Benefit-cost analysis of transitional employment programs.
With growing shortages of reliable unskilled labor, employers are continuing to expand efforts to hire individuals with mental and physical disabilities previously believed to be unemployable in the competitive labor market. Properly trained and placed in the right job, there are significant societal benefits to be reaped from tapping this productive secondary labor market. Indeed, such placements produce indirect benefits that go beyond the direct benefits of increased earnings for those individuals with disabilities successfully placed into the competitive labor market. For example, mentally retarded but competitively productive individuals who are earning 50 cents an hour in sheltered workshops could potentially earn the $3.35 hourly minimum wage or more in the competitive market. Moreover, the transition into competitive employment releases that portion of society's resources that are devoted to sheltered workshops. Thus, more of other kinds of goods and services can be produced. In addition, taxpayers indirectly benefit from the reduced tax burden associated with reduced sheltered workshop dependency.
However, ensuring the successful transition from noncompetitive to competitive employment is not costless. It requires training, careful job placement, and many times follow-up services that can only be provided by trained professionals. Transitional employment programs developed to provide these services are costly to society.(1)
In addition, the federal government has attempted to encourage the competitive employment of people with disabilities through demand-side subsidies, e.g., the targeted job tax credit available for employers who hire people with disabilities that expired in 1985 but was reinstated in the 1987 federal tax bill. Such subsidies add to the cost of placing and maintaining people with disabilities in competitive jobs.
Do the benefits of programs designed to transition people with disabilities into competitive employment outweigh the costs? Are these programs economically efficient? Because they are growing in number and importance at the same time funding is becoming scarce and subject to more political scrutiny, it is clear that these programs be properly evaluated. This article presents a discussion of a framework for evaluating transitional employment programs from an economic perspective. Specifically, the paper demonstrates how benefit-cost analysis can be used as a logical and practical technique for evaluating the merits of transitional employment programs.
This article also presents two applications of benefit-cost analysis to transitional employment projects and makes inter-project comparisons. The first project is a benefit-cost analysis of Project Employability. The second is a benefit-cost analysis of the Structured Training and Transitional Services demonstration project. Overview
Transitional employment programs (and social programs in general) require the use of society's resources, including labor, capital, materials, and land. These resources are scarce and therefore to employ them in one activity necessarily implies that they are not available for use in other activities. Consequently, using these resources involves an opportunity cost because productive services in other sectors of the economy must be sacrificed. The opportunity cost of resources employed in a project is the foregone value of the next best alternative use of these resources.
Placing a dollar figure on these costs is in many cases much easier than placing a dollar figure on the benefits arising from a project. Usually these costs are approximated by what must be paid to attract resources away from other activities and into employment in the social program. That is, it is assumed that these payments equal the value of goods and services that society must sacrifice as a result of implementing the program. For example, the obvious costs of a transitional employment program include the costs associated with administrative, professional, and operational personnel. Other costs which may not be so obvious, but which are important to include in a benefit-cost analysis, will be discussed in the following sections. Presently, for a particular program it is sufficient to recognize that larger projects are more costly in terms of total costs incurred than smaller projects simply because they require more of society's resources. This is illustrated graphically in Figure 1 where project size is measured on the horizontal axis, dollar costs are measured on the vertical axis, and total costs are shown increasing as the project size increases.
Transitional employment programs also produce desirable results and as such give rise to benefits. These can be primary benefits -- the direct effects or outputs of the program -- or secondary benefits -- those indirectly generated by the program. For example a primary benefit would be the increased output of goods and services available to society as a result of the program increasing the productivity of people with disabilities. A secondary benefit would be the reduced use of alternative programs such as a day center or sheltered workshop. Generally, as is illustrated in Figure 2, larger projects produce greater total benefits.
Once the costs and benefits have been identified and measured they can be compared and used to evaluate the economic efficiency of the program. The primary issue to be addressed in a benefit-cost analysis is the question of whether society is better-off with the program than without the program. In other words, do the total benefits exceed the total costs? The framework is illustrated graphically in Figure 3, where the total costs and total benefits previously illustrated in Figures 1 and 2 are now compared. By comparing the total benefits (B) to the total costs (C) for a given project size, we define the net benefits (NB): NB = B - C. If NB is positive, the program is economically desirable. If NB is negative, society is better off without the program.
As illustrated in Figure 3, for projects up to a size of S3, the NB is positive and therefore economically desirable. Note that when the program is economically desirable the benefit costs ratio exceeds a value of 1. For example if the benefits are estimated to be $1.75 million and the costs to be $1.0 million, then the benefit-cost ratio is 1.75. This means that each dollar of cost provides benefits worth $1.75 on the average.
