Appendix B
Estimating Typical Expenditures on Drug Consumption

This appendix discusses the methodology used to develop estimates of weekly expenditures on cocaine and heroin by arrestees who used either or both of these drugs on more than 10 days during the month before their arrests. The estimates reported here are based on self-reports by arrestees in 24 cities. These self-reports, which are for 1989 and later, are from the Drug Use Forecasting (DUF) program.¹

The Data

DUF respondents reported how much they spent on all drugs combined (during a typical week) but not how much they spent on each individual drug. They also reported the number of days they used any of 22 kinds of drugs during the month before their interview. We used regression analysis to infer expenditure patterns for cocaine and heroin based on these data.

The greatest obstacle to accurate reporting is a respondent's denial of drug use.² Therefore, drug use is underreported. Once a respondent admits drug use, however, he or she would seem to have less incentive to underreport or overreport consumption. To be included in this analysis, the respondent had to have admitted some illicit drug use during the last 30 days and had to have admitted some drug expenditure during the typical week. (These different time periods were required because of the wording of the DUF questions.) We estimated expenditure patterns for each year separately.

The dependent variable (EXPEND) was the weekly expenditure on all drugs. This variable was skewed (a few individuals reported very high amounts). Dealing with this skewed distribution required two steps. The first step was to trim the data by excluding all cases where a respondent claimed to spend more than $500 per day on drugs. Reasoning was that large reported amounts resulted from hyperbole, or else respondents were buying for others or for resale. Even $500 per day is a large level of expenditure. This trim to the data eliminated 2–3 percent of observations. The second step was to convert weekly expenditures to a logarithm before estimating the regression discussed subsequently. We then converted the predictions back to the original dollar scale.

The number of days that a respondent consumed each of four categories of drugs were the independent variables. We collapsed drugs into four general categories: COCAINE (powdered and crack), HEROIN (black tar and other), MARIJ (marijuana and hashish—combined in the DUF interview), and OTHER. Cocaine, heroin, and marijuana were the only drugs consumed by a large percentage of the arrestee population. OTHER comprised a large number of infrequently consumed substances. Except for MARIJ, each variable comprised at least two drugs.

The category variable represents the maximum number of days any one of those drugs was consumed. For example, if powdered cocaine had been consumed on 15 days and crack cocaine had been consumed on 20 days, then COCAINE was coded as "20 days."

We expected the relationship between expenditures and days of consumption to be nonlinear, but the logarithmic translation may not have been adequate to capture that nonlinearity. Consequently, each of the above category variables was raised to the second power, creating additional independent variables: COCAINE2, HEROIN2, MARIJ2, and OTHER2.

Cocaine, heroin, and other drugs are frequently consumed in combination. For example, heroin users often use cocaine, a stimulant, to moderate the effect of heroin, a depressant. However, someone who uses a combination of heroin and cocaine on a daily basis is unlikely to consume the same amount of heroin and cocaine that is consumed by two people who are daily users and exclusive in their drug use.

Consequently, two interaction terms were added to the regression. COKEHER equals COCAINE x HEROIN. COKEHER2 = COKEHER²/300. The division by 300 facilitates the computing algorithm, but otherwise has no substantive importance for the analysis. The consumption of other drugs was relatively infrequent, so we did not add an interaction term to the regression for this variable.

Starting in 1990, DUF respondents were asked whether they had consumed any drugs in addition to those listed in the interview. A variable OTHERDRG denotes that some other drug had been consumed (1=yes, 0=no). This question was not asked during 1989.

Roughly midway through 1995 a new DUF survey instrument was introduced. While many of the questions remained the same between the two surveys, some questions important to our analysis changed. Most notably, the old survey instrument asked about the amount of money spent on illegal drug in the past week, while the new survey asked about the amount spent on illegal drugs in the past month. After extensive testing of the data, we concluded that these questions were not comparable, and forced us to use only the data from the old survey.

Estimation

We used ordinary least squares to estimate the regressions. Results are presented in Table B–1.

Table B–1
Statistical Results for Regression Analysis of Drug Expenditures

Click for larger image

The model's explanatory power appears remarkable given the presumed measurement error in these data. Residuals were plotted against the number of days that the respondent reported using cocaine, heroin, marijuana, other drugs, and the interaction term. These plots indicate that the logarithmic transformation does a sufficient job of inducing normality among the residuals and that the model specification does not systematically distort the relationship between days of use and amount of money spent.

Interpretation

We converted predictions based on the regression reported in Table B-1 from logarithms to natural units using two approaches. When Ln($) is the predicted value of the original regression, then the median value in the original units is Median($) = Exponential(Ln($)), and the mean value in the original units is Mean($) = Exponential(LN($)+/2).

When cocaine is the only drug consumed, estimating expenditures on cocaine is straightforward. First, substitute zeros for all independent variables other than COCAINE and COCAINE2. Second, use the regression results to make predictions when COCAINE = 1, COCAINE = 2, ... COCAINE = 30. Similar calculations yield estimates for expenditures on heroin when heroin is the only drug consumed.

