Funding Diversity: A Case Study of A State-initiated, Funding-driven Program to Diversify Mental Health Service Provision in Minnesota

Abstract

Human service organizations are reconfiguring to address diversifying populations and widening inequality. However, institutional change is challenging to implement and fund; resource scarcity and stakeholder buy-in are barriers. In this case study, we analyze a funding-driven, state-initiated program that supports mental health professionals who are people of color in order to decrease health disparities. Analyses of interviews and documents depict how the program struggled with high turnover and uninspired, halfhearted messaging, but was nevertheless well loved. Findings illustrate how the ‘pitch’ and leadership matter in programming for institutional change and its contested nature, a contestation that funding alone cannot temper.

An emerging line of research on human service organizations pertains to transformative change programming, including that of direct service nonprofit organizations (Gates, 2014; Mellinger, 2014; Netting, O’Connor, & Fauri, 2007; Shier & Handy, 2015). Such inquiries are reflective of on-the-ground trends, emergent in response to heightened social issues, such as inequality, social exclusion, oppression, militarization, and racism. Human service and direct service organizations are reconfiguring intervention approaches and organizational practices to meet the dynamic and complex issues and challenges associated with two concerns: population diversification and heightened social problems. Such organizations are also responding to organizational fields that are increasingly resource-poor (Gates, 2014; Mellinger, 2014; Minkoff, 2002).

Planning “transformative change programs” is challenging, however, because such programs are, simply, challenging to fund (Netting et al., 2007; Pearce, 2010; Shier & Handy, 2015). And they are difficult to pitch to policymakers and powerful stakeholders. Netting et al. (2007), in an insightful article, made funding challenges the premise of a study that sought to build theory and formulate “how-to” guidelines for planning transformative programs. Most nonprofits conduct “focused programs” that seek within-system change that is also typically at the individual level for clients (i.e., case advocacy). “Transformative programs,” however, aim to enact broader, long-term, structural interventions that challenge the status quo. “And because (transformative programs) upset the status quo, it will meet much resistance” (Netting et al., 2007:60). Given that transformative programming challenges systemic/institutional barriers and racism, it is typically structured bottom-up, emerging out of actions of resistance from clients and service providers who are adversely impacted by inequities. Thus, to contend with issues of funding, planners and administrators must address the difficult task of translating cause-oriented language into measurable outcomes that will resonate with funders (Netting et al, 2007). Funding or resource scarcity and stakeholder buy-in are thus considered key challenges to transformative change programming, as examined in scholarship.

This study looks at it from a different angle: we provide analyses and insights on fund-driven, top-down implementation of transformative change programming, using a case study of a decade-long, state-wide effort in mental health services to decrease racial disparities and increase representation of people of color. Whereas transformative change programs are typically conducted by grassroots organizations from the bottom-up (Netting et al., 2007:63), our case presents a unique situation, wherein transformative change programming is top-down.

This case study examines the Cultural and Ethnic Minority Infrastructure Grant (CEMIG) program of the State of Minnesota, created to increase diversity among mental health professionals. Increasing diversity, in turn, aims to increase access to and quality of culturally congruent mental health services, as part of the state’s 2007 Governor’s Mental Health Initiative (Department of Human Services, 2006). Minnesota is the only state in the nation to have implemented such a statewide program, as of the time of writing. While diversifying the workforce is an important and innovative approach, it is but a first step needed for systemic change. Structural racism is deeply embedded within public institutions, and there are many dimensions to structural barriers to mental health care for nonwhite populations. Over eight years from 2008 to 2016, Minnesota’s CEMIG program invested approximately 8.86 USD million to train, mentor, and provide direct resources to mental health providers from cultural and ethnic minority communities. “Transformative change” is defined as “long-term, structural interventions designed to change the status quo at a broad community, state, regional or even national level” (Netting et al., 2007), thus befitting the aims of Minnesota’s CEMIG program.

Resource scarcity and stakeholder buy-in as challenges are commonly encountered or ubiquitous for transformative change programs, and typically occur from the bottom-up. Our case thus provides a unique perspective to a counterfactual example. How might implementation processes look and what salient elements emerge when transformative change programming is top-down, fund-driven, and broad-based, as spearheaded by a human services institution at the state level?

Background

Transformative programming

The planning and implementation of “transformative programs” attend to ideas/ideologies, organizational culture, and structural logic within institutions. Transformative programs are dissimilar to “focused change programming,” which is more attuned to outcomes (Netting et al., 2007). In terms of problem definition/dentification and needs assessment, transformative programming hones in on the root causes of problems and engages in their complexity in a collective process; whereas focused change programming hones in on specific conditions that can be reasonably managed by way of direct services. This fundamental difference at the outset sets the terms of the implementation. Strategies for intervention in transformative programming aim at large system levels, are dynamic or evolving, and pay attention to power. As such, they are challenging and extend over a long period of time (Netting et al., 2007).

Diversity in Minnesota’s mental health service sector

Significant racial disparities in mental health services, including access to care (Cook, Trinh, Li, Hou, & Progovac, 2017), have been associated with a lack of cultural congruence between therapists and clients (Smith, 2010). Therapist–client matches in ethnicity and language were found to be related to length of treatment and beneficial to clients who were ethnic minorities (Sue, Fujino, Hu, Takeuchi, & Zane, 1991). Studies illustrate cultural congruence as a predictor of treatment outcomes (reduced symptoms) regardless of type of treatment. Cultural congruence also moderates the effects of depression, suicidality, anxiety, and physical health criteria among immigrant populations, for example (Costantino, Malgady, & Primavera, 2009).

