Becoming resilient: Community‐driven change and the Civic Capacity Index

Abstract

One of the principal features of successful community governance is that it is collaborative and thus dependent on a community's ability to work together. However, there are no valid, comprehensive means to assess a community's capacity to respond to civic challenges in collaborative ways, and that are predictive of community resilience and well‐being. In this study, we describe the conceptual foundations of the Civic Capacity Index (CCI) and the multiple methods used to validate it. Findings from a large statewide sample show the CCI's scales to have high internal reliabilities and conform to a bifactor model, using confirmatory factor analysis. As hypothesized, CCI scores are predictive of multiple measures of community processes that in theory should be related to civic capacity; e.g., trust in governance, social control and cohesion, and community connectedness; public health indicators such as child abuse and suicide rates; and especially community resilience and well‐being, even a year after baseline. The empirical findings align with the conceptual basis of the CCI, in particular, the association between a community's civic capacity and resilience. This tight‐knit relationship has profound implications for promoting community resilience by building civic capacity.

Key Highlights

• Although collaborative community governance is often valued by diverse stakeholders, especially by those invested in the civic capacity of communities that make them more resilient, there is a paucity of valid measures of civic capacity.

• This article describes the theoretical foundation of the Civic Capacity Index (CCI) and the multiple ways that its reliability and validity were assessed and documented.

• The utilitarian value of the CCI is described in terms of how communities' resilience and well‐being may be promoted by using it as a springboard for forward‐looking inclusion of diverse perspectives and partners.

What form of governance helps communities solve exigent problems while building their capacity to address future challenges? In the prevention science field, research indicates that when residents are actively involved in civic affairs, communities are more successful in addressing civic challenges (Chilenski et al., 2007). Similarly, the literature on accountability in governance suggests that communities with durable and trusting relationships among elected officials, organizations, institutions, and citizens are more likely to attain shared goals (Lee & Ospina, 2022). From a broader perspective, the manifold features of civic muscle precipitate social change by mobilizing people and resources, which fosters community resilience and stability (Roulier, 2020). Successful governance in this sense is collaborative and dependent on a community's ability to work together.

Missing from these complementary perspectives is a valid means to assess a community's capacity to respond to civic challenges in collaborative ways. Communities have little insight into what capabilities are needed or on how they can be enhanced (Chandra, 2020). As Lee and Ospina (2022, p. 64) observed, the literature on collaborative governance and civic capacity does not have “… an overarching framework that allows for systematic assessment of accountability and illumination of the processes through which collaborators hold each other accountable.” Our applied research aims to fill this gap by developing a theory‐driven measure, the Civic Capacity Index (CCI), that assesses multiple facets of collaborative governance and civic capacity. In the sections that follow, we describe the civic arena and the conceptual foundations of the CCI before focusing on how it was developed and validated.

THE CIVIC ARENA: CHALLENGES AND CONTEXT

Think of the civic arena as a crossroads where diverse people with divergent interests contend for power and influence. It is distinct from traditional descriptions of the political sphere because it must take on issues that go beyond the jurisdictional boundaries of government, or because those in political positions either fail to or cannot provide effective leadership on challenges of common concern. The challenges in the civic arena are both collective and adaptive. They are collective in the sense that they affect everyone, that each of us bears some responsibility for how we respond, and to make progress, they require a collective rather than a narrow parochial response. The challenges are adaptive in the sense that they require engaging others in mutual learning, problem‐solving, and shared accountability. The COVID‐19 pandemic and the varied responses to it illustrate such a collective and adaptive challenge.

Context matters in leadership and governance. Today's challenges have become more intractable, and the context more divisive and contentious (Head, 2018). The decline in trust of expertise, institutions, and authority figures undermines faith in traditional sources of leadership. The emergence of performative politics—a politics of symbolic gestures—that does not solve problems, or try to, distracts from addressing challenges. An enervating and insular civic culture with little sense of shared responsibility, agency, or urgency erodes the collective capacity to respond. This zero‐sum factional contest of interests cannot respond to collective challenges.

