Job Training: The Impact on California of Further Consolidation and Devolution
Robert T. Reville and Jacob Alex Klerman
Since federal involvement in job training began in the early 1960s, great expectations have been attached to job training as a policy that can reduce poverty and unemployment at minimal cost to the government. It has been a policy that has failed to reach its anticipated potential. This paper attempts to assess the impact of pending legislation on job training in California. Can the increased discretion provided under the proposed legislation lead to more effective job training, so that job training becomes a significant policy to alleviate poverty and unemployment?
Legislation to overhaul the job training system passed both houses of Congress this fall and is currently in conference committee. The stated purpose of the legislation that passed the House of Representatives is to transform the current system of "fragmented and duplicative categorical programs" into a "Federal workforce development and literacy system that is designed to meet the education, employment, and training needs of the workforce and the competitiveness needs of employers of the United States, both today and in the future." (HR 1617, title I, section 3b). Both bills provide block grants for training, consolidate vocational education and job training programs, and transfer most authority for training policy and administration to local and state governments. Therefore, the bill reflects the federal vision that meeting the education, employment, and training needs of the workforce is largely a local responsibility, though with federal financial support. The financial support will be substantially reduced in both of the bills.
Since 1973, the federal job training system has been transformed by continual consolidation and devolution. The result is a system that is locally designed and administered. Therefore, the first question is whether further decentralization will have any impact. To answer this, the next section reviews the history of job training; the following section describes the decentralized structure of the current job training system. Despite previous reform legislation in 1973 and 1982, the current job training system has been heavily criticized, which has led to calls for further reform. The third section reviews the critique of the General Accounting Office (GAO) and its suggestions for further consolidation. The fourth section reviews the critique from the job training program-evaluation literature. This critique has not been directly influential in shaping legislation, but it will help inform California in its reaction to proposed legislation. The fifth section describes the bills for job training reform now in conference committee. The final section concludes with predictions on the effect of the legislation on job training and presents recommendations.
A History of the Devolution of Job Training
Federal job training assistance is targeted primarily at four groups: disadvantaged adults voluntarily seeking employment services, disadvantaged youth, dislocated workers, and welfare recipients. These groups are not mutually exclusive, which has led to a tension in job training policy between ending the provision of duplicative services and focusing on the unique needs of each group. Since the goal of job training policy is employment, ideally at earnings above poverty level, another policy tension arises from the roles of the federal and local government. While the former provides the funding, the latter is arguably more sensitive to the skill demands of employers. Therefore, the history of federal job training programs has been a history of consolidation and devolution. In this section, the history of job training is reviewed to provide context for the current proposed changes.
The first major training program was the Manpower Development and Training Act (MDTA) in 1962. The MDTA was intended to retrain workers displaced by technological change and to improve the earnings capacity of disadvantaged workers. It was a federally funded and administered training program. The Economic Opportunity Act of 1964 initiated another major job training program: the Job Corps. The Job Corps offers training and counseling in a residential environment to highly disadvantaged youth. A third major job training program arose in reaction to the increase in the Aid to Families with Dependent Children (AFDC) caseload over the 1960s. The federal government determined that training would allow AFDC recipients to find work and leave welfare. As a result, the Work Incentive Program (WIN) was established in 1967 to provide training to welfare recipients (O'Neill, 1973).
In addition to MDTA, Job Corps, and WIN, a number of other job training programs and public employment programs were created in the 1960s, including Neighborhood Youth Corps, Operation Mainstream, New Careers, Concentrated Employment Program, Older Americans, Model Cities, and Foster Grandparents (Bassi and Ashenfelter, 1986). In response to the increasing complexity of the job training system, Congress passed the Comprehensive Employment and Training Act (CETA) in December 1973. Title I of CETA consolidated the various programs and turned the administration and operation of the programs to local officials. It replaced the "categorical" system of funding the programs, whereby each program is targeted at a particular population and federal administrators determined the amount of money allocated to each, with "revenue sharing," whereby state and local governments would receive a manpower training grant, allocated by formula, which was to be spent at their discretion on various job training programs for the disadvantaged (O'Neill, 1973). This change was intended to rationalize the administration of workforce development and at the same time increase local control, allowing the programs to be responsive to local conditions.
The federal role in job training, dramatically reduced by CETA, was further reduced in 1982, when Title I of CETA was replaced by Title II of the Job Training Partnership Act (JTPA). While CETA delegated the responsibility of administration to local officials, local administration was subject to federal oversight in the form of regulations and plan approval as well as designation of the local entities to receive funds. States were largely left out of the process. JTPA placed primary responsibility for oversight with the states. The private sector, too, was not directly involved in the federal job training system under CETA. In 1978, Private Industry Councils (PICs) were established, but their role was purely advisory. JTPA included a direct role for PICs in the formulation of job training programs at the local level. A further criticism of the job training system under CETA was that it was not sufficiently coordinated with other related educational and social programs. JTPA made states responsible for promoting coordination, and funds were directly allocated to the governors for this purpose (Levitan and Gallo, 1988). Finally, JTPA reduced the federal expenditure on job training. In 1981, $4.4 billion was spent on job training for the disadvantaged. After reductions throughout the 1980s, this amount had fallen to $2.4 billion by 1994 (Lalonde, 1995).
A similar pattern of decentralization characterizes the history of welfare-to-work programs. While welfare programs themselves continue to have considerable federal oversight and administration, training programs established under the programs are designed and administered by states. In 1981, the Social Security Act was amended to permit WIN demonstration programs (WIN Demos). These programs gave states the option to assume responsibility for the administration and oversight of the WIN program for their state. By 1988, 29 states had established WIN Demos (Barnow and Aron, 1989). In 1988, Congress passed the Family Support Act, which replaced the WIN program with the Job Opportunities and Basic Skills (JOBS) program. JOBS provided federal matching funds for states to design their own welfare-to-work programs. Like JTPA, it specified service options and included performance standards, but allowed states the discretion to determine how to meet them (Gueron and Pauly, 1991). In 1985, the Food Security Act established the Food Stamps Employment and Training (E&T) Program, which funds training programs for recipients of food stamps. This program, too, provides states a great deal of flexibility in designing and administering the training services provided (Barnow and Aron, 1989).
