CHAPTER 4. 
THE SOCIAL COST OF ALCOHOLISM AND DRUG ABUSE

4.1        Introduction

            Alcoholism and drug abuse impose a substantial burden on individuals and on society as a whole.  Alcohol and illicit drug consumption are widespread in the United States (Saxe 1983 and Rice et al. 1991).  Heavy alcohol consumption and drug abuse are known to cause significant health problems.  Alcoholism leads to organ damage (particularly the liver), brain dysfunction, cardiovascular disease, and mental disorders.  Alcoholics have higher morbidity rates and lower life expectancy (by 10 to 12 years) than the average person in the United States (Saxe 1983).  The costs of alcoholism are not confined to the alcoholic.  Alcoholism is known to be a major factor in cases brought to family courts; divorce; automobile, home, and industrial accidents; crimes such as assault and spouse and child abuse.

            Similarly drug abuse has been linked with a number of diseases, most notably are sexually transmitted diseases (STDs) and human immunodeficiency virus (HIV).  French et al. (1996) identify 30 separate diseases thought to be influenced by drug abuse.  Drug abusers also impose high costs on society through their criminal activities.

            A number of studies have attempted to estimate the direct and indirect costs of alcoholism and drug abuse in recent years and their estimates are quite simply staggering.  The total social and economic costs from alcoholism and drug abuse include the use of medical resources, productivity losses, criminal justice costs, the cost to victims of crime, the pain and suffering of substance abusers and family members.  As discussed in chapter 2, the relevance of these types of costs depends on the viewpoint of the analysts.  There are different methods that can be used for estimating these costs so the overall estimates of the social costs of alcoholism and drug abuse will vary from study to study.

            The outline of this chapter is as follows.  In Section 4.2, we discuss the types of costs associated with alcoholism and drug abuse in some detail.  The following section discusses the methods that have been used to estimate these costs.

4.2              The Social Cost of Alcohol and Drug Abuse

      A number of researchers have attempted to estimate the national cost to society from alcohol and drug abuse (Cruze, et al. 1981; Harwood, et al. 1984; Rice, et al. 1990; Harwood, 2000).  All of these researchers agree that the fundamental concept that should be used in making these estimates is that of opportunity cost.  As noted in chapter 2, opportunity cost refers to the value of the next best opportunity foregone as a result of undertaking a particular activity.  In general, the opportunity cost of an activity tends to equal total expenditures on that activity if it occurs in a competitive market.  If the activity occurs in a non-competitive market, modifications may have to be made to market prices to bring them in line with “true” opportunity costs.  If the activity change does not involve a market transaction, then the opportunity cost may have to be imputed.

      While the general concept of opportunity cost is uniformly accepted, researchers disagree over which items to include in social cost calculations and whether some items should be counted as a cost or a benefit.  There is not a consistent theoretical framework for estimating the social cost of alcohol and drug abuse (French, Rachal, and Hubbard 1991). 

      Social cost researchers often fail to make a distinction between private, external, and social costs.  The private cost of an activity is that incurred by the individual (i.e., alcohol or drug abuser) engaging in the activity.  External cost is the economic burden imposed on non-participants as a result of the activity.  The social cost of the activity is simply the sum of the private and external cost after adjusting for transfers within society (French, Rachal, and Hubbard 1991).

      Economists generally assume that consumers are rational and that they will not engage in any activity unless the expected benefits are at least equal to the expected private costs.  In fact, the consumer’s willingness to pay (i.e., price) represents a lower bound estimate of the value of the benefits a consumer expects to receive from a good.  Many buyers receive an additional value over-and-above the money price they pay, which economists call their consumer surplus (Anderson 1992).  Studies of the social cost of alcohol and drug abuse do not recognize the private benefits associated with the consumption of these substances.  They do not include the actual amounts of money that consumers spend on alcohol and illicit drugs.  As currently estimated, social costs calculators implicitly assume that resources presently allocated to the production and distribution of alcohol and illicit drugs yield no benefits and therefore they should be reallocated to something else instead.  According to this view, the concept of social cost is inherently normative (French, Rachal, and Hubbard 1991).  It is not clear who decides when an activity becomes sufficiently dangerous and/or morally wrong to be classified as a social cost.  Since social cost researchers see no private or direct benefits in the consumption of alcohol or illicit drugs, they certainly do not consider the possibility of indirect benefits or positive externalities associated with such drugs.  For these reasons, social cost studies simply add up the estimated private or direct costs and the indirect or negative external costs for alcohol and illicit drug consumption.  They make no attempt to account for or to net out the social benefits associated with the consumption of alcohol and illicit drugs.  From an economist’s point of view, the estimated social costs of alcohol and drug abuse are upwardly biased and not economically meaningful because they ignore the social benefits associated with the consumption of these drugs (Anderson 1992).

