Money Moves the Mare: The Response of Community‐Based Organizations to Health Care's Embrace of Social Determinants

Abstract

Policy Points.

• Health policies that encourage health and social integration can induce community‐based organizations (CBOs) to adopt new ways of working from health care organizations, including their language, staffing patterns, and metrics. These changes can be explained by CBOs’ perceptions that health care organizations may provide new sources of revenue.

• While the welfare implications of these changes are not yet known, policymakers should consider balancing the benefits of professionalizing CBOs against the risks of medicalizing them.

Context

Recent health policies incentivize health care providers to collaborate with community‐based organizations (CBOs), such as food pantries and homeless shelters, to address patients’ social determinants of health (SDOH). The perspectives of health care leaders on these policy changes have been studied, but the perspectives of CBO managers have not.

Methods

Our research question was: How are CBOs in Massachusetts perceiving and responding to new Medicaid policies that encourage collaboration between health care organizations and CBOs? We interviewed 46 people in leadership positions at CBOs in Massachusetts for approximately an hour each. We analyzed these data abductively, meaning that we iterated between inductively coding transcripts and consulting existing theories and frameworks.

Findings

We found evidence of a knowing‐doing gap among CBOs. Even though CBOs value their distinctiveness and autonomy from health care, they have undertaken a series of organizational changes in response to the new Medicaid policy that make their organizations appear more like health care organizations. These changes include adopting new performance metrics, hiring clinical staff to the board and senior management positions, and using medical language to describe nonmedical work. Drawing on institutional theory, we suggest that the nonprofits undertake such changes in an effort to demonstrate legitimacy to health care organizations, who may be able to provide new sources of critically needed revenue.

Conclusions

Massachusetts CBOs perceive health systems as potential sources of revenue, due in part to an ongoing Medicaid redesign that encourages the integration of health and social services. This perception is driving CBOs to appear more like health care organizations, but the impacts of these changes on welfare remain unknown.

Health care providers are increasingly incentivized by health policies to address the social determinants of health (SDOH). ¹ , ² , ³ , ⁴ The growing literature on this subject indicates that interventions to provide housing, nutritional support, and transportation can both improve health and, in some settings, save costs for health care providers and payers. ⁵ , ⁶ , ⁷ , ⁸ , ⁹ , ¹⁰ These interventions often require collaboration between health care providers and community‐based organizations (CBOs) with expertise in social service delivery. Addressing patients’ social needs, even in partnership with CBOs, requires health care delivery organizations to make potentially substantial changes, including hiring new workforces and collecting new patient data. ¹¹ , ¹² , ¹³ , ¹⁴ , ¹⁵ The challenges and rewards of health care's efforts in this regard have begun to be studied. ¹⁶ , ¹⁷

Despite the large role that CBOs are presumed to play in meeting patients’ SDOH needs, ¹⁸ , ¹⁹ CBOs’ perspectives and potential responses to SDOH‐focused health policies have received little attention. In this article, we address this gap in the literature by reporting findings from a qualitative study of CBOs in Massachusetts during an SDOH‐focused Medicaid redesign in 2017/2018. Through qualitative data collection and analysis, we sought to understand how CBOs were perceiving and responding to health care's increasing efforts to address SDOH.

This analysis first describes a knowing‐doing gap among CBOs, in which they express appreciation for their distinctiveness from health care and their wish to avoid medicalization while at the same time making organizational changes to make themselves appear more like health care organizations. These changes include adopting new performance metrics, hiring clinical staff to board and senior management positions, and using medical language to describe nonmedical work. We then provide an explanation of how rational managers maintain this knowing‐doing gap. We suggest that the CBOs make specific organizational changes in an effort to signal legitimacy to health care organizations. ²⁰ , ²¹ They do so based on their belief that health care organizations are in a position to provide new sources of revenue in a resource‐scarce environment.

We present our findings not merely as summaries of what CBO managers said in interviews but also within a theoretical context in order to make the health management and health policy implications of the research more salient. To establish a knowing‐doing gap among CBO managers, we draw on managerial literature that has long recognized that managers (and organizations) face constraints in executing their values and visions, even when the content of those values and visions is clearly articulated. We use the term knowing‐doing gap to capture an element of the data that was clear from the very early phases of our coding and analysis, namely, the tension between CBO managers’ valuation of their distinctiveness and behaviors to adopt ways of working from health care. In explaining the knowing‐doing gap, we draw on a subset of sociological literature referred to as institutional theory, in which organizational changes are understood as efforts to increase organizational viability within an environment. Within a resource‐scarce environment, changes are successful in increasing viability to the extent that they signal legitimacy to those audiences who can provide the resources necessary for survival. The usefulness of the institutional theory literature became clear considerably later in our analytic process when we considered how to explain the changes that CBOs were undertaking. When we reconsidered our data with this literature in mind, we found ample support for a legitimacy‐based explanation of CBO behavior.

This article has several parts. In the first, we characterize the health policy context in which this study took place, both in Massachusetts and nationally. We then describe our methods, including our data collection and analysis procedures. Next we present our findings, in which we first establish a knowing‐doing gap and then draw on institutional theory to explain how rational CBO managers can maintain such a gap. Finally, we outline the limitations of our work, including plausible alternative interpretations of the data, and comment on the potential welfare implications.

Motivation

With an eye toward population health and value‐based care, health policymakers are recognizing the role of social and environmental factors in determining the health and health care utilization of low‐income and other vulnerable populations. ¹ , ⁴ This recognition was particularly visible beginning in the early 2000s in the movement in both domestic and global health to encourage a “health in all policies” approach. ²² , ²³ More recently, payers and policymakers have begun financially incentivizing health care delivery systems to identify and address SDOH, including housing, nutrition, transportation, and social isolation. ³ , ²⁴ , ²⁵ At least partly in response to these new incentives, health care delivery organizations nationwide have begun experimenting with various strategies to meet patients’ SDOH needs. These efforts are often divided between those in which health care “builds” new in‐house capacity and those in which health care “buys” services or capacity from CBOs. ²⁶ In the case of food insecurity, for example, several health systems have built an in‐house capacity by establishing on‐site food pharmacies, while others have partnered with CBOs to provide home‐delivered (and, in some cases, medically tailored) meals. ⁵ , ²⁷ This build‐versus‐buy framework is useful but oversimplified. In reality, represent two ends on a continuum of potential approaches.

