From employment optional to “Employment First”: Explaining two cases of state-level disability policy change

Abstract

Over the last three decades, the United States has increasingly devolved social policy decisions from the federal to the state level. These changes have resulted in substantial variation in policy decisions and related outcomes. Just as the changes allow states to act as policy laboratories, they also offer a window into the process by which organized interests take advantage of such opportunities to influence state-level policy. This study uses the Advocacy Coalition Framework to illuminate the black box of policy change with a comparative study of two states, Washington and Pennsylvania, which adopted “Employment First” policy aimed at prioritizing employment services for individuals with disability. The study reveals that policy change in both states was associated with organized stakeholder mobilization, strategic framing and narrative, and bureaucratic activism, all in an environment of heightened stakeholder attention to the issue. That said, the two states followed distinct paths, with early policy change in Washington stemming from service provider mobilization, suggesting the importance of policy feedback mechanisms.

Keywords

Advocacy Coalition Framework bureaucracy disability policy federalism policy process

Introduction

Over the last three decades, the United States has increasingly devolved social policy decisions from the federal to the state level. This shift has resulted in increased use of federal waiver programs and block grants, and substantial variation in policy adoption decisions and related outcomes (Bruch et al., 2018; Bunch et al., 2018; Dilger and Boyd, 2014; Guzman et al., 2013; Thompson and Burke, 2007). Just as these changes allow states to act as policy laboratories, they also offer a window into the process by which organized interests take advantage of such opportunities to influence state-level policy. Several policy process frameworks (e.g. Advocacy Coalition Framework (ACF); Innovation and Diffusion models; Policy Feedback Theory (PFT)) and theories of bureaucracy (e.g. iron triangles, regulatory capture, bureaucratic activism, deliberate discretion, and contracting) attend to the influence of interest groups in state-level policy change. Some of these traditions are more focused on exposing the black box of interactions and conditions that lead to policy change than others. The ACF, in particular, has substantial potential to illuminate the black box of policy change by taking a causal-mechanistic look (van der Heijden et al., 2019) at the secondary components of policy change (Pierce et al., 2017). In response, this study applies the ACF in a comparative study of one policy domain, drawing on other theoretical insights to further illuminate pathways to policy change. Specifically, this study examines the factors leading to the adoption of state-level “Employment First” policy changes in Medicaid-funded Day Habilitation and Employment (DH&E) services and supports for individuals with Intellectual and Developmental Disabilities (I/DD).

Medicaid-funded DH&E services and supports have grown exponentially since the introduction of Medicaid Home- and Community-based Services (HCBS) waivers in the early 1980s. Medicaid HCBS waivers were introduced to encourage and support deinstitutionalization of the I/DD population (Agranoff, 2013; Gettings, 2011). DH&E services and supports are intended to enable adults with I/DD to successfully live outside of institutions by supporting participation in a variety of community programs and settings. Such programs include day habilitation facilities, facility-based employment (i.e. sheltered workshops), sub-minimum wage jobs, and competitive, integrated employment (CIE). Since 2000, over 30 states have adopted policy prioritizing CIE¹ outcomes among the population of working-age adults with I/DD served by DH&E systems. These policy changes, which are often labeled as “Employment First” policy (Nord et al., 2013; Racino, 2015), vary considerably in terms of timing, content and venue, and are closely associated with interest group activity (Giordono, 2019; Racino, 2015). These observations give rise to the primary research questions: (1) What does state-level CIE-focused policy change look like? (2) Why and how did CIE-oriented policy change occur?

Among various policy process frameworks, the ACF is particularly compatible with these research questions. The ACF is predicated on the idea that organized interests (in the form of advocacy coalitions) compete to achieve policy change (Jenkins-Smith et al., 2017; Sabatier and Jenkins-Smith, 1993). Exploration of the ACF’s secondary components (Pierce et al., 2017) and related causal mechanisms (van der Heijden et al., 2019) remains important lines of ACF-based inquiry (Jenkins-Smith et al., 2017). Moreover, Giordono (2019) finds evidence of at least two competing advocacy coalitions, “Employment First” and “Choice,” in the DH&E Services subsystem, using the ACF.

Other frameworks also offer well-established windows into policy change, such as Innovation and Diffusion models (Berry and Berry, 1990, 2017; Rogers, 1995; Shipan and Volden, 2008, 2012) and Policy Feedback Theory (Béland and Schlager, 2019; Mettler and SoRelle, 2014; Pierson, 1993; Skocpol, 1995). Relatedly, the few disability-focused policy process studies suggest the importance of political resources and strategies (e.g. Itkonen, 2007, 2009; Jeon and Haider-Markel, 2001; Nagel, 2006; Pettinicchio, 2013; Shapiro, 1994). Theories of bureaucracy, including theories of bureaucratic activism (Eaton and Weir, 2015; Pacewicz, 2018), deliberate discretion (Huber and Shipan, 2002), interest group influence (Adams, 1981; Krause, 1999; Stigler, 1971), and competing theories of contracting (e.g., Bertelli and Smith, 2010; Davis et al., 1997; Jensen and Meckling, 1976; Van Slyke, 2007; Walsh and Seward, 1990), also offer insight into influential actors and policy change mechanisms, from an administrative perspective. Finally, disability researchers have extensively studied systems-level conditions that yield changes to individual employment outcomes in this specific policy domain (Butterworth et al., 2017; Hall et al., 2007). This study argues that accommodating relevant insights from those literatures under the umbrella of the ACF provides a more comprehensive strategy for addressing the research questions of interest.

Using a comparative research design, this study examines policy changes observed in two states, Washington and Pennsylvania, to address the research questions of interest. The application ultimately draws heavily on the ACF, leveraging relevant expectations and evidence from these other theoretical traditions. Study results reveal a long-term policy change process considerably more complex than the binary, lightbulb-style policy adoption process often depicted by traditional frameworks. Instead, CIE-focused policy change in both states involved multiple attempts to change administrative rules and legislative statutes (some successful, some not). The findings also indicate that in both states, common conditions for policy change include organized mobilization of stakeholders, strategic framing and narrative, all of which occurred in an environment of heightened attention to the issue. In both cases, high levels of bureaucratic activism and discretion made a distinct difference to the timing and sequence of social policy decisions. The timing differences (early vs. late policy adoption), however, highlight the important role of service provider coalition membership and defection. Relatedly, the study findings suggest that feedback mechanisms influenced the timing of the policy adoption process, namely via antecedent CIE-related service conditions and service provider support for (or rejection of) CIE-oriented policy.

The study explicitly leverages state-level variation to highlight conditions under which policy change is likely to occur and provide insight into relevant causal mechanisms. Insights from the study are likely to have implications for public policy scholarship as well as for domain-specific policymakers and advocates. This study contributes to the set of ACF applications, which typically attend to policy change only within a single subsystem (Pierce et al., 2020), yielding less comparative information about how such interactions yield policy change in the context of a decentralized system. Moreover, the study offers an application of the ACF to “nontraditional settings” (Jenkins-Smith et al., 2017: 158), specifically one representing low salience conditions. Finally, the findings suggest potential directions for theoretical improvements and lines of inquiry.

From a policy perspective, this study’s results are likely to be of interest to policymakers and advocates with an interest in disability policy formulation. These policies admittedly focus on a relatively small target population, with around 610,000 individuals with I/DD receiving DH&E services and supports nationwide in 2015 (Butterworth et al., 2016). However, the shifts that have occurred in this policy community are aligned with international shifts in disability policy, which increasingly emphasize inclusion and independence (e.g. Aucante and Baudot, 2018; Edzes et al., 2013; Lindsay and Houston, 2013; Vornholt et al., 2018). Finally, these findings will likely prove relevant to other similar social policy domains that are subject to waiver- and grant-based systems, such as the systems of long-term supports for older adults and income assistance programs for low-income households.

Policy context

Day Habilitation and Employment services and supports are intended to enable adults with I/DD to successfully live and work in their communities. The interest in providing access to employment-related services for individuals with I/DD is not new. Sub-minimum wage work has been available since 1938 via the Fair Labor Standards Act. However, the development of sheltered workshops (i.e. sub-minimum wage work, typically in segregated settings) during the 1970s and 1980s provided publicly funded access to training and employment opportunities for adults with disability outside of the formal education system (Agranoff, 2013; Gettings, 2011; Wehmeyer, 2013).

DH&E services are managed at the state level, but are nested in the broader system of cooperative federalism via the system of Medicaid HCBS waivers. Prior to 1985, most individuals with I/DD receiving waiver-funded services were served in segregated, unpaid day habilitation settings (McGaughey and Mank, 2001). In the mid-1980s, supported employment services initially became available via Medicaid waivers. Such services included job coaching, specialized job training and customized supervision, intended to support individuals with disability in competitive employment positions in community (i.e. integrated) settings. These services were offered and promoted by both the federal and state governments, especially via HCBS waivers for long-term services and supports (Agranoff, 2013; Friedman, 2016; Friedman and Rizzolo, 2016; Gettings, 2011; LeBlanc et al., 2000). Medicaid long-term services waivers, the result of negotiations between the state and federal governments, are the primary source of funds for DH&E services (Agranoff, 2013; Gettings, 2011).

