Grubb, W. N., Badway, N., Bell, D., Chi, B., King, C., Herr, J., Prince, H., Kazis, R., Hicks, L., & Taylor, J. C. (1999). Toward order from chaos: State efforts to reform workforce development systems (MDS-1249). Berkeley: National Center for Research in Vocational Education, University of California.

SECTION  1

Recent State Reforms:
Visions And Strategies

      Our ten states have taken different directions in reforming their workforce development systems. Fortunately, that variety is instructive about the efficacy of different approaches. In this section, we concentrate on what states have intended to do, leaving the implementation issues to Section II and the effects on local programs to Section III. We first review what caused states to initiate reforms. We then describe the reforms themselves and the mechanisms, or "instruments," that states have used to change their workforce development system. The progress that states have made turns out to depend heavily on the stability and consistency of their efforts over time, as we clarify toward the end of this section, and so the importance of political factors is clear.

The Impetus for Reform

      During the 1980s, very few states had thought much about reforming their workforce development systems in systematic ways. Since then, several developments have prompted states to become more active:

• Declining economic conditions forced some states to take a more active role in improving their economies, and enhancement of education and training became one of the policies to reverse declining conditions and enhance competitiveness. For example, Oregon's timber and wood processing industries suffered badly in the 1980s, and unemployment rose into double digits. In response, a strategic planning process led by Governor Neil Goldschmidt produced a 20-year plan,Oregon Shines,with workforce development as one of its three priorities. Massachusetts released a report called Choosing To Compete,also stressing the role of education and training. Similarly, the decline in Iowa's agricultural base, along with a major reorganization of state government, caused the Department of Economic Development to prepare a strategic plan to include business and industry, educational institutions, and local economic development officials. The initial plan, New Opportunities for Iowa,released in 1987, included various recommendations to refine the state's education and training programs. And when Arizona identified ten economic clusters for recruitment and expansion in the early 1990s, education was earmarked as one of seven mechanisms to achieve this kind of economic development.
  
  
• Some states discovered on their own the bewildering variety of workforce development programs, and took steps to reduce the chaos and duplication well in advance of any federal initiatives. For example, a study conducted in 1985 by the Wisconsin Department of Industry, Labor, and Human Relations found over 40 programs scattered in 12 agencies. In response, Governor Tommy Thompson created the Wisconsin Jobs Council in 1987 to develop a framework for coordination and began supporting Job Centers, which were early versions of one-stop centers. The earlier mentioned studies in Arizona, North Carolina, and Oregon are other examples.
  
  
• Some states began their coordination efforts in anticipation of federal consolidation in 1996. In Oklahoma, a weak-governor state, state agency directors had been discussing the need for some kind of coordination; the prospect of federal consolidation strengthened their resolve to initiate a voluntary Workforce Quality Compact starting in January 1996. Their efforts have continued with gubernatorial support despite the failure of federal legislation.
  
  
• Some states reflect a mixture of influences. In Texas, for example, a legislative mandate and some "good government" Republicans recognized the lack of coordination among programs and generated several reports in the early 1990s. Legislation was then introduced in 1993 and finally passed in 1995, spurred in part by anticipation of consolidation. The reforms enjoyed bipartisan support, first with a Democratic governor (Ann Richards) and a Democratic legislature in 1993, and then under a Republican governor (George Bush Jr.) and a more mixed legislature in 1996.

      To some extent, the states that began their reforms in anticipation of federal consolidation were left high and dry when consolidation failed to pass. They found themselves trying to coordinate federal programs but still operating under federal rules and regulations. For example, the Texas Workforce Commission's (TWC) effort to provide ES funds to local agencies, to use as they saw fit, was thought by Texas policymakers to require permission from the regional office of the U.S. Department of Labor, which refused this permission. More generally, state efforts to develop flexible programs that are oriented to employer needs and reliant on employer advice have been impeded by the continuation of federal rules and regulations. As a community college administrator in Florida mentioned, "It's really hard to tell someone at Southern Bell we can't do that because our regulations won't allow it, or we must wait for approval which takes four to six weeks." However, we note that the federal efforts at consolidation have still had a positive effect on reform because they have forced states to consider what they might do if they were free of federal regulations.

Overall State Strategies

      The dominant state strategy can be simply described, though it has an infinite number of variants. In general, a state agency or office is created with a certain set of statewide responsibilities. Then--partly because state offices can rarely be in close touch with local labor market conditions--local or regional counterparts are established with responsibilities for implementing state policy in local labor markets. For example, in one of the earliest such efforts, Massachusetts created the Massachusetts Jobs Council in the late 1980s to review all work-related education and training programs, replacing the State Job Training Coordinating Council (SJTCC). Then Private Industry Councils (PICs) were transformed into Regional Employment Boards (REBs) in each of the 16 Service Delivery Areas (SDAs), with a wider range of responsibilities than PICs had. Similarly, Michigan created a HRIC called the Governor's Workforce Commission, an advisory body with administrative responsibilities carried out by the Michigan Jobs Commission; then local Workforce Development Boards (WDBs) were created to plan the allocation of funding for workforce development. Oregon has the state Workforce Quality Council (WQC), with 15 regional councils. Texas initiated the TWC, replacing the Texas Employment Commission and incorporating other programs such as JTPA, JOBS, and child care, along with local WDBs. In North Carolina, the governor created the Commission on Workforce Preparedness by executive order, abolishing the state councils for vocational education, JTPA, and basic skills and literacy. Most of these local and state bodies were created by legislation, replacing earlier SJTCCs with HRICs and expanding their mandates. A few--for example, advisory councils like Michigan's Governor's Workforce Commission, and North Carolina's Commission on Workforce Preparedness--were instituted with executive authority.