Health philanthropies as well as the federal government have undertaken a series of high‐profile experiments in an effort to identify effective means of responding to patients’ most common social needs. ²⁸ , ²⁹ , ³⁰ The Robert Wood Johnson Foundation's Culture of Health framework has prompted several new grant‐making initiatives that aim to catalyze health care delivery organizations to become more involved in the development of local housing as well as other community‐level determinants of health. ³¹ The Kresge Foundation has followed suit, publishing a report on how community health centers can leverage the social determinants and also making priorities of both “institutional investment in community health” and “integration of health and human services.” ³² With similar goals in mind, the Center for Medicare and Medicaid Innovation has aimed to create new pathways for communities, rather than only health care organizations, to receive funding for health improvement efforts. Chief among these efforts is the Accountable Health Communities program, which was launched in 2017 and has provided funding to 29 communities to screen and address health‐related social needs. ³³ Housing and nutrition have been particular foci for these and other grant makers, owing to a somewhat circular logic that interventions with some preexisting evidence base may be the safest bets for further investment and study. Nevertheless, this focus has resulted in useful evidence reviews in these areas. ⁷ , ¹⁶ , ³⁴

State Medicaid offices have been particularly active in pursuing ways to incentivize providers to address SDOH. Several high‐profile 1115 waivers are intended to address social determinants in order to reduce health care utilization, and more than a dozen states, including Massachusetts, have created Medicaid accountable care organizations (ACOs). ³⁵ Massachusetts, Washington, Oregon, and New York have used Delivery System Reform Incentive Payments (DSRIP) approval to craft some of the most far‐reaching waivers. ³⁶ , ³⁷ Massachusetts is clearly seen as a leader in developing Medicaid ACOs and creating both carrots and sticks to draw attention to SDOH. Even so, Massachusetts is hardly the only pioneering state in this regard. Rhode Island, Minnesota, and others are also pursuing Medicaid redesigns with the enhancement of SDOH activities as a goal. ³⁸ Specific to CBO involvement, New York took a considerably more formal approach by requiring that provider performance systems (ACO‐like entities) have formal, paid contracts with CBOs in order to receive DSRIP dollars. ³⁹

Massachusetts is an appropriate setting in which to study CBOs’ relationship to health care organizations because of an ongoing 1115 Medicaid waiver implementation that prioritizes SDOH. This waiver was approved by the Centers for Medicare and Medicaid Services in 2016 and will have to be renewed in 2021. The waiver provides both requirements and incentives for health care organizations to address SDOH for enrolled Medicaid patients and to collaborate with CBOs to do so. Three aspects of the waiver are particularly important. First, health care organizations that wish to be designated as a MassHealth (Medicaid) ACO are required to demonstrate an ongoing “effort to address members’ health‐related social needs including expanding community linkages to social service providers.” Here the term social service providers is synonymous with community‐based organizations. Second, newly created MassHealth ACOs must screen for health‐related social needs in eight domains and are encouraged, but not required, to create relationships with CBOs to respond to positive screens. ⁴⁰ Third, MassHealth has allocated $150 million in new funding for a “flexible services” program in which MassHealth ACOs can use Medicaid dollars to pay for a limited set of evidence‐based nutrition and housing interventions that previously have fallen beyond the scope of Medicaid services. ⁴¹ These aspects of the waiver have created an opportunity to study the role of CBOs in pursuing health policy goals.

The health policy literature to date has focused on the challenges and successes that health care leaders face in responding to Medicaid and other payer SDOH‐focused incentives. The potential for health care institutions to derive financial and social benefits from investing in community partnerships, particularly for high‐need, high‐cost populations, has already been described. ⁴² , ⁴³ In most cases, commentators have encouraged partnerships between health care organizations and CBOs by highlighting the potential for health care delivery organizations to derive value from the relationship. ⁴⁴ , ⁴⁵ This article adds to these analyses by considering the perspective of CBOs on SDOH‐focused health policies and the prospect of closer collaboration with health care organizations.

Methods

Data Collection

We relied on qualitative data analysis to gain insight into CBOs’ perspectives. In 2019 we described some of these methods in an article in the American Journal of Preventive Medicine. ⁴⁶ Here it may be worth expanding on the rationale for our approach to data collection and analysis. In 2012, Karen Staller described the need for qualitative researchers to consider their methods as being nested within their methodologies, epistemology, and ontology. ⁴⁷ Studies using a particular ontological worldview must be consistent in the assumptions and choices they make at every analytic level. Accordingly, we proceeded from an ontological stance that the world is imperfectly knowable through study. We chose interviews to collect CBO managers’ perspectives based on a belief that people can be knowledgeable, if incomplete, reporters of their own lived experience.

Our sample population included the leadership and senior program staff of various CBOs in Massachusetts (n = 46). We defined CBOs as nonprofit organizations that provide direct social services and are not focused primarily on promoting or delivering health care. Given this definition, federally qualified health centers (FQHCs) and advocacy organizations were excluded from the sample. Organizations that focused on mental health or long‐term services and supports (LTSS) were also excluded because their role as health care providers was contractually specified in Massachusetts’ Medicaid ACO requirements.

We recruited CBO leadership and staff that had attended listening sessions and other events related to the integration of health and social services in Massachusetts. First, we drew from publicly available attendance rolls from open meetings hosted by Massachusetts Medicaid during the early redesign process in 2016/2017. Second, we drew from the Blue Cross Blue Shield of Massachusetts Foundation's participant list from a social equity convening in 2014. We continued to recruit eligible CBO leaders using snowball‐sampling techniques, in which all the interviewees are asked whether they know of other people who may have critical perspectives. We offered each one a $10 Amazon gift card as a token of gratitude for their participation. The resulting sample was 46 key informant interviews, representing 44 CBOs in Massachusetts.

After the interviews were complete, we sent a brief survey to all the interviewees asking for additional information about their employers. Forty of the 46 interviewees responded to this survey. Table 1 highlights the demographics of the CBOs represented by the interviewees (n = 46) and survey respondents (n = 40).

Table 1.