4.3              The Social Cost Model

      According to French, Rachal, and Hubbard (1991), the social cost model is composed of three distinct steps.  The first step is to summarize and classify the adverse consequences of alcohol/drug abuse.  Both physical and mental health problems can be directly or indirectly linked to alcohol or drug abuse.  Alcohol and drug abuse can also lead to a variety of social problems affecting many levels of society.  Social problems could include crime, homelessness, lower education, drug testing, problems in schools, greater use of social services, greater use of criminal justice system, family problems, community disruptions, and averting behavior.

      The second step involves the transformation of the adverse consequences into the costs of alcohol/drug abuse.  The adverse health and non-health consequences of alcohol/ drug abuse lead to costs incurred by abusers and non-abusers (family, community, and society).  French et al. (1991) divide these costs into four overlapping categories:  tangible or intangible and private or external.  Private tangible costs are incurred by the individual drug abuser.  These costs include “out-of-pocket” medical-related costs, reduced earnings, property damage, incarceration costs, and costs of averting behavior.

      External tangible costs are incurred by the non-abusing population.  This includes abusers medical costs paid by others, crime-related costs in the form of property damage, medical care for injuries, and lost wages to crime victims, similar costs to victims of automobile or workplace accidents attributable to drug abuse.  Also, the accident or crime averting behavior costs of non-abusers, communities, schools, and businesses would be included in this category.

      The intangible costs of alcohol/drug abuse do not typically involve a market transaction or any type of direct payment.  They are characterized primarily by physical and mental pain and suffering.  Private intangible costs refer to the problems such as physical disability, depression, anxiety, isolation, loss of family and job, and other emotional distress incurred by abusers.  External intangible costs refer to the physical and mental pain and suffering inflicted on crime and accident victims and on family and community members by abusers.  French et al. (1991) present a very complete itemization of the potential consequences and costs of alcohol/drug abuse.  In the next section, we shall see that many of these costs are not included in the existing studies of the social costs of alcohol/drug abuse.

      The third and final step in the social cost model is to determine the method to be used to estimate the types of costs chosen to be included in the analysis.  Several methods can be used to estimate the different cost components of a social cost model, but no single method is capable of estimating all the cost elements of alcohol/drug abuse.  The cost elements are sensitive to the methods used to estimate them.  Initial estimates of the social cost of alcohol/drug abuse varied depending on the types of costs included in the model and the methods chosen to estimate them.  Over time the types of costs included and the methods for estimating them have become more standardized so that the estimated social costs of alcohol/drug abuse can be more easily compared.  In the next two sections, the most recent estimates of the social cost of alcohol and drug abuse in the United States and Canada will be presented and evaluated.

4.4              Estimates of the Social Cost of Alcoholism in the United States

      Estimates of the social costs of alcohol abuse in the United States have risen over time.  Rice, et al. (1990) estimated the cost to be $101.8 billion in 1985.  Harwood, Fountain, and Livermore (1998) placed the figure at $148,021 billion in 1992.  More recently, Harwood (2000) increased the figure to $184,636 billion in 1998.  Based on these estimates, we cannot conclude that the problem of alcoholism is becoming more acute.  Most of the increase in the estimated cost can be attributed to population growth and to increases in prices and wages (NIAAA 1991).

      Table 4.1 presents a breakdown in the social costs of alcohol abuse in the United States for 1992 and 1998.  Social cost researchers make a distinction between core costs and related costs of alcohol/drug abuse.  Core costs are those resulting directly from the illness, whereas other related costs are the costs of secondary, non-health effects of illness (Rice, et al. 1991).  Within each category, there are direct and indirect costs.  Direct costs are those for which payments are actually made, and indirect costs are those for which resources are lost.  This classification scheme leads to four general types of cost for alcohol/drug abuse.

      Direct core costs are mainly expenses incurred in the treatment of alcohol abuse-related illnesses.  Expenditures for alcohol abuse-related treatment, prevention, research, training, and insurance administration also are part of the direct core costs.  As shown in Table 4.1, the medical consequences of alcohol abuse account for approximately 9 to 10 percent  of the total social cost of alcohol abuse.  Specialty Alcohol Services account

Table 4.1  United States Social Costs of Alcohol Abuse: 

1992 Estimates and Updates for 1998

(Millions of current-year dollars)

^aHarwood, H.; Fountain, D.; and Livermore, G. (1998).  The Economic Costs of Alcohol and Drug Abuse in the United States 1992.  Report prepared for the National Institute on Drug Abuse and the National Institute on Alcohol Abuse and Alcoholism, National Institutes of Health, Department of Health and Human Services.  NIH Publication No. 98-4327.  Rockville, MD:  National Institutes of Health. 

http://www.nida.nih.gov/EconomicCosts/Index.html.

