Printer Friendly

Using benefit-cost analysis in special education.

Using Benefit-Cost Analysis in Special Education

Are we "getting our money's worth" from special education programs in the public schools? With the increasing application of benefit-cost analysis to other social service programs, the public has come to expect that similar economic analysis can somehow be applied to special education.

The "need" for special education is well established in the literature and political arenas, based upon concerns regarding the importance of providing children and youth with handicapping conditions access to social and learning opportunities similar to those afforded their nonhandicapped cohort. Most analysis of the merits of special education services focuses upon issues of access, equal opportunity, features of program models, and evaluation of program effects. The efficiency and productivity of special education services has received only minimal attention from policy makers, researchers, or practitioners. Although some recent attention has been directed at examining the costs of special education (see, for example, the review on this issue in Lewis, Bruininks, & Thurlow, 1987), little attention has been given to linking these costs with outcomes. To date, not a single study has attempted to systematically examine aspects of school-based special education services from a program efficiency or benefit-cost perspective.

On the other hand, during the past decade there has been a number of benefit-cost applications to postschool employment training alternatives for young adults with developmental disabilities. Recent reviews of these efforts are provided by Nobel and Conley (1987) and Rhodes, Ramsing, and Hill (1987).

Especially noteworthy has been the work by Cronin and Cuvo (1979); Hill, Hill, Wehman, and Banks (1985); Hill and Wehman (1983); Hill, Wehman, Kregel, Banks, and Metzler (1987); Kerachsky, Thornton, Bloomenthal, Maynard, and Stephans (1985); Lam (1986); Rhodes (1982); Schneider, Rusch, Henderson, and Geske (1981). Successful applications of benefit-cost analysis to other educational and social service programs can also be found in the broader literature (see, for example, Kemper, Long, & Thornton, 1981; Psacharopoulos & Woodhall, 1985; Thompson, 1980; Weber, Foster, & Weikert, 1978; and Weisbrod, 1981). Beyond employment data, little information of a systematic nature has been collected on the economic benefits of school-based special education programs. Admittedly, the evaluation of special education services for children and youth with handicaps is complex and should be focused upon the broadest form of analysis. Nevertheless, one useful, but by no means complete, strategy of special education evaluation involves the development of program outcome information in economic terms and more systematic benefit-cost assessments.

In most basic terms, benefit-cost analysis is concerned with assessing the relative economic efficiency of alternative programs through comparing their benefits against their costs of services. Thornton (1984) succinctly describes the primary aspects of this analytic approach:

The basic technique . . . is to assign dollar values to all estimated effects and costs. These values are them summed together to yield an estimate of the program's net present value (i.e., the difference between the benefits and costs where all dollar values are adjusted to reflect their value in a specific base period). This process is called "discounting" and is used to adjust the value of benefits or costs that accrue over several time periods to reflect their value at a specified base period. . . . A positive net present value indicates that resources are being used efficiently. A negative net present value indicates that (at least at its current scale) the program's resources could be used more efficiently elsewhere. (p. 226)



The basic question being addressed in this article is whether formal benefit-cost analysis might allow us to determine whether the monetary outcomes of particular special education services for school-age children and youth are worth their resource costs. In attempting to answer this economic question our focus is upon measuring as many of the outcomes as possible in monetary terms and in illustrating the relevance and value of this evaluation methodology to the field of special education.

As in all benefit-cost analyses, the analyst must first specify the program being evaluated and the comparison against which it will be judged. Second, all costs and benefits must be identified in an appropriate accounting framework. Third, the analyst must value (or measure) all possible costs and benefits for subsequent analysis.

Since benefit-cost analysis attempts to assess all effects and alternatives in terms of monetary cost and benefit values, pecuniary measurement becomes an obvious and very challenging hurdle. This issue has been particularly acute in the field of special education, where traditionally most benefits have been assumed to be based upon achieving important social and educational values and, therefore, to be largely unmeasurable in monetary or economic terms. This is undoubtedly why many evaluators of special education services have never attempted this technique.

The application of benefit-cost analysis within educational programs has also generated concern about a number of critical assumptions concerning the use of discount rates, length or period(s) of time for estimating benefit streams, and, most critically, the alternative or counterfactual comparison group(s) against which the treatment or program is being compared. In every formal study which employs this form of analysis, critical assumptions need to be made concerning these matters; they sometimes can and do lead to inconsistent or varying results.

