The Economic Costs of Alcohol and Drug Abuse in the United States - 1992
Other Effects on Society - Continued
6.3 Social Welfare Costs
Conservatively, 3.3 percent of persons involved in the social welfare system in 1992 received services because of problems, impairments, or difficulties resulting from their abuse of alcohol or other drugs. That is, this proportion of cases had eligibility because of a direct administrative determination involving alcohol or drug diagnoses. Recent welfare reforms are largely terminating eligibility due to alcohol or drug dependence per se. Social welfare beneficiaries only had modestly higher rates of alcohol or drug use and diagnoses than the general population. Accordingly, the amount of total social welfare costs (benefits plus administrative costs) attributable to alcohol and drug abuse was 3.3 percent of the $347 billion in annual expenditures, or $11.4 billion in 1992 (see table 6.10). Of these total expenditures, the value of benefits is $10.4 billion, or more than 90 percent, while administrative costs account for $1.02 billion, or a little more than 9 percent of the total. Of the $11.4 billion spent on social welfare that is attributable to alcohol and drug abuse, it is estimated that 67 percent of costs are related to alcohol abuse and 33 percent to drug abuse ($7.6 billion and $3.8 billion, respectively). However, because cash transfer programs do not represent a cost to society per se, but rather a redistribution of resources, only the administrative expense is included in these estimates. Accordingly, $683 million in social welfare administration/service costs are attributed to alcohol abuse, and $337 million in costs are attributed to drug abuse. Applicant and beneficiary costs to establish eligibility have not been estimated; data are not available to estimate such expenses.
6.3.2 The U.S. Social Welfare System
Social welfare spending is large, relative to the United States' net national product (NNP). This spending was equal to about 20 percent of the NNP in 1991, or $1.17 trillion (U.S. Bureau of the Census 1994). This section of the study focuses on those types of social welfare that assist persons unable to work because of disability; the low-income population eligible for cash or cash equivalent assistance; and direct human services, such as child welfare services, Head Start, special education, and vocational rehabilitation. Programs in these categories had expenditures of $347 billion in 1992.
For the social welfare programs included in this study, appendix table D.1 shows the number of persons enrolled in each program, total expenditures for each program, and the proportion of total costs that are transfer payments versus those that are administrative expenses. Approximately 90 percent of program expenses are transfer payments in the form of either cash or noncash aid. Administrative and direct service expenses, which account for around 9 percent of total social welfare expenditures, do imply a net loss of resources to society.
The fundamental hypothesis of this analysis is that persons with severe alcohol and other drug problems may have reduced earnings or may become disabled and thereby become eligible for social welfare programs. Addiction to alcohol or other drugs can affect the likelihood of gainful employment, and it can increase the chance that persons may qualify and utilize one or more income-eligible programs. In addition to causing welfare eligibility, alcohol and drug abuse may be factors that contribute to "welfare dependency" - that is, alcohol and drug abuse may be barriers for some people to getting off welfare.
Insofar as alcohol or drug use may be the cause for receiving social welfare services, this study's estimates of the proportion of alcohol and drug users who qualify for services is derived from data currently available from Federal programs in 1992. Some Federal and State social welfare programs designated alcohol or drug abuse as a disability, provided a person met certain clinical standards for addiction. Therefore, a percentage of social welfare recipients are considered disabled
because of alcohol or drug problems. The cost estimates reported in this study are derived from analysis of the following:
- Relevant social welfare programs and their total expenditures;
- The proportion of expenditures devoted to direct benefits, transfer payments, and administrative costs; and
- The estimated proportion of social welfare recipients that receive services because of their abuse of alcohol or drugs.
This study organized social welfare programs into three categories: social insurance, public aid, and direct human services. Social insurance programs provide protection against wage loss resulting from retirement, prolonged disability, death, or unemployment. Disability benefits are payable to an insured worker under age 65 with a prolonged disability and to the disabled worker's dependents on the same basis as dependents of retired workers.
Public aid includes both cash and noncash benefit programs. Public aid cash benefits include State-administered public assistance programs, emergency and general assistance, and the Federal Supplemental Security Income (SSI) program. These programs provide assistance to persons based on income eligibility. These benefits include public transfers (food stamps, school lunch, and public housing) and employer- or union-provided benefits to employees.
The third type of social welfare program reviewed here is direct human services, such as child welfare, Head Start, vocational rehabilitation, and special education. Direct human service programs provide services to individuals, and with the exception of maintenance payments for foster care and adoption assistance, expenditures for these programs consist of the costs of providing services.