      One other exception to the pattern of legislatively developed state councils is the Oklahoma approach, which is a voluntary effort so far. In January 1996, a group of state agency heads began meeting to start addressing the obvious need to coordinate the state's programs, regardless of what the federal government did. Based on "on-the-street research" about other state efforts, the participants decided not to seek formal consolidation because this just creates "another layer of government, bureaucracy, and cost--plus people don't like to be taken over, and that's a takeover, and in most cases we found it was a hostile takeover." Instead, the group formed the voluntary Workforce Quality Compact, which met informally on Saturdays and developed a "Memorandum of Agreement Regarding the Creation of Oklahoma's Workforce Development System" in June 1996. The state group hopes to develop at least 21 local Compacts to bring together various education and training providers, but--since there is no legislation with enforcement power--these local Compacts must be voluntary, at least for the moment.

      While the overall pattern of a state agency with local counterparts is widespread, they vary substantially from state to state in their responsibilities, their influence, and the programs they incorporate. (Table 1 summarizes many of these differences among states.) One important distinction is between state agencies that are advisorywith responsibilities for planning and oversight, and those that are administrative,with the authority to allocate funds. In some states, both an advisory council and an administrative agency coexist. For example, Texas has the advisory Texas Council on Workforce and Economic Competitiveness (TCWEC) alongside the administrative TWC; and in Michigan, the Governor's Workforce Commission is advisory and the Michigan Jobs Commission (MJC) has administrative responsibilities. In North Carolina, the Interagency Coordinating Council, comprised of representatives from state agencies, provides advice to the Governor's Commission on Workforce Preparedness, which is comprised mostly of businessmen (sic).

      In general, advisory councils are less powerful and influential than administrative agencies. Several states with purely advisory state councils were not included in this study when it became clearer that the councils were powerless; and Oklahoma's voluntary efforts have found it difficult to make changes at the local level. But some advisory councils have considerable power nevertheless; for example, while Michigan's Governor's Workforce Commission is strictly advisory, its creation by the governor and its roles in reviewing state policy and in approving local workforce development plans have made it highly influential. In contrast, administrative agencies often have their hands tied, and therefore have less authority to reform programs than one might expect.

      State-level agencies vary substantially in the programs they include, and in their real authority over these programs. For example, the TWC was originally conceived as an agency that operated an integrated job training and employment-related education system; but, in practice, the TWC includes JTPA programs; the ES, which DOL claims cannot be integrated with other programs; the JOBS program, which because of the state's emphasis on "work first" provides minimal education and training; the UI system, which provides income support for the unemployed but has little connection with education and training; the Child Care Management System, which consolidates child care services from various other programs but again has little to do with training; and several smaller programs like the Food Stamp Employment and Training Program (FSETP) and the Senior Texans Employment program. TWC has placed different programs in separate administrative divisions so that the decisions across programs that might enhance coordination are difficult to make. Furthermore, the large education programs--adult education, the community colleges, and vocational rehabilitation--are under TCWEC's purview for planning and evaluation, but they continue to be administered independently of TWC. In effect, then, TWC has gained the authority to coordinate JTPA with a number of small and peripheral programs, but, in general, the large programs it consolidates are either too dissimilar to coordinate to any great extent, are protected by continuing federal regulations, or have been insulated by administrative structures.

      Similarly, in Maryland, the Governor's Workforce Investment Board (GWIB) has limited administrative authority, and administrative power rests with separate state agencies. Unlike other states (including Michigan), the current governor has been unwilling to put his authority behind the GWIB, which, therefore, has neither statutory nor political influence. In addition, all but one of the community colleges are locally directed, with relatively weak state oversight (as is true in most states), again making it difficult to coordinate across state agencies.

      In the dominant pattern, local or regional councils are responsible for carrying out policy at the local level and, therefore, are responsible for creating the "system" that potential clients (or students) and employers encounter. At this level, a number of metaphors govern what localities and states are trying to accomplish, which, like all metaphors, provide a simple though incomplete vision of what a state is trying to accomplish. One is the image of the "seamless system." In Iowa, this means that the provision of a variety of services to particular individuals is invisible to them, even though the provider agencies themselves may have to go through substantial contortions to combine services from different programs with different eligibility and regulations. The idea is the same in North Carolina's one-stop centers:

We don't talk to clients about programs, but it is up to our staff to find the pots of money that might pay for child care, transportation, which part the client should invest in. We keep up with changes in regulations.

      Other popular metaphors have relied on the imagery of doors. In Michigan, the notion of "no wrong door" implies that, no matter what state agency individuals first approach, they will be directed to the appropriate provider; Michigan has called its one-stops No Wrong Door Centers and, similar to Oklahoma, has specified three models for achieving this: (1) one-stop integration, in which agencies are physically located in the same building as well as being linked electronically; (2) multiple points of entry, in which agencies are linked electronically, even though they are separated geographically, and may serve as feeders to other programs; and (3) the hub and cluster plan, in which there is a main center of co-located programs as well as programs located elsewhere that refer clients to the "hub." During the 1980s, Massachusetts developed a similar image of "one system, many doors," assuming that individuals finding their way to any program could get the right information about appropriate services (Grubb, Brown, Kaufman, & Lederer, 1990b, p. 24).

      Florida has developed a "zipper" metaphor to describe its coordination effort which links agencies together at both the local and state levels:

Consolidation is trickier . . . . We decided not to strip the money out of the agencies and we decided to encourage the agencies to collaborate at the state level. At that point, you say that an employee over at Labor, for example, is going to have a stake in the success or failure of an employee over at the Education Department, and they will both share what is the interest of two or three agencies at the local level working together. This is called the "zipper" strategy.