For example, when cocaine is consumed 10 days a month, the median weekly expenditure is somewhat more than $80. It is about $200 a week when cocaine is consumed on 20 days a month, and it is about $300 a week when cocaine is consumed on 30 days a month.

When broken down by daily expenditure, spending on heroin and cocaine is about the same. However, this does not mean that when cocaine and heroin are consumed in combination, expenditures on each are equally divided. More likely, one of the drugs is the drug of preference, and the other is used frequently, but at a lower dosage.

When cocaine and heroin were consumed in combination, we attributed greater expenditure to what appeared to be the dominant drug. Let $ represent the predicted dollar expenditure on drugs by individuals who consume cocaine and heroin, but no other drugs. Let Nc represent the number of days a month that an individual consumed cocaine, and let Nh represent the number of days a month that individual consumed heroin. Expenditures on cocaine and heroin are estimated as:

where ADJ = 0.5 when Nc > Nh and ADJ = 2.0 otherwise. According to this formulation, when cocaine is consumed on more days than heroin, at least two-thirds of the drug expenditure is attributed to the purchase of cocaine. When heroin is consumed on more days (or the same number of days) as cocaine, then at least two-thirds of the drug expenditure is attributed to heroin. As a practical matter, this rule dictates that respondents who say that they use both heroin and cocaine daily spend two-thirds of the money on heroin and one-third on cocaine. This division seems appropriate given evidence that such individuals typically are long-established heroin users who add a small amount of cocaine to their consumption.³

Typical Expenditures

Using results from the above regression, coupled with assumptions about how joint expenditures on heroin and cocaine should be apportioned, we estimated the median and mean expenditures for cocaine and heroin for every respondent who used either drug heavily. We averaged those estimates over all respondents who admitted using cocaine or heroin on more than 10 days during the month before their arrest. Those averages were weighted by the number of arrests reported by the FBI in each MSA represented by the DUF city. Results are reported in Table B–2.

Two additional adjustments were made to the statistics before transfering them to Table 4. A dollar in 1996 is worth roughly $0.80 in 1989 purchasing power, based on the Consumer Price Index. although it is arguable whether heroin and cocaine purchases should be deflated by the CPI, it seems misleading to ignore inflation over the eight years from 1988 through 1995, and the CPI is the best reflection of how purchasing power has changed. Thus, expenditures have been adjusted to take inflation into account.

The second adjustment was to convert the estimates to a three-year moving average centered on the year in question. For example, the estimate for 1993 is the average of the estimates for 1992, 1993, and 1994. We used a three-year moving average to reduce year-to-year measurement errors, but an examination of Table B–2 will reveal that averaging will not make major changes to the estimates.

The chief problem in interpreting these numbers is that the medians are so different from the means. Which should be used as "typical" expenditures? Evidence presented later seems to indicate that the median is preferable, but the evidence is not persuasive. Before turning to this evidence, the matter of earnings from income in kind must be considered.

Table B–2
Mean and Median Expenditures on Cocaine and Heroin, 1989–1995

Click for larger image

Accounting for Income in Kind

Hardcore drug users support their drug use through legitimate sources and through crime, especially drug dealing. Returns from dealing are often in the form of drugs as a payment for services rendered, or "income in kind." How does income in kind affect our estimates?

It is sometimes asserted that most hardcore users pay for their drug use by dealing or assisting others who deal in illicit substances. For example, Johnson and colleagues⁴ report that in their sample of New York City heroin users, daily users spent an average of $7,601 a year on heroin, but consumed about $13,189 worth of heroin a year. Regular users (defined in this study as those who use heroin between three and six times a week) spent $4,019 a year on heroin, but consumed about $6,431 worth of the drug a year. The difference between expenditures and consumption represents in-kind earnings in the form of drugs. If this were a typical pattern, then the expenditures on drugs computed from DUF should be inflated by ($13,189/$7,601) or 1.73 for daily users, and ($6,431/$4,019) or 1.6 for regular users. For reasons reported here, such multipliers seem much too large.

First, consider a hypothetical illustration of a drug market. Suppose that mid-level dealers have 100,000 units of drug X to sell and, at $1 a unit, demand $100,000 for their drugs. Suppose initially that there are 100 hardcore users, but no other users. Then, each hardcore user must generate $1,000 of income for the mid-level dealers.

The 100 hardcore users could support their use partly from dealing, but clearly they cannot support their use entirely from dealing. Selling exclusively to each other would not raise the $100,000 expected by mid-level dealers. The $100,000 must come from some other source.

Now, these 100 users could support one-third of their consumption by selling 66,666 units for $1.50 a unit to each other. This is a Pyrrhic marketing success, of course, but it would be effective if alternative income sources sometimes provided sufficient funds to make purchases and sometimes did not. The total revenue generated is $100,000, enough to satisfy the middlemen. This division of drugs would approximate what was observed by Johnson and colleagues in New York.

This solution requires that the street price of drug X be 50 percent higher than the price to mid-level dealers. In reality, retail prices for cocaine and heroin are about one-third higher than wholesale prices,⁵ so income in kind would seem more likely to add about 33 percent to our estimates of drug consumption based on transactions in which money was exchanged.