Rather than ruing this predicament, we can turn to what we know about how to respond to civic challenges in more productive ways. More than a century ago, Mary Parker Follett (1918 [1998]) and John Dewey (1916) began to explore how to make democracy more inclusive, constructive, and responsive. Contemporary scholarship builds on these foundations. These ideas–resilience, collaborative governance, civic capacity, collective leadership, and social capital–illuminate what we know about the theory and practice of community governance (see Conceptual Foundations below). We know that collective and adaptive challenges require collaborative problem solving and mutual learning to create collective responses. We know that communities that have the capacity to work together make more progress on these kinds of challenges than those that do not. We know what it takes to make collaboration work. We know that when collaboration works, it solves problems, builds civic capacity, enhances the civic culture, and deepens democracy. The CCI assesses our ability to do this necessary work.

CONCEPTUAL FOUNDATIONS

Community resilience

The concept of community resilience frames our work. Resilient communities are not only able to withstand stressful incidents; they also are able to make the situation better than it was before the crisis while strengthening the community's ability to address future challenges in more productive ways (Chandra, 2020; Norris et al., 2008; Pares et al., 2017). Many features of community resilience align with processes related to collaborative governance and civic capacity. For example, a systematic review of how 80 articles defined community resilience identified nine core elements, including aspects of civic capacity (local knowledge, community networks and relationships) and collaborative governance and leadership, as well as communication (Patel et al., 2017). A different perspective was provided by Gillespie‐Marthaler et al. (2019), who defined the highest level of community resilience in terms of preparedness: The civic capacity to “systematically anticipate, prepare, respond … through structured approaches … to achieve desired community performance under both expected and unexpected circumstances” (p. 2490). As White et al. (2015) noted, resilience relies on dynamic processes that help coordinate efforts from diverse groups and institutions to marshal the resources for cross‐sector engagement. Given this close relationship, one strategy for validating the CCI is to assess its correlation with measures of community resilience and well‐being.

The relation between civic capacity and resilience can also be understood by deconstructing measures of community resilience. For example, Lindberg and Swearingen's (2020) thrive‐oriented scale of perceived community resilience is designed to assess the influence of community leadership or social capital; i.e., civic capacity is a key predictor. Notably, they use “thrive” instead of “cope” or “recovering” to describe the potential for positive outcomes. This choice of terms foreshadows our finding that in no counties where we collected CCI data were the average scores above the midpoint on the 1 (Never) to 6 (Always) scale; the CCI content is aspirational and constructive. Banyard et al. (2025) shed additional light on what constitutes the most informative measures of community resilience. In their scoping review, they found that (a) the content of community‐level measures often focused on deficits or risk factors, and (b) community characteristics predominantly were assessed at the individual level (e.g., depression symptoms; subjective well‐being) rather than the community level. With the expansion of community research, they are forecasting that collective efficacy and healthy democratic processes are likely to be revealed as vital strengths.

A resilient community that is high in civic capacity is a healthy community in part because of better access to health care and other resources, and social connectivity (Morton & Lurie, 2013). For example, when communities have strong social networks and infrastructure, there is better self‐reported health and lower mortality; higher resilience buffers against poverty‐related health disparities (Camacho et al., 2026). Community‐level protective factors are also associated with significantly lower suicide risk and greater mental well‐being (Stacy & Schulkin, 2023). As well, a comprehensive literature review identified some community protective factors that promote resilience in abused and neglected children, especially school connectedness, neighborhood quality, and informal support from an extrafamilial adult (Jean‐Thorn et al., 2022). Finally, resilient communities that are high in social capital and collective efficacy, as well as a sense of community, have significantly lower crime rates (Stuart & Taylor, 2021). Thus, another strategy we used to validate the CCI was to examine the association between counties' civic capacity and measures of health, mental health, crime, and child abuse.