The Minnesota State Demographic Center has reported that Minnesota’s population will increase by over a million people from 2005 to 2035, reaching 6.4 million (McMurry, 2009); and projected that the white population will fall from 86% to 75% of the total population, while the Latino, African American, and Asian American populations will double within that timeframe (McMurry, 2009). While data describing the demographics of the mental health workforce was unavailable in 2006 when the CEMIG legislation was written and passed, the most current information regarding the diversity of the Minnesota behavioral health workforce is available on the Minnesota Department of Health (MDH) website. MDH gathered and synthesized data from discipline-specific voluntary workforce questionnaires, as depicted in Table 1. As of the end of 2017, approximately 1,659 nonwhite individuals were licensed social workers, psychologists, marriage and family therapists, and clinical counselors in Minnesota (MDH, n.d.). If we consider the provider-to-person ratio within each racial group in Minnesota, lack of representation is evident; for instance, for every one Black provider there would be 756 people from the Black community for a 1:756 provider-to-person ratio, and a staggering 1:1,680 ratio for the Hispanic community, compared with a 1:432 ratio for the white community (see Table 1).

Table 1.

Diversity of Minnesota’s Mental Health Workforce.

Discipline	Total Number	American Indian	Asian	Black	Hispanic	Multiple Races	White	Other
Clinical Counselors	1,565	2.0%	1.7%	2.3%	2.4%		94.9%	1.8%
Marriage & Family Therapists	2,187	0.3%	3.1%	1.9%	0.9%	3.5%	89.3%	1.0%
Psychologists	3,824	1.7%	2.4%	1.5%	1.8%		88%	1.8%
Social Work	11,069	1%	2%	2%	0.2%	3%	92%	1%
Provider-to- Person ratio, within group	NA	1:267	1:552	1:756	1:1680	NA	1:432	NA

Information Provided by Minnesota Department of Health: Mental and Behavioral Health Workforce Reports (MDH, n.d.)

In response to reports that individuals from cultural and ethnic minority communities receive less and lower quality health care (Holm-Hansen, 2006; Snowden, 2012), government entities have sponsored legislation, research, and grant programs to investigate issues of access to mental health care (Snowden, 2012). Over the last 20 years, Minnesota has tried various strategies to increase the diversity of licensed social workers (Bibus & Boutte-Queen, 2011; Minnesota Board of Social Work, 2008) and undergone multiple investigations to uncover both the barriers to licensure and potential solutions (Bibus & Boutte-Queen, 2011).

Minnesota’s Cultural and Ethnic Minority Infrastructure Grant (CEMIG)

In 2007, the State of Minnesota, within its Governor’s Mental Health Initiative, created the Cultural and Ethnic Minority Infrastructure Grant to increase access to and quality and availability of culturally congruent mental health services (Department of Human Services, 2006). The CEMIG program stems from Minnesota’s evolving conversation about the disproportionately low representation of licensed social workers from cultural and ethnic minority groups (Alexander & Johnston, 2010; McDonald, 2006; Minnesota Board of Social Work, 2008). The CEMIG program emphasized developing training programs for mental health providers from cultural and ethnic minority communities. The theory behind the grant program was that if the mental health workforce more closely resembled the state’s population, Minnesota’s racial disparities in access to mental health services would decrease.

Between 2008 and 2016, the CEMIG program allocated 8.86 USD million to fund 21 agencies through 33 contracts that were in effect from 1 to 5-year periods, administered within the Adult Mental Health Division and the Children’s Mental Health Division. Each division received the same state allocation of funds, but financed and administered them differently in terms of contract length, funded activities, and potential enrollees (Table 2). The Children’s Mental Health Division restricted participants to masters-prepared trainees from ethnic or minority backgrounds, and allowed for funds to be used for direct service activities, clinical supervision, evidence-based practice training, and infrastructure development to support third-party reimbursement activities and expansion into under-resourced areas. The Adult Mental Health Division was less restrictive in participant guidelines, targeting bachelor-prepared participants, interns, and non-clinical workers, while focusing on trauma-informed care through longer-term grant contracts with fewer agencies. Grant funds facilitated study groups and assisted cultural and ethnic minority licensure applicants with test taking fees, licensure costs, and enrollment in courses to enhance test taking skills. While the allowable activities were different in each request for proposal (RFP), they used the standard state process for publishing the RFP, accepting bidders, reviewing proposals by committee, and engaging in contracts with approved agencies. Contracts ranged from 31,000 USD to 300,000 USD per state fiscal year. Each division within the Department of Human Services had its own administrator, who processed payments and coordinated quarterly or annual grantee meetings. Each contracted agency had its own agency grant manager who coordinated activities for the participants, such as supervision and payment for continuing education credits and licensure test fees. As each agency proposed its own work plan and program, some focused on internal staff and others targeted interns or community participants. Evaluation of contracts and grantee performance was limited based on the diversity of programs, the modification of allowable activities, various contract lengths and terms, and grant-program tracking by multiple DHS administrators. To that end, DHS hired external evaluators to help provide documentation of the use of state funds.

Table 2.

CEMIG RFP allowable activities.