Alongside employment services, most states also offer a variety of DH&E services, including non-employment services (i.e. day habilitation or respite), as well as non-competitive employment-related supports (i.e. sub-minimum wage) and/or services in non-integrated settings (i.e. facility-based/sheltered workshop services) (Friedman, 2016; Friedman and Rizzolo, 2016). States vary substantially in the degree to which eligible individuals participate in integrated employment services, from about 1% to 87% (University of Massachusetts Boston, Institute on Community Inclusion, n.d.). However, only about 18% of DH&E participants received integrated employment services nationwide (Butterworth et al., 2016). State-level “Employment First” policy changes observed during the last two decades vary considerably by state. They take multiple forms, including legislative statutes, executive orders and administrative directives (LEAD Center, n.d.; Nord, 2014). They also include varying combinations of symbolic language, improved access to employment supports, and/or incentivization of employment services and restrictions on facility-based and day activity services (Butterworth et al., 2016; Nord, 2014).

There is considerable evidence that interest group activity is an important driver of Employment First policy change. Employment First has been described as a “grassroots movement with progressive rehabilitation professionals as its core” (Racino, 2015: 160) and as a national movement by the Department of Labor’s Office of Disability Policy (U.S. Department of Labor, n.d.a.). Moreover, Employment First serves as an organizing principle for professional association activities (e.g. Association of People for Supported Employment First) and related communities of practice (e.g. State Employment Leadership Network). These observations, as well as evidence of state-level advocacy coalitions (Giordono, 2019), suggest the presence of organized activity to promote policy change.

Theory: Drawing on diverse traditions

Given the dual focus on organized interests and policy change, the ACF is an appropriate scaffold for examining CIE-oriented policy change. However, in light of this study’s objective to go deeper into the causal mechanisms of policy change, it is appropriate to consider the insights from several complementary traditions, including other policy process frameworks, theories of bureaucracy, and disability policy research. This section describes the ACF and complementary traditions that inform the conceptual framework used in this study.

The ACF is particularly well suited to addressing the research questions of interest. The ACF is a longstanding framework based on the idea that policy change primarily stems from competition between advocacy coalitions (Jenkins-Smith et al., 2017; Sabatier and Jenkins-Smith, 1993). One of the ACF’s central objectives is to better understand the occurrence of policy change. To that end, the ACF hypothesizes several pathways to policy change: (1) external perturbations; (2) internal events; (3) policy-oriented learning; and (4) negotiated agreement (Jenkins-Smith et al., 2017; Sabatier and Jenkins-Smith, 1993).² While one or more of these conditions are hypothesized as being necessary for major policy change, they are not expected to be sufficient. A number of other conditions, including contextual factors, political opportunity, and coalition resources and strategies, are also likely to influence pathways to change. While this hypothesis has largely been substantiated, exploration of the secondary components (Pierce et al., 2017) and causal mechanisms (van der Heijden et al., 2019) by which policy change occurs remains an important line of inquiry (Jenkins-Smith et al., 2017).

While the ACF is particularly well suited to addressing the research questions of interest, several other theoretical traditions, including other policy process frameworks and theories of bureaucracy, are also worthy of consideration. Innovation and Diffusion (I&D) models, for example, focus on the diffusion of similar policy innovations among jurisdictions over time. I&D models hypothesize diffusion as being the result of several potential factors, including imitation, learning, normative pressure, competition, and coercion (Berry and Berry, 1990, 2017; Rogers, 1995; Shipan and Volden, 2008, 2012). Recent I&D studies substantiate the hypothesis that interest group activity matters to both state-level innovation (Boehmke, 2005) and diffusion processes (Garrett and Jansa, 2015).³ However, I&D studies tend not to linger on the role of interest groups or coalitions, in part due to the constraints of standard I&D methods (Garrett and Jansa, 2015). PFT, in turn, is a mechanism-focused approach, which seeks to illuminate feedback processes by which policy influences politics, which in turn shapes subsequent policy (Béland and Schlager, 2019; Mettler and SoRelle, 2014; Pierson, 1993; Skocpol, 1995). Policy feedback researchers offer considerable evidence that resources provided by social welfare policies can impact identity, political participation, and social movement participation, through both resource and interpretive effects (Mettler and SoRelle, 2014).

Theories of bureaucracy are also relevant to this study. In addition to a longstanding recognition of iron triangles (Adams, 1981) and regulatory capture (Krause, 1999; Stigler, 1971), newer theories of bureaucratic activism (Eaton and Weir, 2015), deliberate discretion (Huber and Shipan, 2002), and competing theories of contracting (Bertelli and Smith, 2010; Davis et al., 1997; Eisenhardt, 1989; Hirsch et al., 1987; Van Slyke, 2007) suggest that both bureaucrats and interest groups are likely to play a key role in policy change.

Finally, disability scholars have devoted considerable attention to CIE-focused changes over the last several decades, although they focus heavily on the achievement of individual employment outcomes (as opposed to policy outcomes) and also tend to disregard political factors. An early effort to use a diffusion model to explain widespread variation in the growth of state-level supported employment outcomes among states suggested that “systems change is complex” (McGaughey and Mank, 2001: 221). Subsequent development of an actor-driven, implementation-focused and goal-oriented Systems Change model highlights context and catalysts, strategies and facilitators (Hall et al., 2007), and reflects researchers’ observations about early adopters of CIE-focused policy (Hall et al., 2003, 2007). The early diffusion model and its successor, however, lack specification around the drivers, conditions or nature of expected policy changes. Moreover, the model turns the lens on employment outcomes for individuals with disability, rather than the policy process itself, omitting reference to political factors frequently highlighted by the policy process literature.

Conceptual framework

When confronted by a number of relevant theoretical traditions, Weible and Cairney (2018) highlight the “real opportunity to advance the field further if we pull together a broader selection of theories and draw practical insights from them.” (2018: 193). In the spirit of that observation, this study is organized as an application of the ACF, integrating relevant insights from other complementary traditions.⁴ As such, the conceptual framework is organized around ACF concepts, but draws upon insights from other theoretical traditions to illuminate the pathway to policy change. The main conceptual elements derived from these traditions include (1) sources of change, (2) context, (3) political opportunity, and (4) resource availability and use. Figure 1 offers a visual presentation of the framework used in this study, followed by a description of each conceptual element, and related discussion about the alignment of the ACF with other theoretical traditions.

Figure 1.

The Advocacy Coalition Framework.

Actors, groups, and subsystems

The ACF is organized around the premise of advocacy coalitions, comprising individuals who “share policy core beliefs and who coordinate their actions in a nontrivial manner to influence a policy subsystem” (Jenkins-Smith et al., 2017: 148). Relatedly, Sabatier (1988) defines the policy subsystem as the interaction between non-trivial policy actors who maintain an interest in a specific policy area. Advocacy coalitions can include a wide range of subsystem actors, including elite policymakers, policy advocates, agency administrators, program participants, and researchers (Sabatier, 1988; Sabatier and Jenkins-Smith, 1993). Both Giordono (2019) and Itkonen (2007) found evidence activity by interest groups or coalitions with opposing policy preferences. Other policy process traditions tend to focus on a narrower group of actors, such as elite policymakers and governments (Innovation and Diffusion models) and program participants (PFT).

Theories of bureaucracy have a long tradition of acknowledging the possibility of interest group influence over agencies, both in the form of iron triangles between congressional committees, interest groups and bureaucrats (Adams, 1981) and regulatory capture wherein an agency prioritizes one or more special interests (Krause, 1999; Stigler, 1971). More recent traditions in public administration also highlight cases of bureaucratic activism and decision-making outside of traditional channels (Eaton and Weir, 2015; Pacewicz, 2018), as well as deliberate loosening of the policy reins by legislators (Huber and Shipan, 2002). Finally, competing theories of contracting, including principal-agency theory (Eisenhardt, 1989; Jensen and Meckling, 1976; Walsh and Seward, 1990), stewardship theory (Davis et al., 1997; Dicke, 2002; Van Slyke, 2007), and relational contracting (Bertelli and Smith, 2010), all acknowledge the challenges associated with monitoring and managing the underlying motives and actions of government contractors. These traditions, therefore, are consistent with the ACF’s assumption that a variety of policy actors may participate, and even dominate, the policy-making process.⁵

Sources of change

In the ACF, external or internal events are expected to increase the likelihood of policy change (Jenkins-Smith et al., 2017; Sabatier and Jenkins-Smith, 1993). External shocks can include events, via heightened public and political attention, agenda change, redistribution of coalition resources, and/or change to venue availability (Heikkila et al., 2014). Similarly, internal perturbations, such as a system failure or crisis, can also heighten attention to program deficiencies and reform opportunities (Jenkins-Smith et al., 2017). Finally, policy-oriented learning and negotiated agreement are alternative pathways to policy change (Jenkins-Smith et al., 2017). Similarly, the Systems Change model developed by disability researchers places a strong emphasis on policy-oriented learning, with little emphasis on the other pathways of change hypothesized by the ACF.

Innovation and Diffusion models also hypothesize multiple sources of policy diffusion, including imitation, learning, normative pressure, competition, and coercion (Berry and Berry, 2017; Shipan and Volden, 2008, 2012). Even though the ACF is not explicitly concerned with diffusion, and only directly focuses on two of these sources (coercion and learning), the other sources of change hypothesized by I&D models are also compatible with the ACF. We might expect imitation to play a role in the collection and use of information by advocacy coalitions, a key coalition strategy in the ACF (Sabatier and Weible, 2007). The role of normative pressure is less clear in the context of the ACF, although internal events (e.g. crises, scandals, and failures) may be driven by violations of system norms. And while inter-jurisdictional competition is not explicitly addressed by the ACF, the framework is predicated on the idea that coalition competition is a primary driver of policy change. Finally, the variation in both the content and form of the changes and evidence of coalition activity (Giordono, 2019) suggests more than a simple policy diffusion process, despite the standard S-shaped curve typically associated with the Innovation and Diffusion literature (see Appendix A1).⁶

Finally, PFT offers the view that policy changes influence politics, including civic engagement and political participation, which then acts to influence policy change (Béland and Schlager, 2019; Mettler and SoRelle, 2014; Pierson, 1993; Skocpol, 1995). Again, while the ACF does not explicitly identify policy feedback as source of change, the framework is clearly articulated as a cyclical process. As such, the possibility that coalition formation, membership and resources might be influenced by a previous policy change is explicitly acknowledged by the ACF, although the connections remain underexplored.