^bHarwood, H. Updating Estimates of the Economic Costs of Alcohol Abuse in the United States prepared by The Lewin Group for the National Institute on Alcohol Abuse and Alcoholism, 2000.

for roughly another 4 percent.  Overall, direct core costs account for about 13 percent of the total cost of alcohol abuse and this share has remained relatively constant for over a decade (Hein and Pittman 1989).

      Indirect core costs of alcohol abuse are those rising from mortality and morbidity.  Mortality costs are the value of lost productivity due to premature death resulting from illness.  Morbidity costs are the value of lost productivity by persons unable to perform their usual activities or to perform them at a level of full effectiveness due to illness.  Table 4.1 indicates that the indirect core costs account for 67 to 69 percent of the social costs of alcohol abuse.  Morbidity costs alone account for approximately 47 percent of total cost and mortality costs account for an additional 20 percent.

      Direct related costs include the costs of alcohol related motor vehicle crashes, crime, and social welfare administration and other (mainly fetal alcohol syndrome).  Collectively, the direct related costs account for approximately 14 percent of the total cost of alcohol abuse.

      Indirect related costs include time lost by crime and accident victims and productivity losses because of incarceration.  These costs account for only about 5 percent of the total cost of alcohol abuse, as shown in Table 4.1.

4.5              Estimates of the Social Cost of Drug Abuse in the United States

      The first study to examine the social cost of drug abuse was by Cruze et al. (1981).  The authors followed the methodological guidelines suggested by the Public Health Service (Hodgson and Meiners, 1979).  Harwood et al. (1984) improved the Cruze et al. methodology and examined more comprehensive data sets.  They included data on the incidence and prevalence of physical health problems, use of medical services, productivity and earnings, and crime effects associated with drug abuse.  Harwood et al. estimated the social cost of drug abuse to be $47 billion in 1980.  Of this, treatment alone accounted for $1.2 billion or 2.6 percent whereas reduced productivity amounted to $25.7 billion or 54.7 percent.

      In a later study, Rice et al. (1991) using method similar to Harwood et al. estimated the social costs of drug abuse to be $44.1 billion in 1985 and $58.3 billion in 1988.  By far, crime costs composed the largest share of the total costs.  Crime costs accounted for 73.7 percent of total costs in 1985 and 72.4 percent in 1988.

      The estimated social cost of drug abuse in the United States has continued to rise in the 1990s.  Table 4.2 presents the most recent estimates for 1992 and 1998.  According to the Executive Office of the President (2001), the social cost of drug abuse rose from $102.3 billion in 1992 to $143.4 billion in 1998.

      The social costs of drug abuse can be divided into four categories.  Direct core costs refer to payments made for the consequences resulting from illness.  As shown in Table 4.2, one component of the direct core cost of alcohol abuse is for Medical consequences, which totaled $5.6 billion dollars in 1992 and $5.7 billion dollars in 1998.  These figures represent only 5.43 percent and 3.99 percent of the total cost of drug abuse in the two years.  The second component of direct core cost is the outlays for drug abuse treatment, prevention, training, research and insurance administration listed under the category Specialty Drug Abuse Services in Table 4.3.  This category accounts for 5.15 percent of the total cost of drug abuse in 1992 and 4.98 percent in 1998.

      The indirect core cost of drug abuse is those resulting from lost productivity.  As shown  in Table 4.2, these consist of lost productivity due to the abusers premature death. 

Table 4.2

Social Costs of Drug Abuse in the United States

(Millions of current-year dollars)

Source:  Harwood, Fountain, and Livermore (1998)

These mortality costs were $14.6 billion in 1992 and $16.6 billion in 1998, which represents 14.2 percent and 11.6 percent of the total cost in these two years.  Morbidity costs of drug abuse includes lost productivity due to drug abuse related illness and institutionalization/hospitalization.  Morbidity costs were $15.7 billion or 15.3 percent in 1992 and $24.9 billion or 17.4 percent in 1998.  The other lost productivity costs reported in Table 4.3 are related to the drug abuser’s criminal activity.  Lost productivity costs to crime victims was $2.1 billion in 1992 and $2.2 billion in 1998.  Abusers lost productivity due to incarceration was $17.9 billion in 1992 and $30.1 billion in 1998.  Finally, lost productivity due to criminal careers was $19.2 billion in 1992 and $24.6 billion in 1998.  The total amount of indirect care cost due to lost productivity was $69.4 billion or 67.9 percent in 1992 and $98.5 billion or 68.7 percent in 1998.  The indirect core costs of alcohol abuse are much greater than the direct care costs.