6.3.4 Alcohol and Drug Abuse and Social Welfare: Prevalence and Costs
Two types of data are generally available concerning the extent of alcohol and drug abuse found among social welfare beneficiaries: (1) data on prevalence of alcohol and drug abuse among beneficiaries and (2) administrative data on cases where alcohol and drug abuse were conditions for eligibility. Unfortunately, there are substantial variations in how alcohol and drug abuse and/or dependence are defined in different prevalence studies and in different program eligibility criteria.
It is important to review the various types of studies because in the public debate over this issue the data are often used almost interchangeably. Prevalence studies are often used in such a manner as to imply that one-third or even more of those who receive social welfare benefits do so because of alcohol or drug problems. Administrative data reveal that most of the social welfare beneficiaries with alcohol and drug abuse problems actually are eligible because of another problem not necessarily related to abuse of alcohol and/or drugs.
18.104.22.168 Prevalence Studies
Several recent studies have estimated the prevalence of alcohol and other drug abuse in various social welfare programs. Since few social welfare programs determine eligibility based on an assessment for alcohol and drug problems, these recent studies focus on the prevalence of alcohol and drug use and frequently use the National Household Survey on Drug Abuse (NHSDA) as the primary data source to calculate their estimates.
Different prevalence studies have applied different criteria for determining use and abuse of alcohol and drugs. Most of the studies have applied some criteria to ascertain alcohol abuse or dependence, but the criteria have varied from study to study. Few studies have applied criteria to determine drug abuse or addiction; instead, they have reported past-month and past-year drug use.
The Center on Addiction and Substance Abuse (CASA) performed a study that focused on alcohol abuse, drug abuse, and women on welfare. CASA estimated that 28 percent of Aid to Families with Dependent Children (AFDC) recipients are addicted to or abuse drugs or alcohol (CASA 1994). However, this study did not apply criteria for determining either abuse or dependence and appears to have overestimated the magnitude of the problem. CASA used NHSDA data.
Two studies conducted by the Assistant Secretary for Planning and Evaluation (ASPE) and the National Institute on Drug Abuse (NIDA) analyzed patterns of alcohol and drug use among persons enrolled in AFDC and food stamps, the first of which (U.S. Department of Health and Human Services [DHHS] 1994b) did not differentiate levels of impairment. The 1994 study estimated that 8.7 percent of persons receiving AFDC benefits and 10.4 percent of persons enrolled in the food stamps program had three or more episodes of binge drinking per month and that 10 percent had used an illicit drug in the last month. Table 6.11 summarizes findings for persons receiving AFDC and food stamps.
The second study by ASPE and NIDA (DHHS 1994b) analyzed the extent of use and impairment relating to that use. ASPE and NIDA found that 5.2 percent of persons in the AFDC population had a "significant" impairment resulting from alcohol and drug abuse, compared with 2.6 percent in the non-AFDC population. Table 6.12 summarizes the results of the study.
A study by Weisner and Schmidt (1993) analyzed alcohol- and drug-related problems among health and social services agency populations, including welfare recipients, in a northern California county. The study found that welfare recipients were 2.2 times more likely to be problem drinkers, 2.1 times more likely to be illicit drug users, and 12.0 times more likely to be multiple-drug users than were members of the general population. Results of the study are summarized in table 6.13.
The prevalence studies reviewed in this section demonstrate that there are a number of different estimates that can be - and have been - seized upon in the debates over the roles of alcohol and drug abuse in social welfare spending. The data support the conclusion that the rates of alcohol and drug abuse are materially higher among the population of social welfare program beneficiaries; however, the data also demonstrate how much the nature and extent of the problem can be misjudged if meaningful criteria are not applied to the determination of estimates. The following section presents more authoritative data about the frequency with which an alcohol or drug abuse problem actually is the basis for receipt of social welfare benefits.
A recent study of this issue found that rates of alcohol and drug problems among social welfare beneficiaries were actually quite similar to rates for the general population. Grant and Dawson (1996) analyzed the 1992 National Longitudinal Alcohol Epidemiologic Survey (NLAES) and applied DSM -IV clinical criteria. This is arguably the most rigorous assessment done to date, based on rigorous diagnostic criteria.