Moreover, many sellers do not themselves consume drugs (at least not at heavy-use rates). To extend the above illustration, suppose that one-half the 100,000 units of drug X are sold by entrepreneurs who do not themselves consume drugs. Then, the 100 hardcore users could support only about 17 percent of their own drug use by selling drug X at $1.5 a unit to other users. The market opportunities for hardcore users to support their own consumption with income in kind is limited by sales by nonusers. Sales by nondrug users may be sizable.

Reuter and colleagues report that only 11 percent of the dealers they interviewed retained one-half or more of the drugs for personal consumption; 30 percent retained less than one-half, "usually only 'a little of it.'"⁶ If "a little of it" means 15 percent, then these figures suggest that about 10 percent of the drugs that were available to these dealers were retained for personal consumption. If Reuter's dealers are typical of those who sell drugs, then the expenditure figures based on dollar transactions should be increased by 0.1/0.9, or about 11 percent to account for income in kind.

In his field study, conducted in 1982 and 1983, of 15 street-level heroin dealers in Detroit, Mieczkowski⁷ reports that dealers are typically not hardcore users: "...although runners appear by and large to be recreational drug users, they are not addicted to heroin." Mieczkowski's findings suggest that income in kind represents a smaller percentage of drug consumption than was reported by Johnson.

Altschuler and Brounstein⁸ interviewed 387 ninth and tenth grade, minority, inner-city boys from Washington, D.C., during 1988. Of the 387, 7 percent used drugs, but did not sell them; 9 percent sold drugs, but did not use them; and 4 percent both sold and used drugs. These findings suggest that many drug sales are made by dealers who are not hardcore users.

Williams tracked the drug (cocaine and crack) dealing of eight New York juveniles who belonged to a teenage drug ring called the Cocaine Kids, or the Kids. Williams reports that "...virtually all cocaine suppliers expect retail dealers to return with cash amounting to about 60 to 75 percent of potential retail sales of their consignment."⁹ If this profit margin is typical for cocaine and crack retailers, and if all this profit is income in kind spent on the retailer's consumption, then estimates based on dollar transactions might be multiplied by 0.66 to 0.33. However, the dealers interviewed by Williams did not take their profits primarily in the form of crack: "All the Kids snort cocaine regularly. This is accepted, but the use of crack is generally frowned upon: those who snort are thought to have more control and discipline than those who smoke crack or freebase. Most dealers see crack smokers as obsessive consumers who cannot take care of business; crack users, they say, tend to become agitated, quickly lose control and concentration, and take one dose after another at the expense of everything else."¹⁰

Skolnick,¹¹ who examined crack sales by gang members in California during 1988, reports two types of dealers: one who sells for profit and one who sells to buy drugs. Interestingly, Skolnick also reports that 75 percent of street sales will be returned to the middleman, a figure consistent with that reported by Williams in New York.

Waldorf and Lauderback interviewed 568 members of 86 different ethnic gangs in San Francisco.¹² They reported that only 16 percent of the crack sellers used crack during the month before the interview, although about one-half of the cocaine sellers and about three-fourths of the heroin sellers used those drugs during the month before the interview. The gang members explained that intoxicated sellers did not make reliable dealers and that drug dependence impaired the gang member's ability to defend the gang. Waldorf and Lauderback reached similar conclusions to Chin¹³ (Chinese gang members who sold heroin did not use it) and Vigil.¹⁴

Mieczkowski, on the other hand, reports that crack sellers in his Detroit sample "appear to conform closely to the 'classic' or 'hustler' view of the drug user."¹⁵ Nearly two-thirds of the respondents said that they sold crack to get money for their own crack consumption.

The important point is that many of the drugs consumed by hardcore users are sold by individuals who do not use drugs heavily. The ability of hardcore users to support their own use through dealing is necessarily limited. Consequently, the amount of drugs that hardcore users receive as income in kind cannot account for much of the cocaine and heroin consumed.

Data are not sufficient to support precise estimates. It seems that a street dealer might be able to retain about one-fourth of the drugs that he markets, and that profit dealers (those taking their profit in cash rather than in kind) are more numerous among cocaine dealers than among heroin dealers. We assume that two-thirds of the cocaine dealers and one-third of the heroin dealers are profit dealers.

Assume that a cocaine retailer must return $3 for every $4 of crack or powdered cocaine that he sells. Also assume that two-thirds of all retail dealers are profit dealers and one-third are users. This means that every $1 spent on crack and cocaine would result in $1 x 0.33 x 0.33 = $0.11 in income in kind, suggesting that the estimates should be inflated by 0.11. This inflation figure equals the 11 percent income in kind figure derived from Reuter and colleagues' study.

Second, assume that a heroin retailer must return $3 for every $4 of heroin that he sells. Also assume that one-third of the retail dealers are profit dealers. This means that every $1 spent on heroin would result in a maximum of $1 x 0.33 x 0.66 = $0.22, suggesting that the estimate should be inflated by 0.22. This inflation figure is lower than the income in kind figure derived from Johnson and colleagues' study but is more consistent with observations that not all those who sell heroin are hardcore users.