A benefit-cost study will generally focus on a given size program or project, e.g., S1 in Figure 3. If the NB for this size project is positive (the B/C ratio exceeds 1), then the project is efficient. However, it is not necessarily the most efficient size project. Society may be better off by devoting more resources to the program and increasing the size of the project. The most economically efficient size project is where total benefits exceed total costs by the greatest amount, i.e., where the NB is the greatest. In Figure 3, the net gain to society would be greatest for the project size identified by S*. This is where NB is maximum. If the project being evaluated is of size S1, then society would do better by allocating more resources to the program and increasing the project size beyond S1 to S*. On the other hand, while project sizes beyond S* have positive NBs, society's net gain would increase by diverting resources away from the program and undertaking a smaller size project until S* is reached.
Similarly, when the results of a benefit-cost study are summarized as a ratio of benefits to costs, the most efficient size project is not necessarily the one with the largest benefit-cost ratio. For example in Figure 3, suppose that the project identified as S1 had a total benefit of $6 million and a total cost of $2 million. Also, suppose that the project identified as S* had a total benefit of $10 million and a total cost of $5 million. In this case, the benefit-cost ratios would be 3/1 and 2/1 for projects S1 and S* respectively. Yet, project S* is more efficient because it provides a net benefit of $5 million compared to $4 million for S1. Choosing the project size with the largest net benefit is clearly the appropriate strategy, e.g., see Thompson (1980) for a detailed discussion.
Although it would be desirable to conduct a benefit-cost study for a large range of project sizes as illustrated in the above framework, the cost of performing such analyses is usually prohibitive. Thus, usually it will not be possible to identify the most efficient size project. However, in many instances it may be possible to compare one transitional employment project of one size to a similar transitional employment project of a different size and use this comparison as a means of evaluating whether more or less funds should be devoted to the project. Frequently even this may not be possible and the benefit-cost analysis must focus on one project size. But even in this instance, the benefit-cost analysis provides valuable information upon which policy decisions can be based and future courses of action can be taken.
Of course, within the general benefit-cost framework presented above, it is assumed that all program benefits and costs are tangible and can be measured in dollars and their magnitudes compared. Unfortunately, many of the benefits and costs associated with transitional employment programs are intangible and cannot be valued in monetary terms -- at least not with available data or empirical techniques. The effect of a transitional employment program on a client's self-esteem or quality of life is difficult to assess in dollar terms. Yet such intangibles are important program effects and certainly relevant to policy decisions concerning the economic efficiency of the program. Given their importance, it is essential that such intangibles be at least qualitatively addressed in benefit-cost analyses. Conducting Benefit-Cost Analysis
Conducting a benefit-cost analysis can be viewed as a five-step procedure. Each of these steps will be discussed: 1. Define the Program and Alternative 2. Specify the Evaluation Criteria 3. Identify Specific Benefits and Costs 4. Estimate Benefits and Costs 5. Determine Present Value Benefits and Costs
Define Program and Alternative
Defining the program in terms of who the program directly and indirectly serves, what its intended effects are, and what unintended effects it may have is the first step necessary in conducting a benefit-cost analysis. While the intended effects of a transitional employment program may be obvious, there may also be unintended consequences which are not so obvious. For example, the placement of individuals with disabilities into competitive employment may displace other workers and therefore impose costs upon them. In another case, the Aid to Families with Dependent Children (AFDC) program intended to provide income support for the needy, but at the same time created a work disincentive. Thus, careful consideration should be given to all the effects of a social program in conducting a benefit-cost analysis.
In addition, it is necessary to define the alternative or alternatives to which the program is to be compared. That is, in the absence of program participation, what alternative activity would the client be engaged in? Ideally, when conducting a benefit-cost analysis a control group would be employed for comparison. For small projects, however, this is usually not feasible. In such instances, a program alternative must be based on conjecture or available evidence from other studies.
Specify Evaluation Criteria
Two basic questions arise when considering the desirability of a social program: (1) Is the program efficient? and (2) Is the program equitable? Efficiency is concerned with the program effect on the value of goods and services to society, i.e., is the net value (total benefits less total costs) of goods and services with which the program greater than without the program. Equity is concerned with the distribution of goods and services among individuals or groups of individuals in society, i.e., what individuals or groups of individuals (program participants and non-participants) receive benefits and what individuals or groups of individuals bear the cost. The focus of benefit-cost analysis is primarily on efficiency. However, equity considerations are also important in the evaluation of transitional employment programs and in some instances may even outweigh the efficiency criterion. Thus, it is important to address the equity question. To a certain extent this can be accomplished within the benefit-cost framework by defining and examining benefits and costs of transitional employment programs from different perspectives of the client (program participants), the taxpayer (non-participants), and society as a whole. It may also be desirable to focus on subgroups such as in-state vs. out-of-state taxpayers.