It translates the same way locally. What you've got at the local level are . . . the boards and commissions that are responsible for strategy and policy. So, we've had to link them together with the "zipper." Mainly what we try to do is to have some kind of linkage at each level that tends to move people in the direction of succeeding if they collaborate well, and without a frontal assault on any one institution try to press them toward adopting the best outcome strategy.

      A third pattern has focused on creating a single point of entry to a state's system. In most states, one-stop centers provide the single point of access. In Massachusetts, the local REB is the central agency that charters the regional one-stop center, which is supposed to be the central point to which all individuals and employers come for referral to any services they need. The creation of a single point of contact is a response to the criticism that the proliferation of programs has made the system so confusing to its potential clients that they cannot find their way to the programs they need. Something similar may be achieved when programs systematically refer individuals to one another. In North Carolina, for example, one local official noted,

Our staff pick up the phone and call other agencies and say, "We think we have a good referral," and we set up an appointment right then. Some agencies have said, "If you get a person in your chair that is a good referral, call us and we will come right out." These are not written in contract agreements, but that is part of our customer perspective.

      The ability of local councils or boards to achieve "seamlessness," or to act as a "door" to all other programs, depends on the responsibility and authority given to these local entities. When certain programs are outside the scope of local responsibility--as is usually the case with the major education programs, as we will see--then local workforce development councils do not formally have the ability either to refer individuals to education, to create vertically integrated programs (e.g., from short-term job-training to longer-term education), or to craft a multiple-services strategy that relies on the extensive resources of programs outside their scope of authority.

      Michigan has taken a new approach in designating three tiers of programs for which the local WDBs are responsible. Tier One programs are under direct control of local WDBs, and, in addition, must be accessible through the No Wrong Door Centers. Local WDBs have only planning authority over Tier Two programs, and only some of these programs must be included in No Wrong Door Centers. Finally, local councils or boards are expected to use their local knowledge and contacts in the community to influence Tier Three programs, which include those that are less directly related to workforce development (like K-12 education and public transportation). The designation of three tiers, therefore, defines a continuum of local control, with local boards having the most authority over Tier One programs and the least over Tier Three.

      In practice, many states are relying heavily on one-stop centers as the linchpin of local coordination.[8] Very often, local workforce development councils have been given responsibility for one-stop centers as a way to begin the process of local coordination, thereby melding a federal initiative with a state process. These centers, initiated in 1994 by the DOL, represent the federal government's most substantial effort to coordinate disparate federal programs. In a system where resources are usually committed to direct services and cannot be reallocated to a coordination role, federal funding for one-stop centers represents some of the only funding for system-building activities. However, it is crucial to identify precisely what these centers do because the rhetoric around them is much more expansive than the reality of what they accomplish. In general, we found that one-stop centers at the local level do one of three things[9]:

1.	Information:By far the most pervasive role of one-stops is simply to provide information about the services available in a local area. Often these are described as "self-service," which means that individuals can gain access to descriptions of local programs and to labor market information--frequently in computer-based archives--that they use on their own, without much guidance from a counselor or caseworker. Much less often, one-stop centers provide workshops about how to use such information, or provide direct help in locating information--particularly to specific client groups like welfare recipients. The provision of information in one place, particularly where information has been hard to find, is a step in the right direction, and is crucial in states that are beginning to use market-like mechanisms that presume a well-informed "consumer" of education and training. But whether the least sophisticated individuals searching for education and training programs can find their way to one-stop centers, and whether they can take advantage of information resources (particularly unfamiliar computer-based resources) without help, are difficult questions that have rarely been asked, much less answered.
2.	Co-Location:A smaller number of one-stop centers co-locate offices of education and training providers in one space. This provides information about the services available, but it represents a step forward because an individual can get personal information and advice (rather than merely text- or computer-based information), can ask questions and explore alternatives, and can fill out applications on the spot. Co-location still places the burden on the "consumer" to find the one-stop center and decide among the services available, but it provides much richer information resources. For example, Iowa's network of workforce development centers was established to provide a range of services in a customer-oriented, business-like atmosphere. These centers, about half of which are administered by community colleges, provide customers with various services (including basic skills) through agencies such as Promise Jobs, JTPA, ES, Veterans Employment Services, Senior Services, UI, and Vocational Rehabilitation. Often adjunct services, such as Goodwill, Green Thumb, and other CBOs, occupy space within the center, providing customers with a true one-stop opportunity for services.
3.	Service Coordination:A still smaller number of one-stop centers have started to coordinate the services in their area. For example, the Newmark Center in Oregon coordinates services provided by Adult Basic Education, Adult and Family Services (the TANF program), and sixteen other public and private agencies within a single building. All these programs participated in developing the mission of the center, which focuses on accessibility, convenience for clients, and providing multiple services in addition to training (like child care and family abuse prevention). Individuals seeking assistance fill out a common in-take form and attend orientation sessions that provide an overall view of all services provided and then direct individuals into particular classes or programs. The philosophy of the center emphasizes "really looking at the whole aspect of moving people into first jobs and then self-sufficiency."

      However, the number of one-stop centers that play a role in coordinating services is still small. In most cases, the local councils are still feeling their way, learning about local participants and programs, and they lack both the moral authority and the administrative power to require coordination. In some cases, state bureaucracies seem to have impeded the development of one-stop centers. In Maryland, for example, some local job training officials were disappointed in the state's one-stop system, and viewed the state bureaucracy as an impediment to greater integration of local efforts. They also viewed the state's welfare initiative as "not helping the system work better together."

