Abstract

Organizational variables, including policies, practices, collaborations, and funding mechanisms resulting in high performance in integrated employment, were described through case study research in 3 states. Findings address how contextual factors, system-level strategies, and goals of the system are related as well as how they sustain systems change. Strategies such as flexibility in funding and practices; communication of values through data, rewards, and funding incentives; and innovation diffusion through relationships and training were most successful when they were embedded within the context of a solid values base, a network of dedicated stakeholders, and clarity about systemic goals. Implications are presented with respect to state systems, community rehabilitation providers as partners in planning, and future leadership in the field.

Enabling people with disabilities to enter the labor market is a “priority concern” for federal and state policymakers (Silverstein, Julnes, & Nolan, 2005). President George W. Bush's New Freedom Initiative articulates the commitment of lawmakers to increase access and achieve better employment outcomes. Policy shifts over 2 decades establish an increasing emphasis on integrated employment, and the federal government has set the tone for broad-based systems change (Rogan, Novak, Mank, & Martin, 2002). From the model of supported employment in the Developmental Disabilities Act Amendments of 1984, the Rehabilitation Act Amendments of 1986, and the Americans With Disabilities Act, to more recent legislation, such as the Workforce Investment Act and the Ticket to Work and Work Incentives Improvement Act of policies and initiatives have been legislated to try to eliminate the barriers traditionally precluding individuals with disabilities from working.

Even with these numerous pieces of legislation, there is a significant gap when comparing employment rates of people with and without disabilities. The 2004 American Community Survey estimates that 37% of working-age adults with disabilities are employed compared with 74% of people without disabilities (Weathers, 2005). For people with intellectual and developmental disabilities specifically, the disparity widens. In Fiscal Year (FY) 2003, only 26% of individuals supported by community rehabilitation providers worked (Metzel, Boeltzig, Butterworth, Sulewski, & Gilmore, 2007), and data suggest that those who are employed often work limited hours with low wages (Mank, 2003). Recent data from the Institute for Community Inclusion (n.d.) indicate that the majority of services provided by state agencies that support individuals with intellectual and developmental disabilities are facility-based (59%) and that the expansion of integrated employment has slowed. Although the number in integrated employment increased from 98,315 in 1996 to over 108,000 in 2001, states reported the same percentage (23%) in integrated employment in FYs 1996 and 2001. At the same time, participation in sheltered employment and nonwork services grew steadily, suggesting that employment services continue to be viewed as an add-on service rather than a systemic change (Mank, 2003).

Although as a nation resources and priorities have not realigned to expand employment, individual states have progressed. In FY 2001, Connecticut, New Hampshire, and Washington reported that 50% of individuals receive day and employment services in integrated employment. Six other states reported more than 40% in integrated employment (Institute for Community Inclusion, n.d.). Fourteen states reported less than 20% in sheltered employment in FY 2001.

The variation in outcomes illustrates the importance of state-level factors regarding integrated employment (McGaughey & Mank, 2001). State agencies can prioritize integrated employment by establishing policies and practices that firmly support it. Institute for Community Inclusion researchers have found that characteristics common among high-performing states include clearly defined goals and data collection, strong agency leadership, interagency collaboration, ongoing training and outreach, communication through relationships, local control and flexibility, and respect for innovation (Cohen, Butterworth, Metzel, & Gilmore, 2003). Washington's County Guidelines (Washington State Department, 1992) provided a clearly defined vision, a foundation for planning, and strategies for using funds to achieve state priorities. Goals related to employment were specific and explicit. Similarly, Tennessee recently introduced an Employment First! initiative, and Florida established a state plan to increase employment participation by 25%. Because states differed in both their approaches and commitment to integrated employment, it is important to understand how change occurs and to identify strategies and factors that facilitate success.

Our purpose in this study was to understand how factors work together to produce successful employment outcomes. In this paper we describe the organizational factors that can lead to high performance among states; explain how contextual factors, system-level strategies, and system goals contribute to systems change; and provide policymakers and advocates with practical strategies for maximizing employment outcomes.

Method

A multiple case study approach was chosen to study 3 states. Case studies provided an understanding of the complex issues facing stakeholders as they worked to provide high quality day and employment supports. We chose this methodology because it enabled us to analyze multiple sources of data, including secondary data, state policy and planning documents, and key informant interviews. Case studies can be used on different analytic levels and can be focused on specific and factual descriptions of each site or produce cross-comparisons between sites (Stake, 2000). A cross-comparative analysis allowed us to focus on themes that emerged across several sites, suggesting patterns and trends in service delivery. In addition, case study research has been successfully employed in other studies in which investigators have specifically studied how organizations are able to transition from facility-based supports to community employment for individuals with intellectual and developmental disabilities (Butterworth, Fesko, & Ma, 2000; Murphy & Rogan, 1995) as well as how organizations can innovatively and effectively support individuals in the community (Taylor, Bogdan, & Racino, 1991). Although the present case studies are focused at a state level, their goals are similar and we can contribute to theory building by comparing case studies of different systems and organizations over time (this is addressed in more detail in the Discussion section).

Sample Selection

All states were ranked based on performance criteria. A high performing state was defined as having a high rate (i.e., number of people in integrated employment per state population) of individuals with intellectual and developmental disabilities in integrated employment, and/or rapid growth in integrated employment over time. (Percentage is based on the total number of individuals enrolled in day and employment services throughout the state.) In this study integrated employment was defined as participation in competitive employment or supported employment. Consistent with the Rehabilitation Act of 1973 (as amended), competitive employment as used here means work in the competitive labor market that is performed on a full- or parttime basis in an integrated setting and for which an individual is compensated at or above the minimum wage, but not less than the customary or usual wage [34 CFR 361.5(b)(11)]. Competitive employment services include time-limited job-related supports or job placement services. Supported employment refers to ongoing job-related supports provided to workers with a disability in order to maintain employment and includes individual jobs and group-supported employment, including enclaves or mobile work crews (Butterworth, Gilmore, Kiernan, & Schalock, 1999).

Using these definitions, we evaluated states based on outcome data from Institute for Community Inclusion's national data collection for day and employment services. We selected 13 states for initial screening interviews (see Table 1). These states rose to the top when we ranked all states according to rate of integrated employment, percentage in integrated employment, and rate of change in integrated employment over time. (For findings related to this preliminary study, see Cohen et al., 2003.)