Children’s Mental Health 2007 RFP	Adult Mental Health 2009 RFP	Children’s Mental Health 2010–2016 RFP	Adult Mental Health 2014 RFP
1. Direct Mental Health Services: Provide mental health services to children from cultural or ethnic minority cultures.	1. Increase number of licensed mental health professionals from cultural and ethnic minority groups.	1. Clinical Supervision Activities: providing clinical oversight of post-graduate clinical experience of clinical trainees working on licensure.	1. Recruiting, supporting, training and supervising students from target communities toward careers in mental health services.
2. Capacity Enhancement Activities: Train providers in culturally appropriate evidence based practices and interventions effective with specific cultural populations.	2. Increase capacity of mental health staff to provide cultural and trauma informed mental health services.	2. “Start Up” or Expansion activities for increasing access for Greater Minnesota: Investments in infrastructure to facilitate culturally specific services outside of Metro area.	2. Expanding and building culturallyspecific, trauma-informed, adult mental health services.
3. Minnesota Health Care Program Billing Activities: Investments in infrastructure that facilitate third-party billing activities or capacity to obtain managed care contracts.	3. Increased opportunities for people seeking mental health professional licensure to work with diverse communities.	3. Evidence Based Practice Training and Implementation: Training and implementation costs for EBPs with culturally specific communities.	3. Expanding activities and services for increased access to adult mental health treatment for ethnic communities in Greater Minnesota.
4. Clinical Supervision Activities: providing clinical oversight of post-graduate clinical experience of clinical trainees working on licensure.		4. Direct Mental Health Services: Provide mental health services to children from cultural or ethnic minority cultures.

In terms of program results, the CEMIG program helped 104 individuals gain independent clinical licensure or clinical supervisor status in the state of Minnesota; 59 individuals are working toward licensure, at the time of writing. A total of 273 individuals received clinical supervision, mentoring, or training that was paid for or provided through the CEMIG program. Data collected was process-based (e.g., number of supervision sessions or therapy sessions for uninsured clients) rather than outcome-based (e.g., client score improvement on standardized measures), according to an evaluation of the program. Each DHS grant administrator created their own grant reporting system, and a standard outcome-based evaluation for the program as a whole was unavailable (Aby, 2020). Grant funds were used to support licensure fees, licensure exam fees, and practice tests. Agencies were also able to use these funds to support infrastructure activities such as creating electronic health records, expanding services to greater Minnesota, and engaging with telehealth technology. In addition, CEMIG paid for therapists to be trained in several empirically supported interventions or treatments, such as Parent Child Interaction Therapy, Narrative Exposure Therapy, Parent Management Training, Eye Movement Desensitization and Reprocessing, Trauma-Focused Cognitive Behavioral Therapy, and Motivational Interviewing.

Methods

To examine implementation processes and what salient elements emerge when transformative change programming is top-down, fund-driven, and broad-based, we apply a case study approach. Case studies are appropriate when participants can share insights and data from a real-life context (Baskarada, 2014; Yin, 2014); the social work practice (Gilgun, 1994) and public health programs (Scheirer & Dearing, 2011) often use case studies as a result. Second, case studies are appropriate for studies that are descriptive and that query the “how,” such as this present study. Case selection was based on the distinctiveness and uniqueness of Minnesota’s CEMIG as the first to implement programming to diversity mental health staff at a state-wide scale; remarkable events provide revelatory insights (Yazan, 2015). The research was designed as a single embedded case study (Yazan, 2015), whereby the adult and children mental health units were examined as distinct but part of a single organization. Analyses included content analysis of grant documents and qualitative thematic analysis of interviews with 22 participants.

Document review

Researchers reviewed all available CEMIG documents, including Request for Proposals (RFPs), RFP responses, grant contracts, grant renewal proposals, and grantee reporting data submitted to DHS. DHS provided all documents related to the CEMIG program currently held within the shared drive for the MHD and any paper or electronic documents in the possession of current program administrators. Using the Dedoose management system (SocioCultural Research Consultants, 2018), researchers organized the documents and analyzed the data using provisional codes (Miles, Huberman, & Saldaña, 2014), including agency name, proposed activities, reported activities, trainee names, and proposed and reported budget information. Researchers then collected and arranged these coded data to create matrices to demonstrate grant contracted outcome measurements and payment information for each contract period during the CEMIG program.

Semi-structured interviews

Researchers contacted current and former DHS grant program administrators, agency grant managers, and Minnesota behavioral health board representatives to participate in the implementation study. Since 2007, there have been nine DHS CEMIG administrators and two Grant Evaluators. Seven current or former DHS employees participated in the study. Out of 20 current or former agency grant managers who were contacted, 13 elected to participate. Several of the grant managers were unavailable due to retirement, a change in position or agency, or contact information that was not up-to-date with DHS. Two representatives from the four behavioral health boards in the State of Minnesota agreed to participate in this study. Researchers contacted all potential participants via workplace e-mails or social media (LinkedIn or Facebook) messaging services. When possible, researchers made follow-up phone calls to solicit study participation a week after the initial invitation e-mail. Researchers scheduled interviews at the time of the participant’s choosing, with several interviews occurring over the telephone due to geography limitations of the researcher. In total, researchers interviewed 22 participants: two board representatives, 13 grant managers, and seven DHS administrators. Please see Table 3 for participant demographics and characteristics. Researchers based interview questions and analyses on implementation literature (Aarons, Hurlburt, & Horwitz, 2011; Raghavan, Bright, & Shadoin, 2008) focusing on the implementation and delivery of the CEMIG program; see Table 4 for details. While scholarship has used implementation frameworks to analyze the implementation and creation of best practices in the dissemination of Evidence Based Practices, recent developments have also used these frameworks to analyze policy implementation (Aby, 2020; Stone, Daumit, Kennedy-Hendricks, & McGinty, 2019; Walker et al., 2019). The first author conducted the interviews, each of which lasted around 60 minutes; and digitally recorded, transcribed, and checked the interviews for accuracy.

Table 3.

Sampling table and participant demographics.

Race	DHS Administrator	Grant Manager	Mental Health Board Representatives
Black	3	2	-
White	3	5	2
Native American	-	1	-
Latino	-	2	-
Asian American	1	3	-

Table 4.

Sample interview questions.