The most serious design problem arises in developing an alternative for comparing the costs and benefits resulting from current special education services. In the field of special education, for example, how does one design a comparison group to special education when both statutory law and social values preclude withholding such services for controlled experimental treatment? In reality, in most special education situations only one program is in place, with no observable alternatives available for comparison purposes. If two or more different programs were in place and available to serve a similar population, then benefit-cost comparisons could be made across the two program groups for assessing relative efficiency. On the other hand, if one wants to assess the economic efficiency to society of a particular program in special education, then such controlled design alternatives are indeed precluded. To examine the latter issue with benefit-cost techniques, it is necessary to employ a post-hoc nonexperimental comparison design with hypothetical rather than actual treatment alternatives. This approach obviously poses difficulties for structuring any study of costs and benefits, and limits to some extent the conclusions that can be drawn from the analysis. Despite such limitations, the use of a benefit-cost paradigm can still yield a number of useful results for researchers, administrators, and policy makers.

In the opinions of many in special education the use of benefit-cost analysis as an evaluation technique to examine the worth (i.e., economic efficiency to society) of special education is moot from the outset. Attempts to examine the value of special education services in economic terms are often viewed as irrelevant, since society is required by both law and existing social values to proceed with such services in any case. Unfortunately, such views miss an important point for at least two reasons. First, benefit-cost analysis can be used to assess both the relative economic value (i.e., efficiency) of providing special education services versus no school-based services and the relative economic value of two or more alternative school-based programs. Both applications give us understandings about the efficient uses of resources in society. There appears to be an inherent bias in the field of special education against examining efficiency considerations, except at the most global expenditure levels. Almost all evaluation efforts in special education focus on questions of maximizing effectiveness, with little or no attention to assessing and linking cost considerations to outcomes.

Second, it is entirely possible that special education might be an economically efficient use of society's resources. Even if negative net economic benefits to society result from special education, policy makers need such information for informed judgments about the consequences of alternative investments to achieve the same public values. Even the process of conducting benefit-cost evaluation can yield important results for structuring and managing resources and programs.

For purposes of illustrating some of these problematic issues and examining prospective uses of this evaluation technique in special education, we have adapted some preliminary empirical data relative to the costs and outcomes resulting from school-based special education services for a sample of youth with mild retardation. Specifically, this article (a) identifies a conceptual framework wherein special education costs and benefits can be comprehensively described, valued, and linked for analytic purposes; (b) presents some benchmark empirical data from a large suburban school district case study relative to the economic costs and outcomes resulting from services for a sample of youth with mild retardation; (c) examines a number of alternative benefit-cost assumptions for illustrating some of the problematic issues identified previously in the use of such procedures; and (d) provides some discussion of the utility of such analyses for improving special education services.


The data employed for the benefit-cost illustrations in this article are drawn from a larger study dealing with the costs and follow-up benefits of special education programs in public schools (Bruininks, Lewis, & Thurlow, 1987). The larger study was a follow-up study of all students who graduated or completed at least 12 years of schooling between the years 1977 and 1984 from two high schools in a large suburban school district in Minnesota. Drawing from data collected in the larger study, certain special education costs and benefits were identified and valued for a sample of former students classified as mildly retarded.

School record information was obtained from students' cumulative files and from special education files. Follow-up outcome data on former students were obtained during the spring and summer of 1984 using a mail questionnaire and a phone interview procedure.

The sample used for data in this article originally included all 52 students with IQs of 80 and below who were classified as mildly retarded and who had completed the 12th grade between the years of 1977 and 1983. The follow-up response rate with complete outcome information was 54% of all students with mild retardation in the sample and totaled 28 young adults almost equally divided between males (54%) and females (46%). The final sample of respondents had a mean IQ of 72.6 (SD = 5.3) as compared with a mean IQ of 73.8 (SD = 5.2) for nonrespondents, which was not a statistically significant difference. Additional detail relative to sample characteristics, statistical tests, and follow-up results are reported in Bruininks et al. (1987).

All students in the sample were mainstreamed in the schools with hourly special education services which averaged 2 hours per day over their 12 plus years of schooling; there were no statistically significant differences between the extent (i.e., total hours) of services received by respondents and nonrespondents in the special education populations in the original study. Almost all of the students in the sample also participated in the vocational (91%) and work (72%) programs in the schools.

The rate of employment for the respondents in the sample was relatively high at 62% (with 17% of those employed holding part-time jobs) when compared to rates of 43% (Edgar & Levine, 1986), 58% (Schalock, Wolzen, Ross, Elliott, Werbel, & Peterson, 1986), and 60% (Wehman, Kregel & Seyfarth, 1985) reported by others for similar populations with mild retardation. A comparable employment rate reported by the Minnesota Department of Jobs and Training for all young adults between the ages of 20 and 24 in the same metropolitan area was 76% in 1980. The average number of hours worked per week was 27 hours (SD = 13.8) for all respondents, while the median years out of school was 4 (SD = 2). The average annual earned income for employed individuals was $6,475 with a median of $5,295 (SD = $4,649); average annual transfer payments (supplemental security income, medicaid assistance, and food stamps) for all respondents was estimated to be $427. The sample presented evidence of relatively high community adjustment. Most of the sample lived with family members (60%) or independently (25%), while only 15% were living within a supervised group or foster home. Moreover, a majority of the respondents reported having a driver's license (58%) and having taken a vacation in 1984 (52%), while 44% reported having checking accounts.