      The direct related costs of drug abuse consists mainly of expenditures reported in the Criminal Justice System category.  These include public spending on police protection, legal adjudication, State and Federal corrections, and Local corrections.  In addition, Federal spending to reduce supply and private costs in the form of legal defense and property damage of crime victims are included in this category as are social welfare administration expenditures.  The direct related costs of drug abuse are quite substantial.  In 1992 they were $21.9 billion or 21.4 percent and in 1998 they were $32.1 billion or 22.4 percent of the total cost of drug abuse.

      The most prominent feature in the estimated social costs of drug abuse is the large amount that is related to the criminal activity of drug abusers.  Table 4.3 summarizes all of  the  crime  related costs of drug abuse for 1992 and 1998.  Crime related costs in 1992

Table 4.3

Crime Related Costs, 1992 and 1998

(in millions of dollars)

Source:  Harwood, Fountain, and Livermore (1998).

were $60.8 billion or 59.5 percent of the social cost of drug abuse in the United States.  In 1998 crime related costs had risen to $88.9 billion or 62.0 percent of the total cost of drug abuse.

4.6              Methods For Estimating the Component Costs Of Alcohol and Drug Abuse

      The methods for estimating the component costs of alcohol and drug abuse in the United States are quite similar and subject to the same sorts of criticisms.  This Section discusses the methods for estimating the component costs of alcohol and drug abuse and the conceptual arguments and criticisms surrounding them.

(1)    Costs of Medical Services and Support Services

      In recent years, social cost researchers have used the prevalence-based cost-of-illness method for estimating the health care costs of alcohol related diseases.  Prevalence refers to the total number of cases at a point in time or over a given period of time.  Because the purpose of the social cost model is to estimate the total burden of alcohol/drug abuse, the prevalence of abuse is the appropriate method (French, et al. 1991).  The prevalence-based medical service cost measures the values of resources used to treat alcohol-related diseases over a specified period, most often a year.

      Researchers generally rely on epidemiological studies that indicate the degree to which specific diseases are related to drug or alcohol consumption.  The cost-of-illness method of estimating the costs of medical services attributable to drug/alcohol abuse is based on the following relationship

C = S B_m E_m

Where C is the total cost of direct medical services, B_m is the proportion of illness m that is caused by abuse and E_m is the total amount of expenditures for direct treatment of persons with illness m in a particular setting.  E_m is estimated as the product of two components:  total utilization and unit prices or charges in each setting.  The amount of alcohol or drug abuse related-spending in each setting is then summed to obtain C.

      Some experts argue that this cost-of-illness method leads to upwardly biased estimates of the costs of medical services because the B_ms are based on correlations rather than “true” causality (Heien and Pittman 1989).  Others believe the cost-of-illness methods estimates of the costs of medical services are understated because of a systematic under-reporting of alcohol/drug abuse diagnoses due to the stigmatism attached to such diseases (NIAAA, 1991 and Rice, et al. 1991).  One special alcohol-disease cost of alcohol abuse is generally reported separately.  This is the cost of fetal alcohol syndrome (FAs).  Costs associated with this disease include the costs of residential care, neonatal care, and treatment for hearing loss, mental impairment, and anatomical abnormalities (NIAAA 1991).  Similarly the costs of drug related HIV/AIDS spending is generally reported as a separate item.

      The costs of support for specialty alcohol and drug abuse related health care services includes prevention, training, research, and health administration.  The estimates of prevention spending by level of government was taken from public records (Executive Office of the President, 2001) .  The costs of training includes initial and continuing education for specialists in drug and alcohol treatment as well as for other health professionals, law enforcement officials, criminal justice professionals and clergy.  Health administrative costs relating to alcohol and drug treatment were based on estimates provided by the Health Care Financing Administration (HCFA).

(2)    Mortality Costs

      Mortality costs are the value of lost productivity due to premature death resulting from drug/alcohol abuse.  Researchers use the cost-of-illness method to determine the drug/alcohol related number of deaths for each disease over the year.  That is, they multiply the annual number of deaths by disease times the proportion of deaths that are caused by alcohol or drugs.  Again the percent of drug/alcohol deaths for each disease is based on correlations rather than “true” causality so the number of alcohol deaths is likely to be overstated (Heien and Pittman 1989).

      Once the number of alcohol related deaths for each age/sex group has been determined, the “human capital” approach is used to calculate the value of lost productivity to society.  That is, lost productivity is calculated as the present discounted value of future market earnings plus an imputed value for housekeeping services (Rice, et al. 1991).  Full-time, year-round earnings and average life expectancy for each age/sex group are used in the estimates of foregone earnings.

