Skip Navigation

Link to  the National Institutes of Health  
The Science of Drug Abuse and Addiction from the National Institute on Drug Abuse Archives of the National Institute on Drug Abuse web site
Go to the Home page
Measuring and Improving Costs, Cost-Effectiveness,
and Cost-Benefit for Substance Abuse Treatment Programs

Explore Cost Benefits

Most interested parties agree that they seek to help patients become less destructive and more productive members of society. In our society, an individual's contribution often is measured in monetary terms -which is why transforming measures of effectiveness into measures of monetary benefits is so important, and why cost-benefit analysis can be so useful for decisionmakers.

According to research by Ball and Ross (1991) and Gerstein et al. (1994), substance abuse treatment can be expected to both save money and produce new income. In California, various drug treatments were estimated to save between $245 million and $1,284 million after subtracting the cost of treatment from cost savings and income generated in a single year in the early 1990s (Gerstein et al. 1994, p. 82). Of course, every treatment program differs in how much (and how quickly) this return on investment occurs, which is one reason to measure the benefits as well as the costs of individual programs.

Typical Benefits of Substance Abuse Treatment

New Income

Real income may be generated by substance abuse treatment due to increased productivity and employment of patients. This does not always occur, however. Researchers have found that employment prospects may not be as positive for former substance abusers as might be hoped (cf. Gerstein et al. 1994). This may be due to the stigma of being a former substance abuser as well as difficulties posed by criminal records. Also, the behavior patterns sometimes acquired in drug abuse lifestyles may need to change radically to meet expectations of potential employers (such as getting to work on time every day and following directives).

Cost Savings

Another benefit of substance abuse treatment is cost savings to society or taxpayers. These cost savings include -

  • Funds that otherwise would have been spent in the illicit economy for drugs.

  • Criminal justice services not required.

  • Social and health services no longer required.

These cost-savings benefits are real and can be quite substantial. Substance abuse researchers (Langenbucher et al. 1993) have found profound reductions in a number of costly events after treatment, including the following decreases:

  • Patients involved in driving while intoxicated/driving under the influence arrests decreased from 18 percent (pretreatment) to 3 percent (posttreatment).

  • Patients involved in accidents decreased from 14 to 1 percent.

  • Patients' families who sought counseling decreased from 31 to 5 percent.

  • Patients' children who missed school decreased from 5 to 1 percent.

  • Patients' spouses who missed work decreased from 10 to 1 percent.

Although different jurisdictions and different methods of assessment may provide different figures, the level of criminal activity patients exhibit can be expected to decrease by roughly two-thirds (Gerstein et al. 1994). Not every program produces a two-thirds reduction, however, so it is essential to measure how much criminal activity changes for each patient.

The reduction in criminal activity following substance abuse treatment may not produce a corresponding reduction in actual costs to society. Although costs to citizens drop in direct proportion to reductions in criminal acts perpetrated on those citizens, public expenses for criminal justice services may not decline in a similar manner. Typically, police, courts, and other components of the criminal justice system are on limited and fixed budgets, while the need for criminal justice services greatly surpasses the ability to deliver those services. For this reason, the impact of substance abuse treatment on criminal behaviors may not result in an actual reduction in criminal justice expenditures. Instead, criminal justice resources saved because of a reduction in crimes committed by former substance abusers may be diverted to other criminal justice services. The entire budget for criminal services probably will still be spent.

Similar problems may occur when cost savings benefits are measured for reduced health, mental health, and future drug treatment services. Because resources in these services typically are very limited, the actual reduction in expenditures may not be as much as might be expected from the reduction in patient use of services.

Nevertheless, transforming effectiveness findings into estimated cost savings still may have considerable value for a program evaluation. In particular, cost savings estimates can show the magnitude of criminal justice and treatment resources that are now available to help other drug abusers who previously could not be helped because of budget restrictions.