Use of a benefit-cost evaluation framework to evaluate programs requires several important steps. First, the program being evaluated needs to be identified along with its alternative comparison(s). Second, an appropriate accounting framework must be developed for identifying all costs and benefits. Third, the costs and benefits need to be measured and valued. Finally, the analyst needs to examine a number of alternative assumptions in the accounting framework to test for their likely effects on the results.

A conceptual framework for benefit and cost comparisons of special education for former students with mild retardation can be constructed if it is assumed that appropriate costs and benefits can be measured and valued for a similar hypothetical sampel that received no special education services. Such an alternative is illustrated and incorporated within the framework of Table 1.

The benefit-cost accounting framework as presented in Table 1 draws heavily upon the framework outlined by Thornton and Will (1986). The framework compares a school-based special education program for individuals with mild retardation with a hypothetical situation such as prevailed in the early part of this century. At that time institutional care was the primary alternative to independent living for mentally retarded adults and special education was extremely limited in the schools. This approach identifies the analytical perspectives of interest to both students and society. It is important to note that this model provides insight into not only those benefits and costs that can be monetized, but also into those effects that cannot be measured in dollars alone. It notes, for example, important other nonmonetary benefits such as preferences for work and prospects for increased self-sufficiency, self-esteem, social integration, and quality of life.

A number of other alternative hypothetical comparison groups were constructed for purposes of illustrating the use of benefit-cost analysis in examining the economic worth question of special education. These hypothetical comparison groups were constructed largely through developing historical data to meet different assumptions. In the most extreme case, for example, one could assume that the "eugenics movement" during the early part of this century was successful in requiring lifetime institutionalization for many individuals with mental retardation. As a second alternative, one could assume that due to technological and occupational changes many individuals with mental retardation require the benefits specific to special education to be employable in today's competitive labor markets. Or third, one could assume that without the extraordinary attention and assistance through special education services most individuals with mental retardation would drop out of school prior to graduation. Differing costs and benefits would derive, of course, from each of these hypothetical alternatives.

Other comparison groups and hypothetical alternatives could also have been developed and illustrated, but the three presented in this article were selected because they most vividly illustrate both the critical nature of our assumptions and the value and flexibility of using benefit-cost techniques to evaluate special education.

The present study used a comprehensive program components approach for identifying, measuring, and valuing the costs of special education services. Estimates for all school-based costs were taken from an earlier cost study in the same school district by Lewis, Bruininks, and Thurlow (1987) and expressed in 1984 present values. Cost estimates were based on all of the resources employed in the delivery of both regular and special education services and were reported for 13 special education service areas and grade levels. Both regular and special education service area costs were representative of other metropolitan school district costs and consistent with similar cost data reported by other studies (Kalalik, Furry, Thomas, & Carney, 1981; Raphael, Singer, & Walker, 1985).

Economics of the Eugenics/Institutionalization


For purposes of illustration it is possible to assume the hypothetical alternative of offering no special education services in the schools and the institutionalization of all individuals with mental retardation at approximately 14 years of age. This hypothetical situation has some basis in reality when one reflects on the status of special education for students with mental retardation in the United States at the turn of the century. During the early 1990s, decisions concerning citizens with mental retardation were most often made on philosophical and political grounds, as opposed to empirically based research or economic considerations. Craig and McCarver (1984) point out in their historical perspective on deinstitutionalization that in the early 1990s citizens with mental retardation were regarded as a menace to society, based on the prevailing notion that moral and mental defectiveness were linked.

It was believed that persons with mental retardation should be isolated from the community in institutions. In a paper delivered in 1912 by a leading professional to the American Association for the Study of the Feebleminded, it was stated that persons with mental retardation were "a parasitic and predatory class never capable of supporting themselves or of managing their own affairs . . . a menace and danger to the community . . .a potential criminal" (Fernald, 1912, p. 88). Moreover, historical evidence on persons in institutional settings shows very high rates for mildly and moderately retarded individuals (Scheerenberger, 1983). Given that over 85% to 90% of all persons with mental retardation are either mildly or moderately handicapped (Grossman, 1983), it is safe to assume that many such individuals in our schools today would have been targeted for such beliefs and subsequent institutionalization during this earlier period.