Table 1

Criteria Used to Determine High Performance in Integrated Employment (IE)

Criteria Used to Determine High Performance in Integrated Employment (IE)
Criteria Used to Determine High Performance in Integrated Employment (IE)

We chose Colorado, New Hampshire, and Washington for further study after considering state size, demographic characteristics, geographic location, and preliminary findings from a thematic analysis of the data collected through this screening process. In addition, the team came to a consensus on the states that would yield the most valuable data for policymakers, administrators, and community rehabilitation providers as they work to improve integrated employment outcomes.

Data Collection

We conducted semi-structured telephone interviews with two to three respondents in each state as part of the screening process in 2001–2002. These respondents were administrators within state agencies that serve individuals with intellectual and developmental disabilities, community rehabilitation providers, leaders, and/or advocates. A snowball sample was used for recruitment. A researcher telephoned potential respondents, explained the study, and invited the respondent to participate. We compiled written field notes from the phone interviews, yielding data that demonstrated some key components of the states' success.

Between 2002 and 2004, two researchers conducted in-person individual and group interviews in each state, with follow-up interviews conducted by telephone. Principal data collection occurred during a 3- to 4-consecutive day period in Washington and Colorado, and in most cases two researchers participated in each interview. In New Hampshire interviews were spread out over several day trips to the state because of its proximity to the researchers. Throughout this time period, we focused on one site at a time. Thus, once data were collected in New Hampshire, we developed a descriptive case analysis (Yin, 1994). Once preliminary analysis was completed at this site, we moved on to the next state, and so on. Once all site visits and initial descriptive case analyses were completed, we conducted a comprehensive cross site analysis. In initially choosing the sites, the team used the most current employment outcomes available and focused subsequent case study research on respondents' views about the factors that had led to their state's progress in integrated employment in recent years. As such, we do not believe that the difference in the reporting period for the employment outcomes and data-collection period compromised the quality or validity of the data.

Interview participants included administrators from agencies that serve individuals with intellectual and developmental disabilities, vocational rehabilitation administrators, community rehabilitation providers, advocacy group members, such as Arc personnel and members of the Association for Persons in Supported Employment, state developmental disabilities councils, parents, and training providers. In total across the sites, 46 individuals were interviewed. For a more detailed description of sample members, see Table 2.

Table 2

Respondents by State and Affiliation

Respondents by State and Affiliation
Respondents by State and Affiliation

Data Analysis

In this study we used a qualitative research design with a case study methodology to gather the richest possible data. The case study method is most useful when conducting a “holistic investigation of some space- and time-rooted phenomenon” (Lofland & Lofland, 1995, p. 21). Four features of naturalistic inquiry aided in designing the study and provided methodological rigor: credibility, dependability, confirmability, and transferability (Lincoln & Guba, 1985). Credibility of the data addresses how accurately findings reflect what is being studied. Prolonged and persistent engagement and triangulation (through multiple sources and multiple investigators) ensured that the data were credible (Lincoln & Guba, 1985). We also used peer debriefing and member checks (having participants review and confirm findings). Dependability of the data was provided through a written record, including transcripts and memos. Field notes where researchers' observations were recorded provided a comprehensive audit trail.

Bogdan and Biklen (1992) described qualitative analysis as “the process of systematically searching and arranging the interview transcripts, field notes, and other materials that you accumulate to increase your own understanding” (p. 153). We used QSR NUDIST 4.0 for the Macintosh (1997), designed specifically for qualitative analysis, to sort data and conceptualize themes. The techniques used to analyze the data included coding (using labels to assign meaning to data) and memo-writing (i.e., conceptual write-ups of ideas about the data and relationships within it that tie together bits of information to show a general concept—Miles & Huberman, 1994). Researchers developed a preliminary list of codes during the initial stages of data collection based on their existing knowledge and the protocol that guided the interviews. This list was continually revised based on emergent data. Operational definitions for codes were developed to ensure a shared meaning. Consensus was reached through the researchers' independently coding the same transcript and then meeting for reconciliation. The authors met regularly to discuss the emerging data. Memo-writing was also a communication mechanism among researchers throughout analysis.

Findings

Participating Case Study States

As a backdrop for the study's major findings, it is useful to provide some contextual information about each state's employment system before comparisons are made. Participating states vary in their geographic location (New England, Southwest, Pacific Northwest) and administrative structure. The total number of individuals enrolled in employment and day services in 2004 varied from 2,100 in New Hampshire to 8,043 in Washington (Institute for Community Inclusion, 2006). The extent to which employment is an explicitly stated goal of the system also differs among states. Policy documents in Washington are the most clear in identifying outcomes, while system goals are less clearly specified in Colorado and New Hampshire.

The three states share a decentralized structure for contracting and management of services, although the mechanisms vary. In New Hampshire services are managed and in some cases provided by 10 area agencies that are autonomous nonprofit corporations. The state Bureau of Developmental Services contracts service delivery functions to the area agencies. In Washington employment and day services are managed by county personnel under the policy directives provided by the County Guidelines. Finally, in Colorado, community-centered boards are private nonprofits that act as the single entry point into the long-term supports; the state contracts with these boards for community-based services (see Table 3).

Table 3

Description of States' Integrated Employment Systems

Description of States' Integrated Employment Systems
Description of States' Integrated Employment Systems

Findings as They Relate to the Conceptual Model

The Findings section is reflected in the conceptual model. The model was developed to explain how factors in the data contribute to successful integrated employment outcomes.

The model demonstrates the multidirectional relationships among elements. In the model, the catalysts and context identified as critical were values, key players, and clarity of goal. Variability in these contextual factors was also observed when a state experienced a shift in employment outcomes, lending more evidence to the importance of these factors in sustaining change efforts. These inputs enabled the success of systemic level strategies that were observed, such as flexibility in funding mechanisms and policy directives; a system of funding incentives; communication of goals through data; and innovation diffusion through training. Relationships and local control were observed as facilitators of these strategies. Because the inputs and catalysts lay the foundation for the strategies, they are set in motion by facilitators, moving the system forward in achieving its goals. Figure 1 depicts the relationship between all of these factors, as well as the synergy and movement necessary for sustaining long-term systems change.

Figure 1

Conceptual model for high performance variables

Figure 1

Conceptual model for high performance variables

Catalysts and Context

Catalysts and context provide the foundation for the model and serve to frame system-level strategies. The circular placement of the values, key players, and clarity of goals suggest a fluid relationship of factors that continually interact in these high-performing systems.

