THE "SCIENTIFIC" JUSTIFICATION FOR URINE DRUG TESTING

John Morgan, Medical professor and acting Chairman of Pharmacology, City University of New York Medical School

KANSAS LAW REVIEW, 1988, Vol. 36, pp. 683-697

INTRODUCTION

Just as drug use is often ceremonial and ritualistic, discussions about those problems have a communal, worshipful character. Individuals have to speak out of strong conviction, and congregational beliefs are strengthened in a confirmation of faith and shared attitudes. In recent years, the experiences described above have been characterized by an important cohesive element, a shared litany.

Since 1984, when urine testing of unimpaired civilian workers began to expand rapidly, its public supporters have used a common set of statements, ideas, and statistics to justify the imposition of the forced detection and testing of urine for illegal drugs. The use of this religious language is not without a purpose. Litany is seldom questioned, nor is the speaker of litany often called upon to prove the truth of his statements, even when they involve things material and not spiritual. I believe this common set of statements used by those justifying and selling urine tests is repeated, polished, and accepted with a minimal offering of proof. In this article, I will identify these statements, search for their sources, and examine their accuracy.

II. THE FIRESTONE "STUDY"

Obviously, statements concerning the job performance of drug users have an impact on their employers. From the earliest days of the debate, a series of statistics about drug-using workers, their poor productivity, and high cost to ther company have been presented. In one of the first proceedings of a meeting directed to the issues of drug abuse in the workplace, Robert Angarola and Judith Brunton stated:

A recent study....suggests that drug users were almost four times as likely to be involved in a plant accident and were two and one-half times as likely to be absent from work for more than a week than employees who did not use drugs. Drug users were five times as likely to file workers' compensation claims and they received three times the average level of sick benefits. (1) Further, the drug users were "repeatedly involved in grievance procedures." The Angarola-Brunton document was published in 1984 but they and others had begun making the remarks frequently by then

In 1987 testimony before the U.S. House of Representatives, by Mark de Bernardo of the United States Chamber of Commerce:

'recreational' drug users are

-----2.2 times more likely to request early dismissal or time off...

--3.6 times more likely to injure themselves or another person in a workplace accident

--5.0 times more likely to file a worker's compensation claim. (3)

For at least five years, these statistics and others like them have been voiced and published by testing advocates, although seldom with de Bernardo's decimal-point precision. Those whom I have quoted seem to be referring to any user of illegal drugs and de Bernardo specifically speaks of recreational drug user. Yet, one might wonder how these data were gathered, particularly since drug testing in the workplace is proposed to uncover the secret user. How were these users described by Angarola and de Bernardo (and many others) found and analyzed? The answer, as is often the case is simple. They were not.

The most important source for dissemination of these statistics seems to be the Drug Abuse and Alcoholism Newsletter of Dr. Sidney Cohen. (4) This monthly newsletter is widely distributed and has been collected and published. In the August 1983 newsletter, Cohen discussed a study by the Firestone Tire and Rubber Company and constructed the quotable statements which have appeared and reappeared: "Drug users were five times as likely to file a Workers' Compensation claim, and they received three times the average level of sick benefits." (5)

The Firestone "Study" has never been published. Thus, Cohen's newsletter seems to be the source of all such statements. Yet, Cohen gave no attribution nor sources, and when I wrote to him in 1985 he answered that he did not have a copy of the study and suggested I write Firestone. After a number of calls and queries I received a two page document from Firestone's Medical Director, E. Gates Morgan. The report apears to be an in-house newsletter. (6) In it, a Mr. Ed Johnson is interviewed about the Employer Assistance Program ("EAP") at Firestone. There are some statements pertaining to absenteeism, but these are not documented, and more importantly, refer only to a few alcoholics who have been served by the Firestone EAP. The statistics generated (if these calculations based on alcoholics were actually made) have nothing to do with drug users, recreational or otherwise.

The statistics cited about absenteeism and workers' compensation claims may have been derived from interviews with alcoholic workers enrolled in the EAP at Firestone. These people were not identified by urine testing for alcohol, but were referred because they or others perceived that their lives were falling apart. They, unlike workers randomly tested for drug use, were dysfunctional. To use them as a justification for testing unimpaired workers is like demanding that all workers have mandatory periodic rectal temperatures taken because a case of tuberculosis was found in the workplace.

