Summary of Findings and Recommendations

Upjohn Press Book Chapters Upjohn Research home page 1981 Summary of Findings and Recommendations Sar A. Levitan George Washington University Garth L. Mangum University of Utah Citation Levitan, Sar A. and Garth L. Mangum. 1981. "Summary of Findings and Recommendations." In The T in CETA: Local and National Perspectives, Sar A. Levitan, and Garth L. Mangum, eds. Kalamazoo, MI: W.E. Upjohn Institute for Employment Research, pp. 1–92. http://research.upjohn.org/up_bookchapters/735 This title is brought to you by the Upjohn Institute. For more information, please contact ir@upjohn.org. Summary of Findings and Recommendations Sar A. Levitan and Garth L. Mangum The basic premise of the Comprehensive Employment and Training Act (CETA) was that local decisionmakers could design and deliver services more appropriate to their local economies and populations than any nationally uniform pro gram. Therefore, the nature, quality, and effectiveness of the programs offered by 476 prime sponsors manifests great diversity. Nevertheless, it is possible to make some generalizations from examination of the national enroll ment, costs, and outcomes data and from intensive study of the training activities of 12 prime sponsorships. First, the conclusions and then the supporting arguments: 1. CETA training is a sound social investment. The na tional data suggest that the social returns for each dollar spent on CETA-funded classroom training amount to $1.14, while each dollar spent for on-the-job training (OJT) returns $2.28. When public service employment is conducted as OJT in the public sector, it too has a substantial payoff, but that is not true of run-of-the-mill work experience projects. 2. The quality of CETA classroom training is primarily a product of the local institutions. Prime sponsors normally buy services from whatever training entities exist in the com munity. Fortunately, more prime sponsors have access to some high-quality training institutions, but the others must settle for what is available. However, though CETA may have had limited impact on the quality of training available in most communities, it has had a major impact on the will ingness of training institutions to enroll and serve economically disadvantaged trainees. 3. The dependence of CETA on the availability and quali ty of the local training system applies to other services. Most often, the CETA prime sponsor functions as a broker and coordinator of services rather than as a service deliverer. The chief elected official and prime sponsor staff can encourage quality services but, with few exceptions, cannot deliver them. The prime sponsors depend, therefore, on the perfor mance of: (1) the education and training institutions; (2) the public employment service, which provides labor market in formation, certification of eligibility, payment of allowances, operation of intake and assessment centers, marketing of OJT, and other linkages with the employer community; and (3) local community-based organizations, which are spotty as to their service delivery capacity but im portant for their client-advocacy role. 4. Judged by the quality of facilities, equipment, cur riculum, and instruction, the quality ranking of training in stitutions in descending order tends to be: (1) private pro prietary schools, (2) public vocational and technical schools, (3) skill centers developed under the auspices of the Man power Development and Training Act (MDTA) between 1963 and 1973, and (4) training activities of communitybased organizations. However, choice among institutions is complicated by the fact that ranking in terms of ability to understand, relate to, and serve the disadvantaged is exactly the opposite. Therefore, prime sponsors are constantly forc ed to compromise between these two sharply differing measures of quality. 5. While CETA training is a continuum of earlier pro grams, significant changes have evolved. The most notable changes developed under CETA are: a. Greater utilization of broad occupational offer ings, achieved primarily through referral of eligible individuals to mainstream training programs in private and public schools and colleges. b. The expansion of nonoccupational training ac tivities such as remedial basic education, English as a second language, prevocational orientation, and job-search training. c. Participation by communities not previously serv ed. d. The additional or expanded use of new service deliverers. 6. Persistent obstacles to improvement of the system are: a. The federal prime sponsor rating system credits procedure rather than substance and neither measures nor rewards quality. b. The data system measures short run rather than long-run outcomes. Longer-duration training has a higher rate of return than training of short dura tion. To date, only the short-run data have been available to decisionmakers. c. CETA has not succeeded in linking its offerings with complementary services offered by the pro grams or sequencing the training and services it of fers. d. The high payoff of OJT is also not apparent to the prime sponsor in the short run. Marketing OJT is expensive for prime sponsors in terms of staff time expended. In addition, the subsidy, equalling 50 percent of the wage for a few weeks, tends to be primarily attractive to marginal employers. e. The allowance system distorts incentives for under taking training. Many of the disadvantaged clientele must have financial support to afford training participation, but uniform allowances at the level of the federal minimum wage for all enrollees appear to encourage some to participate more for the sake of the allowance than for the training. f. The federal budgetary process denies prime spon sors opportunity for long-range planning. g. The federal staff is ill-equipped to provide useful technical assistance. h. Both prime sponsor and federal staff lack training backgrounds. Still, there is reason to take satisfaction in CETA training accomplishments, but there is ample opportunity to make a sound system better. The key steps are: 1. Training and employability development should be made the highest priority of the CETA system. 2. The system should encourage and support a two-tiered structure of classroom instruction, consisting of a remedial career-entry phase, designed to serve the disadvantaged with remedial education, career ex ploration, job-search training, and occupational skills that can be acquired in no more than 36 weeks; and a career-development phase available to those who prove themselves in the first phase. 3. To induce employers to provide on-the-job training, subsidies should cover a "try-out" period during which the trainee is in the workplace but receiving a CETA stipend until the employer has an opportunity to assess the trainee's worth. Public service employ ment should, in most cases, function similarly, as public sector OJT. 4. The training allowance system should be reformed to include: (1) reimbursement for the out-of-pocket ex penses of training participation, paid to all, (2) a sub sistence component based on family income, and (3) a cash incentive for high performance. 5. The federal funding commitment should be made to prime sponsors at least 2 years ahead. 6. An aggressive personnel exchange system should give federal, state, and local staff substantial experience at each of those levels. 7. A technical assistance, staff training, and curriculum development entity should be developed that is capable of bringing meaningful help to the local level. 8. A rating system should be devised stressing qualitative factors and rewarding positive long-range results. The system should reflect data derived from the Con tinuous Longitudinal Manpower Survey upon which the first point of this summary was based. 9. There should be constant experimentation in search for improvements in the quality and outcomes of training. The adaptation by local training institutions of computer-assisted and computer-managed instruction, already proven in Job Corps to be effective with a disadvantaged population, deserves serious considera tion. 1. The Scope of CETA Training Antecedents Congress passed the Comprehensive Employment and Training Act in 1973 with the aim of consolidating federal employment and training resources. Chief elected officials, known in CETA parlance as prime sponsors, representing local units of government with populations of more than 100,000 and states on behalf of smaller areas, were eligible to participate in the program. Prime sponsors were to plan, design and administer local programs consistent with general federal guidelines, subject to local advisory council review and regional Department of Labor approval. CETA is best understood and appraised as a continuation of employment and training programs that had existed since the early 1960s under the Manpower Development and Training Act (MDTA) of 1962 and the Economic Opportuni ty Act (EOA) of 1964. It made available under one authorization a range of services aimed at improving the employability and earnings experience of workers from lowincome households. On the administrative side, CETA transferred decisionmaking authority at the state and local level from the state employment service, state boards of vocational education, and local community action agencies to elected officials identified as prime sponsors. This study is illuminated by the recognition that CETA in herited from its predecessor programs the institutions or ser vices they established, though significant changes have been made in the ways in which those institutions are used and in 8 the mix of services provided. Labor market interventions under CETA's predecessors were limited to classroom and on-the-job training, remedial education, work experience, subsidized public employment, and supportive services to make participation in the other components possible. CETA has added to this list a range of nonoccupational training. As the program's title indicates, its offerings can be divided be tween those that stress job creation and those that emphasize employability development. The latter is the focus of this study. Prior to CETA, training occurred primarily in three in stitutional settings, with minor use of a fourth. Initially, most training was offered to class-size groups in existing public vocational education facilities. This arrangement suf fered from serious drawbacks. First, the training was often offered at inconvenient hours because the facilities were fully utilized during regular school schedules. Second, the schools offered the trainees a limited range of occupational choices. Third, the out-of-school population of generally limited education was taught by instructors accustomed to, and with pedagogy designed for, mainstream in-school youth. An alternative soon emerged that attempted to correct the first and second problems. The remedy was tuition payment for, and referral of, individuals to ongoing postsecondary vocational and technical school programs (known as in dividual referral). But that was useful only for those who could compete with the regular student body. The third alternative was the skill center. This new institu tion was developed specifically to meet the needs of MDTA enrollees drawn mostly from among the poor and deficiently educated, classified as the "disadvantaged" population. Modular training curricula were structured for adaptations to individual instruction, so that trainees could enter im mediately upon enrollment, rather than await the start of a new class, and progress to some extent at their own pace. Emphasizing individual needs, trainees could begin without prerequisites and leave for employment when their potential, their endurance, or their resources were exhausted. Remedial education, either preliminary to or integrated with occupa tional training, was available to those who needed it. The skill centers also provided supportive services, centering on the personal problems faced by the trainees. The services in cluded advocacy counseling to help with personal problems, transportation, minor health care, child care, job develop ment and placement. The skill centers utilized whatever low cost facilities were available, burdening the centers and their trainees with an unfortunate stigma. Observers expressed concern that the student body included a mixture of the up wardly yearning and those attracted mainly by the stipends. Since the state boards of vocational education were responsible for providing the training sites, limited use was made of private training institutions. Occasionally training was contracted with private proprietary schools. Additional possibilities that emerged out of the antipoverty movement were the community action agencies (CAAs) and the community-based organizations (CBOs). Only a few CAAs, the Opportunities Industrialization Centers (OIC), and the Operation Service-Employment-Redevelopment (SER), and organization oriented toward serving Mexican-Americans, played significant training roles during that period. A technical assistance operation known as AMIDS (Area Manpower Institutes for the Development of Staff) provided inservice staff training, curriculum development, and other help directly to the federally supported training sites, but these disappeared with the advent of CETA. Given the limited resources and the clamor for enrollment by the eligible population, program administrators attemp ted to ease the conflict by cutting the duration of training 10 and making it possible to enroll more applicants in training programs. The problem was compounded by the fact that enrollees were paid stipends, consisting of the average state unemployment compensation payment plus allowances for dependents and training-related out-of-pocket costs for adults, and lesser amounts for unmarried youths. The stipends consumed half of the resources. The law also re quired restricting training to occupations having "reasonable expectations of employment." This combination of a disadvantaged clientele, pressures for immediate placement, and the short duration of courses limited the training choice to entry-level preparation for oc cupations requiring rudimentary skill and characterized by high turnover rates. Three-quarters of all pre-CETA classroom enrollments were in seven occupational categories and all at the entry level: clerical, health care, automotive repair, machine operation, welding, building service, and food service. On-the-job training began slowly and accelerated so that by 1968 about half of MDTA's enrollments were in OJT slots. Representatives of the public employment services or community-based organizations offered employers an average of $25 a week for 26 weeks (the equivalent of 15.6 hours pay at the minimum wage in the early 1970s) as a sub sidy for hiring MDTA eligibles. Predictably, only the small marginal employers were attracted by the subsidies and then only if the most qualified among the eligible were selected. Since these employers normally lacked training capability, the subsidy primarily bought a placement in a low-paying job. Nevertheless, the costs were low and the placement was built in, with retention likely, so a high ratio of benefits to costs was assured. CETA was introduced, not because MDTA and EGA were failures, but because it was thought that local political ac- 11 countability would produce a service mix, including training, better adapted to the local scene. In addition, it was assumed that quality would improve if local institutions had to com pete for the right to provide GETA services. Focusing on what has happened to the quality and effectiveness of train ing under CETA should help answer whether those assump tions were correct. Magnitude of CETA Training Under CETA, the nation's commitments to employment and training programs rose dramatically, but employment generation captured the lion's share of dollars and people. There were over 2 million new enrollees in CETA programs in fiscal 1980. Over two-thirds of these new participants were in work components, or nearly half if the youth summer pro gram enrollments are excluded. Job creation components also accounted for two-thirds of the 1 million service years of employment and training activity estimated for fiscal 1980. The number and distribution of service years by program components follow: CETA component Total Local programs Classroom training On-the-job training Summer youth work experience Youth transition services Nonsummer work experience Public service employment Job Corps National programs Service years (000) 1,097.5 Percent distribution 100.0 1,041.8 95.2 219.9 54.3 126.2 41.3 252.3 347.8 35.7 20.0 4.9 11.9 3.7 23.0 31.7 3.2 20.0 1.8 12 Whereas training expenditures predominated in the policy mix prior to CETA (accounting, for example, for nearly 2 of every 3 dollars expended in 1969), they fell to less than a fourth of CETA outlays by 1974 and to only 15 percent in 1978, before rebounding to 21 percent in 1980. The source of the relative shift was massive allocations to job creation rather than decreases in training funds. In fact, with plen tiful employment funding under other titles after 1975 and with public sector employment proscribed under Title IIBC after 1978 (the title giving prime sponsors the most discretion to choose among alternative services), more of that title's funds could be expended on classroom training and OJT, as follows: Fiscal year 1975 1976 1977 1978 1979 1980 Outlays (in millions) Classroom OJT $ 77.9 $ 309.6 168.4 606.2 739.8 207.5 872.6 257.8 941.5 224.0 1,224.6 216.1 The relative distribution of training dollars has also shifted during the first 6 years of CETA. OJT has accounted for between 8.7 percent and 13.8 percent of Title I or IIBC expenditures, while the share of classroom training during the same period rose from 34.4 percent to 57.1 percent of IIBC allocations. Fiscal year 1975 1976 1977 1978 1979 1980 Percent of Title IIBC (former Title I) outlays for Classroom OJT 34.4 8.7 35.7 9.9 42.1 11.8 46.5 13.8 12.4 52.3 57.1 10.4 13 The rising expenditures for training have provided oppor tunities for an increasing number of persons in need. The new participants, or annual flow through training programs each year, rose rapidly over the early 1960s to 336,000 an nually in fiscal 1967. The participant levels reached 481,000 annually by 1972, and declined during the next 4 years. Growth resumed in 1976, reaching a peak of 773,000 in 1978, followed by a decline over the succeeding 2 years. Most of the secular increase resulted from the growth of local classroom training enrollments. In fact, average OJT par ticipants during the first 6 years of CETA were only twothirds the average from 1969 to 1974. With the erosions of inflation and the beginnings of budget cuts, the pattern of Title IIBC (or its predecessor) training enrollments has been as follows: (in thousands) Service years Individuals Fiscal year Classroom OJT 1975 1976 1977 1978 1979 1980 291 514 537 580 569 494 74 148 170 193 157 132 Classroom 69 146 173 188 163 152 OJT 17 39 44 54 40 36 This study focuses on the regular and continuing funding that Congress appropriated for training under CETA, ex clusive of Job Corps. In addition to direct allocations to prime sponsors, funds were also allocated under Title IIBC to state governors for vocational education. Other training funds came from youth training programs, a special ap propriation for a demonstration skill training improvement program (STIP), a private sector initiative program (PSIP), and other titles of CETA. The Title IIBC formula allocation has accounted for about half of the CETA training, as the 14 percentage breakdown of fiscal 1980 enrollments and expen ditures indicates: Source of Funds Total training IIBC Supplemental vocational education HD III STIP IV Youth employment Job Corps VI VII (PSIP) (percent distribution) Expenditures by Participants by program program components components 100.0 100.0 49.9 61.0 4.7 1.4 1.5 2.6 * 2.1 3.3 1.3 6.5 28.0 0.7 4.7 11.8 9.3 0.6 10.6 'Participants counted under Title IIBC in CETA MIS reports. 