      This approach has been criticized on a number of grounds.  Some argue this method produces upwardly biased estimates of mortality costs because the drug/alcohol related death percentages overstate the number of deaths caused by alcohol and drug abuse and that abusers generally earn lower wages than non-abusers (Heien and Pittman 1989).  In contrast, the NIAAA (1991) and Rice et al. (1991) contend that the “human capital” approach understates the value of life for children, women, and retired people.

      A few scholars have challenged the idea that loss productivity due to the premature death of an alcoholic or drug abuser represents a cost to society.  Anderson (1992) contends that some rational individuals routinely make choices that sacrifice health at the margin to increase their consumption of other, non-health related goods.  He says the social cost model is wrong in assuming that rational individuals want to maximize life-span or health status rather than utility.  If an individual chooses to abuse drugs or alcohol knowing the risks to health and life-span, their lost earnings due to a premature death is a cost to the individual not to society.  Heien and Pittman (1989) agree that premature productivity is a cost to the abuser and not to society.

      A middle ground is possible on this issue.  Even if one accepts the concept that society loses some output due to the premature death of an abuser, it needs to be recognized they have lost the abuser’s future consumption as well.  In personal liability or wrongful death cases, it is standard practice to reduce the amount of the award to account for the future decreased consumption.  In this regard, Cheit (1961) argues that the decrease in consumption associated with the death of the family head depends on the size of the family unit.  In the case of two adults, he puts the loss in consumption at 0.30 percent of family income.  For a family of two adults and four minor dependent children, he puts the loss in consumption at 0.18 percent of family income.  To my knowledge, no one has attempted to calculate the loss in future consumption associated with the premature death of an alcoholic or drug abuser.  But the cost to society from such deaths should be based on the loss in net, rather than gross productivity.

(3)    Morbidity Costs

      Morbidity costs are the value of reduced or lost productivity associated with alcohol/drug abuse.  Reduced productivity is estimated as the product of the number of individuals affected times the average income loss per individual due to drug abuse.  The number of alcohol abusers is the size of the reference population times the alcoholic’s prevalence rate.  The average income loss per alcoholic is the percentage loss due to the disorder per individual times the average income level the individual would have earned if he or she did not have the disorder.  Data are collected by age and sex and then aggregated to obtain the total loss in national income.

      Lost productivity from alcohol and drug abuse is estimated by multiplying the number of alcohol or drug clients using inpatient hospital or residential treatment services by the average annual earnings for their age/sex group.

      Heien and Pittman (1989) have criticized this method for assuming that alcoholics would earn the same amount as non-alcoholics if they did not drink and for assuming continuous full employment, which is not true for alcoholics or any other large group of people.  For these reasons, Heien and Pittman believe the estimated value of alcohol morbidity cost is upwardly biased.

      Consistent with his argument on mortality costs, Anderson (1992) contends that reduced or lost earnings are a cost to the abuser not to society.  If the abuser is rational, he or she would have considered these costs when choosing to purchase and consume the alcohol or drug.  Presumably, their expected benefits equaled or exceeded the total cost (i.e., price plus expected discounted lost productivity) at the time of purchase.  The reduction in productivity from abusive behavior was a cost to the individual not to society.  Of course, this argument assumes the abuser is a rational consumer.

(4)    The Cost of Crime and Accidents

      Harwood, Fountain, and Livermore (1998) estimated that in 1995 about one-third of the costs of alcohol and drug abuse in the United States was due to criminal victimization and criminal justice expenditures.  Because drug abusers engage in criminal activity far more often than do alcoholics, the crime related costs of drug abuse are much higher than these costs are for alcoholics.  As reported in Table 4.3, crime related costs constituted 59.5 percent in 1992 and 62.0 percent in 1998 of the social cost of drug abuse in the United States.

The Direct Cost of Crime and Accidents

      Crime costs can be placed in two general categories:  direct costs and indirect costs.  In the social cost of drug abuse studies, the following tangible direct costs are generally estimated.

(a)   Victim Medical Costs and Property Damage

      Victims of crime face two potential direct costs:  the medical costs associated with physical injuries and property damage.  To estimate victim medical costs, researchers need to calculate three components:

(1)   The number of victims by type of crime

(2)   The percent of each type of crime related to alcohol/drug abuse

(3)   The average cost of medical care per victim by type of crime

In 1998, drug abuse crime victim medical costs were estimated to be $127 million (see Table 4.3).  Cohen (1999) believes such estimates are too low because they are based exclusively on short-term health care expenditures.

      Unless stolen property is damaged or destroyed, it is not counted as a social cost because it is transferred from one society member to another (Rajkumar and French 1997).  As reported in Table 4.3, victim property damage from drug abuse related crime was estimated at $186 million in 1998.  That same year, alcohol related crime property damage was estimated to be $325 million (see Table 4.1).