III. THE RTI STUDY FOR DIMINISHED PRODUCTIVITY

The Research Triangle Institute ("RTI") of North Carolina issued a report in 1984 that attempted to calculate the economic cost to society of alcohol and drug abuse and mental illness. (7) It appears to be the source of the claim that drug abuse costs the United States thirty-four billion dollars a year in lost productivity. (8)

This claim seems to indicate that somehow individual worker productivity was assessed by some quantitative measure and was noted to be seriously diminished. I often ask my students how they would design a study which might quantify the effect of drug use or abuse on productivity and they generate interesting ideas and designs. None, however, have ever come up with anything as creative as the RTI report.

The National Institute on Drug Abuse ("NIDA") Household Survey is the actual source of the RTI statistics. (9) The RTI analysis itself does not include any studies about productivity, dor does it document that drug use was associated with any diminished productivity. The RTI report relied on the 1982 version of the Household Survey, which included a question to the adult (older than eighteen years of age) respondents about daily marijuana use. If the respondent stated that he or she had ever used marijuana daily for twenty or thirty days in one month, this respondent's household income was compared to another household without a daily user. The user households were estimated to have a household income approximately twenty-eight percent less than those households that did not have such an interviewee. According to the RTI report: "These result have been used to calculate the reduced productivity due to daily marijuana use. Once again, the prevalence rates of ever using marijuana daily in the labor force by age and sex groups....were applied to the number of persons in the labor force from those groups that experienced reducedproductivity. This number was multiplied by the expected productivity in the workplace and in the household and then by the impairment rate (27.9 percent) to produce the final value: The loss due to marijuana abuse was estimated at $34.2 billion for 1980." (10)

This somewhat opaque quotation means that the $34 billion "lost" to drug-induced diminished productivity and the one-third reduction is productivity of the drug user come from a very unusual calculation. The homes with daily users had a reduced household income. Because the daily users all were between eighteen and thirty-four years of age, the proportion of the work force represented by them was multiplied by total income expected and this was multiplied by the reduction figure to generate the $34 billion guess. I have never seen a published analysis of the RTI study. It is a complex and detailed document of one hundred thirty-seven pages whose appendices number nearly one hundred thirty more. Although many of its assumptions regarding the cost of alcohol and drug abuse are questionable, I have focused only on the loss of productivity formulation.

The reduced income figure itself may be incorrect. It relies upon reports from interviewees that they used drugs and relies upon their estimates of household income. The income data were grouped in seven broad categories with intervals as broad as $10,000. (11) To generate the reduction figure with categories like this amounts to measuring centimeters using a device graduated in meters. The control households were generated by selecting a household whose respondent resembled the marijuana use respondent. This meant that the two respondents resembled each other in regard to age, education, marital status, and occupation. (12) The similarities did not include race, location, or nonchildren resident(s) in the household. Therefore, a twenty-four year-old single black, high school graduate brick-layer who smokes marijuana and lives with his grandmother and uncle in Harlem, could have his family income compared to a twenty-four year-old single white, high school graduate, brick-layer living with his parents in an affluent New Jersey suburb. If the household in Harlem had a lower income, it would be attributed to diminished productivity caused by marijuana use.

As the RTI authors noted, however, the association between marijuana use and reduced income is not necessarily causal. (13) Although age and occupation were similar, the user may not have concerned himself with income and may have made other job choices. (14) Moreover, the heavy use of marijuana could have been secondary to job loss in a poor economy.

This manipulation, which assigned lost productivity to sometime past heavy use of marijuana, did not show any association with current use of marijuana (or any other drug), nor could it show any productivity decrease due to past use of any other drug. It also could not show any effect of any drug use in any person greater than age thirty-five. (15) Thus, according to the study, regular use of marijuana in the past on the part of females aged twenty-five to thirty-four cost the nation $2,396,000,000. (16) However, marijuana use by females greater than thirty-four, cost the nation nothing. (17) In males aged twenty to twenty-four, regular marijuana use in the past would have cost the economy $6,723,000,000. (18) On this basis, however, any use of cocaine, prescription drugs, or phencyclidine (PCP) by this group cost the nation nothing. Of course, the study really does not indicate this because its inadequacy prevented any such calculations. Its inadequacy regarding the marijuana calculations should be viewed in the same fashion. To stand in front of an audience and say that drug use cost the nation $34 billion dollars in decreased productivity is foolish and improper.

