"We are making a difference. Drug use is down 50% over the last decade." -
President William J. Clinton, Preface to The National Drug Control
Strategy, 1998 [1]

"When assessing evidence, it is helpful to see a full data matrix, all
observations for all variables, those private numbers from which the public
displays are constructed. No telling what will turn up." - Edward R. Tufte [2]

DRUG PROHIBITION AND PUBLIC HEALTH

S Y N O P S I S

FOR THE PAST 25 YEARS, the US has pursued a drug policy based on
prohibition and the vigorous application of criminal sanctions for the use
and sale of illicit drugs. The relationship of a prohibition-based drug
policy to prevalence patterns and health consequences of drug use has never
been fully evaluated.

To explore that relationship, the author examines national data on the
application of criminal penalties for illegal drugs and associated trends
in their patterns of use and adverse health out-comes for 1972-997.

Over this 25-year period, the rate at which criminal penalties are imposed
for drug offenses has climbed steadily, reaching 1.5 million arrests for
drug offenses in 1996, with a tenfold increase in imprisonment for drug
charges since 1979. Today, drug enforcement activities constitute 67% of
the $16 billion Federal drug budget and more than $20 billion per year in
state and local enforcement expenditures, compared with $7.6 billion for
treatment, prevention, and research.

Despite an overall decline in the prevalence of drug use since 1979, we
have seen dramatic increases in drug-related emergency department visits
and drug-related deaths coinciding with this period of increased enforcement.

Further, while black, Hispanic, and white Americans use illegal drugs at
comparable rates, there are dramatic differences in the application of
criminal penalties for drug offenses. African Americans are more than 20
times as likely as whites to be incarcerated for drug offenses, and
drug-related emergency department visits, overdose deaths, and new HIV
infections related to injecting drugs are many times higher for blacks than
whites.

These outcomes may be understood as public health consequences of policies
that criminalize and marginalize drug users and increase drug-related risks
to life and health.

WE ARE BY NOW accustomed to sharply opposing view-points and conflicting
claims about our national drug policy and its results. A succession of
Presidents and Congresses have led the field with calls for a "drug-free"
America and "zero tolerance" and have enacted drug prohibitions with
ever-harsher criminal penalties and more militant (and more expensive)
enforcement tactics. In contrast, libertarian reformers like Nobel Prize
winner Milton Friedman or conservatives like William F. Buckley, Jr., call
for outright legalization of all drugs. And others (this author among them)
call for a public health or "harm reduction" approach, [3] reasoning that
dangerous drugs will always be with us and that we had better learn how to
live with them in a way that minimizes their adverse health and social
consequences.

While this debate rages, we see continued (even rising) drug availability
and ever-shifting patterns of drug use: crack and cocaine use are down, but
marijuana and heroin use are becoming more popular among young people. [4]

And, over the last decade, new and more lethal consequences of illicit drug
use have emerged--including infectious disease epidemics (AIDS, TB,
hepatitis B, and hepatitis C) linked to unsafe injecting and to the
marginal life of the criminalized addict.[5] Meanwhile, of course, huge
numbers of people continue to be arrested and imprisoned for drug offenses,
the most specific expression of a policy based on prohibition and a
punitive approach to drug users.

Yet despite constant appeals for more and better drug treatment, we still
see severe shortages in treatment programs [1] as well as limited success
in dealing with the severest forms of addiction, that is, to heroin and
cocaine. There is new and important Federal support for Methadone [6] (the
drug treatment of greatest proven efficacy for heroin addiction [7] ), but
public opinion remains sharply divided on the use of narcotic maintenance
with New York's Mayor Guiliani recently calling it "enslavement" and taking
steps to end treatment for thousands of patients currently under care in
the city.[8]

Further, while AIDS has refocused our attention on drugs as a public health
problem, raising the stakes for epidemiologic research and demanding
effective interventions to reduce the spread of HIV infection, even massive
international documentation of the effectiveness of needle exchange
programs has failed to shift a hostile

Federal policy that bans funding for such programs because they give the
"wrong message," that is, something other than "zero tolerance."[9]

What then are our goals in drug policy? And what should they be?