22.214.171.124 Administrative Data on Alcohol and Drug Abuse in Welfare Programs
Two recent studies have estimated the proportion of persons receiving disability benefits because of an explicitly diagnosed and recorded alcohol or other drug problem. The first was a 1994 report conducted by the U.S. General Accounting Office (GAO) on disability benefits that examined the number of alcohol and drug addicts receiving disability benefits in two social welfare programs: Disability Insurance (DI) and Supplemental Security Income (SSI). The report estimated an annual cost of $1.4 billion for individuals with alcohol and/or drug addictions, including both benefit payments and administrative costs (GAO 1994). Addiction is one of many conditions that determine eligibility for both of these programs. Persons qualifying with addiction as either a primary or secondary diagnosis have increased to 250,000 persons in 1994, up from 100,000 in 1989. However, recent welfare reforms are terminating alcohol and/or drug dependence as a basis of eligibility for many Federal and State social welfare programs.
The DI program determines eligibility for benefits using clinical criteria and medical evidence, such as symptoms and laboratory test results. To qualify for benefits, recipients must have physical or mental impairments expected to last a minimum of 12 months or result in death. Alcohol or drug addiction could be the primary criterion for eligibility (which was terminated effective March 1996), or it could be a secondary reason. In cases where addiction is a secondary diagnosis, the recipients must have a primary problem sufficient to justify benefits, whether or not they have a secondary addiction diagnosis.
The GAO report analyzed Social Security Administration data bases and found that although nearly 100,000 recipients of DI had a determination of alcohol and/or drug addiction in 1993, only 26,000 cases were primary diagnoses, and the rest were secondary diagnoses (where eligibility was primarily determined based on another disabling condition). The study did not distinguish between alcohol and other drug addiction.
The study also analyzed participation of persons addicted to alcohol and other drugs in the SSI program. SSI provides income support to indigent persons 65 and older, blind or disabled adults, and blind or disabled children. Eligibility for SSI on the basis of alcohol or drug addiction is fundamentally the same as it is for DI (and has likewise been terminated). Analysis of the SSI program showed that at the end of 1993, more than 150,000 SSI recipients (4.8 percent of the adult SSI population) had a primary or secondary diagnosis of alcohol or drug addiction. About 90,000, or 60 percent, had a primary alcohol or drug addiction diagnosis, but the rest were eligible because of another primary diagnosis (GAO 1994). Table 6.14 summarizes data on alcohol and drug abuse in the DI and SSI programs.
A 1994 report from the Office of the Inspector General of the Department of Health and Human Services (DHHS) also analyzed the proportion of SSI recipients who qualified because of alcohol and/or other drug problems. This study only looked at recipients in the Drug Addiction and Alcoholism (DA&A) program, in which treatment is mandated as a condition for receiving benefits. The Inspector General reported that 80 percent of alcoholics and drug addicts continued to receive SSI payments 3 years after qualifying for the program despite treatment mandates. The study did estimate that 84 percent of the addiction population was alcoholic or both alcohol and drug addicted (DHHS 1994d). The results are summarized in table 6.15.
The results from the various studies reviewed above are summarized in appendix D.
There is a difference between alcohol and drug use that causes a person to become eligible for a social welfare program(s) and alcohol and drug use that occurs during receipt of benefits but is not severe enough to be considered the primary cause for eligibility. Two important conclusions are drawn from the preceding review of prior research:
- About 3.3 percent of welfare recipients in 1992 had alcohol and drug abuse problems severe enough to establish eligibility for selected welfare programs.
- About 67 percent of the alcohol and drug abuse-related welfare cases appear to be attributable to alcohol and about 33 percent to drug abuse (with overlap in the two groups).
The estimates selected to calculate persons in social welfare programs whose alcohol and/or drug use have qualified them for services are based on data obtained from programs or studies that apply rigorous criteria for assessing whether alcohol and drug abuse is a significant reason for qualifying or receiving social welfare benefits. The review of the literature determined that four programs employ clinical definitions and have data from which to make estimates. Estimates for other social welfare programs are developed by applying the values for the four "data rich" programs:
- Disability Insurance (DI),
- Supplemental Security Income (SSI),
- Aid to Families with Dependent Children (AFDC), and
- Food stamps.
This paper modified the estimates of alcohol- and drug-related impairment for SSI and DI (which were both about 5 percent). For SSI, the number of alcoholics and drug addicts reported in the GAO study included persons in the Drug Addiction and Alcoholism program and all primary and secondary diagnoses; similarly, the number of alcohol and drug abusers in the DI program included all primary and secondary diagnoses. This study assumes that only one-half of the cases with a secondary alcohol and drug abuse diagnosis are attributable. Although all the secondary diagnoses could be dropped, it is plausible that the alcohol or drug abuse problems may have caused the primary nonalcohol and drug diagnosis. However, no data were available to test this assumption.