Organizational Characteristics of Participating Community‐Based Organizations

Interviewees	n = 46
Sector
Food	13% (6)
Housing	33% (15)
Community Centers	17% (8)
Legal Services	4% (2)
Multiservice Centers	15% (7)
Transportation	2% (1)
Workforce Development	11% (5)
Domestic Violence	2% (1)
Early Childhood Education	2% (1)

Survey Respondents	n = 40
Geographic Location
Greater Boston Metro	75% (30)
Western Mass	18% (7)
Eastern Mass/Cape	8% (3)
Number of Employees
<15	20% (8)
15–60	23% (9)
61–145	10% (4)
>145	48% (19)
Annual Operating Budget FY17
<$1 million	10% (25)
$1–5 million	9% (23)
$5–10 million	9% (23)
$10–20 million	2% (5)
>$20 million	16% (40)
Government Funding as a Total of Budget
<25%	43% (17)
25–50%	18% (7)
50–75%	10% (8)
>75%	10% (8)
Has a National Provider Identification Number	28% (11)

We asked the interviewees questions about (1) the mission and services provided by their organization, (2) their perspectives on health care organizations entering into social services provision, (3) their organization's funding structure, (4) past experiences with health care referrals or partnership, and (5) potential risks and benefits of health care entering into social service delivery. Our interview guide is shown in Online Appendix 1. We targeted CBOs from several social service sectors, with a greater emphasis on housing and nutrition, which the Medicaid redesign had identified as a priority. ⁴¹ Other sectors represented were domestic violence, community centers, multiservice organizations, legal services, early education, workforce training and job development, and transportation.

We conducted all the interviews between September 2017 and March 2018; they lasted between 40 and 75 minutes; and they were audio recorded and professionally transcribed verbatim. The research team conducted the interviews until they agreed that thematic saturation had been reached. ⁴⁸ This study was deemed exempt by the Tufts Medical Center Institutional Review Board.

Data Analysis

We pursued an abductive approach to our data analysis, meaning that we sought insight by iteratively consulting the data and existing theoretical frameworks (more later). ⁵⁰ , ⁵¹ We borrowed the term abductive from the work of sociologists Stefan Timmermans and Iddo Tavory, who argued that allowing theory to inform qualitative data analysis may be more likely to yield novel insights than following a purely inductive (in most cases, grounded theory) approach. ⁴⁹ Health services researchers specializing in qualitative methods have suggested a very similar approach without using the term abductive. For example, the seminal paper on qualitative methods by Elizabeth Bradley, Leslie Curry, and Kelly Devers recommended taking an “integrated approach [that] employs both inductive (ground‐up) development of codes as well as a deductive organizing framework for code types.” ⁵¹

We operationalized an abductive approach by first coding the data using a grounded approach without any particular guiding frameworks or taxonomies in mind. We both independently analyzed transcripts, generating an initial set of open codes to summarize key ideas. The two of us then reviewed the initial codes in tandem, reevaluated the codes, and combined them into larger concepts and coherent themes. We both iteratively reviewed all transcripts in this manner, incorporating and refining themes, and then adding and combining new codes as needed through a series of weekly consensus meetings over three months.

Only after this initial coding did we consult the theoretical literature to identify extant frameworks for understanding the pressures on organizations to conform. Because inductive coding indicated that CBOs were adopting ways of working that made their organizations more closely resemble health care organizations, the institutionalization theory literature provided ready frameworks. This literature has extensively studied such pressures from both a theoretical and an empirical perspective. With the basic premises of that theory identified, we returned to our data to consider whether the theory fit. We found that the codes that had already been identified were given new explanatory power, and we also developed a handful of new theoretical codes. Developing these theoretical codes allowed our analysis of the data to have explanatory rather than purely descriptive power. In our final analysis, we agreed on all codes and themes and systematically applied them to the transcripts.

To ensure the valid representation of perspectives from CBOs, the study team also empaneled an advisory group composed of a subset of interviewees as well as health care professionals. The advisory group was convened twice during the study period to review and provide additional feedback on study materials and preliminary results. In particular, the advisory board was consulted on the representativeness of the sampling frame, the clarity of the interview guide, and the extent to which inductive codes comported with their own experiences. The advisory board also helped disseminate findings within the Massachusetts CBO and health care communities.

Findings and Analysis

We present our findings in two stages: We first present data to establish a knowing‐doing gap among CBOs. We then draw on institutional theory to provide an evidence‐based explanation for how rational CBOs are able to maintain this knowing‐doing gap.

Establishing a Knowing‐Doing Gap: CBOs Value Their Distinctiveness and Yet Adopt New Ways of Working From Health Care

Our first finding is a knowing‐doing gap among CBOs (Figure 1). Although CBOs value their distinctiveness from health care and want to avoid medicalization, key informants described a series of organizational changes that they are undertaking in response to the new Medicaid policy, such as adopting new key performance metrics, hiring clinical staff to board and senior management positions, and using medical language to describe nonmedical work. These changes appear to directly contradict CBOs’ interest in remaining distinctive from health care and avoiding medicalization.

Figure 1.

CBOs’ Knowing‐Doing Gap

CBOs Describe a Desire to Maintain Their Distinctiveness and Autonomy From Health Care

CBO leaders routinely described their organizations’ work as distinct from that of health care organizations. While health care organizations were viewed as an industry built to address medical illness, CBOs viewed themselves as addressing social challenges like poverty, economic immobility, displacement and gentrification, and institutional racism. They viewed these issues as intersecting with but also distinct from the work of the health care sector. CBO leaders were quick to point out the differences in perspective between themselves and CBOs: “The medical world is the medical world, and I think the hospitals have a view of the world that is very different than I think the community‐based organizations have” (Identifier [ID] 21, Housing).

In some cases, CBOs aimed to highlight the distinction between health care organizations and CBOs by suggesting that whereas health care was intended to address sickness, communities (and, by extension, CBOs) were a source of health. CBOs described their view as distinct from health care's in that they focused on geographic neighborhoods rather than patient panels, took a long‐term view of social change rather than focusing on meeting short‐term metrics, and created change through tailored relationships with unique clients rather than scaling standardized interventions. One CBO leader raised this point by reflecting on a recent meeting they had had with health care personnel: “What we learned from our meeting a couple months ago is that we speak an entirely different language [from that of health care]. Sitting with a couple of the hospital people having lunch and we were talking for ten minutes and then we're like, ‘Actually we have no idea what you just said’” (ID 33, Housing).

Because CBO leaders value their distinctiveness from health care institutions, they voiced their fears that the Medicaid redesign and the larger SDOH movement would medicalize their work. Although the CBO leaders did not offer precise definitions of medicalization, they used the term medicalized to describe a CBO that had narrowed its focus to match that of health care organizations. This is consistent with the standard academic usages in health services research and sociology. ⁴⁶ , ⁵³ , ⁵⁴ In the academic literature, medicalization is the process by which human or social problems become seen as medical problems and are designated as such. For an organization to become medicalized, therefore, is to give its work a medical framing. This view was reflected in a statement on SDOH released by the Society for General Internal Medicine, which defined medicalization as the process by which “non‐medical issues become defined and treated as medical problems.” CBOs viewed medicalization as problematic because it would mean narrowing CBOs’ values and commitments. One CBO leader's concern about narrowing came through in the following quote: “I also think we shouldn't medicalize this broadly, because if we do, we will view this through a medical lens and not a human lens” (ID 13, Adult Day Service).