If "winning the war on drugs" was once the battle anthem of national drug
policy, that metaphor is now rejected by many, including Gen. Barry R.
McCaffery, Director of the White House Office of National Drug Control
Policy (ONDCP), as fostering "unrealistic expectations for a speedy victory
and a specific end to the campaign."[10] The General now believes the fight
against cancer to be a better analogy--"stressing prevention and
treatment."[10]

Notwithstanding this more health-oriented view and the growth in Federal
support for treatment programs, prohibition remains the major strategic
goal of our national drug policy, under which treatment continues to be
"backed up by a high level of social and legal disapproval" 10 and the
strict enforcement of drug laws. This is most evident in the allocation of
expenditures in the National Drug Control Budget for fiscal year 1998. Of a
$16 billion total, more than $10.7 billion (67%) was devoted to drug law
enforcement, interdiction, and supply reduction in the US and abroad.[1] In
addition to representing the lion's share of current Federal funding,
enforcement expenditures have shown almost two decades of steady
growth--increasing tenfold since 1981. [1] (See Figure 1.) In the same
period, Federal support for treatment and prevention has grown by only half
that amount.[11]

Even the recent innovation of drug courts, which steer arrested nonviolent
users to treatment, represents an extension of Federal enforcement policy
and funding priorities. This approach is still based on the continued
vigorous prosecution of drug users, while using the criminal justice system
to enforce compulsory treatment. Further, Federal budgets reflect only a
small part of all public expenditure for drug control. In this country,
most law enforcement occurs at the municipal and state levels, where annual
enforcement expenses are estimated at more than $20 billion,12 compared
with approximately $7.6 billion for treatment from all government and
private sources. [13]

Thus, as we follow the money for the past 25 years, it is clear that
enforcement has been the centerpiece of our drug policy, far outstripping
other approaches to the problem. The consequences of disproportionate
spending for enforcement are most visible in our society in the high rates
of arrest and incarceration for drug offenses [14] (Figure 2), the
increasing proportion of criminal justice activities devoted to drug
offenses, and the rise in both over the past 25 years.

While overall crime rates today are at their lowest in the past 25 years,
arrests for drug law violations have reached a record high--more than 1.5
million in 1996, the latest year for which complete data are available.[14]
State and Federal prisons and local jails today hold more than 400,000 drug
law violators--60% of all Federal prisoners and more than 25% of state and
local inmates.[14] (See Figure 2.)

Although rates of drug use were already declining rapidly by 1980, between
1980 and 1990 there was a 1055% increase in new commitments to state
prisons for drug offenses (from 8800 to 101,600).[15] New commitments
continued to rise into the 1990s (Table 1).

In 1980 there were 51,950 drug law violators behind bars in state and
Federal prisons (8% of all inmates). By 1995 this number had increased more
than 700% to 388,000 (25% of all inmates in a prison population now four
times as large). This growth represents the clearest expression of a policy
based on prohibition and the vigorous application of criminal sanctions for
the use and sale of illicit drugs.

The surge in incarcerated populations in the 1980s was due to harsher
enforcement policies and longer mandatory sentences for possession of
smaller quantities of drugs, including disproportionate penalties for
possession of crack cocaine. This resulted in progressively longer prison
terms for drug offenses and a widening gap in sentence length between drug
offenders and those convicted of violent crimes [16] --which has helped
increase the proportion of the prison population behind bars for drug
offenses (Figure 2).

And while some individuals are in prison for major trafficking offenses or
violent crimes, more than 90% of drug offenders are arrested for possession
or for low-level drug deals to support their personal use.[16]

It is clear from these data that we have practiced what we preach,
literally with a vengeance. There are more drug offenders behind bars today
than the total incarcerated population of 1970. [17] Indeed, drug
enforcement has accounted for such a large increase in our prison
population that the US is now the Western democracy with the highest per
capita rate of imprisonment.18 What have been the effects on the patterns
of drug use of this vast natural experiment in drug control policy?

Proponents of a drug policy based on prohibition and its rigorous
enforcement claim that their approach is working. See, for example, Figure
3, reprinted here from the ONDCP's 1998 National Drug Control Strategy,[1]
which is used to support this contention. It shows that self-reported past
month use of any illicit (that is, illegal) drug, and specifically of
cocaine and marijuana, have declined sharply since 1985.

While Federal drug control officials admit that the problem is still
serious, costing at least 14,000 lives and $110 billion a year,1 they
assert that our approach has increased societal disapproval of drug use and
lessened the extent and severity of the drug problem.

Citing reductions in "casual use" of all illegal drugs by 50% (and of
cocaine by 75%) since 1979, [1] in its 1998 National Drug Control Strategy,
the ONDCP claims that we will do even better in the future and sets a new
10- year goal of a 50% reduction in overall drug use in America, to a level
below the lowest point attained in the last 30 years.[1]

These claims are greeted with some skepticism given the growing world
market in illicit drugs. We are seeing greater availability of higher
purity drugs at lower prices; from 1981 to 1996 the average price per pure
gram of cocaine fell by 66% and the average purity of street heroin rose
from 6.7% to 41.5%.[1]

Increased crop acreage and expanded international traffic have driven a
steady rise in the number of consumer and producer nations to at least 140
countries and a $500 billion world market, as has been well documented by
the ONDCP, the US Drug Enforcement Agency, Interpol, and the United Nations
Drug Control Program.[1]

In a world awash in drugs, with widespread economic hardship and social
dislocation to motivate their continued production and distribution, can we
succeed in protecting our nation from drugs and their dangers by the
application of our current policies?