Based on the above evidence, we assume that $0.22 worth of heroin is retained as income in kind for every $1 of heroin sold. For cocaine, we assume one-half that amount, or $0.11, for every $1 sold.

Choosing the Median As the Typical Expenditure

If weekly expenditures on drugs were reported with perfect accuracy, there would be little justification for using any number other than the mean. After all, regardless of how the data are skewed, the mean is the average expenditure, and total expenditures will equal the average expenditure multiplied by the number of hardcore users.

However, another interpretation seems more reasonable. Suppose that the average expenditure is about the same for everybody who uses drugs on a specified number of days a month, but that the amount spent on drugs is reported with great inaccuracy. From this view, the median is the best measure of the average expenditure.¹⁶ Some other sources suggest that the median expenditure is more accurate for our analysis.

Other reports of expenditures on drug use

Other studies, primarily of hardcore users involved with the criminal justice system, estimate expenditures on heroin and cocaine that are broadly consistent with the medians reported here (Table B-2). Reviews of those studies appear here. In each study, the authors' estimates of expenditures have been inflated to 1996 dollar equivalents using the CPI.

Johnson and colleagues¹⁷ interviewed 201 subjects who were street-level heroin users in East and Central Harlem; all were involved in some form of criminality and spent most of their time on the streets. Subjects were interviewed for five consecutive days, and then were interviewed weekly for the following four weeks. About 132 of these subjects were interviewed four more times at three- to six-month intervals. The average user spent $6,200 a year on heroin. These users often sold drugs, and when they did, payment was usually in the form of drugs as income in kind. When income in kind was taken into account, these users spent about $10,300 a year on heroin. Daily users directly purchased $12,000 worth of heroin a year, but when income in kind is taken into account, they spent about $19,000. Regular users (those who used at least weekly, but less than daily) made cash payments of $5,900 for heroin over the course of a year, but with income in kind payments, their annual expenditures were $6,431. These estimates are comparable with those based on the median responses in the DUF data, which suggest that hardcore users of heroin spend about $18,000 a year on heroin in 1989, the year most comparable to the Johnson study.

Johnson and Wish¹⁸ recruited 105 male New York "hard-drug abusers" who had committed one or more relatively serious non-drug crimes (such as robbery, burglary, grand larceny, or assault) in the past 24 hours. Those who had committed recent robberies spent an average of $74 a day on illicit drugs. Those who had committed other crimes spent an average of $46 a day. Thus, for those who had just committed crimes, the expenditure on drugs was $319 to $518 a week.

For those who bought both heroin and cocaine, daily expenditures totaled $368 to $518 week. Those who bought only cocaine spent $249 to $329 a week on cocaine. Those who bought only heroin spent $219 to $358 a week on heroin. It is notable that 86 percent of these subjects reported using some illicit substance on 28 of the past 30 days, so the majority could be considered hardcore drug users. These figures seem to be high estimates of consumption, however. Because all these users had recently committed serious crimes, they had money available from illegal sources to buy drugs. Nevertheless, the average expenditures were about the same as those based on the median values from the DUF data—$283 in 1989.

Reuter and colleagues report results based on interviews with 186 males on probation in Washington, D.C., who had sold drugs during the mid-1980s. About one-half reported purchasing drugs for their own use. This half had a median expenditure of $580 a month; the mean was $2,300. However, about 40 percent of the respondents consumed some of the drugs that they acquired for dealing, representing income in kind spent on drugs; about 10 percent reported that they consumed one-half the drugs that they acquired by dealing. The median and mean are much smaller than their counterparts in DUF, but the Reuter subjects are not necessarily hardcore users.¹⁹

Mieczkowski²⁰ also interviewed 97 crack users who were arrested in Detroit, who tested positive for cocaine, and who admitted to using crack. On average, this group used two dozen rocks of crack, at a cost of $329 per week. Although these were not necessarily heavy users, all used crack cocaine at least weekly. Mieczkowski makes an important point: "It also appears true that as levels of use rise and approach 20 to 30 rocks a week, respondents have increasing difficulty in estimating their levels of consumption. This is apparently due to sharing crack with friends, increased personal consumption, and accepting in-kind services as payment for crack." (p. 8)

Mieczkowski²¹ also asked 190 "chronic users of crack cocaine" in Detroit about their crack consumption by appending questions to the DUF interview. About one-fourth of these users consumed four or fewer rocks²² a week ($50 or less a week); most of the others clustered at 10 to 20 rocks a week ($125 to $250 a week) and 40 to 50 rocks a week ($500 to $625 a week). Only 5 percent used over 100 rocks a week. Mieczkowski speculated that the upper range included dealers who could not distinguish between their own consumption and what they sold, as well as individuals who were sharing with friends.