Context

The ACF also expects that relatively stable parameters, including basic attributes, sociocultural values, and constitutional structure may enhance (or detract from) the subsystem’s conduciveness to policy change (Jenkins-Smith et al., 2017; Sabatier and Jenkins-Smith, 1993). The disability literature offers similar insights. For example, early research by McGaughey and Mank (2001) shows that state-level ideology and economic conditions are predictors of integrated employment implementation. They also find that investment and participation levels may influence the likelihood of change, which they attribute to state implementation commitment (McGaughey and Mank, 2001).

The overall level of subsystem collaboration is also expected to influence the use of expert-based information and learning, with collaborative subsystems less likely to use expert-based information for political purposes than more adversarial subsystems (Weible, 2008). Collaboration and partnership have been key elements in the Systems Change model since its development (Hall et al., 2007), and Butterworth et al. (2017) identify interagency collaboration as a key predictor of positive change in employment outcomes.

Finally, I&D models also predict that the policies in geographically contiguous and/or ideologically similar jurisdictions can influence state-level policy innovation (Shipan and Volden, 2008, 2012). In the context of the ACF, for example, we might imagine that policy experiences from other states could be used as information, for imitation, learning, or even coalition mobilization purposes.

Political opportunity

In these subsystems, we do not expect to see major differences in the national-level long-term opportunity structures highlighted by the ACF (Sabatier and Weible, 2007). However, variation in the decentralization of authority between state and local control may be expected to make a difference (Kübler, 2001). Indeed, the Systems Change model proposes that local control of service contracts (often at the county level) is more likely to yield positive changes to employment outcomes (Hall et al., 2007), suggesting that local authority increases political opportunities for policy change. On the other hand, low service provider capacity and competition, especially in rural areas (Van Slyke, 2007), might reduce the local policy-makers’ willingness and capacity to influence policy change.

We may also expect levels of policy conflict and legislative professionalism to influence the degree to which politicians afford statutory discretion to bureaucrats (Huber and Shipan, 2002). Specifically, high levels of technical complexity and legislative professionalism in the context of low policy conflict are expected to yield an incentive for politicians to cede control to bureaucrats (Huber and Shipan, 2002). In the DH&E context, which requires substantial technical knowledge of appropriate service provision strategies, we might expect bureaucrats to have more opportunities for making policy decisions, especially in the context of a unified and professional legislature.

Strategic use of available resources

The ACF further suggests that coalitions strategically use available resources and related strategies (Sabatier and Weible, 2007), which is expected to result in policy change via heightened attention, agenda change, redistribution of resources and opening/closing of policy venues (Jenkins-Smith et al., 2017: 145; Nohrstedt, 2011).⁷ The ACF’s hierarchy of resources points to formal legal authority as one of the most powerful coalition resources (Sabatier and Weible, 2007), but it is not the only available resource. For example, public opinion can serve to help sway decisions by officials in positions of authority, and coalitions typically invest heavily in efforts to gain the support of the public (Sabatier and Weible, 2007). In low salience policy subsystems like the DH&E subsystems, however, we might expect public opinion to play a less important resource role (Jones and Jenkins-Smith, 2009).

Relatedly, the ACF anticipates that information acts a key resource that can take relatively neutral forms of information- and data-sharing, as well as more politicized use of framing and narrative strategies for persuasion purposes (Sabatier and Weible, 2007). Indeed, goal communication and training are both key methods of information-sharing in the Systems Change model (Hall et al., 2007). Sabatier and Weible (2007) also describe mobilizable troops as an inexpensive resource often used by coalitions with few financial resources. Relatedly, Hall et al. (2007) highlight the importance of relationships between policy actors to promote and support systems change, while Butterworth et al. (2017) acknowledges the importance of family members in key service-related decisions.

Skillful leadership (Sabatier and Weible, 2007) by key players (Hall et al., 2007) is another key coalition resource for developing a vision, attracting resources, and using resources strategically. Relatedly, federal anchoring agencies can be an important source of leadership and belief system maintenance (Ellison and Newmark, 2010). A number of recent federal actions suggest renewed federal prioritization of CIE-focused services, while some federal regulations that are arguably contrary to CIE priorities remain in place. As such, we might expect to see national partners serving as coalition resources.⁸ The few disability-focused policy process studies also suggest that disability policy change is frequently characterized by the use of standard political strategies to achieve policy change, including agenda-setting (Pettinicchio, 2013), problem definition (Itkonen, 2009; Jeon and Haider-Markel, 2001), venue-shopping (Nagel, 2006), framing and narrative (Itkonen, 2007; Jeon and Haider-Markel, 2001) and interest group mobilization (Shapiro, 1994). All of these strategies are aligned with the ACF’s resource hierarchy.

Finally, PFT predicts that the resources associated with policy decisions are also expected to influence subsequent policy change by conveying additional capacity upon specific populations to engage in political participation and advocate for their needs, otherwise known as resource effects (Béland and Schlager, 2019; Mettler and SoRelle, 2014). These resources, along with existing rules and procedures, may also generate “interpretive effects” that influence individuals’ propensity for civic engagement via personal experience with government participation and broader shifts in social construction (Béland and Schlager, 2019; Mettler and SoRelle, 2014). We might see a similar feedback effect from the resources invested into employment-oriented programs and policies in the context of DH&E services and supports, especially with respect to growth and/or reduction of advocacy coalition efficacy.

Methods

This study takes a qualitative, comparative approach to addressing the research questions of interest, including: (1) What does state-level CIE-focused policy change look like? (2) Why and how did CIE-focused policy change occur? The study uses multiple sources of data from two states (Washington and Pennsylvania) and a comparative analytic approach in the context of the adapted ACF framework.

Operationalization

Pierce et al. (2017) note the importance of specifying operationalization decisions and related adaptations. In alignment with that guidance, this section describes three distinctive elements of this ACF application, including, (1) subsystem delineation; (2) definition of policy change; and (3) identification and characterization of advocacy coalitions, to achieve a context-specific ACF operationalization. First, the policy subsystem is a key concept in the context of the ACF. In this study, the subsystem is delineated as the DH&E system in each state, from which relevant policy actors are expected to include (at a minimum) representatives from government agencies, politicians, advocacy organizations, service providers, and professional associations. Second, the dependent variable Major policy changes include “significant shifts in the direction or goals of the subsystem” (Jenkins-Smith et al., 2017: 145), while minor policy changes primarily reflect a change in the means of achieving acknowledged goals (Jenkins-Smith et al., 2017; Sabatier and Jenkins-Smith, 1999).⁹ This study further refines the distinction, classifying major policy changes as those that impose restrictions on the status quo, and minor policy changes that simply provide continued support for the provision of CIE services without any accompanying change to non-CIE services.¹⁰ Policies that supplement the status quo are categorized as minor change because the policies function predominantly as a means of achieving a previously stated goal. Evidence for subsystem advocacy coalitions relies on findings from Giordono (2019), who identifies two advocacy coalitions in each subsystem, “Employment First” and “Choice,” which exhibit distinct policy core beliefs and policy preferences. The Employment First coalition would like to prioritize CIE-focused services and supports and minimize reliance on non-CIE services and supports. In contrast, the Choice coalition prefers to maintain a full range of options and allow individuals (and their families) to select the services and supports that best meet their needs. The Choice coalition tends to be loosely organized and reactive and has formally emerged in the early- to mid-2010s, even though it represents the status quo. In contrast, the Employment First coalition is well organized and proactive.

Each conceptual element includes multiple conditions and sources of evidence, which were used to inform data collection, coding and analysis methods. See Appendix A2 for a list of conceptual elements.

Case selection

The two diverse cases were selected purposively to maximize variation in the timing of policy adoption¹¹ and policy-related outcomes¹² during the 18-year window from 1999 to 2017. The use of two states provides a counterfactual to assess the conditions that lead to policy change (Gerring, 2008; Seawright and Gerring, 2008). Table 1 presents the sampling frame used to select the two states, Washington and Pennsylvania.

Table 1.

Sampling frame.

Supported employment and EF policy adoption by US state
Proportion in supported employment 1999	Percentage point change in proportion in supported employment 1999–2014	EF policy adoption	Timing of EF policy adoption	Number of states	States^a
High	High	Yes	Early	1	WA
High	High	Yes	Late	1	OK
High	High	No	None	4	MA, MD, MO, VT
High	Low	Yes	Early	7	DE, FL, IA, KS, OH, TX, UT
High	Low	Yes	Late	2	IN, NV
High	Low	No	None	13	AL, AZ, CT, GA ID, LA, KY, MI,MN, NE, NY, SC, SD
Low	High	Yes	Early	1	VA
Low	High	Yes	Late	1	MS
Low	High	No	None	4	NC, NM, TN, WI
Low	Low	Yes	Early	3	CA, IL, OR
Low	Low	Yes	Late	2	PA, WY
Low	Low	No	None	2	CO, NH
n/a in one or more of the selection indicators				9	AK, AR, HI, ME, MT, ND, NJ, RI, WV
Total				50

Note:

High level: 1999 proportion above median (high initial SE proportion).