The interviewees shared their concerns that engaging with health policies and organizations would result in their work being commodified into discrete, and perhaps billable, products and services. CBO leaders emphasized that their work required taking a tailored approach to individuals, and in many cases, the path to success could not have been predicted ahead of time. The CEO of a community center suggested that health care organizations based their work on protocols but that protocols would not necessarily be successful in his line of work. He continued: “Sometimes it really comes down to the individual and the dedication that our staff have. There's no way that you could've written that out. There's no way you could've known that that was what would've ended up being able to … close that case” (ID 29, Community Center).

CBO managers expressed fears that the flexibility that defined their work would be lost, including one CBO leader who used the metaphor of “putting a box” around social determinants: “Social determinants have to be fluid to the needs of individuals. I think it's that fluidity that scares health care organizations. I am concerned that it will start to become this box that defines this” (ID 3, Community Center).

In their desire to maintain their autonomy, CBOs expressed concern about following in the path of behavioral health (BH) and LTSS providers, whom they viewed as once having been “community‐based” nonprofits but now had become part of the health care system. This was a notable comparison because in the Massachusetts waiver, the BH and LTSS providers were called Community Partners (note capitalization), whereas the CBOs were called community partners (lowercase). ⁴¹ The former received substantially more money and a larger role in managing the care of complex patients, but they did so, in the view of some CBOs, in exchange for some degree of independence. CBOs’ concern about a loss of independence was especially evident in our conversation with one interviewee, who offered the following comments about medicalization:

• I don't have some abhorrence to medicalizing things. That doesn't bother me the way I see it trigger other people. I do think, again, just speaking frankly, if you have a bunch of pompous asses running about anything, then you have a bunch of pompous asses running stuff.

• Interviewer: Is that an allusion to the fact that pompous asses run health care?

• Yeah. (ID 24, Housing)

Here the interviewee's reference to not having “some abhorrence to medicalizing things” underscores how widespread among CBOs are the concern and language regarding medicalization. The description of health care personnel suggests, among other things, a desire to avoid being subjected to their authority.

In addition to general expressions of concern about being medicalized, CBOs shared several specific fears about the risks of becoming more deeply enmeshed in health care delivery and financing systems. First, several of the CBOs with comparatively more experience working with health care in research projects and contracts voiced their fears about potential conflicts of interest between health care organizations and CBOs. The concern they raised was about an organizational‐level conflict in which the CBO has an incentive to change its policies or exert pressure on staff to act in health care's favor, rather than an individual‐level conflict in which a single employee's judgment or behavior is compromised by the opportunity to self‐deal. CBO leaders in housing, legal services, and community health worker organizations described the conflicts most pointedly and worried that their professional obligations to clients could be compromised if health care organizations were paying for their time. For instance, one CBO staff member anticipated local health systems’ wanting to gain access to low‐income housing for their highest‐cost Medicaid enrollees. To do this, the interviewee feared that health systems would attempt to negotiate access to the front of a several‐thousand‐person waiting list in Boston for low‐income housing. If health care organizations were to do so, this would put this CBO leader's staff in a challenging position, given their professional obligations to abide by federal fair housing laws. Concern about conflicts of interest also came through in discussions of community health workers: “I'm sorry but the minute [a hospital] hires you as their community health worker and they're paying the bills, how are you supposed to be the conflict‐free community health worker?” (ID 18, Legal Aid). This acknowledgment of the potential for conflict existed alongside, and in tension with, a more general sense that health care and CBOs serve overlapping populations and that each has complementary values to bring to people's lives. ⁴⁶

Second, the CBOs shared concerns that Medicaid and other health policies that incentivized health care to contract with social services would lead to a proliferation of unfavorable contracts for their organizations. We heard this both as a fear about the future and in stories that CBO leaders and staff shared about their experiences with health care organizations to date. The most commonly cited concern regarding contracting was that CBOs would be asked to sign exclusive contracts with one Medicaid ACO or health system. This concern was rooted in what CBOs observed of other health system “partners’” experiences, in particular specialty medical practices, BH outfits, and LTSS. “I think [these policies] raise some really interesting questions about whether in ten years there are going to be exclusivity expectations of social services providers the way that there currently are with these other Community Partners and what that means” (ID 18, Legal Aid).

Interest in pay‐for‐performance contracts was more lukewarm, as CBOs were aware that risk‐based contracting was becoming standard in health care and were eager to demonstrate their value according to these terms, despite their worry about how the terms of those contracts would be set. As one CBO leader said about pay‐for‐performance: “I'm probably not sophisticated enough to do it” (ID 5, Multiservice). A more general fear of many of the CBOs was that they would have few means of assessing their own value and negotiating with larger organizations with large legal teams. One CBO leader referred to health care's holding a trump card that it could play at any time in discussions with CBOs: “Doctors in particular, it's like, hey I'm all about the community… . And it's all good until somebody disagrees. Then the whole facade changes and it's like—here's my trump card” (ID 6, Housing).

Third, interviewees shared some fear about whether the scale of health care's operations would create a need for CBOs to consolidate. The CBOs recognized that health care organizations were wary of having to contract with multiple CBOs in order to secure coverage for all ACO enrollees in the state. A single statewide CBO partner was thought to be preferable. In response, CBO leaders and staff described the potential for both voluntary, interorganizational collaboration and the threat of involuntary consolidation. Although they regarded low‐intensity forms of collaboration, such as the formation of “collaboratives,” as positive, the CBOs worried that these low‐intensity forms of collaboration might one day lead to the loss of some of the CBOs’ organizational identities. Key informants described a particular concern about the viability of smaller, culturally specific CBOs that may provide little obvious use to a Medicaid ACO.

CBOs Adopt New Ways of Working From Health Care

Despite their concerns, the CBO leaders described a series of organizational changes under way. These changes made CBOs’ structures, processes, and language appear more like those of health care organizations. Viewed alongside CBOs’ emphasis on the importance of differences between health care and social services as well as their fears of medicalization, these changes appear discordant.