The growth of welfare-to-work programs like JOBS and Food Stamps E&T and the expansion of programs for dislocated workers over the 1980s have led to renewed calls for consolidation of the job training system. However, compared with the system that existed prior to the passage of CETA, twenty-five years of federal job training reform has created a system in which federal administration and oversight are virtually nonexistent, local infrastructure for designing and administering the programs is already in place, the funding for these programs is largely in block grants allocated by formula to the states, and private sector participation is substantial.
The Current Job Training System
The largest federal job training program is the Job Training Partnership Act (JTPA). The JTPA is a highly decentralized block grant program administered by local communities with minimal federal regulation and oversight. There are a number of other programs with overlapping clientele and similar purposes, but there is also an extensive set of incentives for coordination among programs.
The JTPA is the primary source of job training funds for California. Total federal funding received by California for 1994-1995 was $542 million (Bugarin, 1995). Title II of the JTPA provides training services to disadvantaged workers, and Title III, referred to as the Economic Dislocation and Worker Adjustment Assistance (EDWAA) Act since being amended in 1988, provides training to dislocated workers. The "partnership" envisioned is between state and local governments, and the local business community.
Under the JTPA, training block grants are allocated to states by a formula based on the unemployment rate and the number of disadvantaged individuals in the state. Within each state, the governor has designated Service Delivery Areas (SDAs) to which 78 percent of the funds are allocated by the same formula (Barnow and Aron, 1989). In California, there are 52 Service Delivery Areas. (Bugarin, 1995). Within each SDA, the chief elected official will appoint the members of the Private Industry Council. At least 51 percent of the PIC must be from the business community. The PICs, in combination with local elected officials, are responsible for formulating the job training plan for the SDA, which includes procedures for selecting training providers and recipients, and details of the administration of the program (Barnow and Aron, 1989). The active involvement of the business community, the ultimate provider of employment for trainees, is intended to ensure that the training provided is of sufficient quality and relevant content.
In response to criticism of CETA that performance was measured by the number served rather than the quality of the service, JTPA included performance standards and incentives to meet them. Six percent of JTPA funds are allocated to provide incentives for SDAs to meet a set of performance standards designed by the Secretary of Labor. The governor can modify and extend the standards. SDAs that exceed the standard can receive a portion of the incentive funds. Those that do not meet the standards can receive technical assistance from the 6-percent set-aside, but if they do not meet the standards for two consecutive years, the SDA can be dissolved and reorganized. Standards include the percentage placed in employment and the wage received (Barnow and Aron, 1989).
Training is available to dislocated workers both through JTPA Title III and through the Trade Adjustment Assistance (TAA) program. The TAA program provides assistance to workers displaced by foreign trade. Because of the overlap among Title II, Title III, and the TAA programs, legislation since 1982 has attempted to increase coordination among them. In 1982, JTPA Title III did not require states to pass funds down to local areas, and as a result Title III programs were not administered by SDAs. In 1988, the Economic Dislocation and Worker Adjustment Assistance (EDWAA) Act amended JTPA Title III to, among other things, require states to pass down at least 60 percent of the funds to local areas. This increased coordination among Title II and Title III training programs. At the same time, the Trade Act of 1974, which authorizes TAA, was amended to require training for workers who receive TAA cash assistance, and to encourage extensive coordination between JTPA and TAA (National Commission for Employment Policy, 1991).
JTPA also directed the states to coordinate job training with vocational education, the Employment Service, and welfare. This was intended to avoid duplication of administrative functions, as well as facilitate ease of referral among agencies. Eight percent of JTPA funds were set aside for "coordination grants." In addition, the governors were directed to create State Job Training Coordinating Councils (SJTCCs) composed of local government, state government, and the business community, as well as other groups such as organized labor and community organizations. The SJTCC has the responsibility of preparing a plan for coordinating the various educational and social welfare agencies of the state, and of reviewing the SDA plans for consistency with the state plan (Levitan and Gallo, 1988).
Coordination with the Employment Service avoids duplication with the job placement function of the JTPA. The Employment Service was created in 1933 under the Wagner-Peyser Act. It provides job search assistance from listings of job openings to resume-writing workshops. It is primarily funded from the Unemployment Insurance Tax. Prior to 1982, the Employment Service was largely a federal program, with the Department of Labor determining office locations and staff size. In 1982, the Wagner-Peyser Act was amended to give states the authority to administer the Employment Service, to encourage greater responsiveness to local conditions. After the 1982 amendments, the Department of Labor provided technical assistance and program oversight (GAO, 1991). The 1982 Wagner-Peyser Amendments allow 10 percent of Employment Service funds to be used by states for coordination with JTPA (Levitan and Gallo, 1988). The 1988 amendments to the Trade Act of 1974 also encourage coordination between TAA and the Employment Service (National Commission for Employment Policy).
In 1984, Congress passed the Carl D. Perkins Vocational Education Act, continuing federal support for vocational education that began in 1917 with the Smith-Hughes Act. Separate federal support for job training and vocational education arose from the belief that the vocational educational system had failed in training the disadvantaged and in retraining adults. The Perkins Act, however, emphasizes increasing vocational training for populations, such as adults and the disadvantaged. Therefore the potential for efficiency gains from coordination and costs from duplication became greater than ever (Roberts and Petrossian, 1987). The 8 percent of JTPA funds set aside for coordination grants is intended to provide encouragement for states to realize these efficiency gains. In addition, both JTPA and Perkins require that state and local plans discuss methods of coordination between vocational education and job training (Grubb, et al., 1990).
