Table 3. Shipping the Good Apples to New York
Source: Alchian and Allen 1972, 71.
An ad valorem tax imposed on a commodity with n characteristics has different results from both the constant-quality model and the constant per-unit tax. The tax would define the commodity as a minimum amount of characteristics 1, . . ., e. The ad valorem tax, however, would tax all the characteristics included in the product. Therefore, the inclusion of a characteristic and its level will depend crucially on whether it is cheaper to include the characteristic or to sell it separately to avoid the tax. The unconstrained attributes (those not defined in the statute) will be reduced, eliminated, or sold separately in order to avoid taxation. The product will sell for less than predicted by the constant-quality model and will result in a lowerquality product (of fewer characteristics). The ad valorem tax therefore reduces taxation by eliminating product characteristics in a way similar to that in which transportation charges eliminate characteristics. For example, fresh oranges and fresh orange juice can be considered higher-quality products than frozen or reconstituted orange juice. Suppliers of orange products, however, can greatly reduce the transportation “tax” on orange juice by shipping frozen concentrate, which is much less bulky than an equal amount of the fresh product. The frozen concentrate is then reconstituted with the addition of water and labor at the point of consumption.
Barzel found support for his hypothesis in the response of cigarette prices to changes in cigarette taxes. Excise taxes tended to increase the tar and nicotine level of cigarettes. Johnson (1978) provides additional evidence that ad valorem taxes result in lower prices and unit taxes result in higher prices than predicted by the traditional model of taxation. Harris (1980) also recognizes that increasing the per-unit taxes on cigarettes leads to a substitution of high-potency cigarettes for low-potency cigarettes. Sumner and Ward (1981) question the applicability of the evidence on cigarette prices by suggesting alternative explanations for diverging prices.
Feenstra (1988) found that the effect of a 25 percent ad valorem tariff on imported pickup trucks was ambiguous with respect to overall quality, while the import quota on Japanese automobiles led, as expected, to quality upgrading. These decisions are based on the cost of including a characteristic versus selling the characteristic separately. The ambiguous result from the ad valorem tax, however, is not unexpected in this particular case. Barzel (1976, 1183n.) notes that the price of the parts of an automobile was two and a half times the price of the same automobile assembled. In this example, the savings involved in purchasing an assembled car dwarf the effects of the 25 percent ad valorem tax. Japanese producers have also responded to the tax incentives by adding extra features to their “truck” products in order to get them classified as vans, which are subject to only a 2 percent tariff.
THE IMPACT OF PROHIBITION ON POTENCY
The type of taxation and the definition of the taxed commodity by the tax statute result in different effects on the quality and attributes contained in a product. The prohibition statutes and their enforcement play a similar role in determining product composition, quality, output, and price. The directives and incentives of law-enforcement officials will influence market outcomes such as product composition.
Prohibition establishes a gambling environment rather than an explicit tax. Participants who are actually caught face huge losses from lost revenue, fines, confiscations, and jail terms. Those not caught reap large monetary profits. All market participants, however, incur large costs of risk bearing. The tax is evaluated as a function of the penalties and the likelihood of capture and conviction.
Prohibition statutes generally consist of three parts. First, to be illegal, products must contain a minimum amount of a certain drug. During alcohol prohibition, products that contained more than 0.5 percent alcohol were illegal. A product containing any detectable amount of heroin is illegal. Second, penalties are generally levied on the basis of weight. For example, maximum penalties for marijuana possession in Indiana are a one-year prison sentence and a $5,000 fine for amounts up to thirty grams. The limits on penalties are doubled for amounts over thirty grams. Finally, penalties are established for production, distribution, and possession.
The prohibition statutes consistently define the product in terms of minimum potency (without constraining the maximum). Also, the heavier the shipment, the more severe the penalty. Since penalties are based on the weight of a shipment, suppliers will reduce the attributes that are not taxed when separated from the product.3 Potency is unconstrained and will likely increase as suppliers raise the value of the shipment to reduce the relative burden of the tax. This aspect of prohibition statutes therefore acts as a constant per-unit tax.
In addition to the prohibition statutes, the probability of capture plays an important role in risk. The efficiency of law enforcement relative to the size of the black market for drugs establishes (in part) the probability of capture. Suppliers of illegal drugs evaluate law enforcement and penalties to determine the risk they face and the allocation of their resources to avoid capture.
A key to avoiding capture is concealment of the shipment. While this can take many forms, the size of the shipment is a basic factor. Size is related to weight and will act as a constant per-unit tax, inducing a higher-than-market potency. Concealment efforts would increase at the margin as law-enforcement resources or efficiency increased. These factors therefore act to increase quality and result in a higher price than that predicted by the constant-quality model. Potency increases, and this increase is largely responsible for the increase in the overall “quality.”
