects of prohibitions.
Clark, Norman H. Deliver Us From Evil: An Interpretation of American Prohibition. New York: W.W. Norton and Company, 1976.
Dills, Angela and Jeffrey A. Miron. “Alcohol Prohibition, Alcohol Consumption, and Cirrhosis.” Manuscript, Boston University, 2001.
Fisher, Irving. Prohibition Still at Its Worst. New York: Alcohol Information Committee, 1928.
Friedman, Milton. “The War We Are Losing.” In Searching for Alternatives: Drug-Control Policy in the United States, edited by M. B. Krauss and E. P. Lazear, 53-67. Stanford, CA: Hoover Institution, 1991.
Levine, Harry G. and Craig Reinarman. “From Prohibition to Regulation: Lessons from Alcohol Policy for Drug Policy.” The Milbank Quarterly 69 (1991): 1-43.
Merz, Charles. The Dry Decade. Garden City, NY: Doubleday, Doran and Co., 1930.
Miron, Jeffrey A. “Some Estimates of Annual Alcohol Consumption Per Capita, 1870-1991,” ISP Discussion Paper #69, Department of Economics, Boston University, 1996.
Miron, Jeffrey A. “The Effect of Alcohol Prohibition on Alcohol Consumption.” NBER Working Paper No. 7130, 1997.
Miron, Jeffrey A. “Violence and the U.S. Prohibitions of Drugs and Alcohol,” American Law and Economics Review 1-2 (1999): 78-114.
Miron, Jeffrey A. “Violence, Guns, and Drugs: A Cross-Country Analysis.” Manuscript, Boston University, 2001.
Miron, Jeffrey A. and Jeffrey Zwiebel. “Alcohol Consumption During Prohibition.” American Economic Review 81, no. 2 (1991): 242-247.
Miron, Jeffrey A. and Jeffrey Zwiebel. “The Economic Case against Drug Prohibition.” Journal of Economic Perspectives, 9, no. 4 (1995): 175-192.
Sinclair, Andrew. Prohibition: The Era of Excess. London: Faber and Faber, 1962.
Schmeckebier, Laurence F. The Bureau of Prohibition: Its History, Activities, and Organization. Brookings Institution: Washington, 1929.
Thornton, Mark. The Economics of Prohibition. Salt Lake City: University of Utah Press, 1991.
Thornton, Mark. “The Potency of Illegal Drugs.” Journal of Drug Issues 28, no. 3 (1998): 725-740.
Warburton, Clark. The Economic Results of Prohibition. New York: Columbia University Press, 1932.
1 Historical accounts cite a range of factors as finally tipping sentiment in favor of national prohibition. One was the huge number of immigrants during the first decade and a half of the 20th century, since popular wisdom held that immigrants were heavy drinkers. A second factor was increasing urbanization, which made the presence of the hard drinking, saloon frequenting, urban poor more visible (Clark, 1976). U.S. involvement in World War I may also have played a significant role, by legitimizing the view that turning grain into alcohol was wasteful (Merz 1930), by creating an air of moral certainty that facilitated passage of prohibition (Sinclair 1962), and by producing a distaste for anything German (i.e., beer).
3 The two key factors usually credited with precipitating Prohibition’s demise (Levine and Reinarman 1991) are the Great Depression, which invalidated dry claims that Prohibition promoted prosperity and produced a need for tax revenue, and the increasing violence associated with Prohibition.
9The data on alcohol consumption are estimates of the per capita consumption of pure alcohol, measured in gallons, computed as a weighted sum of separate estimates for beer, spirits, and wine, assuming a particular pure alcohol content for each component. The cirrhosis death rate is measured as the number of deaths per 100,000. Miron (1996, 1997) and Dills and Miron (2001) provides details of the construction of these series.
Citation: Miron, Jeffrey. “Alcohol Prohibition”. EH.Net Encyclopedia, edited by Robert Whaples. September 24, 2001. URL http://eh.net/encyclopedia/alcohol-prohibition/