How will we know a wise and workable drug policy when we see one? How do we know that legalization will be betteror worsethan our current policy? To know what is likely to happen when and if the currently illegal drugs are legalized, it is necessary to know something about the extent of the use of the relevant drugs. How harmful are they? To know this, we have to have a picture of how frequently they are used, and by how many people. How many users will the change in legal policy affect? Again, to answer this question, we have to have an overview of the extent of drug use in America. Will legalization result in a change in patterns and frequencies of drug use? To address this question, we need baseline comparisons against which to compare what the new use rates and frequencies could be. Specifically, we have to know about the extent or prevalence of the use of the most commonly taken drugs, their changes over time, and crucial correlates or accompaniments of the various levels of use. Then, perhaps, we can address the question of whether our current policy is making these things worse or keeping them under control. The drug legalization debate is practically meaningless without a fairly accurate snapshot portrait of current patterns of use.
How many experimenters, users, abusers, and addicts of each drug are there in the United Statesand how do we know? To answer these questions, it is necessary to stress two points. First of all, when estimating drug use in the general population, or major segments of it, we need systematic information. That means that it is not permissible to rely exclusively on examples, cases, anecdotes, or small, unrepresentative samples; examples are illustrative, not definitive. We need a cross-sectional view: evidence that is drawn from and which reflects the society as a whole, that gives us a picture of what Americans generally are doing. And second, to have confidence in the data we do have, we need a triangulation of sources. As a general rule, the greater the number of independent sources of information that reach the same conclusion, the more confidence we can have in that conclusion. This is what we mean by triangulation: getting a factual fix on reality by using several separate and disparate sources of information To the extent that several independent data sources say the same thing, we can say that their conclusions are more likely to be true or valid.
STUDYING DRUG USE
The data we have on legal drugs is a bit different from the information that comes to us on illegal drugs. In principle at least, the sale of all legal products is recorded for tax purposes; we know how many bottles of alcoholic beverages and cartons of cigarettes are sold each year in the United States; we also know how many prescriptions are written annually for each pharmaceutical drug. It is true that a certain proportion of these legal drugs are purchased by private parties abroad and brought back to the United States; hence, they are not included in the sales total even though they may be consumed here. Quantities of some legal drugsfor instance, bottles of alcohol purchased as a gift for someone elsemay be placed on a shelf and never consumed; and some of what was consumed one year was purchased during the previous year. But nearly all experts agree that taking these unrecorded sources of legal drugs into account does not add up to a major change in the picture basically, we can take the sales of legal drugs in a given year as synonymous with the use of these drugs. Thus, if want to know about the level of alcohol, tobacco, and prescription drug consumption in the United States, and whether they have increased or decreased over time, we need only look at the sales figures for these three legal products. However, in order to get a closer look at legal drug consumptionfor instance, what segments or categories in the population use these legal drugswe must rely on surveys. Do men drink more than women? Which racial and ethnic categories smoke tobacco cigarettes the most? The least? Which age categories are most likely to use which prescription drugs? Again, surveys help us answer questions such as these.
For a picture of the extent and frequency of illegal drug use and who uses which drugs, experts have to rely mainly or exclusively on surveys. Common sense tells us that most people lie about their participation in illegal or criminal activities but, in fact, common sense is at least partly wrong. The fact is, if respondents are assured of confidentiality, that is, that their names will not be used, that they cannot be tracked down after the survey, and that they will not get into any trouble as a result of telling researchers the truth, most give fairly honest answersto the best of their ability. Some will lie, of course, and some do not give accurate answers because of problems in recall and interpretation, but most answers will be reasonably honest and accurate. How do we know? Here is where triangulation comes in. We compare the answers respondents give in surveys with independent and fairly "hard" information, such as hospital or medical records, arrest or prison records, and blood or other objective tests; most of the time, the two sources of data correspond fairly well, although far from perfectly (Chaiken and Chaiken, 1982; Hindelang, Hirschi, and Weis, 1981; Johnson et al., 1985, p.23). To put the matter another way, the discrepancies tend to be only moderate rather than huge. The figures we get in such surveys are close enough to be useful.