Interview Topic	Sample Question
Exploration	Could you describe your agency’s process in applying for CEMIG funding? How did you determine you met criteria for the grant? How did you build your proposal? What were the key factors in your agency that made applying for the CEMIG program possible? (Look for organizational characteristics, culture, leadership, and values.)
Preparation	Were there items that you asked to be a part of the grant contract that were not approved? What were they and how did that change the implementation of your program? How were the contract goals and objectives/outcomes created? How were the contract deliverables created?
Implementation	What was your interaction like with your State grant manager during your CEMIG contract? How often did you interact? What was the interaction like? Have your agency’s processes changed because of the CEMIG program (like billing insurance, contracting, service programs)?
Sustainability	Has your agency sustained the CEMIG funded initiative at your agency? What has that process been like? What have been the key factors to sustaining? (Look for internal and external items.)
Social/Political Contexts	What were the environmental factors that influenced the implementation process? What were the other system changing initiatives occurring at the time?

Analysis

One coder (first author) analyzed transcripts and grant documents in two coding cycles, using Dedoose. The first cycle utilized provisional (Miles et al., 2014) or sensitizing codes (Gilgun, 2015) pertaining to broad external and internal organizational factors related to program implementation (Aarons et al., 2011). The first author conducted code mapping and operational model diagramming strategies as the second analytical step, which involved reorganizing codes and connecting and distilling concepts into central themes (Miles et al., 2014; Saldaña, 2016). The author utilized pattern coding for the second round of coding, which generated overarching themes that arose in the process and that were created from explanatory or inferential codes (Miles et al., 2014; Saldaña, 2016). The author used memo writing and member checking throughout data collection, analysis, and report writing in order to facilitate reflection, idea generation, and concept development (Gilgun, 2015). This study’s findings come from various data sources so as to develop ideas of replication and convergence (Gilgun, 1994; Palinkas et al., 2015). Using all of the data sources and information, researchers created constructs that are supported by a logical chain of evidence (Miles et al., 2014) and that are reported through main themes within the findings (Coffey & Atkinson, 1996; Gilgun, 2015). In the presentation of findings below, the authors have modified some quotes to enhance readability by eliminating repeated phrases, qualifiers such as “you know,” and phrases of habit such as “um” or “like”; and to eliminate dialect or speech patterns that could indicate a person’s ethnicity.

Reflexivity

Reflexivity refers to the researcher’s introspection about their position within the research process, and how that position may influence the research process and findings (Stige, Malterud, & Midtgarden, 2009). Reflexivity does not aim to “remove” the researcher’s influence, but can be better conceptualized as a self-aware consideration of the researcher’s positionalities and insights, in a way that develops and enriches the analysis rather than clouds it (Wiley, Jen, Storer, & Gonzalez Benson, forthcoming). The first author is a white female, licensed to practice clinical social work, and was a DHS administrator for this grant. The first author interacted in numerous ways with this grant program and clinical licensure within the State of Minnesota. As such, several interviewees or study participants were former coworkers and colleagues of the first author, in her DHS administrator capacity. The first author chose to not include an interview with herself to limit her voice and perspective within the report’s findings. The first author’s previous experiences as a clinical social worker and employee at DHS, as well as her relationships with the study participants, could have altered responses provided by participants and influenced their understanding of the research process (Finlay, 2002). Aiming toward reflexivity and rigor in implementing this study, the authors took several precautions throughout the research process (Wiley et al., forthcoming). First, the first author consulted with peers, some of whom had expertise in qualitative methods, and worked closely with the second author in analyses and theoretical formulations. Second, the coder/first author memoed after each interview, reflecting about her own positionality – particularly as a white woman and as a previous administrator – in relation to the data, participants, and research. She was intentional in considering how social location, experiences, and knowledge may be impacting the research process, specifically data collection and analysis. The first author also used memoing at key junctures in the coding and analytical process. Finally, the author engaged interview participants in a few areas: in member checks regarding theme development; in helping to shape the framing of themes; and in reflecting on how their experiences aligned with the overall theme. This reflective process brought additional perspectives and weight to the analysis.

Limitations

As the case study focused on one grant initiative, findings are not generalizable. Researchers contacted all grant managers, DHS administrators, and behavioral health board representatives for participation. Participation might have been limited by prior interactions with the researchers, fear of future funding repercussions, or the limited time window for data collection. Also, this analysis was retrospective in nature and thus subject to hindsight. Clinical trainees or practitioners who received support through the CEMIG program were not contacted for interviews for this study. Future research should include the voice of the clinical trainees, so as to address barriers to licensure and the individual experience of the process.

Findings

The theme programmatic value shows underlying support for the program and what it means, while two emergent themes, the revolving door and no buy-in, portray the struggles in program implementation. Programmatic value pertains to the overarching message of significance and benefits of the program; the revolving door relates to the high managerial turnover during the life of the grant, and no buy-in points to the lack of vision, promotion and prioritization of the program. These main themes are presented in this section with quotations from interview participants and grant documents; further conceptual discussion is presented in the Conclusion.

Programmatic value

An overarching message from the grant managers and DHS administrators was one of value. Regardless of the difficulties surrounding the program, all interviewees emphatically supported it. Participants emphasized how the grant opened up opportunities for advancement for both clinical trainees and agencies through financial incentives, infrastructure development, and increased accessibility to DHS.

Positive micro influences that CEMIG has provided for clinical trainees include building motivation to attempt the exam. An agency grant manager shared the following:

… In the past, some people may not have the money to even (attempt the exam). They’re afraid because when they don’t pass they don’t feel like they spent so much money. But this grant allowed people to be comfortable to say, “You know what? I’m going to even attempt it.” And our evaluation is not just based on how many people passed, but also how many even attempted it because by attempting it, they learned something. Maybe the next time around, they will do better. But before, it was almost gambling for people to say “Well, I don’t have 500.00 USD to waste.”