Identify Specific Benefits and Costs
The next step in conducting a benefit-cost analysis is to identify and list the specific benefits and costs. This can best be accomplished in an accounting type framework proposed by Thornton (1985). As an illustration, Table 1 presents a shopping list of potential benefits and costs that can be applied to the evaluation of a transitional employment program. Benefits and costs are listed from three perspectives. The symbols A +,-,o indicate whether a component is a net benefit, net cost or neither, respectively. Here direct benefits can be expected in the form of increased output (or earnings) and indirect benefits in the form of reduced use of alternative programs. Program costs are those associated with program operation and administration, as well as an offset to costs in the form of the in-program output of participants.
Note that Table 1 is quite comprehensive in listing expected benefits and costs. It includes both tangible and intangible benefits and costs. Preferences for work, increased self-sufficiency, and quality of life are intangibles that will be difficult to assign a dollar value to, but nonetheless are recognized as an important component of the benefit-cost analysis.
Estimate Benefits and Costs
Once the expected benefits and costs have been identified and placed into the accounting framework illustrated in Table 1, the next and probably most difficult task in the benefit-cost analysis must be undertaken. Dollar values must be assigned to the benefits and costs. Because it is unlikely that the analyst will have access to the exact dollar values of these benefits and costs, estimates must be obtained from a variety of sources.
Estimates of the value of post program output may be obtained by assuming that this value is equal to the program participant wages and fringe benefits. Post program wages are generally obtainable directly from accounting records. Fringe benefits can be estimated using a variety of reports from the U.S. Department of Labor. For example, in 1980 it was estimated that fringe benefits for minimum-wage workers were 15 percent to 18 percent of gross wages. In Table 1, this estimate of post program output (earnings plus fringe benefits) is clearly a benefit to the participant and society.
Table : 1. POTENTIAL BENEFITS AND COSTS OF A TRANSITIONAL EMPLOYMENT PROGRAM
Table : 2. AVERAGE OPERATING EXPENDITURES FOR ALTERNATIVE PROGRAMS FOR MENTALLY RETARDED YOUNG ADULTS
Table : 3. Average Monthly Payment and Administrative Costs for Transfer Programs: 1980 Dollars Per Recipient
An increase in postprogram earnings implies an increase in postprogram taxes, including payroll, income, and sales taxes. A study conducted by the Brookings Institution in 1974 estimates that the effective tax rate for low-wage workers is approximately 23 percent of their total income. Thus, postprogram taxes may be estimated by applying this rate to the estimated postprogram earnings of participants. In Table 1, increased taxes are entered as a cost to the participant and a benefit (reduced burden) for the taxpayer (nonparticipant). Note that this is merely a distributional effect (i.e., a redistribution of income from nonparticipants to participants) of the program and therefore does not enter as a cost or benefit to society.
Participation in a transitional employment program can be expected to affect an individual's use of other substitute programs. This will free-up some of society's resources and therefore generate indirect benefits. Generally, it is possible to obtain at least average cost estimates for alternative programs. Table 2, for example, provides estimates of average operating expenditures for alternative programs for mentally retarded young adults. Such estimates can be used to place dollar values on the benefits due to the expected reduced use of alternative programs identified in Table 1.
As with the reduced use of alternative programs, transitional employment programs can be expected to result in a corresponding reduction in the participant's receipt of transfer payments. This would primarily be associated with Social Security Disability programs, but may also include the Medicaid, Food Stamps, and AFDC programs. Table 3 presents estimates of the average payment and administrative cost per month for these programs. As indicated in Table 1, administrative cost savings associated with the reduced use of these programs represent a real savings to society. However, the payment savings serve to benefit taxpayers while imposing losses on participants. Therefore, payments savings represent a redistribution that cancels out from a social perspective.
The task of estimating program costs is usually not as difficult as the task of estimating program benefits. Program operation and administrative costs are fairly straightforward and can easily be obtained from project accounting records In addition to these costs, costs associated with in-program wages received by participants should be evaluated if such costs are incurred and the benefit-cost analysis is cast in the participant-taxpayer framework. The receipt of these wages is a benefit to the participant and a cost to the taxpayer. From the social perspective, however, such in-program payments represent a redistribution that cancels out. On the other hand, as indicated in Table 1, the value of in-program goods and services (output) produced by participants is a real benefit to the taxpayer and society and serves as a partial offset to program costs. This benefit can be estimated based on the revenue generated by the sale of goods and services produced by in-program participants.