Prohibition may act like an ad valorem tax as a result of the directives of legislatures and the incentives of law-enforcement officials and judges. Lindsey (1976) showed that bureaucrats have the incentive to produce goods and services with attributes that are easily monitored and desired by Congress, while shirking the production of attributes that are not or cannot be monitored by Congress. For example, Lindsey found that whereas Veteran Administration hospitals produced measurable output (patient-days) at lower cost, proprietary hospitals provided more staff per patient, “better” physicians, shorter stays, less crowding and waiting, and more environmental amenities. Thus, in the same sense that taxation alters the attributes of products, the incentives of bureaucrats can alter the attributes of products.
With prohibition, the type of service provided by law-enforcement bureaucrats to Congress and state legislatures can have important effects on the mix of attributes in illegal drugs. One example of bureaucratic behavior is the technique of estimating the dollar value of drug confiscations. The value of drug confiscations is estimated using average “street price.” The street price is the highest per-unit price because it represents the last step in product distribution. This estimation is akin to a farmer’s ascertaining how much flour is required to make a loaf of bread and then multiplying the retail price of bread (per pound) by the total weight of his freshly cut wheat, including chaff, to determine the value of his crop.
Likewise, law-enforcement bureaucrats might view the capture of large or high-potency shipments as promoting their self-interest and satisfying the directives of the legislature. The capture of large shipments provides the bureaucracy with publicity about the effectiveness of their work and may help stimulate demand for their product (enforcing the drug laws). The concentration of law-enforcement resources on the interdiction of high-potency drugs would reduce the risk of shipping low-potency and less dangerous drugs such as marijuana and increase the risk of shipping high-potency and dangerous drugs such as heroin. Because the risk would increase so would their market value. This incentive would have an effect similar to that of either an ad valorem or ad potere (according to potency) tax.
Similar incentives may exist in the court system. High-potency shipments and more dangerous drugs may influence the probability of conviction. The court system has some discretion in determining penalties. According to current federal senten
cing guidelines, higher-potency shipments may bring longer prison sentences and larger fines. The discretion of the court could therefore act as a constraining factor on potency and as an ad valorem tax.
The incentives of law enforcement and the court system, like ad valorem taxes, result in lower than expected quality and price, but do they constrain or reduce potency? Are these incentives—or ad valorem taxes—empirically relevant, as in Feenstra’s (1988) examination of the import tax on Japanese pickup trucks?
Several points must be made about the existence or strength of the ad valorem effect on potency. First, it is likely that law-enforcement officials benefit more from capturing larger shipments that increase estimated street value than from higher-potency shipments that do not. Second, there is no reason to believe that law-enforcement officials have the means to discriminate between a high-potency and a low-potency shipment of a particular drug, except with respect to where in the chain of distribution it is confiscated. Third, while judges do have some sentencing discretion according to drug potency, this discretion is limited to interpreting the intent of the defendant within the penalty structure based on weight. Last, the pure ad valorem tax is not expected to have a marked effect on potency but rather a proportional effect on all product attributes.
Law enforcement may create an ad valorem tax in a multidrug illegal market. Concentration on more dangerous drugs such as heroin would have effects similar to those of an ad valorem tax on the entire illegal drug market, since the probability of capture from a unit of heroin would be greater than from a similar unit of marijuana. This aspect of prohibition enforcement does not constrain potency as a product attribute—it merely focuses greater resources and penalties on the more dangerous drug types. The expected results are similar to the situation of higher penalties for heroin relative to penalties for marijuana. We would expect that the higher penalties for heroin would result in higher potency relative to the increases in potency of marijuana, which is subject to lower penalties.
Many factors, such as the definition of illegal drugs, penalties based on weight, and probability of capture, do not constrain potency and therefore result in higher-potency drugs. Only the limited discretion of the courts was found to place a small constraint on potency. In a multiple illegal-drug market, differences in penalties and the incentives of law-enforcement bureaucrats intensify the effects of prohibition on heroin and reduce the effect on potency for drugs such as marijuana.
The evaluation of risk therefore places a strong incentive in increasing the potency of illegal drugs. Empirically, the cost of bearing the risk of prohibition is high relative to the cost of production and distribution in a legal environment. Edward Erickson (1969) estimated the prohibition tax on heroin at 20,000 percent. The cost of one ounce of marijuana is well over one hundred times the market price of an equal weight of cigarettes. Such high rates of taxation obviously have a major impact on the attribute mix of products.
POTENCY IN PROHIBITED MARKETS
The lack of reliable data concerning prohibited markets makes rigorous econometric testing impossible. Nonetheless, history provides several instructive illustrations concerning the potency of products in such markets. Two prominent episodes, Prohibition and the modern “war on drugs,” are presented here.
These episodes provide enough information to present the correlations among prohibition, relative prices, and potency, as well as the correlations among changes in law-enforcement resources, relative prices, and potency. This information illustrates the product alteration and innovation that occurs during prohibition. In terms of regression analysis, even these limited goals are difficult to establish because adequate data are simply not available. In these cases, using proxies for variables is like substituting Ping-Pong balls for turtle’s eggs in a recipe.