Values

“Mostly I believe that the only safeguard for people with developmental disabilities is how people think about them. … if we don't have impact on values you put people at risk,” said one study respondent. Respondents believed that early and explicit system-wide values were responsible for their state's employment outcomes. Values were a blueprint for the service delivery model, transmitted through an early period of widespread training. States used their values system to redirect and continually refine the foundation for service delivery. Community inclusion was cited as the driving value for the service system, with work as one way for inclusion to occur.

Early values change efforts, particularly in New Hampshire and Washington, sprang from the coinciding forces of deinstitutionalization and the principles of normalization in the 1970s. These values were instilled through intensive values training and the integration of people with intellectual and developmental disabilities into their communities. Informants related how early training, specifically Wolfensberger's teachings on normalization, led to a shared understanding of the importance of community involvement.

Program Analysis of Service Systems, or PASS training (Wolfensberger & Thomas, 1983), was noted by respondents as a building block for underlying values in Washington and New Hampshire specifically. Workshops focused on the importance of community inclusion and were widely attended over several years. Many current leaders in Washington participated as facilitators in early PASS training. An administrator from Washington's Division of Developmental Disabilities noted, “We wanted people singing off the same sheet of music. And you needed to go through PASS training in order to have that real solid basis.” This evolving environment continually refined its focus on supporting individuals in inclusive communities. It was a time of change, growth, and greater expectations about the types of supports that should be made available.

Key Players and Their Relationship to a Shared Value System

Values were sustained and transmitted through key stakeholders who were committed to the system. This was evidenced in part by their longevity and strong working relationships, specifically in Washington and New Hampshire. A group of leaders gelled; they were connected to the values and, consequently, to one another. Individually and collectively, they held steadfast to the ideals of community integration and meaningful contributions. One informant commented that “the way things develop[ed] can be based on a handful of people and a handful of personalities that share the same values.” Important state and local staff have stayed through the development of the community system and have worked to continue its progress. A long-standing network of stakeholders in state and county government, community rehabilitation providers, and the advocacy community emerged as a result of values-based training. In Washington, many leaders began their careers concurrent with the introduction of integrated employment and experienced similar indoctrinations to the field. This created a community of long-term change agents. A respondent said of this formative time, “It was new and different and we had a charge and a mission.”

Clarity of Goals

Goals emerged through the values base that each state established. The clearer the goal, the more likely the system seemed to achieve it. In New Hampshire, driven by the Laconia State School class action lawsuit in 1978, the state invested heavily in values-based training as the transformation to a community-based service system began. This remained the principle focus of service delivery. Since 1984, New Hampshire has not provided funding to open new sheltered employment programs, and in 1985, a systems change grant with the state vocational rehabilitation system helped to spur the closure of sheltered workshops.

In Washington, the consistency of stakeholder commitment to employment as the primary goal of day supports was striking. The county system, which manages day and employment supports, did not support traditional facility-based nonwork day programs. Currently, less than 40 individuals statewide receive Adult Day Health services. Moreover, some counties in Washington, such as Clark County, had explicit goals to reduce or eliminate sheltered employment.

Also apparent was that when goal clarity was absent, there was an observable negative effect on outcomes. Through the mid 1990s, Colorado was a national leader in integrated employment. In 1997, momentum began to wane. Several factors contributed to this decline, including the elimination of a state staff person dedicated to promoting integrated employment, an abandoned moratorium on new funding for sheltered work, and a decline in funding for training and technical assistance. In addition, a statewide systems change effort altered the funding structure significantly, eliminating a previous financial incentive for employment. The systems change initiative was also part of the impetus for the growth of a new service option, community participation, which was conceived as community connection programs but became what most agreed was “van therapy.” In many cases, community participation was an outings program valued for its perception by families as a safer and more stable alternative to employment.

Following a 2002 statewide tour of the system, the director of Colorado's Division for Developmental Disabilities developed a strategic plan to address concerns related to service provision. One result was a committee focused on employment. The group's goals centered on raising the priority for integrated employment, ensuring equality of opportunity for all individuals to participate in paid employment, and promoting the use of natural supports in the workplace. The group developed a consensus definition of integrated employment, an important step in moving towards a clarified vision.

System Level Strategies

States used a variety of strategies to further the goal of integrated employment, including flexible policies, flexible funding, funding incentives, communication of goals through data, and innovation development and diffusion through training.

Flexible policy guided service expectations

Both Washington and New Hampshire had minimally perceived regulations for employment. They took a flexible approach guided by the state's overall values in creating expectations for services. Washington's integrated employment system follows their County Guidelines. This document details the requirements for how Washington's Division of Developmental Disabilities contracts with employment service providers. The guidelines have a clear focus on employment as the preferred outcome but function as a flexible policy framework and not a contractual obligation: “The County Guidelines are used to set policy and direction—there [are] not a lot of other mandates [from the state].”

Washington's Division of Developmental Disabilities recently issued a new policy that took effect July 1, 2006. The Working Age Adult Policy:

designates employment supports as the primary method of furnishing state-financed day services to adult participants … services for persons under the age of 62 that do not emphasize the pursuit or maintenance of employment in integrated settings can be authorized only by exception to policy (Washington: Department of Social and Health Services, Division of Developmental Disabilities, “County Services for Working Age Adults” Policy 4.11).

Representing an evolution of the policy framework established in the County Guidelines, the Working Age Adult Policy represents a clear commitment of state resources.

Informants also characterized New Hampshire's Bureau of Developmental Services as operating under minimal state regulations. Respondents perceived New Hampshire as a state based on small towns and local control, and Bureau of Developmental Services tried to operate accordingly. Said one Bureau of Developmental Services official, “We have a couple of general regulations which … set out the parameters for how the system should be run. But we really delegate all the decision-making to the local area agencies.”

Flexible policy-supported innovation

Due to the flexibility of the states' systems, providers felt free to manage employment supports to best meet individual needs. One New Hampshire provider commented, “The state doesn't really tell us. They ask us. Honest to God, they sometimes ask us. … the state doesn't dictate what we do.” New Hampshire's and Washington's providers felt that the state supported innovative or nontraditional approaches. Both states created an environment where the fear of doing something wrong was replaced by a visible support for creativity and risk-taking in service delivery.

