The effect of variation in high-level domestic drug enforcement on variation in drug prices
Introduction
This paper studies high-level domestic drug enforcement. Such enforcement is perhaps the least studied of the four principal supply-control programs. Souce-country control8, 10, 19, interdiction1, 14, 16, 17, and street-level enforcement4, 11, 13, 15 have received considerable attention, while relatively less is known about the effects of high-level domestic enforcement.
High-level domestic enforcement typically consists of undercover purchases and drug seizures directed at drug dealers, especially high-level dealers. These efforts seek to increase the cost of supplying drugs, to raise retail prices, and, therefore, to reduce consumption. Thus, in principle, high-level domestic enforcement can be assessed in terms of its effects on supply, price and/or consumption. Since detailed time series on supply or consumption of illicit drugs are hard to obtain, this paper focuses on whether enforcement raises drug prices, as observed through undercover purchases.
Enforcement can affect prices by increasing the cost of supplying drugs or by creating short-run scarcity. In the first case, enforcement acts like a tax. The greater the enforcement, the more expensive it is for dealers to bring drugs to market and, as a result, the higher are prices. Hence, changes in enforcement would be expected to induce changes in prices of similar direction after some delay. In the second case, seizures beyond what suppliers anticipated could lead to temporary shortages. Retail prices might then rise in the short term by enough to bring the quantity demanded in line with the quantity available during that period. Unfortunately, no data exist on suppliers’ expectations of seizures; however, changes in seizures relative to the previous period would again seem to be relevant.
The one previous empirical study of high-level domestic enforcement of which we are aware is DiNardo’s[5] investigation of the effect of cocaine seizures on cocaine price by analysis of variance. He found that variation in cocaine seizures did not contribute significantly to variation in cocaine prices. This finding is important, but it invites further study for several reasons. First, heroin attracts considerable enforcement, and the heroin and cocaine markets interact. Hence, heroin enforcement also needs to be assessed. Second, DiNardo used the total weight of cocaine seized to measure enforcement, which does not take into account the substantial difference in value per unit weight of drugs seized in different sizes2, 12. Third, analysis of variance does not test the direction of causality. One might expect enforcement to change prices, but it is also possible that changes in seizures may be caused by changes in availability, which may in turn affect prices.
This paper seeks to address these limitations by: (1) incorporating heroin prices and enforcement, (2) using four different measures of enforcement, and (3) using a vector-autoregressive methodology.
Section snippets
The data
The data employed here are from the Drug Enforcement Administration’s (DEA) system to retrieve information from drug evidence (STRIDE). STRIDE contains records of drug seizures, undercover purchases, and ‘other collections’ analyzed in DEA laboratories[7]. For seizure records there is information regarding the drug’s identity, weight, purity, location and date of seizure. Purchase records also include the amount paid. The subset of STRIDE available to us covers the period from January 1977 to
Granger-causality between enforcement and drug prices
Granger-causality is essentially a correlation between variables over time. Granger-causality from enforcement to prices would be a correlation between prices this month and enforcement in past months. If there is such a correlation, information about past enforcement can help predict prices, enforcement is then said to Granger-cause prices.
Granger-causality in no way implies causality in the traditional sense. It is possible for an unmeasured third variable to cause changes in enforcement and,
Does enforcement increase prices?
The previous section established that there is a relationship from enforcement to the prices of cocaine and heroin when enforcement is measured in terms of numbers of seizures. This section explores the nature of the relationship. With prices of retail cocaine, wholesale cocaine, and heroin measured at the 1, 80, and 6 gram levels respectively, estimating the VAR(1) model gives
Robustness of results
We are concerned with the robustness of the results with respect to the method of detrending, the selection of data, the treatment of drug purity, the neglect in variation of variances of the series, and changes in federal sentencing policy.
First differences were used to detrend the price and enforcement series. Other possible detrending methods include linear detrending and the Hodrick–Prescott (HP) filter[9]. Each method has potential pitfalls if the real structure of the innovation processes
How price effects of seizures can be negative
Conventional wisdom holds that increasing enforcement increases prices. This expectation is predicted on a straightforward model of drug distribution as a hierarchy of interrelated markets. Enforcement at one level raises the cost of doing business at that level.
Part of the increase in cost is passed along to lower levels, leading to higher prices. It is assumed that enforcement, especially higher level enforcement, has no effect on demand except through changes in price.