2. Local Vantage The Comprehensive Employment and Training Act pro vided funds in 1980 to 476 state and local govern ments—known as prime sponsors—for the delivery of ser vices designed to improve the skills and the employment op portunities of low-income individuals experiencing dif ficulties in the labor market. While not necessarily represen tative, the 12 intensive case studies are a microcosm of the CETA system. This can be demonstrated by first comparing them on a few key variables with national averages. The quality and effectiveness of the training provided by each prime sponsor is assessed, and the aspects of regulation and administration that have an impact on quality are reviewed. The national and local reviews then become the source of conclusions and lessons related to the entire system. Classroom Training Allocation Determinants Nationally, prime sponsors enrolled half of their Title IIBC participants in classroom training. Most of the prime sponsors interviewed indicated a preference for classroom training, and all but Baltimore, San Francisco, and North Carolina balance of state enrolled a higher proportion than the national average in classroom training. But because of 15 16 variations in costs per enrollee, expenditure patterns did not coincide with those for enrollments, as indicated below: Classroom training under Title IIBC, fiscal year 1980 Percent of Percent of total expenditures total participants Location 48.4 50.3 U.S. average 10.1 10.4 Baltimore 53.0 75.1 Dallas 65.7 54.5 Indianapolis 48.4 62.4 Montgomery North Carolina balance 33.6 39.6 of state 18.9 62.5 Ottawa 2L6 53.2 Penobscot 53.1 41.0 San Francisco 68.0 55.7 Seattle 81.3 63.6 Tucson 65.8 57.0 Utah 60.3 71.9 Worcester Nationally, neither the local unemployment rate nor the age structure of the participant population explain the dif ferences in service mix. Also, in making cross-prime-sponsor comparisons, it is not certain that common definitions were used. It might also be expected that classroom training would decrease with rising unemployment; however, the op posite has tended to be true. As unemployment has risen, more Title IID and VI PSE funds became available, thereby freeing up Title IIBC money for added training. On a crosssectional basis, there is a tendency for those prime sponsors with the lowest unemployment rates to show the greatest commitment to classroom training, a relationship which does not hold true for on-the-job training. The major depar tures from this generalization are those less-populated areas with limited access to training institutions or private employers. 17 Location U.S. total Baltimore Dallas Indianapolis Montgomery North Carolina balance of state Ottawa Penobscot San Francisco Seattle Tucson Utah Worcester 1980 civilian labor force (000) 102,908 734 510 397 309 1,684 70 87 399 776 141 554 158 1980 unemployment rate (percent) 5.8 6.3 3.8 5.5 3.5 7.1 4.7 6.4 6.6 5.6 5.1 3.8 4.7 Prime sponsors have also complained about the require ment of maintaining the same level of service to youth under Title IIBC after the addition of new Title IV youth programs in 1977. This provision forced them to emphasize work ex perience under Title IIBC, since this was the most common youth activity in this title. Yet, the studies found no consis tent correlation, positive or negative, between the share of IIBC funds devoted to youth activities and the ratio of train ing to work experience under that title. Institutional Setting The providers of classroom training varied widely among the 12 sites studied. In Seattle, Tucson, and Utah, the major providers were the skill centers, which originated under MDTA. In San Francisco, the use of community-based organizations was favored, along with minor use of a skill center and a substantial amount of individual referral to private schools and the community college system. In North Carolina, the community colleges and technical institutes had most of the action, as was the case in Penobscot, where 18 private colleges rounded out the classroom training facilities. In Dallas, two school districts and two CBOs shared most of the training activity. The overriding concern was with pro viding low per-capita cost training. Baltimore and Worcester used a mix of public and private schools and some employers to provide classroom training. Rural Ottawa County, lack ing training facilities of its own, provided living expenses and tuition aid to send many of its trainees out of the area. Some of the more rural areas of the Penobscot consortium follow ed a similar practice. Montgomery County enrolled its trainees in public and private schools, as well as local col leges. Indianapolis relied wholly on individual referrals to public and private colleges. To generalize from the case studies, public postsecondary schools seem to provide most of the training, with skill centers in second place, followed by CBO-run schools and then private proprietary schools. In contrast to the national picture where individual referral purportedly predominates, enrollments in class-size groups were more common than in dividual referrals in all of the study locations, but this was true primarily because of the unusual CBO delivery system in San Francisco, Seattle, and Dallas. Though skill centers played the predominate role in three locations, they were important in two others and supplemen tal in one. In Seattle, Tucson, and Utah, the skill centers pro vided the bulk of occupational training for the prime spon sor, but, in Baltimore and San Francisco, the skill center had a limited CETA role. These MDTA institutions still seem to be characterized by the same advantages and shortcomings they offered during the earlier period. Their staffs are typically dedicated to serving a disadvantaged population, and a range of on-site services required by disadvantaged trainees is generally available. On the other hand, the train ing tends to be limited to an average of 26 weeks and a max imum of 36 weeks (except for health care occupations where 19 licensing requirements require longer training periods); the facilities are often poor, though the equipment is generally adequate, and the institutions are likely to have low prestige in the community. Most skill centers make use of existing structure. For in stance, the Tucson skill center was initially housed in aban doned department store, garage, and restaurant buildings. In Utah, one skill center is in a former laundry building and another, in what was formerly a high school. The Baltimore skill center is also in an abandoned high school, as is the Dallas facility, although the latter has been lavishly remodel ed. The San Francisco skill center is housed in an abandoned elementary school. However, the latter institution has lost most of the attributes that typify a skill center. The Seattle facility, which was the major provider of CETA occupa tional training in that city, was exceptional in being the only center built for that purpose. Owned and operated by the Seattle Opportunities Industrialization Center, it has the allegiance of its enrollees and the black community, but although it is well-designed and -equipped, it still does not appear to command the general prestige that its quality deserves. Beyond these examples of MDTA continuance, the oc cupational training locus seems to have shifted to standard public vocational schools and proprietary schools, sup plemented by specialized programs run by CBOs. In the mainstream public and private institutions, which serve the general population, CETA enrollees are referred individually to the regular program, with any necessary fees paid from CETA funds. This approach provides the CETA enrollee with training as good as that available to the taxpaying or feepaying public and allows participants to report to peers or prospective employers, "I'm at the technical college," rather than, "I'm in CETA." 20 The drawback is that only those who are fairly selfmotivated and who can compete with the regular clientele in the school will last. Some schools do provide remedial educa tion, but that is not the norm. A few prime sponsors follow a sequencing process in which an enrollee can spend time in a remedial component at one institution and then be referred to a skill training institution. However, almost every CETA prime sponsor has far more applicants than available slots, and the case studies found waiting lists for training ranging from 3 to 6 months. Thus, most sponsors considered it in equitable to invest heavily in some eligible persons, thereby leaving no funds for others. The regional office staff, and national policy as well, reinforced this concern by inveighing against high per-capita cost and setting up an assessment system that commends program for combining low costs and high immediate placements but largely ignores the quality of training offered. Community-based organizations also play a larger role under CETA than they did under MDTA. The Opportunities Industrialization Centers (QIC) have been in the skill train ing business since the mid-1960s and in many cities provide high-quality training in a wide range of occupations. The Seattle OIC, as noted, is an outstanding example. However, in many other cities the OIC offering is limited and, in some places, of low quality. OIC provided a narrow range of training in Dallas and North Carolina, and it had been drop ped for poor performance in San Francisco and Mont gomery County. Often, as in Dallas, the prime sponsor must share responsibility for not simultaneously supporting and spurring the CBO to higher-quality efforts. While OIC offers occupational skills, other CBOs tend to restrict their offerings to remedial prevocational components and to sponsorship of on-the-job training. However, the trend is for local community-based organizations to add entry-level skill training components. Facilities tend to be 21 unsatisfactory, but the training emphasizes those occupa tions that require little training equipment. CBOs are par ticularly effective where the primary concern is with the in struction of English as a second language. In some cases, the support of CBOs may be motivated by a desire to "give them a small contract to keep them alive." However, the San Francisco and Seattle CBOs are both politically potent and offer highly competent instruction. Training Occupations High turnover occupations predominated among the training activities of the 12 case studies. Most of the enrollments seem to be in clerical, health care, automotive and auto repair, welding, machine operation, building ser vice and food service, just as they have been since the days of MDTA, but some significant change has occurred. Clerical and health care occupations are, in generally high demand so that, depending upon the level of investment in time and training costs, jobs can be found at lower or higher levels within the occupational cluster. Given the backgrounds of the enrollees and the relatively short training time allowed in most CETA programs, the clerical trainees tend to attain on ly marginal levels of skills. Yet the demand is sufficient in most locations to assure a respectable placement rate in the 60- to 80-percent range. Admission to health programs tends to be more selective. Training for licensed practical or voca tional nurses, an MDTA creation, continues to be the outstanding CETA health care program. Training extends from 36 to 52 weeks, and placement is almost guaranteed if state certification standards are met. Yet, within the health care category, the Nurse's Aide Academy program in Dallas is evidence that programs of short duration can be produc tive when attuned to the needs of the client population and the demands of the labor market. 22 There seems to be a small and sporadic, but still signifi cant, trend toward a broadened occupational range in classsize projects. Training for computer-related occupations and office machine repair are on the increase. Other occupations are added from time to time in response to local economic conditions, but the conditions and the programs rarely con tinue over long periods of time. In fact, CETA-funded train ing activities appear to be more flexible in phasing in or out according to community need than is common among train ing institutions. The trend toward increased use of individual referral has expanded the range of occupations for which training is of fered and is nearly as wide as the total offering of the local training institutions, but the numbers trained add up to few in each occupation. Examples from the 12 case studies in clude radiologic, biomedical, psychiatric and graphic reproduction technicians, dental hygiene, drafting, air con ditioning, computer programming, graphic reproduction technician, and cosmetology. However, many of these are 2-year programs. CETA tries to limit enrollment to 36 weeks, with an average, scheduled course duration of about 26 weeks, and it rarely pays for more than 52 weeks of train ing. In a few cases, in Penobscot especially, CETA was found to be paying tuition allowances for the first half of a 2-year training program, leaving those who can afford it to finish at their own expense. Apparently there is interest in long-term training but an unwillingness to pay the price of reducing the numbers served since for every person enrolled in an expensive long-term training course, others are left without service. To demonstrate the importance of this kind of activity, the skill training improvement program (STIP), initiated in 1977, earmarked funds for both high level skills and longterm training. The funds were allocated to prime sponsors on a competitive basis. To support the same training for the 23 same number of enrollees from regular prime sponsor funds would have absorbed a high proportion of available funds and resulted in the rejection of many applicants. However, the last grants for STIP projects were awarded in the first quarter of fiscal year 1979 from fiscal year 1978 funds, and no additional funds have been allocated for the program. Other problems are associated with long-term training. Not only are these training programs more costly, but they tend to have higher dropout rates and no higher immediate placement rates than short-term courses, even though the analysis of national data indicates that their payoff in the long run is greater. Many eligible people lack the persistence and financial ability required to last through a long-term training program. They are apparently willing to train for the relatively low-paid jobs plagued by high turnover rates that seem to be available even during economic slumps. As a matter of general policy, the Baltimore prime sponsor had opted for expensive, longer-term offerings in its regular training program. That was achieved at the price of con siderable selectivity among applicants. The Seattle skill center had historically shown a preference for long-term training, thereby contributing to criticisms of high percapital cost. Worcester's largest skill training activity was in low-level clerical skills, but it had also carefully selected enrollees for electronics and computer programs. Mont gomery County tried high-level skills on an individual basis without stipends but had to screen 1,100 applicants to find 33 persons qualified for a biomedical technician program. Several others had similar experiences with STIP. While the completion rates in these long-term offerings were generally lower than in short-term courses, those who completed were placed in jobs with considerably higher pay. Success at running long-term training for high-level oc cupations presupposes either concentration on a better- 24 prepared segment of the CETA-eligible population or a greater investment in remedial activities. The Baltimore prime sponsor has clearly made the selectivity choice for oc cupational training, as has Worcester for some of its pro grams. Utah and Penobscot had CETA-eligible populations, which in general had these characteristics, but they had not chosen to invest in long-term training. Tucson has chosen to serve a population ready and willing to qualify for minimum-wage jobs, while San Francisco stressed relatively low-cost English as a second language for predominately female immigrants. Dallas chose to focus on a needy black population but then failed to adjust the curriculum to meet those new needs. For others, the choice between lower-level enrollees and higher-level skills remains a troublesome one. On-the-Job Training Seven of the prime sponsors studied enrolled a larger pro portion of their IIBC participants in on-the-job training than the national average, including Penobscot and Ottawa, which enrolled more than double the national average. The variation in the percent of funds and participants enrolled in OJT was even greater than in classroom training. Most of the prime sponsors studied praised OJT but then tended to make relatively little use of it. Enrollment in OJT ranged from 0.2 to 24.6 percent of the total IIBC enrollment in the 12 sites, despite the favorable results already indicated and the cost advantages documented below. The key deter minant in the use of OJT appeared to be the relative ag gressiveness of the agencies accepting the OJT marketing assignment, which is usually delegated to the public employ ment service, CBOs (especially Urban League), and local Na tional Alliance of Business (NAB) chapters. 25 On-the-job training under Title IIBC, fiscal 1980 Percent of Percent of Location total participants total expenditures U.S. average 11.4 10.4 Baltimore 3.8 4.1 Dallas 16.4 9.3 Indianapolis 0.2 0.5 Montgomery 1.2 0.6 North Carolina balance of state 13.7 12.1 Ottawa 24.6 13.4 Penobscot 23.7 15.9 San Francisco 12.4 22.5 Seattle 19.0 10.7 Tucson 2.1 1.7 Utah 17.1 17.3 Worcester 5.3 7.5 Penobscot used the state federation of labor as one of its OJT developers and had established a well-designed and carefully monitored system, linked with other services. Worcester had a high-quality, coupled classroom/OJT pro gram sponsored by local banks and administered by the local NAB chapter. In Seattle, Utah, and North Carolina, the employment service marketed OJT far more energetically than did the prime sponsor. Ottawa County had no classroom training institutions, and OJT served as an alter native to sending trainees outside the area at a high cost for transportation and lodging; Dallas had the advantage of a tight labor market to help "sell" the advantages of on-thejob training to employers. Indianapolis used OJT only in isolated cases because employers had earlier criticized the 4 'red tape" that it involved. Private industry councils ex pressed interest in OJT promotion but were only beginning to get involved in the effort when the case studies were prepared. 