The Office of National Drug Control Policy sponsored a study of hardcore drug users in Cook County, Illinois, during 1995.²³ A hardcore users was one who: (1) used heroin or cocaine on at least eight days during either of the two months just before the interview, or (2) used heroin or cocaine during the last 60 days and scored high on a scale of addiction. Practically, the first criterion was determinative, so these users satisfy the heavy user criterion used in this report on retail sales.

Each subject in this study was asked separately about his or her use of heroin, powdered cocaine, and crack cocaine. Regarding heroin, he was asked: "In the past 60 days, how much did you usually spend on heroin for your own use on those days when you were using?" The interviewer clarified the question by saying: "I know the days can vary quite a bit, but when you were using, how much did you spend for sure on heroin? If you got it in trade, on credit, or for some services, just give me your best guess as to what it was worth." Because of this last qualifying statement, consumption includes heroin gotten as income-in-kind, so these estimates do not need to be adjusted to account for heroin that was gotten without being paid for in currency.

Interviewers also asked subjects about how many days they used heroin during the 60 days before the interview. The 304 subjects who used heroin said they used it an average of 5.5 days per week. They spent an average of $61 per day on those days when they used. They spent an average of $369 per week during the two months that predated the interview.²⁴ The median amount was $290.

Interviewers asked similar questions about crack and powdered cocaine use. The 164 subjects who said they used powered cocaine said they used the drug an average of 3.8 days per week, and they spent an average of $91 per day of use. The average expenditure was $393, and the median expenditure was $240. the 637 subjects who said they used crack cocaine admitted using it on 4.6 days per week and spent an average of $89 per day. They said they spent an average of $422 per week, and a median amount of $300.

Some of these subject used just powdered cocaine, some used just crack cocaine, and some used both. Without distinguishing between powdered and crack cocaine, we estimate that hardcore users of cocaine spend an average of $476, and a median of $290, per week.

The expenditure patterns are skewed. A few hardcore users have relatively high reported expenditure patterns, which causes the average amount spent on cocaine to be much more than the median amount. To illustrate this, suppose that no hardcore users could spend more than $500 per week on cocaine, so that any reports of over $500 should be truncated at $500. Then the mean expenditure on cocaine would be $299. If the upper limit were $750, then the average would be $357. An upper limit of $1,000 yield an average of $388, and an upper limit of $1,250 leads to an average of $412.

Three other studies provide partial support for the estimates used in this report. Fagan²⁵ interviewed 1003 drug users and sellers, whom he recruited from two sites—Central Harlem and Washington Heights, New York using snowball sampling techniques. Crack users were almost two-thirds of the sample, and cocaine/crack users were almost three-quarters of the sample, in both sites. During the year following initiation of crack use, males in Central Harlem spent $1,380 dollars per month and females spent $800 per month. This expenditure patterns were not much different than those assumed in this report. However, in Washington Heights, the averages were $1,950 per month for males and $2,100 per month for females. These expenditure patterns are more like the mean expenditures estimated from DUF.

A study commissioned by ONDCP examines purchase patterns by heroin users. Rocheleau and Boyum²⁶ located about 50 heroin users in each of three cities (New York, Chicago, and San Diego) using snowball sampling. These users were asked about their purchasing habits and, how much they spent on heroin and other illicit drugs. Across the three sites, heroin users spent an average of $290 (median = $230) per week on heroin and an average of $380 (median = $270) on all drugs. These estimates are consistent with the dollar expenditures assumed in this report, but Rocheleau and Boyum also computed expenditures by multiplying the amount of heroin used by the average price paid, to estimate that heroin users spent an average of $500 (median = $315) per week on heroin. This is more than we have assumed in our calculations.

The studies cited above are consistent with the conclusion that the median values based on DUF data typify spending patterns for those arrestees who admitted using cocaine or heroin on at least 11 days during the past month. However, some of these studies report expenditures hearer the mean; moreover some of the studies described below report larger expenditure patterns.

Mieczkowski²⁷ reports on interviews with "100 self-reported dealers and user/dealers of crack cocaine" who were in residential treatment facilities in Detroit. All can be considered to be hardcore users. The amounts reported on weekly drug usage were highly skewed. The estimates were: $1,232 mean; $1,150 trimmed mean; $789 median; $716 M-estimator. These estimates are considerably higher than those we report, although they are not inconsistent with estimates for the very heaviest users. One explanation of this variation may be that these users had especially high use patterns, as evidenced by their seeking treatment.

Other studies of treatment populations indicate that expenditures can be much higher for the typical hardcore user than is assumed here. Schnoll and colleagues²⁸ report on expenditures by 172 men and women who received treatment for cocaine abuse in Chicago primarily during 1982 and 1983. Average expenditures were reported as $1,270 a week.

Gawin and Kleber²⁹ describe cocaine use in a sample of 30 consecutive admissions to a cocaine treatment program in New Haven. Thirteen intravenous drug users used an average of 5.6 grams a week, six smokers used an average of 9.1 grams, and 11 who snorted used an average of 5.3 grams a week. If these users paid $100 a gram, they must have spent $500 to $900 a week for cocaine prior to entering treatment.