(b)   Criminal Justice System and Other Public Costs

      Criminal justice system direct costs of drug abuse include police protection, the costs of running the criminal justice system, private legal costs, correctional costs, and public spending to reduce the supply of drugs.  These costs are estimated using the offense-specific methodology developed by Cruse et al. (1981) and Harwood et al. (1984), in which causal factors that represent the proportion of offenses or arrests considered to be due to drug abuse are multiplied by the number of known offenses and this product is then multiplied by the costs per offense.  Critics have pointed out the weaknesses in this method.  Cohen (1999) believes the proportion of crimes attributed to drug abuse is overstated because the estimates of the relative risk of criminal activity for alcohol or drug abuse do not control for other factors.  They generally use upwardly biased self-reported assessments by prisoners.  On the other hand, Rajkumar and French (1997) reported that between 50 and 80 percent of individuals who were arrested for non-drug crimes tested positive for drugs at the time of their arrest.  Finally, Cohen (1999) notes that the number of crimes reported to police is significantly less than the actual number of crimes committed.  Given these conflicting forces, it is impossible to know whether the estimates of the crime costs of drug abuse are too high or too low.  The impact of drug abuse related crime on the criminal justice system and other public costs is staggering.  As reported in Table 4.3, $31.6 billion dollars was spent on this category in 1998, which represents 35.6 percent of the total social cost of drug abuse for that year.

(c)    Alcohol Crime and Accident Direct Costs

      As noted in Table 4.1, direct crime costs of alcohol abuse are quite small compared to the direct crime costs of drug abuse.  In 1998, $6.3 billion was spent on alcohol related Crime-Criminal Justice, Property Damage, etc.  This represents 4.4 percent of the total cost of alcohol abuse for that year.  The direct costs of alcohol related motor vehicle crashes and fire destruction increase the social costs of alcohol abuse quite substantially.  In 1998, $17.3 billion was spent on these accidents, which accounts for 12.1 percent of the total cost of alcohol abuse for that year.  As we shall see in the next section, the indirect costs of these accidents are much larger.

The Indirect Costs of Crime and Accidents

      The indirect costs of crime are much larger than the direct costs.  As reported in Table 4.3, there are three major types of productivity losses associated with drug abuse criminal activity.

(a)   Crime Victims Lost Productivity

      This category represents the value of time lost by victims of crime.  Since crime is caused by alcohol/drug abuse, the victims lost time is a social cost of alcohol/drug abuse.  As reported in Table 4.3, the victim lost productivity from drug abuse crime was estimated to be $2.2 billion in 1999, which represents 1.5 percent of the total social cost of drug abuse for that year.

      Lost productivity to victims of alcohol-related crime was estimated at $988 million in 1998, or 0.5 percent of the social cost of alcohol abuse for that year (see Table 4.1).  Victim lost productivity estimates are based on four components:  (1) the number of victims for each type of crime; (2) the percentage of each type of crime attributable to alcohol/drug abuse; (3) the average number of lost days for each type of crime; and (4) an average lost wage based on age/sex groups in the general population.

      Alcohol abuse studies include estimates of the productivity losses due to premature deaths in alcohol motor vehicle and other accidents.  In 1998, the estimated value of this lost productivity was $36.5 billion, which represents 19.8 percent of the social cost of alcohol abuse for that year (see Table 4.1).

(b)   Lost Productivity of Incarceration

      Productivity losses for incarcerated drug abusers is one of the largest components of the social cost of drug abuse.  As reported in Table 4.3, the costs of incarceration of drug abusers in 1998 caused productivity loss of $30.1 billion dollars, which represented 21.0 percent of the total social cost of drug abuse for that year.  The productivity losses of incarcerated alcohol abusers also is quite substantial.  As shown in Table 4.1, such losses were valued at $9.1 billion in 1998, which represents 4.9 percent of the total cost of alcohol abuse for that year.

      Estimates of incarceration lost productivity are based on the product of the total number of individuals incarcerated by primary offense times the percent attributable to alcohol/drug abusers for each type of crime multiplied by average earnings of the general population weighted by the age-sex of incarcerated offenders.  Heien and Pittman (1989) and Cohen (1999) believe the estimate for the incarceration costs of drug abuse are too  high because drug abusers generally earn lower incomes than the population as a whole for the same age/sex groups.  There is also the question raised earlier whether the percentages of crimes of each type attributable to drug abuse are accurate.

(c)    Crime Career Productivity Losses

      Crime career productivity losses are included as a social cost of drug abuse on the grounds that if the abuser could be cured, his labor effort in illegal activities would be redirected to legal production that would benefit society (Cartwright 1998).  According to Table 4.3, drug abuse crime career productivity losses were 24.6 billion in 1998, which accounted for 17.2 percent of the total social cost of drug abuse that year.