The grant allows the clinical trainees the freedom to attempt the test without being financially hindered. Grant managers discussed the competing financial interests of their clinical trainees: often the choice was taking the test or paying their mortgage.

The grant also created supervision infrastructure within the agencies. As agencies were struggling with meeting billable hour expectations and with funding, CEMIG provided a needed financial buffer to insulate a supervision system, allowing supervisors more leeway in budgets, and therefore more time with supervisees. One supervisor shared the following:

Well I think that the supervision issue is probably the biggest, most helpful thing because I think as a smaller agency, we really at that time—and even still a little bit, but not as much now—we really had trouble paying for the supervision piece and hours that that was taking. Those were potentially billable hours, but we have no income for that, and that actually really helped us in a large-scale organizational way. Those funds helped to support that. That’s the infrastructure piece that we really needed help with.

CEMIG had benefits that were tangential to the program as well. Agencies utilized their clinical trainee programming to recruit and retain staff. In a competitive market, especially for clinicians of color, smaller nonprofits are continually worried about staff turnover or recruitment by larger, better paying agencies. This grant allowed agencies to be competitive:

I would also say, that for us, we have been able to remain competitive because we can offer the support to staff. And sometimes when staff are thinking, do I go to [agency] and get paid 20,000 USD less than if I were to go to a hospital or the county, and we can say, and we’ll support your growth, and we’ll do this, and we’ll do that, it feels more attractive. So it at least bumps our ability to recruit and then retain, especially staff of color.

Recruiting and retaining staff who are consistently sought by higher paying agencies (such as government or for-profit entities) was a positive side effect of the grant program. Agency grant managers reported that even though the smaller, culturally based agencies pay less, with the grant, they were able to keep staff and retain clinicians – even recruiting from out-of-state:

For us, the need is so huge in having culturally responsive providers, so this grant has been perfect. I will give you an example of how—what a perfect fit this is. At some point, I was importing clinicians, right, because we didn’t have enough in Minnesota. There was a woman who’s now a clinician here who graduated from the University of Chicago. She was considering coming here but she could stay in Chicago. Her family was here but there were so many opportunities for her in Chicago. What the grant was offering made the difference, and then she decided to come because of the perks.

A final positive remark regarding the CEMIG program came from grant managers who expressed that this grant decreased stigma within the mental health community about DHS. Grant managers discussed a community-wide hesitancy to call DHS and a fear of getting into trouble with Medicaid rules or grant billing:

A lot of entities are really scared of DHS, and a lot of my colleagues or even my supervisors during my time at [agency] were very like “don’t call them.” But I guess I made really good connections, and I would always seek out individuals to make sure we were doing the right thing.

It was the connections with DHS administrators that could change the perception of DHS as unhelpful. Some grant managers described feeling better about the state agency because of the relationships they developed with their assigned DHS administrator, and the opportunity to learn about state rules and regulations and contract negotiations with a big state system:

I remember (that relationship) was a really positive experience. I actually learned a lot about what to look for and what questions to ask. It made me less afraid to ask questions, I guess, too, and know what questions to ask.

Lastly, each participant stated, in their own way, that while there were imperfections with the system, the grant program was beneficial to the state overall.

What I really want to emphasize is that the difficulties, which have been real, are nothing, pale in comparison to the benefits of the grant. When I travel out of the state I use this grant as probably the best example, or one of the two best examples, of how the State of Minnesota is so progressive and walks the talk.

Regardless of any internal difficulties DHS has with prioritizing racial equity work, the community external to DHS really saw the CEMIG program as evidence of both progress and commitment to improving culturally specific mental health services across the state. Throughout all the grant documents, there is uniform appreciation of what the grant offers from the voice of clinical trainees. One participant voiced the following:

My hope will be for this grant to be extended and expanded as it proved invaluable to our organization. The ripple effect translates broadly in the clients we serve, partners established, and the community at large. The product of this grant deeply had a positive astronomical outcome on me personally, my colleagues, our clinical team, staff, and interns; we became ambassadors and advocates for clients that were underrepresented and marginalized.

The revolving door

A symptom of the program’s lack of priority also came through in the large amount of turnover of grant administrators, which was mirrored in the grant agencies themselves. Over 10 years, there were nine grant administrators across the two Divisions, although some of the departures were due to natural job change. In the Adult Mental Health Division, administrators often changed positions every 9 to 12 months.

Although some DHS administrators enjoyed managing the CEMIG program and sought out the opportunity, others were passive about being selected to manage the grants: “I don’t think I was really asked for the opportunity as much as I was given it. They needed somebody that would oversee it, with not a lot of complaint or concern.” Or the opportunity was seen as a burden: “All I wanted was it off my desk.” Grant administration duties for the CEMIG program were “handed off with such dismissiveness, and that was also reflection of how they were viewed within our Division. ‘I’m going to just give this to whoever else is available.’” Several grant administrators speculated they were selected because of their own race or ethnicity rather than how the grant matched their job duties. They were also given the message that they shouldn’t do too much with the grantees, “It wasn’t directly shot down, but just that’d be too proactive for the environment. Just stay more on the reactive side or the neutral side of things.”

Lack of consistent leadership was also felt at the management level, as one participant states:

In terms of people, first of all, when you have people that just are disappeared as we like to call them now, that’s very weird. People just disappear. And you have that. I mean in my [time at DHS], I’ve had seven different directors, seven different—I mean, it’s incredible. So that leaves you with who do you go to? Who do you talk to about this?

The disappearing of staff and the revolving door of management all contributed to information isolation from one manager to the next. This created a perception that the DHS administrators were not helpful and lacked trust in the grantee agencies. Knowledge transfer was incomplete as the grants bounced from one desk to another.