After more than 50 years of increasing institutionalization, a peak period was reached in 1965 (Lakin, Hill, Street & Bruininks, 1986). Current policy has dramatically shifted to one of deinstitutionalization. Recent and current deinstitutionalization efforts have been incorporated into law and practice, and in many states it is very nearly an accomplished fact. This important trend was accompanied by increased efforts to provide special education and other services in community settings (Lakin, Hill, Hauber, & Bruininks, 1982).

If institutionalization with its attendant and exceedingly high costs is viewed as the hypothetical comparison, the resulting benefit-cost analysis would clearly favor special education in the schools and deinstitutionalization even if post-school competitive earnings were zero. In Minnesota, the 1984 average annual per capita institutional care costs for persons with mental retardation was $44,986 (Lakin et al., 1986). As noted in Table 1, when these costs, with their attendant assumptions, were factored into the accounting framework, the net per capita monetary benefit to society for school-based special education was conservatively estimated to be $667,927 in 1984 present value.

In Table 1 we are assuming a hypothetical comparison group of individuals with mental retardation who received no special education and who would probably have been institutionalized at approximately age 14 with no lifetime earnings. Prior to age 14 they would have participated in regular school instruction. This is obviously the most extreme comparison for evaluating the possible benefits of special education services. This hypothetical alternative assumes the most pessimistic political and social outcome for youth and adults without special education services. The model and illustration presumes that the availability of local special education services would operate to prevent possible institutionalization and that the sampel would live with family members or independently. With these conditions and critical assumptions, data for the algorithm have been derived and computed for the benefit-cost illustration in Table 1.

Accounting for Benefits and Costs

With respect to the estimated impact on benefits, increased earnings represent average annual earnings of all respondents with mental retardation in the sample ($5,319). An assumption of no earnings is made for those institutionalized. When extrapolated over a work-life of 35 years and discounted at 6%, these earnings totaled $77,115 in 1984 present value for each respondent. Increased fringe benefits represent annual gross earnings multiplied by a factor of 15% (U.S. Department of Labor, 1980). Increased taxes represent annual gross earnings multiplied by a factor of 23% (Pechman, 1985). Work preferences are expressed as unmeasured but clearly represent a positive outcome of most individuals and public opinion. Institutional care costs represent average annual per-capita costs of care in Minnesota state-operated residential facilities for persons with mental retardation ($44, 986), extrapolated over a life-span of 44 years beyond the age of 14 (Balakrishnan & Wolf, 1976) and then discounted into present value.

When institutionalization is prevented, the state (i.e., taxpayers) saves the money that would have been spent on that person in the institution (measured by the average daily cost of care in that setting). However, the costs of basic board, room, and care for that individual must now be paid by someone else within the community, usually the person or their family. These basic home care costs have been estimated to be $2,346 per year (U.S. Department of Labor, 1980). When projected over a life-span and discounted into present value, these costs were estimated to be $36,081 for each individual or family. Other benefits are expressed as unmeasured in this monetary algorithm, but represent current research findings and professional consensus relative to the positive benefits of community residency and employment of persons with mental retardation.

With respect to estimated costs, special education costs represent average annual costs for special education services for students with mental retardation ($3,652) compounded over 12 years and expressed in present value terms. All of the students with mild retardation in the sample who received special education services were also mainstreamed within the regular school

curriculum. Regular instruction costs represent the added costs of secondary education for the special education students in our sample who remained in school past 14 years of age. Use of social services is unmeasured in this framework because many of the survey responses were ambiguous on these questions and the resulting small sample did not produce reliable results. Undoubtedly there are some modest resource costs, but it was not possible to estimate their amounts from the available data. Increased use of transfer programs represents the increased use of medicaid and SSI when individuals with mental retardation are living within the community.

Altering the Rate of Institutionalization

In a separate analysis (Bruininks et al., 1987), the assumptions of Table 1 were modified regarding the possible rate of institutionalization for this sample. As an additional alternative, it was assumed that the hypothetical comparison group again received no special education in this school district; however, it was also assumed that the comparison group of youth with mental retardation were not completely institutionalized, but rather we institutionalized at the peak 1965 rate only after completing school. It was further assumed that the comparison group (a) received no special education, (b) did not drop out of school but completed all 12 years solely in a mainstreamed course of regular instruction, (c) had the same employment and earning prospects as those with special education, and (d) had no other additional need or access to the use of social services in the community.

From these assumptions, it was estimated that at least 1 of the 28 respondents in the sample and living in the community today would previously have been institutionalized. This estimate is based on the 1965 rate of institutionalization of 1 out of 17 (see historical rates in Lakin et al., 1986) as compared to the current rate of institutionalization of 1 out of 40. With 62% of the current sample gainfully employed, one can also estimate that this institutionalized individual probably would have been employed. Given these changed assumptions, one can re-estimate the net benefits in monetary terms. Implementation of these assumptions for our hypothetical comparison group of 28 individuals with mental retardation indicates that the hypothetical group, without access to special education services, is favored on a per capita basis by $35,020.