From this perspective,
Conclusions
Time series constructed with data from the Drug Enforcement Administration were analyzed for relationships between intensity of high-level domestic enforcement and the prices of cocaine and heroin. No evidence was found that increasing enforcement Granger-causes increases in drug prices. Hence, this analysis is pessimistic about the prospects for driving up drug prices by increasing high-level domestic enforcement.
Note that does not imply that such enforcement has no effect on prices. Cocaine
Acknowledgements
We wish to thank John DiNardo, Patrick Larkey, and Fallaw B. Sowell for comments on earlier versions of this paper. The research was supported in part by the RAND Corporation’s Drug Policy Research Center with funding from the Ford Foundation.
References (22)
- et al.
Mathl and Comp. Model.
(1993) Mathl and Comp. Model.
(1993)- et al.
Mathl and Comp. Model.
(1993) - et al.
J. Am. Stat. Assoc.
(1993) - Caulkins, J. P., Rydell, C. P., Schwabe, W. L. and Chiesa, J., Mandatory minimum drug sentences: throwing away the key...
- Chaikan, M. R., Street Level Drug Enforcement: Examining the Issues, National Institute of Justice, Washington D.C.,...
- Everingham, S. S. and Rydell, C. P. Modeling the demand for cocaine. RAND, MR-332-ONDCP/A/DPRC, Santa Monica, CA....
J. Analytical Toxicology
(1987)- Greenfield, V. A., Bolivian coca: a perennial leaf crop subject to supply reduction policies. Ph.D. Dissertation,...
- Hodrick, R. J. and Prescott, E. C., Postwar U.S. business cycle: an empirical investigation. Working Paper, No.451,...
Cited by (27)
Import substitution in illicit methamphetamine markets
2023, Journal of International EconomicsCitation Excerpt :We then use this policy experiment to estimate a simple structural model of methamphetamine production, which allows us to provide the first quantitative estimates of the magnitude of balloon effect.3 This paper also contributes to a larger literature examining the effectiveness of supply-side policies in curbing the production and distribution of illicit drugs more generally (DiNardo, 1993; Yuan and Caulkins, 1998; Miron, 2003; Kuziemko and Levitt, 2004; Becker et al., 2006; Dell, 2015; Mejia and Restrepo, 2016; Mejía et al., 2017; Castillo et al., 2020). Of these studies, Mejia and Restrepo (2016) is the most closely related to our work.
Retail margins in illegal drug markets
2019, Economic Analysis and PolicyCitation Excerpt :Kuziemko and Levitt (2004) find that the increase in imprisonment for drug related offenses leads to 5% to 15% increases in the price of cocaine and slightly fewer property crimes. DiNardo (1993) and Yuan and Caulkins (1998) focus on incarceration at the wholesale level and find no relationship between incarceration rates and drug prices. Pollack and Reuter (2014) provide a summary of studies investigating the relationship between incarceration rates and drug prices.
Substance abuse treatment centers and local crime
2018, Journal of Urban EconomicsAre supply-side drug control efforts effective? Evaluating OTC regulations targeting methamphetamine precursors
2014, Journal of Public EconomicsCitation Excerpt :Despite the fact that enforcement efforts are the primary approach to reduce drug use in the U.S., there is little consensus on their effectiveness. Some studies find no impact of enforcement on prices (DiNardo, 1993), others a negative relationship (Yuan and Caulkins, 1998), and still others a positive relationship (Miron, 2003). This is likely due to the difficulty in finding credible research designs.
Retail price as an outcome measure for the effectiveness of drug law enforcement
2010, International Journal of Drug PolicyPrice and purity analysis for illicit drug: Data and conceptual issues
2007, Drug and Alcohol DependenceCitation Excerpt :Those analyses would benefit from: (1) more precise estimates, particularly for heavier use of harder substances by more troubled populations, (2) cross- as well as own-price elasticity estimates, and (3) divisions of the overall elasticity into conditional elasticity of consumption per current user and the elasticity of participation, with the elasticity of participation further divided into the elasticity of initiation, elasticity of desistance, and, for multi-state models, the elasticity of escalation and de-escalation between states reflecting different intensities of use. The second application (loosely, regressing changes in price on changes in enforcement) has engendered a smaller but still important literature (e.g., DiNardo, 1993; Crane et al., 1997; Yuan and Caulkins, 1998; Kuziemko and Levitt, 2004), and there has been a flurry of papers on the effects of the Australian heroin drought (e.g., Weatherburn et al., 2003; Smithson et al., 2004). A principal limitation is that many price estimates are annual or at best quarterly, whereas markets adapt so quickly to all but the largest disruptions that higher-frequency price series would be more useful.