The combination of having administrators who did not prioritize or have time for the program, and lack of departmental vision, created confusion and at times distrust with grantees:

DHS needs to trust the provider [in how] they approach us. When people start feeling that way, they may not participate. They need to be more trusting. They need to walk people through the system and give people the benefit of the doubt.

Meanwhile, the grantee agencies were also fraught with turnover at the clinical trainee, supervisor, and director levels. The time period from graduate degree to clinical license can be three to four years, which is a long time for an entry-level employee to have the same position. This was especially true because as clinical trainees benefited from the grant program, they would become more marketable. Agency staff stated, “I panicked every time someone talked about leaving”; and quarterly grant documents show a similar story. There was significant turnover because bilingual and “minority clinicians are in high demand, and for-profit organizations offer salaries and benefits out of reach for not-for-profit organizations.” The revolving door of clinical trainees made it difficult for agencies to feel like they had accomplished the goals of the grant:

Turnover is always challenging, and I feel like that was probably one of the biggest challenges for us being able to see somebody from start to finish go through and get their license instead of having this person advance from this to this, and this person advanced here to here. But—and most people, I think, eventually did go on and get their licensure–but there wasn’t one person, I don’t recall that—(who went) from start to finish through the grant.

Anxiety was not limited to staff turnover; many agency managers were also worried about maintaining their agency’s financial health as well. Several awarded agencies closed, merged, or barely survived during their contracted period. DHS administrators voiced difficulty in doing audits: “No, I couldn’t find payroll. In the case where I couldn’t see a payroll journal, I started asking questions. I couldn’t get answers. Then I find out that they really didn’t do any of their own books, that this other organization did the books. And they were kind of a pass through.” There were difficulties in billing and payment systems, and monitoring and auditing processes. One agency closed two weeks after receiving their first contract and advance. A review of financial records across agencies found a lack of proper documentation and fraudulent behavior. Agencies closed, merged, and cobbled together partnerships with fractured accounting practices that were not always aided by grant program requirements. It is important to note as well that CEMIG was launched at the beginning of the Great Recession and many agencies closed up shop during these years; these externalities may have been an additional stressor for this program (Graaf et al., 2016) At the same time, particularly during the precarious economic context of the Great Recession, the job security and stability gained from being public employed via CEMIG should have also incentivized less turnover rather than more.

Normal rates of turnover within direct service were augmented by the reality that there were so few individuals from communities of color that such employees were constantly in demand and being recruited. Agencies themselves closed, merged with each other, and went in and out of partnerships throughout the terms of the grant. The revolving door within DHS and the agencies themselves created situations that fractured relationships through distrust and confusion about roles. However, when the relationships between grant managers and DHS were nurtured, CEMIG created a community where managers and clinical trainees could problem solve, commiserate, and share training strategies. These notions are developed in the last theme.

No buy-in

DHS administrators described how CEMIG was not seen as a priority by leadership. Those outside of DHS described a lack of clarity concerning who was a grantee and how to appropriately access DHS resources. Lastly, there was confusion surrounding how to define the grant program, the social problem it was designed to address, and the appropriate stakeholders.

The CEMIG program was not a flagship program though it fell under the umbrella of the 2007 Governor’s Mental Health Initiative. Although other initiatives were allocated millions of state dollars each year (Department of Human Services, 2006), the combined financials of CEMIG was under 600,000 USD for both divisions; and while both divisions added Federal Block Grant dollars to their CEMIG programming, the grant amounts continued to be small in comparison to other DHS initiatives.

One administrator pointed to a lack of vision and community engagement, which kept the Adult Mental Health Division from fully developing the grants: “There was never a collective visioning of what the aim and purpose of this was. It was communicated down, ‘We have this funding now, let’s get it out.’ I don’t know that there was deliberate thought of what the intent is here.” The lack of intent manifested in a two-year delay for writing the first RFP, which was “indicative of how we value this work – an organizational culture that did not place value on the work of these grants,” the same administrator shared. The program flew under the radar of the division’s strategic plan and was not connected to the other infrastructure initiatives.

Community participants were frustrated that the grant program seemed detached from other initiatives. The only DHS-sponsored announcement about CEMIG grantees occurred in 2008 for the first round of the Children’s Mental Health Division grants, and there is no available information about the CEMIG program on the MHD DHS website. “I wish that the grant was probably a little more well known in the community or who are the recipients of these grants and if they’re doing something that other organizations can participate in or learn more about.” There was such a lack of communication between grantees that 30 clinical trainees were on multiple grant contracts without anyone realizing it. This created suspicion between agencies when recruiting participants. Relationship building across grant agencies and within the larger community was difficult: “Agencies that have never had anything to do with you before reaching out and wanting your people. That’s suspect.” Instead of creating partnerships across agencies, the grant program sometimes fractured relationships because motivation for recruitment and outreach was unknown.

Lastly, the purpose of the grant was not well understood or considered relevant to the constituency that the grantee agencies considered the most important – the national associations who create the exams and the state boards. An agency manager discussed how agency managers were told by a national exam preparation instructor that agency staff needed to “just think like a white man” to pass the test. National exams were described as seemingly “really culturally biased.” Agency managers reported feeling frustration that the Boards were not communicative about the cultural issues with the exam; they were told “this is a national test, and this is what we’re going with.” A behavioral health board representative stated that “85 to 90% of all our licensees report being Caucasian” and that “it’s usually in response to a conversation happening elsewhere” when legislation changes occur surrounding issues of language and accommodation. Behavioral health boards described themselves as regulators and protectors of the public, not as institutions that help facilitate licensure or address racial inequities. The representatives stated that they had received anecdotal reports about the difficulties with the tests and about barriers to licensure experienced by applicants of color (especially those for whom English was a second language), but they also felt these problems were not appropriately documented through data. (Some boards and national licensure exam organizations do not record demographic data of their applicants.) Further, when asked how the Board prioritizes cultural and ethnic work, the response shows that there is interest in such an effort, but limited commitment to it:

This board is very, very conscious and really has taken some initiatives and some pretty deliberate steps to really be conscious of diversity and to try to support more diversity in the workforce, to reach out to those communities. I mean, so there’s—there is that. In terms of passing the examination, where might that fit on the priority? I think, at this point in time, I don’t know that—I don’t know that it’s a real high priority.