Determine Present Value Benefits and Costs
For many social programs, the time at which benefits are received and costs incurred becomes an important consideration. Transitional employment programs may generate benefits and costs over several years. To the extent that they do so, it is important to recognize that a dollar benefits to be received or costs incurred in the future (or in a different year) does not have the same value as a dollar in benefits or costs today. Benefits produced and costs incurred in different years are not directly comparable (cannot be added or subtracted) due to the effects of inflation and the fact that individuals prefer dollars today to dollars in the future. All benefits and costs must be evaluated at their worth in today's (or some base period) dollars. To do this, future benefits and costs are discounted using a discount (or interest) rate to arrive at a present value measure. After discounting benefits and costs. a net present value can be determined by subtracting the discounted (or present value) costs from the discounted (or present value) benefits. Having done this, the project can be evaluated as was done in Section 2. If the net present value is positive, then the program is said to be economically efficient and society is better off with the program than without it.
Table 4 illustrates the concepts of discounting using a hypothetical example of a program in which costs are incurred and benefits produced at different points in time. In general, a cost of y dollars incurred or benefit of y dollars produced n years into the future has a present value of y/(1+d)n where d is the discount or interest rate. It is also important to note that, as illustrated in Table 4, the costs and benefits listed in years 2 and 3 are expressed in real terms (i.e., year 1 dollars) and discounted using a real discount rate (the nominal interest rate less the rate of inflation). As an alternative, one could first list benefits and costs in nominal dollars (i.e., including inflation) and then discount benefits and costs using a nominal discount rate (inflation now included). Either method is appropriate as long as one is consistent in using all nominal values or all real values.
The major issue in discounting future costs and benefits concerns what discount rate to use. The results of the benefit-cost analysis can depend critically on the discount rate selected by the analyst. Because transitional employment programs will usually incur costs in earlier years and produce benefits in later years, higher discount rates (i.e., discounting the future more heavily) will result in a smaller net present value. Thus, higher discount rates will have the effect of making the program appear less attractive while lower discount rates will raise the net present value and will have the effect of making the program appear more attractive.
What is the appropriate discount rate to use? Unfortunately, this is a complex issue which cannot be easily resolved. Most government programs are evaluated using a 10 percent discount rate. It is not uncommon, however, to find other studies based on a three percent discount rate. To help eliminate the uncertainty concerning the appropriate rate to use, one could easily calculate the net present value of the project using a base case rate such as five percent and then check the sensitivity of the results using a lower rate (e.g., three percent) and then a higher rate (e.g., 10 percent). Applications
This section briefly reviews and compares the benefits and costs of two recent transitional employment projects (also see Sav (1986) for a detailed discussion). The first is a benefit-cost analysis of Project Employability in which 90 clients moderately and severely disabled were placed into competitive employment during a 47-month period from May 1978 through July 1982. In a recent study, Hill and Wehman (1983) presented a benefit-cost analysis of Project Employability from the nonparticipant (taxpayer) perspective. Here, the analysis is extended to include the additional social and participant perspectives and to construct the analysis in a framework that allows direct comparison to a second transitional employment project.2
This second project is the Structured Training and Employment Transitional Services (STETS) demonstration. In STETS 284 young adults with IQ scores between 40 and 80 who had no work-disability secondary handicaps received training, support services, and follow-up services designed to transition them to unsubsidized, competitive employment. Under the direction of the Manpower Demonstration Research Corporation, STETS was implemented in five U.S. cities between November 1981 and December 1982. Kerachsky, Thornton, et al. (1985) have examined the impacts of STETS. Here, the benefits and costs of STETS are summarized and structured in a framework that enables interproject comparisons between STETS and Project Employability.
Table 5 lists the dollar benefits and costs of each project from the three perspectives of society, participants, and non-participants. For comparative purposes, benefits and costs are reported on a per participant basis in 1982 dollars. The individual components comprising the benefits and costs are listed from the perspective of society, i.e., under the social column, benefits appear as positive and costs appear as negative.
Both programs produce substantial benefits in terms of output produced by participants and reduced use of alternative programs. Because STETS consisted of several phases in which participants were gradually moved into competitive employment, participant output is reported for both in-program and post-program periods. While to some extent this output decomposition may be appropriate for Project Employability, data were simply not available for comparison as indicated by n.a. in Table 5.