Collaborative governance

Collaborative governance equates governing with organizing collective efforts to address shared challenges (Chrislip et al., 2023). Governance is a process rather than a structure, and collaboration is the prevailing method. Multiple stakeholders–governments, institutions, organizations, community members–engage as equal partners in addressing civic challenges (Boyte, 2005; Lasker & Weiss, 2003). In this sense, democracy is not restricted to the provision of public services and the popular vote. Rather, it manifests as shared responsibility for co‐creating solutions to issues that affect the community writ large (Lee & Ospina, 2022). This shift from hierarchical to collaborative governance requires embracing different values and more complex processes. In practice, collaborative governance often emerges when no single institution has all of the authority or resources to address the complex challenge alone. Within these contexts, various stakeholders must develop new forms of shared leadership and inclusive processes for discussion and decision making.

Through trustworthy processes, purposeful inclusion, and meaningful dialog, communities discover common purpose, identify challenges, create and implement solutions, and hold each other accountable (Briggs, 2008; Pares et al., 2017; Trust for Civic Life, 2025). They achieve equitable outcomes by including diverse voices, especially those who historically have been excluded from power, who are better able to confront issues of discrimination and social justice (Boyte, 2005; Chrislip et al., 2016; Hicks, 2020; Lasker & Guidry, 2009). Accountability depends upon open sharing of information about how the collaboration is working, achieving goals, and generating public value. These processes hold all members of the collaborative accountable. This sense of mutual obligation and accountability serves as a foundation for civic capacity and community resilience (Lee & Ospina, 2022).

Civic capacity

Conceptually, civic capacity can be defined as the “capability of neighborhoods, communities, and regions to plan, learn, and act together to solve problems in the public interest. … It is a shared resource that can be amplified for a specific purpose, squandered if not attended to, or intentionally built up” (Chrislip et al., 2023, p. 17). It consists of multiple ingredients: individual and collective agency, a civic culture open to including diverse voices, informal networks, formal institutions, and knowledge and skill for leadership and collective engagement (Briggs, 2008; Chrislip et al., 2023; Pares et al., 2017; Stone et al., 2001). Shinn (1999) described it as collective action, the key components of which are social capital, civic competence, and civic enterprise. Civic capacity is not static. Communities develop it over time through repeated collaborative efforts to problem solve and engage in collective learning about what worked and did not work. With enough civic capacity, communities can identify and make discernible progress on public challenges through their collective ability to solve problems in an inclusive and constructive way.

Collective leadership

In the civic arena, no one has the absolute authority to act, and community members do not necessarily follow. It takes leadership from multiple factions and perspectives to create the collective credibility to act (O'Malley & McBride, 2023). In this conception, leadership is a collective, dynamic process that helps communities solve problems and build a collaborative civic culture (Couto, 2016; Uhl‐Bien & Ospina, 2012). Rather than emphasizing leaders and followers, collective leadership stresses the relational activities of people within the social system, especially communication across sectors (Lee & Ospina, 2022). It expands on the idea that leadership is an activity, not a position or authority (Chrislip & O'Malley, 2013; Heifetz, 1990; O'Malley & McBride, 2023). One person's leadership activity invigorates others' involvement, leading to unpredictable and synergistic actions that build the collective credibility to act in the service of a common purpose (Hicks, 2020; Pares et al., 2017). For example, in coalitions, institutions work with other sectors, and community members are equal collaborators; thus, leadership inhabits systemic dynamics (Ospina et al., 2020).

A related term, civic muscle, considers community members' sense of belonging and agency to shape their common world through collaborative engagement (Chilenski et al., 2007). With the active involvement and leadership of many community members, communities are more likely to possess the civic infrastructure to create a healthy, resilient community (Roulier, 2020). Kretzmann and McKnight's (1993) guidelines for effective community interventions provide an example of this approach. They distinguish between a deficiency‐oriented treatment of residents as “clients”—consumers of services—as opposed to a bottom‐up linkage of diverse stakeholders that binds communities together through mutual support and problem solving (also see Trust for Civic Life, 2025). This notion of collective efficacy reliably predicts success in making changes to organizations (Armenakis et al., 1993) and communities (Chilenski et al.; Kretzmann & McKnight). As Powell and Menendian (2024) emphasized, fostering true belonging requires dismantling systems of “othering” that exclude or marginalize individuals instead of cultivating spaces where all people are recognized as co‐creators of the community. In this way, building civic muscle is not only about strengthening collaboration and resilience, but about inclusion as an essential condition for thriving (Powell & Menendian, 2024).

Social capital

Social capital concerns the social resources available to communities to address civic challenges. It is a key contributor to community resilience because it concerns how communities navigate crises by coordinating responses (Fan & Lyu, 2021; Kyne et al., 2026). It focuses on the networks and relationships among community members and the quality of these ties as reflected in norms of trust, reciprocity, and cooperation. Trust can be in reference to local government officials or neighbors, or it can be manifest in leaving doors unlocked and the number of hours volunteered (Aldrich & Meyer, 2015). These connections—particularly “bridging” social capital that supports information exchange and cooperation among diverse groups—have been recognized as critical social determinants of health, economic mobility, and community resilience and well‐being (Kyne et al., 2026; Putnam & Garrett, 2020). However, government agencies still prioritize investment in physical infrastructure as a foundation for community resilience despite evidence that social capital has a greater impact (Aldrich & Meyer). Success in collaborative governance depends on mobilizing these assets to act together to make progress on civic challenges.

Framing the research

Our intent is to create a more comprehensive understanding of community‐driven change in practice, not just in the abstract. We want to build on the strengths and address the limitations of the foundational ideas we described. For us, collaborative governance and civic capacity are the most capacious and robust of these conceptions of civic change. These two theoretical conceptions provide a powerful and practical framework for our work.

We created the CCI in part because there were no other validated assessment tools grounded in contemporary scholarship to measure civic capacity. We examined measures that assess concepts related to civic capacity and found them wanting in terms of diagnostic capability. These measures often describe static conditions, reflect attributes instead of actions, or collect subjective impressions in place of identifying concrete behaviors (Fischer et al., 2020). Fischer et al. also noted that in most of the research on leadership, the predictor often is conceptually close to, if not indistinguishable from, the outcome, which makes refutation unlikely. Their critique boils down to the following: “Questionnaire measures of leadership do not capture what leaders actually do or what their observers actually see, feel, or think, or how they would actually react to a certain leader behavior or information environment in a given context” (p. II). Our construction of the CCI takes account of this critique by describing concrete behaviors and processes related to civic culture. These are attributes of collaborative governance distinct from the outcomes of interest in this study—community resilience and public health indicators.

Furthermore, these extant measures do little to assess the dynamic processes related to collaborative civic engagement. They are also not comprehensive or specific enough to provide guidance for building and mobilizing civic capacity in communities (Chandra, 2020). For instance, the Civic Health Index (National Conference on Citizenship, n.d.) focuses on more traditional measures of civic health, such as volunteerism, public participation, voting, and social connectedness, not on the dynamics of civic engagement. We also recognized that the existing qualitative research on civic capacity (Briggs, 2008; Pares et al., 2017; Stone et al., 2001)—essential as it is—offers little insight into how civic capacity changes over time. We wanted to know what civic capacity looks like in sufficient detail to assess the extent to which it was present or not in a community and measure how it evolves over time.

To do so, we needed to know what civic capacity looked like in practice, so we framed our research around the question: What would you see if community‐driven change is occurring? We posed this question to an expert panel and engaged them in brainstorming, clustering, and analyzing their responses to our query (Chrislip et al., 2023). Through collaborative inquiry, we developed and validated a CCI and created a normative framework for civic engagement. We describe this process and its results next.

Development and preliminary validation of the Civic Capacity Index

The development of the CCI is described fully in Chrislip et al. (2023). Briefly, we convened a panel of experts who had extensive experience in community governance as elected officials; academic researchers (e.g., in political science, public administration, civic leadership, and community psychology); and/or well‐regarded directors or practitioners from community nonprofit agencies. They participated in a concept mapping procedure (Trochim, 2017), the first step of which entailed generating effective versus ineffective features of community‐driven change. In the next step, panelists sorted nonduplicative descriptors of community‐driven change into categories of their choosing. With the removal of redundant items and editing them for clarity, the resulting beta version of the CCI had 52 items in seven categories.

The current study describes multiple methods of documenting the psychometrics of the CCI that resulted in the current 42‐item version. These include its (a) internal reliability and stability over 1 year, (b) factor structure to determine alignment with the concept mapping results, (c) convergent validity with multiple measures of community processes that in theory should be related to civic capacity, (d) criterion validity in assessing the relation between CCI scores and community resilience and well‐being, and (e) construct validity in terms of group differences. Given that there may be a halo effect of relying on citizens' perceptions of multiple community dimensions (e.g., Ruddle et al., 2023), we also assessed (f) correlations between CCI scores and county‐level data from Census Bureau and public health sources.

METHOD

Participants

Study 1

The 52‐item CCI beta version was field tested with 484 community members and policy makers in 16 Colorado and Kansas counties. Participants and cultural brokers were identified through online resources and snowball sampling, with the online resources including government officials, boards and commissions, volunteer and human service organizations, community foundations, school and college staff, advocacy and leadership organizations, businesses and news outlets, and religious, political, and minority‐serving organizations. Invitations to participate were emailed to 10,254 individuals, resulting in a 4.72% return rate. The only eligibility criterion, as stated in the email invitation, was that participants be moderately familiar with local governance. Consistent with this criterion, 92% of respondents said that they had been involved in some form of community collaboration, 39% were or had been in paid leadership positions, and 50% said that they were or had been in an unpaid leadership position. Almost all consumed local news (paper, TV, online) daily (59%) or once or twice a week (30%).

Study 2

The same method as in Study 1 was used to recruit participants for validation of the 42‐item version of the CCI. Invitations to participate were emailed to 39,317 individuals, resulting in a 4.32% return rate (N = 1699). The demographic profile for both studies did not differ in terms of income (M = $60,900), age (M = 49.86 years; range = 18–94), gender (69% female), ethnicity (81% nonHispanic White), and education (M = 16.23 years; range = 6–24). Respondents also were similar (within 2%) to the Study 1 sample in their involvement in community leadership.

Demographics for the combined samples (N = 2183) indicated that those who completed the CCI are not entirely representative of their community, with the exception of income (M = $59,786). All other demographics differed significantly between the county CCI samples and their corresponding county census profiles, t(58) > 3.10, p < .003, with the sample being older, having a higher percentage of females and nonHispanic Whites, and higher education. This bias toward more advantaged respondents is likely a product of asking them to be at least moderately familiar with and engaged in local governance (see Baum, 2015).

Longitudinal Follow‐Up

Stability and change on the CCI and validation scales were assessed 1 year after the initial surveys were completed by those in the early waves of data collection (n = 428 completers; 56% return rate). All of them completed the 42‐item version; 329 were from the Study 1 cohort. There were no significant demographic differences between those who completed and those who did not complete the follow‐up.

Missing data

Little's MCAR test showed that CCI and validity scale scores were missing completely at random, χ ² = 0.32, p = .57; 1.2% to 13.6% of scores were missing, with few missing a CCI scale (3%). Therefore, missing data were imputed using the fully conditional MCMC strategy recommended by Schafer and Graham (2002), with 20 iterations specified.

Recruitment

For all components of the study, emailed participants were asked to complete a Qualtrics survey—available in multiple languages—related to local governance. The university's Institutional Review Board (IRB) approved the study, the IRB's contact information was provided in the email soliciting participation, and implicit consent to participate was given if they responded to the Qualtrics survey. Respondents were anonymous unless they provided their email to receive a report about the findings, which was optional. Qualtrics fraud detection software was used to identify and remove bots and cheaters.

Measures

Civic Capacity I ndex

The six scales of the 42‐item CCI are briefly described below. One of the CCI's original scale titles, “confronting racism and injustice,” evoked antagonism from some participants who perceived a political agenda. With input from our expert panel, we revised the scale titles to be more neutral and descriptive:

• Leadership, or how well the community collectively exercises leadership to address shared challenges.

• Inclusion, or the extent to which the community addresses challenges—including those of race, discrimination, and social justice—in inclusive and constructive ways.

• Institutions, or how well institutions in the community support addressing community challenges.

• Civic Culture, which is the shared attitudes, practices, and values of the community that help or hinder addressing shared challenges.

• Coalitions, or how well the community coalesces around addressing shared challenges.

• Collaboration, or how well community members work together to address shared challenges.

Social justice

The Racial Equity Map (Adimu et al., 2020) is a matrix of 18 items in six categories of an organization's culture, for example, power, programs, public policies, and systems. With guidance from the Racial Equity Map authors, we developed a six‐item Social Justice short form based on the matrix's scale anchors. Items such as, “We have a process in place to have safe and brave conversations about race and injustice,” are rated from 1 (Strongly disagree) to 5 (Strongly agree). This Social Justice short form has good internal reliability, α = .86.

Collective efficacy

The Collective Efficacy Scale (CES; Sampson et al., 1997) assesses perceived community social cohesion, trust, and ability to achieve socially important tasks. The five‐item Social Cohesion scale (α = .74) asks respondents to rate their agreement with statements that assess neighborhood cohesion and trust (e.g., “People around here are willing to help their neighbors”), rated from 1 (Strongly disagree) to 5 (Strongly agree). The Social Control scale (α = .82) has five items, rated from 1 (Very unlikely) to 5 (Very likely), that ask respondents to rate how likely another community member would intervene to benefit the neighborhood if various adverse community‐based events occured. Much evidence supports the CES's validity (Sampson et al.), including positive correlations with indicators of neighborhood resources and inverse relations with measures of neighborhood problems.

Community connectedness

The 10‐item Social Connectedness Scale (Lee & Robbins, 1995) assesses sense of belonging to one's community (e.g., “I don't feel I participate with anyone or any group,” reversed) and feelings of respect (e.g., “In my community, my opinion is taken seriously”). Each item is rated from 1 (Strongly disagree) to 5 (Strongly agree). The scale has adequate internal consistency (α = .81) and good factorial validity.

Trust in governance

This scale includes four items adapted from the Grunig Relationship Instrument (Hon & Grunig, 1995); e.g., “I am very willing to let my local government make decisions for people like me.” An additional item was included from the Pew Research Center (2022): “Just about all of the time I can trust my local government to do what is right.” Each item is rated from 1 (Strongly disagree) to 5 (Strongly agree). The 5‐item scale has good internal reliability (α = .84) and is correlated with the CCI and other validity scales (Table 1).

Table 1.

Correlations between Civic Capacity Index and validation scales, with scales' Cronbach alphas.

Scale	2.	3.	4.	5.	6.	7.	8.	9.	10.	11.	12.	13.	α
Civic Capacity Index
1. Leadership	.79	.72	.75	.70	.73	.61	.61	.39	.42	.39	.61	.53	.89
2. Inclusion	‐‐‐‐	.74	.79	.70	.74	.69	.57	.37	.43	.41	.63	.53	.91
3. Institutions		‐‐‐‐	.84	.75	.78	.56	.79	.48	.44	.39	.69	.56	.91
4. Civic Culture			‐‐‐‐	.83	.84	.64	.72	.50	.50	.44	.75	.60	.92
5. Coalitions				‐‐‐‐	.85	.54	.65	.47	.46	.39	.67	.53	.92
6. Collaboration					‐‐‐‐	.60	.74	.46	.44	.40	.70	.56	.96
7. Social Justice						‐‐‐‐	.55	.34	.46	.15	.59	.49	.86
8. Trust in Governance							‐‐‐‐	.52	.56	.51	.78	.64	.84
Collective Efficacy
9. Social Control								‐‐‐‐	.57	.47	.66	.52	.82
10. Social Cohesion									‐‐‐‐	.71	.65	.53	.74
11. Comm. Connectedness										‐‐‐‐	.59	.51	.81
12. Comm. Resilience											‐‐‐‐	.74	.90
13. Comm. Well‐Being												‐‐‐‐	‐‐‐‐