      As we will see more clearly in Section III, the development of locally coordinated systems is something that, as local administrators all acknowledge, requires a long period of time. Stability in program rules and regulations; good personal relationships among administrators; programmatic flexibility; and, ideally, support from the state level are all necessary (Gula & King, 1990). It is, therefore, unrealistic to think that one-stop centers, which are relatively recent and provide limited funding without any direct leverage over other programs, can instantly improve local coordination. As a mechanism to enhance information about existing programs, and as a process for getting local programs to begin working together, they are universally regarded as steps forward. But if they are confined to a role in information provision, then they will never be the mechanisms of coordination that most states envision.

The Instruments of State Policy: The Variety of Approaches

      In reforming their unwieldy systems of workforce development, states have used a variety of mechanisms. In this section, we contrast several different approaches to state action: institutional approaches that try to strengthen the quality of public institutions, as distinct from market-oriented approaches that use competition and other market-like mechanisms to induce change; and mechanisms that use various inducements, including funding and technical assistance, contrasted with those that try to impose change via mandates, sanctions, and other punitive measures.

Institutional Versus Market-Like Mechanisms

      A decade ago, most efforts to improve state systems emphasized institutional mechanisms of change--efforts to improve the quality of public programs and the connections among them, through public policy that required government programs to reform in particular ways. Since then, a greater interest has developed in the policies that mimic markets instead, through price-like incentives for programs to change through competition, and through voucher-like mechanisms that make consumers rather than governments the real decisionmakers. The rise of market-like mechanisms has been due, we surmise, to dissatisfaction with the slow pace of reform via institutional mechanisms, and to the ascent of free-market ideologies over faith in government.

      Among the institution-building mechanisms that states commonly use are the following:

• Advisory committeesare designed to provide information to programs, particularly about what employers want most. Following the example of PICs, a clear trend is to require the majority of advisory committees' members to be from the private sector. For example, Florida's regional WDBs require 51% participation by local business leaders at the Chief Executive Officer (CEO) level. These regional WDBs also have the option of including the local WAGES (Work and Gain Economic Self-Sufficiency) Board, and often use the same membership to address the needs of welfare recipients as well as the local workforce. In other cases, business groups have a substantial influence. In Maryland, for example, the Business Roundtable has been particularly active in informing the state's school-to-work (STW) efforts.
  
  
• Consolidation may improve the quality of public institutions by allowing particular state agencies to make decisions involving more than one program, so that judgments can be made about relative strengths and weaknesses, and about appropriate patterns of cooperation or integration. Of course, as we pointed out above, the effectiveness of consolidation depends on which programs are included and excluded, on whether some programs included cannot be changed for political or regulatory reasons, and on whether a consolidated state agency, in fact, makes joint decisions--rather than programs in different divisions being administered separately as in the TWC
  
  Oregon has taken the idea of consolidation one step further by making all state decisions with crossfunctional teams. These teams have representation from all the major state agencies concerned with workforce development, on the assumption that "by combining together we have better results than we do alone." Teams focus on particular issues to do "the detailed policy work that's necessary, rather than throwing people together and hoping it works out." Each state agency has members on such crossfunctional teams; for example, the Human Resources team in the Office of Community College Services represents community colleges in all initiatives, and thereby develops expertise on statewide issues. These teams have developed the state's plans for innovations like the WQC, one-stop centers, STW, and welfare reform. In addition, state personnel tend to work in several agencies during their careers, again supporting crossfunctionality in their approaches to decisions.

• Technical assistance,or advice from state agencies to local providers, is comparatively rare. Technical assistance presumes that state agencies have developed some expertise about quality to share with local agencies, and--as we will see in greater detail in Section VII--state officials often have little idea about the key components of a "good" education or training program, aside from those expectations embedded in state performance measures. However, North Carolina has increased its technical assistance, particularly in the form of staff development provided through the Workforce Development Institute (WDI), as described in Box I.2. Oregon has also developed staff training for program managers on various statewide priorities; this provides not only information about emerging issues but also a "signal that everyone needs to change in order to be successful." Maryland has provided professional development through its Department of Education, and convened local leadership to provide opportunities for shaping strategies, under the assumption that it is a local responsibility to define models of high-quality services. In addition, the Department of Human Resources (DHR), which oversees welfare reform, describes itself as a "technical assistance/consulting unit and partner" rather than a monitoring agency, and has provided some support services, initiated pilot projects around collaboration, and facilitated local partnerships.
  
  State agencies are in good positions--certainly better than local programs--to collect information about effective practices around the state (and from other states), to evaluate claims of effectiveness, and to disseminate information about these practices. However, technical assistance requires a different kind of expertise than the "traditional" state activities of approving, monitoring, and coordinating local plans, and many state agencies have found it difficult to make the shift. While a few states are trying to be more service-oriented rather than regulatory, this approach to improving quality has been tried in only a few states.
  
  As states move toward greater accountability, technical assistance becomes, somewhat paradoxically, more important. If programs are held accountable to state standards, but lack the capacity to change, then accountability can only identify poorly performing programs without improving them--a politics of humiliation rather than improvement. But if local programs can develop the capacity to change through technical assistance, then accountability can stimulate improvement in the workforce development system. In effect, there are two tenets of the new approach: there shouldn't be public funding without accountability; and, in turn, there shouldn't be accountability without technical assistance. Or, to put it simply: no funding without accountability; no accountability without technical assistance.
  
  
• Cultural changelies at the heart of Oregon and Oklahoma's efforts. In Oregon, the state is trying to move away from a traditional model--funding for relatively specific services with regulation and compliance review to make sure that funds are appropriately spent. In its place, the Oregon Option has articulated a "redesigned system" with four principles: (1) focusing on results rather than inputs; (2) promoting integration of services across agencies; (3) investing in prevention; and (4) delegating responsibilities for service to front-line, local providers. The approach may be described as one of policy leadership, rather than mandated coordination or consolidation. As one state official noted, what the state has done is to "make the most of the state policy leadership's interest in a quality workforce in Oregon through the influence of how we make decisionsrather than how we organize." This vision provides greater authority to local WQCs, under the assumption that they can be more effective in following these four principles rather than simply obeying regulations.
  