26 For the prime sponsor, a major obstacle to greater use of OJT is the high staff costs involved in persuading individual employers to contract for OJT slots. Many employers are reluctant to take on a CETA enrollee. They tend to question the qualifications of the eligible population and are fearful about the amount of red tape that may be involved in the contracting process. The result is tnat, in most cases, only small marginal employers, to whom a half wage subsidy for 3 to 6 months may be attractive, sign up for one or two trainees. Moreover, in some rural areas, an absence of even small employers makes the development of OJT contracts nearly impossible. One much-advocated approach to serving the disadvantaged is a remedial education/classroom instruction/on-thejob training sequence, which Penobscot had achieved for some enrollees. However, given the limited funds available, the outside pressures to keep per-capita costs low, and the logistical difficulties involved in developing such complex programs, most prime sponsors considered it more ap propriate, if not more politically expedient, to run three parallel programs. The expansion of job-search training seems to be adding a fourth parallel track for the job-ready, who may need only placement services, but for whom the public employment service does not have an appropriate job listing. The tendency is for prime sponsors to decide on some division of available funds for each of these tracks, contract for the needed services, and then monitor the activities under each contract. San Francisco is an example of a prime sponsor that ad vocates high-support OJT and has had some positive ex perience with it. (The emphasis is reflected in the expenditure of a high proportion of funds to support a rather low pro portion of its participants.) The program is an example of the remediation/classroom training/OJT sequencing pro cess. It begins with a commitment from the employer to take 27 on one or more of those persons who have successfully com pleted the first two training phases—usually with some limited right of selection. Often the employer provides a training facility at the workplace, along with some equip ment and instructional assistance from the company person nel office for a work-simulation training phase. Those who attain a prescribed performance level, or a predetermined proportion of them, are then offered permanent employ ment by the firms involved. Well-paid jobs on a structured promotion ladder, with ac companying job security and fringe benefits, have been at tained through this route. Examples include employment with a grocery chain, a public utility, and an engineering association. Nevertheless, the level of enrollments fluctuated with the needs of the companies—frequently involving affir mative action requirements—and could not be sustained. Not only must a willing employer be found, but uncommit ted dollars must be available at the appropriate time—and these programs tend to be expensive. Both community and federal pressures are aimed at committing every expected program dollar at the beginning of the fiscal year and thereafter keeping costs low. Yet, a cooperating employer may find it difficult to adjust his labor needs to the schedule of federal fund allocations. Thus, launching a successful OJT effort usually involves the fortuitous availability of funds that are either uncommitted or deobligated from other contractors or prime sponsors. The San Francisco prime sponsor also offers an illustra tion of an approach that is related to the high-support OJT program, but is less expensive and therefore has a chance for greater continuity of funding. Advocacy organizations for minorities, women, and the handicapped are funded for the sole purpose of making placements, without offering any ac companying remedial education. As a prime sponsor staffer put it, representatives of these advocacy groups "follow the 28 equal employment opportunity enforcers around offering companies relief by providing selected, though CETAeligible, people with the right affirmative action characteristics.*' North Carolina used an alternative approach, called "work site assessment," which was funded by the state rather than CETA. In this program, workers are paid a sti pend by the employment service while they are assigned to a host employer's establishment. After a trial period, the employer may reject the worker, offer unsubsidized employ ment, or accept an OJT contract. Beyond these few exceptional programs, OJT was a useful, if routine, component of each prime sponsor's reper toire of program strategies. Its high benefit-cost payoff, clearly noted at the national level, was unknown to local operators and not readily apparent from any of the data available to them. The difficulty of marketing OJT, the high administrative cost of the personalized contracting process, and the overrepresentation of marginal employers among the contractors acted as disincentives to heavy sponsor involve ment in that activity. The fact that planned OJT expen ditures were substantially greater than actual expenditures for most of the prime sponsors indicates both the desire for OJT and the difficulties of its promotion. Nonoccupational Training The expansion of nonoccupational training is one of the most intriguing developments among CETA training ac tivities. Included among these efforts are instruction in basic education, English as a second language, prevocational orientation, coping skills, motivation, and job search train ing. These services are sometimes supplied separately and sometimes, in any combination, included as adjuncts to oc cupational skill training. 29 The steady increase in the educational attainment of the work force and the decline of jobs that do not require at least some reading, writing, and simple arithmetic skills make lack of basic education an increasingly serious handicap. General education development (GED) certification as a substitute for a high school diploma is, therefore, a component of growing importance. Many CETA eligibles cannot succeed in occupational training without first or simultaneously par ticipating in remedial education and often in English as a second language. Some programs have demonstrated their ability to raise reading and arithmetic performance three grade levels during a 12-week course, though the usual pro gress is less dramatic. The persistence of Spanish as the mother tongue among rural peoples of the Southwest and Puerto Ricans on the East Coast, together with the flood of more recent im migrants from Latin America and the Far East, has made English as a second language (ESL) the largest training pro gram of some prime sponsors. Increasingly, it has become vocational English as a second language (VESL) with the ad vent of curriculum materials teaching language in a job con text. ESL (or VESL) tends to be highly successful in terms of placement rates, retention rates, and the gain between preand post-training wage rates. San Francisco had the widest range of ESL offerings among the 12 prime sponsors studied, enrolling persons whose native tongue was Spanish, Chinese, Korean, Filipino, Vietnamese, Cambodian, In donesian, and Russian. Seattle also had a growing ESL com ponent, serving various Asian refugees. Dallas and Tucson offer ESL for Hispanics, with Operation SER as contractor. Other primes had more limited ESL activities. For all of these groups, their very presence so far from their points of origin is evidence of their motivation, and many have substantial skills that can be applied only after language competence is obtained. 30 North Carolina had the heaviest investment in prevocational orientation, with a state appropriation of $2.3 million in addition to CETA funding. In 47 of its 58 community and technical colleges, CETA and non-CETA disadvantaged enrollees are given 8 weeks of a combination of remedial education, career exploration, and various life-coping skills. The numbers enrolled exceed the numbers enrolled in CETA-funded occupational skill training. This human resource development (HRD) program emerged as a natural consequence of reliance on public postsecondary institu tions, which were not prepared to deal with a disadvantaged student body. Though it functions within the same institu tions, it is separate from and not linked with occupational skill training. The Baltimore prime sponsor was unique in establishing its own in-house faculty for conducting most nonoccupational training as an alternative to contracting it out. The hottest new entry among the nonoccupational pro grams is job search training. A person who knows how and where to search and how to impress an employer can usually find at least a high-turnover, low-paying job. Typically, the job search program enrollees are taught during a 2-week course to prepare resumes, fill out applications, and practice answers to the most common questions asked by interview ing employers. The enrollees then learn to use the telephone yellow pages and classified advertisement sections in the newspapers in order to identify employers likely to have the job they want and for which they are qualified. They practice telephoning skills and then spend long, supervised hours call ing for interview appointments. Reported placement rates tended to range from 70 to 90 percent among the cases studied, though parallel programs for work incentive pro gram (WIN) enrollees in the same cities were observed to have as low as 30 percent success rates. Moreover, there is as yet no substantial data on retention rates and, more impor- 31 tantly, no data on whether or not the participant has obtain ed a permanent job-seeking skill, which works in subsequent spells of unemployment. The nonoccupational offerings rest on two assumptions: that a high proportion of jobs require no specific pre-entry skills and that turnover produces a flow of jobs, even during less-than-prosperous times. However, few programs indicate how to distinguish between primary and secondary labor markets and how to choose and attain jobs with promise. Unlike occupational skill training, which is usually provided by public education institutions and proprietary schools, private contractors appear to dominate nonoccupational training, except for ESL where community-based organiza tions predominate. Training Costs Training costs varied widely across the study sites, but gross cost data are of limited value because the scope and content of the training differed widely (Table 1 which is drawn from national DOL data sources includes only in structional and administrative costs but not allowances). On ly Indianapolis and Seattle of the 12 prime sponsors exceed ed the national average cost per positive termination of $3,170, with only Indianapolis recording a higher-than-theaverage national cost per placement of $6,508. Important factors in determining the costs per participant and per outcome include the ratio of nonskill to skill train ing, the average length of training, the occupational mix, the institutional mix, the client characteristics, the dropout rate, and the extent of local subsidies. Costs per placement or costs per positive termination depend, of course, upon the characteristics of the participants, the state of the local economy, and the availability of positive alternatives to placement. According to national data, the positive termina- Table 1 Program costs, fiscal 1980 (IIBC, excluding allowances) Cost per participant United States Baltimore Dallas Indianapolis Montgomery North Carolina Ottawa Penobscot San Francisco Seattle Tucson Utah Worcester Classroom $1,328 682 928 2,299 1,435 1,008 376 513 1,898 2,099 1,566 1,143 952 OJT $1,130 766 747 4,082 941 1,056 679 844 2,662 897 933 1,001 1,620 Per positive termination $3,170 1,098 2,332 5,973 2,847 2,137 2,159 2,035 2,111 3,582 1,828 1,735 1,924 Rate 67.0 87.7 69.2 72.8 78.9 78.4 78.6 76.6 88.9 72.4 77.2 75.9 68.7 Cost per outcome Per private placement Rate $INA 1,298 3,539 INA 3,884 5,799 5,112 5,270 3,679 6,660 2,397 3,903 3,505 INA 75.9 46.3 INA 59.7 32.3 42.6 39.3 59.5 42.3 63.5 40.3 41.5 Per placement Rate $ 6,508 1,257 2,780 33,306 3,406 4,921 4,408 4,286 3,222 5,704 2,224 3,432 3,089 40.2 78.4 59.0 24.9 68.0 38.0 49.4 48.3 67.9 49.4 68.4 45.8 47.1 33 tion rate of all prime sponsors averaged 67 percent. All 12 prime sponsors studied had a better record than the national average by the positive termination criterion, but In dianapolis and North Carolina fell below the national average in placement rates. The prime sponsors studied con sistently reported higher outcomes to the researchers than were found in the national data source. The national average cost per placement was more than double the outlay per positive termination, but only three of the 12 prime sponsors exceeded that ratio. Nationally, costs per participant in classroom training ex ceeded OJT costs by 18 percent, but half of the 12 prime sponsors expended more funds per OJT than per classroom enrollee. The presumed OJT cost advantage was offset in many places by the locally subsidized and sometimes tuitionfree use of public training institutions. Public community colleges, technical institutes, and area vocational schools are generally eligible for the same average daily attendance fund ing whether the student is matriculated or not, and, if tuition is not completely free to enrollees, then the cost to the prime sponsor is minimal except for the stipend. Another offset against the apparently low comparative costs for on-the-job training are the promotional costs; because there may be one trainee per employer, the staffing requirements for either prime sponsor or contractor associated with OJT may exceed that of classroom training. The costs obtainable from national data sources often dif fer from data obtained directly from the local level. The following locally obtained data are examples of the wide variations in costs. At the Baltimore skill center, costs per trainee ranged from $1,169 for clerical training to $3,344 for construction trades. Individual referral costs were uniform at $2,858 because the schools charged the same tuition rates for different occupations. When allowances are added, the total cost range grew from $2,041 for clerical to $4,696 for 34 computer and office machine repair, which involved training of longer duration than the construction trades course. Most were open entry-open exit programs, and therefore costs are determined by the amount of time the average trainee takes to complete the programs (or to drop out) rather than any ar bitrary, maximum course length. Since individual referral courses were fixed in length and tended to be longer and in higher skills than skill center courses, the average total cost rose to $5,173 including allowances, despite subsidized tu ition. In Tucson, for example, a skill center and a private pro prietary school provided occupational training alternatives. The annual average cost per slot at the skill center was $3,324 in 1980. However, costs per enrollee averaged $351 and costs per completer were $640, ranging from $95 per trainee receiving only adult basic education or job search training to $4,144 for a full year of skill training as a licensed practical nurse. Also, included in the mix were 2-week pre-OJT basic skill courses for electronic assemblers and an 8-week course for bank tellers. The private trade school charged a tuition of $5,200 for an academic year but, since CETA enrollees were automatically eligible for basic education opportunity grants, the cost to CETA was $2,000 per enrollee for the academic year. That private school advantage was lost in early 1981 when enrollees at the skill center became eligible for BEOG grants in consequence of the center becoming a unit of the community college. Because the private school was highly selective in its entrace requirements, the prime sponsor decided to put all its fiscal year 1981 classroom training funds into the skill center, whereupon the private school unsuccessfully sued. Where the classroom training program was accomplished primarily in public postsecondary vocational and technical training institutions, either through individual referral or in class-size groups (as in North Carolina), CETA pays only the heavily subsidized 35 tuition any individual student would pay, thus shifting substantial costs to the state. Indianapolis is an example of a program that has operated classroom training totally on an individual referral basis. One private college charged $1,690 per slot for tuition and books for two academic quarters, while a cosmetology school charged $1,890 for 1,500 hours of training over a 9-month period. At one technical institution, the cost by oc cupation ranged from $1,500 to $1,800 per academic year, while another charged $4,283 for a 1,000 hour program, and a truck driving school charged $2,675 for 10 weeks. These costs did not include the allowances, which were paid direct ly to the enrollees. Montgomery County relies primarily on class-size courses at private and public institutions. The Penobscot prime sponsor obtains basic education at no cost and occupational skill training at subsidized tuition rates at public institutions but pays full tuition costs, less BEOG's, at private institu tions. By carefully selecting its enrollees, Montgomery County negotiated $3,000 tuition costs for 26-week hightechnology programs at a private university and a private technical school, while a CBO had been charging $4,000 for training in much lower-skill occupations. Occupational training at a community college was obtained for $1,000 per enrollee, while non-skill training consisting of English as a second language, basic education, and assertiveness training for women ranged from $800 to $1,800 per person. In general, costs appeared in line with costs of occupa tional training outside CETA. Allowances were the major additions over the costs considered standard to training in stitutions. They were a necessity for many of the low-income clientele, but some of the non-family heads probably could have managed tuitionless training without stipends, had such allowances not come to be expected. However, Montgomery 36 County had tried nonstipended training and had ended by serving non-family heads from households at the upper socioeconomic levels of CETA eligibility. Dallas had manag ed a compromise, paying stipends at an hourly rate that was an average of 80 cents below the federal minimum. A per suasive case can be made for changing the allowance struc ture to prevent those stipends becoming a major incentive for enrollment. A major cost for CETA has been the deci sion to set the training allowance at the minimum wage for the hours attended, plus add-ons for training expenses and dependents. Those training for low-skill occupations may ac tually have a higher take-home pay during enrollment than after placement, making many reluctant to leave the shelter of the program. Since total allowance costs are determined by duration of training, whereas training costs