Flexible funding

The Washington Division of Developmental Disabilities functions as a bifurcated system. State level staff are responsible for residential and case management services, while county staff focus on employment services. Each county has the authority to designate its own funding structure. Several counties in Washington provide funding for integrated employment through block contracts. Stakeholders felt that the structure of these contracts allows providers the flexibility to allocate agency resources based on the changing needs of individuals without having to convene formal meetings with case managers to request a change in a person's funds. The ability to be adaptable with funding allocations has permitted agencies to support a variety of people in integrated employment, including those with more intense needs. Mirroring Washington's outcomes, Colorado's funding system prior to the state's systems change initiative was designed to allow providers flexibility in meeting individual needs. Although the formal dollar amount per person allotted to the agency was determined by the state, at the local level providers could pool their financial resources and support each person individually.

In Washington, respondents spoke of some counties that used block funding, and some that were moving towards individualized funding, where each person controls his or her own budget. Some interviewees expressed concerns about individualized funding actually limiting agencies' flexibility to support all individuals well. Respondents felt that individualized budgeting did not allow for organizational costs, such as marketing, which is critical to building relationships with employers. The variability in approaches suggest that both funding mechanisms have the potential for success as long as they are accompanied by clear outcome measures and a willingness to hold providers to that standard.

Funding across service categories

Counties in Washington that used block contracts emphasized the flexibility that this structure provided in shifting resources to meet individual needs and to simultaneously fund multiple services, including job development and sheltered employment. This funding was important in encouraging providers who were hesitant to accept payment only for employment services while concurrently providing job development and sheltered employment services, in transitioning individuals from sheltered to integrated work.

Similarly, New Hampshire's flexible approach to funding has permitted access to multiple funding sources. At a contract level, residential and day service funds can be combined to integrate resources across major life areas. The integration of funding streams allows the state to overcome the typical barriers that arise when services need cross-categorical funding.

Multiple funding sources

Cross-agency funding was another strategy that New Hampshire used to ensure flexibility in its funding structure. The directors of the Department of Vocational Rehabilitation, the Division of Developmental Disabilities, the Department of Mental Health, the Job Training Council, and the Developmental Disabilities Council collectively developed a policy to allow their organizations to work together at the local level to fund integrated employment. One respondent commented that this “created a ‘synergy’ of money that was more effective at meeting individuals' needs than if each agency had tried to solely fund supported employment services.” The ability to easily integrate multiple funding sources signals both states' commitment to providing and supporting comprehensive services. In Colorado, from 1988 until 1997, a program connected Community Centered Boards (local agencies responsible for authorizing services for individuals with intellectual and developmental disabilities) with their vocational rehabilitation offices for jointly supported individuals. This included blending or braiding the agencies' funding to share the cost of employment services. Despite the end of this formal agreement, some Community Centered Boards and local vocational rehabilitation agencies have continued this collaboration.

Funding Incentives

Data showed that states were willing to differentially support providers who were successful in helping people find and maintain employment. This was accomplished through informal rewards and up-front funding for services.

Informal rewards

Washington counties that used the block contract funding structure informally awarded providers who had high rates of quality outcomes with larger blocks of funding. This prestige and financial flexibility further increased their desire for innovative employment programs. Similar to Washington, New Hampshire had an informal understanding that the state allocated additional resources to providers that were more effective at meeting the broad goals of community inclusion and employment.

Up-front funding of services

In Colorado, the system used financial incentives to encourage employment growth. Prior to changes in Colorado's funding structure in 1997, providers that enrolled individuals in integrated employment supports received a 25% higher rate of funding from the state than they received for sheltered workshop services. Colorado's traditional funding guidelines were outcome-based, and providers did not receive payment for employment services until after an individual obtained a job. Under these guidelines providers were expected to simultaneously fund job development and day supports without additional payment, leaving many financially hesitant to pursue integrated employment. In response to this, the 25% funding incentive was provided up-front to offset the cost of job development and initial employment supports. As one respondent noted, “As a provider, the more people I could get successful jobs, the more money I would have to maneuver.” More recently, Colorado instituted a bonus system of $1500 to providers for each individual whom they supported in a job of his or her choice, for 20 hours or more per week, that paid at least minimum wage and lasted for 6 months. This incentive was not perceived to be as effective as the previous up-front reimbursement because agencies had to allocate funds for the costs associated with job development with no guaranteed return on their investment.

Communication of Goals Through Data

The use of individual and provider level outcome data is one way in which high performing states have communicated the importance of employment. A respondent in Colorado noted that the state's decline in integrated employment has in part been caused by minimal data collection and an inadequate system for tracking progress and creating measurable goals. Another respondent reported that advocacy agencies were particularly interested in the data from Colorado's Division of Developmental Disabilities in order to understand their recent decline in employment and increase in community participation services. The importance of data collection was two-fold:

The main benefit is to hold people accountable. … The other is to market our efforts to people, to Vocational Rehabilitation, to the legislature. If there's one way we could get more money out of them [it would be if we said] “This is a system that's working. This is a system that's putting people to work. They're becoming tax-paying citizens.”

New Hampshire used individual outcome data as the basis for discussion with providers. Since the mid-1990s in New Hampshire, the data collection system has tracked individual earnings, hours of employment, benefits, level of workplace integration, and satisfaction every 6 months. Multiple stakeholders, including several local providers and all 12 area agencies, helped develop the extensive employment data system. This joint effort created a sense of ownership for the provider community. Data are used routinely as the basis for discussion about how to improve outcomes. Provider informants consistently spoke positively about data collection and genuinely appreciated the state's communication through the data: “[It] certainly had an impact on us because … without the data, we would not have been able to say whether [we were improving].”

Innovation Diffusion Through Training

Formal training supported by state agency administrations helped to diffuse innovative employment techniques throughout state systems. Colorado's Rocky Mountain Resource and Training Institute was developed in the 1980s to provide statewide technical assistance and training in employment. Although this group is no longer operating, the Center for Technical Assistance and Training located in Denver has provided training through the Region VIII Community Rehabilitation-Continuing Education Program since 1995. The Center for Technical Assistance and Training focuses upon the employment of people with disabilities and organizational development and management in agencies that provide services to people with disabilities. The Colorado chapter of the Association for Persons in Supported Employment, an advocacy organization, also conducts training on integrated employment and holds conferences to educate stakeholders on the benefits of and strategies in successful integrated employment. Specialized training is designed for vocational rehabilitation counselors, Community Centered Boards, casemanagers, families, school systems, and employers.

Similarly, in Washington several groups have been active in shaping training and technical assistance. The Washington Initiative for Supported Employment was created in 1985 as part of the Rehabilitation Services Administration-Funded Supported Employment Systems Change grant. This group receives funding from the counties and the state to provide training and individualized technical assistance. O'Neill and Associates is an additional training resource funded by the state and counties to encourage innovative employment techniques. At the time of these interviews, O'Neil and Associates was receiving funds to provide values-based and integrated-employment training, and they served as a broker for external expert resources. This group has held state and county training contracts for over 20 years. The concentration of training and technical assistance funds has allowed Washington to import nationally known trainers and coordinate employment forums. For many years the state has hosted the Ellensburg Employment Conference as a chance for all levels of staff members to learn about innovations in the field. Typically attracting from 600 to 800 attendees per year, this conference has provided an opportunity for people across the state to share new techniques, ideas, and strategies.

New Hampshire's 1986 award of a 5-year systems change grant through Rehabilitation Services Administration also provided funding exclusively for training and conferences at the national, state, and regional levels for area agency and state administration staff members and providers. Attendance at national conferences by local case managers and vocational rehabilitation counselors exposed these groups to innovative employment strategies and allowed staff to identify new types of training they wanted to bring to New Hampshire. One respondent described meeting an advocate from Montana with extensive experience in developing employment opportunities in rural areas. New Hampshire used systems change funding to arrange for this individual to train providers in remote regions of the state.

In 1998, New Hampshire's structure for formalized training was cemented in the Northern and Southern Training Collaborative. Previous training had been conducted through the Bureau of Developmental Service's central office, and although successful, the structure made it difficult to address the specific strengths and weaknesses of all communities. Through the Collaborative, 6 southern and 6 northern regions met periodically to identify their training needs for the fiscal year and determine regional priorities. Funding from the Bureau of Developmental Services enabled area agencies to bring in national trainers that they would otherwise be unable to afford.

Goals

The goal of the system is the model's most fundamental aspect. Washington State as a whole is noted for its emphasis on the importance of work as the preferred service outcome. Officials in the state made the assumption that everyone could work and contribute to their communities. All work is highly valued in Washington, and even sheltered work is preferred over community-based nonwork activities. The combination of the values of work and inclusion sets the standard that integrated work is the priority service option for people with intellectual disabilities. Respondents noted that the “state culture supports the idea of being individually productive” and together with the idea of inclusion, “the argument becomes about where a person works and not if he works.”

In New Hampshire although the goal was broader, employment was seen as a key element. According to the New Hampshire Bureau of Developmental Services' director, “Certainly employment is an important part of that, but it hasn't been the primary mission of the various area agencies as much as making somebody's life meaningful.” Another respondent said:

The goal isn't to get people jobs. A job is an objective. The real goal is to help people become more interdependent and valued members of their communities. A job is perhaps the most powerful vehicles [sic] for doing that, but it's only a vehicle. … I think the state recognizes the difference between the goal and the objective.

Colorado did not have as clear a distinction in its primary goals of work compared to community inclusion but, as stated earlier, was working to articulate its goal and mission for day and employment supports. What was apparent through the data in Colorado, however, was an emphasis on self-determination, both as a service paradigm and overall goal for all agency-provided supports, including day and employment.

Facilitators

Several elements were identified that enabled the strategies to be successful. These factors were thought of as “facilitators” of the process; they provided a context and environment that was hospitable to increasing employment. These factors were local control and relationships.

Local control

Although a state-level commitment to integrated employment is critical, practical implementation of employment supports happens at the local level. In Washington, the bifurcated system allowed one level of government to concentrate on integrated employment. Day supports are managed at the county level, while case management and living supports are managed at the state level. This has helped County Coordinators focus their efforts on supporting local, community-based employment opportunities. The county structure has also provided a source of funding for innovation. County property tax dollars, representing a small percentage of the county Division of Developmental Disabilities office budget, provide a flexible resource that has been used for demonstration projects and technical assistance activities.

County tax dollars have been a boon to integrated employment in Colorado as well. At the time of the site visit, Denver received $7 million through this funding stream, which was under local discretion and not monitored by the state.

In addition, pilot programs were established by counties to increase the knowledge base in Washington. These programs were an opportunity to try new and creative ideas and because of their inherently preliminary nature, were allowed the time to evolve before their outcomes were evaluated. When pilots were evaluated positively, these initiatives spread to other counties. One example of this is King County's program to employ people with disabilities in county jobs. Lobbying by Provider 2020, a Washington provider advocacy group, created legislation that ensured these workers did not count against the county's fulltime employee allotment, meaning that departments could hire an employee with a disability despite not having an official position vacancy. The county also hired a full-time job developer who could train county employees to support coworkers with disabilities. Similar initiatives are active now in other counties and within state government positions.

In New Hampshire, an emphasis on local control has contributed to opportunities for innovation in the design of its service system. Services are managed by 12 area agencies that are independent nonprofit corporations. Area agencies and local providers are important sources of creativity and commitment to integrated employment. The state encourages these ideals but assumes that services are best organized at the local level.

Colorado's system is comprised of 20 local level Community Centered Boards, some of which are actually providers of services, while others contract their services to other local organizations. Community Centered Boards serve as a single point of entry for a geographic area where persons with intellectual and developmental disabilities can receive services, eligibility determination, and ongoing case management. The autonomy of Community Centered Boards is indicative of a larger service philosophy rooted in local control. As one respondent from Colorado said, “Every Community Centered Board does it differently and decides how they're going to contract with providers and provide quality assurance and oversight.” For instance in Denver, providers receive a higher rate for supported employment services than any other service. Another respondent noted, “Local Community Centered Boards set their funding rates, decide how they're going to pay … really decide how much of their resource is going to go to which program.”

Relationships

One innovative aspect of New Hampshire's service system is its use of personal relationships to facilitate employment. Informants indicated that the state structure and its size (approximately 22 state personnel worked for Bureau of Developmental Services) made state leaders very accessible. Because program specialists spent so much time with the area agencies, the state was able to personally help area agencies and providers identify strategies for employment innovations. One informant described the program specialists as being “very, very good that way. … they really push [providers] to excel. They get people to think the hard questions.”

The development of strong working relationships between program specialists and providers has been a key element in New Hampshire's ability to improve the quality of its employment outcomes. One provider commented on the capacity of program specialists to praise providers for their progress, yet continually challenge them to do more. A Bureau of Developmental Services staff member said:

You guys [providers] have come a long way in the last 5 years. Things have really been improving. There is a lot more that can be done and this is a way we think we can help you in this process. … And I don't think when they present it that way, there's no way providers can say, “We really don't want to do more.”

In Washington, the longevity of both the county coordinators and the providers, as well as the development of mutually supportive relationships between the two, have each influenced growth in employment outcomes. People have noted that early growth in employment was related to trust within the system. Pilot programs were funded by counties to meet specific local needs, with the independence to try new and creative ideas. It appears that the exploration of nontraditional strategies through these pilots is the result of the state's trust of the providers to produce good outcomes. This trusting relationship has grown over time so that providers who consistently produce quality outcomes are given more freedom for creativity.

In 2000, realizing that growth in integrated employment in New Hampshire had stalled, the Bureau of Developmental Services aggressively invested in the use of personal relationships by recruiting an external stakeholder to become a consultant resource. John Vance, the director of ACCESS, a small individualized support organization in New Hampshire, was hired on a half-time basis to conduct individualized technical assistance to providers across the state. Vance's role at Bureau of Developmental Services was to establish working relationships with providers to expand the rate and quality of employment services, in part through demonstrating how job development was done on a person-by-person basis. Vance emphasized real-life demonstrations, one Bureau of Developmental Services staff person said:

John just does it [employment] and people see. He will go in and work with one individual–and this is often the most challenging person for the agency to find employment for—and he does it and the providers see it and so then he will work with another person. He doesn't have to work with too many before they get the idea. So it's leadership of a different kind, it's modeling. Getting the message out. I think it's far more powerful than us just lecturing to people or all the trainings that have been provided.

Vance was noted for working with organizations over an extended period of time, addressing both values and strategic change. He defined his role as building excitement and commitment. Speaking about his experience with one provider he commented, “She was reluctant at the first meeting—but—I think because I would leave her with more ideas than she had when I came, she saw it as a net gain.”

Discussion

Within the framework of federal legislation that promotes the inclusion of individuals with disabilities into the workforce system, states vary widely in their commitment to integrated employment. Despite the clarity of the Developmental Disabilities Act, there is not a well-defined mandate for employment in federal or most state funding sources that support persons with intellectual and developmental disabilities (Wehman, Inge, Revell, Butterworth, & Gilmore, 2005). Research findings suggest the factors that influence whether states comply with federal mandates include leadership, the alliance between agency mission and federal priorities, communication, resource allocation and limitations, and competing priorities (Rogan et al., 2002). Underlying these factors is the context of values within which each system operates; it is these values that move agencies toward systemic goals.

What is clear from the current data is that these 3 states have begun to prioritize integrated employment over sheltered work or other segregated day supports for individuals with intellectual and developmental disabilities. Findings presented throughout this paper and in the conceptual model highlight concrete strategies that state systems have used to increase the commitment of stakeholders to expanding work. This is in light of the fact that the federal government spends four times more money on segregated adult day programs (day habilitation and sheltered work) as compared to supported employment ($488 million vs. $108 million, respectively) (Rusch & Braddock, 2004). Strategies used by these 3 states proved to be the most successful when they were embedded within the context of a solid values base, a network of dedicated stakeholders, and clarity about the system's goal. These strategies were facilitated by relationships, and a structure of local control.

The themes discerned from this research are comparable to case study work conducted within the field of intellectual and developmental disabilities more generally. Case studies conducted in the early 1990s of organizations supporting people with disabilities did not focus on service approaches or models; rather, they described organizations that were responsive to the people they served (Bogdan, Taylor, & Racino, 1991). Bogdan et al. spoke of an organization's “world view” that guided its service delivery, including a vision of greater community inclusion. The organizations were also learning environments that carefully examined and continually evaluated their services and embraced a sense of creativity, uniqueness, and innovation. These themes are mirrored in the importance of local control and clarity of goals to fostering innovation and effective employment supports. Responsive organizations also fostered a sense of communication and personal relationship building among staff, individuals they supported, and community members.

In Murphy and Rogan's (1995) case studies of organizational change, they found that agencies successfully transitioning their services from facility- to community-based had clearly defined values that drove their mission; strong, consistent leadership; and ongoing staff training. Hulgin's (2004) case study work of organizations that had implemented person-centered services found that successful agencies demonstrated adherence to clear ideology and values and that connective relationships among stakeholders were key as were connections with innovators in the field. Moreover, these agencies had positive relationships with policymakers and had “openness in interpreting the limits of the policies” (Hulgin, 2004, p. 173) suggesting that flexible interpretation of policies was a factor in creating more person-centered supports. Butterworth et al. (2000) had similar findings pertaining to the importance of organizational values and openness to risk-taking and innovation. However, Butterworth and colleagues found that catalysts for change were largely internal, and funders and state policy were rarely reported to be significant factors in decision-making. One respondent was even quoted as saying: “I think we changed in spite of the system, not because of the system” (Butterworth et al., 2000, p. 26). In the present study we document states that engage in activities and initiatives that move their systems beyond that of barrier, or even witness to, the growth of community employment. Rather, in tandem with community rehabilitation providers, these states seek the role of facilitator as they push the integrated employment agenda forward. As a larger organizational system, state agencies also thrive in an environment that is based on values, clear goals, and innovation. As at the provider level, a state's decision to pursue integrated employment and the larger goal of community participation have been driven internally by a values clarification process. There was little evidence of federal policy as a catalyst for state decisions.

Considering these findings that build upon results of past case study research, we can extrapolate implications for state systems, community rehabilitation providers, and future leadership in the field.

Implications for Systems: Clarifying the Goal

Rogan and colleagues (2002) called for states to articulate their values and use these to develop and evaluate their policies and practices. There is an explicit challenge for state systems to seek creative approaches to policy formation, organizational change, and service provision that directly influence the development of employment opportunities. Past research has shown that an infrastructure of supports including training, technical assistance, and policy and funding initiatives have helped to drive the expansion of state-supported employment (Novak, Rogan, Mank, & DiLeo, 2003). Although in the current research we found such mechanisms to be important, the most critical component seemed to be the clarity in which the state system identified, transmitted, and maintained commitment to the goal of community inclusion and integrated employment. In their study of the characteristics of community rehabilitation providers, Metzel et al. (2007) called for recognition of the need for private and public funding resources to set integrated employment as the goal. In a study of local state agency service areas, Connolly (1999) found that a clear statement that integrated employment is the preferred outcome for individuals, combined with clear outcome standards for integrated employment in contracts predicted higher rates of employment. Indeed, findings from the present study reinforce what has been found previously in the literature and emphasize the importance of an unambiguous goal statement that drives service delivery.

We found that these 3 states developed structures through training, collecting data, and reporting outcomes back to community rehabilitation providers and limiting or eliminating funding for sheltered work, which allowed for explicit transmission of their goals of including people with disabilities in the workforce. One state in particular, Colorado, offered evidence of the implications of losing focus, which was demonstrated by a decrease in integrated employment outcomes. The state's renewed clarity about its mission resulted in a core group of stakeholders that were dedicated to reinvigorating the system. Whether the states' explicit goals were the expansion of work (which was the case in Washington) or greater community inclusion in part through integrated employment (New Hampshire), states unequivocally sent the message of preferred outcomes for individuals with intellectual and developmental disabilities.

Implications for Community Rehabilitation Providers: Partners in Planning

Review of the literature suggests that a large number of individuals in sheltered employment could successfully transition to integrated employment through aggressive change efforts at the local and state level. However, the continued reliance on facility-based settings for the majority of day and employment supports shows that community rehabilitation providers have yet to reach the goal of community membership (Metzel et al., 2007). It is clear that the commitment and buy-in of community rehabilitation providers towards increasing integrated employment remains a formidable current and future challenge.

In the states involved in this case study research, as in all states, there were community rehabilitation providers who supported integrated employment as well as a significant number who valued the status quo. Recognizing the difference in the values needed to provide sheltered as opposed to integrated employment, advocates and system partners made a conscious decision to focus their efforts on those who supported change. This was very clear both in Washington and New Hampshire, where strategies that were initially focused on the entire provider community, eventually became narrowly focused on those who demonstrated commitment to change and dedication to moving the system forward.

Findings show the influence of providers' actions. In Colorado, community rehabilitation providers were an impetus in ensuring that the state published its data that detailed the state's decline in the number of people becoming employed. Consequently, these providers played an active role on the committee that was focused on expanding employment. In Washington, a provider group consisting primarily of community rehabilitation providers' executive directors effectively lobbied the state legislature in order to increase access to state jobs for people with developmental disabilities. Respondents noted that action-oriented agendas, as opposed to “gripe sessions” helped to ensure that missions of the groups were within reach. Through collaborating with multiple stakeholders, including local and state administrators and the state legislature, community rehabilitation providers were able to help shape the agenda and influence real change. At a federal level, policymakers could support leadership networks of providers at multiple levels to partner with funding agencies with the goal of the expansion of integrated employment.

Implications for Future Leadership: Sustaining the Momentum

As documented throughout this article, one state struggled with sustaining the momentum of forward movement in community supports. Interview data showed that respondents from other sites were concerned about this as well, although outcome data did not yet reveal a real decline in integrated employment. Most commonly cited were concerns about leadership. A network of leaders and key players in the system were pivotal in moving systems towards their goals. Respondents expressed widespread concern that experienced and knowledgeable leaders, those who had undergone intensive values training and experienced deinstitutionalization, were beginning to retire:

People who have always cared about this issue are in their 50s. Where would new and passionate leaders come from? It's different for the new professionals. They are not having the great collective experience that occurred in the 1970s and 80s. They don't know the leaders of those times.

There was the sense that younger leaders had not developed the sense of moral outrage and commitment that could support radical change.

The work involving developing core competencies and performance indicators for staff contributes to a systemic approach in preparing staff to fulfill their roles, remain in the disability field, and obtain the experience and knowledge to move into future leadership roles (Hewitt et al., 2004). Local and state agencies could provide support for certification programs, such as the PATHS program in Ohio funded by the Developmental Disabilities Council (M. Taylor, 2002) that are available for direct support staff. These initiatives could increase retention and provide opportunities for career ladders and longevity in the field (Cohen, 2000). Finally, investment of professional organizations, such as the American Association on Intellectual and Developmental Disabilities, which currently has a working group dedicated to the needs of new professionals in the organization, could do important work in supporting young professionals in the field to develop into future leaders. The intellectual and developmental disabilities field needs to come together to harness the history, experience, and dedication of current leaders and transfer this energy to the next generation:

Future leaders, like their predecessors, must have a deep sense of mission and passion guided by strong moral, ethical, and spiritual values. … They must not only be able to create and articulate a compelling, community-inspired vision of a desired future, but must also know how to engage people in the process so they feel invested in and trust that shared vision. (Foster, 2000, p. 89)

Only a widespread and cross-generational commitment to more inclusive communities will sustain the momentum and further the progress and vision that can lead to full membership for all. The charge for all leaders—past, present, and future—is to find a common medium through which to channel the deep passion and history of experience with the creativity and energy of tomorrow.

Conclusion

Findings from this study are not intended to be a blueprint for states as they work towards expanding integrated employment for individuals with disabilities. Our findings showed that states operate within unique contextual frameworks that shape culture, service delivery, and priorities. However, the respondents from the 3 states highlighted in this article (Colorado, Washington, and New Hampshire) suggest that clarity of goal, an unwavering commitment to training and technical assistance activities that support the goal, flexible and individualized supports, and a long-standing values base are key elements in a system with successful employment outcomes.

References

References
Bogdan
,
R. C.
and
S. K.
Biklen
.
1992
.
Qualitative research for education: An introduction to theory and methods (2nd ed).
Boston: Allyn & Bacon
.
Bogdan
,
R.
,
S.
Taylor
, and
J.
Racino
.
1991
.
Conclusion.
In S. Taylor, R. Bogdan, & J. Racino (Eds.), Life in the community (pp. 253–258). Baltimore: Brookes
.
Butterworth
,
J.
,
S.
Fesko
, and
V.
Ma
.
2000
.
Because it was the right thing to do: Changeover from facility-based services to community employment.
Journal of Vocational Rehabilitation
14
:
23
35
.
Butterworth
,
J.
,
D.
Gilmore
,
W.
Kiernan
, and
R.
Schalock
.
1999
.
State trends in employment services for people with developmental disabilities: Multiyear comparisons based on state MR/DD agency and vocational rehabilitation (RSA) data.
Boston: Institute for Community Inclusion/University Affiliated Program, Children's Hospital
.
Cohen
,
A.
2000
.
Focus on the frontline: Perceptions of workforce issues among direct support workers and their supervisors.
Boston: Governor's Commission on Mental Retardation
.
Cohen
,
A.
,
J.
Butterworth
,
D.
Gilmore
, and
D.
Metzel
.
2003
.
High-performing states in integrated employment.
Research to Practice Series, 9(1). Boston: University of Massachusetts Boston, Institute for Community Inclusion
.
Connolly
,
P. M.
1999
.
The impact of systems-level policies and practices on integrated employment programming for persons with mental retardation and other developmental disabilities.
Unpublished Doctoral dissertation, Boston College, School of Social Work
.
Foster
,
R.
2000
.
Leadership in the twenty-first century: Working to build a civil society.
National Civic Review
89
:
87
94
.
Hewitt
,
A.
,
S.
Larson
,
C.
Lakin
,
J.
Sauer
,
S.
O'Nell
, and
L.
Sedlezky
.
2004
.
Role and essential competencies of the frontline supervisors of direct support professionals in community services.
Mental Retardation
42
:
122
135
.
Hulgin
,
K.
2004
.
Person-centered services and organizational context: taking stock of working conditions and their impact.
Mental Retardation
42
:
169
180
.
Institute for Community Inclusion.
2006
.
National survey of day and employment programs for people with developmental disabilities.
Boston: Author
.
Institute for Community Inclusion. (n.d.).
StateData.info. Retrieved January 14, 2006, from http://www.statedata.info
.
Institute for Community Inclusion. (n.d.).
Statedata.info. Retrieved February 22, 2007, from http://www.statedata.info
.
Lincoln
,
Y.
and
E.
Guba
.
1985
.
Naturalistic inquiry.
Newbury Park, CA: Sage
.
Lofland
,
J.
and
L.
Lofland
.
1995
.
Analyzing social settings: A guide to qualitative observation and analysis.
Belmont, CA: Wadsworth
.
Mank
,
D.
2003
.
Supported employment outcomes across a decade: Is there evidence of improvement in the quality of implementation?
Mental Retardation
4
:
188
197
.
McGaughey
,
M.
and
D.
Mank
.
2001
.
Empirical evidence of systems change in supported employment.
Journal of Disability Policy Studies
11
:
210
222
.
Metzel
,
D. S.
,
H.
Boeltzig
,
J.
Butterworth
,
J. S.
Sulewski
, and
D. S.
Gilmore
.
2007
.
Achieving community membership through community rehabilitation provider services: Are we there yet?
Mental Retardation, 45, 149–160
.
Miles
,
M.
and
A. M.
Huberman
.
1994
.
Qualitative data analysis (2nd ed).
Thousand Oaks, CA: Sage
.
Murphy
,
S. T.
and
P.
Rogan
.
1995
.
Closing the shop: Conversion from sheltered to integrated work.
Baltimore: Brookes
.
Novak
,
J.
,
P.
Rogan
,
D.
Mank
, and
D.
DiLeo
.
2003
.
Findings from a national survey of state vocational rehabilitation agencies.
Journal of Vocational Rehabilitation
12
:
157
166
.
Rogan
,
P.
,
J.
Novak
,
D.
Mank
, and
R.
Martin
.
2002
.
From values to practice: State level implementation of supported employment.
Journal of Vocational Rehabilitation
17
:
47
57
.
Rusch
,
F.
and
D.
Braddock
.
2004
.
Adult day programs versus supported employment (1988– 2002). Spending and services practices of state mental retardation and developmental disabilities agencies.
Research and Practice for Persons with Severe Disabilities
29
:
237
242
.
Silverstein
,
R.
,
G.
Julnes
, and
R.
Nolan
.
2005
.
What policymakers need and must demand from research regarding the employment rate of persons with disabilities.
Behavioral Sciences and the Law
23
:
399
448
.
Stake
,
R. E.
2000
.
Case studies.
In N. Denzin & Y. Lincoln (Eds.), Handbook of qualitative research (2nd ed., pp. 435–454). Thousand Oaks, CA: Sage
.
Taylor
,
M.
2002, April 25
.
The road not taken. Path to careers in human services.
Paper presented at the Innovation out of Crisis, Boston. Retrieved on April 25, 2002, from www.hsri.org/ddworkforce/articles.html
.
Taylor
,
S.
,
R.
Bogdan
, and
J.
Racino
.
1991
.
Life in the community. Case studies of organizations supporting people with disabilities.
Baltimore:Brookes
.
Washington State Department of Social and Health Services, Division of Developmental Disabilities.
1992
.
County Guidelines.
.
Weathers
II.,
R. R.
2005
.
A guide to disability statistics from the American Community Survey.
Retrieved July 3, 2006, from Cornell University, Employment and Disability Institute, http://digitalcommons.ilr.cornell.edu/cgi/viewcontent.cgi?article=1123&context=edicollect
.
Wehman
,
P.
,
K.
Inge
,
G.
Revell
,
J.
Butterworth
, and
D. S.
Gilmore
.
2005
.
Community rehabilitation programs: Current status and perceptions of agencies holding special wage certificates.
Unpublished manuscript, Virginia Commonwealth University
.
Wehman
,
P.
,
W. G.
Revell
, and
V.
Brooke
.
2003
.
Competitive employment: Has it become the “first choice” yet?
Journal of Disability Policy Studies
14
/
3
:
16
26
.
Wolfensberger
,
W.
and
S.
Thomas
.
1983
.
PASSING (Program Analysis of Service Systems' Implementation of Normalization Goals): Normalization criteria and ratings manual (2nd ed).
Toronto: National Institute on Mental Retardation
.

The authors thank the participants from New Hampshire, Colorado, and Washington for their time, expertise, and invaluable perspective on employment for people with intellectual and developmental disabilities in their respective states. This document was supported in part by Cooperative Agreement 90ND0204 from the Administration on Developmental Disabilities, Administration for Children and Families, U.S. Department of Health and Human Services. Points of view or opinions do not necessarily represent official Administration on Developmental Disabilities Policy.

Author notes

Authors:

Allison Cohen Hall, PhD (Allison.Hall@umb.edu), Research Associate; John Butterworth, PhD, Research Coordinator; Jean Winsor, MA, Research Fellow; Dana Gilmore, MA, Research Associate, Institute for Community Inclusion; and Deborah Metzel, PhD, Assistant Professor, Department of Environmental, Earth & Ocean Sciences, University of Massachusetts, 100 Morrissey Blvd., Boston, MA 02125