  Similarly, in Oklahoma, the state Workforce Quality Compact has developed a set of principles and goals that each member agrees to, though there are no sanctions except peer pressure. The principles represent an effort to get all state agencies--and, over the long run, local programs--to operate in different ways. These are cultural changes because they attempt to influence the basic approach of local and state agencies without coercion; they do not rely on the usual model of funding and regulation, and there are no sanctions if local governments do not comply.

      Increasingly, however, market-like mechanisms have become popular--mechanisms that try to mimic the incentives in markets, where profit incentives, competition and customer choice, and the fear of going out of business increase the effectiveness and efficiency of providing various goods and services. Among the most widely used market-like incentives are the following:

• Performance measures and standards create measures of performance--often, in workforce development programs, placement rates, earnings levels, or rates of leaving welfare--and then establish appropriate standards that local programs should meet. First introduced for JTPA in 1983, measures and standards simultaneously provide information to policymakers about performance, to administrators for use in program improvement, and to consumers about effectiveness. They shift the emphasis from inputs to outcomes--at least in theory. Florida has developed one of the most extensive system of performance measures, described briefly in Box I.3--as befits its continuing interest in creating market-like incentives for high performance--and has now moved to performance-based funding. North Carolina has urged the formation of a state skill standards board through its Workforce Preparedness Commission.
  
  However, performance measures can be valuable even if they are not used as the basis of funding. Texas has developed a system to evaluate and manage its workforce system, enacted in 1997, that applies to all education and training in the state, not just JTPA or TWC-administered programs. Based on a project carried out in six states by the National Governors' Association (Trott & Baj, 1996), and work by the LBJ School's Center for the Study of Human Resources (King & McPherson, 1997), two key performance measures have been adopted for every program in the system: (1) employment rates and (2) post-program earnings measured by UI wage record data.[10] For the moment, these measures are being used only to monitor strengths and weaknesses in the state system, not to fund or eliminate programs. Yet another example is Oregon, which developed a set of performance measures in 1989 for statewide priorities such as education, economic development, and environmental protection. Called benchmarks, the measures provide a common language to determine progress. Despite this system of measures, state agencies in Oregon do not allocate funding based on local performance as Florida does.

• Performance-based budgeting is the next logical step once performance measures have been created. In a sense, JTPA incorporates performance-based funding because programs that fail to meet their performance standards two years in a row face reorganization of the PIC. In practice, however, states rarely impose these sanctions, and they affect organization rather than funding.
  
  In the ten states we examined, Florida has gone the furthest to implement performance-based budgeting for its community colleges. In 1996, colleges "contributed" $12 million (out of a $40 million discretionary fund), and colleges could then earn these funds back on the basis of students completing associate degrees and vocational certificates, with double incentives for economically disadvantaged students, dislocated workers, and welfare mothers. Initially, participation by community colleges and technical centers was voluntary; however, Florida is planning to phase in Performance-Based Budgeting over the next few years, which will make participation mandatory and will increase the amounts subject to performance-based funding to 15% of all funding--though the measurement of performance is still under discussion. In addition, for its welfare "reforms," Maryland has introduced pay-for-performance contracts to private providers at both the local and state levels.
  
  In other states, performance-based funding is still under discussion. In 1995, the Governor's Commission on Workforce Preparedness in North Carolina recommended shifting from enrollment-based to outcome-based funding. The state community college system is conducting a self-study for recommendations about how current funding formulas based on enrollment might be modified by performance-based mechanisms.
  
  
• Competition among providers is a way of breaking up the monopoly of public providers and the supply-driven pattern of education and training programs. Florida has taken this the furthest, with provisions in 1996 legislation that tried to create a level playing field, encouraging competition among programs. If a local provider is not meeting the needs of business, another program (public or private) may offer a similar program and--potentially, at least--drive lower-quality programs out of business. As one community college administrator noted, "Under this bill, there isn't a hiding place for programs of low quality." Similarly, Iowa would like to introduce competition among service providers, with the state government shifting from being a provider of services to being a broker of services provided by a variety of public and private agencies. However, early efforts to introduce competition raised the ire of state employees unions, and progress has been slow.
  
  
• Competition through subcontracting is a mechanism for providing services under more competitive conditions, where public and private providers compete for public funds to provide specific services. Subcontracting has a relatively long history in local and state government, where highway construction, garbage collection, janitorial services, bill collection, and other similar services have been contracted to private firms. In the education and training sphere, a few school districts have made headlines by subcontracting some schools to private firms like the Edison Project, or services to other businesses like Score! and Sylvan Learning Centers. In workforce development programs, local JTPA programs commonly use subcontracting mechanisms, presumably choosing among different bidders for specific services on the basis of price and quality. Currently, subcontracting is being extended to welfare-to-work programs, where certain states (like Maryland and Florida) and localities (like Milwaukee) have subcontracted with private providers for specific services. In addition, as part of its promotion of competition, Florida and Texas are requiring their regional boards to subcontract for services as a way of increasing competition. Many local boards have been designing a "simulated block grant" approach, consolidating separate funding streams (including STW grants, welfare-to-work funds, and HS/HW Incentive Funds) and contracting out the provision of direct services. For example, in Palm Beach and Dade Counties in Florida, Lockheed Martin has contracted to provide case management services for WAGES recipients. Michigan requires its local Michigan Works! boards to subcontract for services, under the belief that this will yield a more efficient delivery system. The core of Massachusetts' coordination policy relies on a subcontracting mechanism: it funds new career centers by redirecting funds away from agencies that have traditionally received them to new centers competitively selected to meet customer needs.[11] The state's experience so far is that, while the one-stop centers established through competition are of high quality, the intra-state conflict created by the process may have undermined their long-run base of support. Similarly, Texas requires its local WDBs to subcontract the management of its One-Stop Career Centers, and in Dallas County that provider is Lockheed Martin, a private, for-profit firm. The common statement that governments are turning into brokersof services, rather than direct providers,is usually a reference to the increasing use of subcontracting.
  