fluctuate to some extent by occupation and by institution, allowances vary from about 40 percent to about 60 percent of total training costs. Because length of training is the major deter minant of both training cost and occupational level, the highest-cost training tends to prepare enrollees for the best jobs. 3. Quality of Training There are no simple criteria for measuring the quality of training; in fact, there is much mystique in the concept. Ef fectiveness might be conceptually easier to measure, but the data do not exist. Placement does not measure effectiveness unless compared against controls because the state of the local labor market may be the more critical determinant. Many factors enter into an assessment of training quality: the adequacy of facilities, equipment, and curriculum; the competence of the instructors; the appropriateness of the training occupations; the adequacy of training duration; and the quality of the needed supportive services provided. The characteristics of the clients and the supportive services they require also affect the quality of training provided. The management of training, including the linkages among ser vices and with employers, may affect outcomes no less than the training itself. Of course, costs are both a determinant of quality and a yardstick against which to measure effec tiveness. Institutional Capacity Because of the limited resources available to them, the prime sponsors included in this study were dependent upon the quality of the training institutions in existence. They were in no position to create new ones. Prime sponsor staff could only demonstrate acumen in choosing between alternatives of higher and lower quality. The quality of the available in stitutions in rural Maine, Michigan, North Carolina, and Utah was generally good, and the prime sponsors utilized the 37 38 available resources. In some isolated areas of these states, however, there were no training institutions available. In those situations, the choices were to rely on work experience or on-the-job training (though potential host employers were also scarce) or to pay costs of transportation and lodging at a distant site. In each of these cases, at least some moneys were allocated to institutions of lesser quality for political reasons, to maintain a range of alternatives, or because of the particular attachments of institutions to race, sex, or other groupings among the eligible population. The types of training institutions have been described but can also be ranked according to quality. The best institutions were those that were created to appeal to the broader, nondisadvantaged population in the community. Private pro prietary trade schools depended upon a combination of highquality training and good placement rates to maintain a clientele and make a profit. Attractive facilities, up-to-date equipment, high-quality staff, sound and motivating cur ricula, and high placement rates were all necessary to con tinue to attract tuition-paying customers. However, this quality came at high cost to CETA, both in budget dollars and selectivity of enrollees. Few CETA enrollees could meet the entry requirements of proprietary schools on an individual basis. Pressures from the Federal Trade Commission to advertise their placement rate make such schools reluctant to accept the hard-to-place. The Montgomery County experience of screening 1,100 CETA eligible applicants to find 33 persons acceptable to a technical school for a biomedical technician program has been noted. On the other hand, the San Francisco experi ment with placing a class-size project in a private business school, where the instructional and administrative staff was never able to establish rapport, illustrates the difficulties for such a school in seeking to adapt itself to an unfamiliar clientele. 39 There are, however, private institutions capable of serving a broader portion of the CETA population, as illustrated by the ABC Trade School in Tucson and the Beal and Hudson Colleges in Maine. For the CETA-eligible persons who can qualify, these appear to be good investments because they train for the mainstream labor market, their training lasts for longer periods of time, and they provide access to jobs that pay enough to assure economic independence. But selec tion must be made with care, and the prime sponsor should maintain a liaison with and access to external supportive ser vices to increase the chances of enrollee survival. Community colleges, technical colleges, and area voca tional schools rank next in quality. In recent years these in stitutions have experienced a vast expansion and qualitative upgrading. The facilities and equipment are generally of high quality, and the occupational offerings tend to be broader than at the proprietary schools. Some CETA prime sponsors may be overly attracted to such institutions because of the comparatively low cost. Most CETA enrollees can gain en trance, but the challenge is to survive. The institutions have adapted to the needs of the average high school graduate. They rarely have available remedial adult basic education, English as a second language, close counseling support, and other supportive services likely to be needed by the CETA population. CETA referrals who can survive in that setting are likely to be brought into contact with the primary labor market. Most such schools have informal and formal employer contacts. The enrollee emerges from a mainstream institution rather than from a stigmatized federal program. The enrollee may well be inspired to a higher self-image as well. As the case studies illustrate, some of the survival prob lems for CETA enrollees in these mainstream institutions can be minimized by the prime sponsor staff working closely 40 with the training institutions. The options are to persuade the institution to mount its own supportive services ac tivities, allow the stationing of agency personnel at the train ing institution to provide the social services, or arrange to provide access to supportive services external to the training institution. North Carolina has attacked the problem with its 8-week human resource development program, given prior to enrollment in occupational skill training. The third institutional tier appears to be the skill centers. They are typically designed around an open entry-open exit concept in which an enrollee can enter regardless of background, obtain remedial education and advocacy counseling, enter individualized training without waiting, progress through a modularized course sequence, and seek employment upon attainment of a skill. They may also leave when they have reached their learning capacity or upon becoming financially pressed. All the necessary remedial and supportive services, including placement services, were designed to be available onsite from skill center staff or outstationed staff of other agencies. The survival chances of the less-qualified CETA eligibles are enhanced in the skill center environment. However, the facilities suffer in quality and tend to be limited to occupations in which entry for a disadvantaged person can be attained in 6 months or less of training. The training provided by community-based organizations offers the fourth tier in quality. As usual, there are excep tions. As noted above, the Seattle OIC occupies modern training facilities and provides superior training. That is a unique case, however. CBOs generally operate out of haphazard facilities, just as skill centers do, and, although high quality training can occur in a substandard facility, there is at least an image handicap. The comparative advan tage of a CBO is racial and ethnic identity. The few CBOs that offer occupational training tend to depend upon the 41 charisma and commitment of staff, but the overall perfor mance is spotty. Contracting with CBOs for CETA services involves more than straightforward judgments of training quality. Aside from political considerations, which may be controlling, CBO personnel can provide the enrollees with a sense of identity and can offer the program greater visibility in the community. In some localities this may be sufficient reason for funding groups that mix training with a good dose of ad vocacy. In general, the messages of these case studies suggest that it is best to leave to CBOs the primary functions of outreach and intake, remedial components, and affirmative action sponsorship on behalf of their particular clientele. Since these activities are generally new to everyone in the community, a CBO can mount them as effectively as any other available institution. Also, the CBOs are more likely to be aware of the needs of the eligible population, and they have greater flexibility than the educational institutions. Oc cupational training is, however, at its best when assigned to schools or employer settings with satisfactory facilities and recognized competence. The case studies demonstrate that the paucity of funds for facilities and equipment is a distinct obstacle to CETA train ing. The institution acting as contractor to provide training normally furnishes its own facilities and equipment. DOL pressures are against providing sufficient funding to upgrade facilities and equipment, and prime sponsors are reluctant to use scarce funds for these purposes. Private schools, of course, charge enough for their training to cover amortiza tion. Public educational institutions provide a substantial subsidy to CETA through free use of facilities and equip ment as well as through tuition charges that are well below costs. Skill centers and CBOs must negotiate with prime sponsors for sufficient funds to improve their facilities and 42 upgrade their equipment but have had to overcome prime sponsors' resistance to cost increases. As noted, prime sponsors lack the resources to provide training in areas where no effective training institution ex ists. The differences were critical in two of the three Penobscot consortium counties, in some of Utah's isolated areas, and in Ottawa County. The choice left to the prime sponsor in these areas was to make do with the available in stitutions, limit the program to work experience, or train ap plicants at some far-away institution involving travel ex penses and living costs. Curricula and Staff The quality of occupational skill training curricula ap peared to be generally good, though each institution was be ing forced to adapt materials prepared for the mainstream. A system of sharing curriculum within CETA training circles would have been of considerable value. Curricula for remedial adult basic education are now reasonably well developed and available "off the shelf" from several commercial publishers and educational systems. English as a second language is approaching that status with an informal network of exchange among practi tioners. Fortunately, a separate curriculum does not seem necessary for each language. San Francisco, where all ESL instructors share community college certification, has its own multilingual informal interchange, while ELS materials for Spanish-speaking trainees have been circulating since MDTA began. Vocational English as a second language (VESL) seems to be the code term for this growing multi lingual interchange. San Francisco, Seattle, and Tucson of fered the best training noted in this field, although no case study author identified ESL as an area of curriculum weakness. The anxiety of the immigrant enrollees for pro- 43 gress and the ethnic identification of the instructors are un doubtedly positive qualitative factors. In contrast, limited curriculum material seems to have been designed for orientation, motivation, and various cop ing skills such as grooming, personal finance, and how to gain access to public services. The goals of orientation are not clear nor do we know how to motivate employees, as the experience of high-priced consultants to industry clearly il lustrates. Job search training is too new to have developed a widely acceptable curriculum. Most of the literature in the field, developed in support of career choice and access to professional jobs by college graduates and displaced ex ecutives, has limited relevance to a CETA-eligible popula tion. Consulting firms have begun to compete vigorously for assignments from CETA prime sponsors, but many unresolved methodological and conceptual issues concerning curriculum approaches remain. Quality at this point is likely to rest more with the charisma and good sense of the workshop leader than with curriculum content. Staff development seemed to be a serious problem in the CETA training system. For institutions outside the training mainstream, including skill centers, there is ordinarily no re quirement for preservice teacher training and no linkage to ongoing inservice teacher training and staff development systems. There are no funds or provisions in CETA contracts for upgrading contractor personnel. There is none of the leisurely pace of public education. Instructors are paid by classroom hour and have no built-in incentives for selfdevelopment. The generally high quality of instruction must be attributed to personal dedication rather than to institu tional incentives. Formal staff development ought to make it even better. North Carolina went furthest of the 12 case studies in attempting to meet that need but the full potential effectiveness of its state-financed Employment and Training Institute was thwarted by political infighting. Still, it is a 44 model worth consideration, since it is close enough to the local level to allow general staff attendance and onsite assistance, yet has a large enough scope for economies of scale. Occupational and O JT Quality The training occupations available to CETA enrollees are limited by the policy of keeping per-capita costs low and, therefore, training of short duration. The average classroom course length is 5.5 months, although the law permits pro grams of up to 2 years. Nonetheless, all of the prime spon sors indicated the imposition by regional officials of more stringent de facto limitations, which prime sponsors tend to enforce in response to pressures for maximum enrollments. In effect, therefore, CETA training is limited to 1-year pro grams even though some prime sponsors, notably Penobscot, enrolled their participants in 2-year programs, with the understanding that they had to pay their own way the second year. The availability of basic education oppor tunity grants facilitated this approach, and the grants were used extensively in several locations to reduce the cost of the initial support. Enrollment of 1 academic year or less in a low-tuition occupational program in a public institution seemed to be generally available for those meeting both CETA eligibility and the institution's entry requirements, but the number who could qualify was limited, except in San Francisco, Seattle, and Utah—areas with a generally high average level of education—which generated waiting lists of qualified applicants. Occasionally prime sponsors attempt to generate class-size CETA projects within mainstream training institutions. Ex amples include Montgomery County's biomedical technician program, electronic technician, computer operation and pro gramming courses in North Carolina, Penobscot, Worcester, 45 Baltimore, and Dallas. Most of these advanced projects were funded under the skill training improvement program, which had its own separate, more-generous funding and lessrestrictive eligibility standards than the run-of-the-mill CETA training. The skill center programs offer occasional departures from the standard meager menu because of the special and temporary employer need, but the pressures for short train ing time, low costs, and immediate placement create a cen tripetal force back to the basic grouping. Tucson's elec tromechanical assembly and bank teller programs, air condi tioning installation courses in Dallas, and Seattle's addition of maritime trades to the skill center's offerings are examples of efforts to meet special local needs. The difficulties of recruiting and retaining a sufficient number of eligible enrollees qualified for the training outside of the customary clerical, health, automotive, welding, machine operations and food and building service are illustrated by the class-size computer programming and operation courses in Baltimore and at the Dallas QIC. Little is known about the quality of on-the-job training. CETA prime sponsors are usually one additional step remov ed from OJT employers. As noted, the prime sponsor con tracts with a CBO, the local public job service, or the Na tional Alliance of Business chapter to contact employers and place CETA-eligible enrollees with them. Some subcontrac tors make periodic visits to employers and some do not, and prime sponsor staff also make spot checks. But the staff making these checks are not training experts. OJT can range from formal inplant courses to working under the watchful eye of a supervisor or being assisted by a fellow employee. The first issue is, did the employer hire a CETA-eligible per son because of the training subsidy rather than an employee not eligible under CETA who would otherwise have gotten the job? The payoff is whether the enrollee is retained in an 46 unsubsidized job. The fact that such retention occurs in most cases undoubtedly explains the highly attractive benefit-cost ratio documented in the national review. Supportive services are an essential ingredient of CETA training that affects its quality. MDTA pioneered in the development of client assistance—child care and transporta tion—to ease participation in occupational skill training. CETA continued to provide these services, although other programs have shared in the responsibility for their funding. Even more than MDTA, CETA has emphasized acquiring the personal attributes of employability, whether or not the client participates in occupational training. Counseling remains a supportive service whose value is taken on faith in the absence of any strong evidence about the extent to which it makes a difference in participant out comes. The same applies to training for job search, which is a more recent development without an articulated common curriculum and to orientation and motivation activities, which are too diverse and amorphous to make assessment of these approaches any more than a matter of faith. Need for transportation assistance was significant only in Dallas and Penobscot. Most child-care facilities and services were provided by non-CETA agencies and did not appear to be a serious problem to prime sponsors. In general, the nontraining supportive services have been taken over by other in stitutions during the past decade, while the nonoccupational labor market skills such as job search have exceeded them in importance within the CETA program. Linkages and Sequence A significant consideration in assessing the quality of a CETA training program might be the degree to which it is linked with other CETA components, with programs in 47 mainstream education institutions, and with employing organizations. The scarcity of such linkages was one of the disappointing findings of the study. An advantage of decen tralized administration should be the ability of the prime sponsor to orchestrate passage of the eligible enrollee through