The apparent mismatch between CBOs’ stated values and their behaviors can be framed as a knowing‐doing gap. We use the term knowing‐doing gap as it is commonly used in the organizational literature, in which Jeffrey Pfeffer and Robert Sutton coined the term in 1999 to refer to a gap between organizational know‐how or values and action. ⁵⁵ This gap can be either latent, in the sense that the organization feels no contradiction or tension, or manifest, in which managers sense the resulting tension. By relying on the organizational literature's conception of a knowing‐doing gap, our usage differs somewhat from the way that health policy researchers typically employ the term in reference to a gap between scientific evidence and common practice. ⁵⁶

Some of the most apparent organizational changes that CBO leaders described involved bringing new kinds of people into the organization to facilitate relationships with health care. Board members with backgrounds in health care were described as valuable for their ability to facilitate introductions to health care organizations.

• I think since the ACA was passed, it's really created these opportunities to reach out in a meaningful way and think about partnerships. So I think it's probably the last four or five years…. We've actually added somebody from the health sector to our board because we can see that this is a real key link, these two areas. (ID 33, Housing)

• The new head [of a local health care organization] and I had lunch, and at the end of that hour and a half, he joined our board and that opened up the entryway for our organization and theirs to start thinking collaboratively and put money together. (ID 5, Multiservice)

The CBOs were equally interested in hiring staff with clinical backgrounds. The work of these professionals had proved to be fundable, and their perspective was given additional weight in a CBO environment.

• There's nurses here; registered nurses who take blood pressure and glucose weekly. It started as a pilot here … there's statistics that show a really marked improvement in both blood pressure and glucose. The first funder was a [hospital] community benefit program. (ID 10, Community Center)

• Our registered dietician can actually bill for services … just to be able to play credibly in the space—we have to. (ID 11, Nutrition)

• Our CEO says as soon as [a physician] was hired and started talking, people just immediately started listening. The MD, for good or for bad … it carries with it a privilege in being able to open up doors. (ID 2, Nutrition)

Some CBOs used nonclinicians to foster links between staff and health care. We spoke to a manager of one home‐delivered‐meals program who described the program's role as “to advocate and integrate our service into health care payment and delivery models” (ID 4, Nutrition). While this person was not clinically trained, she had previously worked in health care administration and believed that to be a key reason that she had been hired.

The CBO leaders also indicated that they were making their organization's metrics and performance indicators more closely aligned with what health care saw as valuable. In some cases, old metrics were being discarded in favor of ones more closely related to health: “We are changing our metrics. We've gone from pounds of food … more toward healthy meals” (ID 2, Nutrition). In other cases, new metrics related to health care utilization were being introduced alongside traditional metrics. “What we hope to do in our model is provide food interventions and then partner with all these fantastic medical and academic centers that are here in [name of city], to determine the impact of food interventions on both health and cost outcomes” (ID 2, Nutrition).

The CBO leaders showed enthusiasm for data and research on their operation and its impact. This enthusiasm took the form of new data collection, record‐keeping systems, and internal analyses, as well as externally funded academic research. They talked about data and research as unequivocally positive goods that should be used to inform decision making.

• We have the [housing] resident fill out this health impact questionnaire, which asks questions about your housing experience before moving in, and the size months after moving in. So questions about frequency of ER visits, connectedness to neighbors, self‐reported health measures like stress and general health. (ID 6, Housing)

• Five years ago, our access to any kind of health outcome data [and] health or school outcome data was close to zero. I think we've advanced some of those conversations in a way that we weren't able to talk [about] health care well enough beforehand to do. (ID 25, Housing)

In addition to buttressing internal operations and decision making, research projects conducted with reputable academics were a means of demonstrating value to health care organizations.

• In order to counter [health care skepticism,] we've engaged in some pretty serious research projects. We just wrapped up a study … and the data show cost savings. It's there. [Health care leadership] just couldn't believe it, and so they asked [the lead researcher] to go back and do further analysis. (ID 4, Nutrition)

The CBO leaders also described new efforts to package their social services in ways that would be understandable to health care providers. These efforts included the adoption of health care–centric language, including terms like clinical, triage, and patient. In several instances, CBOs also described the creation of “menus” of CBO services for health care managers, particularly from the Medicaid ACOs, to review. The CBO leaders created these menus to allow health care organizations to retain a sense of choice and control over their engagement with the CBOs, which, the CBOs believed, would facilitate the creation of partnerships. Although pricing was not mentioned, references to creating a “menu” also signaled at least an implicit awareness that a prospective health care partner might be interested in knowing the prices of various CBO services.

• So the idea is to provide training [to local health care organizations,] and we have a menu that we have created. (ID 36, Community Health Worker)

• I'm not asking [staff] to start something new or develop a brand new program or bring it new staff or new training. I'm asking you to repackage what you already have. (ID 3, Community Center)

The CBOs’ embrace of health care's ways of working existed alongside their desire to maintain their distinctiveness and autonomy. The coexistence of these statements and actions thus can be understood as a knowing‐doing gap. ⁵⁵ No CBO leader acknowledged the tension between wanting to maintain distinctiveness but also wanting to become more similar to health care. In fact, it was the same CBO leader who voiced the strongest caution about health care's dangerous interest in “putting borders and boundaries” around SDOH to create “a service line” that, later in the same interview, revealed that he was urging his staff to consider “repackaging” what his organization had done in the past so that health care might become interested in purchasing this CBO's services.

Explaining the Knowing‐Doing Gap

The remainder of this article aims to provide an explanation for the apparent contradiction, or knowing‐doing gap, that is created by the CBOs’ pride in their differentiation from health care and their adoption of health care's ways of working. Insights from organizational theory, specifically institutional theory, shape this analysis. Excellent summaries of institutional theory are available in the managerial literature and have been previously brought to bear in the work of health services researchers Mary Dixon‐Woods and Kenneth White. ⁵⁷ , ⁵⁸ The key tenet of institutional theory, sometimes referred to as new institutionalism, is the idea that organizational characteristics are modified to increase the compatibility of an organization with its environment. ²¹ , ⁵⁹ Viewed through this lens, the organizational modifications just outlined are intended to increase CBOs’ ability to survive in a resource‐scarce environment.