Collins, Hubbard, and Rachal (1985) studied annual drug expenditures of 3,276 drug users who entered publicly-funded drug treatment in 1979.³⁰ For daily heroin users, the median drug expenditure was $17,000, and the mean was almost $28,000. For weekly users (exclusive of daily users), the median was $6,700, and the mean was about $7,400. There were fewer regular cocaine users. On a yearly basis, daily users spent a median of $23,000, and a mean of almost $32,000. Those who used on a weekly (but not daily) basis had median expenditures of $10,000, and a mean of almost $20,000.

Comparing these estimates with those based on DUF is complicated. Because these estimates are from 1978 and 1979, an adjustment for inflation leads to estimates that are higher than those based on DUF. However, the street price of heroin and cocaine have fallen so much since 1978–1979 that, even controlling for inflation, that 1978–1979 is a questionable benchmark. Another limitation is that cocaine users who sought treatment in the late 1970s may have little resemblance to crack users of the late 1980s and early 1990s.

Although the latter studies, all of which are based on a population in treatment, indicate that hardcore users spend more on cocaine than is assumed in this study, users in treatment probably have use patterns that are atypical of hardcore users in general. As Waldorf and colleagues³¹ report, most hardcore cocaine users are able to control their consumption, avoiding the ruinous expenditure patterns that often drive other users—those who have the least control—into treatment.

Evidence from the NHSDA

Additional evidence comes from the NHSDA. The number of individuals who admitted using cocaine on more than 10 days during the month before the interview and reported how much they had spent on cocaine was small, thus the estimates for cocaine expenditures are fairly unreliable (Table B–3). For example, in 1993, only seventy-eight individuals admitted hardcore cocaine use and reported cocaine expenditures. Average expenditures were $81 a week. These average expenditures reported by hardcore cocaine users in NHSDA were much lower than those reported in DUF (Table 2). Although the NHSDA estimates appear to show a decrease in average expenditures by hardcore cocaine users from 1988 ($130) to 1993 ($81), this trend is neither consistent nor reliable.

Table B–3
Average Expenditures on Cocaine by Hardcore Cocaine Users, 1988–1993

Estimates based on amount consumed

Another way to validate the median as a measure of expendi-tures is to infer how much hardcore cocaine and heroin users could spend given their consumption patterns.

It is difficult to shoot heroin more frequently than four times a day, and many DUF respondents used heroin less often than daily.³² According to Division of Substance Abuse Services (DSAS) in New York City, a hardcore user might use one to two bags of heroin a session, and each bag would cost $10. These approximations suggest that a hardcore user could not spend much more than $420 a week.

Although $420 is close to the mean expenditure estimated based on the DUF data, even those heroin users who are hardcore consumers cannot shoot heroin every day, four times a day. Clearly, $420 a week should be considered more as an upper limit than an average for weekly expenditure on heroin.

Cocaine is different. A study commissioned by the National Institute on Drug Abuse³³ provided estimates of the amount of pure heroin used by heroin addicts in San Francisco, Baltimore, and Newark. The estimates, which are summarized below, pertain to addicts with high, average, and low habits.

Table B–4
Heroin Usage Patterns by Heroin Addicts in Three Cities

The study does not report dollar expenditures, but some estimates are possible based on the apparent price paid per milligram in San Francisco and Baltimore, and an assumption that the price paid per milligram in Baltimore is the same as that paid in Newark. We also assume that these users purchased heroin on 5.5 days per week, a number that is consistent with observations by Simeone, Rhodes, and Hunt in Chicago. These calculations suggest that the average weekly expenditure on heroin was about $98 in San Francisco, $462 in Baltimore, and $380 in Newark. While heroin seems to be self-limiting (the user's craving can be satisfied much as a diner is satiated after a full meal), cocaine is notable because it immediately engenders a desire for more cocaine. As discussed earlier, there are reports of very hardcore consumption patterns just prior to seeking treatment.

Nevertheless, cocaine consumption has two limitations. The first is physiological. Binge use exhausts the body, so rest is necessary before binge use can recur. The second is that hardcore cocaine use can quickly exhaust financial resources. One way to check estimates of drug spending patterns for cocaine is to assess the user's income from legal and illegal sources.

Estimates Based on Financial Resources

One way to assess the practical upper limit for cocaine use (excluding the minority of users who drain personal resources prior to entering treatment) is to estimate the amount of money that hardcore users have available to spend on cocaine.

According to DUF, most hardcore users who are arrested and questioned have limited legal incomes. Much of their earnings comes from crime. How much do such individuals earn from combined sources?