a sequence of locally available services, starting from the enrollee's initial need and completing with having attained employability and a job. Such sequencing was rare, however. Only three of the 12 prime sponsors centralized their total intake, and two others did so for part of their clientele. The norm was for a service deliverer to be responsi ble for its own recruitment, selection, and assessment of clients whose access was limited to the services provided by that contractor. The general tendency was to make one pro gram referral per enrollee and provide nothing further, ex cept perhaps placement services after completion. There is no legal or regulatory obstacle to starting an in dividual in ESL and adult basic education at one institution, moving to skill training at another, followed perhaps by an on-the-job training stint and job search training from another institution, as needed. The obstacles are conceptual, logistical, and financial. Prime sponsor staff generally lack interest in and capability to design comprehensive delivery systems, to chart client flows, and to generate realistic and meaningful employability development plans. A systematic client tracking system is necessary to monitor the enrollee's progress, and it is difficult to assign accountability and measure contractor performance in an interdependent system. Reporting requirements further discourage mixing components. The financial drawback is the amount of money to be spent per individual. The longer amount of time necessary to traverse the complex programmatic terrain in creases the amount spent for allowances. The standard practice of the prime sponsors studied was to provide no employment and training services directly. Only 48 the skill centers and the North Carolina community colleges and technical institutions provided both nonoccupational and occupational training within the same training institu tion. Most other institutions performed only one or a limited range of services. Absent centralized intake centers, there were no means of arranging a needed sequence of services to be supplied in turn by service deliverers. Each contractor had every incentive to keep its applicants inhouse and little or no incentive to refer them to other institutions for alternative or supplemental services. But centralized intake is not enough. There must be some means of assessment to determine enrollee capability and need. There were some exceptions to these generalizations about sequencing and relations with the education community. Penobscot operated its own intake and assessment centers and tried to develop a realistic employability development plan for each participant. It was able to determine who need ed adult basic education, refer them to that service, and then on to occupational skill training when the desired level had been achieved. Through the assessment process, only those with good work habits were referred to OJT. PSE was visualized as being OJT in the public sector with persons referred there first to learn and practice skills and then to be referred to either classroom training or private sector OJT. Penobscot was the only prime sponsor studied that seemed to be able to use the employability development plan as an instrument for sequencing individuals through multiprogram involvement. Despite operating its own intake centers, there was little attempt in Baltimore to provide sequential services beyond the first referral. Dallas had contracted with a CBO for a centralized intake and assessment center but there was no continuing linkage between that center and the organizations to which the individuals were referred and no arrangement for later sequential referrals. The Washington state employ- 49 ment service operated a well-staffed assessment center to recommend choices among OJT, individual referral, the skill center, and occupations within them. Tucson had con siderable movement of individuals from adult basic educa tion and ESL and other nonoccupational activities at CBOs into occupational training at the skill center. This was ac complished by allocating a fixed number of slots at the skill center to each CBO that assumed the responsibility for allowance payments and placement efforts on behalf of those individuals. The skill center then provided vocational assessment to help in the choice of training occupation. Beyond these examples, assignments were based upon availability of openings or enrollee choice. Except in Seattle, all skill centers studied were units of the public education systems. Yet the latter invariably treated their skill centers as stepchildren and established few, if any, linkages. Tucson and Baltimore seemed to have the most significant linkages with the employer community. In Tucson, a joint and overlapping economic development council, prime spon sor advisory council, and private industry council was a useful device for program planning and for development of linked classroom and O JT programs but not for direct place ment. Baltimore, more than the other sponsors, seemed to have worked out effective continuing program planning and marketing relationships for direct placement with cooperating employers. By and large, CETA appeared to be a mechanism for plac ing resources into the hands of training institutions and channeling eligible individuals into training programs. However, with the possible exception of Penobscot and to a lesser degree Tucson, nowhere did any of the 12 prime spon sors studied develop the institutional arrangements that are necessary for a sequential training system. 50 The Quality of Evaluation The 12 case studies indicate that federal administrators displayed little concern about training quality and few prime sponsors appeared to have staffs qualified to assess the quality of training offered and to recommend improvements in it. The only prime sponsor staff with a continuing assign ment for direct onsite observation are the program monitors, but theirs is an entry level position characterized by high turnover—either up or out—which prevents the accumula tion of experiences upon which to make valid judgments. Training quality appeared to be more often an accident of in stitutional availability. Fortunately, the accident happened more often than not, so that one can report favorably on the general quality level of CETA training, including the facilities, equipment, curricula, and staff. But what about the results? Regrettably, the information does not exist to measure accomplishment at the local level. Not one of the prime sponsors studied could produce reliable and comprehensive statistics proving the long-run results of their training efforts. All maintained placement and positive termination rates and pre-entry/postparticipation wage rates because those were required by the DOL reporting system. A few maintained followup data for up to 6 months after train ing, but most did little followup. Baltimore had the most ex tensive evaluation program of the 12, with a separate office of program evaluation and research to carry it out. Rather than evaluate on a contract-by-contract basis, the evalua tions were structured around service components in order to ascertain which are most effective for whom. However, the evaluation effort necessarily relied upon measures of shortrun rather than long-run gains. Dallas is an example of a prime sponsor which contracts for 3-, 6-, and 12-month followup, but obtains no information on program impact. The purpose is to assess contractor performance but not to 51 test the worth or outcomes of the program. Penobscot follows up its enrollees but does not record results separately for each service component. Completion rates were generally reported by the prime sponsors to be in the 80 and 90 percentiles, a remarkably high achievement if the data are correct. Placement rates typically were reported to range from 65 to 85 percent, but was the determining factor the quality of the training pro gram, the state of the local economy, or the competence of the job development and placement functions? No prime sponsor had conducted any controlled study to determine how the gains to the participants fared compared with those experienced by a like group of nonparticipants. Only San Francisco had attempted to calculate cost-benefit ratios to be used as a management tool for the allocation of funds among contractors and services. Participant costs were compared to the annualized first postenrollment placement wage. In some of the ESL programs particularly the gains were spectacular, but hardly surprising considering, for ex ample, a Cambodian pre-entry wage compared with a San Francisco post-participation wage. Since the ESL placement rates were consistently over 80 percent and retention rates even higher 6 months later, there is no reason to doubt the positive thrust of the findings. Even excluding the unique San Francisco situation, the average wage gains were significant, considering that many participants had to settle for jobs in secondary labor markets. The boosts in average hourly rates from $3.36 to $3.97 in Dallas and from $3.60 to $4.25 in Baltimore are typical examples. The natural concern of a prime sponsor is the relative ef fectiveness of service alternatives—does classrom training, on-the-job training, work experience, or some other alter native service provide the most placement per dollar of ex- 52 penditure? Whether the unserved did as well as the served was a question which they felt no obligation to answer. In volved as they are in day-to-day operations, prime sponsors do not place high priority on determining whether the benefits from CETA services exceed the costs of those ser vices. Process evaluation of the CETA system is con siderable, perhaps too much so. Outcomes evaluation at the local level is rare and the national system, though thorough, involves long-term lags. Because they were limited to im mediate placement data, prime sponsors are often led into less than cost-effective strategies. Work experience and short-duration training produce equal immediate placement rates at lower costs in comparison with longer-duration training, which is nonetheless shown by the national longitudinal data to have the greater ultimate payoff. 4. The Management of Training The CETA system involves a partnership of federal, state, and local governments, with advice from other labor market participants, for the delivery of services designed to improve the employment and training experiences of eligible unemployed and economically disadvantaged persons. While this study focuses on training, it is necessary to assess how training fares in this milieu. Understanding and appraising the system in which training decisions are made requires an examination of the decisionmaking process. The Decisionmakers The CETA decisionmaking process involves elected of ficials, prime sponsor staff, advisory bodies, contractors and subcontractors, client groups, and the public at the local level, all interacting with state officials and agencies, federal Department of Labor officials, and, ultimately, the Con gress. Each has varying impact on the nature and quality of the training delivered, but not necessarily on the outcomes. Elected Officials Few elected officials were deeply involved in the dozen cases under scrutiny. Some ignored CETA's existence. Some asked for periodic briefings to assure that no unforeseen political dangers were lurking within CETA's complexities. Only one, the mayor of Baltimore, perceived CETA as play ing any significant role in his plans for his jurisdiction and administration. For him, CETA was a key resource for ser vices to his constituents, a welcome linkage between social 53 54 services and economic development activities, and its direc tor was a valued lieutenant in his administration. Most chief elected officials were satisfied to leave the management of the prime sponsorship to the staff, provided that CETA operations did not cause political embarrass ment. In the two cases where the governors were the chief elected officials, their role was even less than that of mayors and county officials. Since decisionmaking was largely decentralized to associations of government in North Carolina and Utah, the systems may simply have been too amorphous for the governors to have means for participa tion. In Utah, the previous governor had resolved to shift from intense personal involvement to extreme decentraliza tion, and the incumbent governor had not reversed that trend. In North Carolina, CETA has often been a political issue, but no governor has been much concerned with its substance. Two-thirds of the North Carolina county com missioners involved would have preferred the abolition of CETA, which seemed also to be the preference of the Dallas city council. The finding that the elected official's involvement is not crucial to effective CETA administration does not mean that the role does not exist. Elected officials were interested and evident where political sensitivities were at stake. They occa sionally overruled their directors after counting political costs. That meant, however, that their interest level was like ly to be high in relation to public service employment and low in relation to training. The Staff Directors The key decisionmaker in a local CETA system is the prime sponsor staff director. This official is most often the prime determinant in deciding how the local prime sponsor ship will respond to the local political, economic, and 55 demographic conditions in the usage of CETA funds. From the 12 case studies, it is possible to derive a profile of the ef fective staff director. Yet there is no evidence linking that profile with outcomes of the program, as measured by employment and income gains of participants. Of the 12 prime sponsors, four stood out as the most effective leaders, whereas two others were so new that there was inadequate evidence of their eventual effectiveness. What marked the leadership effectiveness of these four was their ability to con ceptualize the CETA system for their locality, derive a set of objectives consistent with the local economic and political mix, design a realistic program consistent with those objec tives, and then direct the human and financial resources of the prime sponsorship toward the achievement of those directives. The conceptualization might not be that which was in the minds of CETA congressional authors, and the objectives might not be those espoused by the national and regional offices. However, survival demands objectives and approaches that are realistic and desirable within the local context. Considering the long-run interests of the CETA-eligible population, an effective program is one in which (1) ap propriate priorities are made for specified reasons among those eligible for service; (2) the barriers impeding the employment of individuals in the target groups are identified as clearly as available data make possible; (3) a mix of ser vices is explicitly selected which takes into account the needs of the target group, the capabilities of potential service deliverers, and the realities of the labor market; (4) service deliverers are chosen on the basis of their ability to deliver quality service; and, (5) the outcomes are as favorable as the state of the labor market and the nature of the employment barriers allow. The staff director must be sensitive to outside pressures and constructively accommodate the prime spon sor's objectives with the mix of political interests dominant 56 on the local and national scene. Effectiveness in each situa tion must be judged in relation to alternatives that the prime sponsor might have adopted. Cross-prime-sponsor com parisons are of limited value because of the widely varying circumstances. Baltimore, San Francisco, Penobscot County, and Tucson are examples of prime sponsors operating under strong and effective leadership. Sex certainly was not the determining factor. Two of the four more-effective leaders were women and two were men. All four have a clear vision of what they believe CETA objectives to be, how much can be ac complished within their political, economic, and budgetary constraints, and are aggressive in pursuing their aims. The stability or the strength of the political leadership does not create the staff leadership. Only Baltimore of the four cited prime sponsors could claim an elected chief executive concerned enough with the employment and training arena to contribute to the creation of a leader in his own image. The San Francisco staff director had served under three mayors, none of whom were especially enamored of CETA or deeply involved with it. The Penobscot director served under a corporate leadership of nine county commissioners, none of whom loaned substantial strength to the CETA operation. One Tucson mayor spanned the entire CETA ex perience, supported his staff director, but did not involve himself in CETA affairs. Except in Baltimore, the chief elected official never added to or subtracted from the effec tiveness of the staff director. The Baltimore staff director could undoubtedly carry the load by herself but has been aid ed by the strengths of her mayor. On the other hand, political leaders established a climate in North Carolina and Dallas, Texas in which no strong leader could have or would have persisted. Eight successive directors served the former, usually until abruptly removed. 