The vocational education system is largely financed by state and local governments, and federal support is a small percentage of the total expenditure. Therefore, like JTPA and the Employment Service, Perkins funds are allocated in block grants. The block grant formula is based upon the age distribution of the state, and the state's relative median per-capita income (with poorer states receiving a larger grant). The funds are then distributed to public secondary and postsecondary schools by a different formula based upon the school's share of the state's total enrollment of disadvantaged and handicapped students (Roberts and Petrossian, 1987).
Gains to coordination have also been perceived in welfare-to-work programs. The Family Support Act, which established the Job Opportunities and Basic Skills program, required state plans for JOBS to be coordinated with JTPA. States are allowed to use the 8 percent education coordination grant to fund JOBS. In California, half of the 8 percent is used to fund the JOBS program GAIN (Greater Avenues for Independence). In practice, GAIN participants use JTPA programs extensively (Grubb et al., 1990).
In 1992, Congress amended the JTPA to encourage states to form Human Resource Investment Councils (HRICs) that would unify in one council the various administrative and advisory bodies involved in employment, training, and education. The HRIC would be involved in formulating a comprehensive human capital investment policy, and then would monitor the implementation of that policy. An HRIC has not been formed in California (Bugarin, 1995).
The most recent federal innovation to encourage increased coordination in education, employment, and job training services is the development of the concept of "one-stop career centers." In late 1994, the Department of Labor began to award grants to states to develop "centers" to unify access to job training, employment, and education for all population groups. The concept envisions recipients of training as "customers" and emphasizes accountability of service providers to maximize "customer satisfaction." Development of a center, according to the Department of Labor, requires integration of the administration of the various employment-related programs at both the state and local level (U.S. Department of Labor, 1996a).
The Critique, Part I: The GAO
Criticism of the current job training system has many dimensions. In this section and the next, criticism of the system from two sources will be reviewed. The first critique, from the General Accounting Office, has been influential in shaping legislation for reform of the system. The second critique, from the academic literature and the research organizations responsible for evaluating training programs, has not directly influenced legislation, but should be considered by California and other states when deciding how to exercise the increased discretion that block grants are expected to provide.
In a series of reports in the last few years, the General Accounting Office has reviewed the current job training system and found that it is fragmented and inefficient. In 1995, it identified 163 separate programs administered by different agencies, often offering identical services to overlapping clientele. With the current popularity of welfare-to-work programs and job training programs, this number has been increasing rapidly. In 1991, the GAO identified 125 programs. By 1994, that number had increased to 154 (U.S. GAO, 1995a). According to Bugarin (1995), California has 22 labor and employment programs located in 13 administering departments.
The GAO has argued that to administer these programs, parallel structures have been created in different departments at both the federal and state levels. Often, separate programs provide services to the same population. Bugarin (1995) found a comparable situation in California. For instance, 12 programs in California target welfare recipients. This duplication increases administrative cost, reducing the amount of money available for training workers (U.S. GAO, 1995a).
Furthermore, the system, according to the GAO, is confusing to clients, discouraging individuals seeking assistance. Since it is not linked to economic development programs (local programs that seek to attract or retain businesses), it is difficult for employers to meet their employment needs. Administrators, too, face conflicting regulations and different planning cycles. Finally, according to the GAO, most programs do not have accurate information on the effectiveness of their programs (U.S. GAO, 1995a).
The GAO made four recommendations:
- The number of programs needs to be reduced, and the ease of access to the remaining programs needs to be increased.
- Conflicting federal requirements and duplicative administrative procedures need to be eliminated.
- A wide array of employment training services needs to be offered, with the programs developed in partnership with employers to ensure relevance to labor market needs.
- Program administrators need to be held accountable, while at the same time states and local agencies need to have the flexibility to design the program to be responsive to local conditions, which requires "clearly defined goals and performance standards."
The GAO critique is familiar. Congress has been trying to implement the four recommendations above since 1973. It is not clear whether they have not yet gone far enough, have reversed themselves in the intervening years, or have failed to diagnose the problem correctly.
The Critique, Part II: The Evaluation Literature
Another criticism of the job training system emerges from a review of the training program evaluation literature. This critique is arguably more fundamental, because the source of the problem is not the administration of the system, but the programs themselves. According to this critique, the modest, short-term job training programs offered by the federal government are largely ineffective. For some demographic groups, they produce small gains in employment and yearly earnings (but almost never in wages--amount earned per hour), while for others they have no effect.
A review of which programs are effective and what groups should be served will assist California in its priorities for funding job training programs. According to a GAO study of the experience with block grants in the Omnibus Budget Reconciliation Act (OBRA) of 1981, two important lessons are that funding priorities tend to change when states have more discretion, and the need to offset reductions in federal funding tends to lead to increased state expenditure on the block-granted programs. California's imprint on programs ought to reflect what is known about job training from these evaluation studies.
Nonexperimental evaluations of the impact of early job training programs were not very reliable. In response, a large number of experimental evaluations of job training programs have been conducted since the 1980s. An experimental evaluation includes a control group which does not receive the programs provided to the experimental group, and the outcomes for the two groups are compared. These experiments have provided a large amount of information about the effectiveness of various training strategies, and about the returns to training for various demographic groups. Several recent studies (U.S. Department of Labor, 1995; Lalonde, 1995; Grubb, 1995) have summarized the literature on job training evaluations, and concluded the following:
- Job training programs, by themselves, cannot eliminate poverty. Even the most-effective programs are not able to raise the participants out of poverty. However, certain programs have been able to increase the earnings, and reduce the welfare participation, of particular groups.
- Modest programs for disadvantaged youth are generally the least effective. Short-term training, classroom training, and summer youth programs have generally been shown to have no long- term benefits. The participants in the programs do not have higher rates of school completion, employment, or wages than the control groups. Programs targeted at young single mothers have not significantly reduced their welfare receipt or poverty. On the other hand, intensive (and expensive) programs such as Job Corps have been effective.