The results were estimates that 3.1 percent of SSI and 1.7 percent of DI cases were attributable to alcohol and drug abuse. The ASPE study's estimate of 5.2 percent of persons enrolled in AFDC because of "significant" impairment due to alcohol or drug abuse was adopted and was also applied to the food stamps program.
This paper then applied these estimates to social welfare programs for which no alcohol and drug abuse-related data existed. These estimates were applied within the three typology areas discussed earlier. For example, the estimate for persons enrolled in DI was 1.7 percent, and because DI is a social insurance program, the 1.7 percent estimate was applied to the other social insurance programs. Similarly, SSI and AFDC are public aid programs, so the simple average of 3.1 percent and 5.2 percent was applied to other public cash and noncash aid programs. Finally, the 3.3 percent average of DI, SSI, and AFDC was applied to all programs under direct human services. Detailed calculations by social welfare program appear in appendix table D.3, and summary calculations are presented in table 6.16. The estimates for transfer payments and administrative expenses are also presented in table 6.16.
Estimates of the proportion of cases related to alcohol and drug problems, respectively, were derived using SSI data. Those records indicated that for persons who qualified for SSI with alcohol and drug abuse as the primary diagnosis, 67 percent had a primary dependence on alcohol and 33 percent were primarily drug addicts.
Applying the above rates to social welfare spending for the programs reviewed in this study yields an estimate of the cost of services (i.e., benefits plus administrative costs) provided to people with alcohol and drug problems of about $11.4 billion in 1992. Of this total, $10.4 billion are transfer payments, which reallocate resources from one segment of society to another. Administrative and direct service expenses accounted for $1.02 billion, and these costs represent a net loss of resources to society. Two-thirds of the total expenditures - $7.6 billion - is attributable to alcohol, and $3.8 billion is attributable to illicit drug abuse. These estimates may be conservative and may underestimate the true extent to which alcohol and drug abuse problems contribute to receipt of social welfare services. This paper's estimates are based on analyses that applied clinical criteria (dependence or abuse) or clinical-like criteria (significant impairment) for determining alcohol and drug abuse-related program eligibility. The estimates are, therefore, much lower than the prevalence estimates summarized in this report and somewhat lower than the number of reported cases in the DI and SSI programs.
Alcohol and drug abuse impose significant costs to the social welfare system. This study conservatively estimates that more than 3.3 percent of current social welfare cases are attributable to alcohol or drug abuse. This number is directly tied to cases for which there is a direct administrative finding of eligibility due to disability or impairment from alcohol or drug disorders. It is possible that the actual number of persons whose alcohol or drug abuse problems led to participation in the social welfare system may indeed be much higher, because both alcohol and drug use prevalence rates are somewhat higher than this estimate (in the range of 5 or even 10 percent). Determining the level at which alcohol and drug abuse impair or prevent persons from working and lead to receipt of welfare benefits is an important policy issue, and this study represents an attempt to analyze available data and construct reasonable criteria for this issue.
The economic burden of alcohol and drug use that is borne by the social welfare system is significant as well. Alcohol and drug abuse-related social welfare costs were about $11.4 billion in 1992, of which $10.4 billion was passed on as transfer payments, and $1.02 billion was an administrative expense. Transfer payments do not represent a net loss to society, but rather a redirecting of resources, and accordingly are not included in these estimates of other impacts of alcohol and drug abuse. Administrative costs, however, are included here. This study estimates that about $683 million was attributable to alcohol abuse, and about $337 million was attributable to drug abuse.
Beneficiaries also bear costs in applying for and obtaining program benefits. It is possible that these costs may be comparable in size to program administrative costs, because a great deal of time can be required to make application and document eligibility for social assistance programs. These costs have not been included because of a lack of suitable data.
Alcohol and drug abuse in the social welfare system represent significant policy concerns. This review of the literature found several studies that estimated that welfare recipients were about twice as likely as members of the general population to have a serious alcohol or drug abuse problem. Another area of concern is that welfare recipients who are drug abusers are more likely to remain in a program for longer periods of time.
6.4 Motor Vehicle Crashes
A significant amount of research and literature indicates that alcohol is a causal factor in motor vehicle crashes. Most of the calculations in this section are based on the Blincoe and Faigin study (1992), which is the most recent comprehensive analysis of motor vehicle crashes and the most recent data on alcohol involvement in crashes. Verbal inquiries were made with the National Highway Traffic Safety Administration (NHTSA), and Traffic Safety Facts 1993 were reviewed regarding data on drug involvement in motor vehicle crashes.
Although it is hypothesized that drug abuse may be a causal factor in motor vehicle crashes, reliable data with which to estimate the value of this cost element still do not exist. Little is known about the role of other drugs of abuse in fatal motor vehicle crashes and the effect of driver impairment from drugs in nonfatal crashes. Orsay et al. (1994) conducted a hospital-based study of injured motorists in the Chicago metropolitan area that linked police reports and hospital data and evaluated the role of other drugs in addition to alcohol in nonfatal, serious motor vehicle crashes. Similar to recent results reported from a Tennessee study (Kirby et al. 1992), this study revealed that other drugs in addition to alcohol are often present in the seriously injured motorist population. These two very limited studies begin to make the case for more intensive study of the involvement of drugs in motor vehicle crashes; however, they do not as yet constitute a sufficient basis for developing cost estimates. Consequently, we include this cost element only for alcohol abuse and have excluded drug abuse costs from this study.
Strong data are available concerning the involvement of alcohol in motor vehicle crashes. Information on the number of fatalities, property damage only, accidents, and number of injuries was obtained from data published by the NHTSA from the Fatal Accident Reporting System (FARS), the National Accident Sampling System (NASS), the Crashworthiness Data System (CDS), and the General Estimates System (GES). According to a recent review of the literature, there is an increased risk of a fatal motor vehicle crash with a 0.04 blood alcohol content (BAC) and a dramatically increased risk at BAC's of 0.10 percent or higher (Perrine et al. 1989).
Blincoe and Faigin (1992) have developed fairly detailed estimates of motor vehicle crash costs for 1990, including a number of different components. Accidents with personal injury are divided by the severity of injury, using a 5-point Maximum Abbreviated Injury Scale (MAIS), with minimum/moderate injuries represented by categories AIS 1 and AIS 2 and severe and critical injuries by categories AIS 3 through AIS 5.
For purposes of this section, only data on direct-cost elements are included - that is, legal and adjudication costs, insurance administration, and vehicle and roadway damage. Costs for alcohol-related motor vehicle health care and productivity losses due to premature mortality and morbidity have been estimated elsewhere in this report. The cost estimates by Blincoe and Faigin are built on detailed NHTSA information about the following factors: (1) the number of accidents classified by the greatest level of injury severity (fatal, property damage only, and accidents with injuries classified by AIS severity level); (2) the costs per type of accident, by category for 1992; and (3) the percent of motor vehicle crashes in each AIS category caused by alcohol abuse. The system attempts to account for all costs associated with an accident, including multiple vehicles and multiple victims. Estimates of alcohol involvement by level of crash severity were determined by using NHTSA data on the proportion of crashes (fatalities, severe/critical injuries, minimum/moderate injuries, and property damage) involving alcohol-impaired drivers with a BAC of 0.10 percent or higher. If the costs related to crashes with drivers who had BAC's below 0.10 were included, then attribution factors and costs would be about one-fifth higher (i.e., the mortality attribution factor would be 50 percent instead of 39.7 percent).
Estimates from Blincoe and Faigin for 1990 were inflated to 1992 using the consumer price index (CPI) all-items index. The 1992 motor vehicle crash costs attributed to alcohol abuse were $13.6 billion (see table 6.17).
6.5 Fire Damage and Control
Although alcohol has been implicated in more than 50 percent of fire-related deaths in a number of studies (NIAAA 1993), it appears that alcohol is much less frequently involved with fire-related losses. Berry and Boland (1973) reviewed the sparse literature and concluded that about 6.1 percent of structural fire loss was alcohol related. This is because alcohol-related fires appear to be primarily due to inebriated smokers falling asleep while smoking. That study estimated that 11.2 percent of economic losses from fire were alcohol related. In lieu of more recent reviews, this study adopts the same attribution factors. These factors were also employed in Cruze et al. (1981) and in Rice et al. (1990). Thus, 6.1 percent of structural fire losses totaling $6.96 billion ($425 million) are attributed to alcohol, and 11.2 percent of fire protection services of $10.4 billion ($1.165 billion) are attributed to alcohol. (Estimates of fire loss and protection costs are from U.S. Bureau of the Census 1994.)