Low level: 1999 proportion below median (low initial SE proportion).

High change: 1999–2014 percentage point change below median.

Low change: 1999–2014 percentage point change below median.

Early = Executive Order or Legislation adopted before 2014.

Late = Executive Order or Legislation adopted 2014 or later.

^aAt the time that the sampling frame was developed, states in bold font adopted EF legislation. All others adopted via Executive Order. Pennsylvania passed legislation just prior to the data collection period.

In qualitative terms, Washington represents early adoption of CIE-focused policy (before 2014), while Pennsylvania is a late adopter (2014 or later). The two states are similar with respect to a number of other broader state-level characteristics, including citizen ideology, welfare generosity, and the presence of a rural/urban divide. The semi-longitudinal nature of the study mitigates the threat to inference through selection on the dependent variable (King et al., 1994). Because both states ultimately experience policy change, the case selection strategy allows for a counterfactual based on the distinct timing of policy change. Selecting cases with different support employment outcomes, however, provides leverage for understanding the pathways to change experienced by each state. See Appendix A3 for a detailed comparison of the two states.

Data collection

The study uses a combination of telephone interview data and publicly available documents as the primary data sources. Using subsystem delineation parameters, an iterative search strategy was used to identify non-trivial policy actors. For each state, a systematic Google search provided an initial source of policy-relevant sources, including organization websites, policy documents, and news articles.¹³ The top 100 returns from each search were downloaded and screened for relevance. Databases for organizations and documents were produced from the search results. Organizational websites were screened for relevant documents and added to the document database, which was coded for stakeholder names, organizations, and stakeholder type. All relevant results were entered into a respondent database. Policy actors with the highest citation frequencies from each stakeholder type were selected for initial contact. During the interview process, the respondent sample was supplemented by snowball sampling based on references provided by stakeholders (Lofland, 2006). A total of 36 semi-guided telephone interviews were conducted in the two states (20 in Washington; 16 in Pennsylvania).¹⁴ The completion rate among individuals who responded to the initial e-mail invitation was 80% in Washington and 61% in Pennsylvania. Interview respondents were assured anonymity to promote candid responses. The average interview duration was 50 min. See Appendices A4 and A5 for the interview guide and number of responses by stakeholder type. All interviews were recorded and transcribed.

Supporting data were retrieved from a variety of publicly available secondary sources. Selected documents from the initial Google search were catalogued for analytic use. In addition, LexisNexis searches were conducted to retrieve relevant items from state newspapers, yielding a total of 47 relevant items from 8 Washington newspapers and 101 items from 27 Pennsylvania newspapers. Targeted searches of state websites yielded official policy documentation, including legislation, executive orders, and policy directives/memos. Supporting documents transmitted by interview respondents were catalogued and reviewed. Extant state-level data related to demographics, socio-economic status, service participation, and federal grant participation were gathered from a variety of sources, including the American Community Survey (U.S. Census Bureau, n.d.), StateData.info (University of Massachusetts Boston, Institute on Community Inclusion, n.d.), State of the States in Intellectual and Developmental Disabilities (Coleman Institute for Cognitive Disabilities at the University of Colorado, n.d.), State Fact Sheets (Center on Budget and Policy Priorities, 2012), the Civil Rights Litigation Clearinghouse (University of Michigan Law School, n.d.), Squire’s Legislative Professionalism Data (Squire, 2007, 2017), Fording’s Citizen Ideology Index (Fording, 2018), and various federal government websites (e.g. U.S. Department of Labor, n.d.b.).

Coding

All interview transcripts were coded in Dedoose 7 using ACF conceptual categories and a multi-staged coding strategy involving initial (open) coding on a sample of the interviews, followed by theoretically grounded focused coding on all interviews (Lofland, 2006). The initial coding strategy was used to become familiar with the data and assess the relevance of ACF conceptual elements for the context and research questions of interest. A focused coding strategy was used to code each interview for the presence of selected ACF conceptual elements, including (1) external and system events; (2) coalition resource availability and use; (3) policy changes; and (4) beliefs and coalitions. See Appendix A6 for a detailed codebook, including primary codes and subcodes.

Analytic approach

The coded and catalogued data were triangulated and used to inform within-case and cross-case comparative analyses. The within-case approach involved preparation of an in-depth case study for each case using theoretically oriented process tracing methods to identify likely causal paths between one or more conditions and the outcome(s) of interest (George and Bennett, 2005). The results from the within-case analyses were then used comparatively, taking advantage of the longitudinal nature of the research design to yield inferences about the contributors to variation in policy change timing and pathways (George and Bennett, 2005). Per Maxwell (2013), threats to validity were primarily managed by employing (1) rich data from multiple sources; (2) triangulation; and (3) a comparison case (i.e. counterfactual).

Case summaries: Pennsylvania and Washington

This section provides a brief overview of the events relating to CIE-focused policy change in Pennsylvania and Washington. Figure 2 offers a side-by-side visual of key events and policy changes in the two states between 1999 and 2017.

Figure 2.

Timeline of key events and policy changes in Washington and Pennsylvania (1999–2017).

Each case is described briefly below, followed by an analytic discussion of related findings.¹⁵ See Appendices A3 and A7 for a detailed comparison of case conditions and a thorough analytic case summary for each state, respectively.

Washington

In the late 1990s, after multiple years of growth in CIE-focused services, the DH&E subsystem entered a period of increased conflict over the quality of services and related priorities, resulting in a legislated mandate to review system goals (Washington Legislature, 1998).¹⁶ Following that mandate, and a subsequent commitment by service providers to providing CIE-oriented services, Washington State became one of the first adopters of a CIE-focused policy directive in 2006. The Working Age Adults Policy (Washington Department of Social and Health Services, 2004) established employment supports as the first use of subsystem program funds for working-age adults (21–61 years) (Hall et al., 2007). The policy effectively displaced day habilitation (“community access”) services with CIE-focused services for the first nine months of service participation and was subsequently codified in 2012 legislation (Washington Senate, 2012). Initial policy changes were heavily promoted by the agency director at the time, as well as members of the service provider and participant communities.

These efforts were followed by a 2015 commitment to eliminate new admissions to sheltered workshops as part of a HCBS transition plan (Washington State Department of Social and Health Services and Washington State Health Care Authority, 2015). The state also adopted several minor changes during the study period, including a comprehensive needs assessment strategy and outcomes-based service provider rates.

The Washington DH&E has a strong history of employment-related service provision and stakeholder collaboration. However, more recently, coalitions have become more active on both sides of the CIE policy issue, and the Choice coalition has repeatedly mobilized in response to Employment First policy changes and related activities. Respondents from both coalitions describe ongoing tension. One Choice respondent states:

If the jobs were available now, then Washington would already be an Employment First state. It's our position that Washington wants to become an ‘Employment Only’ state under the guise of ‘Employment First’. (WA_ind30 telephone interview)

Similarly, an Employment First respondent describes ongoing discontent with the Employment First approach:

We've got some legislators who continue to oppose it. [Who] say, ‘Why would we force people to try our most expensive service.’ And we live that every year in legislation, and again, this year. So, even though we put Employment First in policy, now that policy is continually threatened… (WA_ind04 telephone interview)

Early CIE-focused changes were spearheaded by the agency director, as well as organized groups of family members and service providers. Notably, the service providers’ professional association took a definitive stance to support Employment First (and reject Choice) in the early 2000s.

Pennsylvania

Pennsylvania was slow to prioritize CIE-focused policy. Since 1990, Pennsylvania has had a written policy supporting access to employment opportunities for individuals with intellectual disability (Commonwealth of Pennsylvania, 1990), but service provision records suggest a low level of attention to integrated employment services. In 1999, only 19% of Pennsylvania participants received integrated employment services, substantially lower than the average of 29% among all states (University of Massachusetts Boston, Institute on Community Inclusion, n.d.). Moreover, no major policy changes occurred until after the 2014 HCBS Final Rule and related guidance, despite early attempts to renew the state’s commitment to employment policies via policy directive in the 2000s. Increased attention to CIE-focused policies began in the early 2010s, with minor changes related to centralization of contracting authority and a related service provider rate restructuring process. After the federal “settings” rule was finalized, multiple changes occurred, including an Executive Order and new HCBS waiver rules that added new employment-focused service definitions and restricted facility-based participation. Agency directors acted as important advocates for CIE-focused change, but also actively worked with other stakeholders, including sympathetic service providers, families, and researchers.

Similar to Washington, Employment First coalition activity preceded major policy changes, beginning in the early 2010s with a stakeholder social media campaign and gubernatorial involvement in the “A Better Bottom Line” CIE-focused national campaign. The Choice coalition did not emerge until the new HCBS rules were announced in 2016. In response to the more restrictive rules, service providers, families, and their legislators mobilized at a protest in Harrisburg, which led the state to reduce the restrictions. Employment First advocates have organized to promote Employment First legislation, which was introduced as HB 2130 in the 2015 session (Pennsylvania House of Representatives, 2015), introduced again as HB 1641 in the 2017 session (Pennsylvania Legislature, 2018). Respondents note that the legislation is intended to codify preceding policy changes, and that organizers have deliberately avoided including dramatic changes to sheltered workshop rules in the legislation. One Employment First supporter states:

I'll be absolutely frank in saying that we did not want to get into the middle of the sheltered workshop fight. We didn't think that was a fight that we could win… As much as we would like it to go further, we think there's only so much appetite politically that we can manage right now. (PA_ind51)

The legislation was ultimately signed into law by Governor Wolf in June 2018, with provisions for state agencies to prioritize supported employment services, establish a Governor’s Cabinet for People with Disabilities, and a set goal for at least 7% of the state agency work force to be comprised of individuals with disability (Pennsylvania Legislature, 2018).¹⁷

Findings

Results from the study are organized around the two main research questions, including: (1) What does state-level CIE-focused policy change look like? (2) Why and how did CIE-oriented policy change occur?