Most researchers feel that the problem of sampling the population is somewhat more serious than the problem of getting truthful answers. Sampling difficulties come into play in an especially problematic way when we study illegal behavior, because the segments of the population that are least likely to be included in a researcher's sample and be surveyed because they are difficult to locate are also most likely to engage in the behavior in which we are interested. For instance, the homeless do not appear in a household survey because they do not live in a household; the same is true of jail and prison inmates (in the United States today, nearly 1.5 million on any given day). In addition, high school dropouts do not appear in surveys on the illegal activities of high school students. It is almost certain that the homeless, the incarcerated, and dropouts are more likely to use, or to have used, psychoactive drugs, than Americans who live in households, are not incarcerated, and/or are attending or have completed high school. Sampling problems usually represent a more formidable challenge to the researcher than problems of responder truthfulnessbut they can be overcome. In addition to the problem of getting a sample that looks like or "represents" the population as a whole, researchers face the problem of interviewer refusal; even where researchers are very clever and persistent, some 20 percent or so of the sample refuses to be interviewed. Are those who refuse distinctly different in important ways from people who agree to be interviewed? At times, refusals may bias our results, although usually not fatally.
The federal government sponsors two surveys of drug use which are based on very large, nationally representative samples. One is conducted every year; high school seniors, secondary school, and college students and noncollege young adults are surveyed about their legal (alcohol and tobacco) and illegal drug use, attitudes toward drug use and drug legalization, whether they think that drug use is harmful, and their perceived availability of illegal drugs. Roughly 15,000 to 20,000 respondents are contacted in each category, that is, high school seniors, secondary school students, and young adults, either noncollege or in college. This survey is known as the "Monitoring the Future" study. (It used to be referred to as the "High School Senior Study," since, originally, it focused its drug questions exclusively on high school seniors.) Another federally sponsored survey asks questions of the residents of households a nationally representative sample of households; it is called the National Household Survey on Drug Abuse. Its sample is quite large; it is made up of roughly 30,000 persons age 12 and older who are members of the selected households.
One crucial warning is in order. Looking at the number of users of the various psychoactive drugs and their frequency of use is one crucial piece of the drug puzzle. It's possible, however, that there is a much more important piece of the puzzle: society's addicts. It can be argued that the occasional or less-than-weekly drug user rarely poses a comparably serious threat of harm to the society. It is entirely possible that the public health issue, as well as the issue of any and all potential dangers that drug use poses to the society, is more or less entirely confined to the heavy or chronic user. The less-than-weekly user of even heroin and cocaine is much less likely to rob to support a drug habit; die of an overdose; contract a serious, life-threatening, drug-induced disease; or kill someone in connection with drug use. It could be that the harms we associate with drug use and abuse are mostly confined to that segment of users we refer to as the behaviorally dependentin a word, to "addicts." And yet, as we know, addicts tend to be far more difficult to locate and study by means of a survey than occasional users. We'll look at society's addicts and drug abusers more or less throughout this book. The fact is, however, we need to look at both the full range of usefrom the experimenter to the addictas well as focus more intensely on the upper end of use: the heavy, chronic, repetitive addict who takes drugs to the point of selfharm and creates major problems for the society. Both are relevant to the question of legalization, but what happens at the upper end of the use spectrum is especially important: The addict is capable of creating far more problems for the rest of us than legal policy can either alleviate or worsen. For this reason, we need to pay special attention to the relatively atypical heavy user. And this denizen is very unlikely to be captured by the conventional sample survey; we need to study him or her through more imaginative research methods. At the same time, we cannot ignore the more typical and common casual user, either. Knowing what the average or typical levels of use are is an important piece of information; among other things, it provides us with a basis of comparison with the abuser. Hence, use in general is a good place to start. What do our surveys tell us about drug use in the population?