The assumptions used in this latter analysis present another extreme case for comparative and illustrative purposes relative to examining the possible benefits of special education. This latter model assumes, for example, that the only monetized benefit to providing special education for the entire sample of respondents results solely from the added employment of one individual and the cost savings from not institutionalizing one hypothetical individual. It assumes that no other monetized benefit (e.g., drop-out prevention, employability, increased earnings, and the like) results from special education. Even with these extreme hypothetical assumptions, almost one-half of special education costs are saved for the entire sample by preventing institutionalization for only one person.

These results point out the exceedingly high costs of institutionalization for persons with mental retardation. Assuming that special education services can prevent institutionalization, the results also point out the exceedingly high economic efficiency effect gained from the provision of special education services and deinstitutionalization. From these data one can estimate that if only 2 more (for a total of 3) of the hypothetical cohort of 28 had been prevented from being institutionalized, the per-student net monetary benefits to society would become positive by $18,157, even if the sample cohort still bore all of the special education program costs. It is not unreasonable to assume that special education has had some influence on those factors that lead society to make decisions regarding institutionalization for this population. From the results in this illustration one could estimate that if special education prevents at least 3 individuals from the hypothetical cohort of 28 (i.e., 1 in 10) from becoming institutionalized, it is cost-beneficial in monetary terms alone.

Economics of Special Education Dropouts and

Technological Unemployment

The occupational picture for persons with mental retardation is far from encouraging even with provision of special education services. Occupational categories once occupied by significant proportions of persons with mental retardation have been decreasing markedly over the past half-century. Experts predict that this trend will persist, although probably at a reduced rate (Rumberger, 1984). Much of the disappearance of jobs for persons with mental retardation can be attributed to technological changes brought on by automation and mechanization in almost all occupational categories, including the service industries. Traditional employment for individuals on farms and in homes has disappeared almost completely. Although there has been a recent proliferation of low wage employment in the United State economy (Blueston & Harrison, 1986), there has also been a decline in the occupational categories that have historically employed persons with mental retardation. With the growing underclass in our society resulting high school dropouts, it has been estimated that competition for these few jobs will become even more keen in the future.

Until recently, special education programs for persons with mental retardation beyond the elementary level were limited and "most dropped out of school at an early age" (Conley, 1973, p. 289). Four decades ago, Kennedy (1948) reported that fewer than 9% of her sample of persons with mild retardation went beyond the eighth grade. As a consequence of these conditions, it is plausible to assume that without special education most individuals with mental retardation would drop out of school before normal completion. It is also possible to construct employment and earnings records of a hypothetical dropout group for comparison purposes.

Data from other studies indicate that the unemployment rates of underclass school dropouts have been a high as 50% to 80% (Wolman, 1987). Earnings for these groups similarly have been found to be as low as 30% of the regular population earnings rate. If, for example, the rates of employment and earnings for dropouts were only 50% of those with special education and high school completion, and at least four of the dropouts required increased use of group homes or other supervised living arrangements, then the accounting framework illustrated in this article provides us with estimates of per capita net benefits for special education of at least $10,000 in 1984 present value. Lifetime per capita group home costs were estimated to be $331,331 in present value based upon available statistics provided by Greenberg, Lakin, Hill, Bruininks, and Hauber (1985).

Recent reports from U.S. census data (O'Neill & Sepielli, 1985) indicate that for members of the regular population the ratio of income for high school completers relative to high school dropouts of 25- to 34-year-old males was 1.47 in 1983, and that this ratio has been growing over time. Assuming a similar earnings disadvantage for persons with mild retardation without special education and with high secondary school dropout rates, one can modify our accounting framework and again estimate the likely benefit-cost effects. With such high dropout rates it is also probable that at least 20% (or 6) more individuals would need increased use of social services such as supervised living arrangements. When these altered assumptions were examined, per capita net benefits were estimated to exceed $14,000 in present value.

The Minnesota Sample Compared

with Data from Vermont

A recent study by Burchard and her colleagues (1986) in Vermont provides a real sample comparison group for this analysis. They reported data from a sample of 109 adults with mental retardation (both men and women between the ages of 23 and 55) living in the community with 75% residing in group homes or supervised apartments. It is important to note that 30% of their original sample was represented by adults with moderate retardation, whereas the Minnesota sample used in this study included only individuals with mild retardation. Their total sample reported an average annual income for $4,665 in 1984 dollars. With average employment rates of only 26% reported for this adult population, it is assumed that these annual incomes included SSI and other transfer payments as well as earned income. Because most of these adults were in school over 20 to 30 years ago, it is also reasonable to assume that many of these individuals had received, at best, only limited special education services.