Two institutional theory papers are frequently cited in the sociological literature. The first, by John Meyer and Brian R. Rowan, argues that within institutionalized environments, pressure to follow high performers leads organizations to adopt structures such as organizational charts and vocabularies. ²⁰ This pressure to mimic the high performers, the authors argued, is what accounts for the homogeneity in organizational fields. The second paper, published six years later by Paul DiMaggio and Walter W. Powell, extended Meyer and Rowan's work by outlining more specifically how pressures from outside an organization can create changes in not only structures but also internal decision‐making processes and behaviors. ²¹ Both papers, and the prodigious literature that has followed, view organizations as willing to adopt rationalized myths in order to be seen as legitimate. Organizations can then use this legitimacy to garner resources. Here we should pause to say that many articles, including this one, that rely on institutional theory combine its insights with those from a second literature in organizational theory, known as the firm's resource‐based view (RBV). The RBV literature assumes that highly rational managers change an organization in order to maximize output or minimize costs in light of technological or market changes. Even though institutional theory comes from a sociological, traditional, resource‐based view of the firm, such articles tend to be written more often by economists. Particularly in the nonprofit sector in which CBOs operate, in which the product's technical aspects are not well defined and therefore rational behavior is difficult to prescribe, institutional explanations of behavior provide additional explanatory power above and beyond what traditional, rationalist views can provide.

Informed by these analyses, our findings indicate that CBOs’ willingness to adopt health care's ways of working can be accounted for when CBOs are viewed as resource‐dependent organizations. In short, CBOs experience themselves as “poor” and health care as “rich,” and so they try to mimic health care's ways of working in an effort to demonstrate their legitimacy to health care organizations and traditional funders that may be able to provide critical resources (Figure 2). One CBO leader summarized their willingness to adapt to the new Medicaid and Medicaid ACO policies: “Money moves the mare” (ID 1, Multiservice).

Figure 2.

Explaining CBOs’ Behavior Through the Lens of Institutional Theory

Next we look at the data to support this analysis, including findings that CBOs experience resource scarcity within their own organizations, perceive health care organizations as being resource rich, and undertake deliberate strategies to secure resources through partnerships with health care.

CBOs Experience Resource Scarcity

CBOs consistently described a sense of profound scarcity of resources. Even among relatively larger organizations with more employees and larger annual budgets, the sense that the organization was failing to fully meet the community's needs and therefore required additional resources was pervasive. All of the CBOs in our sample operated on the basis of more than one revenue stream, often including federal, state, and local government grants or contracts, as well as foundation grants and private donations from people in the community. One CBO leader summarized the impact of resource scarcity on organizational leaders by describing them as operating in a “mindset of poverty” (ID 1). Other CBO leaders echoed the sense of resource scarcity:

• We are stretched to the max … always looking for hidden pots of money. (ID 12, Nutrition)

• Our individual shelters run about a $1 million deficit every year from state and federal funding, so we have to make that up with private donations. (ID 32, Housing)

Among the CBO leaders in housing with whom we spoke, many mentioned that people placed on a waiting list for affordable housing in Boston could expect to wait nearly a decade, which they saw as a systemic and moral failure. CBOs held municipal and state governments partly responsible because they centrally managed public and affordable housing waiting lists, but all who worked in housing felt a sense of urgency to prevent evictions and avoid unnecessary gentrification, to work to find creative solutions for currently homeless and housing‐insecure people, and to ensure that new housing units were being developed. In many cases, this work required resources that CBOs did not have available.

The experience of resource scarcity led many CBOs to be agnostic about the kinds of funding for which they would apply. Some CBO leaders indicated that this led CBOs (including their own organizations) to be un‐strategic about how they procured resources. Out of a concern for their organization's own survival, CBO leaders were prone to applying for funding that required the organization to creep beyond its stated mission.

• Anything that looks like new, viable funding sources will be very interesting to anybody because the contracts that we work with now are very, very lean and everyone is always looking for additional sources of funding. (ID 26, Housing)

• The reality is there's never going to be enough funding for what we would like to do, so any funding is always gonna be helpful. I think one of the challenges that nonprofits across the board have to figure out is that it can be very easy to fall into the trap of just chasing money. (ID 32, Housing)

CBOs Perceive Health Care Organizations as Resource Rich

In contrast to their own poverty, CBO leaders perceived health care organizations as large, resource‐rich environments. Descriptors related to size often preceded mentions of local health care organizations (e.g., “the big hospital”). The resources that interviewees cited were primarily, but not exclusively, financial.

• We will never have the resource ability or large number of [inaudible] that a hospital will have. (ID 5, Community Center)

• They have a certain amount of money sloshing around that they can invest, unlike most human service agencies. (ID 1, Multiservice)

• I think anything we can do to get service providers, particularly hospitals that have pretty significant resources at hand to connect with community‐based organizations to work together … is a good thing. (ID 11, Nutrition)

CBOs believed that recent changes in health policy, and most notably the introduction of DSRIP dollars as part of Massachusetts’ 1115 waiver, were just the latest in a line of past instances in which health care had received large injections of cash from the state.

In addition to financial resources, CBOs also described health care organizations’ professional and political reputations as key resources. In interviews, CBO leaders were critical of themselves (e.g., “Maybe I'm naive” and “We are the lowest rung on the ladder”) but spoke of health care personnel as powerful (e.g., “ACO bigwigs” and “phenomenally sophisticated”). CBOs also described health care entities as having access to policy discussions with MassHealth personnel in which the CBOs were not included, in part because of their size.

We're a fraction of the size of institutions like that. So we can sometimes feel like we're scrambling … and also feel like things can change quickly, at a [policy] level that we don't have access to, which might have a real impact on the way our business runs. (ID 29, Community Center)

CBOs Aim to Secure Resources From Health Care

Some CBO leaders described efforts to “be like” health care, particularly using health care language and metrics, but they did not explicitly link this desire to a strategy for attracting new resources. In this case, therefore, we must infer their motivation to adopt health care's ways of working. Other CBO leaders were explicit about how the changes they were undergoing were driven by a desire to win additional resources. Several CBO leaders indicated that changes to Medicaid policy had prompted a race among CBOs to win the attention of health care organizations, with the goal of joining health care organizations in commercial relationships.

• Right now, the big thing is trying to jockey and get your service to be seen as valuable by health services. (ID 1, Multiservice)

• Some of our [CBOs] perceive the [health systems] more as a source of potential income. (ID 3, Community Center)

If CBOs were successful in having their services be seen as valuable and entering into paid relationships with health care organizations, their leaders recognized the potential for positive spillover effects. CBOs also reported that even if their relationships with health care organizations did not generate financial resources for the CBO, they still could provide symbolic power that might help them generate funding from other funders. “[Hospital name] is such a big player that it enhances [CBO name] to have a relationship with [hospital name]” (ID 43, Legal Aid).