- The most promising results have been achieved for disadvantaged adult women. Earnings gains from both voluntary programs (like JTPA) and welfare-to-work programs (like JOBS), while still leaving the participants in poverty, have been shown to last for several years. For example, the difference between the experimental and control group in annual earnings in the second year after participation in JTPA is $900 for women, a difference of 15 percent (U.S. Department of Labor, 1995). The evidence from welfare-to-work programs for adult women is that they can reduce welfare participation rates, and they are cost-effective insofar as the reduction in welfare payments outweighs the cost of the programs (see Gueron and Pauly, 1991). For example, AFDC payments to experimental group members in California's GAIN program were an average of $350 per year lower than the control group in the three years after program participation (U.S. Department of Labor, 1995).
- The evidence is not as conclusive for other groups. In particular, disadvantaged adult males have been helped by some programs, including JTPA (though to a lesser extent than women), and not by others. Training for displaced workers has also been shown to have mixed results. However, results for both of these groups, in particular displaced workers, are tentative because of a relative scarcity of studies.
It is difficult to summarize the evidence with regard to the effectiveness of various services provided by job training programs, and the conclusion of Grubb (1995) and U.S. Department of Labor (1995) is that offering a wide variety of services is the best strategy. Job search assistance, the least expensive program, has been able to increase earnings for all groups by reducing the duration of unemployment. However, it tends to raise only earnings and not wages, and the effect does not last. Short-term classroom training of basic skills is generally ineffective for all groups.
The modest successes for displaced workers may even be less significant than reported, because the benefit of training programs to the worker (and to society) ought to be counted against the value of time spent out of the workforce and in training. Displaced workers tend to be of prime working age and the value of this time is not insignificant, even if the wages they would receive are lower than what they received on their previous jobs. Unlike with disadvantaged youth, there are unlikely to be benefits from reduced crime, or, unlike with disadvantaged women, from reduced welfare participation (Lalonde, 1995).
Despite this somewhat bleak picture of the success of job training programs, it should be noted that certain programs have been extremely successful, and the benefits to replicating these programs may be very high indeed. Two examples of successful programs are located in California. The Center for Employment and Training (CET) in San Jose has been successful in training two groups that have seen poor results elsewhere: young female single parents and young high school dropouts. For example, CET was evaluated as part of a program called Jobstart for disadvantaged youth. The program at most other locations was unsuccessful, but at the CET program participants had 40 percent higher earnings than the control group, an annual earnings increase of $3000 (U.S. Department of Labor, 1995). The success of this program has been attributed to (among other things) the close connections between the center and local businesses, and to an innovative training program that integrates basic skills with vocational training. Another successful program is the Riverside County welfare-to-work program (part of the Greater Avenues for Independence (GAIN) demonstration). At this site, an emphasis on job search and job placement has proven very effective. The Riverside GAIN participants had annual earnings $1000 higher than the control group (U.S. Department of Labor, 1995). Grubb (1995) argues that the success of the CET and the Riverside GAIN program derives more from the fact that both provide comprehensive and interrelated employment services than from the quality of any particular services offered by them. The CET, in particular, with a broad array of training programs, is a prototype "one-stop shop."
While the GAO critique has figured prominently in the plans for federal job training reform, the difficult questions posed by the evaluation literature critique have been delegated to the states. In designing a job training system for California, the evaluation literature has shown that inexpensive short-term job training programs may lead to modest increases in earnings and employment, but they do not reduce poverty on their own. For some demographic groups, they do not even increase earnings. Given reduced federal funding for job training, California should consider targeting job training on groups for which the training is effective. Training for disadvantaged adults, and particularly women, is the most effective. Resources for training youth need to be directed toward expanding intensive job training programs like Job Corps, and toward experimenting with innovative programs like those offered by CET. Continued experimentation, carefully evaluated, should be funded by California as well, since if training is to be a useful poverty alleviation measure, new methods need to be developed.
Careers and the Workforce Development Act
Federal job training support is about to enter a new era. On October 11, 1995, the Senate passed by a vote of 97-2 the Workforce Development Act, which provides block grants for job training, adult education, and vocational education. A month earlier, the CAREERS (Consolidated and Reformed Education, Employment and Rehabilitation Systems) Act passed the House of Representatives by a vote of 345-79. The two bills are currently in a House-Senate conference committee where differences are being resolved (Havemann, 1995).
The House CAREERS Act consolidates JTPA and Perkins, as well as the Adult Education Act and the School-to-Work Act, into three block grants. The first block grant consolidates youth programs into the Youth Development and Career Preparation Consolidation Grant ($2.325 billion for 1997). The second consolidates adult programs into the Adult Employment and Training Consolidation Grant ($2.183 billion). The third is the Adult Education and Family Literacy Consolidation Grant ($280 million). States would be permitted to transfer up to 10 percent between the youth training grant and the adult training grant.
The Senate Workforce Development Act would also consolidate vocational education and job training, using one block grant instead of three. The block grant would be allocated 25 percent for workforce employment activities, 25 percent for workforce education activities, and 50 percent for a "flex account" that is allocated to employment or education by the governor. The flex account money can also be spent on welfare-to-work job training programs.
Several provisions are common to both bills, and are therefore likely to be retained in the version that emerges from the conference committee. Both bills
- allow states considerable flexibility in the allocation of funds between vocational education and job training for adults,
- eliminate eligibility on the basis of unemployed or disadvantaged status,
- emphasize accountability and the use of performance standards,
- require that states make information on available services easily accessible ("one-stop career systems") and maintain a labor market information service (Wagner-Peyser Employment Service),
- allow for the use of vouchers so that adults can choose among providers, and
- include provisions for the use of block grant money for economic development, such as firm retraining of incumbent workers.