Because of 30% of the Vermont sample was represented by individuals with moderate mental retardation, adjustments for the exclusion of these individuals are necessary in both earnings and community living arrangements. For purposes of adjustments in the data, it was assumed that the full cohort of individuals with moderate retardation in Vermont were living in supervised facilities and were not employed. As a result, the Vermont employment rates were adjusted upward to 36% and family living arrangements were adjusted upward to 26% when all individuals with moderate retardation were excluded from the sample.

In Table 2 the results from this group are compared with those from the Minnesota sample. Drawing on data from the Vermont sample, the model assumes (a) no special education costs for this comparison group, (b) all Vermont respondents completed high school, (c) the competitively employed individuals in Vermont received no extra income from SSI while the unemployed received their full entitlement, (d) the employed Vermont respondents received the same level of average earnings as the Minnesota sample, and (e) community support services were the same for both groups. All of these assumptions make the economic analysis very conservative for estimating possible benefits from special education services in the Minnesota sample.

With respect to the estimated impact on benefits, increased earnings represent the difference between the two samples in average annual earnings for all respondents. If one assumes that the employed respondents in Vermont (adjusted 36% of sample) made the same average annual earnings ($6,475) as those employed in Minnesota, the Vermont earnings would average $2,331 for all resodents. The difference between the Vermont sample average and the Minnesota sample was estimated to be $2,988, and when extrapolated over a work-life of 35 years and discounted into present value, these earnings totaled $43,296. SSI was similarly adjusted for the differences in employment rates and earnings; the reduced use of supervised living facilities was estimated to be 59% with 16 individuals no longer using group or supervised apartments.

As a result of these assumptions and adjustments, estimations from Table 2 indicate that those who received special education during the past 12 to 15 years in Minnesota had average per capita net benefits of $157,351 in 1984 present value. It is important to note that in this comparison model, if no adjustments or cost saving estimates had been made for different community living arrangements as result of special education, the net benefits would be negative by $11,819. On the other hand, if special education assists only one individual from the entire sample to move into independent living arrangements, it is estimated that cost savings would be over $12,000 in present value and result in positive net benefits.

It is obvious that the use of the Vermont sample, even with adjustments for the presence of persons with moderate retardation, is still not without major design problems. The Vermont sample, even with the adjustments, could still be more handicapped in employment and community living skills. Moreover, the per capita income and employment rates in the two states may differ. On the other hand, the older age of the Vermont cohort provided more time for enhancing their employment status, a primary consideration in employment rates and earnings. The value of the comparison rests upon illustrating the potential of special education services to increase school retention, reduce dependency in supervised living, and enhance prospects for community living and competitive employment. These positive outcomes, in turn, yield reduced use of public resources and increased educational productivity.


This article has focused on the use of a benefit-cost conceptual framework in estimating the benefits and costs of a public school special education program for persons with mild retardation. Results indicate that with appropriately identified, measured, and valued costs and benefits, it is possible to employ a formal benefit-cost framework to assess the efficiency of special education services. Such a model provides insight into not only those benefits and costs that can be monetized, but also into many other effects that cannot be valued in terms of money. It notes, for example, important other benefits such as work preferences and prospects for increased self-sufficiency, self-esteem, and quality of life.

In the context of this model, along with data collected from a local school district and subsequent follow-up surveys and interviews, the basic question examined was whether special education might be worth its cost when compared with a number of hypothetical comparison groups presumed to be without the benefit of special education services.

Unfortunately the results of these comparisons are severely limited for making reliable generalizations because of design problems with the comparison groups. Nevertheless, the illustrations presented do give vivid detail to the importance of both the technique and many of its critical assumptions. Moreover, the illustration do give us tentative estimates that special education services for persons with mild retardation may be cost-beneficial when compared with a number of alternatives. When historical data were used for hypothetical counterfactual comparison groups, it was possible to examine a number of alternative hypotheses concerning the likely postschool effects for individuals without special education services. Various rates of institutionalization, school dropout, and unemployment were examined as hypothetical comparisons to provision of special education services for a sample of youth with mild retardation. The resulting benefit-cost estimates, despite some obvious methodological problmes, indicate the likely economic efficiency of special education for the individuals in or sample.

Clearly more work needs to be done with the use of benefit-cost analysis in examining these questions. The most promising applications of this approach would involve comparisons of alternative treatments or strategies in which individuals have been assigned randomly to different conditions (see, for example, Kerachsky et al., 1985). Although a weaker paradigm, some useful results can accrue from the application of benefit-cost analyses using hypthetical comparisons, provided that historical data are available to assess possible alternatives. The principal benefit to special education is the focus on the usage of resources in relationship to outcomes. Such analyses force administrators and policy makers to address questions of resource usage in relation to expected postschool benefits for students with handicaps.