As would be expected, it is no easier to estimate the amount of money earned from crime than it is to estimate the amount of money spent on drugs. There are, however, a few studies of earnings from property crime. Johnson and colleagues, in a study described above, report that daily heroin users earn an average of $13,000 a year from non-drug criminal activity, and regular users earn $9,200 a year. Total criminal income, including drug income in kind, is $28,000 for daily users and $16,000 for regular users. After living expenses are subtracted from these incomes, the amount available for heroin expenditures could not be much greater than is assumed in this report.³⁴

Anglin and colleagues³⁵ describe the income of 279 male heroin addicts who were selected from those who had first entered a methadone program between 1971 and 1973. The period of time described is the 12 months prior to their first period of legal supervision. Chicano respondents averaged $6,708 in illegal income a year (not counting $924 a year from drug dealing), and whites averaged $8,580 a year (not counting $1,320 a year from drug dealing). Legal incomes were $1,984 to $2,672 a year. Even when inflation is taken into account, these incomes could not support drug use habits far in excess of what is assumed in this paper.

Reuter and colleagues report results based on interviews with 186 males on probation in Washington, D.C., who had sold drugs during the mid-1980s. They report an average income of $4,000 a month, all but $1,200 from illegal activity, mostly drug sales. However, most of this income was spent on expenses other than drugs—drug expenditures averaged $883 a month.³⁶

Fagan also reports monthly income. His table 4.7 is a cross-tabulation with multiple breakdowns and is hard to summarize because he does not report the number of observations in each cell. Nevertheless, patterns are apparent. Those who do not sell drugs regularly make less than $1,100 per month on average, sometimes considerably less when they are disabled. Those classified as sellers make about $3,000 in Central Harlem and roughly $4,000 in Washington Heights. There is considerable variation about these averages, however, and the reported earnings for drug sellers are inflated because they are gross estimates that do not subtract the cost of obtaining drugs for sale.

Conclusions

The evidence is not compelling, but it seems best to assume that the median expenditures on cocaine and heroin—as measured from DUF data—provides the best basis for computing dollar expenditures on cocaine and heroin. The uncertainty surrounding this assumption is best handled through sensitivity analysis, which we execute in the main report.

The evidence in support of the percentage of drugs earned as income in kind is also meager. We assume that for every dollar spent on cocaine another $0.11 of cocaine is consumed as income in kind. We assume that for every dollar spent on heroin another $0.22 of heroin is consumed as income in kind.

Endnotes

¹ No question was asked about the amount of expenditures on drugs in the 1988 DUF data, so no analysis was performed for that year.

² Drug users also have difficulty recalling how often they used a drug, how much they used, and how much they paid for it. However, this inaccuracy, unlike intentional denial or deception, probably averages out when the data are aggregated.

³ D. Hunt, "Tracking the Prevalence of Heroin Use," paper prepared for ONDCP, July 1992.

⁴ B. Johnson et al., Taking Care of Business: The Economics of Crime by Heroin Abusers (Lexington, Massachusetts: Lexington Books, 1985).

⁵ W. Rhodes, R. Hyatt, and P. Scheiman, "The Price of Cocaine, Heroin, and Marijuana, 1981–1993," Journal of Drug Issues, 24, no. 3. (1994): 383–402.

⁶ P. Reuter et al., Money from Crime: A Study of the Economics of Drug Dealing in Washington, D.C. (Santa Monica, California: Rand Corporation, 1990) RAND publication R–3894–RF, p. 61.

⁷ T. Mieczkowski, "Geeking Up and Throwing Down: Heroin Street Life in Detroit," Criminology, 24, no.4. (1986): 645–665.

⁸ D. Altschuler and P. Brounstein, "Patterns of Drug Use, Drug Trafficking, and Other Delinquency among Inner-City Adolescents in Washington, D.C.," paper presented at the Annual Meeting of the American Society of Criminology, Reno, Nevada, 1989, 9; and Table 3.

⁹ T. Williams, The Cocaine Kids: The Inside Story of a Teenage Drug Ring (Reading, Massachusetts: Addison-Wesley Publishing, 1989), 36.

¹⁰ Williams, The Cocaine Kids, 47.

¹¹ J. Skolnick, The Social Structure of Street Drug Dealing (BCS FORUM, Office of the Attorney General, State of California, undated).

¹² D. Waldorf and D. Lauderback, "Don't Be Your Own Best Customer—Drug Use of San Francisco Ethnic Gang Drug Sellers," Crime, Law and Social Change—An International Journal, 19 (1993): 1–15. The published article was based on a population of 300; however, an updated version surveyed 568 gang members.

¹³ K. Chin, "Chinese Triad Societies, Tongs, Organized Crime and Street Gangs in Asia and the United States," unpublished Ph.D. dissertation, University of Pennsylvania, Philadelphia, Pennsylvania, 1986.

¹⁴ J. Vigil, Barrio Gangs: Street Life and Identity in Southern California (Austin, Texas: University of Texas Press, 1988).

¹⁵ T. Mieczkowski, "Crack Distribution in Detroit," Contemporary Drug Problems, (Spring 1990): 23–24.

¹⁶ An analogy helps make this point. Suppose that a grocery store clerk were to ring 1,000 $1 candy bars individually on his register. Suppose that he was inaccurate but unbiased as he occasionally registered too many or too few zeros: 100 bars were priced at $0.10, 800 were priced at $1, and 100 were priced at $10. The total expenditure on candy bars would be $1,810, or an average of $1.81 per candy bar. Here the clerk's random errors do not balance out.