      To estimate crime career productivity losses you first have to estimate the number of chronic hardcore drug users and the amount of income they receive in illegal activities.  It is difficult to determine how much income drug abusers receive through criminal activity and whether they would earn more or less than this in legal employment (Rajkumar and French 1997).  Social cost researchers generally assume that criminal drug abusers could earn the same income as that received by their age/sex group in the general population.  Cohen (1999) argues this causes an upward bias in the estimates of crime career productivity losses because criminals’ earnings are usually lower than non-criminals’ earnings.  Several studies have interviewed drug abusers and found that they do earn smaller average annual income from crime than the opportunity legal income generally used in social cost studies (see Deschnes et al. 1991 and French 1995).

(5)    Social Welfare Administration Costs

      In dealing with the welfare costs attributable to alcohol/drug abuse, care must be taken to distinguish between the real resource cost of abuse, which is the administrative costs for substance abuse-related welfare cases, and costs which are simply transfer payments that do not constitute a true social cost.  It is important to avoid double counting of costs or benefits.  If a person previously in the workforce now receives welfare benefits as a result of abuse-related illness, it would be double counting to also include in the estimate of external costs the productivity loss (Single 1996).  The welfare costs included in most social cost studies are limited to administrative costs (i.e., the value of administrative resources used in welfare case loads associated with alcohol and drug abuse.

      In 1998, the social welfare administration costs of alcohol abuse was estimated at $484 million, which accounted for only 0.26 percent of the total cost of alcohol abuse for that year (see Table 4.1).  For that same year, the social welfare administration cost of drug abuse was estimated at $249 million or 0.17 percent of the total cost of drug abuse (see Table 4.2).

(6)    Omitted Social Costs of Alcohol and Drug Abuse

      Alcohol and drug abuse impose tangible and intangible costs on individuals and society that are generally not included in social cost studies because they are conceptually or empirically difficult to measure.  Tangible costs that are usually omitted include reduced property values in drug-ridden communities; educational programs and mass media efforts to inform the public about the hazards of drug abuse; and “averting behavior” by families, communities, and victims of crime.  French, Rachal, and Hubbard (1991) argue that averting behavior is a widespread reaction to drug abuse and it should be included in social cost calculations.

      No attempt is made to value the services of family members and friends who care for alcohol and drug abusers.  These “informal care” costs are likely to be significant, but there are no reliable data from which to make estimates (Rice, Kelman and Miller (1991).

      Intangible costs are probably the most important costs that are omitted in social cost studies of alcohol and drug abuse.  Intangible costs for alcohol and drug abusers “are manifested in depression, isolation, heightened anxiety, loss of companionship, loss of job, physical disability and other forms of pain and suffering” (French, Rachal, and Hubbard, 1991, p. 2).  The intangible costs of pain and suffering also extend to family members, community residents, crime victims, and other members of society.

      The intangible costs of crime are difficult to measure because individual well-being or utility is a theoretical concept that does not easily lend itself to monetary equivalents.  Rajkumer and French (1997) identified a number of conceptual methods that could be used to measure crime victim’s intangible costs.  First, the numerical crime-ranking method attempts to estimate the cost of victims’ pain and suffering by asking survey respondents to attach numerical rankings to each type of crime.  The numerical rankings are then converted to monetary values through a crime valuation scale to estimate the total cost of crime.

      Second, the willingness-to-pay method could be employed using contingent valuation techniques to directly elicit respondents’ valuation of crime compensation or their willingness to pay to reduce crime risk.  Third, also following the willingness-to-pay method, the property-value approach could isolate the effect of neighborhood crime rates on property values through statistical techniques.  The intangible costs of crime to potential victims should be captured in the property values.

      Also based on the concept of willingness-to-pay, the quality-of-life approach can be used to estimate the amount of crime victim’s pain and suffering.  First, one must rank the severity of the physical and psychological injuries.  These rankings are translated into monetary values by comparing the loss in quality of life due to the injury with the value of an index state such as perfect health or a statistical life.

      Finally, the jury-compensation method can be used to measure crime victim’s pain and suffering.  Under this method the cost of pain and suffering from a particular crime is measured by the total amount of jury awards to crime victims minus the observable costs of medical expenses and lost wages.  The pain-and-suffering component is not observable, but it can be indirectly estimated.

      Rajkumar and French (1997) used the jury compensation method to estimate crime victim intangible pain and suffering costs for aggravated assault and robbery in 1992 dollars.  They estimated that an aggravated assault victim’s tangible cost excluding risk of homicide was $305 compared to pain and suffering cost of $8,753.  Similarly, the estimated tangible cost to robbery victims excluding the risk of homicide was $300 compared to pain and suffering cost of $4,944.  Based on these estimates, it appears that social cost studies that omit intangible costs seriously understate the cost of alcohol and drug abuse in the United States.