Table 4. Federal Expenditures upon the Enforcement of Prohibition (thousands of dollars)
Source: Warburton 1932, 246.
ALCOHOL PROHIBITION
An examination of the prohibition on alcohol during the 1920s provides useful and interesting evidence on “shipping the hard liquor in.” During Prohibition a variety of law-enforcement resources were mobilized through the Volstead Act in an attempt to curtail the production, sale, and consumption of alcohol. Enforcement created risks for alcohol suppliers, risks that had pervasive effects on how, when, where, and what kind of alcohol was consumed.
Table 4 provides information concerning the federal prohibition enforcement effort.4 Total expenditures grew from less than four million dollars in the second half of 1920 to almost forty-five million in 1930. The annual budget of the Bureau of Prohibition doubled during this decade, with the greatest growth occurring between 1920 and 1925. In 1925 the Coast Guard’s budget was augmented to enforce prohibition, doubling the resources devoted to interdiction and enforcement. The indirect cost, which included such expenditures as the costs of criminal prosecutions, also grew throughout the decade.
Fines, penalties, and net expenditures are not relevant for the purpose of determining law-enforcement resources. Fines and penalties do, however, provide some evidence of the effectiveness of law-enforcement resources. This effectiveness appears to have been weakened by the development of specialists in illegal production, the development of rigidities within the bureaucracy, and the corruption of public and law-enforcement officials.
Clark Warburton and Irving Fisher were opponents in the academic debate over Prohibition, but they both presented evidence concerning the dramatic change in relative prices that occurred during Prohibition. We would expect the change in relative prices to result from the risk imposed by law enforcement.
Warburton demonstrated that the price of spirits fell relative to the price of beer. Based on the average of four separate estimates of probable prices had National Prohibition not been enacted, the price ratio of spirits to beer would have been 15.42 to 1. The actual estimated ratio of retail prices in 1929–30 was 11.78 to 1, while the estimated cost ratio of homemade alcohol to homemade beer was 3.33 to 1 (Warburton 1932, 148–66). Estimates of full cost would lower these price ratios under Prohibition. Buyers faced the risk of confiscation, but this risk was lower for spirits because of its compact size.
As noted earlier, Irving Fisher was a major proponent of Prohibition. He used observations of increased prices to claim that Prohibition was drying up the supply of alcohol. Fisher used the data in table 5 to support his case. Calculations in parentheses are mine. Fisher’s calculations to the right are of unknown origin. He did in fact show in his “alcohol price index” that alcohol increased in price. He also showed that lager beer increased in price by 700 percent while rye whiskey increased by only 312 percent, again supporting the case that the relative price of high-potency alcohol fell.
It is generally agreed that as a result of the increase in the price of alcohol, the absolute quantity of alcohol purchased declined, a fact confirmed by Warburton. Consumption of high-potency-alcohol products, however, rose relative to low-potency-alcohol products, such as beer. The effect of lowering the relative price of spirits during Prohibition on expenditures and consumption is shown in table 6. Without explicitly making the connection to the change in relative prices, Warburton noted that “Prohibition has raised the amount spent for spirits to three and a half billion dollars, and reduced that for beer to less than a billion dollars” (1932, 170). Fisher was also aware of the “well known fact that Prohibition has been more effective in suppressing the drinking of beer than of whiskey” (1927, 29). T. Y. Hu, who lacked an understanding of relative prices, doubted this finding.
Table 5. Fisher’s “Alcohol Price Index,” 1916–1928
Source: Fisher 1928, 91.
Prohibition’s impact on consumption is further illustrated by placing it in historical perspective (table 7). Although alcohol consumption is related to various factors, such as income and alcohol taxes, certain trends are suggested. First, the consumption of beer increased during Prohibition, partly as a result of the decl
ine in its relative cost of production and distribution. Second, total consumption of pure alcohol was more or less stable throughout the period.
Table 6. The Effect of Prohibition on Alcohol Expenditures (millions of dollars)
Source: Warburton 1932, 170.
Third, expenditures declined before Prohibition because of increased taxation, wartime prohibitions, and state prohibitions.
Expenditures on alcohol as a percentage of national income declined by 2 percent from 1890 to 1910, more being spent on beer than spirits (55:45), a ratio that continued from 1911 to 1916 (Warburton 1932, 114–15). This pre-Prohibition consumption pattern was reestablished after the repeal of Prohibition, distilled spirits again accounting for only about half of all alcohol expenditures. During Prohibition (1922–30) expenditures for distilled spirits as a percentage of all alcohol expenditures grew to 70–87 percent. For the period 1939–60 distilled spirits accounted for 42–53 percent of total alcohol sales.
The Economics of Prohibition Page 12