What does CIE-oriented policy change look like?

Washington and Pennsylvania both experienced major and minor policy changes over the course of the study period, but differed in the degree, timing, and sequence, as shown in Figure 2.

Types of change

Washington and Pennsylvania both made major policy changes during the study period. The Working Age Adults Policy in Washington clearly favors participation in employment services. Limitations on participation in day habilitation services and sheltered workshop entry foreclose on the use of such services. In Pennsylvania, new waiver-related rules requiring at least 25% participation in a community setting also limits the use of segregated employment services. However, Washington changes are more restrictive to status quo service options and place a greater emphasis on employment.

These major policy changes might otherwise be described as incremental, in the sense of introducing marginal changes to existing rules. However, they clearly reflect a shift in the policy core beliefs, namely related to the importance and availability of traditional non-CIE services, and are described as meaningful by respondents. While the ACF does not preclude incremental change, major change is not typically described by ACF scholars in those terms.

Both states also made a variety of minor policy changes throughout the study period, such as new service definitions (Pennsylvania) and an outcomes-based rate structures (Washington and Pennsylvania), both of which reflect a shift in the means of encouraging participation in employment services. The Systems Change model explicitly anticipates these types of changes (e.g. flexible funding, incentive structures), but does not as clearly articulate the types of major changes observed in these subsystems.

Timing and sequence

There are clear differences in the timing and sequence of policy changes. As expected, Washington experienced early major policy change (before 2014), while Pennsylvania did not experience major policy change until after 2014. In both states, the major policy changes are associated with a combination of decisions by government authorities and changes to institutional rules. The sequence of events differs in the two states, with policy directives and institutional rulemaking unexpectedly preceding decisions by government authorities in Washington State.

Why and how did policy change occur?

As shown in Figure 3, and described in the remainder of this section, the evidence indicates that in both cases, initial events and other key conditions yielded opportunities for Employment First supporters to proactively use coalition resources in pursuit of policy change. Choice supporters tended to act more reactively in response to the adoption and/or proposal of restrictive CIE-focused policies. Pathways to policy change in both states involved several common conditions, including coalition mobilization and use of framing/narrative, bureaucratic activism and heightened stakeholder attention. Several differences also emerge. Washington State, an early policy adopter, presented with high antecedent CIE-oriented service outcomes, and internal discord, which was followed by service provider defection from the status quo (Choice) coalition. In contrast, Pennsylvania, a late policy adopter, faced low antecedent CIE service levels and followed a quasi-external pathway to policy change only under conditions of federal guidance and other state policy innovations. In Pennsylvania, the status quo coalition (Choice) had more strategic resources at its disposal than in Washington State.

Figure 3.

Conditions associated with change in Washington and Pennsylvania.

See also Appendix A8 for a more detailed matrix of conditions associated with major policy change.

Sources of change

The ACF posits that multiple conditions can act as necessary, but not sufficient, sources of policy change. In both states, subsystem conditions provided the spark for subsequent coalition-based activity and policy change. In Washington, the late 1990s subsystem breakdown heightened political and subsystem attention to CIE-related issues and was closely followed by the (mandated) 2002 Stakeholder Workgroup report that articulated support for a Pathway to Employment, as well as the 2003 dismissal of the Arc of Washington vs. Quasim lawsuit that argued for a full range of services (University of Michigan Law School, n.d.). The later events represented a negotiated agreement and a judicial venue closure, respectively, offering an opportunity for administrative leadership to formulate and implement the initial Working-Age Adult Policy. Pennsylvania also experienced some conflict over service provision in the 2000s, targeted at waitlist issues and problems with the county-based contracting, rather than the service mix.¹⁸ However, the 2014 federal HCBS “settings” rule served as an important motivator for major policy changes in both states. In Pennsylvania, observations of Olmstead-related litigation in other states also provided an impetus for change. These observations suggest that Washington experienced an internal pathway to policy change, while Pennsylvania experienced a quasi-external pathway. Pennsylvania’s pathway to change is also consistent with both coercion and imitation mechanisms articulated by Innovation and Diffusion models.

There is also evidence that policy-oriented learning contributed to minor policy changes in both Pennsylvania and Washington. The process of developing a comprehensive needs assessment process (WA) and setting outcomes-based rate structures (both PA and WA) were both accompanied by substantial reference to research and extant data. There is less evidence of policy-oriented learning before the major changes, as indicated by technical analysis and related analytic debate. However, selected service providers in Washington refer to survey data (showing that individuals would prefer to work) as an instigator for subsequent efforts to transform their service structures.

Context, political opportunity, and coalition resources/strategies

Relatively stable state and subsystem contextual attributes can also influence the favorability of the environment to policy change. The two states displayed slight differences in political climate and key economic conditions. Washington’s government is historically more liberal and unified, although the two states are roughly equivalent with respect to citizen ideology (Fording, 2018). Pennsylvania has had consistently higher legislative professionalism scores than Washington during the study period, although both states were in the top 40% of states (Squire, 2007, 2017). Economic conditions, including unemployment and labor force participation responses to the Great Recession, were similar during the study period (US Census Bureau n.d.). Moreover, subsystem service structures were similar—both states relied on fee-for-service models (i.e. not managed care) and HCBS services were not unionized in either state. Finally, despite differences in government ideology, the states’ overall welfare commitment, including TANF participation and support and Medicaid expansion, has followed similar trends over time.

The two states’ portfolio of CIE-focused service provision prior to the study period was considerably different. In Washington, the proportion of participants receiving integrated employment services in 1999 was 58%, compared with only 19% of Pennsylvania participants. Furthermore, the proportion grew to 85% in Washington, while Pennsylvania’s proportion remained consistently low. It is challenging to disentangle the influence of service levels from other conditions. There is little evidence that service levels alone influenced policy; instead, they interact with coalition resources and strategies to yield a pathway to policy change. The evidence in this study suggests that service levels matter, in these cases, at least in part because of the influence on coalition membership, especially among service providers.

Service providers advocate for policy change on their own behalf and can also act as mobilizers of families and individuals. In the late 1990s, major Washington service providers formed an advocacy network (P2000, subsequently renamed P2020) separate from the traditional service provider association Rehabilitation Enterprises of Washington (REW). In the 2000s, P2020 and REW merged to form the Community Employment Alliance and issued a statement declaring CIE as its preferred policy and yielding an important source of support for proposed policy changes. In contrast, service providers in Pennsylvania played an active role in mobilizing individuals and families against proposed rule changes, suggesting that service provider coalition membership can play an important causal link between service levels and policy change.

These observations neither substantiate nor refute competing theories of contracting. However, they do suggest that contractor mobilization can go both ways, providing support for policy goals or actively resisting them. Since the finalization of the 2014 federal HCBS rule, however, most service providers (and their professional associations) in both states anticipate eventual implementation of federal regulations, and articulate some combination of support, neutrality, or resignation toward Employment First policy positions. One representative from a Pennsylvania professional association states “The state's really careful to say they're not mandating closure of these workshops or adult training facilities, but the landscape is really changing” (PA_ind45 telephone interview). That said, in both states, a contingent of service providers have mobilized against CIE-focused policy in the form of Choice coalitions, and in Pennsylvania there is little of a mass coalition defection by service providers similar to that experienced in Washington in the 2000s.

Political opportunity refers to the influence of long-term opportunity structures and political environment on the resources and behaviors of subsystem actors and coalitions (Sabatier and Weible, 2007). In these cases, bureaucratic advocacy stands out as a key contributing factor related to policy change in both states. Major CIE-focused policy change in both states was initially achieved via policy directive (in Washington) and rulemaking (in Pennsylvania) by administrative leaders widely recognized as vocal and skilled Employment First champions. These findings are aligned with both ACF suggestions that formal legal authority is one of the most powerful coalition resources (Sabatier and Weible, 2007). The results also align with literature acknowledging the active role of and discretion accorded to bureaucrats in the policymaking process. That said, an attempted CIE-focused change via policy directive in Pennsylvania during the late 2000s failed, suggesting that bureaucratic activism alone is not sufficient for policy change, but also requires strong and committed support from other fronts.

The contribution of decentralized authority (to the county level) is unclear, although multiple respondents attributed Washington policy changes to county authority. However, authority in Pennsylvania was similarly decentralized during the same period, and no change occurred there, even under the leadership of a sympathetic state administrator. Moreover, by the time that Pennsylvania began to explore policy change, authority had been centralized at the state level, in response to federal concerns about inconsistent county contracting practices. These findings suggest that goal alignment between state and county administrations offered the opportunity and support for change, rather than the structure itself.

There is little evidence that macro-level long-term opportunity structures influenced policy change. Both subsystems are pluralist subsystems with high consensus requirements and formal system openness. Moreover, both states are subject to overlapping rural/urban and ideological divides, suggesting that such cleavages do not explain observed differences in the policy process, although there is some evidence that such cleavages may contribute to coalitional divergence in policy preferences. However, it would be ill-considered to deny the influence of the national stage on state-level choices, even though national events, support, and guidance are broadly applicable to all states. National events and support offered opportunities for policy actors, especially those from the Employment First coalitions, to raise the profile of CIE-focused policy change on the subsystem agenda and to provide a coalition-based interpretation. Furthermore, stakeholders in both states have access to the formal, federally supported Protection and Advocacy system, and there is evidence that the P&A-supported organizations (e.g. DD Council, Disability Rights Network, UCEDD) are active in both states. Similarly, there is strong evidence of both formal channels of communication and advocacy (e.g. public comment periods) in both states, as well as informal networks. However, there is some evidence that the Employment First coalition in Washington had closer ties with national supporters in the 2000s, while Pennsylvania ties with national supporters were relatively weak until the mid-2010s. Again, these observations are consistent with both ACF and I&D expectations.