Overall, Project Employability generated significantly greater societal benefits compared to STETS -- 75 percent more benefits per participant. The differential is due almost entirely to the difference in benefits accrued by participants in the two programs, i.e., $3,020 under Project Employability vs. $599 under STETS.3
The participant benefit difference arises from the fact that STETS evaluation includes at most 10 months of post-program competitive employment, whereas Project Employability is evaluated on the basis of potentially 47 months of competitive employment. Thus, post program output differentials are significant. If STETS participants maintained their competitive employment for 37 more months, then the post program output could potentially be on par with Project Employability.
From society's perspective there is little difference in the cost of the two programs ($5,891 vs. $6,657 per participant). Yet, when viewed from the perspectives of the participant and non-participant, the cost burdens are significantly different. Although Project Employability participants do not bear any of the program costs, STETS participants actually receive benefits of $2,669 per participant due to the receipt of in-program payments. From the perspective of the non-participant, STETS imposes nearly 60 percent more cost (per participant) on taxpayers compared to Project Employability.
When program benefits are compared to costs, Project Employability is economically efficient, producing net societal benefits of $4,025 per participant or on the average $1.68 in benefits per participant for each dollar society invests in the program. On the other hand, from a purely monetary evaluation, STETS does not fare as well. From a societal perspective, STETS' costs outweigh benefits by nearly 20 percent, producing a net societal benefit of -$1,039 and a benefit-cost ratio substantially less than 1.0. However, it is important to consider the investment nature of STETS as well as other transitional employment projects similar to STETS. Substantial costs for training people with disabilities are incurred upfront with the expectation that such investments will generate long-term benefits. If STETS and other similar programs are effective in increasing the output and earnings of people with disabilities and reducing their dependence on alternative programs, then it is likely that the long-run net benefits will be positive. Of course, this issue still awaits empirical testing, but the Project Employability results appear to lend strong support to the notion that transitional employment programs are economically efficient from society's perspective. Moreover, the inter-project analysis presented here would have excluded an analysis of the intangible benefits contributed by each project. They are excluded here because it is reasonable to assume that each project would produce similar intangible benefits and therefore make an insignificant difference in the interproject comparison. However, on the basis of an individual project evaluation, such intangibles are important and should at least be qualitatively addressed for public policy purposes. Summary
Benefit-cost analysis provides a technique that is helpful in weighing the advantages and disadvantages of transitional employment programs. This paper has outlined the general technique of benefit-cost analysis and the five steps necessary to conduct a successful benefit-cost analysis. Two examples of how the technique can be successfully applied were provided and discussed. However, the article has only touched on the major issues involved in identifying and measuring benefits and costs and discounting them to obtain a present value measure of the desirability of a project. Unfortunately, each project will likely present a unique set of problems and difficulties associated with measuring benefits and costs. And while the steps required in benefit-cost analysis may seem simple, an adequate analysis may require a great deal of ingenuity. It also requires the combined talents of many disciplines to correctly enumerate and evaluate the benefits and costs of a particular project.
Footnotes (1) The term transitional employment has been used to describe programs differing widely in structure. Strictly speaking, the term refers to short-term interventions that achieve employment in competitive jobs paying at least the minimum wage. In the present article, the term is extended to include supported work where intervention is gradually withdrawn as the program participant acquires greater independence. In the latter case, there are increased tax burdens associated with supported work. (2) Hill and Wehman (1983) provide an excellent analysis of Project Employability. However, their focus on the non-participant (taxpayer) could potentially mislead the reader in the sense that the net benefit of the project reported from the non-participant perspective is approximately one-fourth of the net benefit computed here from the social perspective. Thus, from a purely dollar benefit-cost framework, the program is significantly more efficient than they suggest. 3 In reference to program benefits (in particular, the monetary valuation of post program output and in-program output), one must naturally wonder whether or not participant productivity is at a level commensurate with wages. Any discrepancy will result in a biased estimate of benefits because for empirical estimation it is sometimes necessary to assume that wages (plus fringe benefits) serve as a proxy for the value of increased output to society. Thus, in theory we must assume competitive markets. While this may appear to be a strong assumption, it becomes an empirical necessity in nearly all benefit-cost studies (including the cases presented here), where worker productivity was not directly measured.
Table : Discounting Benefits and Costs: Hypothetical
Table : BENEFIT-COST ANALYSIS OF PROJECT EMPLOYABILITY AND STETS DEMONSTRATION: 1982 DOLLARS PER PROGRAM PARTICIPANT
|Printer friendly Cite/link Email Feedback|
|Author:||Sav, G. Thomas|
|Publication:||The Journal of Rehabilitation|
|Date:||Apr 1, 1989|
|Previous Article:||Implementing a computerized job-matching program with a hierarchical occupational structure.|
|Next Article:||A qualitative study of work stations in industry: employer, professional and service recipient views.|