Crime-Related Cost Savings

Other research provides evidence for numerous cost savings that result from drug abuse treatment. For example, Rajkumar and French (1996) found that although total costs of crime averaged $47,971 per patient in the year prior to treatment, that figure dropped to an average of $28,657 per patient in the year following treatment. That drop of $19,314 was far more than the cost of treatment, making cost savings in terms of crime alone worth the cost of treatment: $2,828 for methadone maintenance, $8,920 for residential treatment, and $2,908 for outpatient treatment (Rajkumar and French 1996).

Employment- Related Cost Savings

French and associates (1990) found that drug treatment improved the employment and earning potential of drug abusers. Although only 31 percent of drug abusers were employed at the start of treatment, almost 45 percent were employed after treatment. There was a similar increase in the number of patients seeking work (from 9 to 13 percent). And, employed patients earned more after treatment. French and colleagues (1990) found that average personal earnings for employed patients rose from $6,158 during the year before treatment to $7,120 during the year after treatment.

The legality of employment and income also can be positively affected by drug treatment. French and Zarkin (1992) found that increasing time spent in methadone treatment by just 10 percent increases legal earnings by 1.5 percent and decreases illegal earnings by 3.2 percent. A 10-percent increase in time spent in residential programs increases legal earnings 2.4 percent and decreases illegal earnings 4.1 percent.

Health Service- Related Cost Savings

French and colleagues (1996) estimated the cost savings if one case of the following health problems could be avoided:

  • $1,100 for avoiding a case of severe venereal disease

  • $74,513 for avoiding a case of severe hypertension

  • $96,005 for avoiding a case of severe tuberculosis

  • $114,796 for avoiding a case of AIDS

Caveats on Benefit Assumptions and Calculations

Reductions in each of the above events are notable in their own right, as well as in terms of monetary savings to the individual and society. For your program, the average cost of each event can be requested from those providing criminal justice, health, or social services locally. It also may be possible to glean this cost information directly from records of expenditures of public funds. The cost savings benefit then can be calculated for each patient as the reduction directly experienced in these events.

Some important changes may be impossible to monetize. For example, patients who interrupted their education decreased from 12 to 4 percent. Although this is a substantial decrease, it is impossible to determine the monetary value of this reduction. Other changes may not occur during the time period used to collect outcome data. For example, patients' financial problems may continue to occur for years after treatment because of the length of time necessary to compensate victims and pay off accumulated debt.

Increased Expenditures From Outcomes

Substance abuse treatment can temporarily increase patients' use of social services, including welfare support, disability payments, and health services. Patients may become well enough to seek help for health problems and to seek financial support from licit as opposed to illicit sources.

According to the CALDATA study (Gerstein et al. 1994), enrollment and payments received from various social services (other than health services) increased 17 to 50 percent during treatment. Being in treatment also may increase eligibility to receive a variety of social support services.

These increases in expenditures need to be included in treatment outcome reports. They should not be excluded simply because they do not seem like benefits. They are monetary outcomes and must be considered. They will likely be canceled out by the cost savings and income generated after treatment.

A case in point: In the CALDATA study, the costs of health services decreased between 1-year periods prior and subsequent to treatment from a mean $3,227 to a mean $2,469 per person. Also, in a study reported by Holder and Hallan (1986), private health insurance costs dropped from approximately $100 per month per patient in the 2 years preceding treatment to less than $14 per month per patient in the fifth year following treatment (which is when most health sequelae of substance abuse should have subsided).

Cost savings and other benefits may vary considerably depending on the type of treatment. In the CALDATA study, residential treatment was associated with a 58-percent reduction in costs to taxpayers, whereas methadone discharge was associated with a 17-percent reduction in costs to taxpayers. Also, longer treatment generally corresponded to greater cost savings, although not for methadone maintenance.

Transform Effectiveness Findings Into Benefits

Effectiveness findings often can be transformed into benefit findings by multiplying effectiveness data by a cost value. For example, to estimate cost savings after treatment, the change in the number of thefts before versus after treatment can be multiplied by the average cost of drug- related thefts in terms of property loss, victim losses, and criminal justice expenses. Statistical analysis of data collected in an experimental design is the best way to determine whether these cost savings are significant and can be ascribed to treatment. Other research designs, including correlational methods, provide guidance and useful estimates. The transformation procedure for figuring benefits from effectiveness findings remains relatively straightforward.

