1. Lack of reliable price data. The development of a demand function consists of using statistical methods to relate observed price changes to changes in consumption. Price data that indicate what consumers really pay are needed to do this. The committee found that existing data on the prices of illegal drugs do not represent actual drug prices in cities.

  2. Price dispersion. The prices of illegal drugs vary greatly among transactions. The same dealer may charge different prices to different buyers, and prices may vary greatly over the distance of a few city blocks. In addition, the prices paid by different buyers may change in different ways over time. Moreover, a buyer may have some control over the price he pays. One way of controlling the price is by choosing the amount to purchase. There is substantial quantity discounting of cocaine and heroin, even within the range of quantities found in retail transactions. Thus, the price per unit purchased is likely to be lower if the quantity purchased is large than if it is small. A consumer may also be able to influence the price by his choice of seller. In particular, a consumer may choose to familiarize himself with several sellers and may choose not to buy from a high-priced seller if he knows that a lower-priced one is nearby. The committee knows of no model of demand for illegal drugs that takes account of potential price dispersion. All studies that have come to our attention use a price index as a proxy for sale prices. The price index typically is an estimate of the cost of one gram of pure cocaine or heroin in a city. The index is constant over all transactions in a given city and year (see Chapters 3 and 5 for a detailed discussion).

  3. Other costs. The sale price is only one of the costs that a consumer pays for an illegal drug. There are also search costs (the time spent looking for a seller), the costs associated with any legal penalty that may be incurred if the consumer is arrested, and, possibly, psychological costs associated with committing an illegal act. The committee found no source of data on search costs and no demand model that attempts to incorporate these costs.5 Data are available on legal penalties for possessing illegal drugs (see Chapter 6 for further discussion of this issue). The committee knows of three studies that have incorporated these into a model of de-


ADAM (Arrestee Drug Abused Monitoring) asks arrested drug users whether there was a time during the 30 days preceding the interview when they attempted to buy drugs and had the cash but were unsuccessful. ADAM also asks the reasons for lack of success. The answers to these questions provide crude indicators of the search costs experienced by ADAM respondents. It is not known whether the difficulties in buying drugs experienced by ADAM respondents are representative of difficulties experienced by consumers in general. No study that the committee knows of has used information from ADAM to construct an indicator of search costs for a demand model.

The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement