Abstract
The ascendance of federal grants to states and localities as a major tool of government action has fueled scholarly interest in building a better theory of intergovernmental management (IGM). It has also spawned an enduring metaphor, “picket-fence federalism,” which has done much to shape thinking about the context and nature of IGM. More recently, however, a competing conceptual lens called “executive federalism” (EF) has emerged. Proponents of this perspective contend that administrative discretion looms increasingly large in shaping who gets what from federal grants, that vastly greater dependence on program waivers has driven this development, and that political executives—both elected and appointed—play a growing role in the administration of grant programs. To the degree that the EF lens accurately captures developments, it challenges the conceptual and empirical underpinnings of both the picket-fence model and IGM. This article provides a preliminary test of the EF perspective by examining the case of Medicaid. It charts a research agenda that would more fully probe the implications of EF for a theory of IGM.
Keywords
Introduction
A bedrock element of the American political system, federalism has long shaped the contours of public administration. It has affected the selection of policy tools for the accomplishment of public purposes. It has posed myriad challenges for the administrative agents of government as they implement programs. Federalism has influenced the degree to which these agents achieve acceptable levels of performance—of efficiency and effectiveness; it has affected the extent to which they are politically responsive and accountable. The ascendance of federal grants to states and localities in the 1930s and accelerating in the 1960s fueled scholarly interest in describing and explaining the intergovernmental dynamics of administration. This development further vanquished “dual” or “layer cake” federalism—a pattern where a well-defined division of labor existed between the national government and the states (Walker, 2000). Instead, “cooperative” or “marble cake” federalism appeared on the table with the national and state governments working together (not always harmoniously) in a panoply of policy spheres.
The proliferation of federal grants energized and reshaped the study of intergovernmental relations (IGR) within political science and public administration. IGR zeroed in on the full spectrum of interactions between elected and nonelected officials in federal, state, and local governments (Wright, 1978). Grant programs constituted an increasingly important focus within this larger domain. In constant dollars, federal grants to states and localities were seven times greater in 2010 compared to 1965. Beyond the dollars involved, the number of grant programs grew from roughly 130 in 1960 to about 1,000 in 2010. 1
The challenges of administering grant programs also fueled a more explicit consideration of intergovernmental management (IGM). IGM directed attention to the interplay between federal and state administrators in attempting to carry out programs. In doing so it drew heavily from the policy implementation perspective that flowered in the 1970s (e.g., Pressman & Wildavsky, 1973). The implementation literature pushed public administration beyond a focus on function (e.g., budgeting) and the organization as units of analysis to the “program.” It stressed the complexities and perils involved in transforming policy promises in Washington into program performance on the ground. The implementation literature downplayed images of hierarchy in probing interaction between the national government and states. Instead, it portrayed these relationships as arenas for negotiation and bargaining (e.g., Ingram, 1977). This emphasis on power symmetry between levels of government became a core tenet of the IGM literature. It ultimately led scholars to argue that the concepts and methods of network analysis could be applied to vertical relationships within the federal system (e.g., Agranoff & McGuire 2004; O’Toole & Meier 2004).
In addition to kindling interest in IGM, the outpouring of federal grants in the 1960s spawned an enduring metaphor—“picket-fence federalism.” As articulated by former North Carolina Governor Terry Sanford (1967, p. 89) and summed up by others (e.g., Wright, 1978, p. 61-63), this metaphor envisioned the growing number of federal grants as pickets connected by wire symbolizing a close alliance between federal and state administrators in implementing grant programs. It underscored how the similar educational backgrounds and outlooks of these administrators facilitated this working relationship. As explicated later, picket-fence federalism has shaped thinking about the context and nature of IGM.
The conceptual lenses embedded in picket-fence federalism and IGM have done much to illuminate issues of governance and public administration. More recently, however, research into patterns of federal-state relations in major grant programs has given rise to another lens called “executive federalism” (EF). Proponents of this perspective present three core propositions. First, they argue that major decisions shaping grant programs have increasingly shifted from the legislative to the executive branch at both the federal and state levels. In essence, administrative discretion looms larger than before in shaping who gets what from federal grants. Second, they contend that political executives—both elected and appointed—play an ever more important role in the administration of grant programs. Third, they highlight the importance of program waivers as an administrative tool in accelerating these two developments.
While insightful and promising, the degree to which the EF perspective in fact captures developments with respect to grant programs has yet to be systematically assessed. So too, its implications for the picket-fence model and IGM await in-depth analysis. This article takes initial steps to address these gaps in understanding. I conduct a preliminary test of EF by asking whether it accurately depicts developments in the case of Medicaid—the largest federal grant program which subsidizes health care for some 65 million low-income Americans. My findings here flow from intensive research into the evolution of that program from 1993 through 2010. 2 Paying particular attention to the dramatic increase in Medicaid waivers, I conclude that the program provides substantial support for the EF perspective. The article shows how EF challenges the conceptual and empirical underpinnings of the picket-fence model and IGM.
This article opens by assaying the defining characteristics of the picket-fence model, IGM, and executive federalism. It then draws on evidence from Medicaid and certain other grant programs to explore the potential and limits of EF. The great outpouring of Medicaid waivers in the period from 1993 through 2010 comes under the spotlight. The article illuminates how the use of these waivers has elevated the importance of the administrative presidency and also what might be termed the “administrative governorship.” Subsequent sections underscore the limits of the picket-fence model, the need for IGM to become more capacious in its conceptual focus, and the issues EF raises for democratic governance. Findings based on Medicaid may not, of course, generalize to other grants. Thus, a concluding section stresses the importance of drawing on a more systematic sample of grant programs to test the EF model. I present a specific research agenda that would advance understanding of the dynamics driving the administration of federal grants.
A key caveat deserves emphasis at the outset. This article analyzes IGM exclusively in the context of federal grants to the states. 3 It does not address the considerable utility of this perspective in studying horizontal relationships among governments, such as how officials in various local agencies attempt to coordinate activities and “network” in the absence of hierarchy (e.g., Agranoff, 2007). Nor does this article focus on IGM in vertical contexts other than grant programs (e.g., Agranoff & McGuire 2004, p. 496).
The Picket Fence and IGM
As articulated by Governor Sanford and scholars writing in the 1970s, picket-fence federalism pointed to two interrelated clusters of problems for democratic governance and public administration. The first involved concerns about excessive bureaucratic autonomy, especially relative to governors and their top staffs. Hence the metaphor assigned a major role to “vertical functional autocracies,” “balkanized bureaucracies,” and “bureaucratic baronies,” in the administration of grant programs (Wright, 1978, p. 63). The propensity of administrators to reach beyond the bureaucracy to forge a supporting coalition also garnered attention. Beer (1978, p. 17-18), for instance, noted that a “professional bureaucratic complex” comprised of “similarly trained professionals” at the national and state levels often worked closely with certain key legislators and groups benefitting from the grant program. He averred that this “tripartite complex or subsystem” functioned as a kind of “iron triangle” at the state level. Others claimed that a comparable triangle shaped grant programs in the nation’s capital (e.g., Walker, 2000, p. 25).
A second problem suggested by picket-fence federalism had less to do with democratic control and accountability than with whether the grant system undercut efficient and effective public administration. Certainly the proliferation of grants created challenges for officials in achieving administrative coordination and integrated service delivery. It generated more red tape. Each grant program involved its own set of bureaucratic players, standard operating procedures, and reporting requirements. Faced with this situation, proponents of “good government” and efficacious public administration generally gravitated to the view that “there were too many narrow, overlapping federal grant-in-aid programs.” Many reformers favored block grants as a remedy. These grants would presumably foster “simplified planning, application, and administrative procedures” as well as “reduced financial reporting and auditing requirements.” 4
The picket fence has lived on in analyses of intergovernmental relations and management (e.g., Beam & Conlan, 2002, p. 364-365; Cho & Wright, 2007). 5 To be sure, some observers have called for tweaking the metaphor. 6 Walker (2000), for instance, has suggested replacing it with “bamboo fence federalism” (p. 25, 140). He believes this view better captures “the Federal grant administrators’ capacity to bend with but not succumb to prevailing political winds.” But the precise implications of this metaphor and how it differs from the original picket fence remain underspecified. The bamboo fence has not gained conceptual traction.
The fact that the number of grant programs had grown to 1,000 and showed no signs of abating naturally reinforced the view that the picket-fence model had continued relevance. Only a few researchers sought to test the veracity of the model in a fuller sense. For example, Hale and Palley (1979) surveyed state administrators to probe the question of “who governs” in the case of federal grants. Consistent with key tenets of the picket-fence metaphor, they found that administrators in agencies which depended more heavily on federal grants reported being more insulated from executive and legislative oversight and direction. They were also more likely to reach out to interest groups to pressure the legislature for support. Still, Hale and Palley concluded that the picket-fence model overstated the tightness of the vertical alliance between national and state administrators and incompletely captured the complexity of their relationships. Subsequent surveys of state administrators tended to reinforce the finding that agencies drawing more of their funding from federal grants attributed less influence to governors and legislators. 7 Aside from these studies, my literature search unearthed no full-blown assessments of whether the governance model embedded in the metaphor continues to be valid. Instead, scholars have persisted in using the picket fence as a heuristic device to capture in a shorthand way the essence of the grant system (Stewart, 1984).
In key respects, the conceptual underpinnings of IGM reflect the thrust of the picket-fence metaphor. In the 50th anniversary issue of the Public Administration Review, Deil Wright (1990) indentified several defining qualities of IGM as distinct from broader theories of intergovernmental relations. Two loom particularly large in this context. First, he noted that IGM reflects a “modest, marginal and moderate approach to the resolution of . . . interjurisdictional issues” (p. 170). It was about coping and incremental problem solving often achieved via networking, negotiation, and bargaining. This imagery fits neatly with the picket-fence model as administrators strive to bring some coherence, coordination, and integration to the plethora of relatively small grant programs. It squares with the picture of federal and state administrators haggling over the details of grant implementation. Second, Wright (1990, p. 170) noted that IGM stresses the “gulf or gap between career personnel and political actors.” While IGM acknowledges the role of political actors in shaping implementation, it contains a “lesser political quotient” (p. 170, 173). This conceptualization also squares with the picket fence and its vision of a relatively autonomous professional bureaucratic complex shaping the specifics of grant programs. Wright’s distinctions continue to underpin discussions of IGM in the context of federalism (e.g., see Agranoff & McGuire, 1999, 2004; Kincaid & Stenberg, 2011).
Executive Federalism
EF presents a vision that differs appreciably from the picket fence and IGM. Historically, scholars have employed the concept of “executive federalism” to describe the working relationship between national and provincial executives in Australian, Canadian, and German federalism. As Walker (2000, p. 24-25) underscores, “It depicts the corresponding capacity of premiers at both levels of each system to meet regularly and arrive at binding agreements.” No such orderly arrangement exists in the United States. Instead, the concept in the American context derives from an essay by Thomas Gais and James Fossett (2005) prepared for a volume on the federal executive branch. As noted earlier, EF possesses three primary features. First, it stresses that major decision sites shaping who gets what, when, and how from grant programs have increasingly shifted to the executive branch at the state and federal levels. Its focus on administrative discretion in the intergovernmental process in part echoes the picket-fence model. But it presents this discretion as much larger—as having the ability to transform grant programs with no significant changes in federal law. “The strategic exercise of executive powers to promote major changes in state policies or administrative practices” goes to the heart of EF (Gais & Fossett 2005, p. 507).
Second, EF portrays grant programs as dominated by potent political executives rather than the relatively autonomous bureaucracies or the “iron triangles” of the picket-fence model. In this vein, EF points to the emergence of a “new intergovernmental dynamics” where “presidents and their appointees have been able to produce significant changes in program management, coverage, and standards without new legislation.” These political executives have successfully worked “to alter the context of state policy making and administration in order to influence state choices.” The program changes that occur (often “important policy innovations”) hinge less on who dominates Congress “than on control over the presidency and a few governorships” (pp. 507, 515). In sum, EF challenges the picket fence and IGM visions of the limited role played by political principals in grants administration. EF portrays grant programs as much more driven by the political masters of the career civil service, specifically top political executives.
Third, EF stresses the growing importance of a particular administrative tool in intergovernmental dynamics—program waivers. To be sure, Gais and Fossett note the relevance of a range of more conventional implements long associated with the administrative presidency. These include federal rule making, executive orders, other formal written directives or guidelines, reorganizations, the appointment of loyal personnel to key executive positions, and more. But they aver that waivers constitute “the most common administrative device” that federal executives use “to change the operation of grant programs” (p. 508). Waivers are a congressional delegation of authority to the executive branch to permit states to deviate from the ordinary requirements of the law. They are not just about interpreting the statute, but also declaring time out from it. By persuading federal administrators to approve their waiver requests, states gain freedom to customize their grant programs. As Gais and Fossett observe, waivers allow efforts “to experiment and evaluate demonstration programs; improve program administration; or provide services or cover groups other than those authorized in the legislation” (p. 508). The EF model sees waivers as front and center in enhancing administrative discretion and as opening the door for presidential and gubernatorial leadership.
In portraying the rise of EF, Gais and Fossett present an original and important thesis with significant implications for our understanding of public administration in the federal system. In doing so, they drew on examples from a range of policy spheres—welfare, health care, education, nutrition assistance, workforce development, and more. But their claims for EF represent provocative propositions rather than documented findings based on an in-depth study of a sample of federal grant programs. Much research remains to be done to determine if and under what circumstances EF applies, and where the picket fence and the assumptions of IGM continue to be germane. With special attention to the role of waivers, the next three sections of this article represent an initial step toward testing the EF model. The analysis draws primarily from my research on Medicaid but includes some reference to other grant programs as well.
The Proliferation of Waivers
Many studies have documented the propensity of Congress to delegate sweeping authority to the executive branch and the inclination of the Supreme Court to uphold these delegations. 8 An initial decision by Congress to achieve public ends through a grant program in itself signals a substantial willingness to allow significant variation in a program. The policy literature focused on Medicaid and other grant programs consistently demonstrates that states vary widely in the benefits they provide and how they deliver them. 9
The congressional decision to go one step further in delegating authority by authorizing waivers constitutes a remarkable and understudied development. No readily available census exists of all waiver authority embedded in the statutes authorizing grant programs. The precise degree to which federal executives use their waiver authority is also unclear. But without doubt waivers have surfaced in a spectrum of policy spheres. Starting in the 1980s, for instance, federal officials used waivers to open the door to state efforts to reform welfare under Aid to Families With Dependent Children (AFDC). By the time the welfare reform law passed in 1996, an estimated 75% of AFDC recipients received benefits through systems operating under waivers (Gais & Fossett, 2005, p. 508). Waivers also came to education. Legislation in 1994 gave the federal executive branch broad waiver authority; the law establishing No Child Left Behind in 2002 reaffirmed it. By the end of 2008, the Education Department had granted close to 680 waivers. 10 In 2011, the Obama administration announced its eagerness to approve major waivers that would exempt states from many of the requirements of No Child Left Behind. Eighty percent of the states promptly announced their intention to seek these waivers (Cavanagh & Klein, 2012).
My research indicates that waivers have played a dramatic role in reshaping Medicaid. Waiver authority varies in the scope of program activity it covers and the amount of funding involved. On both counts the federal executive branch has huge latitude to exempt states from Medicaid’s statutory requirements. Medicaid waivers assume two basic guises—comprehensive demonstrations and targeted initiatives focused on long-term care. Both have loomed large in the evolution of Medicaid—a program that costs federal and state governments close to US$400 billion annually and provides health care to over 65 million enrollees.
Section 1115 of the Social Security Act, which Congress approved in 1962, provides the primary provenance for Medicaid demonstration waivers. Established prior to Medicaid’s birth in 1965, the provision gives the federal executive branch broad authority to experiment with alternative state approaches to Medicaid. It explicitly envisions these waivers as a tool for policy learning by requiring that the demonstrations be formally evaluated. Prior to 1993, the federal bureaucracy used Medicaid demonstration authority sparingly, approving just 50 waivers in nearly three decades. Waiver activity during this period tended to reflect a top-down approach, with the national government typically driving the topical agenda and then inviting selected states to participate. Federal officials generally stressed the research component and limited duration of the waivers. With the arrival of the Clinton administration in 1993 and continuing to the present, this approach changed and Section 1115 waivers multiplied apace. By late 2006, the federal government had signed off on nearly 150 demonstration waivers. 11 Forty-four states and the District of Columbia operated some facet of their Medicaid program under a Section 1115 waiver. In doing so, federal officials downplayed the importance of rigorously evaluating the demonstrations. Instead waivers became a vehicle for accommodating state preferences (Thompson & Burke, 2007).
Transformation via demonstration waivers
The demonstration waivers tended to be broad in scope. 12 States used them to move Medicaid enrollees into managed care, expand coverage to new adult populations, and embed market-based features in the program (e.g., cost sharing). The G. W. Bush administration attempted to use them as a vehicle for shifting Medicaid from an open-ended fiscal entitlement to the states to a capped block grant. 13 The waivers also loomed large in efforts to deal with emergencies, such as Hurricane Katrina, and in the attempts of some states to restructure their health care systems. The waivers vividly illustrate how the exercise of federal administrative discretion can unleash major transformations of a program at the state level.
Consider, for instance, cases from New York and Massachusetts. In 2003, New York Governor George Pataki (R) appointed a Healthcare Reform Working Group to consider ways to streamline and improve the state’s overall health care system. Ultimately, the group concluded that the state needed to move beyond piecemeal reform in reducing excess capacity in the hospital and nursing home sectors. Policymakers understood, however, that this downsizing effort stood little chance of success unless they could obtain additional federal Medicaid funds to coax health care providers to cooperate. The Pataki administration submitted a waiver proposal to accomplish this end and, after considerable negotiation, federal administrators signed off on it in 2006. The federal government pledged up to US$300 million more Medicaid dollars annually over 5 years to promote the reform. This case vividly illustrates how administrators can use waiver discretion to promote broad basic reforms in the health care sector that have little to do with the nitty-gritty of the Medicaid program itself (e.g., its eligibility criteria, service packages, delivery modes).
The use of demonstration waiver in Massachusetts was even more transformational. That state had obtained a waiver in the mid-1990s to expand Medicaid eligibility and place enrollees in managed care. When the waiver came up for renewal during the administration of Governor Mitt Romney (R) federal administrators indicated that they would no longer sign off on certain waiver provisions that the Clinton administration had originally approved. This led to intense bargaining between the Romney administration and federal officials. Ultimately, Governor Romney used the occasion to negotiate a major overhaul of the Massachusetts health care system with the legislature. His plan sought to guarantee that nearly every resident of the state would have health insurance. Medicaid became the foundation for the new system which also included the creation of insurance exchanges, a mandate that Massachusetts residents obtain coverage, and tax subsidies to facilitate their purchase of insurance. Federal administrators approved the greatly transformed waiver in 2006. The continuation of the Massachusetts plan for near universal coverage depends on the federal executive branch periodically renewing the state’s Medicaid waiver.
Targeted waivers for home and community-based services
A second kind of more targeted waiver has also profoundly shaped Medicaid. These waivers differ from the comprehensive demonstrations in that they apply to a narrower scope of program activities. Nonetheless, their cumulative impact has been substantial and in some states transformative. From its inception, Medicaid had an “institutional bias” in the delivery of long-term care to the elderly and people under 65 with disabilities. These cohorts tended to receive care in nursing homes or other large institutions. To be sure, Medicaid law gave states the option to provide such care in the home and the community. But few states chose to do so. In order to promote a rebalancing of Medicaid long-term care toward home and community-based services (HCBS), Congress authorized “Section 1915c” waivers in 1981. When approved by federal administrators, these waivers would allow states to circumvent key requirements of Medicaid law. They could, for instance, limit HCBS to enrollees living in specific geographic areas rather than face the customary Medicaid obligation to provide services statewide. The waivers also allowed states to dodge Medicaid’s customary requirement that all individuals who met certain eligibility criteria (e.g., certain income levels, a certain degree of physical or intellectual disability) were entitled to benefits. Instead waivers permitted states to cap the number of participants and establish waiting lists for Medicaid HCBS.
Granted the opportunity to customize HCBS to their particular circumstances in these and other ways, states moved apace to apply for these waivers. During the 1980s, however, the federal bureaucracy took a guarded approach to approving state proposals paying particular attention to whether they were “cost neutral.” During the Clinton and G. W. Bush administrations, however, federal officials relaxed these standards and it became easier for states to gain approval for HCBS waivers. By the time the Obama administration took office, the number of 1915c waivers had mushroomed to 280. All but two states were providing some HCBS through these waivers. (Arizona and Vermont had no 1915c initiatives because they relied on demonstration waivers to support HCBS.) Medicaid outlays through HCBS waivers had risen to over US$27 billion, more than 10 times the amount in 1992. So too, some 1.2 million Medicaid enrollees received HCBS under the banner of 1915c waivers, over 5 times the number in 1992. Forty-two percent of all Medicaid enrollees receiving HCBS did so via waivers rather than through the regular Medicaid program. Even more dramatic, about two thirds of all Medicaid spending on HCBS occurred through waivers with 21 states spending more than 90% of their HCBS Medicaid outlays this way. 14
The administrative discretion wielded through Medicaid waiver processes greatly exceeds that generally associated with the picket-fence metaphor. But it does not automatically undermine the metaphor’s emphasis on the role of a relatively autonomous professional bureaucratic complex in shaping these waivers. Conceivably, career civil servants at the federal and state levels could be front and center in fueling the outpouring of waivers. As the next two sections of this article demonstrate, however, the Medicaid experience strongly supports the EF thesis on the importance of political executives in grant programs.
Waivers and the Administrative Presidency
Students of American governance have over the last three decades stressed the importance of the “administrative presidency’’—the ability of presidents to shape who gets what from federal programs through executive action rather than the legislative process. In this regard, early studies zeroed in on such factors as the president’s appointment powers, influence over the budget, capacity to forge reorganizations, and ability to shape the rulemaking process (Durant, 1992; Nathan, 1983; Waterman, 1989). More recent analyses suggest that direct presidential action—a kind of “power without persuasion”—has come to “constitute the distinguishing mark of the modern presidency.” 15 These works stress that presidents increasingly take unilateral action through executive orders, proclamations, directives, signing statements, or other written guidance. Research rooted in this “unilateral-action” perspective has, however, paid little attention to federal grant programs and (as near as I can find) completely neglected waivers as a tool of presidential leadership. The role of presidents in the Medicaid waiver process suggests this is a serious omission.
The White House Role in Fueling Medicaid Waivers
Both the Clinton and G. W. Bush administrations strove to use waivers as vehicles for achieving their policy and political goals. President Clinton’s impact was particularly great. During his terms as Governor of Arkansas, Clinton had chafed over federal Medicaid mandates and his inability to shape the program. Within a month after taking office in January 1993, President Clinton appeared before the winter meeting of the National Governors Association (NGA) to promise that states would have greater discretion to launch new approaches to Medicaid via waivers. Soon thereafter, the president ordered federal administrators to work with the NGA to streamline the processes for dealing with state waiver requests. By 1994, federal officials in the Health Care Financing Administration (HCFA; 1994) promulgated a new set of regulations designed to speed up the review of state waiver requests. Among other things, they promised to establish a well-defined schedule for reaching a decision on waiver proposals and to expand preapplication consultation and technical assistance.
The Clinton administration also took two other significant steps to make it easier for states to win approval for demonstration waivers. First, it eased pressures on states to prove that their waivers were “budget neutral.” Concerned that federal administrators and the states might use demonstrations to drain the national treasury, the Office of Management and Budget (OMB) had imposed this requirement in 1983 (Anderson, 1994, p. 227). This meant that a demonstration could cost the federal government no more over its duration than if the state had continued to operate its existing Medicaid program. OMB imposed a stringent methodology on states for estimating cost-neutrality that made it difficult for states to win waiver approval. In 1994, the Clinton administration announced that it would ease OMB’s strict approach. Among other things, it pledged to be more open to state ideas on how to calculate future Medicaid spending under the waiver. Within a year, the new federal approach had sparked criticism from fiscal watchdogs who thought that HCFA had interpreted budget neutrality far too loosely (Pear, 1995).
Second, the Clinton administration relaxed requirements that the approved demonstrations be rigorously evaluated. Prior to his administration, federal administrators had typically worked with states to plan careful assessments of the demonstrations using state-of-the-art social science methods. They perceived the waivers as vehicles for policy learning in a formal sense. Eager to allow states more discretion to shape their programs, the Clinton administration signaled that it would be more flexible and less exacting in reviewing the research protocols states submitted as part of their waiver proposals. It backed away from universally requiring an independent evaluator for each waiver and increasingly relied on state self-evaluations instead. While some states took this obligation seriously, these self-assessments often failed to cast much light on the efficacy of the demonstrations (Thompson & Burke, 2007, p. 984).
The G. W. Bush administration continued to promote waivers. After changing HCFA’s name to the Centers for Medicare and Medicaid Services (CMS), it continued to interpret the budget neutrality of waiver proposals so flexibly as to draw criticism from General Accounting Office (GAO; 2004). It also broadened the substantive scope of waivers that it would approve. 16 States had even greater prospects of obtaining waivers that departed appreciably from statutory requirements. So too, the G. W. Bush administration downplayed the research component of the demonstration waivers even more than its predecessor. CMS further reduced funding for evaluation and placed little pressure on the states to document their research plans in their waiver proposals.
The Clinton and G. W. Bush administrations also worked to make it easier for states to obtain Section 1915c waivers to foster home and community-based services. As noted earlier, Congress had authorized the HCBS waivers in 1981. But while significant numbers of states had obtained the waivers, their requests often precipitated intense and protracted negotiations with federal administrators. For instance, it took Minnesota officials 4 years to obtain a federal sign-off on an HCBS waiver proposal, and Texas administrators 3 years. Delay tended to spring from differences between national and state administrators over how to estimate the costs of the waivers. The law had required the HCBS waivers to be cost neutral. This generally meant that the average cost per enrollee receiving HCBS under the waiver had to be less than or equal to the projected spending for these enrollees in a nursing home or other institution. But as with the Medicaid demonstrations, OMB intervened to put further pressure on states to reduce costs. In the mid-1980s, it imposed the “cold-bed rule” on states seeking waivers. This rule required a state to demonstrate that for each HCBS waiver participant, it had emptied a bed in a nursing home or other institution.
As with the demonstration waivers, the Clinton administration made it easier for states to meet the test of budget neutrality in their waiver proposals. In this regard, it ordered OMB to desist from applying the cold-bed rule. The G. W. Bush administration also proved receptive to the HCBS waivers. Among other things, it expanded the scope of Medicaid activities a waiver could authorize. For instance, it permitted states to use Medicaid monies to pay security deposits on apartments and fees to hook up utilities for enrollees transitioning from institutions to the community.
Waivers as Leadership Tools
The case of Medicaid highlights the degree to which waivers can appeal to presidents and top political executives as tools to reshape domestic grant programs. Waivers possess distinctive features as instruments of presidential leadership. Their explicit authorization by Congress in federal statutes makes waivers harder to challenge in court than other presidential directives. In contrast, for instance, the G. W. Bush administration issued a directive to state Medicaid administrators that would have greatly restricted their ability to cover more low-income children. Several states resisted in court 17 and GAO advised Congress that the president had failed to observe the requirements of the Administrative Procedures Act. Faced with this resistance, the G. W. Bush administration abandoned the initiative. Waivers reflect an alliance between the national government and a particular state. Unlike many other unilateral actions, this means that the president has a willing partner as an administrative agent from the outset.
Ultimately, of course, it takes two to tango. Presidential administrations can use waivers to invite states to dance. But states decide whether to accept. In the case of Medicaid, many states were willing partners especially for waivers that expanded coverage to new enrollees and enlarged the share of long-term care provided outside of institutions. The Clinton administration realized many of its policy preferences in this way. In the case of the Bush White House, the expansion of HCBS via waivers fit squarely with its priorities.
But on other fronts, the G. W. Bush administration did not have as much success in using waivers to promote its vision. To a greater degree than under Clinton, top Bush appointees signaled their willingness to approve demonstration waivers that would reduce Medicaid spending by paring services, constraining eligibility, and imposing market-based approaches on enrollees (e.g., more cost sharing). But few states applied for these waivers. Instead, most sought to renew or modify existing waivers to expand Medicaid coverage. The experience of the G. W. Bush administration may seem to make waivers a less useful tool of unilateral presidential action relative to executive orders and proclamations in other contexts. But this difference can be overdrawn. In his study of presidential “power without persuasion,” Howell (2003, p. 22) notes with respect to executive orders and proclamations: “when it comes to (their) implementation . . . the power modern presidents wield very much depends on their ability to persuade.” These directives do not implement themselves but depend on the cooperation and capacity of administrative agents. Much the same applies to the federal executive branch in its dealing with states over Medicaid waivers. Waivers tend to be more valuable to the White House when goal congruence exists between a presidential administration and state policymakers, or states are at least open to federal persuasion.
To underscore the role of the administrative presidency in shaping waivers is not to gainsay that Congress, interest groups, and career civil servants also play a role. 18 Members of Congress have at times made their views known on the waiver requests coming from their states urging the executive branch to approve, deny or modify them. More broadly, Congress and its analytic agencies (such as the GAO) have devoted considerable oversight to waivers. On rare occasions Congress has altered statutes to constrain executive branch discretion in granting certain Medicaid waivers. So too, interest groups from the states affected by the waiver frequently press their case with federal officials. The expertise and views of career civil servants factor into the waiver process as well. Still the large role for political executives and the vast administrative discretion exercised via waivers depart sharply from the picket-fence model and IGM.
Is Medicaid unique among grant programs in the degree to which waivers are a central tool of the administrative presidency? Only additional research can fully address this issue. Clearly, however, other cases of presidential reliance on waivers exist. In February 2012, for instance, President Obama held a press conference attended by educators and key lawmakers to announce that 10 states had obtained waivers from many of the provisions of No Child Left Behind. He presented the action as flowing from congressional inaction in reauthorizing and reforming the program. Obama averred that the waivers would enable states to meet educational goals that his administration had enunciated since taking office (Hu, 2012). By late May, Secretary of Education Arne Duncan announced the approval of eight more waivers (Perez-Pena, 2012).
Waivers and an “Administrative Governorship”?
Medicaid waivers have, then, emerged as a valued tool of the administrative presidency. But does the same apply at the state level? Do governors and top political executives employ waivers to reshape their Medicaid programs? Any effort to address these questions must confront the reality that Medicaid policy and administrative processes vary appreciably from one state to the next. My research on Medicaid, however, suggests the following: (1) Governors have often used waivers to transform their Medicaid programs as the EF model contends. (2) State legislatures with some frequency play a more significant role in shaping waivers than Congress does at the national level. This suggests a need for adjusting the EF framework. (3) The professional bureaucratic complex plays a larger role in shaping Medicaid HCBS waivers than the more encompassing demonstrations.
Governors and medicaid waivers
Unfortunately, inquiry in public administration and political science has been virtually silent on the concept of an “administrative governorship”—how governors use administrative tools to promote their policy agendas without changes in law. To be sure, many studies have examined the institutional strength of the governor’s office, including his or her authority to appoint top political executives. 19 More recently, gubernatorial strategies in issuing executive orders to shape the administration of state programs has garnered attention (Ferguson & Bowling, 2008). So too, a few studies have illuminated how governors and their top appointees lead or manage state bureaucracies (e.g., Abney & Lauth, 1986; Behn, 1991; Weinberg, 1977). But the literature on leadership by governors via administrative means pales in significance when compared to the extensive, nuanced research focused on the “administrative presidency.” More specifically, the role of governors and their top political executives in conceiving, incubating, hatching and ultimately implementing waivers awaits systematic scrutiny.
The case of Medicaid, however, suggests that waivers loom large among the administrative tools governors employ to shape grant programs. An initial assessment of demonstration waivers approved during the 1993-2010 period found that substantial, visible gubernatorial involvement occurred in nearly 20 states, including some of the most populous ones. Medicaid demonstrations have often been the highly publicized signature initiatives of governors. For instance, Governors Jeb Bush (R) in Florida, John Kitzhaber in Oregon (D), Ned McWherter (D) in Tennessee, George Pataki (R) in New York, Mitt Romney (R) in Massachusetts, and Christine Todd Whitman (R) in New Jersey all initiated major changes in their Medicaid programs through Section 1115 waivers. It also deserves note that the absence of gubernatorial support for a demonstration waiver almost certainly spells its demise.
In contrast to the demonstrations, the more targeted HCBS waivers are seldom the highly visible signature initiatives of governors. But it would be a mistake to overlook the role governors play with respect to HCBS waivers. While usually not as involved with specific waivers, many governors have sent strong signals that they sympathize with HCBS and the move to deinstitutionalize Medicaid enrollees. The National Governors Association has supported HCBS and promulgated best-practice models.
State legislatures and medicaid waivers
While governors and the executive branch possess vast administration discretion to shape waivers, state legislatures should not be portrayed as uniformly quiescent. Legislatures vary greatly in their general ability to control executive branch discretion. All else being equal, legislatures with higher salaries, larger staffs, longer legislative sessions, and no term limits stand a better chance of holding their own vis-à-vis the executive branch (Kousser, 2005; Huber & Shipan, 2002). Various rankings suggest that ten legislatures (primarily though not exclusively in the most populous states) achieve high levels of professionalism and that 12 rank low. 20 Other evidence suggests that state legislatures with at least a moderate level of professionalism have the potential to impact waiver processes meaningfully (Edwards, Garcia, Lashbrook, & Flowers, 2007).
At times, legislatures have acted to shape waivers. On rare occasions, such as in Minnesota, these legislatures have become pivotal players in initiating waiver proposals that the executive branch has then formally proposed to the national government. 21 More commonly, legislatures have responded to major demonstration proposals from the executive branch by negotiating concessions. In this regard, state legal arrangements matter. In New Jersey, for instance, the executive branch proposes and implements waivers without obtaining a sign-off from the legislature. In other states, however, both houses of the legislature must approve major demonstrations. This is an institutional advantage that Congress does not have in the waiver process at the national level. The federal bureaucracy does not need to obtain authorization from Congress to approve a waiver request. State legislatures have at times used their sign-off authority on major waivers to extract concessions from the executive branch. When Governor Jeb Bush of Florida proposed a comprehensive waiver in the early 2000s, for instance, the state’s house of representatives largely deferred to his preferences. But leaders in the state senate greeted his waiver proposal with skepticism and negotiated many modifications before finally approving it. Governor Mitt Romney’s much publicized waiver in Massachusetts also required considerable bargaining with legislative leaders to obtain their approval. The level of legislative involvement in waiver processes in all 50 states is unclear. But a study of eight states found substantial legislative engagement in three states, occasional involvement in three, and a limited role in two (Edwards et al., 2007, p. 12, 21).
The EF model should more fully acknowledge that governors may well face greater legislative intrusion in the waiver process than presidents do. Still, the overall picture that emerges from states tends to affirm the role of governors and political executives in initiating and sculpting the contents of major demonstration waivers.
The professional bureaucratic complex and medicaid waivers
Compared to the comprehensive demonstrations, the professional bureaucratic complex tends to play a larger role in shaping the more targeted HCBS waivers. While cumulatively transforming the delivery of long-term care in many states, the individual HCBS waivers tend to be narrower in scope. Specialized program professionals in state agencies do much to influence the specifics of the HCBS waivers. In doing so, they typically work with a bevy of advocacy groups and service providers for the elderly and people with disabilities. The exact relationship between different interest groups and state agencies in forging HCBS waivers varies from state to state (Castellani, 2005). At times entrepreneurial professionals in the Medicaid bureaucracy lead in persuading stakeholders outside government that certain waiver proposals have merit. In other cases, states feature a more client-driven pattern which advocates and providers more markedly shape the waivers; state administrators serve more as representatives than their partners. Since HCBS waivers frequently apply to some areas of the state and not to others, legislators from the affected regions may also voice their views to administrators.
Career professionals at the federal and state level tend to dominate intergovernmental interaction over the HCBS waivers. Over time, these waivers have increasingly come to embody a kind of “intergovernmental licensing” (Weissert & Weissert, 2008). While some negotiation between national and state officials occurs, much interaction is routine. State administrators present requested information on a boilerplate application form. Federal officials check to make sure the application is complete and conforms to program requirements. Once approved, federal administrators engage in little monitoring and almost automatically renew the waivers. Occasionally, national and state officials have sufficient differences to prompt negotiation and bargaining.
The dynamics embedded in the HCBS waiver processes fit more easily with the picket-fence metaphor and IGM. The changes promoted by a single waiver tend to be more incremental and political executives less involved in molding their specifics. Still, the HCBS process departs from the picket-fence model in significant ways. Of particular importance, these waivers do not represent the triumph of “autonomous bureaucracies” over governors frustrated by their inability to control them. As noted previously, governors have tended to support the initiative to shift Medicaid long-term care from nursing homes and large state institutions to a home and community setting. Moreover, governors have typically favored waivers rather than regular Medicaid channels to promote this end. This is because waivers afford them greater flexibility and budget control (e.g., the ability to cap enrollment if demand for HCBS soars). In sum, the alliance between the bureaucracy and interest groups to promote HCBS waivers usually proceeds under the watchful eye of the state budget office and with the approbation of the governor.
The degree to which Medicaid waiver processes at the state level differ from those in other policy arenas awaits systematic inquiry. Clearly, however, strong gubernatorial leadership via the administrative process has at times emerged in other arenas. For instance, in the 1990s several governors from the Midwest employed waivers to pursue major welfare reforms (Weissert, 2000). Governors and their commissioners have also been front and center in calling for education reform and seeking waivers from No Child Left Behind.
Toward a Revised Perspective
In sum, evidence from Medicaid provides substantial support for the EF model. This model points to the limits of the picket fence and IGM in three primary ways. First, EF more explicitly recognizes the asymmetrical character of the federal grant system and follows the money. As of the early 2000s, the 20 largest grant programs distributed nearly 80 percent of all grant monies to states and localities (Posner, 2003, p. 5). 22 EF tends to extract its lessons from these programs, which usually attract attention from a potpourri of stakeholders primarily because of their budget implications. 23 These larger grants emerge less as pickets than as major slats in a fence or, in the case of Medicaid, a pillar. Second, EF underscores the vast scope of administrative discretion associated with waivers and the program transformations they can galvanize. Many states used demonstrations to transform their Medicaid programs in relatively short periods of time. While narrower in scope, the great numbers of HCBS waivers have over a five to ten year period markedly changed long-term care in many states. The change facilitated by EF tends to be greater than the incremental adjustment associated with the picket fence and IGM.
Third, EF features a much higher quotient of “the political.” To be sure, career administrators continue to play a significant role. They provide expertise and advice as their political superiors pursue new initiatives via waivers and other means. They stand front and center in implementing the details of these initiatives. Their acumen in handling the everyday management of waivers matters greatly. Still, the rise of EF means that top elected executives and their political appointees loom much larger in shaping federal grant programs relative to the professional bureaucratic complex.
It deserves note that EF’s emphasis on political leadership squares with other empirical assessments of trends within the executive branch at the federal and state levels. Metaphors possess normative implications. They point to deficiencies in governance and frequently suggest remedies (Stewart, 1984, p. 18). Writing over three decades ago, Wright (1978) captured the prescriptive thrust of the picket-fence model (p. 211). In his words, the model “dramatizes the need for powerful assertions of gubernatorial influence if state policies and programs are to be more than a hodgepodge of independent and functional fiefdoms.” Since Wright made this observation, the institutional powers of governors have grown appreciably (Beyle, 2006). Among other things, governors have gained enhanced control over the budget process, greater authority to veto legislation, and the opportunity to serve consecutive terms. Of particular importance, states have boosted gubernatorial leverage over personnel processes. The number of appointments governors can make to the top echelons of state bureaucracies has increased. The reinvention movement of the 1990s led many states to scale back the personnel system protections that had helped empower career civil servants in their dealings with political executives. Georgia went so far as to abolish its merit system (Hays & Sowa, 2006; Nigro & Kellough, 2008).