The exact cost value used to transform effectiveness findings into benefit findings is ascertained by surveying local criminal justice and social and health service agencies. Ideally, you would find the cost of each criminal act, the cost of each health service used, and so on, for each patient individually. If you cannot get that information, you may be able to use estimates of average costs per patient for these effectiveness variables.

For example, suppose you know that the number of theft convictions for a patient dropped from three in the year preceding treatment to one in the year following treatment. Suppose, too, that the estimated cost of a theft totaled $1,200 after adding costs of arrest, holding, and conviction to the cost to citizens of lost property and mental anguish. The total savings that could be attributed to treatment would be the cost of thefts during a period prior to treatment, minus the cost of thefts during a similar period following treatment. For this patient, that would be:

(3 x $1,200) - (1 x $1,200) = $3,600 - $1,200 = $2,400 in cost savings.

It would be more accurate to find the actual cost of each theft. It is conceivable that the one theft following treatment was quite minor compared to the thefts preceding treatment. On the other hand, that one theft after treatment could have cost more than all the thefts before treatment.

There also may be too much variation between jurisdictions (and over years) to allow a set cost for social services, health services, criminal justice services, and other cost items to be established for all drug treatment programs throughout the country for all time.

When cost savings and benefits involve health services, welfare, and other services for which cost data are available for individual patients, the cost for each patient needs to be contrasted for different periods of treatment. These services can vary greatly between patients; an estimate of the average health care cost per patient could result in over- or underestimation of cost-savings benefits.

Table 24 lists examples of the types of costs and potential cost savings that can be included in the survey. It is not meant to be complete. Note also that room for a range of estimates is provided, in recognition of the variability in costs of these services between patients and over time for the same patient. Costs of the specific criminal behaviors of individual patients then can be contrasted for the periods -

  • Before versus after treatment.

  • Before versus during treatment.

  • During versus after treatment.

These costs can be examined separately for each category of potential cost savings or actual income produced and then summed across all categories to find the total benefit.

Table 24. Types of Costs and Potential Cost Savings

  Effectiveness measure Effectiveness-benefit transformation Benefit measures
Possible Cost Savings Criminal acts not performed Thefts at $___ / misdemeanor
$___/ felony
Assaults at $____
Savings to potential victims due to income loss avoided, property not damaged or lost, and health and mental health services not needed
Drugs not purchased Opiates at $___ to $___/day
Cocaine and crack at $___ to $___/day
Other at $___ to $___/day
Money not spent on drug purchases
Criminal justice services not used Arrests at $___/ arrest
Jail at $___/day
Prosecution at $__/ day
Expense of criminal justice services avoided
Drug treatment no longer needed $__ per patient per day for the mixture of treatments provided Cost of drug treatment no longer needed
Welfare payments not provided $__ per patient per day in welfare payments Amount of welfare payments not provided
Disability payments not made $__ per patient per day in disability payments Size of disability payments not made
Health services not used Sum health care cost use for 6 - 12 months before treatment and 6 - 12 months after treatment Cost of health services not used
Possible Benefits Produced Employment (licit) Income earned from licit sources
Entrepreneurship (licit) New income (profit) from enterprise
Income taxes paid on licit income Amount of Federal, State, and local taxes paid on licit income
Increased productivity in an existing job Increased profit for employer, company, and sole proprietorship

Net Benefit

Cost-benefit analysis answers the question of whether the outcomes of a program are worth the costs by -

  • Measuring outcomes in the same units -dollars, usually -as costs.

  • Seeing whether the value of outcomes exceeds the value of costs (by subtracting total costs from total benefits, which is called the net benefit).

To calculate the total benefit per patient for a program, simply add up the benefit figures for each of the specific measures. Similarly, to calculate the total cost per patient for a program, add up the cost figures for each procedure. Then you can calculate the net benefit (total benefits minus total costs) for the patient. Add these up for all patients to find the net benefit for the treatment program.

To make cost-benefit analysis more specific, list the specific costs of achieving the benefits on each measure. Instead of adding up benefits for all measures for one patient, and then summing or averaging across patients, add up or average for all patients the benefits attained by a program for one measure.

Present-Value Benefits

Immediate positive outcomes are more valuable than delayed positive outcomes. Nonmonetary outcomes rarely are adjusted for the amount they are delayed, but monetary benefits often are. If costs and benefits are to be compared, monetary benefits delayed by more than a year from the time that costs occur can be adjusted for their delayed value.

The adjustment divides benefits by the sum of 1 plus a discount rate (often 0.08, 0.10, or 0.14). The discount rate closely resembles the interest rate that could be earned if the money spent on treatment were invested in another activity (such as a money market fund). Benefits delayed by 2 years are adjusted by dividing them by the result of multiplying the sum 1 + (discount rate) by itself once (squared). Benefits delayed by 3 years are adjusted by dividing them by the result of multiplying the sum 1 + (discount rate) by itself and then by itself again, and so on.

The result of applying net present value to delayed benefits can be striking. Consider, for example, a stream of cost-savings benefits of $10,000 that occur at the end of the year for each of 3 years and a discount rate of 0.10. It is tempting simply to sum the benefits for a total of $30,000. The net present value of the first end-of-the-year return is, however, $10,000 ÷ (1 +.10) = $10,000 ÷ 1.10 = $9.091 following the calculation guidelines given above.

The net present value of the second year's cost-savings benefit is $10,000 ÷ [(1 + .10) x (1 +.10)] = $10,000 ÷ [1.10 x 1.10] = $10,000 ÷ 1.21 = $8,264. The net present value of the third year's cost-saving benefit is $10,000 ÷ [(1 + .10) x (1 + .10) x (1 + .10)] = $10,000 ÷ [1.10 x 1.10 x 1.10] = $10,000 ÷ 1.331 = $7,513. The total of these net-present-value benefits is far less than $30,000. It is only $24,868.

The resulting present-value benefits reflect the declining value of benefits that take longer to occur. The difficulties of making this adjustment are minor, although two to three discount rates (say, 0.08, 0.10, and 0.14) should be used. The resulting benefit adjustments provide a quantitative advantage of alternative procedures (and alternative treatment programs) that produce benefits more rapidly.

Time to Return on Investment

Net benefit is the result of subtracting costs from benefits. Present valuing benefits reduces the value of benefits. Using present-value benefits gives an appropriate advantage to programs that achieve their benefits sooner. Present valuing benefits still, however, gives an advantage (appropriately) to programs that take longer but achieve better benefits than programs that produce quick but small benefits.

Time to return on investment is the time at which investment equals monetary outcomes. The time it takes benefits to begin to exceed costs for substance abuse treatment is of concern to funders and other interest groups. Each patient can be monitored for the time actually elapsed before the monetary value of the outcomes achieved equals the monetary value of the resources used. The average time to return on investment then can be computed for all patients.

One way to do this is to keep each patient's figurative "bill" on a lined piece of paper or on a spreadsheet, such as the one shown in table 25. "Investment" is the cost of treatment services delivered. "Return on Investment" is the monetary or monetized benefit resulting from treatment services. "Cumulative Investment" is the running total of all treatment and other service costs. "Cumulative Return on Investment" is the continuous total of all benefits (monetary and monetized) resulting from treatment. "Net Benefit" is the result of subtracting the Cumulative Investment from the Cumulative Return on Investment. An advantage of keeping these data on a computer spreadsheet is that the cumulative total and the net benefit can be automatically updated by the computer each time you enter new cost (investment) or benefit data.

Table 25 could be completed just from the perspective of the present treatment program, or from the perspective of past as well as present treatments, or for society as a whole. In the "Return on Investment" column, one could add the patient's debt to society -restitution owed victims or the cost of criminal justice services. The balance unpaid from previous treatment programs also could be added here.