Readers may have noted that the past paragraphs followed the mode of the RTI report and used income and productivity as equivalents. I do not believe they are equivalents. Are those who choose to work in positions where they earn less than the maximum possible less productive? If so, a judge is less productive than a practicing lawyer, a medical school professor is less productive than a practicing physician, a farmer is less productive than a florist, and an elementary school teacher is less productive than an owner of a day-care center.

IV ILLEGAL DRUG USE IS DECLINING

In an affidavit filed in support of the federal government's attempt to promote urine testing of certain federal employees under the President's executive order, J. Michael Walsh characterized the drug problems in America as pervasive: "Overall, 70.4 million Americans age 12 and over 37% of the population have used marijuana, cocaine or other illicit drugs at least once in their lifetime. Nineteen percent of the population age 12 and over 36.8 million have used illicit drugs at least once in the past year, and 12 % at least once during the month prior to being surveyed." (19)

Walsh himself later notes that of the 70.4 million Americans who have tried an illicit drug at least once in their lives, for most (sixty two million) that drug was marijuana. (20) Many of the proponents of testing use such data. Proponents, however, do not often seem to notice that this same data reveals that almost fifty percent of those who have ever used an illegal drug at least once have not used an illegal drug in the past year and perhaps never will again. (21)

The sources of these data are important and interesting and merit careful attention. There are two important surveys funded by the National Institute on Drug Abuse from which most of our knowledge about the use of illegal drugs in the United States is derived. These surveys are the National Household Survey and the annual High School Senior Survey. (22) Both surveys, however, were not constructed to signal serious heavy drug use: "The unique contribution of serveys of the general population lies in their ability to furnish prevalence estimates that include many forms of drug use that never come to the attention of medical and legal authorities.....Survey responses permit classification of drug-using behavior along a continuum of involvement, substituting empirical for arbitrary definitions of medically or socially pathologic conditions, and emphasizing the typical nature of the phenomenon rather than its most dramatic manifestations." (23)

These words from the forward to the National Household Survey published in 1983 seem particularly calm in the light of the frequent repetition of its glaring numbers to induce employers and others to join in the zealot's game. The National Household Survey is a general population survey of household members aged twelve and above and has been conducted every two to three years since 1971. It excludes individuals in institutionalized settings (colleges, prisons, military bases) and, therefore, as conservative. Because each respondent gives general data, it can monitor trends by age and sex and other demographic variables.

The annual High School Senior Survey obtains self-reported information from approximately 130 high schools in the continental United States. Between 16,000 and 18,000 seniors are surveyed. In addition, the investigators construct a sub sample of individuals from each class who are followed, yielding a longitudinal study. These two surveys do not yield a horrible story of an American decline into profligate drug use, but exactly the opposite. Table I depicts data from the Household Survy regarding marijuana use over time. (24) Since 1979, the data indicate a significant decline in use among the two important youth groups. In the eighteen to twenty-five group, the percentage of the population that had ever used marijuana declined from 68.2% to 60.5% Those who had used marijuana in the year preceeding the survey declined from 46.9% to 37% and in the last month preceeding the survey, from 35.4% to 21.9%.

(Unable to duplicate Tables, charts, and graphs which follow in original text)

These national trends also are reflected in the High School Senior Survey (Table II). (25) The apex of marijuana use in this group also was noted in 1979 and has steadily declined. Marijuana use in the month preceeding the survey fell from 33.7% to 23.4%. Daily use has steadily declined from 10.3% to 4.0%

During the growing clamor to test working Americans, interest in marijuana, the most widely-accepted illegal drug, has declined, and declined significantly. This fact is not stressed by those frequently quoting the two survey instruments. The use of cocaine shows a different but still hopeful pattern. Between 1974 and 1985, the lifetime prevalence of cocaine use increased from 5.4 million users to 22.2 million users. (26) However, the trend in the High School Senior Survy showed a distinct leveling off between 1982 and 1985 with a slight decrease. (27) In 1987, cocaine use by high school students and young adults dropped twenty percent. (28) Again, a quarter of those reporting cocaine use had not used the drug in the last year. The decline noted for marijuana has occurred in most drug categories in the High School Survey including a decline in current and daily use of alcohol. (29)

FOOTNOTES

1 Angarola & Brunton, "Substance Abuse in the Workplace: Legal Implications for Corporate Actions," in SUBSTANCE ABUSE IN THE WORKPLACE 35-36 (1984) (hereinafter Angarola & Brunton).

2 id. at 36.

3 "Proficiency Standards for Drug Testing Laboratories: Hearings Before a Subcomm. of the House Comm. on Government Operations, 100th Cong., 1stt Sess. 91-92 (1987) (statement of Mark de Bernardo, Special Counsel for Domestic Policy and Manager of Labor Law. U.S. Chamber of Commerce) [hereinafter Statement of de Bernardo].

4 See Angarola & Brunton, supra note 1, at 35.

5 Cohen, "Drugs in the Workplace," 12 DRUG ABUSE & ALCOHOLISM NEWSL. I. I (Aug. 1983).

6 Photocopy sent from E. Gates Morgan to Dr. John Morgan (Nov. 1986) (discussing Firestone "Study") (copy on file at the Kansas Law Review).

7. RESEARCH TRIANGLE INSTITUTE, ECONOMIC COSTS TO SOCIETY OF ALCOHOL AND DRUG ABUSE AND MENTAL ILLNESS: 1980 (1984) [hereinafter RTI STUDY].

8. See Statement of de Bernardo, supra note 3, at 89. "$60 billion is the annual cost to the business community for drug abuse, a 30 percent increase in only three years. One half of that cost is in lost productivity." Id. (emphasis in original). de Bernardo did not cite any source for this statement. See also Church, "Thinking the Unthinkable," TIMEm May 3, 1988, at 14 (chart).

9. See RTI STUDY, supra note 7, at 67-68 (citing NATIONAL INSTITUTE ON DRUG ABUSE HIGHLIGHTS OF THE 1982 HOUSEHOLD SURVEY ON DRUG ABUSE (1983) [hereinafter 1982 NATIONAL HOUSEHOLD SURVEY].

10. RTI STUDY, supra note 7, at A-22. Interestingly, the data relied upon showed that adults thirty-five and older effectively had a zero prevalence rate of ever using marijuana on a daily basis. Id. at A-20.

11. Id. at A-20.

12. Id. at A-9.

13. See id. at A-24.

14. Id.

15. See supra note 10.

16. See RTI STUDY, supra note 7, at 68.

17. Id.

18. Id.

19. Declaration of J. Michael Walsh, Ph.D., Director of Workplace Initiatives, National Institute on Drug Abuse, at 2, submitted with Department of Justice Memorandum of Support for Summary Judgment, American Fed'n of Gov't Employees, Civ.No. 87-1797, 87-2350 (D.D.C. March 1, 1988) [hereinafter Declaration of J. Michael Walsh, Ph.D.].

21. id.

22. NATIONAL HOUSEHOLD SURVEY, supra note 9; NATIONAL INSTITUTE ON DRUG ABUSE US DEPT OF HEALTHE AND HUMAN SERVICES, NATIONAL TRENDS IN DRUG USE AND RELATED FACTORS AMONG AMERICAN HIGH SCHOOL STUDENTS AND YOUNG ADULTS, 1975-1986 [hereinafter HIGH SCHOOL SENIOR SURVEY].

23. 1982 NATIONAL HOUSEHOLD SURVEY, supra note 9, at 1.

24. Kozel & Adams, "Epidemiology of Drug Abuse: An Overview," 234 SCIENCE 970, (1986) [hereinafter Kozel & Adams] (citing NATIONAL INSTITUTE ON DRUG ABUSE, HIGHLIGHT OF THE 1985 HOUSEHOLD SURVEY ON DRUG ABUSE (Nov. 1986).

25. HIGH SCHOOL SENIOR SURVEY, supra note 22, at 47-50.

26. Kozel & Adams, supra note 24, at 973.

27. HIGH SCHOOL SENIOR SURVEY, supra note 22, at 47-50

28. Statement by Dr. Lloyd D. Johnstone, U.S. Dept. of Health and Human Services News Conference (Jan. 13, 1988) (copy on file at the KANSAS LAW REVIEW). Dr. Johnston is one of the principal authors of the High School Senior Survey.

29. Id.