There are also several key provisions included in one bill but not in the other. The Senate bill restricts training to individuals who have received a high school diploma or the GED, unless currently enrolled in a course to obtain one. The House bill restricts training to individuals who cannot find employment through job search assistance. The House bill requires states administer training through a system similar to JTPA's Private Industry Councils, but the Senate bill makes this system optional.
As noted above, in addition to seeking increased coordination with vocational education and the Employment Service, Congress has sought to increase coordination between JTPA and displaced worker programs, and between JTPA and welfare-to-work programs. However, the JOBS program and the Food Stamps Employment and Training program were not included in either of the block grants. Both bills retain the Job Corps as a federally administered program and exclude the Trade Adjustment Assistance program from the block grants.
The Implications for Job Training in California
CETA consolidated federally administered programs into a block grant and delegated authority to design and administer those programs to local areas. CAREERS and the Workforce Development Act consolidate several block grants into larger block grants. The programs are locally administered and designed, and there is an extensive set of mechanisms and incentives in place to encourage coordination among them. CETA did not revolutionize job training. It is hard to imagine that the new legislation will either. In this section, some of the potential implications of the workforce legislation for California will be reviewed.
According to the GAO, there are numerous potential gains from block grants (GAO, 1995b). They include
- savings through the elimination of duplicative administrative activities,
- promotion of interdepartmental coordination at both the federal and state level,
- decentralization of decisionmaking, so that problems are identified and solved at the source,
- encouragement of innovation, and
- better targeting of resources, since distributing funds by formula reduces the power of federal administrators to control grants.
The first and second gains are related to consolidation. If the new legislation leads to consolidation of job training and vocational education in California, these gains may be realized. However, there is reason to expect that this will be difficult, given the incentives already in place for consolidation. Unlike consolidating two job training programs for the disadvantaged that are administered by two different agencies, job training and vocational education often have different clientele and offer different services (for example, on-the-job training by JTPA) (Grubb et al., 1990). The third and fourth gains are related to decentralization. Since job training and vocational education are already highly decentralized, it does not seem likely that the new legislation will lead to further gains in these areas. Finally, the fifth gain is related to administering a block grant program, but since both JTPA and Perkins are block grants, this gain has already been realized.
Since the block grant and federalism provisions of the job training legislation do not hold the potential to create a job training system significantly different from the current one, it is likely that the criticism in the evaluation literature will still be relevant. There are, however, several provisions in the current legislation that could potentially lead to a very different job training system--one that might be significantly improved. These are the provision of job training through one-stop shops, the potential for increased statewide coordination of job training programs, the ability to combine job training with economic development, and the potential for innovation and efficiency through the use of vouchers.
Both versions of the legislation emphasize service delivery through one-stop shops. The conclusion of the evaluation literature is that access to a wide variety of services, as in one-stop shops like the center for Employment and Training (CET), is the most effective way to provide training. Unfortunately, there are two potential problems with the one-stop shop provisions that may outweigh the benefits associated with creating them. The first is that the legislation may not provide enough money to implement one-stop shops effectively. The second is that to eliminate administrative barriers to one-stop shops, both versions of the legislation eliminated criteria for eligibility such as welfare status or low income, which may lead to a job training system that neglects the disadvantaged.
It is unclear whether the benefits from access to a wide variety of services require the services to be provided at a single location, or only that information on a wide variety of services be available. As envisioned by the Clinton Administration, one-stop centers with services provided at a particular physical location are likely to be expensive, and, if only a limited number can be constructed in a service delivery area, the need to travel considerable distances may actually reduce access to services for many individuals. Both the Senate and House bills have emphasized one-stop systems, rather than centers, with the implication that it is access to information about services, rather than the services themselves, that is provided at one stop. While this would certainly be cheaper, there is no evidence that it would lead to more effective job training.
One barrier to one-stop shops in the current job training system is the complex set of eligibility criteria, each created under separate legislation. Both of the bills in conference committee eliminate eligibility criteria, which presumably will reduce the cost of providing job training through one-stop shops. However, at current funding levels, only a very small fraction of the eligible populations can be served by existing job training programs, which are carefully targeted. Without targeting, the beneficiaries of one-stop systems with improved services may not be the targeted population. California should be careful to ensure that the disadvantaged continue to be served.
Increased Statewide Coordination of Job Training Programs
Given the size of California, there may be an advantage to reconsidering a fundamental assumption of block grants: the advantages of local control. At least in the Senate version of the legislation, California would have the option of constructing a system with more statewide coordination than the current 52 Service Delivery Areas (SDAs). The advantage of local control is that the training can be oriented toward the needs of local employers. This advantage may be even more significant in the future if job training programs are coordinated with economic development programs, as current legislation permits. However, since devolution to the local level has led to a reduction in statewide coordination, rethinking this devolution could lead to improvements in the job training system.
The simplest advantage of greater statewide coordination is that the current system leads to wasteful administrative difficulties. For example, a large employer in Long Beach had to spend several months negotiating an arrangement with the Long Beach and Riverside SDAs whereby the firm's employees living in Riverside could receive training near home. The problem was that it was not clear which SDA would be allowed to receive credit for the training. The job training system in California should, at a minimum, have sufficient coordination to ensure that individuals eligible for training in one SDA should be able to receive training in any SDA. A second advantage to increased statewide coordination is that a statewide system can provide a larger market for trainees and a larger range of occupations for which to train. For example, the national scope of Job Corps is considered one of its advantages, since trainees can choose from different training programs in centers across the country, and graduates are placed in a national market. A third advantage of increased statewide coordination is that in a local area with a troubled economy, it may be impractical to expect the local industry to hire the graduates of training programs while other parts of the state are in need of skilled workers. Finally, a statewide program might provide standardization in training for particular occupations. With standardization, the program is recognized in other areas, so that completion of a job training course can be a credential, and the skills can be portable.