So too survey data from the federal bureaucracy point to a more commanding role for political executives. From the early 1990s to 2005, an appreciable decline occurred in the number of senior career officials who felt they exerted “a great deal of influence” over policy in their agencies. These officials also reported fewer contacts with interest group representatives and with key congressional personnel. Their “insulation” from these contacts may well indicate that presidential appointees have increased their control over administration relative to the career civil service (Aberbach, 2007, p. 183-185). The degree to which these survey findings hold among career Medicaid officials remains an open question. But the findings are at least consistent with the strong role of the administrative presidency in shaping Medicaid waivers.
In noting the rise of executive federalism, I do not mean to gainsay the continuing relevance of the picket fence and IGM for understanding federal grants. These lenses may be especially pertinent in illuminating the dynamics embedded in the plethora of smaller grant programs. As of the early 2000s, over 600 grant programs annually allocated US$50 million or less each. While these programs accounted for only 6% of all federal grants to states and localities, they present significant public administration issues that deserve scholarly attention (Posner, 2003, p. 5). The picket fence and IGM may also be germane in the context of larger grants. As noted previously, for instance, they apply in part to the dynamics evident in HCBS waiver processes. Only additional research can uncover the degree to which the larger grant programs testify to the rise of the EF model relative to the picket fence, IGM, or some alternative.
In pursuing this research agenda, the available evidence suggests that EF may be less manifest in the case of certain developmental grants 24 that fund physical or social infrastructure under the banner of economic development. Members of Congress have special interest in these grants because they provide abundant opportunities for claiming political credit. For instance, a recent analysis points to an important congressional role in allocating federal highway demonstration grants among the states (Gamkhar & Hamid, 2008). Congress can limit administrative discretion through statutes compelling states to target projects in certain ways. Or congressional influence may surface during implementation whereby administrators eager to rally support for their programs disproportionately allocate projects to the districts or states of those members of Congress who sit on powerful committees or engage in more oversight of agency actions (e.g., Arnold, 1979; Duffin, 1999). Future research should appraise whether EF primarily flourishes in formula-based, redistributive grant programs, like Medicaid, that provide income, services or other sustenance (e.g., food, housing) to the disadvantaged.
To the degree that subsequent research affirms the rise of EF, it will become important to hone the conceptual underpinnings of public administration in the context of federalism. EF fits someplace between intergovernmental relations and the current constrained view of IGM. Intergovernmental relations speak to a sphere of interaction between federal and state officials that goes well beyond the boundaries of EF. In turn, IGM with its emphasis on incremental adjustment in relatively apolitical contexts falls short of capturing pivotal developments in the governance and administration of grants. EF therefore suggests the need for a more refined typology (or metaphor) under the general umbrella of “administering federalism.” 25 We need a vocabulary to describe a capacious IGM that more fully embraces the interplay between politics and management as well as the potential for accomplishing large program changes via the administrative process.
Executive Federalism and American Democracy
The rise of executive federalism not only intersects with efforts to hone an empirical theory of grant programs, it also raises significant normative issues concerning public administration and democracy. For some, EF would seem to be yet one more step toward the erosion of a rule of law and the checks and balances built into the American political system. Luton (2009, p. 573), for instance, calls for a rejection of the administrative presidency strategy because it leads to excessively ambitious efforts to reinterpret statutes (p. 573; see also Lowi, 1969, 1987). Focusing more specifically on waivers, Kincaid (2001, p. 22) echoes these concerns and goes on to argue that waivers “pose issues of democratic accountability insofar as they are negotiated and implemented by executive officials outside of the floodlit legislative process. They raise questions of equity, as well, insofar as they introduce variability in the implementation of law and, thus, equal-protection concerns. . .”
In a related vein, EF in general and waivers in particular have kindled concern that public administration has become excessively “politicized” (e.g., Kincaid, 2001, p. 22). The thrust of the picket-fence model was that a professional bureaucratic complex with career administrators at its core had too much autonomy. EF brings matters full circle. It raises the question of whether the career civil service has too little influence in grant processes—whether “neutrally competent” administrators committed to law and due process as well as efficiency and effectiveness have insufficient control over grant programs. 26 Politicization could play out as a willingness of the White House to approve waiver proposals from governors of the president’s political party independent of their substantive merits or their consistency with the goals of the authorizing statute. It turn, perceptions of such favoritism can yield complaints that waivers are the handmaiden of a “new” partisan, or “crony politics.” 27
These criticisms certainly deserve continued attention as executive federalism, and waivers in particular, come under scrutiny in diverse policy areas. For present purposes, however, two caveats deserve note. First, my research on Medicaid casts waivers in a more favorable light. Considerable evidence suggests that in the period from 1993-2010, waivers enhanced the performance of Medicaid. States seized the opportunity provided by waivers to informally test an array of policy hypotheses concerning how to achieve a better balance among three contending values—cost containment, access to services, and quality care. On balance, waivers played to state strengths as laboratories of democracy and facilitated policy learning. Nor did waivers become a covert vehicle for program retrenchment. 28 In sum, a strong case exists that Medicaid was in many states a more efficient and effective program because of waivers than it would have been otherwise.
My research on Medicaid has also illuminated some issues of democratic process with respect to EF. While the federal bureaucracy exercised vast discretion in approving Medicaid waivers from 1993 through 2010, this did not mean that Congress was completely shunted to the sidelines. As discussed earlier, Congress conducted considerable oversight of the waivers and occasionally intervened to constrain the executive branch. At neither the national nor the state level, was there much evidence of the executive branch proceeding with a waiver over the strong opposition of key members of the legislative branch. Nor did waiver processes egregiously thwart due process and transparency. To be sure, Medicaid waivers are not governed by the Administrative Procedures Act. Nonetheless, stakeholders typically have the opportunity to review and comment on the waiver proposals as they move from the state decision arena to review by the federal government. Waiver processes tend to be at least as transparent as those involved in the state plan amendment process that governs program changes under the regular Medicaid program. So too, waivers may be seen as reflecting federalism’s contribution to democracy. They fostered a closer match between the policy preferences of elected officials in particular states and the character of their Medicaid programs.
As for the politicization of waivers, the White House did at times intervene to expedite approval of a waiver request submitted by a governor from the president’s own political party. For instance, when the Centers for Medicare and Medicaid Services questioned certain aspects of a waiver proposal submitted by the administration of Governor Jeb Bush in Florida in the early 2000s, the White House stepped in to assure that it obtained rapid approval. 29 Presidents have strong incentives to build supportive relationship with governors of their own party by approving their waiver requests. But it deserves emphasis that states with Republican governors under Clinton and Democratic governors under Bush also succeeded in winning approval for their proposals. Moreover, my research uncovered no evidence of a governor from the president’s party proposing the waiver equivalent of a “bridge to nowhere.” States differed in their orientations and approaches to Medicaid and their waivers reflected that. But the financial stakes and general importance of the program led governors of both parties to shun approaches that overtly represented the triumph of waivers as “spoils.” Waivers almost always reflected a serious effort to achieve goals related to cost containment, enhanced access or improved quality.