¹⁷ B. Johnson et al., Taking Care of Business: The Economics of Crime by Heroin Abusers (Lexington, Massachusetts: Lexington Books, 1985).

¹⁸ B. Johnson and E. Wish, "The Robbery-Hard Drug Connection: Do Robbers and Robberies Influence Criminal Returns and Cocaine-Heroin Purchases," paper presented at the Criminology Section of the American Sociological Association, August 17, 1987.

¹⁹ P. Reuter et al., Money from Crime: A Study of the Economics of Drug Dealing in Washington, D.C. (Santa Monica, California: Rand Corporation, 1990) Rand Publication R–3894–RF, p.61.

²⁰ T. Mieczkowski, Understanding Life in the Crack Culture: The Investigative Utility of the Drug Use Forecasting System (National Institute of Justice, Research in Action, November–December 1989).

²¹ T. Mieczkowski, "Crack Distribution in Detroit," Contemporary Drug Problems, (Spring 1990): 18–20.

²² Crack is made by heating powdered cocaine until it crystallizes. "Rocks" are then broken off the chunk of crack produced. Crack is often bought by the rock. Although this purchase unit varies in weight and size, rocks tend to be fairly small and inexpensive.

²³ R. Simeone, W. Rhodes and D. Hunt, A Plan for Estimating the Number of "Hardcore" Drug Users in the United States: Preliminary Findings, January 1997. Estimates reported above were based on previously unreported tabulations from that study.

²⁴ This average comes from multiplying the average number of days per week that the subject used heroin times the average expenditure per day for each heroin users, and then taking the average over all users. This is not the same as multiplying the average expenditure per day for all users by the average number of days used heroin for all users. Thus, the reader cannot confirm the weekly expenditure estimate from the number cited above. The tabulations were weighted to reflect the population of hardcore drug users in Cook County.

²⁵ J. Fagan, Drug Selling and Licit Income in Distressed Neighborhoods: The Economic Lives of Street-Level Drug Users and Dealers, in Drugs, Crime and Social Isolation, edited by A. Harrell and G. Peterson (The Urban Institute Press, Washington, D.C., 1992): 99–146.

²⁶ A. Rocheleau and D. Boyum, Measuring Heroin Availability in Three Cities (Office of National Drug Control Policy, Washington, D.C., November 1994).

²⁷ T. Mieczkowski, "The Economic Dimensions of Crack Use and Distribution: Some Preliminary Data," paper presented to the American Society of Criminology Annual Meetings, Reno, Nevada, November 1989.

²⁸ S. Schnoll et al., "Characteristics of Cocaine Abusers Presenting for Treatment," in Cocaine Use in America: Epidemiological and Chemical Perspectives, ed. N. Kozel and E. Adams (Rockville, Maryland: National Institute on Drug Abuse, 1985), NIDA Research Monograph 61, 171–181.

²⁹ F. Gawin and H. Kleber. "Cocaine Use in a Treatment Population: Patterns and Diagnostic Distinctions," in Cocaine Use in America: Epidemiological and Chemical Perspectives, ed. N. Kozel and E. Adams, (Rockville, MD: National Institute on Drug Abuse, 1985), NIDA Research Monograph 61, 182–192.

³⁰ J. Collins, R. Hubbard, and J. Rachal, "Expensive Drug Use and Illegal Income: A Test of Explanatory Hypotheses, Criminology", 23, no. 4. (1985): 743–763.

³¹ D. Waldorf et al., Cocaine Changes: The Experiences of Using and Quitting (Philadelphia: Temple University Press, 1991).

³² Kahn reports an average of three fixes a day for 453 clients before their participation in a methadone maintenance program. Only 16 percent of his clients reported more than four fixes daily. Anglin reports that during the 12 months before entering criminal justice supervision, 279 heroin addicts (who had been identified through a methadone maintenance program in the early 1970s) had injected heroin an average of 2.7 times a day. R. Kahn, "The Frequency of Narcotic Use Before and After Admission to a Methadone Maintenance Program," International Journal of the Addictions, 14, no. 8. (1979): 1157. M. Anglin et al., Effects of Legal Supervision on Narcotics Use and Criminal Behavior over the Addiction Career (Los Angeles, CA: UCLA Drug Abuse Research Group, December 1988).

³³ National Institute on Drug Abuse, Heroin Addict Habit Size in Three Cities: Context and Variation (NIDA, Rockville, Maryland, downloaded from the Internet, March, 1996).

³⁴ B. Johnson et al., Taking Care of Business: The Economics of Crime by Heroin Abusers (Lexington, Massachusetts: Lexington Books, 1985).

³⁵ M.D. Anglin et al., Effects of Legal Supervision on Narcotics Use and Criminal Behavior Over the Addiction Career (Los Angeles, CA: UCLA Drug Abuse Research Group, December 1988); and Table 3.

³⁶ P. Reuter et al., Money from Crime: A Study of the Economics of Drug Dealing in Washington, D.C. (Santa Monica, California: Rand Corporation, 1990), Rand Publication R–3894–RF, p. 61.