Increased Coordination Between Economic Development and Job Training Programs
Given the evidence on the ineffectiveness of job training programs, particularly for dislocated workers and for disadvantaged youth, the best way to serve these groups may not necessarily be by providing job training courses. An alternative method may be available through the Senate provision that allows Flex Fund money to be spent on economic development job training. Use of Flex Fund money for programs to attract employers who provide jobs to the disadvantaged may be a better way to serve the disadvantaged.
California's Employment Training Panel (ETP) is an example of an economic development job training program that could potentially be expanded with Flex Fund dollars. ETP provides money to businesses to retrain workers while they are still employed. It may be that using flex fund money to expand ETP is a better way to assist dislocated workers than training them after they become unemployed.
Ventura County has initiated programs that combine economic development with job training by attempting to assess the needs of small businesses and placing disadvantaged individuals in employment in these businesses. These programs are based on the idea that most jobs are created by small businesses, but small businesses are not served by JTPA and other government job training programs (including ETP). A common criticism of federal job training assistance is that the jobs provided to trainees are simply displacing other job seekers, and that businesses are using government assistance to train workers they would have trained without assistance. Ventura County has argued that this criticism is more likely to be true for large businesses, and less likely to be true for small businesses who are underserved by the current federal job training system.
Whether programs like ETP and the small business development programs in Ventura County are more likely to reduce unemployment and poverty than the current job training system is unknown. In fact, it should be emphasized that these programs should be rigorously evaluated and not assumed to be better simply because they are different. However, if after evaluating these programs, they are found to be more effective than the current system, the pending job training legislation is likely to provide the ability to expand them using federal money.
The creation of a voucher system for job training is perhaps the most promising provision in the new legislation. Through the use of vouchers, individuals can choose the service provider that offers the training best suited to their abilities and goals. Since training is most effective when a wide array of services is available, this could be provided by the market through the use of vouchers. However, voucher programs for job training have not been rigorously evaluated. Therefore, this discussion is necessarily speculative and subject to the qualification that the program should be evaluated when implemented.
One objection to vouchers is that to be effective, the customers must have sufficient information to make a well-informed choice among service providers. Since job training is not likely to be a frequently purchased commodity, experience is not likely to provide the information needed. To address the information problem, the House bill requires that training providers report extensive performance-based information and that they be eligible to receive funds under the Higher Education Act or have met acceptable levels of performance set by the governor. Performance standards under JTPA have a somewhat different purpose--to ensure that SDAs provide high-quality training with federal funding. They have been criticized for leading to "creaming," or the selection of the most trainable instead of the most in need. The combination of vouchers and performance standards available at the service provider level may allow better targeting of training resources.
Another advantage of vouchers is that they may lead to innovative new training programs provided by the private sector. New ideas need to be developed if job training is going to be an effective poverty alleviation measure. Finally, vouchers may help resolve the tension between local, state, and national job training. If vouchers are portable across service delivery areas, individuals eligible for training in one SDA would be eligible in others. Since vouchers will be paid for with federal taxes, this could be true not only within California but across states as well. Furthermore, with vouchers, national private training firms could compete with local training firms to attract workers in need of training, and workers could be placed where their skills are most in need (and their wages the highest).
Barnow, Burt S., and Laudan Y. Aron, "Survey of Government-Provided Training Programs," Background Paper 9 in U.S. Department of Labor, Investing in People: A Strategy to Address America's Workforce Crisis, Washington D.C., September 1989.
Barnow, Burt S., and Jill Constantine, Using Performance Management to Encourage Services to Hard-to-Serve Individuals in JTPA, Research Report Series, National Commission for Employment Policy (DOL), April 1988.
Bassi, Laurie J., and Orley Ashenfelter, "The Effect of Direct Job Creation and Training Programs on Low-Skilled Workers," in Sheldon H. Danziger and Daniel H. Weinberg (eds.), Fighting Poverty, What Works and What Doesn't, Harvard University Press, Cambridge, Massachusetts, 1986.
Bugarin, Alicia, Restructuring Workforce Preparation Policy, California Research Bureau, California State Library, October 1995.
Grubb, W. Norton, Evaluating Job Training Programs in the United States: Evidence and Explanations, National Center for Research in Vocational Education, Technical Assistance Report, May 1995.
Grubb, W. Norton, Cynthia Brown, Phillip Kaufman, and John Lederer, Order Amidst Complexity: The Status of Coordination Among Vocational Education, Job Training Partnership Act, and Welfare-to-Work Programs, National Center for Research in Vocational Education, Berkeley, 1990.
Gueron, Judith M., and Edward Pauly, with Cameron M. Lougy, From Welfare to Work, Russell Sage Foundation, New York, 1991.
Havemann, Judith, "Senate Votes to Shift Job Training to States," Washington Post, October 12, 1995.
Heckman, James J., "Is Job Training Oversold?" Public Interest, Spring 1994.
Lalonde, Robert J., "The Promise of Public Sector-Sponsored Training Programs," Journal of Economic Perspectives, Vol. 9, No. 2, Spring 1995.
Levitan, Sar A., and Frank Gallo, A Second Chance: Training for Jobs, W. E. Upjohn Institute for Employment Research, Kalamazoo, Michigan, 1988.
Lynch, Lisa (ed.), Training and the Private Sector: International Comparisons, University of Chicago Press, Chicago, 1994.
Mirengoff, William, and Lester Rindler, CETA: Manpower Programs Under Local Control, National Academy of Sciences, Washington, D.C., 1978.
Moffitt, Robert A., "The Effect of Employment and Training Programs on Entry and Exit from the Welfare Caseload," Journal of Policy Analysis and Management, Vol. 15, No. 1, pp. 32-50.
Moffitt, Robert A., "Incentive Effects of the U.S. Welfare System: A Review," Journal of Economic Literature, Vol. 30, March 1992, pp. 1-61.
National Commission for Employment Policy, Assisting Dislocated Workers: Alternatives to Layoffs and the Role of the Employment Service under the Economic Dislocation and Worker Adjustment Assistance Act, Special Report Number 30, Washington, D.C., October 1991.
O'Neill, Dave M., The Federal Government and Manpower: A Critical Look at the MDTA--Institutional and Job Corps Programs, American Enterprise Institute Evaluative Studies 9, Washington, D.C., August 1973.
O'Neill, Dave M., "Voucher Funding of Training Programs: Evidence from the GI Bill," Journal of Human Resources, Vol. XII, No. 4, Fall 1977.
Roberts, Robin, and Natica Petrossian, Navigating the Labyrinth: The Connection Between the Carl D. Perkins Vocational Education Act and the Job Training Partnership Act, Texas State Department of Community Affairs, Austin, August 1987.
Savner, Steve, Devolution, Workforce Development, and Welfare Reform, Center for Law and Social Policy, Washington, D.C., 1996a.
Savner, Steve, Summary of Senate Passed Workforce Development Act (S. 143), Center for Law and Social Policy, Washington, D.C., 1996b.
Savner, Steve, Summary of House Passed CAREERS Act (HR 1617), Center for Law and Social Policy, Washington, D.C., 1996c.
United States Department of Labor, "Economic Dislocation and Worker Adjustment Assistance Act (EDWAA)," Employment and Training Administration, Fact Sheet, 1996a.
United States Department of Labor, "Economic Dislocation and Worker Adjustment Assistance Act (EDWAA)," Employment and Training Administration, Fact Sheet, 1996b.
United States Department of Labor, What's Working (and What's Not): A Summary of Research on the Economic Impacts of Employment and Training Programs, Office of the Chief Economist, January 1995.
United States General Accounting Office, Employment Service: Improved Leadership Needed for Better Performance, GAO/HRD-91-88, 1991.
United States General Accounting Office, Major Overhaul Needed to Create a More Efficient Customer-Driven System, GAO/HEHS-95-70, 1995a.
United States General Accounting Office, Block Grants: Characteristics, Experience, and Lessons Learned, GAO/HEHS-95-74, 1995b.
United States House of Representatives, Committee on Ways and Means, Overview of Entitlement Programs, U.S. Government Printing Office, Washington, DC, 1995.
 We have benefited from discussions with Norton Grubb, Jim Hosek, Joe Hotz, Robert Moffitt, Bob Schoeni, and the participants in the RAND Fiscal Federalism Conference. Helpful comments were received from Barbara Williams, Cathy Stasz, and Ron Komers.
 The focus of this paper is on job training provided or subsidized by the federal government. See Lynch (1994) for a collection of papers on employer-provided job training.
 There are also job training programs for veterans, disabled persons, older workers, and inmates, although these programs are relatively small.
 The first federal program was part of the Area Redevelopment Act in the late 1950s (O'Neill, 1973).
 Job Corps continued under federal administration. The WIN program was not part of CETA.
 CETA is best known as a public employment program. In 1973, CETA included a modest public employment program that was intended to be implemented whenever unemployment rates were above a certain level. At the same time, the OPEC oil embargo struck, and unemployment rates increased from 5 percent in early 1974 to over 7 percent by the end of 1974. In response, President Ford agreed to a large increase in the public sector employment component, transforming CETA from a job training act to a job creation act. Under President Carter, the number employed by CETA was more than doubled. CETA's public sector jobs were eliminated in 1981 (Levitan and Gallo, 1988).
 The Job Corps continued as a federally administered job training program under Title IV of the JTPA.
 Another significant change from CETA to JTPA was that stipends paid to trainees were eliminated. The Reagan Administration argued that since a substantial portion of training funds were used for stipends, job training was just another form of welfare (Levitan and Gallo, 1988).
 According to U.S. House of Representatives, Committee on Ways and Means (1995), disadvantaged is defined as "being a member of a family whose total income for the 6-month period prior to application (exclusive of unemployment compensation, child support payments, and welfare payments) does not exceed the higher of the poverty line or 70 percent of the Bureau of Labor Statistics' lower living standard. Members of families receiving Aid to Families With Dependent Children (AFDC) or other cash welfare payments and those eligible for food stamps are also defined as economically disadvantaged."
 Dislocated is defined as "Workers who have lost their jobs and are unlikely to return to their previous industries or occupations." (U.S. Department of Labor, 1996b)
 As will be discussed further below, an additional 8 percent is allocated to state education and coordination grants, and 6 percent for incentive grants and technical assistance. The remaining 8 percent is allocated for training programs for older workers (3 percent), and administration (5 percent) (Barnow and Aron, 1989).
 The incentives provided by this system are complex. On the one hand, SDAs are encouraged to establish a training program that meets the needs of local employers so that the standards can be met. On the other hand, contrary to the JTPA's goal of providing training to the most needy, SDAs may attempt to meet the goals by selecting trainees who are easier to place in high quality jobs (Barnow and Constantine, 1988).
 A distinction is sometimes made between one-stop "centers" and one-stop "systems." The term "center" usually connotes services provided at a particular physical location, whereas a one-stop "system" refers to electronic access to information about available services.
 The "one-stop center" concept was foreshadowed by the "skill center" concept in job training policy in the 1960s. The services initially delivered by MDTA were classes in particular occupational skills taught at public vocational education schools. This approach was criticized because the occupational classes and times available to workers were limited. In response, the "skill center" concept was developed. A "skill center" is a single facility where a broad array of employment and training services are provided to all types of trainees on a continuous basis. In 1968, the MDTA was amended to promote the skill center as the preferred service delivery method, but the amendment was not implemented (O'Neill, 1973).
 California received a grant of $400,000 from the Department of Labor to develop a One-Stop Career Center System.
 The GAO has issued a number of studies. The studies are referenced and summarized in U.S. GAO, 1995a. That study also includes the recommendations for policy below.