In both states, the Employment First coalition was more proactive in its pursuit of major policy change, while the Choice coalition tended to mobilize more reactively in response to proposed or actual policy changes. The character of mobilization in both states suggests that the Employment First coalition has access to more resources than the Choice coalition, especially during the period(s) prior to policy change. However, both coalitions attempted to influence policy via heightened political and subsystem attention, especially through the strategic connections to allies in positions of authority, information, and mobilizable troops.

The Employment First coalition maintained a strong connection in the form of an ally in agency leadership in both states. In Washington, for example, while respondents noted the important contributions of skilled advocacy leadership, a dedicated organization (WISE) and county-based support, they attributed the initial policy directive to the Director of Developmental Disabilities Division Linda Rolfe, who held that position until 2013. Similarly, Deputy Secretary Nancy Thaler in Pennsylvania has been given much of the credit for recent changes in Pennsylvania, although respondents also noted the contributions of her predecessor Steve Suroviec and other advocacy leaders. In contrast, the Choice coalition does not appear to have had formal support among state agency leadership, instead focusing on establishing connections with legislative allies, often relying on existing advocacy channels and frequent in-person contact. The Employment First coalition also invested in establishing and maintaining relationships with state legislators to influence CIE-focused policy change.

Relatedly, both coalitions strategically used information and mobilizable troops in tandem in both states. For example, the Pennsylvania Choice coalition relied heavily on these strategies, mobilizing sheltered workshop participants and their families to protest proposed rule changes at the Capitol, providing insight into the path-dependent nature of subsystem policy change and stasis. Similarly, the Employment First coalition in Pennsylvania mobilized individuals with I/DD to develop a social media campaign (#iwanttowork) that respondents attribute with focusing political attention on CIE-focused policy. The coalitions typically rely on civil rights framing (employment as a right, choice as a right) and narrative (stories of hope, stories of decline), targeting subsystem stakeholders and selected government authorities, to influence major policy change. In Washington, administrative data have also been used by both coalitions since the initial policy change, both politically and instrumentally. Thus far, Pennsylvania coalitions show little evidence of using administrative data to influence major policy change, relying more heavily on framing to convey information. In contrast, administrative data and expert-based opinion were used in both states to influence minor changes, as anticipated by the ACF. Interestingly, neither coalition in either state focused its resources on influencing broader public opinion.

Returning to Figure 3, several distinguishing subsystem conditions stand out as having an influence on the timing of major policy change. First, as was previously acknowledged, the two subsystems faced very different initial service levels. Second, the mass defection of service providers from the Choice coalition in Washington, as indicated by reform of the main professional association, signaled support for major policy change from that stakeholder group. While service providers in Pennsylvania are cognizant of the federal changes, they have not publicly announced a preference for Employment First policy en masse. While they are not necessarily opposed to all Employment First policy, they are protective of traditional services and mobilize as necessary to resist reform.

Third, long waiting lists for services emerged as a major issue in both states in the early 2000s. In Pennsylvania, advocates formulated a major mobilization campaign to address the issue with direct advocacy. In contrast, Washington waiting list issues were addressed by the judicial system, freeing up stakeholders to pursue CIE-related issues. These observations suggest that in a system with limited resources, competing subsystem agenda items can also influence the timing of policy change.

Finally, even though the study timeframe extended only to 1999, multiple Washington respondents stated without prompting that policy decisions made in the 1980s and 1990s were integral to the policy changes that occurred in the 2000s. One respondent states, for example,

I think the only other unique thing is just, and I'm biased of course, is just that Washington has invested so strongly in training and TA for the last 30 years… that’s one key piece of their success over the years. (WA_ind01, telephone interview August 2017)

While this observation is aligned with those made by disability researchers (e.g. Hall et al., 2003, 2007; Winsor et al., 2006), neither the Systems Change framework nor the ACF provides a strong causal mechanism for linking previous policy decisions, especially minor ones, to subsequent policy change. It is, however, well aligned with Policy Feedback expectations. Specifically, the findings suggest a relationship between these earlier policy decisions, which heavily targeted service providers, subsequent service levels and the eventual Washington service provider coalition defection observed in the 2000s.

Discussion and conclusion

From a public policy perspective, considerable federal government resources have been invested to prioritize CIE outcomes for individuals with I/DD. And yet there continues to be variation in state-level policy change timing and content, even within the context of federal Medicaid rules and related guidelines. Shining a spotlight on the policy process helps refine our understanding of the types of policy change (minor and major) and conditions that are conducive to major policy change. Moreover, modeling policy change as a coalition-based competition provides information about the mechanisms that drive such change. This study also highlights the role of organized interests in a federal system that has devolved decision-making authority to the state level.

This study offers insight into conditions conducive to state-level policy change expected to be useful to both policymakers and researchers. While some of these findings are unlikely to come as a surprise to DH&E stakeholders, they are likely to offer insight into an ongoing nationwide change process. This application points to a number of coalition-based strategies and resources that have the potential to accelerate or inhibit change. Moreover, the application highlights the potential conflicts that arise in the context of policy change. “Minor” changes (e.g. changes to rate structures and assessment tools) may be as effective and meaningful for achieving policy goals without introducing conflict. In a system that prides itself on collaboration and trust, reformers may need to weigh the costs and benefits of pursuing major change. Finally, the ACF shines light on the implications of the subsystem agenda and attention for policy change; limited stakeholder attention and capacity means that strategic decisions about priority issue areas matter. In a small, mobilization-reliant subsystem like DH&E, choosing to prioritize gains in one area (e.g. waitlist) may come at the expense of gains in another area (e.g. CIE policy).

This study also offers several contributions to the policy process literature. First, this study illuminates the black box of interactions between context, actors, and events, while substantiating core hypotheses related to policy change and interest group activity from a number of theoretical traditions. Several of the common conditions, including bureaucratic advocacy, stakeholder mobilization, and heightened attention, tell a story of change at the margins of public attention, which is aligned with ACF expectations, as well as theories of bureaucratic activism (Eaton and Weir, 2015; Pacewicz, 2018) and delegated discretion (Huber and Shipan, 2002). The differences in sources of change (and timing), however, underscore the equifinality of the policy process, or the possibility of reaching the same outcome via more than one path. Finally, the findings provide insight into how organized interests can yield positive or negative feedback in the context of social service provision.

That said, there are limitations to what can be inferred from the study findings. First, the overall research design limits causal inference. Although the findings are likely to be analytically generalizable, findings from a two-case study are not generalizable to the larger population of US states, especially given the possibility of other pathways to change (e.g. hierarchical imposition). Furthermore, despite the use of variation in policy change timing as a counterfactual, the two cases selected for inclusion did not include any states that had not experienced policy change. The sample of respondents is small and non-representative and the sources of external information (e.g. media and policy documentation) are limited, although the study triangulates data from multiple sources wherever possible. Finally, the study’s focus, which was to illuminate the causal pathways to policy change using a multi-tradition policy process framework. While some theoretical expectations were acknowledged, no formal propositions were articulated or tested, thwarting Sabatier’s guidance to be “clear enough to be wrong” (Sabatier, 2000: 135).

Looking forward, there are opportunities for the findings from this study to contribute to future applied and scholarly research. The Systems Change model, for example, is used as the basis for decision-making and technical assistance in a variety of government initiatives (e.g. Butterworth et al., 2017). While this study does not explicitly recommend changes to that model, it offers insights into categories of policy change, the influence of political factors, and the role of coalitions, which would likely yield a more transparent and inclusive model of change. The findings also yield questions. The role that service providers played in Washington State highlights questions about theories of contracting—when contractors mobilize in the context of relational contracting (or stewardship), how do agencies maintain control over goals? Under what conditions do contractors mobilize, and (when) is it appropriate for them to engage service participants in those efforts?

The study also highlights opportunities to refine our conceptualization of policy change. The ACF and I&D models, for example, tend to treat policy change as dichotomous (minor/major and adoption/non-adoption, respectively) like a lightbulb with two settings. In contrast, this study reveals a more complex process of change that acts more like a dimmer switch, occurring at multiple time points, in multiple forms, and in varying sequences. Moreover, this study highlights potential connections between pre-1999 efforts to promote CIE, and subsequent changes in service provider coalition membership (i.e. defection) that are a main condition for early or late adoption. These results point to policy feedback as an important (though not exclusive) mechanism for policy change; interestingly, in this case, the feedback loop operates through service providers. These observations offer opportunities for theoretical refinement and generation of new hypotheses related to conditions under which bureaucrats and/or contractors are likely to act as important coalition actors.

Finally, the findings from this study enrich our understanding of the process by which policy changes occur in a federal system that continues to undergo decentralization in a variety of policy domains. On the one hand, such decentralization enables states to function as so-called laboratories of innovation (Thompson and Burke, 2007). On the other hand, states also have the opportunity to bow out, or at least delay, their entry into adoption of promising practices. In either case, the study offers a spotlight into the black box of the policy process, and the role played by a wide variety of organized interests therein.