Alcohol is the drug that is consumed by the greatest number of usersand by a considerable margin. Roughly two-thirds of the American population age 12 and older (in 1994, this figure was 67 percent) say that they have used alcohol once or more in the past year; 54 percent did so in the past month; and in 1993, just over one in five (21.5 percent) say that they drank once a week or more during the past year (HHS, 1994a, p.119; 1995a, p.85). The "Monitoring the Future" study of secondary, high school, and college students and young adults also shows high levels of alcohol use. Nearly half of eighth-graders (47 percent for 1994) had consumed alcohol in the past year, and nearly a fifth (18 percent) admitted having being drunk at least once during that period of time. Half of high school seniors (50 percent) said that they had drunk alcohol in the past month; over seven out of 10 college students (72 percent) and noncollege young adults (70 percent) had done so (Johnston, O'Malley, and Bachman, 1994, pp.85, 162; 1995, p.43).
Sales of alcohol average out to roughly 2.3 gallons of absolute alcohol per person for the population age 18 or older per year, or just under one ounce per person per day (Williams, Clem, and Dufor, 1994, p.15). This means that the American population as a whole consumes about 60 to 70 billion "doses" of alcohol per year. (Keep in mind that distilled beverages are 40 to 50 percent alcohol, wine is 12 percent, and beer is about 4 percent; thus, how much alcohol is consumed in a given quantity of a beverage has to be calculated from its potency.) However, there is great variation from one person to another in the amount of alcohol consumed. There is a kind of polarization in use: While one-third of the American population is made up of abstainers, and over half are moderate or "social" drinkers, that very small one-tenth of the population which is made up of the heaviest drinkers imbibes more than half the total alcohol consumed. Thus, the category "drinker" or alcohol "user" represents an extremely mixed bag. It should be emphasized that the concept "alcoholic" is extremely controversial; different experts define it radically differently, and the field cannot agree on how many alcoholics there are in the population (Hilton, 1989). However, taking as our handy working definition of addiction the use of a psychoactive substance on a frequent, repetitive, and compulsive basis to the point of physical or psychological dependence, one researcher estimated that there are between 10 and 15 million alcohol addicts in the United States today (Goldstein, 1994, pp.7, 263).
Far fewer Americans smoke tobacco cigarettes than drink alcohol; slightly more than a quarter of the adult respondents questioned in recent surveys (28.6 percent) smoked a cigarette in the past month (HHS, 1995a, p.89). However, while fewer people smoke than drink, cigarettes are used vastly more often than alcohol. Though three-quarters of the American population abstain completely from cigarettes, those who do smoke consume far, far more "doses" of their drug than drinkers consume of theirs. A typical drinker will have one or two drinks during an evening of drinking, and will do so once a week or once or twice a month, whereas a typical smoker will smoke 20, 30, or more cigarettes a day. During the early l990s, roughly 500 billion cigarettes (or "doses") were purchased in the United States each year, nearly eight times as many as drinks of alcoholic beverages. (Is it a bit arbitrary to refer to one cigarette as equivalent to a one-ounce drink of alcohol? Sure it is, but what "dose" of tobacco would be equivalent?) Thus, while alcohol is the drug that is taken by the greatest number of people, during a given stretch of time, tobacco is the drug that is used the greatest number of times. If drug dependence (or "addiction"), as with alcohol, is measured by repetitive and compulsive use of a psychoactive substance (Goldstein, 1994, p.3), there are far more persons dependent on or addicted to tobaccoroughly 50 million in the United States (p.263)than all other drugs combined! Let's put it another way: While most drinkers use alcohol in moderation, almost all smokers are addicted to tobacco.
Since the sale of the prescription drugs, like the sale of alcohol and tobacco products, is legal, it is tabulated fairly accurately. In the United States each year, roughly two billion prescriptions, each authorizing the purchase of 30 or so doses (tablets, capsules, or "pills"), are written. A minority of all prescription drugs, roughly one out of seven, are psychoactive, even by an extremely broad definition; the rest work more or less exclusively on the body and do not affect the mind. Eliminating the nonpsychoactive drugs from the total, we see that roughly eight or nine billion doses of psychoactive prescription drugs are administered in the United States each year, or less than one tablet or capsule of a psychoactive prescription drug per adult American per week. This sum is a bit more than one-tenth the total for alcohol, and between 1 and 2 percent of that for cigarettes (again, taking each cigarette, each one-ounce alcoholic drink, and one tablet or capsule as a "dose"). Even so, the sale of prescription drugs represents a substantial, even hugemore than $40 billion a yearindustry in America (Ray and Ksir, 1997). And certainly in terms of absolute numbers, many Americans take many doses of prescription drugs. Nonetheless, relative to these two other legal drugsalcohol and tobaccoprescription drugs do not make up a hugely prodigious source or category of psychoactive drug use. (Of course, some drugs that are dispensed via prescription are also manufactured illegally, but, in this case, we would be discussing illegal, not prescription, drug use.) With respect to number of users and total volume of use, the use of psychoactive pharmaceuticals is much more comparable to the use of some of the more popular illegal drugs than to legal use.