This relatively sanguine assessment of EF and waivers may not apply to other grant programs. It also may not be germane for Medicaid in the future if political dynamics change. For instance, a new presidential administration committed to retrenching Medicaid might pursue waiver practices that slammed the door on state innovations to expand insurance coverage. It might approve state waiver requests to undercut the program that radically departed from the thrust of Medicaid law. Whatever the future holds, my research on Medicaid does underscore the need to assess the degree of “threat” to democratic governance posed by waivers on a program by program basis over a specific period of time.
A second caveat also applies to considerations of EF and democratic process. Waivers must be assessed empirically and normatively in the context of broader forces that vitiate policymaking processes in the United States. Well-functioning democracies possess an array of attributes. One of them is the ability of legislative bodies to aggregate interests effectively—to forge theoretically plausible policy that establishes priorities among competing demands and responds to pressing problems in a timely way (Weaver & Rockman, 1993, p. 10). The growing partisan polarization in the United States along with the propensity of elections to yield divided government has undercut this capacity (Abramowitz, 2010). At the national level, gridlock in Congress has weakened prospects for timely and effective policymaking. This development has kindled concern among political scientists that Congress is “broken” (e.g., Mann & Ornstein, 2006). State legislatures have also manifested increased partisan polarization. While states vary in this regard, evidence from one study indicates that 15 state legislatures are even more polarized than Congress (Shor & McCarty, 2011, p. 546). Given the high public expectations of presidents and governors, it is hardly surprising that they are tempted to respond to legislative sluggishness or gridlock through executive action. 30
It is, of course, one thing to note forces that encourage waivers and unilateral executive action, and another to condone this development normatively. Still, assessments of EF from the perspective of democratic governance cannot sensibly ignore a core question: To what degree can waivers and EF be seen as a legitimate mechanism for coping with dysfunction in the legislative branch? Addressing this question will afford a more nuanced argument about the role of checks and balances and the proper scope of administrative discretion in American democracy. It will intersect with the enduring issue of tradeoffs between performance and accountability in American public administration.
Future Research
This article has taken an initial step in documenting the rise of executive federalism. In so doing, it has called into question the picket-fence metaphor and the conceptual foundations of IGM as applied to federal grants. But much research awaits completion to determine the degree to which EF captures the dynamics of federal grant programs. In this regard, four foci seem particularly promising.
First, we should monitor the contours of the federal grant structure and target the largest grants for systematic comparative analysis. In its purest form, the picket fence summons up images of vast numbers of small grant programs roughly equal in size. As noted previously, however, asymmetry marks the federal grant structure with the 20 largest grant programs (2% of the total number) allocating nearly 80% of the monies to states and localities (Posner, 2003, p. 5). Given the substantial sums of money involved, EF seems most likely to manifest itself in these larger grant programs. Hence, comparative analysis of these substantial grants drawing on qualitative and quantitative methods constitutes one important test of the pervasiveness of the EF model. It also directs students of public administration to focus on the most “important” grant programs—those that meaningfully impact the most people.
Second, this comparison should spotlight the degree to which these larger grant programs feature delegation of authority to the executive branch. EF stresses the role of substantial administrative discretion in shaping grant programs. Hence, analysis of how specific, or precise, federal statutes are in prescribing the goals and means of grant programs becomes important. How much authority does Congress delegate to federal administrators to determine how the grants will be allocated among states for specific activities? What menu of options does the federal statute explicitly afford states? These and related questions deserve attention as does the role of state legislatures. How active are state legislatures in passing laws that prescribe how administrators will implement a particular federal grant? Insightful research in this tradition already exists. For instance, Huber and Shipan (2002) focus on the “institutional foundations of bureaucratic autonomy” by analyzing the precision of state statutes shaping Medicaid managed care. They found considerable variation with legislatures in states featuring divided partisan control of the branches of government more likely to pass precise statutes. 31
In focusing on legislative delegation, a census of waiver authority in the 20 largest federal grant programs should be conducted and the scope of action permitted by these waivers assayed. Examination of waiver authority should not end with the federal executive branch. As noted earlier, states vary considerable in the degree to which they assign legislatures a meaningful role in waiver processes. Some states require legislative approval of certain waivers while others do not.
Third, the comparison should target how aggressively, or expansively, administrators exercise discretion in grant programs, especially via waivers. In implementing grants, administrators at times play by the book. They interpret the statute narrowly paying homage to “legislative intent.” Alternatively, administrators may push the envelope pursuing activities at the outer limits of what the statute may permit. Expansiveness in exerting executive branch prerogatives can manifest itself in multiple ways; the propensity to use waivers is a particularly compelling indicator. The presence of formal waiver authority does not mean that officials will use it. The federal bureaucracy had the authority to encourage comprehensive demonstrations since the birth of Medicaid in 1965. It took three decades, however, before the Clinton administration opened the flood gates to state requests to refashion their Medicaid programs via this tool. In appraising the level of waiver activism in grant programs, several questions loom large. How many waivers have states obtained? What is the scope of program activities the waivers cover and how much do they depart from the program authorized by the statute? What proportion of grant funds get spent through waivers relative to regular program channels? What percentage of enrollees receives services under waivers? Addressing these and related questions can do much to illuminate the degree and nature of executive branch assertiveness in grant programs.
Fourth, the comparison of the 20 largest federal grant programs should probe the relative influence of political executives and the professional bureaucratic complex in molding administrative discretion. EF rests on the idea that presidents, governors, and other political executives will be at center stage—that grant programs will at times provide vivid evidence of the “administrative presidency” or “administrative governorship” at work. To what degree does evidence from the largest grant programs mirror Medicaid in supporting this proposition? Or, alternatively, do they point to the dominance of Beer’s (1978) complex of strong public agencies, interest groups, and key members of the legislative branch? Going further, are there cases that fit uneasily with both the picket fence and EF? For instance, do some of the larger grants feature legislative bodies in the driver’s seat vigilantly monitoring and influencing the exercise of administrative discretion? Once the patterns of influence embedded in each of the largest grants are mapped, the following question naturally emerges: What accounts for variation in these patterns?
In addressing these questions, possible variations on the EF theme should be considered. Full-blown EF would apply if vast administrative discretion mingled with strong political leadership at both the federal and state levels in a grant program. Federally-driven EF would be present when the administrative presidency played a large role in grant administration (primarily through waivers) but the professional bureaucratic complex dominated processes at the state level. In turn, state-driven EF would be manifest when governors and political executives strongly influenced waiver proposals and related program decisions while interacting with a professional bureaucratic complex in the nation’s capital. Still other variations may also be pertinent as we proceed with the important work of testing still further the potential and limits of EF, IGM, and the picket-fence model.
Footnotes
Acknowledgements
The author thanks John Kincaid, Carl Stenberg, and the editors of ARPA for helpful comments on an earlier draft of this article.
Declaration of Conflicting Interests
The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The Investigator Award Program of the Robert Wood Johnson Foundation provided financial support for the author’s study of Medicaid.