  
• Voucher mechanisms are ways of providing individual consumers with the purchasing power necessary to make choices. These vouchers, of course, have little effect if choices do not exist, and, therefore, their effectiveness depends on the existence of competition among public and private providers. The largest voucher mechanisms in the national education and training system include federal grants and loans to postsecondary education, including Pell Grants to low-income students, the new Hope Tax Credits, and the Lifelong Learning Credit, all of which can be spent in community colleges and proprietary schools (but not for adult education or short-term job training). The WIA of 1998 also places greater emphasis on Individual Training Accounts, which may turn out to be like vouchers. Voucher-like mechanisms have been contemplated at the state level as well, including a pilot voucher project under JTPA in Baltimore that provides "customized management accounts" to eligible clients, allowing them to select among training providers with a lifetime cap of $7,000 per individual.
  
  
• Consumer information is a widely accepted requirement of market-like systems, since uninformed consumers cannot possibly make rational choices. One-stop centers are the most obvious mechanisms for increasing consumer choice. In addition, several states are planning to make information available from their performance measures. However, it is important to note that one-stop centers and other public sources of information usually refer individuals only to certified public providers, rather than to a broader array of providers as a true voucher system would (Barnow & King, 1996). It is possible, however, that this will change under the WIA.
  
  
• Other market-like policies outside the realm of education and training programs include tax incentives for business and regulatory relief, both of which figured prominently in Governor Engler's reforms in Michigan. The tax mechanisms included the creation of tax-free Renaissance Zones in economically depressed areas, with businesses being promised ten to fifteen years exemption from local and state taxes. However, the effect of such policies on education and training programs is indirect and difficult to trace, and, in any event, is contingent on whether such policies have any effect at all, which is often quite doubtful. Particularly in market-oriented states, workforce and economic development can be promoted through tax and regulatory mechanisms rather than--or in addition to--interventions into education and training programs.

      There is, of course, a long history of debate about the efficacy of each of these mechanisms of state policy. Many of the debates restate the conventional claims for and against market mechanisms, without providing much empirical evidence about how they operate in practice, particularly in the education and training field. In the case of subcontracting, there is evidence that the presumed efficiencies of subcontracting are often illusory, since the costs of contract monitoring often outweigh any savings. In the areas of education and training, the evidence from England--which has tried a broad variety of market-like mechanisms under the Thatcher and Major governments--indicates that such devices do not increase the exercise of choice except among the most sophisticated and well-informed consumers, and that they tend to exacerbate inequalities just as their opponents in this country claim (Finkelstein & Grubb, 1998). We note, therefore, that many arguments in favor of market-like mechanisms are really untested claims--though state practices are now providing opportunities to evaluate these claims.

      Of course, some states combine both mechanisms of institutional improvement and market-like practices. One-stop centers are examples of mixed policies since they provide information to consumers, both directly and through co-location, but they may also facilitate coordination--a form of institutional improvement--through co-location and through more direct activities with providers. Performance measures can be ways of improving consumer information and choice, though when they are used exclusively by state administrators to scrutinize local programs they are primarily mechanisms of accountability rather than consumer information.

      In general, states have been moving toward greater use of market-like mechanisms, though only a few states have adopted the most powerful forms: competition and performance-based funding. The ten states we examined can be roughly placed along the continuum depicted in Figure 1. At one extreme, Oregon has depended almost entirely on institution-building practices, including some consolidation; coordinated services for

welfare clients provided by JTPA and community colleges; crossfunctional teams to develop new state initiatives, described above; the development of indicators of quality (e.g., customer satisfaction, retention, and skill gains), not for punitive or funding purposes but to signal the need for improving quality; the development of statewide benchmarks, used not for mandates and sanctions but to "create an environment that supports performance and measurement," a "shared set of expectations" (Oregon's Benchmarks: Setting Measurable Standards for Progress,1991, pp. 1-2); and the development of long-range plans (like the 20-year strategic plan in Oregon Shines[1989] and Oregon Shines II[1991]), recognizing that workforce development and coordination are complex issues requiring long periods of time and stable leadership. In general, Oregon tends to distrust competition and other market-like mechanisms as methods of driving reform. Several pilot projects using market mechanisms--focusing on vouchers and competitive purchase of services using JTPA funds--have been tried, but the dominant sentiment is that "throwing it open to competition gets us just another version of chaos."

      At the other extreme, Florida has consistently followed the approach of enhancing market-like mechanisms. It has introduced competition into the system by developing three tiers of performance measures, which are at different stages of development and still require refinement:

1.	Customer outcomesas measured by employment in an occupation, demonstrating growth, continued employment, reduction in and elimination of public assistance, employer satisfaction with level of preparation, and return-on-investment.
2.	Program outcomeswith specific outputs in each of the four workforce strands: STW, welfare-to-work, one-stop centers, and the HS/HW initiative. Each component will calculate such output measures as the number of dropouts, leavers with marketable skills, program completers, and job listings filled.
3.	Program measuresthat measure services to enrollees, service recipients, and individuals moving among service components.