4.7              The Uses and Abuses of Social Cost Estimates

      The major function of social cost studies is to provide estimates of the burden of the illness on society.  Based on the concept of opportunity costs, these studies indicate the value that the illness subtracts from the productive potential of society.  Social cost studies have been criticized for making comparisons between the real world of alcohol/drug abuse with its many imperfections to an ideal world in which no abuse exists (Heien and Pittman 1989 and Single 1996).

      Social cost researchers often use the highest estimates of the social cost of alcohol and drug abuse to justify increased public spending on research, treatment, education and prevention of alcohol and drug abuse and for law enforcement efforts to control the supply of illegal drugs (Gerstein 1991).  There is an element of circularity built into these arguments.  For example, the more money that government spends on controlling drug-related crime, such as law enforcement, the judicial system, and the prison system, the more the calculated costs of drugs to society.  This would imply then, according to the social cost viewpoint, that more money should be spent on drug policies (Sindelar 1991).  Furthermore, the more money spent on treatment, research, education and prevention, and law enforcement, the greater the costs saving to society.

      The above argument is fallacious.  The social cost studies provide estimates of the upper limit on the cost savings that could be obtained if alcohol/drug abuse were completely eliminated.  Such estimates do not provide a sound basis for determining the amount of public funds and the manner in which they should be allocated to reduce the level of alcohol and drug abuse.  To properly allocate scarce public funds, decision makers need to know the relative effectiveness of policies, in addition to the overall cost of a problem (Sindelar 1991).  In devising a national strategy to reduce the costs of alcohol and drug abuse, decision makers must decide how much money should be dedicated primarily to law enforcement (supply reduction) and how much should be allocated to treatment and prevention (demand reduction).  To make this choice properly, decision makers need information on the relative cost-effectiveness or cost-benefit ratio for these different types of interventions.  Based on economic criteria, the optimal balance between supply and demand reductions is achieved when the expenditure of an additional dollar on either alternative yields an equal benefit (Harwood 1991).  Similarly, decision makers need the results of rigorous CEA or CBA studies of alcohol and drug abuse treatment programs to determine what types of programs should be expanded and what types of programs should be abandoned.  The existing studies will be reviewed in the following chapters. 

4.8              The Cost of Substance Abuse to State Government

      Columbia University’s Center on Addiction and Substance Abuse (National Center 2001) recently estimated the costs of substance abuse and addiction (for alcohol, tobacco, and illegal drugs) to State governments.  It only considers State costs.  It does not cover federal matching funds that States spend (e.g. on Medicaid and welfare) or other federal spending.  Nor does it consider the spending of local governments on law enforcement or other items or private sector costs for health care or lost productivity.  Like social cost studies, it ignores the pain and suffering of addicts, abusers and their families and friends and other members of the community.  The study looks at the costs of substance abuse exclusively from the viewpoint of state government and their taxpayers.

      The major findings of relevance to the current study are reported in Table 4.4.  By far, the single drug linked to the largest percentage of state costs is alcohol.  In 1998, states spent $1.1 billion on illicit drugs only, $7.4 billion on tobacco only, $9.2 billion on alcohol alone, and $63.6 billion on alcohol and other drug abuse.

      States spend relatively little trying to prevent or treat substance abuse.  Of every dollar states spend, 96 cents went to shoveling up the wreckage of substance abuse and addiction and only four cents was used to prevent and treat it.  The National Center (2001) argues that substance abuse treatment programs are effective and that states could save a lot of money if they spent more money on prevention and treatment programs.

      The largest component of state spending goes to the criminal justice system.  In 1998, states spent $30.7 billion on the burden of substance abuse on the justice system for incarceration, probation and parole, juvenile justice and criminal and family court costs.  In 1998, Louisiana spent $419.5 million on criminal justice related to substance abuse, which represents 4.0 percent of the state budget (National Center 2001).  For this reason, we shall pay close attention to the CEA and CBA studies on alcohol and substance abuse that examine the effect of treatment programs on patient’s criminal behavior.  The next largest spending components in order are for Elementary and Secondary Education ($16.5b); Health ($15.2b), and Child and Family Assistance ($7.7b).  You will recall that the latter category of spending is ignored in social costs studies because transfers of income are considered economically neutral to the total society, since one person’s loss equals another’s gain.  From the government or taxpayer’s viewpoint these are a cost.  For comparative purposes, Louisiana’s substance abuse spending for 1998 is also shown in Table 4.4.

Table 4.4

State Substance Abuse Spending by Category, 1998

Source:  National Center (2001).