Supplemental Material

sj-pdf-1-ppa-10.1177_0952076720942822 - Supplemental material for From employment optional to “Employment First”: Explaining two cases of state-level disability policy change

Supplemental material, sj-pdf-1-ppa-10.1177_0952076720942822 for From employment optional to “Employment First”: Explaining two cases of state-level disability policy change by Leanne S Giordono in Public Policy and Administration

Footnotes

Acknowledgements

I would like to acknowledge the support I received during the preparation of this article, including members of my Oregon State University dissertation committee, including Edward Weber (Chair), Hilary Boudet, Michael Jones, Gloria Krahn, and David Rothwell. Thanks to Chris Weible and the IPPA Spring 2018 Policy Workshop for early encouragement, and to two anonymous reviewers for feedback and guidance during the publication process. Finally, many thanks to the interview respondents who generously shared their experiences and insights in the interest of this research.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by the U.S. National Science Foundation Graduate Research Fellowship Program Grant No. 1314109-DGE.

ORCID iD

Leanne S Giordono

Notes

Supplemental material

Supplemental material for this article is available online.

References

Adams

(1981) The Politics of Defense Contracting: The Iron Triangle. 1 ed. New York: Transaction Publishers.

Agranoff

(2013) The transformation of public sector intellectual/developmental disabilities programming. Public Administration Review 73: S127–S138.

Aucante

Baudot

P-Y

(2018) Introduction implementing disability rights in national contexts: Norms, diffusion, and conflicts. Social Policy and Society 17(1): 87–100.

Béland

Schlager

(2019) Varieties of policy feedback research: Looking backward, moving forward. Policy Studies Journal 47(2): 184–205.

Berry

(1990) State lottery adoptions as policy innovations: An event history analysis. The American Political Science Review 84(2): 395.

Berry

(2017) Innovation and Diffusion models in policy research. In: Weible

Sabatier

(eds) Theories of the Policy Process. 4th ed. Boulder: Westview Press.

Bertelli

Smith

(2010) Relational contracting and network management. Journal of Public Administration Research and Theory 20(suppl 1): i21–i40.

Boehmke

(2005) Sources of variation in the frequency of statewide initiatives: The role of interest group populations. Political Research Quarterly 58(4): 565–575.

Bruch

Meyers

Gornick

(2018) The consequences of decentralization: Inequality in safety net provision in the post-welfare reform era. Social Service Review 92(1): 3–35.

10.

Bunch

Liebertz

Milita

(2018) Outcomes and outputs: Long-term effects of Temporary Assistance for Needy Families on caseload enrollment and poverty. Public Policy and Administration 33(3): 290–310.

11.

Butterworth

Christensen

Flippo

(2017) Partnerships in employment: Building strong coalitions to facilitate systems change for youth and young adults. Journal of Vocational Rehabilitation 47(3): 265–276.

12.

Butterworth

Smith

Winsor

, et al. (2016) StateData: The 2015 National Report on Employment Services and Outcomes. Institute for Community Inclusion, University of Massachusetts Boston. Available at: http://www.statedata.info/sites/statedata.info/files/files/statedatabook_2015_F.pdf (accessed 12 August 2016).

13.

Cairney

(2013) Standing on the shoulders of giants: How do we combine the insights of multiple theories in public policy studies? Policy Studies Journal 41(1): 1–21.

14.

Center on Budget and Policy Priorities (2012) State fact sheets: How states have spent funds under the TANF block grant. Available at: https://www.cbpp.org/research/family-income-support/state-fact-sheets-how-states-have-spent-funds-under-the-tanf-block (accessed 26 October 2018).

15.

Coleman Institute for Cognitive Disabilities at the University of Colorado (n.d.) State of the states in intellectual and developmental disabilities. Available at: https://www.colemaninstitute.org/state-of-the-states-in-intellectual-and-developmental-disabilities/ (accessed 26 October 2018).

16.

Davis

Schoorman

Donaldson

(1997) Toward a stewardship theory of management. Academy of Management Review 22(1): 20–47.

17.

Dicke

(2002) Ensuring accountability in human services contracting: Can stewardship theory fill the bill? The American Review of Public Administration 32(4): 455–470.

18.

Dilger

Boyd

(2014) Block grants: Perspectives and controversies, 15 July, Congressional Research Service. Available at: http://www.fas.org/sgp/crs/misc/R40486.pdf (accessed 7 December 2014).

19.

Eaton

Weir

(2015) The power of coalitions: Advancing the public in California’s public-private welfare state. Politics & Society 43(1): 3–32.

20.

Edzes

AJE

Rijnks

Dijk

(2013) Spatial implications of using firm level quotas to employ low productive workers. Tijdschrift voor economische en sociale geografie 104(5): 621–629.

21.

Eisenhardt

(1989) Agency theory: An assessment and review. The Academy of Management Review 14(1): 57–74.

22.

Ellison

Newmark

(2010) Building the reservoir to nowhere: The role of agencies in advocacy coalitions. Policy Studies Journal 38(4): 653–678.

23.

Fording

(2018) State ideology data. In: Richard C. Fording. Available at: https://rcfording.wordpress.com/state-ideology-data/ (accessed 1 August 2018).

24.

Friedman

(2016) Day habilitation services for people with intellectual and developmental disabilities in Medicaid Home and Community-Based Services Waivers. Research & Practice for Persons with Severe Disabilities 41(4): 244–255.

25.

Friedman

Rizzolo

(2016) ‘Get us real jobs:’ Supported employment services for people with intellectual and developmental disabilities in Medicaid Home and Community Based Services Waivers. Journal of Vocational Rehabilitation 46(1): 107–116.

26.

Garrett

Jansa

(2015) Interest group influence in policy diffusion networks. State Politics & Policy Quarterly 15(3): 387–417.

27.

George

Bennett

(2005) Case Studies and Theory Development in the Social Sciences. 4th printing edition. Cambridge: The MIT Press.

28.

Gerring

(2008) Case selection for case study analysis: Qualitative and quantitative techniques. In: Box-Steffensmeier JM, Brady HE and Collier D (eds) The Oxford Handbook of Political Methodology. Oxford handbooks of political science. Oxford, New York: Oxford University Press.

29.

Gettings

(2011) Forging a Federal-State Partnership: A History of Federal Developmental Disabilities Policy. Alexandria: American Association on Intellectual and Developmental Disabilities.

30.

Giordono

(2019) Advocacy coalitions in low salience policy subsystems: Struggles under a smooth surface. Policy Studies Journal. DOI: 10.1111/psj.12330.

31.

Guzman

Pirog

Seefeldt

(2013) Social policy: What have we learned? Policy Studies Journal 41: S53–S70.

32.

Hall

Butterworth

Gilmore

, et al. (2003) High-performing states in integrated employment. Research to Practice 32, February. Institute for Community Inclusion, University of Massachusetts, Boston. Available at: https://www.communityinclusion.org/article.php?article_id=121 (accessed 20 July 2018).

33.

Hall

Butterworth

Winsor

, et al. (2007) Pushing the employment agenda: Case study research of high performing states in integrated employment. Intellectual and Developmental Disabilities 45(3): 182–198.

34.

Heikkila

Pierce

Gallaher

, et al. (2014) Understanding a period of policy change: The case of hydraulic fracturing disclosure policy in Colorado. Review of Policy Research 31(2): 65–87.

35.

Hirsch

Michaels

Friedman

(1987) ‘Dirty Hands’ versus ‘Clean Models’: Is sociology in danger of being seduced by economics? Theory and Society 16(3): 317–336.

36.

Huber

Shipan

(2002) Deliberate Discretion: The Institutional Foundations of Bureaucratic Autonomy. Cambridge studies in comparative politics. Cambridge, UK; New York: Cambridge University Press.

37.

Itkonen

(2007) Politics of passion: Collective action from pain and loss. American Journal of Education 113(4): 577–604.

38.

Itkonen

(2009) Stories of hope and decline interest group effectiveness in national special education policy. Educational Policy 23(1): 43–65.

39.

Jenkins-Smith

Nohrstedt

Weible

, et al. (2017) The Advocacy Coalition Framework: An overview of the research program. In: Weible

Sabatier

(eds) Theories of the Policy Process. Boulder: Westview Press, pp.183–224.

40.

Jensen

Meckling

(1976) Theory of the firm: Managerial behavior, agency costs and ownership structure. Journal of Financial Economics 3(4): 305–360.

41.

Jeon

Haider-Markel

(2001) Tracing issue definition and policy change: An analysis of disability issue images and policy. Policy Studies Journal 29(2): 215.

42.

Jones

Jenkins‐Smith

(2009) Trans-subsystem dynamics: Policy topography, mass opinion, and policy change. Policy Studies Journal 37(1): 37–58.

43.

King

Keohane

RORO

Verba

(1994) Designing Social Inquiry: Scientific Inference in Qualitative Research. Princeton paperbacks. Princeton: Princeton University Press.

44.

Krause

(1999) A Two-Way Street: The Institutional Dynamics of the Modern Administrative State. Pittsburgh: University of Pittsburgh Press.

45.

Kübler

(2001) Understanding policy change with the Advocacy Coalition Framework: An application to Swiss drug policy. Journal of European Public Policy 8(4): 623–641.

46.

LEAD Center (n.d.) Employment First webpage. Available at: http://drivedisabilityemployment.org/ (accessed 6 July 2018).

47.

LeBlanc

Tonner

Harrington

(2000) Medicaid 1915 (c) Home and Community-Based Services Waivers across the states. Health Care Financing Review 22(2): 159.

48.

Lindsay

Houston

(2013) Disability Benefits, Welfare Reform and Employment Policy. Work and welfare in Europe. New York: Palgrave Macmillan.

49.