Therefore, although the board representatives were open to discussions about the difficulties individuals from cultural and ethnic minority groups experience in the licensure process, there was limited active awareness of this issue, of the grant program itself, or of ways to address this problem within their scope of practice.

In sum, the original mission seemed cloudy or unknown, and there was a lack of public awareness about the grant program and the problem within the community, as shared in interviews. Findings suggest a lack of large-scale and meaningful buy-in across state systems on the importance of funding workforce development programs for mental health providers of color, and the program remained largely unknown and unsupported.

Concluding discussion: leadership matters and buy-in matters

Analysis demonstrates how the CEMIG program struggled, but at the same time was well loved and valued by the community. Drawing upon literature on planning transformative change (Netting et al., 2007) and theories of discursive institutionalism (Boswell & Hampshire, 2017; Schmidt, 2008), we posit two conceptual points as informed by our empirical findings: in funding diversity and implementing programs that aim to institutionalize diversity, leadership matters and buy-in matters. These points highlight the contested nature of transformative change within institutions, a contestation that funding alone cannot temper.

First, leadership matters: we argue for leadership as a crucial element in transformative change programs, as teased out in discussions around the theme revolving doors. Revolving doors describes the high amount of turnover within both the DHS administrator role and the grantee agencies, with nine people filling the grant administrator position over the ten-year life of the program. The lack of consistent leadership affected relationships with grant agencies and sometimes led to distrust. Leaders in grant agencies, as well as participants themselves, were also transient over the decade.

This in-and-out of leaders is thus posited as a primary challenge to program success, indicating more embedded structural challenges. Consistent leadership is crucial for overall grant management and the delivery of measurable outcomes. However, for transformative programming, such as CEMIG, leadership takes on compounded dimensionalities (Shier & Handy, 2016; Walter et al., 2017) because such programming entails engaging with complexities, and long-term, collective, dynamic processes (Netting et al., 2007). “An effective leader must have the capacity to envision what will be necessary going forward, rather than looking back to customary ways to solve problems or reach new goals,” (Walter et al., 2017, p. 213). Furthermore, leadership at CEMIG was largely white; having diverse leaders and decision-makers, who represent the aims of the program precisely, would help generate approaches and plans that are not only creative, innovative, and diverse, but are also reflective of the communities served. That is, especially for programming that aims for transformative goals, leadership is not only about being effective, but about having vision and generating and administering an organizational culture for social change to be carried forward (Mustafa & Lines, 2013; Shier & Handy, 2016; Walter et al., 2017). Leadership that is visionary, committed, trusted, and steadfast is crucial in leading such a transformation in organizational culture, a transformation that can be instituted only over time. Such leaders – including opinion leaders and those who courageously champion innovations and alternative ideas – are crucial in facilitating buy-in or ideological shifts among staff, communities, and the public. “Individuals (are) involved in the presentation, deliberation, and legitimation of … ideas to the general public,” in enacting deliberative communication (Schmidt, 2008, p. 310). In transformative planning, leadership is thus taken together as co-constitutive with broad-based ideological shifts, discussed subsequently as a second element.

Secondly, the buy-in matters; we argue that the ideas or values that underpin transformative change and their effective (or ineffective) communication are crucial elements in programming. No buy-in illustrates the lack of collective and broad-based shifts within CEMIG in terms of ideas or awareness about diversity and the fundamental principles that buoy diversity, while also indicating poor morale and burnout. “The transformative power of ideas and discourse (…) show how they exert a causal influence in political reality and, thereby, engender institutional change” (Schmidt, 2008, p. 306). Ideas, hidden biases, unrecognized prejudices, and taken-for-granted values (norms, standards, assumptions, principles, beliefs) circulate within institutions (Walter et al., 2017). These “background ideas” follow cognitive, emotive, cultural, rational, and logical rules that underpin how leaders and staff “make sense of things” within a given institutional context or setting (Boswell & Hampshire, 2017; Schmidt, 2008). Furthermore, sentiments also reflect weariness and poor morale, both crucially problematic for the important tasks at hand. This relates to the first point above regarding leadership, as preventing staff burnout and maintaining morale go hand-in-hand with initiating and pursuing big organizational shifts.

Lack of diversity among mental health providers was the explicit, targeted issue for CEMIG, and programmatic outcomes pertained to diversifying personnel. However, it seems that a more fundamental idea underpinning the problem of diversity was institutional racism, that set of “polices, practices, or procedures embedded in bureaucratic structure that systematically lead to unequal outcomes for groups of people” (Walter et al., 2017, p. 216). There was discussion within CEMIG about “cultural bias” in the content of the exams and in access to education and licensure, but these ideological, structural aspects were not tackled as part of the problem. For CEMIG, programming did not go deep enough in addressing and altering this fundamental logic, as “the language of social change is not easily converted into evaluable program planning products,” (Netting et al., 2007, p. 64). The diversity aims of CEMIG perhaps belied something more fundamental, the institutional racism embedded within institutional practice. Lack of diverse representation among service providers results from issues with recruitment or outreach in the present or recent context, but also from historical legacies of racism and the suppression of “ways of doing and thinking” of racialized communities within the context of mental health systems. Herein lies a key difference between focused change (diversifying providers) and transformative change (addressing institutional racism), as discussed above (Netting et al., 2007). Indeed, (re)defining the problem as agenda setting is most crucial but also a most challenging aspect of changing policies and programs (Kingdon, 1984; Schmidt, 2008).

In considering CEMIG, the works of Netting and colleagues (2007) and Schmidt (2008) offer insights for transformative planning, as detailed in Table 5. A first step in identifying and defining root problems and assessing needs would be acknowledgment of the institutional racism that buoys diversity problems, and then digging into and planning around the complexities and dynamics around that. Second, structural forms of intervention strategies may then go beyond recruitment of diverse staff, also tackling the predominance and privileging of white/dominant modalities and knowledges in service provision and organizational culture. Third, processes of goal setting and program design would be sufficiently collective and broad based (Netting et al., 2007). This would facilitate cross-learning and increased trust among community providers. Also, increased outreach to the behavioral health boards would lead to more concrete information concerning the struggle of applicants from cultural and ethnic minority backgrounds in relation to passing board exams. And beyond the providers and administrator within the closed network of CEMIG, intervention strategies would also extend communications and collaborations with other systems and institutions that are distal in proximity and temporality but crucially important in systemic change about diversity. Such institutions include higher education and civil society organizations that directly target marginalized communities of color as well as housing, transportation and other social services across state systems that would support individuals of color who are aspiring and working to be providers.

Table 5.

Transformational opportunities.

Planning Elements	CEMIG Barrier	Transformational Opportunities
Problem definition, identification and analysis	Recognized lack of diversity in mental health field, but did not assess systemic, underlying potential causes Internal DHS process without community involvement Disjointed definition and analysis across divisions.	Conduct a root cause analysis with stakeholder groups Identify structural barriers and involved systems Create ongoing advisory group that continues the root cause analysis and systemic issue identification throughout the grant process.
Needs assessment	Unclear focus of grant program across divisions Focus on the individual clinical trainee rather than the system of licensure Lack of community involvement.	Incorporate a needs assessment after root cause analysis Include system change needs in assessment Involve communities and other systems in needs assessment.
Intervention strategies	Lack of consistent allowable activities Focus of intervention was on the individual clinical trainee rather the system of licensure Disjointed strategy across DHS administrators and divisions.	Focus intervention on systemic needs, rather than individuals Involve community and other stakeholders in creating grant expectations Include external messaging regarding grant participants and programs.
Setting goals and objectives	Lack of specific, structural goal setting Lack of consistent program evaluation criteria Lack of community involvement in goal and performance expectations Lack of communication cross programs.	Utilize advisory group and community-based stakeholders to set systemic goals and program objectives Create reflective process for program evaluation that includes qualitative/qualitative components Create communication opportunities across programs Include policy makers and leadership in the evaluation and change process.
Program design decision-making	Focused on the individual clinician rather than systemic change Wide variability in contract deliverables Lack of consistent leadership across program years.	Broadly define parameters, including larger systemic interventions Include stakeholders in decision-making processes; Communicate knowledge gains and program information widely across systems.

Based on Netting et al. (2007).

Finally, as an overarching modality across these various elements of planning, deliberative communication allows for thinking beyond institutional practice, as institutional actors reason and debate among themselves about the very structures they use and aim to change (Schmidt, 2008). Communication in this sense is not so much an expression of normative values, or a directive, or a negotiation that entails exchange or compromising, but is more like persuasion and argumentation that “brings people in” to meaningfully see the necessity and appropriateness of the alternate vision or course of action proposed. Leaders are crucial here, as discussed above; transformative change leaders are not directive or normative. Rather, they mobilize and inspire staff and communities to bravely consider new paths. Also important in such communications are “critical junctures” or “windows of opportunity” for facilitating ideological shifts in policy and programming (Kingdon, 1984; Schmidt, 2008). These are key or momentous events that are often unexpected and external to the institution itself but place attention and support for a new, alternative or different idea that would otherwise be neglected; examples are legal cases, media stories, policy shifts in another domain or programming shifts in neighboring places. For example, the Black Lives Matter movement was external in the broader local or national context that can be utilized as a “window of opportunity” to galvanize transformative change within an institution. This is all the more pertinent within the specific context of Minnesota, where the killing of an unarmed Black man, George Floyd, by police led to a local national and global anti-racism movement in 2020.

At the same time that there were challenges with leadership and buy-in, there were examples from within the community that demonstrated the value of CEMIG, as suggested with the concept of programmatic value. Managers and DHS administrators recognized the importance of the program. Grantee agencies stated that the CEMIG program has been invaluable in recruiting and retaining employees, and in providing a much-needed safety net for a financially strapped industry. CEMIG was touted for providing key assistance to participants, but also for establishing relationships and opening communication between DHS and small community agencies. The program was regarded as demonstrating Minnesota’s level of progressiveness and commitment to battling disparities. The Minnesota case shares insights within a localized context or case; further research is needed about distinct elements across regions, to then theorize and/or generalize about processes for diversifying social services and tackling institutional change programming.

Programmatic value reveals the capacity for ideological shifts (Boswell & Hampshire, 2017; Schmidt, 2008). Indeed, even as institutions’ hidden biases and values are dogged and unyielding, ideas can nevertheless shift. Ideas are constraining, but they are also malleable and open to interpretation, reflecting “people’s ability to think and speak outside the institutions in which they continue to act” (Schmidt, 2008, p. 315). While CEMIG perhaps focused on the symptomatic problem of staff diversity, programmatic value illustrated an undercurrent that valued what CEMIG stood for. What was needed was buy-in at the structural, broader levels and an end to the revolving door of CEMIG; these changes would allow the kind of visionary, inspirational leadership and the ideological shifts needed to address institutional racism – that dogged, unyielding logic from which problems of diversity and disparities stem.