57 The pattern of the latter was to view the role as a temporary assignment to be moved beyond as soon as possible. In other cases, the political environment was essentially neutral. Leadership can drag a program down as well as build it up. Both the Utah and Seattle prime sponsors began under nationally recognized leadership and their programs were considered among the best in the nation. The original Utah leadership left when the governor backed away from support of a strongly centralized state program. The Seattle director left because he felt the Congress and the incoming ad ministration in 1977 was unduly restricting local autonomy. He has not been replaced by leaders of equal stature as the Seattle program subsided into mediocrity. Concern has been expressed about the staff director's stability of tenure. However, these case studies should be reassuring. No strong and effective staff director was replac ed as a result of a lost election, though some were removed in internal political squabbles. On the other hand, those with effective leadership in the small group of cases have never ex perienced a change of political party in an election or have been protected by a consortium structure in which all of the principals did not change simultaneously. Academic credentials appear to be irrelevant and ex perience seems to be paramount. All four of the most im pressive leaders preceded CETA passage in their manpower program involvement. Typically, the less-effective leaders were more recent entrants, leaving one to wonder whether experience built strength or only the strong survived. Salary and job security are not the explanation of strong leadership, though they may often encourage its absence. CETA directors and staff seem more responsive to challenge than to salary. But why would an effective leader accept the CETA assignment? The rewards are totally in a sense of ser vice. CETA salaries are usually no more than and often less 58 than those for comparable jobs in state and local govern ments. Why would persons give up other alternatives to function in the unstable and uncertain world of CETA? The motivations appear primarily intrinsic and often seem to in volve a commitment to improve social conditions and con tribute to the alleviation of unemployment, poverty, and discrimination. The answer may have to come from a psychiatrist, but even in these cynical times dedicated people can still be found. That leaves a conundrum: How can a program attract strong leadership? Must it remain a happy accident? The search must be for independent spirits with sound ad ministrative skills. They will always be scarce but not nonex istent. Ultimately, for a system to survive, the extrinsic motivations must be strengthened to support the intrinsic ones. Staff Conventional wisdom has it that one of CETA's major problems is the high turnover of prime sponsor staff. The 12 case studies do not support that generalization. Of the 12, all but two, Dallas and North Carolina, had in fact experienced remarkable stability, considering the limited access to the customary devices for job security. In fact, the Dallas CETA staff were city merit-system employees and that seemed to contribute to, rather than minimize, turnover. Two prime sponsors, Montgomery County and Seattle, had each ex perienced a substantial one-time turnover consequent to changes in directors but had experienced staff stability before and after. Utah experienced a major turnover in the leadership of its pioneering state manpower planning effort, but that was on the eve of CETA passage and stability had prevailed since. All of the others had experienced continued stability in all of the key management and technical levels. Turnover of lower-level and nonpolicy staff occurred but did 59 not significantly affect policy or operation. Possibly one ma jor contribution to staff stability has been the fact that over the first 6 years, CETA was a growth industry and able staff members had ample opportunity for upward mobility within the system. Dallas reflected a hostile political environment where political leaders apparently wished that CETA went away, and at least on one occasion the city council narrowly defeated a proposal to refuse CETA funds. A CETA assign ment was just another job to staff who expected to be pro moted and transferred soon, and many were. There was no noticeable commitment to the CETA mission. North Carolina has suffered from fluctuating political perceptions of what CETA could and should be and failure of anyone to visualize a consistent and viable mission for the sprawling balance-of-state structure. Yet the high turnover was limited to the director with the staff experiencing no higher turnover than typical in state and local government. On the other hand, stability existed in situations with and without merit-system protection, with low pay and high. Two situations seemed to contribute to staff stability: (1) ef fective staff directors and an organizational sense of mis sion—being part of an effective and committed organization was apparently attractive enough; and, (2) an abundant sup ply of the college-educated who depended upon CETA for scarce job opportunities. Utah and Ottawa County were ex amples. The living environment was attractive and there were few alternatives for college-educated people without technical and professional skills. That does not mean these staffs lacked competence. They were able people but without a notable sense of mission. Dependence of staff stability on leadership strength and agency commitment is in some ways a disappointing, though not surprising, finding. Putting in place higher salaries, job security, or a training program for staff development is much easier than finding and develop- 60 ing competent leaders or generating from the top down a sense of mission. Not turnover but accretion had been the staffing problem of the 12 prime sponsors. All the prime sponsors report rapid increases in staff following the passage of the Youth Employment and Demonstration Projects Act in 1977; it was not unusual for the prime sponsor staffs to increase fivefold or more within several months of enactment. Advisory Councils The case studies were reassuring as to the role and con tribution of the advisory councils mandated by CETA law, again in contrast to what has been reported from other CETA studies. In about half of the prime sponsors studied, staff followed the advice of the advisory councils on most issues. In fact, a council vote was considered the final word by several prime sponsor staffs. The conditions that divided the prime sponsors into two groups on this issue are instruc tive. The effective leaders seemed to put the greatest respon sibilities on their advisory councils, and advisory councils seemed to respond when they had decisionmaking power. Councils did not generally work well where the CETA operations were spread beyond the local community lines, as in Ottawa County, North Carolina, and Utah. However, the Penobscot staff director had been able to develop a useful council role by having three councils, one for each county in the consortium, and some of the separate advisory councils to substate planning regions have worked effectively in North Carolina. Advisory councils also did little where their function was not considered important by the chief elected official and staff, or where their advice was ignored, or where little of significance was happening under CETA. Council subcommittees were active in recommending the relative priority to be given to the various target groups and 61 in reviewing and ranking program proposals, but had no means for judging quality by other than gross outcomes. They supplied a buffer for the staff in a politically sensitive function. They tended to show sound judgment and firmness in choosing and rewarding contractors and programs capable of demonstrated effectiveness, as measured by costs, placements, and other outcomes. They were not sufficiently knowledgeable to judge the quality of program content and conduct. By design, the case studies paid little attention to youth programs and the youth advisory councils established by the 1977 law. In general, it appears that these councils had not found a meaningful role, Private industry councils had a more programmatic than advisory role, but were relatively new when the case studies were prepared. A problem was the lack of rationale for three councils per prime sponsor. The Tucson model, which effectively combined the prime spon sor advisory council and the private industry council with the local economic development advisory council, seemed to have the most to offer. A persisting controversy in CETA has been whether representatives of service delivery agencies should be allowed to serve on advisory councils, which make recommendations concerning choice among service delivery agencies. This con troversy involves particularly the job service and community-based organizations. To avoid conflict of in terest, federal regulations forbid representatives of organiza tions that deliver services to vote on decisions affecting their own funding, but membership on advisory councils is not forbidden. Most of the prime sponsors studied had taken steps to limit the proportion of service deliverers among ad visory council membership, and at least one relegated them to nonvoting status. However, the restriction did not prevent the representatives of service deliverers, particularly private contractors, to affect decisions by lobbying council members 62 and elected officials. The job service and the public educa tion officials were sometimes not sufficiently interested to pursue a role aggressively, or in other cases appeared to con sider such pursuit demeaning. In general, private service deliverers, by aggressive lobby ing, were able to preserve roles for themselves as long as they were at least moderately effective in performing their respon sibilities. Eventually, all the prime sponsors in the cases studied tended to "bite the bullet" and dumped clearly inef fective contractors, despite political pressure. Marginal con tractors, however, were often continued rather than do political battle over debatable issues. The Feds The prime sponsor perception of the federal role in CETA ranged from highly to mildly negative. Few favorable com ments were heard concerning any federal decision. That many of the decisions complained about were products of congressional action, not the U.S. Department of Labor, did not appear to be recognized by many at the prime sponsor level. However, it appeared to be universal judgment that the department's regulation writers tended to compound con gressional restrictions. The national office was perceived as having no concept of the impact of its administrative deci sions at the service delivery level. Meddling threats—rarely carried out, but disruptive of operations—rather than technical assistance appeared to the prime sponsors to be the preferred remedy for any apparent transgression. Judgments as to national office competence and intentions were no less harsh in Labor Department regional offices than in the offices of the prime sponsors: "They don't know what they want or what life is like outside of Washington." A par ticular complaint of regional staff and a source of derision from prime sponsor staff was that any subscriber of the na- 63 tional reporting services knew of Washington decisions days to weeks before the word reached officially the regional of fices, and the latter could not (or would not) take action until informed through formal channels. Then, higher regional levels were perceived as compounding the delay by rewriting the national directives into regional directives before sending them on to the field, often, it was claimed, distorting the meaning and compounding the confusion in the process. One must interpret such complaints with care. It is to be expected that subordinate agencies and staff will complain about those who set the rules and hold the purse strings. It is also to be expected that national oversight agencies will con sider the subordinate as insubordinate and incompetent, especially to the extent those who oversee have more respon sibility than authority. Nevertheless, there does appear to be a real problem in contrast with the past. In pre-CETA days, a substantial number of federal operatives at regional and national levels had "come up" through the federal-state employment service and through state vocational education agencies, which were also the major providers of federally mandated services. No important national office positions are currently held by former members of prime sponsor staffs. The same is true, by and large, at the regional level, though a few are beginning to emerge at the lower levels there. Thus, the federal staff is responsible for functions they have never performed and that are foreign to their ex periences. When the alternatives are remedial technical assistance or threats, they are incapable of offering the former and resort to the latter. For the prime sponsor, the visible test of federal com petence is the regional office field representative—in CETA parlance the "fed rep." This individual, who is expected to provide onsite supervision and technical assistance is at the bottom of the regional office hierarchy, rarely has any previous relevant experience, and is given little helpful train- 64 ing. In the 12 case studies, only one "fed rep" gained praise—from the North Carolina observer—as a consistently positive influence on the prime sponsor. Two other prime sponsors felt encouraged that after a long series of bad ex periences, new fed reps appeared more helpful than those of the past. Significantly, in one of the latter cases, the federal representative had come to that post after several years with a competent prime sponsor. Otherwise, the strongest praise was "he doesn't bother us very much." After North Carolina, the most sanguine of the prime sponsors in their attitudes toward the fed reps and the regional offices were the Baltimore, San Francisco, and Utah prime sponsor directors. The first two both had the security on the local scene and the reputation nationally to feel in vulnerable to regional criticism. Both also were sufficiently close to the national scene to realize that Congress and the Washington officialdom were the source of unwise decisions rather than the regional staff. In Utah, the state CETA of fice shielded the associations of government, responsible for program operations, from direct federal contact. Just where the balance of truth lies in the federal-local and national-regional relationship may be difficult to ascertain, but the relationship is clearly not a productive one. The Dallas case study provides a summary of intergovernmental relations that characterizes well the frustrations of the cur rent situation. A study made from the prime-sponsor and national-office levels can provide little insight into the congressional decisionmaking process. However, it may be useful to contrast apparent congressional perceptions of the local decisionmakers with those gained by knowledgeable observers mak ing indepth studies. Common complaints against Congress included: (1) overloading the system with too much public service employment too fast before the prime sponsor system 65 was securely in place; (2) adding to that overload a continu ing stream of complex new programs; (3) compounding pro gram complexity by the detailed 1978 legislative re quirements; and, (4) being invariably late with its appropria tions. The latter presents a serious obstacle to businesslike opera tions. A prime sponsor must plan for a year's service delivery without ever knowing within even a reasonable range what the funding levels will be. To have contracts in place by Oc tober 1, the start of the federal fiscal year, a prime sponsor must begin the planning process by January or February. The Labor Department promises budget estimates and na tional policy constraints by May 15 but almost never delivers on that promise. When the funding estimates arrive, they are no better than a prime sponsor could do from reading the na tional reporting services. The Labor Department appropriation is almost never in place before the start of the fiscal year, with continuing resolutions governing for one, two, or three quarters or even through an entire fiscal year. But that is not the end of fiscal uncertainty. Deobligations of unspent funds make redistribution possible throughout the year. New congres sional initiatives often provide supplemental appropriations. The Labor Department persistently vacillates over whether and how much of the carryover funds from the previous year the prime sponsor will be allowed to spend. Only after the fact can the prime sponsor determine how much money has been available. Planning and Performance This does not mean that CETA planning does not exist. Planning is the management function that sets the direction for future activities of the organization. Like Moliere's hero who spoke prose all his life and did not know it, prime spon- 66 sors must and do plan whether they know it or not. But it is contingency planning, fraught with uncertainty and laden with frustration. In addition to the delays in key data elements, the compliance review system and the geographical scope of the prime sponsorship present formidable obstacles to prime sponsor planning. Ironically, the formal planning document submitted to the regional office may have little relation to the prime sponsor's realistic intention. A persistent complaint is that the Labor Department denies itself any meaningful oversight of pro gram substance by fractionalizing the programs. At the regional office, pieces of the local planning document are distributed for checking on compliance, but not for its coherence or substance. It is inevitably approved condi tionally and then frequently rejected for some procedural technicality. Meanwhile, the prime sponsor's real plan has been written into the contract documents with service deliverers, which specify who is to be served, what services are to be delivered, and what the performance criteria are to be. The Montgomery County case study provides an example of a situation in which the regional office failed to take of ficial notice year after year not only of poor plans but also of poor performance. As long as the form was observed, the substance was ignored. Only when the prime sponsor failed to spend its full allocation because it was incapable of serv ing the more-disadvantaged population mandated by the 1978 amendments did the regional office blow the whistle. Then, rather than offer technical assistance to solve the pro blems, the response was a punitive restriction of funding flows, which made planning and administration even more difficult and certainly did not help the population entitled to the services. The experience seems not to be uncommon. The regional office reviews annually the operations of each prime spon- 67 sor. The published "report cards" offer a revealing assess ment of what is deemed important by the national ad ministrators. The fiscal 1980 Title IIBC review revealed that 28 prime sponsors experienced "serious problems," involv ing "major barriers to the accomplishment of program goals." Seattle and Ottawa were included in the list. The lat ter rural county was found guilty of not assigning a staff per son to perform EEO functions—not discriminatory ac tion—and for not having taken "corrective action regarding outreach, training and advancement of the handicapped." Seattle was cited for having "difficulty in operating the eligibility verification system . . . ." Nationally, training quantity was mentioned as a source of difficulties in five cases. Of these, three prime sponsors had not spent 15 per cent of Title IID funds on training. Another was faulted for inadequate IIB performance reporting; and the fifth prime sponsor was apparently guilty of underutilizing vocational education setaside funds. By comparison, inadequate EEO compliance systems (again, not necessarily lack of actual performance) were mentioned nine times, and 21 of the 28 sponsors were cited for inadequate monitoring or eligibility determination systems. The quality of training was not men tioned as a serious problem for any prime sponsor. The priority concerns of the Department of Labor are reflected in the point values assigned in the 1981 assessment package for IIBC: Management Independent monitoring unit Eligibility determination, verifica tion, and tracking Financial management Planning (composition of council and process, only) Subagent management Equal opportunity Complaints Corrective action 6 6 8 4 8 6 5 7 68 Program Design Recruitment and selection of participants (says nothing about assignment) Assessment and employability development plans Job development and transition services Services to youth Program activities (assess two categories) OJT Classroom training Upgrading and retraining Work experience Corrective action followup Numerical Performance Indicators Positive termination Entered employment Indirect placement Private sector placement Cumulative enrollment Cost/positive termination Cost/entered employment Cost/indirect placement 7 7 7 4 18 (9) (9) (9) (9) 7 5 8 10 5 5 4 4 4 The message the prime sponsor receives is that quality, especially of training, as well as long-run results, has a relatively low priority and may not even enter into the deter mination of the report cards that the feds issue to the prime sponsors. The possibility of negative local publicity is more of a driving force than any available rewards for good management or quality programming. Yet advance in dicators of training quality are not readily available or easily derived. To prescribe input measures would limit diversity. Ultimately, long-run outcomes will have to be the basis for judgment. 