 The number is smaller in California because the GAO study counts all programs operated under a particular law, such as the JTPA, separately, whereas Bugarin does not. JTPA alone, in the GAO study, contributes 20 programs. For instance, the GAO counts the 3 percent of JTPA funds set aside for older-worker grants, the 8 percent set aside for state education and coordination grants, and the 6 percent set aside for incentive grants as separate programs.
 For example, there are different definitions of "low income," or "household" depending upon the program involved, which leads to confusion and difficulties of coordination (GAO, 1995a).
 The extent of increase varied depending upon the block grant and the state. The overall increase in state expenditure on the programs in the Preventive Health and Health Services Block Grant was 9 percent. The increase in state expenditure on the Maternal and Child Health Block Grant was 24 percent. Some other block grants, however, saw no increase in state expenditure.
 See Bassi and Ashenfelter (1986) for a discussion of the early nonexperimental estimates and the need for experimental evaluation of job training programs.
 These evaluations are expensive, and the vast majority of job training programs have not been evaluated in this manner. The enormous value of the information from these evaluations for the formulation of policy suggests that one use of funds that should not be sacrificed under block grants is the rigorous experimental evaluation of training programs.
 Though job training is often suggested as a solution to the increasing inequality in earnings, Heckman (1994), using optimistic assumptions, notes that the amount that would be required to reverse these trends is prohibitive. For instance, to restore average earnings for high school dropouts to their 1979 levels would require spending $214 billion on job training.
 Some programs, such as summer employment, may be effective in reducing crime, but should not be viewed as investment in the human capital of the participants. See Heckman (1994).
 While Job Corps is generally viewed as a successful program, there is also some controversy. Lalonde (1995) discusses the difficulties in measuring the benefits attributed to the program. Young single mothers have not been shown to be helped by Job Corps (Department of Labor, 1995). Another intensive, though nonresidential youth training program called Quantum Opportunities Project has also been shown to be beneficial (Department of Labor, 1995).
 It is possible, however, that if welfare- to-work programs substantially increased the skills of welfare recipients, and that if they were available to a large fraction of the caseload, welfare participation would actually increase (Moffitt, 1996).
 The earnings increase for men participating in JTPA is $900, or 10 percent higher than the control group.
 For this reason, Lalonde (1995) notes that the earnings gains from job search assistance, more than any other service, may be at the expense of another disadvantaged worker not receiving assistance, who does not get the job because the assisted worker does. If this is the case, widespread implementation of job search assistance may be ineffective. The extent of this problem cannot be measured by experimental evaluations.
 According to Department of Labor (1995), "CET staff have extensive knowledge of local labor markets, which they use to determine which technical skills will be taught in their courses. In each new community CET enters, an industrial advisory board is set up to assist in skill selection and curricular review. CET courses are taught by experienced technicians from industry. Many area employers are also on the board." (p. 18)
 As noted by the Department of Labor (1995), even this successful program had only 23 percent of participants still employed and off AFDC three years after participation in the program.
 According to Bugarin (1995), the following are available: "JTPA, JOBS (GAIN), Food Stamps Employment Training, adult basic education, GED certification, alternative high school, skills training, labor market information, vocational assessment, foster care independent living skills, infant day care, and job placement services for employers." (p. 41)
 Concise summaries of the two bills are available from Savner (1996b, 1996c). These summaries, together with the text of the bills and documentation received from the Department of Labor, were used in writing this section.
 The total allocation to the state would include the amount allocated for the Employment Service under the Wagner-Peyser Act. The quarter of the block grant allocated to workforce employment activities would include the amount from the Wagner-Peyser Act, and Wagner-Peyser requirements would be paid for from this block grant. In eight states (though not in California), the current Wagner-Peyser allocation exceeds the amount allocated for workforce employment activities, leaving no money for training activities unless flex fund dollars are used.
 The House bill gives priority to the unemployed or disadvantaged but does not specifically target these groups.
 Vouchers are mandatory in the House bill, provided a sufficient number of certified providers are available. In the Senate bill, vouchers are optional.
 The Senate bill permits up to half of the worker employment grant be used for economic development, or 25 percent of the entire block grant. The House bill is more restrictive.
 Welfare-to-work programs can be funded through the Flex Fund provided in the Senate act.
 An example from California can illustrate the difficulties of coordination of job training and vocational education. According to Bugarin (1995), in March 1993, the State Council on Vocational Education and the State Job Training Coordination Committee attempted to jointly contract for a study of the coordination of job training and vocational education in California but were unable to agree to the terms of the contract (p. 35).
 Barbara Boxer, in a speech defending the Job Corps from consolidation in the Workforce Development Act, noted that "Even if California agrees to continue to operate these centers under a State program, the centers would still lose if the national program is eliminated. Job Corps trains students to get jobs in the national market, not just the region. Enrollees can choose centers across the country that best match their career plans." (Congressional Record, October 11, 1995)
 While innovative, ETP is regarded with hostility by many employers. It is considered overly bureaucratic and its regulations overly burdensome. Furthermore, ETP is considered to be poorly targeted, funding job training for firms that are not likely to lay off workers. An expansion of ETP should be coupled with an evaluation and reforms.
 The first federal job training program that contained elements of a voucher system was the "Individual Referral" program of MDTA. Under this program, an individual was permitted to choose a vocational program provided by the private sector, and the Employment Service would enter into a contract with the vocational program to provide the service. The Individual Referral program was not widely implemented. In addition, CETA directed the Department of Labor to develop and evaluate voucher programs for job training (O'Neill, 1977).
 O'Neill (1977) examined vocational-technical training taken by disadvantaged individuals eligible for the GI Bill, a program analogous to a voucher program, and compared their earnings outcomes with those of MDTA and CETA recipients. He concluded that the GI Bill vocational training led to significantly better results than the MDTA and CETA training programs. However, even disadvantaged GI Bill recipients are likely to differ from MDTA or CETA recipients, so the comparison may not be valid.