Institutional theory is a useful framework for interpreting the changes described by CBOs, by highlighting the ways in which organizational changes are efforts to increase CBOs’ viability in a resource‐scarce environment. Conceptually, the mechanism by which the adoption of health care's ways of working would result in increased resources is through an increase in legitimacy. Sociologists constantly debate the definition of legitimacy, and neither Meyer and Rowan nor DiMaggio and Powell define the term. For our purposes, legitimacy can be roughly defined as the perception that CBOs are socially valuable and organizationally capable.

DiMaggio and Powell outline three ways in which an environment can cause an organization (such as a CBO) to change in order to appear similar to another organization (such as a health care organization). ²¹ They call these changes isomorphisms, formally defined as “a constraining process that forces one unit in a population to resemble other units that face the same sets of environmental characteristics.” The authors then discuss coercive, mimetic, and normative varieties, each of which are visible in our data (see Table 2).

Table 2.

Mimetic, Normative, and Coercive Isomorphisms

Type	Definitiona	Relevance to CBOs’ Strategies	Illustrative Quotation from CBO Leaders
Mimetic	Changes intended to model one's own organization on other successful organizations.	CBOs aim to mimic health care organizations that they view as successful in generating resources.	And so that's a big part of what we try to do, is to get involved with folks before . . . an absolute crisis. And that's a big part of what we do is, prevention first is our motto. (ID 27, Housing)
Normative	Changes that come with “professionalization.”	CBOs undertake changes in an effort to persuade health care organizations or other funders that they are capable partners.	Our national organization has said that this should be a priority for us. [CBOs] should be concerned about this and thinking about ways to do a better job integrating with health care organizations. (ID 8, Community Center)
Coercive	Changes intended to comply with laws, regulations, or expectations or that “stem from pressures on the organization by other organizations on which the former depends.”	CBOs that are in a relationship with health care organizations are forced to comply with health care expectations and preferences or risk losing resources.	The [health care organization] that was funding us found out that we were training home health care aides and placing them in agencies. Then they came back and said to get rid of all the other employers and they could be the only one. (ID 16, Job Training)

^aDefinitions from DiMaggio and Powell. ²¹

Organizational theorists commenting on DiMaggio and Powell's work have critiqued the distinctions among the three types as being overwrought. ⁵⁹ , ⁶⁰ , ⁶¹ Indeed, differentiating among the three mechanisms in our findings related to CBOs is a challenge. For instance, distinguishing the changes that CBOs undertake as a means of professionalizing from the changes that they undertake in direct response to coercive health care requirements may be impossible, particularly when CBOs are trying to anticipate what health care organizations will want in a partner. Similarly, changes undertaken for normative purposes may be virtually indistinguishable from those undertaken as part of mimesis, given that health care organizations are perceived by CBOs as being strongly professionalized.

Limitations

Several types of limitations deserve note: those related to our study setting, those related to our methodological approach, and those related to our analytic assumptions. Our study was set in Massachusetts during a large‐scale redesign effort by Medicaid. We investigated the impact of changes in Medicaid policy on CBOs and recognize that shifts in other kinds of health policy (Medicare, commercial insurance markets) may affect the CBO landscape differently. Moreover, the insights generated from a study of Massachusetts Medicaid changes may not be generalizable to other states’ Medicaid reform efforts. Massachusetts has a large number of socially vulnerable people using Medicaid owing to its effectively universal health insurance coverage, and its Medicaid waiver has specific features that emphasize the development of relationships between health care delivery organizations and CBOs.

We selected our methodology to generate exploratory insights into a new, policy‐relevant topic area rather than to test hypotheses, make comparisons, or systematically survey a population. Because our research design was not constructed to generate insights into differences among types of CBOs (e.g., food pantries versus emergency shelters), comparative work should be considered for further inquiry. Our research also was collected at a single, critical point in time, making it impossible to describe longitudinal changes. An additional limitation that is unique to this study deserves note as well: because we began sampling with a set of CBO leaders who had attended public events related to health policy and SDOH, we may have an overrepresentation of CBO leaders who are actively following and planning for changes in this arena, as opposed to those leaders who are (and may remain) unengaged.

Furthermore, our analysis rests on two analytic assumptions to which readers may object. The first is that when CBO managers described an interest in maintaining their distinctiveness from health care along with changes under way to make their organization more closely resemble health care, this indicated some level of dissonance—which we termed a knowing‐doing gap. An alternative explanation may be that there is no “gap” between CBO managers’ statements and their behaviors, because CBOs are only adopting behaviors from health care that do not threaten their distinctiveness and autonomy. This explanation would lead to a characterization of CBO managers who are surgical in adopting the “good” parts from health care that enable CBO legitimacy while shielding their organizations from the parts of health care of which they are wary.

The second analytic assumption is that CBOs are adopting health care's ways of working in order to demonstrate legitimacy to health care organizations rather than for some other reason.

An alternative analysis might suggest that the new ways of working described would have been attractive to CBOs no matter where they came from and thus were not necessarily attractive to CBOs because they were coming from health care. This explanation would emphasize the efficacy or efficiency of such changes considerably more and diminish the role of legitimacy. In this interpretation, the motivation to adopt would not be related to CBOs’ perceptions of health care at all.

We chose not to pursue these explanations because of the specific comments that CBO managers made about how health care organizations are viewed and funded in comparison to CBOs and the value that they placed on adopting ways of working that were well established in health care circles. The latter point can be seen most clearly in CBOs’ hiring health care personnel and adopting language that is used almost exclusively in health care (e.g., triage). Nevertheless, we recognize that ours is not the only available interpretation of the data. Analysts who have alternative theoretical frames available to them could reasonably identify other explanations for the behaviors that CBO managers reported.

Implications

We have endeavored to discuss the changes CBOs in our sample are undertaking without judgment. We cannot predict whether adopting health care's ways of working will ultimately be a positive or negative development for CBOs or the health and social service ecosystem. As our article is among the first to document the changes under way in CBOs as a result of the Medicaid redesign, many of the most salient implications remain unclear and open to future research. Two questions about the implications of these findings are to be expected: (1) How widespread are the changes among CBOs? (2) Should we interpret these changes as normatively good or bad? Although we cannot answer either of these questions fully, we will comment on how we might approach them in the future.