      In addition, Florida has forced all public and private providers of VET to compete with one another; it has had performance-based funding in vocational education since the late 1980s,[12] now extended to community colleges as described above and being phased into other VET programs; the state's Jobs and Education Partnership is developing still other performance indicators, to be used to develop future strategic plans; and FETPIP, which links records from various institutional sources, and provides information on outcomes. The state has been quite consistent, over more than a decade, in developing these market-oriented initiatives and--while it still has a range of more institutional mechanisms, including a local-state structure to administer programs like STW and one-stop centers--it uses a wider variety of market-oriented mechanisms than any other state.

      In between, as Figure 1 indicates, states use a wide combination of institutional and market-oriented mechanisms. For example, North Carolina has placed great emphasis on its technical assistance, but it has also stressed information through one-stops as a way of enhancing coordination among programs. In general, many of these states are moving toward limited coordination of programs (particularly those considered trainingrather than education) and reconstituted advisory committees on the institutional side. At the same time, many have adopted outcome measures and information provisions (particularly through one-stop centers) on the market-oriented side. Whether through planning or happenstance, most states have avoided the extremes of one approach or another, preferring a mixture of policies to improve their workforce development systems.

      We stress, then, that states have developed a variety of approaches to reform and coordination. Aside from the nearly ubiquitous use of a state-level agency with local or regional counterparts, there is hardly a single practice that has been adopted by all these states, and it is certainly incorrect to think that simple consolidation--the combination of programs with some incorporated into others--is a dominant strategy. Furthermore, the exceptions--the innovations that particular states have undertaken, like North Carolina with its technical assistance, Florida with its market orientation, and Oregon with its array of institutional mechanisms--are in many ways more interesting than the common practices. There is certainly no notion among the states that one size fits all--even in a relatively limited number of states.

Economic Development Efforts: Recognizing the Demand Side

      In addition to reforming their workforce development systems, many states have initiated policies to stimulate the demand side of the labor market. The underlying logic is that developing a more educated labor force, without increasing the demand for well-trained workers, is likely to be ineffective. Among the most important demand-side strategies are the following:

• Most states have programs to provide short-term customized training to specific employers, partly as a way of luring employers from out of state or keeping existing industry from leaving. Many of these take the form of state grants for firm-specific training, often operated through community colleges. In Maryland, the Department of Business and Economic Development (DBED) is collaborating with the DHR to assist businesses locating in Maryland, doing "whatever is needed to get businesses to locate here," and has also initiated collaboration with local SDAs and Job Service offices to assist with corporate expansion and recruiting. Iowa has taken a novel approach in allowing its community colleges to float bonds for firm-specific training; the bonds are then paid off by earmarking the income taxes paid by the corporations participating. This is essentially a way of allowing local colleges and employers, rather than the state legislature, to determine the spending for such activities, and Iowa's efforts are thus better funded than in most states.
  
  
• North Carolina has been particularly active in promoting economic development as the goal of all state policies, and has created community colleges that are specifically active in occupational education, in workforce training for employers, and in providing training for JTPA and welfare clients. Numerous community colleges have become active in such workforce development activities, including small business development centers funded by state resources in Oregon and North Carolina, providing information and (if necessary) training to small businesses about new technologies, business procedures, quality requirements, health and safety issues, and other issues relating to profitability that might otherwise escape their attention.[13]
  
  
• As part of its efforts to diversify the state's economy, Oregon's Economic Development Department has designated Key Industries, and each local Regional Strategies Board has selected three of these to emphasize in its six-year economic development plan. Some Key Industries have formed industry associations that, in turn, have become important in communicating sectoral needs to providers of education and training. For example, the Semiconductor Workforce Consortium of Oregon and Southwest Washington has defined a standardized set of job skills and developed industry-specific courses at the high school and community college levels. Several other associations are developing their own statements of skill requirements, and the state has begun to develop a database that would give VET providers access to information about job openings described by standardized skill and training requirements.

      There are, then, some steps being taken by states to stimulate demand for well-trained workers as well as to supply more education and training. Like state efforts to reform workforce development systems, these demand-side strategies hardly existed a decade ago. However, the effectiveness of these policies remains unclear. Many of these are still efforts to engage in "smokestack chasing," or efforts to lure mobile industry from other states--a tactic that is not very effective. In many cases, states do not discriminate between cases that might justify public funding for training and cases where employers should provide their own training because there are unlikely to be public benefits; often, a great deal of state money is being spent to little effect.[14] Certain other policies seem more promising, like the Oregon efforts to organize industry associations, but these are still not widespread. Finally, some community colleges' efforts in economic development (including small business centers to improve the efficiency and longevity of local businesses) provide help to employers while they supply any necessary training to new employees (Grubb, Badway, Bell, Bragg, & Russman, 1997). For example, the Maricopa Community College District brought together employers in several sectors, and now promotes its Semetech Network of semiconductor producers and suppliers:

We tend to be the state's neutral meeting ground for a number of employers. Instead of competing with one another, they're working together with us in a safe zone to pool resources in order to develop the technical capacity in our community to serve them with employees. Instead of stealing employees from one another, they work with us to develop technical capacity within their firm to train their employees. We focused on areas of mutual concern, mainly marketing the industry as a viable career. We needed students in the front door, and the firms needed students out our back door, so we cooperated.

      In such approaches, the conditions for economic expansion are effectively coordinated with the provision of an educated workforce.