Balakrishnan, T. R., & Wolf, L. (1976). Life expectancy of mentally retarded persons in Canadian institutions. American Journal of Mental Deficiency,

80(6), 650-662.

Bluestone, B., & Harrison, B. (1986). The great American job machine: The proliferation of low wage employment in the U.S. economy (A study prepared for the Joint Economic Committee). Washington, DC: Joint Economic Committee, U.S. Congress.

Bruininks, R., Lewis, D., & Thurlow, M. (1987). Assessing outcomes, costs, and benefits of special education programs. Minneapolis: Minnesota University Affiliated Programs, University of Minnesota.

Burchard, S. N., Hasazi, J., Gordon, L., Rosen, J., Yoe, J., Toro, H., Dietzel, L., Payton, P., & Simoneau, D. (1986). The community adjustment and integration of adults with mental retardation living in group homes, supervised apartments, and with their families. Burlington, VT: Department of Psychology, University of Vermont (Mimeograph).

Conley, R. W. (1973). The economics of mental retardation. Baltimore: The Johns Hopkins University Press.

Craig, E. M., & McCarver, R. B. (1984). Community placement and adjustment of deinstitutionalized clients: Issues and findings. In N. Ellis and N. Bray (Eds.), International review of research in mental retardation (pp. 95-122). New York: Academic Press.

Cronin, K., & Cuvo, A. (1979). Teaching mending skills to mentally retarded adolescents. Journal of Applied Behavior Analysis, 12, 401-406.

Edgar, E., & Levine, P. (1986). Follow-up of Washington State special education graduates 1978-1984. Unpublished paper, University of Washington, Seattle.

Fernald, W. (1912). The burden of feeble-mindedness. Journal of Psycho-Asthenics, 17, 87-111.

Greenberg, J., Lakin, C., Hill, B., Bruininks, R., & Hauber, F. (1985). Costs of residential care in the United States. In C. Lakin, B. Hill & R. Bruininks (Eds.), An analysis of Medicaid's intermediate care facility for the mentally retarded (ICF-MR) program (pp. 7.1-7.82). Minneapolis: Center for Residential and Community Services, University of Minnesota.

Grossman, H. J. (Ed.) (1983). Manual on terminology and classification in mental retardation. Washington, DC: American Association of Mental Deficiency.

Hill, M., Hill, J. W., Wehman, P., & Banks, P. D. (1985). An analysis of monetary and nonmonetary outcomes associated with competitive employment of mentally retarded persons. In P. Wehman & J. Hill (Eds.), Competitive employment for persons with mental retardation: From research to practice (pp. 110-133). Richmond: Rehabilitation Research and Training Center: Virginia Commonwealth University.

Hill, M., & Wehman, P. (1983). Cost-benefit analysis of placing moderately and severely handicapped persons into competitive work. Journal of the Association for the Severely Handicapped, 10, 25-40.

Hill, M., Wehman, P., Kregel, J., Banks, P. D., & Metzler, H. (1987). Employment outcomes for people with moderate and severe disabilities: An eight-year longitudinal analysis of supported competitive employment. Journal of the Association for the Severely Handicapped, 12, 182-189.

Kalalik, J. S., Furry, W. S., Thomas M. A., & Carney, M. F. (1981). The cost of special education. Santa Monica, CA: Rand Corporation.

Kemper, P., Long, D., & Thornton, C. (1981). The supported work evaluation: Final benefit-cost analysis. New York: Manpower Demonstration Research Corporation.

Kennedy, R. J. (June 1948). The social adjustment of morons in a Connecticut city. Storrs: Commission to Survey the Human Resources of Connecticut.

Kerachsky, S., Thornton, C., Blomenthal, A., Maynard, R., & Stephans, S. (1985). Impacts of transitional employment on mentally retarded young adults: Results of the STETS demonstration. Princeton: Mathematica Policy Research.

Lakin, K. C., Hill, B. K., Hauber, F. A., & Bruininks, R. H. (1982). Changes in ages of first admission for mentally retarded people. Mental Retardation, 20, 216-219.

Lakin, K. C., Hill, B. K., Street, H., & Bruininks, R. H. (October 1986). Persons with mental retardation in state-operated residential facilities. Minneapolis: Center for Residential and Community Services, University of Minnesota.

Lam, C. S. (December 1986). Comparison of sheltered and supported work programs: A pilot study. Rehabilitation Counseling Bulletin, pp. 66-82.