Moreover, for most psychoactive prescription drugs, use in the United States is declining over time. There are exceptions, of course. For instance, the sales of Prozac, introduced in 1987, increased five times from its first full year of sale to its third; in 1994, Prozac was the nation's ninth most frequently prescribed pharmaceutical and its most popular antidepressant, by far. But for the vast majority of the psychoactive drugs that were widely prescribed in the 1960s and 1970sespecially those whose use leaked out into the illegal, recreational arenasales have plummeted. The number of prescriptions written in the l990s for barbiturates is one-twentieth of that written in the 1960s; for the amphetamines, it is less than one-tenth. In the early to mid- 1970s, Valium was the nation's number one prescription drug. Today, it posts sales less than one-quarter of its peak; in 1994, Valium ranked 145th in sales among all prescription drugs. (IMS America, of Ambler, Pennsylvania, tabulates the sales of prescription drugs each year. In addition, in the April issue of the journal Pharmacy Times, the nation's 200 top-selling prescription drugs are ranked; these 200 drugs make up slightly more than half of all prescription drug sales.) Some prescription drugs that were extremely popular 20 years ago are no longer being prescribed at all, such as methaqualone (Quaalude) and Benzedrine, an amphetamine. (Consider this, however: While certain prescription drugs have declined sharply in popularity, with some, substitutes that did not exist a generation ago are now being sold. Tabulating sales over time for an entire drug type is not a simple proposition.) A generation or more ago, prescription use represented a far more substantial source of all psychoactive drug use; today, although it cannot be dismissed as inconsequential, prescription drug use does not rival that of our two legal drugs, alcohol and tobacco.
An extremely important measure of drug use is something known as the continuance rate. This is the "loyalty" of users to each drug: Of all the people who have ever taken a given drug, even once, what percentage is still using it? What percentage "sticks with" a given drug? For some drugs, the "continuance" rate is high, while for other drugs, most of the people who have used it once or more have given it up or use it extremely infrequently. Of all drugs, the "continuance" rate for alcohol is the highest roughly six out of 10 people who have ever taken at least one drink of alcohol in their lifetime are still using it, that is, have taken a drink in the past month. Put another way, only four out of 10 of all the people who have taken at least one alcoholic drink were abstainers from alcohol during the past month. With cigarettes, in contrast, the "continuance" rate is roughly four out of 10: Over six out of 10 of all persons who have smoked at least one cigarette are no longer smoking (HHS, 1995a, p.91). The issue of the continuance rates of different drugs is so important that we'll encounter it several more times down the road.
Illegal drugs are used far, far less often by far, far fewer people than our two legal drugs. Roughly one-third of the American population age 12 and older (31 percent) say that they used marijuana at least once during their lifetime; fewer than one in 10 (8.5 percent) say that they did so during the past year, however; and only one out of 20 (4.8 percent) says that he or she did so in the past month (HHS, 1995a, p.23). The "Monitoring the Future" study shows essentially the same picture: While marijuana is, by far, the most widely used illegal drug, its use is far lower than that of alcohol or even tobacco. Just over one in eight (13 percent) claimed to have used marijuana in the past year; one in 13 (7.8 percent) used it in the past month. These figures increase as the person grows older: Three high school seniors in 10 (31 percent), and a bit more than one out of four college students (28 percent) and noncollege young adults (25 percent), used marijuana in the past year; and between one-fifth and one-seventh (19, 14, and 13 percent, respectively) did so in the past month (Johnston, O'Malley, and Bachman, 1994, p.6; 1995, p.59).
Although these figures are much smaller than those for alcohol, the percentages for the illegal drugs other than marijuana are, in turn, much smaller than those for marijuana. For instance, cocaine, the second most popular illegal drug in the general population, is used by roughly one-fifth as many Americans as marijuana. Fewer than 2 percent of Americans age 12 and older (1.7 percent, to be exact) say that they have used cocaine even once in the past year; less than 1 percent (0.7 percent) say that they did so a dozen times or more (HHS, 1995a, p.31). In the "Monitoring the Future" study, 2 percent of eighth-graders, 3 percent of high school seniors, 3 percent of college students, and 5 percent of noncollege young adults used cocaine at least once during the past year; the comparable figures for use in the past month was 1 percent across the board (Johnston, O'Malley, and Bachman, 1994, pp.84, 85; 1995, pp.60, 62).
In addition, only about 1 percent or less of the National Household Survey said that they used one or more hallucinogens, such as LSD (1.3 percent); a stimulant, such as amphetamine (0.7 percent); a sedative, such as a barbiturate (0.4 percent); a tranquilizer, such as Valium (1.1 percent); or an analgesic, such as Darvon (2 percent) nonmedically at least once in the past year (HHS, 1995a, pp.49, 61, 67, 73, 79). The annual figures for eighth-graders to college students and young adults which are generated by the "Monitoring the Future" study are several times the national totals. For instance, 3 percent of high school seniors and 6 percent of college students say that they took a hallucinogen during the past year; for the stimulants, the comparable figures are 4 percent for both groups (Johnston, O'Malley, and Bachman, 1994, p.l61; 1995, p.62). Heroin is perhaps the least often used of the well-known illegal drugs. These surveys specifically ask about "nonmedical" drug use. When prescription drugs such as amphetamine or barbiturates are used for recreational purposes, of course, they are illegal; in addition, most recreational use of such drugs stems from illicit or illegal manufacture as well. However, let's keep in mind the three warnings I issued earlier about relying on surveys to estimate illicit drug use in the population: One, some respondents (a minority) lie, forget, or even exaggerate; two, some of the target population (a minority) refuse to participate in the study; and three, some categories (such as the homeless, the institutionalized, and school dropouts) do not appear at all in most surveys like thisand, chances are, they are more likely to use drugs, and more likely to do so heavily, than the sample we draw and interview.
In addition to vastly lower levels of use for the illegal drugs than for the legal, it is also true that the continuance rates of the illegal drugs are far lower. Persons who try or experiment with the illegal drugs are much more likely to give them up or use them extremely infrequently than is true of the legal drugs. Remember that about 60 percent of all persons who have ever imbibed an alcoholic beverage in their lifetime are still drinking, that is, they drank at least once within the past month; fewer than four out of 10 of all persons who have ever had at least one puff of a cigarette are still smoking. For marijuana, the comparable figure is 15 percent; for cocaine it is 7 percent; and for hallucinogens such as LSD, it is less than 6 percent (HHS, 1995a, pp.25, 31, 47). As a general rule, the continuance rates for illegal drugs are significantly lower than they are for legal drugs. This is true in the United States, and it is true in the Netherlands as well (Sandwijk, Cohen, and Musterd, 1991, p.25).
Even though most experts regard these survey data on casual or recreational illegal drug use as fairly accurate and reliable, our estimates of heavy, chronic users are far less reliable. In fact, we have to rely on what is referred to as "guesstimates," based on information sources such as hospital admissions, admissions to drug treatment programs, surveys of street-corner addict populations, anthropological-style observation, and so on. A commonly cited figure for the number of heroin or opiate addicts in the United States is between 500,000 and one million (Goldstein, 1994, p.241); for persons using cocaine weekly or more, it is between two and three million (Kleiman, 1992b, p.288). It should be pointed out, however, that there are different definitions of what defines someone as dependent or addicted; they are not always in complete agreement with one another. Most experts see the use of marijuana and the hallucinogens as less of a public health problem than the use of heroin, cocaine, and amphetamine. For all practical purposes, alcohol, cocaine, heroin, and nicotine constitute the bulk of the public health issue when it comes to psychoactive drugs.
Chapter 3. Drug Abuse: Definitions, Indicators, and Causes