In addition, economic development efforts in most states are not well-coordinated with education and training. Usually they take place in different departments. In Maryland, for example, Governor Glendenning's administration has deliberately separated economic development from employment and training: job training is administered by the Department of Labor, Licensing, and Regulation, while the new DBED handles all economic development. There is no consensus within the state about the rationale for this division, and many administrators are simply mystified. In Michigan, the extensive economic development efforts through Renaissance Zones and the Economic Development Job Training Program are uncoordinated with workforce development. As one researcher mentioned about the division between the Economic Development Service Division, which routinely calls on businesses but fails to develop job slots for JTPA clients, and the Workforce Development Division,

In theory, the workforce development people within the Job Commission essentially run their own programs, and they don't interact with the people who are calling on business. It's a sham. It's an agency running economic development and workforce development in parallel, without any synergy between them.

In theory, the great variety of state-funded training for employers supports economic development and, therefore, employment demand, while the training simultaneously creates the supply of trained workers necessary. In practice, however, states have given little thought to the conditions under which such programs stimulate demand and, therefore, they often provide subsidies for training (supporting the supply side) without enhancing demand.

      In general, therefore, there are no mechanisms in place to match supply with demand, even in a rough way. For example, most states would like to see high-productivity, high-wage employment move in or expand, but that is inconsistent with the need for more jobs for the least skilled welfare recipients going through "work first" programs. Similarly, most states have paid more attention to coordinating their short-term job training programs, but these are again aimed at relatively low-level jobs rather than the high-productivity jobs that most states want. Most states continue to talk about the skill requirements of the 21st century, and most state reports have a section repeating the need for higher-order skills, communications skills, and problem-solving abilities, as well as better preparation in basic verbal and mathematical competencies. However, aside from STW initiatives and standards-based reform in K-12 education, there has been almost no attention in state efforts to the kinds of curriculum and teaching methods that would ensure that such competencies are taught, especially in the short-term job training programs and in "work first" programs that provide no education or training whatsoever. At several different levels, therefore, there is a real mismatch between the emphasis of state economic development efforts and the reforms in workforce development systems.

      As always, there are important exceptions, usually in conditions where the same program participates in both economic development and in providing education and training. For example, small business centers in community colleges often identify technology and training needs within small firms, and also identify sources of supply. The area vocational-technical centers in Oklahoma support economic development by providing firm-specific training, and they have become one of the main strategies to attract business into particular regions of the state. The Regional Strategies Boards in Oregon identify Key Industries for local areas, and then are expected to link with regional WQCs to develop integrated plans for both workforce and economic development. In Maryland, DBED is trying to better integrate its economic development with education and training needs, trying to bring a business focus and demand orientation to the workforce development system. Texas has a process of identifying labor market information into planning models issued by local boards and community colleges; this process is intended to help them choose high-demand occupations and industries. Iowa community colleges can both identify needs for training among new and expanding industries and provide that training directly. North Carolina has tried to integrate the two by including community colleges whenever there are efforts to lure industry to the state; as a local official noted, "Economic development used to be location; now it's education." This is often accomplished by creating job training programs within community colleges, both through the short-term noncredit vocational programs called occupational extension and through special centers. As one administrator said,

We try to sell a unified vision of what economic development means in Columbus County, including an industrial park, good-paying jobs, and our Industrial Skills Training Center we built with the community college. The lack of a skilled workforce is a weakness when we try to attract new industry.

      Baltimore County provides an interesting local example of the integration of demand-side and supply-side policy. In the past, the Baltimore County office had created "speculative" training programs, without having jobs in hand. Now, they initially meet with employers through economic development partnerships, and then develop customized training programs to ensure that trainees have the competencies employers need so that they will be hired. The office contracts out its training to outside providers, believing that it can thereby serve better as an "honest broker," bringing together a range of services (including training) for employers.

      There are, then, a number of fledgling efforts to coordinate demand-side and supply-side policies. In the absence of such mechanisms, states are all too likely to engage in economic development efforts that bear little relationship to reforms in workforce development systems.

      On a more positive note, however, these state economic development strategies are quite new. Just as coordination among education and training programs takes considerable time, it may be unrealistic to expect that economic development efforts are at this point well-integrated with education and training reforms. As states continue to develop their "systems," however, this is one form that future coordination could take.

[8] See also the evaluation of one-stop centers by Kogan, Dickinson, Fedrau, Midling, and Wolff (1997), who came to similar conclusions about the central roles of one-stops.

[9] Again, our findings are relatively consistent with those of Kogan et al. (1997). While they did not state how common various kinds of services in one-stops are, their discussion emphasizes different kinds of information provision, from which we infer that this has been the most common approach by far.

[10] Similarly, the state of Washington has development several excellent reports on the employment benefits of its programs using UI data; see especially Workforce Training and Education Coordinating Board (1996).

[11] Traditional agencies and programs can still win these contracts, but they must compete for such funding.

[12] Florida has required all vocational programs, both in community colleges and technical centers, to have 70% rates of placement in related occupations or face having that program eliminated. There have been various debates over the years about the data necessary to enforce this provision.

[13] On the roles of community colleges in workforce and economic development, see Grubb, Badway, Bell, Bragg, and Russman (1997) and Dougherty (1998).

[14] For an analysis of the conditions under which states should and should not subsidize firm-based training, see National Center for Research in Vocational Education and the Center for Labor Research and Education (1994), a study prepared for the Employment Training Panel in California. Aside from Iowa's prohibitions on subsidies to wholesale and retail trade and professional services (since they are not mobile in response to subsidies) and California's restriction of funding to firms engaging in export out of the state (or import substitution), from what we can ascertain from this study, states do not appear to distinguish worthy from unworthy subsidies.

Grubb, W. N., Badway, N., Bell, D., Chi, B., King, C., Herr, J., Prince, H., Kazis, R., Hicks, L., & Taylor, J. C. (1999). Toward order from chaos: State efforts to reform workforce development systems (MDS-1249). Berkeley: National Center for Research in Vocational Education, University of California.