First, the scale of the changes described here is unknown. CBOs in Massachusetts have no doubt faced pressures to institutionalize and professionalize that predate the Medicaid redesign. For years before the most recent 1115 waiver, the “health in all policies” movement certainly paved the way for CBOs to conceptualize their work as contributing to public health, and the rise of “impact investing” may have pressured CBOs to adopt more rigorous measurement standards. ²² , ⁶² We contend that the arrival of the Medicaid redesign and the explicit focus placed on health and social service integration amplified and extended such pressures. But knowing precisely what proportion of the changes documented can be traced to what source is difficult if not impossible. Furthermore, our sampling of CBOs in the early phases of the statewide Medicaid redesign may have led us to over‐ or underrate sectoral changes within the CBO landscape. Our goal for this research was to bring the changes among CBOs to health policymakers’ attention rather than to estimate quantitatively how widespread these changes are. Longitudinal research will be necessary to understand whether our findings detected an initial flurry of excitement that will taper off over time or are early harbingers of sectorwide trends.

Second, it is also not yet clear what the welfare implications of the CBO changes described here will be. We have used the term medicalization throughout because the CBO managers with whom we spoke frequently used it. Many readers will likely see the term “medicalization” and view the changes described here in terms of what is to be lost. In 2019, Paula Lantz took this view in a Milbank Quarterly editorial. ⁵³ However, one woman's medicalization (a term with clear negative connotations) can be another woman's professionalization (a term with mostly positive connotations). Particularly for those concerned about inefficiencies among CBOs, the potential for a Medicaid redesign to prompt professionalization among CBOs may be welcomed. For instance, whether the changes that CBO managers discussed regarding the development of research and evaluation capacities should be viewed as positive or negative developments will almost certainly be open to interpretation. The same may be said for hiring clinical social workers or health care–related staff.

Our research highlights the way in which policies that incentivize the integration of health and social services stand to create new markets. If a market emerges in which health care organizations are the buyers and CBOs are the sellers of social services, economic theory may be the most useful in hypothesizing the implications for welfare. We might imagine that the creation of a market would improve welfare, as the market would make an otherwise confusing social service landscape more understandable to health care managers. The beginnings of such a market were seen in efforts by health care policymakers in North Carolina to institute a fee schedule for CBOs addressing housing instability, food insecurity, transportation insecurity, interpersonal violence, and toxic stress. ⁶³

In order for such a market to deliver on the promise of welfare enhancements, CBOs would need to compete on the attributes of their services that are most directly tied to service quality or effectiveness. Previous economics literature has argued that when some of the many relevant outcomes are measured better than others, the markets will become distorted. ⁶⁴ , ⁶⁵ If one dimension (e.g., organizational form and practices) is measured better than others (e.g., quality), organizations will compete more aggressively on the better‐measured dimension, thereby creating a dysfunctional market. Standardized quality measures for social services are underdeveloped at the moment, making quality among CBOs virtually unobservable. However, our findings indicate that observable attributes such as the presence or absence of health care professionals, the presence or absence of performance metrics related to health care, and the presence or absence of menus of services may be identifiable to buyers. If these observable qualities are good proxies for unobservable qualities, economic welfare could be enhanced by the emergence of a market. In such a scenario, it would be reasonable to anticipate that lower‐quality CBOs may be out‐competed and face closure. It may also be reasonable to expect considerable consolidation among CBOs who would be looking for market power in negotiations with health care buyers. Conversely, if these observables are poor proxies for—or in competition for resources with—an unobservable quality, then the emergence of a market may be inefficient and diminish welfare. For decades, health care has been embroiled in an effort to identify reliable indicators of quality that would allow consumers and payers to select higher‐value purchasing. ⁶⁶ Despite these efforts, critics continue to warn that hospitals compete on amenities like lobby fountains rather than dimensions of care that indicate quality. ⁶⁷ This is only the most recent iteration of a long‐standing concern about health care competition being premised on an “arms race” that fails to yield welfare enhancements. ⁶⁸ , ⁶⁹ It seems now that the same debates may be coming to the social services sector.

The implication of changes within CBOs on dimensions of welfare that are not related to health should also be monitored over time. For example, the literature offers several analyses of CBOs’ impact on political representation and social capital. Jeremy Levine's 2016 sociological analysis of Boston's CBOs found them to be more legitimate representatives of urban poor neighborhoods than even elected representatives. ⁷⁰ In Levine's fieldwork, CBO leaders echoed our findings presented here by talking about neighborhoods as “my neighborhood” even when the speaker was not a legal resident. Several decades earlier, in his work on the medicalization of AIDS service organizations, sociologist Roy Cain decried the loss of political representation and engagement that medicalization had brought, writing, “Political engagement [in AIDS service organizations] is difficult when working within a circumscribed job description and when one is subject to the supervision and control of others who may not share these ideological commitments.” ⁷¹ As CBOs draw themselves closer to health care organizations, the loss of their ability to represent clients politically is a consideration that should not be overlooked in favor of purely utilitarian analyses of health‐related welfare.

Conclusions

We began by asking, How are Massachusetts CBOs perceiving and responding to a Medicaid redesign that incentivizes health care delivery organizations to address SDOH? What we found is that CBOs perceive this policy (and others like it) as a strategic opportunity to develop closer relationships with health care organizations. Many hope that such relationships could yield a new source of critically needed revenue for CBO operations.

Although we found evidence of a knowing‐doing gap, we must emphasize that CBO leaders may be acting rationally and, in many cases, appear to be quite managerially sophisticated operators. Many have led the same organization for decades, stewarding tight budgets through political shifts, managing inadequately compensated staff with high turnover rates, and providing unglamorous services to the most vulnerable among us—often out of jerry‐rigged physical spaces lacking in technological supports.

CBOs operating in a health policy environment in which SDOH are prioritized face a catch‐22. They value the fact that their mission is different from health care, both on principle and because capabilities in areas like food, housing, and transport make them potentially complementary partners to health care organizations. But in order to remain a viable organization that can maintain or grow its mission, they need resources. In order to secure resources from traditional funders or health care, they need to signal legitimacy and are doing so by adopting health care's ways of working. Given this construction of the problem, CBO leaders risk straying from their missions if they tether their organizations to health care resources and risk underperforming and forgoing critical resources if they do not.

Funding/Support: Blue Cross Blue Shield of Massachusetts Foundation provided funding for this research.

Conflict of Interest Disclosure: Both authors completed the ICMJE Form for Disclosure of Potential Conflicts of Interest. No conflicts were reported.

Acknowledgments: The authors would like to thank the Blue Cross Blue Shield of Massachusetts Foundation for funding this research and several anonymous reviewers whose feedback strengthened the manuscript considerably.

Supporting information

Online Appendix 1