Lofland

(2006) Analyzing Social Settings: A Guide to Qualitative Observation and Analysis. 4th ed. Belmont: Wadsworth/Thomson Learning.

50.

Maxwell

(2013) Qualitative Research Design: An Interactive Approach. 3rd ed. Applied social research methods series, vol. 41. Thousand Oaks: Sage Publications.

51.

McGaughey

Mank

(2001) Empirical evidence of systems change in supported employment. Journal of Disability Policy Studies 11(4): 210–222.

52.

Mettler

SoRelle

(2014) Policy Feedback Theory. In: Sabatier

Weible

(eds) Theories of the Policy Process. 3rd ed. Boulder: Westview Press, pp.151–182.

53.

Nagel

(2006) Policy games and venue-shopping: Working the stakeholder interface to broker policy change in rehabilitation services. Australian Journal of Public Administration 65(4): 3–16.

54.

Nohrstedt

(2011) Shifting resources and venues producing policy change in contested subsystems: A case study of Swedish Signals Intelligence Policy. Policy Studies Journal 39(3): 461–484.

55.

Nord

(2014) Employment First across the nation: Progress on the policy front. Policy Research Brief, 1 March. University of Minnesota, RTC on Community Living. Available at: http://rtc.umn.edu/prb/214/ (accessed 14 September 2015).

56.

Nord

Luecking

Mank

, et al. (2013) The state of the science of employment and economic self-sufficiency for people with intellectual and developmental disabilities. Intellectual and Developmental Disabilities 51(5): 376–384.

57.

Pacewicz

(2018) The regulatory road to reform: Bureaucratic activism, agency advocacy, and medicaid expansion within the delegated welfare state. Politics & Society 46(4): 571–601.

58.

Commonwealth of Pennsylvania (1990) Pennsylvania Bulletin. Available at: http://www.pacodeandbulletin.gov/Display/pabull?file=/secure/pabulletin/data/vol31/31-31/1430i.html&d=reduce (accessed 14 July 2020).

59.

Pennsylvania House of Representatives (2015) HB 2130. Employment First Act. Session of 2015. Available at: http://www.legis.state.pa.us/cfdocs/billinfo/billinfo.cfm?syear=2015&sind=0&body=H&type=B&bn=2130 (accessed 19 October 2017).

60.

Pennsylvania Legislature (2018) Act 36 (HB 1641). Employment First Act. Session of 2018. Available at: http://www.legis.state.pa.us/cfdocs/billinfo/billinfo.cfm?syear=2017&sind=0&body=H&type=B&bn=1641 (accessed 19 October 2017).

61.

Pettinicchio

(2013) Strategic action fields and the context of political entrepreneurship: How disability rights became part of the policy agenda. Research in Social Movements, Conflicts and Change 36: 79–106.

62.

Pierce

Giordono

Peterson

, et al. (2020) Common approaches for studying the Advocacy Coalition Framework: Review of methods and exemplary practices. Social Science Journal. DOI: https://doi.org/10.1016/j.soscij.2019.06.005.

63.

Pierce

Peterson

Jones

, et al. (2017) There and back again: A tale of the Advocacy Coalition Framework. Policy Studies Journal 45(S1): S13–S46.

64.

Pierson

(1993) When effect becomes cause: Policy feedback and political change. In: Esping-Andersen G, Hall P, North DC, et al. (eds) World Politics 45(4): 595–628.

65.

Racino

JAR

(2015) Public Administration and Disability: Community Services Administration in the U.S. Boca Raton: CRC Press, Taylor & Francis Group.

66.

Rogers

(1995) Diffusion of Innovations. 4th ed. New York: Free Press.

67.

Sabatier

(ed.) (2000) Clear enough to be wrong. Journal of European Public Policy 7: 134–140.

68.

Sabatier

(1988) An Advocacy Coalition Framework of policy change and the role of policy-oriented learning therein. Policy Sciences 21(2/3): 129–168.

69.

Sabatier

Jenkins-Smith

(1999) The Advocacy Coalition Framework: An assessment. In: Sabatier PA (ed.) Theories of the Policy Process. 1st ed. Boulder: Westview Press, pp.117–168.

70.

Sabatier

Jenkins-Smith

(1993) Policy Change and Learning: An Advocacy Coalition Approach. Theoretical lenses on public policy. Boulder: Westview Press.

71.

Sabatier

Weible

(2007) The Advocacy Coalition Framework: Innovations and clarifications. In: Sabatier

(ed.) Theories of the Policy Process. 2nd ed. Boulder: Westview Press, pp.189–222. Available at: http://200.41.82.27/108/1/Sabatier,%20Paul,%20Weible%20C_The%20Advocacy%20Coalition%20Framework.%20Innovations%20and.pdf (accessed 8 November 2017).

72.

Seawright

Gerring

(2008) Case selection techniques in case study research: A menu of qualitative and quantitative options. Political Research Quarterly 61(2): 294–308.

73.

Shapiro

(1994) Disability policy and the media: A stealth civil rights movement bypasses the press and defies conventional wisdom. Policy Studies Journal 22(1): 123–132.

74.

Shipan

Volden

(2008) The mechanisms of policy diffusion. American Journal of Political Science 52(4): 840–857.

75.

Shipan

Volden

(2012) Policy diffusion: Seven lessons for scholars and practitioners. Public Administration Review 72(6): 788–796.

76.

Skocpol

(1995) Protecting Soldiers and Mothers: The Political Origins of Social Policy in the United States. Reprint edition. Harvard: Belknap Press, an imprint of Harvard University Press.

77.

Squire

(2007) Measuring state legislative professionalism: The squire index revisited. State Politics & Policy Quarterly 7(2): 211–227.

78.

Squire

(2017) A squire index update. State Politics & Policy Quarterly 17(4): 361–371.

79.

Stigler

(1971) The theory of economic regulation. The Bell Journal of Economics and Management Science 2(1): 3–21.

80.

Thompson

Burke

(2007) Executive federalism and medicaid demonstration waivers: Implications for policy and democratic process. Journal of Health Politics, Policy & Law 32(6): 971–1004.

81.

University of Massachusetts Boston, Institute on Community Inclusion (n.d.) StateData.info: Population Data from the American Community Survey (Post 2007), Cognitive Disability. Available at: www.statedata.info (accessed 28 October 2018)

82.

University of Michigan Law School (n.d.) Civil Rights Litigation Clearinghouse. Available at: https://www.clearinghouse.net (accessed 15 February 2018).

83.

U.S. Census Bureau (n.d.) American FactFinder - Results. Available at: http://factfinder.census.gov/faces/tableservices/jsf/pages/productview.xhtml?src=bkmk (accessed 16 October 2015).

84.

U.S. Department of Labor (n.d.a.) Employment First webpage. Available at: http://www.dol.gov/odep/topics/EmploymentFirst.htm (accessed 15 March 2015).

85.

U.S. Department of Labor (n.d.b.) Subminimum wage. Available at: https://www.dol.gov/general/topic/wages/subminimumwage (accessed 20 August 2018).

86.

van der Heijden

Kuhlmann

Lindquist

, et al. (2019) Have policy process scholars embraced causal mechanisms? A review of five popular frameworks. Public Policy and Administration: 0952076718814894. DOI: 10.1177/0952076718814894.

87.

Van Slyke

(2007) Agents or stewards: Using theory to understand the government-nonprofit social service contracting relationship. Journal of Public Administration Research & Theory 17(2): 157–187.

88.

Vornholt

Villotti

Muschalla

, et al. (2018) Disability and employment – Overview and highlights. European Journal of Work and Organizational Psychology 27(1): 40–55.

89.

Walsh

Seward

(1990) On the efficiency of internal and external corporate control mechanisms. The Academy of Management Review 15(3): 421–458.

90.

Washington Department of Social and Health Services (2004) WA Policy Directive 4.11. County services for working age adults. Available at: https://www.dshs.wa.gov/sites/default/files/DDA/dda/documents/policy/policy4.11.pdf (accessed 23 May 2017).

91.

Washington Legislature (1998) Chapter 216 (SSB 6751). AN ACT relating to stabilizing long-term care for persons with developmental disabilities living in the community and in residential habilitation centers. 1998 Regular Session. Available at: http://app.leg.wa.gov/billsummary?BillNumber=6751&Year=1997 (accessed 15 November 2017).

92.

Washington Senate (2012) Chapter 49 (SSB 6384). AN ACT relating to ensuring that persons with developmental disabilities be given the opportunity to transition to a community access program after enrollment in an employment program. Laws of 2012. Available at: http://app.leg.wa.gov/billsummary?BillNumber=6384&Year=2012 (accessed 4 May 2017).

93.

Washington State Department of Social and Health Services and Washington State Health Care Authority (2015) Washington State Transition Plan for New HCBS Rules.pdf. Available at: https://www.dshs.wa.gov/sites/default/files/DDA/dda/documents/Washington%20State%20Transition%20Plan%20for%20New%20HCBS%20Rules.pdf (accessed 16 November 2017).

94.

Wehmeyer

(2013) The Story of Intellectual Disability : An Evolution of Meaning, Understanding, and Public Perception. Baltimore: Brookes Publishing Co.

95.

Weible

(2008) Expert-based information and policy subsystems: A review and synthesis. Policy Studies Journal 36(4): 615–635.

96.

Weible

Cairney

(2018) Practical lessons from policy theories. Policy & Politics 46(2): 183–197.

97.

Winsor

Hall

Butterworth

, et al. (2006) Pushing the integrated employment agenda: Case study research in Washington State. Available at: http://www.communityinclusion.org/article.php?article_id=173 (accessed 2 May 2017).

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.42 MB