Lewis, D. R., Bruininks, R. D., & Thurlow, M. (1987). Cost analysis for district level special education planning, budgeting and administrating. Minneapolis: Department of Educational Psychology, University of Minnesota.

Noble, J. H., & Conley, R. (1987). Accumulating evidence on the benefits and costs of supported and transitional employment for persons with severe disabilities. Journal of the Association for Persons with Severe Handicaps, 12, 163-174.

O'Neill, D. M., & Sepielli, P. (1985). Education in the United States: 1940-1983 (CDS-85-1, Bureau of the Census, U. S. Department of Commerce). Washington, DC: Government Printing Office.

Pechman, J. (1985). Who paid taxes, 1966-85. Washington, DC: Brookings Institute.

Psacharopoulos, G., & Woodhall, M. (1985). Education for development: An analysis of investment choices. New York: Oxford University Press.

Raphael, E., Singer, J., & Walker, D. (1985). Expenditures on special education in three metropolitan school districts. Journal of Education Finance, 11, 69-88.

Rhodes, L. (1982). Alternative investment analysis of services for severely handicapped people. Unpublished doctoral dissertation, University of Oregon, Eugene.

Rhodes, L., Ramsing, K., & Hill, M. (1987). Economic evaluation of employment services: A review of applications. Journal of the Association for Persons with Severe Handicaps, 12, 175-181.

Rumberger, R. W. (1984). High technology and job loss (Project Report No. 84-A12) Stanford: Institute for Research on Educational Finance and Governance, School of Education, Stanford University.

Schalock, R. L., Wolzen, B., Ross, I., Elliott, B., Werbel, G., & Peterson K. (1986). Post-secondary community placement of handicapped students: A five-year follow-up. Learning Disability Quarterly, 9, 295-303.

Scheerenberger, R. C. (1983). A history of mental retardation. Baltimore: Brookers Publishing.

Schneider, K., Rusch, F., Henderson, R., & Geske, T. (1981). Competitive employment for mentally retarded persons: Costs vs. benefits. In W. Halloran (Ed.), Funding and cost analysis. Urbana-Champaign, IL: Leadership Training Institute, Vocational and Special Education, University of Illinois.

Thompson, M. (1980). Benefit-cost analysis for program evaluation. Beverly Hills, CA: Sage Publications.

Thornton, C. (1984). Benefit-cost analysis of social programs. In R. H. Bruininks & C. K. Lakin (Eds.), Living and learning in the least restrictive environment. Baltimore: Paul H. Brookes.

Thornton, C., & Will, J. (January 1986). Benefit-cost analysis and special education programs. Minneapolis: Department of Educational Psychology, University of Minnesota.

U. S. Department of Labor. (1980). Employee compensation in the private non-farm economy. Washington, DC: Bureau of Labor Statistics.

Weber, C., Foster, P., & Weikert, D. (1978). An economic analysis of the Ypsilanti Perry preschool project (Monograph #5). Ypsilanti, MI: High Scope Educational Research Foundation.

Wehman, P., Kregel, J., & Seyfarth, J. (1985). Employment outlook for young adults with mental retardation. Rehabilitation Counseling Bulletin, 29, 90-99.

Weisbrod, B. A. (Fall 1981). Benefit-cost analysis of controlled experiment: Treating the mentally ill. Journal of Human Resources, 16, 523-548.

Wolman, C. (1987). Special education drop outs: A literature review. Minneapolis: Department of Educational Psychology, University of Minnesota.

DARRELL R. LEWIS is Professor of Educational Policy and Administration; ROBERT H. BRUININKS is Professor of Educational Psychology; MARTHA THURLOW is Research Fellow in Educational Psychology; and KEVIN MCGREW is Research Assistant in Educational Psychology, University of Minnesota, Minneapolis.
COPYRIGHT 1988 Council for Exceptional Children
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1988 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Lewis, Darrell R.; Bruininks, Robert H.; Thurlow, Martha; McGrew, Kevin
Publication:Exceptional Children
Date:Nov 1, 1988
Previous Article:Cultural diversity and the test of leadership.
Next Article:Special education teacher preparation: a synthesis of four research studies.

Related Articles
Integrated classroom versus resource model: academic viability and effectiveness.
Supplemental/replacement: an alternative approach to excess costs.
Relationships among states' fiscal and demographic data and the implementation of P.L. 94-142.
20/20 Analysis: taking a close look at the margins.
Counterpoint: "20/20 Analysis" - Phase 1 only.
Inclusive education: what do we know and what do we still have to learn?
20/20 analysis: estimating school effectiveness in serving students at the margins.
Comparing instructional contexts of students with and without severe disabilities in general education classrooms.
Why we can't say much about students with disabilities during education reform.
The Quality of Life Argument in Public Funding of Special-Needs Students in North Carolina.

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters