From a “Perfect Storm” to “Smooth Sailing”

Method

Study Context

We examine the use of SafeCare in the child welfare systems of two U.S. states referred to here as State A and State B. According to the 2010 U.S. Census Bureau, State A has a population of approximately 3.7 million residents, almost 42% of whom live in rural areas. In State A, SafeCare is implemented through a state-operated child welfare system with all services guided and contracted by the state government. State B includes a mix of 10 urban and rural counties involved in implementing SafeCare. These counties range in population from approximately 3.2 million residents to just over 150,000. In State B, each county is accountable to the state government via the use of system improvement plans (SIPs); the state is then accountable to the federal government through PIPs. However, there is considerable discretion exerted by county policymakers regarding the development of SIPs and in the selection of and support for child welfare services to be provided to children and families.

The main advantage of including State A and B in this study is that they together provide ample opportunity to assess, at different levels of government, a range of service systems in which SafeCare has been provided for varying lengths of time. Indeed, we consider a total of 11 separate implementation sites: 1 statewide site in State A and 10 countywide sites in State B. Training in the EBP began between 2 and 10 years prior to this study. Three sustainment levels existed across sites. We define SafeCare “sustainment sites” as ones where certified providers were currently implementing SafeCare services with active caseloads, convening regularly as SafeCare teams, and receiving coaching in accordance with NSTRC policies. We define “partial sustainment sites” as ones where certified providers were currently implementing SafeCare with active caseloads but without regular team meetings or coaching. We define “nonsustainment sites” as ones where there were no longer certified providers implementing SafeCare with fidelity. Three sites were nonsustainment sites (total time implementing ranged from 1.5 to 2.3 years), one was categorized as a “partial sustainment site” (total time implementing was 4 years) and seven were “sustainment sites” (total time implementing ranged from 2 to 10 years).

Participants

We recruited and then interviewed 24 child welfare policymakers at state and county levels about EBPs and SafeCare. The participants were purposefully selected based on their roles, knowledge of, and direct involvement in SafeCare implementation within their respective service systems. We defined policymaker as an employee of a state or county government who was instrumental in championing, developing, monitoring, or strongly influencing the course of events related to the provision of an EBP. The policymakers in our sample were widely recognized by the leadership of CBOs and other stakeholders as playing pivotal roles in implementation. Categories of state- and county-level social service stakeholders included directors (n = 3), deputy directors (n = 6), division directors (n = 5), program managers and administrators (n = 7), and analysts (n = 3). We invited policymakers via both phone and e-mail; all agreed to participate, resulting in a 100% response rate. The sample included 22 females and 2 males. The sampling method, research design, and consent procedures were approved by the Human Research Protections Program of the University of California, San Diego. All participants also read and signed an official written informed consent document that clearly specified that pertinent identifying features (e.g., names and locations of employment) would not be included in publications to protect their anonymity. Thus, to maintain the confidentiality of participants, the specific names of states and service systems were withheld in the research reported subsequently.

Data Collection

Two anthropologists (first and third authors) undertook 1-hr, semistructured interviews to elucidate policymaker experiences. The interviews occurred between 1.5 and 10 years postinitial EBP training and implementation of the SafeCare model. The interviews took place in the 11 service systems and at two state capitals between 2012 and 2013. At the outset of the interviews, policymakers were asked to describe their particular involvement in SafeCare. The remaining questions were informed by the EPIS model and covered several topics: positive and negative influences on SafeCare implementation and sustainment (e.g., How are things working out with SafeCare at this point in time? What is working well? What is not working well?); leadership (e.g., How have “leaders” at all levels, including the state, county, and CBO, influenced the initial implementation of SafeCare? How are they influencing the ongoing use of SafeCare?); decision-making and policy making, and contracting processes (e.g., Who are the most important decision makers, or stakeholders, to influence whether SafeCare continues [in this system]? Can you explain to me how the current SafeCare contract works? What policies are in place to support the use of SafeCare?); and the prospects for SafeCare within each implementation milieu (e.g., What might prevent SafeCare from being delivered in the future?)

Data Preparation and Analysis

All interviews were digitally recorded, professionally transcribed, and checked for accuracy by at least one of the authors. We employed an iterative process to review the textual data from interviews and utilized NVivo 10 qualitative data analysis software to facilitate this work (QSR International, 2012). Data analysis proceeded first by engaging in open coding to locate the themes and issues that emerged from the interview transcripts (Corbin & Strauss, 2008). The transcripts were independently coded by the first three authors to condense the data into analyzable units. Segments of text ranging from a phrase to several paragraphs were assigned codes based a priori on the particular topic areas and questions that made up the interview guides (Patton, 2002). These codes thus centered on key constructs of the EPIS framework (e.g., “leadership,” “policy,” and “collaboration”). The use of such constructs enabled us to examine both the salience and the meaning of these constructs for policymakers through the provision of descriptive data based on the actual words of participants; the resulting narratives directly reflected their perceptions and experiences related to SafeCare sustainment issues. We employed these constructs to help make sense of the qualitative data but not “to dominate, strain, or force the analysis” (Patton 2002, p. 457). During our review of the transcripts, new codes were subsequently identified and defined to capture information on emergent themes (e.g., “cultural issues related to SafeCare” and “SafeCare referral processes”). Focused coding was then used to determine which of these themes surfaced frequently and which represented unusual or particular concerns to the research participants.

In the staged approach to analysis, each author coded sets of transcripts, created detailed memos that both described and linked codes to each theme and issue, and shared their work with one another for review. Through the process of constantly comparing and contrasting codes with one another (Corbin & Strauss, 2008; Glaser & Strauss, 1967), we grouped together those with similar content or meaning into broad themes linked to segments of text. The final set of codes, constructed through a consensus of the authors, consisted of a numbered list of themes which, for this analysis, placed SafeCare in a framework for understanding how policymakers identify and describe outer-context characteristics that shape implementation and sustainment.

EBP Adoption Decision

Policymakers identified several factors influencing their decisions to support implementation of EBPs in child welfare systems. In particular, they linked such decisions to national trends that prioritize federal funds for EBP implementation. Yet, funding opportunism was not the driving force, with 71% of policymakers expressing interest in EBPs since these interventions include methods for systematically measuring and demonstrating success. Because they had been “tested,” policymakers believed that EBPs provided them with some assurance that the attainment of positive outcomes was possible. Outcomes of particular interest to policymakers included reduced out-of-home placement for children impacted by neglect and recidivism or reentry into the child welfare system. Echoing a commonly held sentiment, one policymaker stated, “We don’t have to look at it as ‘Will it work or will it not?’” The proscriptive structure and emphasis on “accountability” also bolstered their support for SafeCare. In the seven systems in which SafeCare had entered into the sustainment phase, policymakers clarified that this emphasis increased their confidence in knowing that public dollars were being spent responsibly and to support evidence-based initiatives at the system level.

Twenty-five percent of policymakers were “attracted” to SafeCare because of its “train-the-trainer” model that encouraged development of both local “experts” and “infrastructure” to educate and monitor home visitors. They noted that this feature limited the need for expensive or ongoing involvement of intervention developers, whom they commonly characterized as contractors “from the outside” of their respective service systems. These policymakers considered the presence of local experts and infrastructure as critical to addressing home visitor turnover and the training needs of new staff hires. They also reported that the cultivation of local capacity for coaching home visitors facilitated appropriate quality control of SafeCare and afforded greater oversight of services delivered.

Leadership

Across the spectrum of implementation sites, policymakers referred to the presence of strong leaders who recognized the benefits of EBPs and advocated for them as a requirement for successful implementation and eventual sustainment. Twenty-five percent of policymakers explicitly stated that state and county leadership must be willing to “champion” interventions. As articulated by one policymaker, these champions also needed to “maintain a steadfast valuing of evidence-based practices.” Policymakers in systems where SafeCare had been sustained described themselves as “networkers” who participated in committees and sought out interactions with individuals outside their respective counties and states to gain knowledge about EBPs and to learn about strategies for supporting both implementation and sustainment. More locally, these policymakers had also committed themselves to participating actively in planning meetings and early training activities related to SafeCare specifically and continued to attend relevant events so that those charged with implementation remained aware of system-level support for the program. Across the board, policymakers argued that successful implementation was not only dependent on their “buy-in” for SafeCare but also on the buy-in of provider agency administrators and home visitation staff. They saw it as their responsibility to foster both buy-in and collaboration among all persons involved in each phase of the implementation process.

In the four partial sustainment and nonsustaining sites, policymakers reportedly lacked strong preferences for EBPs. They also admitted to being preoccupied with other competing priorities and child welfare projects. Thus, implementation of SafeCare fell short. One such policymaker disclosed, “As important as initiatives [like SafeCare] are, the bottom line is we have to get our mandates done.” In contrast to their peers in the sustaining sites, these policymakers also indicated that they had invested little effort into developing contracts to ensure sustainment of the Safecare model, that is, building in provisions for appropriate resources to maintain the quality of the intervention model or systematizing a referral process for SafeCare.

More than half of policymakers (58%) reported that changes in leadership in key administrative positions at either the county or the state levels could compromise implementation and possible sustainment of Safecare. A policymaker in a service system where SafeCare had been discontinued lamented, “Having someone inherit this [responsibility for SafeCare] and then try to shepherd it through was problematic. It seemed to be a struggle.” This policymaker also commented on the lack of sustainment of SafeCare within her or his system, “It was that perfect storm where several elements came together at the same time, and initial investments, initial people who were involved, changed, in terms of leadership. And you lost that vision and the investment piece at the start.” Another policymaker in a different service system worried about leadership changes at the highest level of government, because new appointees lacked knowledge of and commitment to SafeCare. In yet another system, policymakers explained that once a key champion of SafeCare implementation had departed from her or his position, funding diminished and child welfare social workers stopped referring clients to the SafeCare program.

In contrast, however, two systems that had experienced turnover in top leadership had very viable SafeCare programs, partly owing to their efforts to build support for the EBP within their own government ranks and in the broader service systems. In one system, the key champion started to “groom” support for the EBP among colleagues and her or his successor 1 year in advance of her or his retirement. This individual was quite confident of ongoing support from leadership. In the other system, a policymaker described her or his predecessor as a “visionary” who was “very much a proponent of extending EBPs” throughout the public sector and envisioned her or his job as helping to start up and ensure the continuation of the program.

Variability in Funding

Policymakers recollected that implementation of SafeCare in State A began over a decade before, with initial evaluation work supported in part by a National Institute of Mental Health (NIMH)-funded efficacy trial led by academic partners. Afterward, SafeCare was implemented system wide by the state through contracted CBOs. Initiated between 2008 and 2011, funds for the county-operated implementations in State B were supported by blended funding streams that varied across locations, usually including county funding streams combined with philanthropic foundation funds and federal research grants from the NIMH, Centers for Disease Control and Prevention, and Administration for Children and Families However, as commonly observed by policymakers, service delivery was supported by customary state and county sources.

Over the years, funding for SafeCare took multiple forms. Service systems depended on various state and federal sources, each with different stipulations tied to how monies were to be spent. Policymakers in State A, for example, depended on long-term funding streams from state general revenue dedicated to child welfare services to fund SafeCare provision. Six counties in State B took advantage of “one-shot” funding opportunities to pay for the training or other specific components of SafeCare (e.g., supplies). For example, in one county, SafeCare adoption and implementation processes were initiated and catalyzed by a philanthropic foundation initiative that brought new but nonrecurring funds into the system specifically for training.

In the five systems in State B where funding for the intervention was diversified, policymakers emphasized creative use of seemingly disparate sources of capital. Policymakers in one of these systems tapped into the Corporation of National and Community Service (AmeriCorps) to hire home visitors at reduced costs to the county. Policymakers in a second county obtained local health council funds designated for child abuse prevention, whereas another used federal public health and state mental health monies. Policymakers in these five systems commonly observed that such funding sources were subject to the changing priorities of councils and gubernatorial administrations or had to be utilized in a very proscriptive way.

In State B, policymakers in systems where SafeCare had entered the sustainment phase underscored the need to be forward thinking about optimally integrating different funding sources to support implementation, institutionalization, and sustainment. Yet, these policymakers still worried about budget cuts at both the state and the county levels possibly threatening the viability of SafeCare in spite of their strategic thinking. Policymakers described funding woes in the four systems where the EBP was said to be “on life support” (partially sustaining) or “dead” (nonsustaining). In one system in which there was a lapse in funds soon after policymakers committed resources to train providers in SafeCare, implementation lagged by several months. This lag reportedly led to dilution of staff skills in SafeCare, a low level of buy-in among stakeholders at all levels, and the discontinuation of the EBP.

Policies and Contracts

Policymakers in the sustaining systems stated that well-specified requests for proposals (RFPs) and contracts facilitated implementation and sustainment of SafeCare. Yet, official policies regarding SafeCare were lacking generally in both States A and B. Only 1 of the 11 systems had official policies concerning SafeCare. For the most part, policymakers in all of State B’s 10 implementation sites suggested that there was no need for policies. Instead, county governments relied on RFPs and contracts that explicated the roles, responsibilities, statements of work, and expected outcomes for service delivery. All policymakers emphasized that the structure of SafeCare was well formulated in the curriculum that providers were contracted to follow. When built into contract requirements, ongoing fidelity monitoring and coaching also augmented quality assurance. In State A, the child welfare system forged contractual relationships with both academic partners and CBOs, whereby the former engaged in evaluation and quality control and delivered training and technical assistance to the CBO-based SafeCare coaches. This formal relationship with the academic partners was written into the contracts between the state government and the CBOs.

A minority (12.5%) of policymakers said it might be useful to create policies around SafeCare to buffer it against threats after key administrators leave their current role or other leadership changes. A policymaker transitioning to a new job stated, “One thing that we did is to [make] SafeCare one of our strategies for a 5-year system improvement plan or SIP. It’s been institutionalized in that way and it’s been highlighted as one of the ways we’re going to improve child welfare outcomes. It doesn’t matter who sits in my position.” Another retiring policymaker in a different system also said that county administrators created de facto policy by including SafeCare into an SIP to ensure that the EBP would continue, despite leadership changes.

Partnerships

SafeCare was sustained in systems where policymakers described their partnerships with local stakeholders (e.g., CBO administrators and staff and academic partners) as already strong. Policymakers in these systems reported that preestablished partnerships eased implementation challenges. One policymaker said simply, “We also contract with the same providers for [another program]. We have a relationship.” Those involved in selecting CBO contractors described robust, productive working relations, yet they also cited resentment among CBO directors of other agencies regarding a perceived lack of competitive advantage. Policymakers in one system who utilized a closed bidding process observed that the directors of CBOs who had never been awarded contracts for SafeCare services were looking forward to the conversion to a new contracting process that would potentially level the playing field for potential bidders.

Close to 63% of policymakers commented on how academic partners help facilitate implementation. Through formal presentations, conference calls, and participation in planning and implementation meetings, academic partners educated policymakers about EBPs in general and SafeCare in particular. Policymakers characterized the academic partners as conduits for information exchange, as they shared feedback on program processes and outcomes on a regular basis, helped to ensure EBP “fit” with local context, and facilitated dissemination of SafeCare across both states. The academic partners also brought financial resources to the table through their grant writing activities that facilitated initial implementation and dissemination as well as fidelity monitoring and evaluation at less cost to county and state child welfare systems.

Although policymakers touched on the utility of having the EBP developers involved in setting up SafeCare in their systems, very few sustained their relationships with the developers over time, partly owing to the costs involved in relying upon their expertise. Instead, they reportedly depended on the academic partners based on in States A and B. Policymakers explained that they had ongoing relationships with these academic partners and depended on them to provide both input and information related to SafeCare implementation. The academic partners, however, maintained their relationships with the model developers and contributed to the design, integration, and evaluation of SafeCare innovations and adaptations within local systems.

SafeCare Staffing

Staffing for SafeCare has varied as policymakers utilized multiple funding sources with different stipulations and limitations. Across States A and B, SafeCare was delivered by a broad workforce with diverse professional backgrounds and credentials. For example, policymakers in 7 of the 11 systems explained that they relied on CBOs to implement SafeCare with case managers or other workers (some Master’s level and others Bachelor’s level) already doing in-home services. In the remaining systems, policymakers decided to use public health nurses, AmeriCorps volunteers, child welfare social workers, or a combination of staff, for SafeCare provision. However, 8% of policymakers raised concerns about the use of certain staff, claiming that reliance on highly educated nurses to deliver the intervention was expensive and unsustainable (despite the federal government reimbursing up to 75% of their salary) and citing the temporariness of the volunteer workforce and the competing demands placed on child welfare workers who were easily sidetracked from administering SafeCare with fidelity. Although the train-the-trainer model reduced the costs of educating new staff, policymakers identified the expense associated with turnover as a pervasive problem.

Policymakers also emphasized the need to ensure access to quality training, coaching and supervision, and project coordination to support the home visitation staff and the broader goals of SafeCare implementation. Almost 67% of policymakers, for example, emphasized that quality training and supervision were imperative to program success. Policymakers working in a sustaining site where SafeCare had reportedly floundered when it was first introduced had spearheaded a change in supervisory staff, incorporating an individual recognized locally as an opinion leader into the service delivery team. Once the change occurred, “Everything started to go better.” According to policymakers, SafeCare languished during the early implementation stage in the nonsustaining sites because staff members able to fulfill training, coaching, and supervision functions were not consistently available to the home visitors.

System Challenges

More than 83% of policymakers suggested that effective SafeCare implementation required the use of careful planning processes that encouraged stakeholders at all levels in a system to be problem solvers, anticipate system challenges in advance, and design and enact solutions. One policymaker stressed all the networking he or she does at state and national meetings to learn from others involved in SafeCare to enhance implementation planning locally. He or she also said his or her job responsibilities entailed listening to CBO contractors and ensuring they had the resources needed to implement SafeCare (e.g., copies of the curriculum to distribute to families).

In general, policymakers emphasized a proactive approach to overcoming barriers to SafeCare in complex service systems. Sixty-two percent, for example, characterized timely referrals as a key to successful implementation, as home visitors needed to practice skills soon after training for effective learning and to enhance their confidence in their ability to deliver SafeCare. Yet, generating referrals to SafeCare providers was a common problem in States A and B, particularly during early implementation phases. Policymakers attributed such problems to a lack of knowledge about SafeCare among the child welfare social workers responsible for submitting referrals. Their insufficient awareness of the program caused them to not refer parents to SafeCare or to refer parents of older children who were not well suited to SafeCare. Policymaker-initiated efforts to rectify referral problems included classes and presentations for child welfare service workers that focused on making appropriate referrals and specifying the SafeCare option on referral forms. Policymakers in one system also colocated home visitation staff in the same office as child welfare social workers to facilitate communication, familiarity with the EBP, and “handoffs” of SafeCare clients. These efforts to adjust referral procedures reportedly helped enhance both awareness and use of SafeCare. In systems where adjustments were made and social workers were provided with needed education and supports, policymakers claimed that appropriate referrals were now the norm, which allowed the EBP to flourish. Policymakers cited a continued lack of referrals as a contributor to the demise of SafeCare in systems where the EBP was no longer practiced.

The policymakers associated with the three systems where SafeCare was no longer practiced indicated that they and their colleagues had not prioritized careful planning for the EBP. Although they reported placing high expectations on frontline workers to administer the EBP with fidelity, they identified themselves as responsible for not sufficiently attending to how these workers could be best supported in intervention practices to meet expectations at the outset. These policymakers were also the most likely to criticize the intervention. They were among the 8% of policymakers who expressed concerns about cultural insensitivity and 20% who discussed limiting system-level support for implementation due to the restricted child age range served by SafeCare. One policymaker, for instance, suggested that because SafeCare specified only a set number of strategies for influencing parenting behavior, it was not necessarily conducive for effecting change across different cultural groups. Others argued that families in the child welfare system represented a larger demographic than SafeCare allowed. These families also faced multiple challenges that frontline workers needed to address but which fell outside the purview of SafeCare (e.g., securing shelter and food). In the sustaining systems, policymakers characterized such issues as less of a problem for delivering SafeCare, possibly because the EBP was part of a broader home-based service array that included case management and discretionary funds for providers to assist families in addressing their basic needs.

The limited age range and potential for cultural insensitivity were acknowledged as challenges but were not considered insurmountable in sites where SafeCare was in the sustainment phase. In these sites, policymakers claimed to have worked with intervention developers to incorporate families with older children and collaborated with their academic partners to create modules to tailor SafeCare to clients with diverse needs. Even in these sites, however, policymakers kept their eyes out for “something better [than SafeCare] to come along,” including the possibility of “more effective” and/or “cheaper” home visitation models.

Legal, Legislative, and Political Pressures

Policymakers described multiple political and legal pressures emanating from the outer context that had implications for SafeCare and its sustainment. For example, policymakers suggested that a new contracting process recently introduced by the gubernatorial administration in State A had prevented them from releasing an RFP focused explicitly on the requirements of SafeCare. Sustainment had been predicated on the ability of the old contracting structure and its accompanying language to institutionalize SafeCare and its cross-CBO quality control system. The new contract bid system reduced the authority of child welfare policymakers to specify a desired target EBP, mandate collaboration with academic partners, or establish more than a few basic service parameters. It additionally diminished the role of child welfare staff in reviewing bids, thus opening the door for low-price bidders proposing to deliver different and less well-supported services. Ultimately, the new leadership of the child welfare system opted to sustain SafeCare implementation but not before this single and ostensibly minor change to state policy (a new contract bid process) threatened to disrupt, if not eliminate, a decade of EBP implementation and sustainment effort within a few months’ time.

Although legal actions affecting service delivery were underway in both states, policymakers in State A explained that a recent lawsuit and the negative press surrounding it had led to a major restructuring of the child welfare service system that had shifted power away from a commission that previously oversaw operations to the executive branch of the government. The lawsuit resulted in the 5-year enforcement of a single statewide child welfare plan (akin to a PIP) that focused more on children in the foster care system than on families receiving home visitation services. Policymakers in State A observed that the new leadership appointed by the gubernatorial administration lacked the institutional memory concerning previous investments in the start-up of SafeCare. Within this evolving context, policymakers expressed concern regarding ongoing state-level leadership support for SafeCare sustainment in the child welfare system.

Policymakers also underscored that although regional and state-level service system dynamics affected SafeCare, so did outer-context factors arising from the federal government. For example, policymakers in State B were disappointed when SafeCare was not initially selected by a key federal funder as an “approved” home visitation model to reduce child neglect. They explained that the federally selected EBPs were not as “flexible” as SafeCare regarding eligible clientele, nor were they as cost effective. Despite the fact that SafeCare has since been endorsed by a national evidence-based clearinghouse for child welfare services and added to the approved list (U.S. Department of Health and Human Services, 2013), it was too late for policymakers in State B to apply for financing from this source to support SafeCare sustainment.

Limitations

This work occurred in two states experienced with SafeCare, which constrains generalizability. We also did not interview the highest level of policymakers with influence over child welfare systems, such as cabinet secretaries, state agency directors, legislators, or governors. We instead interviewed a diversity of policymakers playing key roles in the two most dominant outer-context environments, namely state- and county-run systems. This adds to our ability to document similarities and differences in system-level strategies affecting delivery of the same EBP. We also interviewed policymakers in systems with long-term sustainment, partial sustainment, and failed sustainment and still achieved relatively high levels of saturation, or consistency of data, with regard to the findings presented here. Importantly, we only interviewed policymakers acknowledged by others for their role in SafeCare implementation, hence the sample was likely more supportive than others may have been. Nevertheless, the policymakers were generally forthcoming in their characterizations of what did and did not go well in terms of implementation. Although some relied on recall for retrospective accounts of these processes, their accounts were corroborated with qualitative and quantitative data collected for two larger studies (Aarons et al., 2009; Hurlburt et al., 2014). We examine the perspectives of policymakers, a population whose views are often absent from research related to EBP delivery, while other articles by our research team delve into the perceptions of other key stakeholders (e.g., home visitors, coaches, and CBO directors) involved in SafeCare implementation (Aarons & Palinkas, 2007; Palinkas et al., 2009; Palinkas & Aarons, 2009). Findings from this research pertaining to strong leadership, proactive planning that considers implementation into the future, and the utility of data, well-specified RFPs, and contracts may enhance EBP sustainment elsewhere and would benefit from additional investigation in different public service systems.

Other research suggests that communication is vital to implementation and sustainment of EBPs in the outer and inner contexts (Aarons et al., 2014; Hurlburt et al., 2014). However, it is notable that the topic of “communication” did not present as a major theme in policymakers’ discussions of EBP sustainment except in one system, although it had emerged as a prominent theme in discussions with inner-context stakeholders involved in SafeCare implementation (Aarons & Palinkas, 2007). In our research, those positioned in mid-management in CBOs and at the front line repeatedly called attention to how problems of communication at the top of system hierarchies affect EBP delivery. Although the absence of such discussions could be an artifact of how we structured our semistructured interview questions, we suspect that having positions as high-level administrators enables policymakers to become accustomed to setting rules, creating contracts, and having their directives followed and less dependent on reciprocal communication. Future studies will examine how these challenges surface and are addressed by policymakers in the outer context.