Table 25. Sample Cumulative Costs and Benefits and Net Benefit

Time Investment Return on investment Cumulative investment Cumulative return on investment Net benefit
DateCost of treatment services deliveredBenefit to society, patient, or other individual Running total of all treatment costsRunning total of all benefits of treatmentCumulative return minus cumulative investment
1/3 start$376 (screening) $3760-$376
1/5$145 (session)$21 (drug-free day)$521$21-$500
1/6 $21 (drug-free day)$521$42-$479
1/8$95 (group)$21 (drug-free day)$616$63-$458
1/8$145 (session) $761$63-$698
1/9 $124 (income for employed day)$761$187-$574
1/9 $21 (drug-free day)$761$208-$553

Total investment in treatment expenses can be compared to the total monetary value of outcomes achieved for a cohort of patients (say, the first 100 patients entering the clinic following the first year of startup and operation).

Time to return on investment can be contrasted for different groups of patients, such as those receiving different procedures or exhibiting different processes. The cost-benefit of different procedures also can be compared by contrasting time to return on investment for patients treated by the different procedures.

Just as calculations of time to return on investment should include present-value benefits, more delayed costs also should be adjusted for present value. The latter procedure quantifies the judgment that programs that delay some costs are preferred over programs that require all expenditures up front.

Potential Problems With Cost-Benefit Analysis

Erroneous Assumptions of Linearity

The strength of cost-benefit analysis also is its weakness or, more accurately, its problem. Because ratios can be calculated very readily (since costs and outcomes are in the same monetary units in most cost-benefit analyses), funders may make all the erroneous assumptions noted earlier that are encouraged by cost-outcome ratios .

Net benefit and time to return on investment forms of cost-benefit analysis encourage similar, and similarly erroneous, assumptions. For example, funders may incorrectly assume that because the benefit for an investment of $100,000 in a substance abuse treatment program is $50,000, doubling the investment to $200,000 will double the benefit to $100,000.

The common pattern of diminishing returns on investment would diminish this anticipated benefit to less than double. It also is possible that increasing the initial investment so much would allow entirely different (and much more effective and beneficial) treatment procedures to be used.

Some funders also may believe that increasing the investment in treatment might yield a quicker return on investment, which might not occur given limitations on how rapidly current treatment technology can modify the behaviors, life skills, and lifestyles associated with substance abuse.

Overemphasis on Monetary and Monetized Outcomes

The major problem with all forms of cost-benefit analysis is that monetary outcomes are the only outcomes considered. Most service providers, many patients, and some other interested parties believe that the most important outcomes of substance abuse treatment can hardly be quantified, much less monetized (translated into monetary outcomes). To note that some nonmonetary outcomes, such as reduced crime, can be monetized does not eliminate, but only reduces, this problem. Many providers are unwilling to consider placing a monetary value on the outcomes of their services. These providers often resent attempts by persons outside the treatment program to monetize their outcomes.

Critics also note that cost-benefit analysis has been used to justify a number of decisions that proved to be not only erroneous but disastrously so. For example, cost-benefit analyses conducted by State mental health hospitals in the 1980s apparently were used to justify sudden deinstitutionalization without preparation of the patient or the community. This removal of many mental patients from hospitals and placement into communities that were not prepared to provide necessary services exacerbated homelessness and amounted to abandonment of some patients.

This unwise decision does not necessarily mean that cost-benefit analysis is itself unwise. Problems arise when only one perspective is considered; it is important to adopt multiple perspectives in cost-outcome analyses. For example, in the deinstitutionalization analysis, only the perspective of the State mental hospital was considered.

Resources for Cost-Benefit Analysis

Several good books discuss the value of using cost-benefit analysis to evaluate programs (Nas 1996; Thompson 1980). A classic cost-benefit analysis performed in mental health (deinstitutionalization of schizophrenic patients) is provided by Weisbrod (1983). The much-discussed CALDATA study (Gerstein et al. 1994) also deserves your attention, as it is directly related to substance abuse treatment.

[Manual Index] [Previous Section] [Next Section - Compare Programs...]

Archive Home | Accessibility | Privacy | FOIA (NIH) | Current NIDA Home Page
National Institutes of Health logo_Department of Health and Human Services Logo The National Institute on Drug Abuse (NIDA) is part of the National Institutes of Health (NIH) , a component of the U.S. Department of Health and Human Services. Questions? See our Contact Information. . The U.S. government's official web portal