69 Geographical Scope and Economies of Scale The original CETA legislation provided a bonus to en courage adjacent jurisdictions to combine into labormarketwide planning units. It is not working and the studies document some of the reasons. One is simply the value of the incentive. Tucson City and Pima County, Arizona lost less than 2 percent of their combined budget when they split up as a consortium. It was not a sufficient threat to dissuade the county supervisors from seeking control of their propor tionate share of the remaining funds. Money is power, and the exercise of power is what politics is about. If consortia are desirable, the incentives must be commensurate with their worth. Baltimore County and Snohomish County (ad jacent to Seattle) both withdrew from consortia even though surrounding counties remained in. The benefits of consortia were not enough to outweigh the attraction of autonomy. The Labor Department claims to be neutral as to the choice between consortia and individual prime sponsors, but the labor market planning concept would require positive en couragement of consortia. San Francisco, Seattle, Worcester, Baltimore and In dianapolis came nearest to having jurisdiction over entire labor markets. The results seemed positive for the first two, but of no particular significance for the others, which made no special efforts to adapt their operations to special local conditions. Tucson and Montgomery County had jurisdic tion over less than a labor market. The loss did not seem serious in the case of Tucson, which drew enrollees from the city but prepared them for jobs in the suburbs as well. However, it unduly limited Montgomery County, which had to compete for access to jobs, training institutions, and employers with other prime sponsors in a complex metropolis encompassing the District of Columbia and a dozen political jurisdictions in Virginia and Maryland. The 70 Penobscot consortium administered jointly and planned separately for each of three rural counties, and the results were impressive. Ottawa lacked training institutions and a budget to offer adquate training. The two statewide operations made no attempt to plan for or administer programs for the areas covered. They delegated most planning and administration to local associa tions of government, which left too few and fragmented resources to obtain optimal results. Apathy at the highest state levels appears to be the primary explanation for the unimpressive performance by Utah. The North Carolina balance of state just seemed too massive and complex to be manageable, even had there been the will to do so. Planning and related decisionmaking is another matter. A state, at least those as geographically large and diverse as North Carolina and Utah, is not a labor market. No single plan is likely to rationalize such diversity. Since the balanceof-state concept eliminates the major employment centers, employment and training planning is unlikely to produce a basis for sound decisions. Breaking up the less populous areas into labor market watersheds surrounding major employment centers might be more effective. Such an ap proach could probably be accomplished with adequate con sortium bonuses and encouragement, T>ut state staffs would have to develop the necessary expertise to provide sound guidance. Related to the issue of geographic scope are potential economies of scale in staffing, planning, evaluation, management information systems, public relations and other functions. In 1980, 77 percent of prime sponsors received less than $5 million in Title IIB funds, 75 percent had less than $3 million of Title IID funds, and 82 percent received less than $5 million from Title VI. Of the remaining prime sponsors having higher funding, half were balance-of-state 71 prime sponsors with some of the problems noted above. More aggressive promotion of consortia would help meet the economy-of-scale criteria as well as further the concept of labor market planning. The Decisions So much for the decisionmakers and the decisionmaking process. What can be said about the quality of the decisions themselves? As noted, those can be generally categorized as: Whom to serve; what services to deliver; and, to whom to assign service delivery responsibility. Whom to Serve? The whom to serve decision is resolved through an interac tion of law, regulation, politics, and objective judgments, probably in that order. The legislation confines eligibility to the "economically disadvantaged," representing a persistent tradeoff between concentrating resources on those most in need at the cost of imposing a negative image on some pro grams. Department of Labor regulations do not direct priorities among the "significant segments" eligible to be served by CETA, nor do the feds direct priorities for fund allocations for the groups who should share in the distribu tion of those funds. Among the prime sponsors studied, the aggressiveness and power of target-group members in the pursuit of services was a major factor in determining the racial and ethnic mix of trainees. Since they were a higher proportion of the CETA-eligible population, minority groups were overrepresented in all cases. However, how high their enrollment was in relation to their proportions of the eligible population seemed to be dependent primarily upon the effectiveness of the organiza tions representing each group. CETA staffs seemed to re spond to groups that applied pressure, thereby leaving less 72 resources for those that did not. In Worcester, for example, black enrollments in IIB exceeded the proportion in the population by a factor of 10. Since no one represented the nonminority poor, they received less than a proportionate share of the resources. Seattle, early in its CETA history, designed a system based on relative need and probability of success, but the effort was abandoned with changes of staff. In most cases, local chapters of national organizations car ried the battle for allocation. For poorly organized groups such as native Americans and, in some locations, Asian refugees, other organizations with service-delivery ambitions often pursued the fight. San Francisco is a unique example of a CETA scene dominated by homegrown communitybased organizations without national affiliation—as evidenced by the success of gays and lesbians, as well as foreign language groups, in gaining special attention. Ag gressive and sophisticated, these local CBOs have been ex traordinarily effective both politically and as service deliverers, and the system has responded accordingly. Seattle also had effective homegrown CBOs, but they were less numerous and, therefore, less obvious than in San Fran cisco. Advisory councils, in most prime sponsorships, were involved in conflicts over the distribution of slots among the eligible populations, and settled potential political conflicts in a setting shielded from the public gaze. Youth everywhere obtained a high proportion of the available program resources. However, this was not a func tion of organizational pressures, but of legislation and federal regulation. Youth had all of Title IV to themselves plus a federal regulatory requirement that the proportion of youth prevailing in Title II before the passage of the Youth Employment and Demonstration Projects Act of 1977 be maintained. Since youth unemployment had been recognized at the local level as a serious problem before the passage of the youth legislation, prime sponsors had been allocating 73 resources heavily in that direction. Most were therefore lock ed in with around one-half of their IIB slots reserved for youth. In general, the experience had been that expensive in vestments in training for youth below age 18 were not justified by their placement and retention rates. Therefore, the tendency was to relegate them to work experience pro grams, absorbing IIB resources that would probably have otherwise been spent on training. There was a wide range of responses to selectivity by education. Areas like Utah and California with relatively high proportions of high school graduates tended not to use education as a selection criterion except in occupations where course content or licensure seemed to require educa tion. In contrast, Baltimore, with a relatively low average of high school completers, was the most selective for its classroom training programs. The intent was to train those most likely to profit from the expenditure, relegating the less-educated to a work experience program containing its own remedial training components. As a result, classroom training concentrated on a higher level of skills. In general, however, the tendency was to spread the training across the educational range, referring the high school graduates in dividually to ongoing vocational and technical training pro grams and to OJT, with the lesser-educated concentrated in class-size projects. Given the additional need and cost of remedial education and the propensity of this population to drop out, this policy tended to restrict the deficiently educated population to short-term, entry-level preparation. In line with legislative requirements, it appears that prime sponsors carried out the intent of the law and drew enrollees from low-income families, although violations occasionally occurred and were played up in the media. With its generally favorable labor market, Dallas was most notable in concen trating its CETA training efforts on an extremely hard-toemploy population. It was also apparent throughout the 74 range of prime sponsors that the causes of disadvantage were many and that a wide range of competence and motivation existed among the eligible populations. It was no surprise to find that immigrants were most likely to combine income eligibility with potential for labor market success. It takes more careful screening to find that combination among the broader CETA population, but there are people who need only a boost to become economically independent. The Service Mix The prime sponsor has discretionary authority to allocate Title IIBC funds for training, work experience, or supportive services. None of the 12 prime sponsors utilized all the train ing funds allocated under the public service employment titles. Vocational education allocations through the gover nors' offices were a significant factor in promoting training, but no use was being made of the upgrading options of Title IIC. Since no separate funds were available for that purpose, there was no incentive to use the existing pool for upgrading the employed in preference to training the unemployed. The targeted jobs tax credit was also not popular in the 12 areas studied, though the Penobscot private industry council pro moted it aggressively during the summer of 1980. Lack of understanding by employers was the usual explanation of the neglect. The need for minimum reading ability and arithmetic comprehension made it essential to include adult basic education as an integral part of IIB training. The flood of immigrants into many of the jurisdictions was the major fac tor for the growth of ESL. The need to provide a transition service for PSE enrollees as shrinking funds led to termina tions, as well as the low costs and high placement rates associated with job search training, served as the motivations for the addition of that component. 75 Throughout, however, the prime sponsors, while declaring their preference for employability development, seemed to feel obligated to maintain balance among the full range of alternative services. They were under more pressure to spread the funds among familiar service deliverers than to adjust the service mix either to client need or to demonstrated effectiveness. Few prime sponsors had developed the technical capability to defend any other alter native. Most of the prime sponsors tended to be more responsive to enrollee need than to labor market demand. Baltimore was the notable exception. Through an extensive system of employer advisory councils, demand was determined and training planned accordingly. The underlying rationale was that there were more eligible applicants than training slots and that no one can profit from training in an occupation that is not in demand by employers. Therefore, the CETAeligible population is best served when employers are best served. The high proportion of clerical and health care occupa tions in the training mix of all prime sponsors was probably both a symptom and an explanation of occupational choice. Since such jobs were available, planners tended to choose those occupations in making training decisions. However, the high proportion of women applicants also pushed plan ners in the same direction. Efforts to place women in nontraditional jobs were few and limited to special projects for that purpose. Little success could be claimed. Training for male oriented jobs seemed to be more strongly supplycentered. That is, there appeared to be a greater tendency to search out occupations in which men might be trained and placed because there were fewer occupations where the de mand for men was obvious. 76 The Service Deliverers Prime sponsors must choose among a variety of deliverers of a variety of services. This report is concerned only with the choice among deliverers of training-related services. For classroom training, the prime sponsor usually has few op tions. No CETA prime sponsor has the necessary funding to create training insitutions. If a skill center is left over from MDTA days, it is used. If not, or in addition, individual referral occurs wherever there are institutions willing and able to accept CETA eligibles at reasonable cost. Where there are CBOs with any political clout, they are used unless they prove to be incompetent. Use of for-profit firms depends upon their costs and their aggressiveness. Never theless, within the limits of the availability of institutions, the studies demonstrate that prime sponsors do drop in competent service deliverers and continue on a small scale and at the margin to add and test new deliverers. They re spond to political pressures, but persistently they tend to ex pand use of the most cost-effective and shrink use of the op posite (to the extent they have dependable measures), all the while sensitive to the need to maintain diversity in offerings and institutions. The most marked changes have been the strong shift to in dividual referral and to private proprietary schools. Another marked change has been the enlarged training role of organizations that are not schools. For the most part, this is both a cause and a consequence of the growth of nonskill training. Community-based organizations and consulting firms live or die by their ability to sell their services. They tend to be much more aggressive than tax supported institu tions or proprietary schools, which appeal to a broad population. They push their wares by resorting to lobbying and salesmanship, thereby expanding their role beyond what sheer competence would have won for them. 77 The on-the-job component has undergone little change. Few employers have ever been interested in participation. An intermediary is necessary—a job service office, a community-based organization, or a prime sponsor itself—to contact employers and persuade them to accept the subsidy of one-half the enrollee's wages, which has become standard. Favorable responses generally come from small, struggling employers to whom the subsidy is attractive and larger firms who use CETA OJT as a recruiting source to meet affirmative action requirements. Whether the establish ment of private industry councils will be able to generate more OJT remains problematic. Based on early experience, there is little reason for optimism. 5. Lessons Learned The purpose of the case studies was to gain insight into the CETA system, not to evaluate the performance of the 12 prime sponsors. The studies illustrate a fascinating diversity. Penobscot showed the co-existence of strong staff leader ship, minor political involvement, effective advisory par ticipation, a weak economy, and limited service options, in ducing the prime sponsor to expand its training capacity. Worcester was characterized by prosaic leadership in a revitalized high-demand economy, but with a modest place ment record despite the favorable economic environment. Baltimore offered strong political and staff leadership, responsible for holding on to a consortium and making good use of generally mediocre service institutions in a redevelop ing economy. Montgomery County was in transition from strong but misdirected leadership to a more promising com bination in an economy of plentiful jobs usually requiring credentials. North Carolina had a sprawling service area and diverse administrative problems too challenging for its politically burdened leadership, but was helped by the availability of sound training institutions. Dallas had a most favorable economic situation with a potential for becoming a CETA showcase, but was thwarted by political leadership that just wished CETA would go away. Indianapolis was putting itself back together after an initially mismanaged and dif ficult period, but was still operating at high unit costs. It relied almost totally on individual referral to community and private colleges to do so. Ottawa County lacked the training 79 80 institutions to provide a comprehensive program. Utah was characterized by unaggressive leadership, but a strong economy and sound institutional support shored up the agency. Tucson had strong continuous leadership and sound in stitutional support in a low-wage economy. Despite effective training, it was unable to demonstrate success measured by earnings gains for lack of evaluation followup. San Fran cisco had enjoyed strong staff leadership over the years, was highly politicized, but by organizations that were also capable service deliverers. It had a favorable job market for women but not for men. Seattle had based a strong, but ex pensive, program on individual assessment by the state employment service, individual referral to community col leges by one CBO, and classroom training at a skill center run by another CBO. The System in Capsule To the extent these prime sponsors are representative, the CETA system clearly can and does serve its clienteles reasonably well and has generally adapted its operations to local social, political, and economic conditions. Yet a number of shortcomings reduce the total payoff. There is considerable operational planning in the use of each year's budget, but not much strategic planning relating the use of CETA resources to the broad and long-term needs and ob jectives of the areas served. However, the uncertain annual funding process makes the operational planning also con tingency planning, as prime sponsors are forced to live with uncertainty and adapt to changing priorities directed from Washington. Political and administrative pressures lead to minimizing per-enrollee costs, despite evidence that longer-duration training tends to pay off better than that of shorter duration. 81 Based on immediate postenrollment observations, results of work experience compare favorably with those from training programs, yet national longitudinal data disclose that the former has long-run negative, and the latter positive, results in cost-benefit terms. OJT payoff is generally highest in the long run but the prime sponsor is rarely aware of this fact. The available evidence of short-run social benefits and in dividual gains is not sufficient to overcome marketing dif ficulties and to attract employers. Local decisionmakers are prone to perceive the weaknesses in the system as being primarily federal in origin. It is true that the federal managers of CETA lack the necessary exper tise to provide guidance and technical assistance to prime sponsors and have tended to emphasize bureaucratic pro cesses rather than encourage creativity and provide substan tive leadership. A more concerted effort is needed to ex change federal and prime sponsor staff to familiarize each with the other's roles and problems. It is also essential that persons with training background be brought into the decisionmaking system at both levels. As the intergovernmental system is presently structured and functioning, the regional office has no meaningful role to play in the system. Its personnel have little policy discre tion, even if they knew what was best for the program. Regional offices are merely a poor link in the communica tion chain from the national to the local level since there are so many routes for information in both directions that bypass the regional office. At the top levels of the Labor Department's Employment and Training Administration, CETA administration began under the guidance of leaders who understood and were committed to the philosophy of the system. National leader ship dissolved into confusion and was just beginning to get itself together again when the 1980 election results introduc- 82 ed uncertainty not only about the future directions of the program, but its very survival. The Congress appears to be committed to local decisionmaking only in its rhetoric. It wants to handle the helm and leave the locals to respond, disregarding the administrative consequences of its constantly changing priorities. Local Staff Development The key determinant of CETA effectiveness is the strength and ability of the prime sponsor staff director. The attention of the elected officials can never be held for long, and when they do focus on the employment and training programs their primary interest is to avoid political embarrassment. Capable staff is essential but an effective leader will attract and develop a competent and stable staff. However, the measure of a leader in the prime-sponsor context is the abili ty to accommodate a wide range of diverse social, political, economic, and personal interests, not all of which are consis tent with maximum payoffs in terms of employment and in come gains for the CETA client population. Of course, good management can be assisted to become better, and staff development can be institutionalized to speed and improve its effectiveness. Labor Department regional training centers have been useful occasionally in providing logistical support for recordkeeping and com pliance with (frequently unnecessary) regulations, but they have not been conducive to sound management, planning, and policymaking. Universities, on their own initiative or in response to institutional grants, have put together degree granting programs for preparing entrants to the field, but opportunities during recent years have been too plentiful in human resource activities, especially in managerial roles, for large numbers to enter the CETA system. Even where they 83 did, preentry preparation can never obviate the need for onsite staff development. The National Association of Counties' manpower unit has played a significant leadership and staff development role, as have the national mayoral and gubernatorial associations to a lesser degree. Attachment to these local and state associa tions has impeded the emergence of a CETA-wide profes sional association, which could set professional standards and promote their attainment. There is badly needed a mechanism through which experience can be shared, technical assistance can be provided, and staff can be developed by the only ones who know how—those who have been through the mill and have learned their lessons. Experimentation is currently underway of a prime sponsor-to-prime sponsor technical assistance and training system. This approach has promise for training prime spon sor managers. The rising stars, trained by effective managers, are the most promising source for leadership in lagging prime sponsorships. The process already works to some degree. Efforts to institutionalize such relationships should be encouraged. The federal officialdom needs training no less than the prime sponsor staffs. The feds are caught between Congress and the locals without the competence to respond adequately to either. There is need for a more effective focus for joining key prime sponsor staff with the federal executive staff in setting overall national directions consistent with local challenges. Only a national/local consensus on mission goals and objectives can serve as a lasting framework for local decisionmaking and as a basis for assessing prime sponsor performance. The compliance issues could be presented to a CETA-wide professional leadership group who could design means of accomplishing the goals without interference with performance. The Employment and Training Administra- 84 tion sponsors a CETA director's work group which is con vened as a sounding board. Orchestrated by the feds, the local administrators are invited to listen to what is going to happen, but are not invited to suggest alternatives. Congress, too, could be more effectively educated and influenced by such an organized body. There is need to work upon and develop the inherent common professional interests that ex ist between national and local staff, substituting it for the adversary relationship which has emerged during the CETA years. Curriculum Development Sorely missed on the CETA scene is the federally con tracted but privately operated technical assistance and staff training system, which once developed and disseminated cur ricula throughout the MDTA system and trained local staff. Times have changed, and different arrangements are needed to take the place of this defunct institution. A possibility to be explored is a computer-assisted and computer-managed instruction being successfully promoted in general education and in Job Corps centers. Terminals linked to national or regional sources could offer access to common curricula in remedial education, English as a second language, and voca tional instruction related to "hands-on" training. But that would meet only part of the need. Person-to-person relation ships are also essential. Whatever approach proves best, there is a crying need for a positive technical-assistance-andguidance approach to replace the adversary relationship, which has grown between prime sponsor and federal agen cies. Such a system need not reduce local autonomy and discre tion. Unlike regulations handed down from above, technical assistance is neither arbitrary nor compulsory. Prime spon sors remain free to accept or reject. But given the hunger for 85 leadership and direction, most prime sponsors would accept help without considering it a threat. Management Information Related is the need for a common, computer-linked management information system. Prime sponsors could feed in day-to-day operational data allowing constant monitoring of the system's inputs and outputs without the burden of useless reports. Performance standards and records of in dividual progress could both be incorporated. The local operators could call up national comparative data to test their own performance, as well as storing their own informa tion for future recall. The same facilities could serve the cur riculum and management functions at lower long-run costs than the current inadequate information flows. Structure and Planning Judging from the 12 cases, prime sponsors cannot respond solely to the needs of either the local labor markets or the CETA clientele. Rather, there must be a delicate balancing of the political needs of the chief elected official, the direc tives of the federal funding source, the needs and desires of the eligible population, the demands of the organizations that represent (or claim to represent) the clients, the marketing pressures from the alternative service deliverers, the limited cooperation available from employers, and the personal predilections of the prime sponsor staff. Judged, however, against the complexities of the environment, the fact that 12 prime sponsors could respond with acts that do accommodate the diverse, and frequently conflicting, in terests and still achieve benefits that exceed the costs, is little short of remarkable. But that does not deny that there are a 86 number of potential improvements suggested by the data, which can provide promise of improved performance without denial of any legitimate interest. The key problem remains that of setting priorities among competing objectives in an uncertain environment with numerous special interest groups demanding attention. Only strong, well-trained, and secure management can meet this challenge. Elected officials seem to play a limited role, but since the alternative within a decentralized system is a rigid and arbitrary bureaucratization, there appears to be no bet ter alternative than the present delegation of authority. When given the opportunity, advisory councils can become active and positive forces. However, three separate advisory councils are too much of a good thing and there is a need for consolidation. More important is the need to organize prime sponsors to cover complete labor markets rather than on the basis of a fragmented political jurisdiction. The consortium incentives need strengthening until they outweigh the political temptations to fragmentation. The state level organization is inherently troublesome. While state participation in policymaking makes sense, most states are too diverse for the planning and delivery of ser vices to local labor markets. The balance of state is a residual concept with no rationale to support it and every experience to cause doubt about its effectiveness. An alternative ap proach might be to assign states the responsibility for helping political subdivisions to deliver CETA services on a labormarket basis. This would require subdividing states into areas surrounding major concentrations of population for program delivery, but maintaining a state role in coordina tion among these units. No simple block grant can encom pass the complex relationships between local and state governments. 87 A Two-Tiered Training System There are and will probably always be more eligible disadvantaged workers willing to be trained in a stipended pro gram for low-level entry occupations than there will ever be funding to support. There will also be a continuing flow of those who, at low per capita cost, can be made more employable by remedial education, English as a second language, and job search training. Many of those may not have the capability, endurance or resources to undergo more extensive training for high-level occupations. Yet there is evidence that there are many within the currently eligible population who could and would profit from the more ex tensive training, and at higher social benefit-cost returns. Involved are two sets of institutions, or at least two sets of institutional services. Some eligible enrollees are capable of undertaking occupational training. Others need a remedial stage to prepare for the advanced training. These latter, as well as those only capable of entry-level work, need the variety of supportive services described earlier, which are not readily available in mainstream training institutions. There should be a two-tiered system. The system would provide short-term remedial and entry-level training accom panied by supportive services available to all who need them. This entry/remedial stage could be affected either by specialized or mainstream institutions. The equivalent of 1 academic year, 36 weeks, should be generally sufficient for this stage. The second-tier of the system would offer the op portunity for extended career training to those initially ready for it or who successfully complete the entry level. It would be offered in mainstream institutions and would require only an additional year's tuition to implement the 2 years now legal, but never provided. 88 Closely related is training allowance reform. Training allowances at the minimum wage level distort the incentives of training participants. Who is enrolled for the training and who for the stipend? Training allowance might be divided in to three components: (1) reimbursement of training-related expenses; (2) a subsistence stipend based on family income; and (3) a motivational component based on performance. On-the-Job Training OJT is the most effective of all for those who have access to it. Increasing that access depends upon attracting more employers by reducing their reluctance to hire CETA-eligible applicants. Aggressive marketing that places the enrollee on a "tryout" basis should be explored to encourage employers to hire disadvantaged applicants. During this period of 3 to 6 months, the prime sponsor would provide a stipend to the employer in lieu of wages. Future Directions The CETA training activity, in general, seems to be an ef fective amalgam of the MDTA inheritance, the local voca tional training system, the inputs of community-based organizations, and the coordination and direction provided by local and state prime sponsors. It is strong in some places and weak in others, but there is no reason to think any cen trally derived pattern could work more effectively than the local designs. It is not the whip but guidance that prime sponsors need. There is no indication of reluctance to accept knowledgeable and positive direction. Technical assistance from successful prime sponsors could strengthen the locally weak systems. Some staff development and technical assistance can be centrally provided, but much will have to be brought to the local scene. Centrally developed curricula could be locally adapted without reducing local discretion. 89 Centralized information does not necessarily threaten local initiative. Specific problems can only be solved at the local labor market level. The temptation for radical surgery as part of CETA's reauthorization in 1982 should be resisted. One luxury CETA has never enjoyed is stability. There is no need to change the basic structure of the system, though there is need to delineate the roles and relationships of the players and the parameters of their managerial responsibilities. That does not preclude changing the service mix or eligibility criteria, but the basic delivery system relationship should remain in place long enough for capacity building to take place in a relatively orderly environment. As a national policy, there needs to be more thought to a human capital development approach to CETA training. Concern for productivity, energy, and "^industrialization" are refocusing attention on developing and upgrading the work force. The rapid fall-off in the number of youth enter ing the labor force during the 1980s will make each new en trant that much more valuable. If they turn out to be sup plemented by a continued immigrant flow, the latter, too, will require an increased investment. There are other systems for other components of the human resource pool. CETA was designed to aid the disadvantaged, but overall national objectives are best ac complished when the disadvantaged enter the mainstream. The prime sponsors in the 12 case studies experimented with a number of approaches, including a hard-nosed selection process that relates the abilities of the potential enrollees to the requirements for training in particular occupations, and a lengthy sequencing of remediation skill training and lowor high-support on-the-job training. Each of these ap proaches appeared promising depending on local situations. Prime sponsors or their successors should be encouraged to 90 continue with their efforts since ultimately programs will have to be developed for upgrading the labor force in their communities. The federally funded GETA system or its replacement can play a major role in achieving this goal. However, since the study was completed, the federal government has opted for cutting fiscal 1982 CETA training funds by more than a third, while totally eliminating PSE as a training option. The record of the training in the 12 prime sponsorships, as well as the national evidence of participant income gains, does not justify the cut. On the contrary, when the national productivity growth has almost ground to a halt, a program that returns $1.14 on every dollar invested in institutional training and several times as much on OJT is a wise investment worth preserving and nourishing. We should have learned from 20 years of employment and training experience that institution building is a slow and painful process. In many ways, demographic and economic developments are recreating the issues out of which MDTA emerged in 1962. Technological displacement and plant clos ings were major motivations for the legislation, which was directed toward "retraining." All of today's robots and computers were on the horizon, but the entry of the babyboom generation and their mothers into the labor force made available an ample supply of low-wage jobseekers, en couraging more labor-intensive processes. Now the economic and demographic conditions portend a swingback to a more capital-intensive level. Intensified international competition and technological advances again threaten increased plant closings and the need for worker relocation. All the signs are for continued influx of immigrants. Geographical barriers will become more, not less, serious for central city residents, even as their numbers decline. Youth entering the labor force, though in shrinking numbers, will require more, not less, training. At 91 the other end of the age spectrum, delayed retirement will be more likely to involve second careers and retraining than continuation of the old jobs. The displaced homemaker has not disappeared from the scene. Clearly, the demographic portents of the 1980s are for in creased training needs if we are to revive the growth in pro ductivity and prepare the labor force for the inevitable changing technology in the years ahead. Disbanding a system which took 20 years to build, inadequate as it still is, is likely to prove in a few years to have been shortsighted and costly.