Apparently not.

Despite reductions in adult use, the latest data from national surveys [19]
show a sharp climb since 1991 in the prevalence of illicit drug use among
American high school students--despite decades of intense enforcement and
powerful anti-drug messages. (See Figure 4.) This primarily reflects
increased use of marijuana, but use of the harder drugs also appears on the
increase.[19] These climbing rates of teen use are a sentinel for the
failure of our current policies to reduce the number of new users of
prohibited drugs. And, interestingly, they are echoed in teen use of legal
drugs—tobacco (despite the anti-tobacco crusades of the last few years) and
alcohol--neither of which may be legally sold to people in this age group.[19]

Are there other ways in which our drug policies are failing us? What do the
data show?

EVALUATING ALL AVAILABLE EVIDENCE

Fortunately, in this country, we are in a position to evaluate the
long-term relationship between drug policy and drug use by examining in
detail some of the public health consequences of that policy. We have more
than 25 years of information on changes in patterns of drug use in the US
population and may hold these up alongside data on the use of criminal
penalties, identifying long-term trends and health and social outcomes.

Sources Of Data On Drug Use.

The United States has the best funded, largest scale, longest functioning,
and methodologically most consistent drug use surveillance and data
monitoring system in the world. There are three major sources of national
survey data on drug use in the United States:

(a) The National Household Survey on Drug Abuse (NHSDA), conducted by the
Federal government since 1973, measures the prevalence of drug and alcohol
use among the US household population ages 12 years and older; expanded in
1991 to include college students, homeless shelters, and the military.

(b) Monitoring the Future (MTF), conducted for the National Institute on
Drug Abuse by the University of Michigan; surveys high school seniors
(since 1972), and 8th through 12th graders (since 1982).

(c) The Drug Abuse Warning Network (DAWN), a data collection program of the
Substance Abuse and Mental Health Services Administration (SAMHSA), in
place since 1972; annually samples more than 400 hospital emergency
departments (ERs), reporting on ER visits in which both legal and illegal
drugs are implicated, and also tallies medical examiner reports of deaths
in which drugs and alcohol are implicated.

Each of these surveys and the data they report have limitations: the
household survey (NHSDA) underrepresents the homeless, and the survey of
high school seniors (MTF) misses school dropouts, both groups with higher
than average rates of drug use (for example, school dropouts are reported
to have two to four times the rate of cocaine use of non-dropouts [1] ).
And DAWN does not capture all hospital ERs. Another limitation, of course,
is that given public law and private sentiment, one would expect a certain
amount of under-reporting of personal drug use to researchers. This is
probably most true for heroin, for which some Federal studies warn of
substantial underreporting.[1] For these reasons, "harder" data on measures
of drug-related morbidity and mortality, which are less dependent on
self-report and more public than use per se, should be closely watched,
recognizing that these reflect the adverse consequences of drug use and not
simply its prevalence.

But, despite these short-comings, data from large, ongoing, national
surveys are very useful because they are consistent in their limitations
and biases and allow us to create a reliable comparative picture of
patterns and time trends in the prevalence of drug use over the past 25
years. They also permit us to see the demographic profile of drug users and
to identify changes in this population over time.

TRENDS IN POPULATION PREVALENCE, 1972-1977

Data on the prevalence of drug use are available by year for the major
social and demographic categories (age, sex, "race") and for each of the
illicit drugs (as well as for tobacco and alcohol use). The NHSDA collects
data on use in the respondent's lifetime ("ever used"), in the past year,
and in the past month ("current use").

NHSDA household survey data show that in 1997, 36% of the adult population
ages 12 years and older reported some illicit drug use in their lifetimes,
but that number dropped to 11% for use in the past year and 6% for the past
month [20] --ratios that have not changed significantly in the national
data in a generation despite changes in prevalence.[21] These data show
that most illicit drug users are not "hard core" addicts and that most
experimental or casual use does not eventuate in continued or regular use.

From a public health perspective, past-month use is the most appropriate
measure for looking at long-term changes in the prevalence of drug use
because it captures all "current" or regular users (including dependent
users) but only a small percentage of the much larger group who may have
used drugs a single time or who are experimental or casual users. Figure 4
shows the NHSDA prevalence data for US population ages 12–17 years for
past-month use of illicit drugs.

As most health risk is associated with regular exposure to the "major"
drugs--cocaine, heroin, stimulants, depressants, and hallucinogens,[22] it
is useful to focus attention on the long-term trends in past month use of
these drugs independently from trends for marijuana, which has consistently
shown a higher prevalence since data collection began in the 1970s than all
other illicit drugs combined.

Unlike the data beginning in the mid-1980s that are presented to support
the claim that our policies are working to reduce the prevalence of drug
use (see, for example, Figure 3), these more complete and specific data on
time trends make clear that the prevalence of drug use in the US has
followed no simple course over the past 25 years. Use of the "major"
illicit drugs rose in the early 1970s from a 1960s level estimated at less
than 2% of the adult population ages 12 and older,[21] peaked at about 6%
in 1985, and declined until 1992, when it started to rise again among teens
(although the 1990s average was still only 2.3% of the adult population)
(Figure 4).

Trends In The Use Of Specific Drugs.

While overall population trends in the use of any illegal drug are
informative, individuals use specific drugs. Figure 6 shows 1979-1996
trends for each of the most commonly used illegal drugs. It is immediately
apparent from this Figure that prevalence levels for the various drugs are
markedly different and that each drug exhibits a different trajectory of
use over the years.

Marijuana dominates the picture, accounting for over 93% of all reported
use of illicit drugs-- more than all other illicit drugs combined. Past-22
month marijuana use reached a peak of 13.2% of the adult (greater than 12
years) population in 1979 and declined until 1993, when it began to climb
again--although to only a fraction of its former level, reaching 4.7% by 1996.

Cocaine use rose most sharply exactly as marijuana use was declining,
peaking at 4.6% of the adult (greater than 12 years) population in 1988 but
declining to the 0.7% to 1% range for 1990-1998. The NHSDA reported heroin
use to be relatively stable, at less than 0.1%, throughout the years from
1972 through 1979.

(Heroin use is particularly covert and subject to rapid local changes in
availability and use, changes not well captured in the household survey
method, and the NHSDA does not claim great accuracy or reliability for its
heroin data.)

While there are no more reliable surveys than the NHSDA from which to
document national levels of the use of heroin, the ONDCP has estimated
(relying on local field studies and modeling techniques) that there are
810,000 chronic users of heroin in the US,1 0.3% of people ages 12 years
and older. According to the ONDCP, this group now includes more younger
(new) users, among whom there is clear evidence of a shift away from
injecting to sniffing, an important change for AIDS risk but one that does
not necessarily make the drug safer.[1] Do these trends in the prevalence
of drug use bear any relationship to the steady rise in enforcement that we
have seen over the same time period? The details about who uses drugs (and
who does not) provide important clues to this relationship.

WHO USES ILLICIT DRUGS?

While the overall prevalence of drug use and the drugs of choice may have
changed over time, the characteristics of the populations using these drugs
has been more stable.

Figure 7 shows the demographics of the population using illegal drugs for
selected years from 1979 through 1997.

Gender. From Figure 7, we can see that over this 19-year period, male use
regularly outstripped female use by about 2:1 and both showed proportionate
rises and declines as overall prevalence changed over time.

Age. Initiation of the use of all drugs, both legal and prohibited, is
principally an event of adolescence, especially ages 12 through 17. But the
18-25 age group, the group most at risk for criminal activity, arrest, and
imprisonment, [16] consistently has the highest prevalence of use. We see
lower rates of use as individuals "age out" of the lifestyles and social
networks in which they used drugs; however, the increase in youthful drug
use in the 1990s created new cohorts, some of whom will continue use as
adults. So, for the present, we see a shift in the age mix of the
drug-using population in the direction of youth. For example, in 1979, only
21% of current drug users in the 12-34 age category were younger than 18
years of age, but by 1997 that proportion was 33%, albeit of a total
population of users half the size.

"Racial" category. A common stereotype, fostered by the media, is that some
"racial" or ethnic groups use drugs more than others. This is not borne out
by the data. There are only small differences across "racial" categories in
the prevalence of illegal drug use. And the declines in drug use seen from
1979 through 1997 are reflected in all groups. Some small age- and
drug-specific differences by "racial" category appear over this 19-year
period--for example, marijuana and amphetamine use has been heavier among
whites, and cocaine use somewhat higher among blacks.

But these differences are neither large nor consistent, and the recent
trend of rising use in the 12–17 age group reflects virtually identical
increases in the prevalence rates for all "racial" categories.[20]

While the prevalence of drug use is an important measure of changing trends
over time, from a public health perspective we are most concerned with
health effects, seen in morbidity and mortality related to drug use. How do
trends in these adverse outcomes correspond to the substantial changes we
have seen in both enforcement and prevalence over this 25-year period? To
answer, we turn to the data from the Drug Abuse Warning Network (DAWN).

(Continued in part 2)