Note: For all r > 0.21, p < .0001

Community resilience

We used 10 items from the Conjoint Community Resiliency Assessment Measure (CCRAM; Braun‐Lewensohn & Sagy, 2014; Leykin et al., 2013) to assess community resilience. An example item is, “My community has had more success than failure in solving local issues,” rated from 1 (Strongly disagree) to 5 (Strongly agree). The scale authors found CCRAM scores to correlate with stress reactions and state anxiety, especially in rural and communal communities in Israel. Cronbach's alpha was high in the current study (α = .90).

Community Well‐Being

We used Gallup (2009) adaptation of Cantril's Ladder to assess participants' perception of “how your community is doing right now,” rated from 1 (suffering) to 10 (thriving) with struggling as the midpoint.

We converted all survey scores, including the CCI, to a 0‐ to 100‐point scale so that they were in a common metric based on the community well‐being scale, and so that scores would be more easily interpreted by community partners.

County‐level health indices and covariates

Current county‐level demographic and community resilience‐related public health data were retrieved from census, state public health, and government repositories. In two of the state's multicounty rural areas, there were too few child maltreatment and suicide cases in half of the six counties per region to include in the analyses. These counties also had fewer than 10 CCI respondents. Therefore, we created weighted means for these two regions' CCI scores, demographics, and public health data and omitted their constituent counties from the analyses. The counties within each rural area had similar demographic profiles.

Child maltreatment rates were based on the mean for the three most recent years available. Three types of crime rates (violent, drug, property) were strongly correlated for adult perpetrators (r = 0.71 to 0.89, p < .001), so the three standardized indicators were combined into a reliable scale, α = .82. Counties' overall health was based on three composited variables (r = 0.74 to 0.85) from the state department of public health's 2020 survey (Colorado Department of Public Health and Environment (CDPHE), 2020a): four weighted indicators of age‐adjusted health outcomes (e.g., premature death), 30 weighted health factors (i.e., health behaviors, clinical care, social factors, physical environment), and adult health rated as fair or poor. Age‐adjusted county‐level suicide rates (CDPHE, 2020b) were based on the three most recent years of available data.

Socioeconomic status (SES) was measured with four correlated variables (r = 0.50 to 0.75, p < .001) that formed a cohesive scale (α = .86): labor force participation, years of education, median income, and poverty rate. The second covariate was percent of the population self‐identified as White, which was correlated with SES (r = .32, p = .02) as well as the composite health indicator (r = −0.58, p < .001) and suicide rate (r = 0.28, p = .02).

RESULTS

Internal reliability

Cronbach's alpha coefficients for each CCI scale (Table 1) are very high, enough so to be suitable for individual decision making (Nunnally & Bernstein, 1994).

Factor analyses

Exploratory factor analysis (EFA)

We conducted an EFA to identify potential scales in the 52‐item version. The overlap of clusters in the concept mapping phase suggested that an orthogonal solution would be inappropriate because it presumes that scales do not overlap conceptually or in content. Therefore, we used an oblique rotation. In analyses of the beta version, items from two of the seven scales loaded on a single factor and were therefore combined. The six factors that emerged (below) largely mirrored the concept mapping results. The EFA indicated that (a) three items had a greater loading on a scale other than the one nominated by a small majority of panelists (i.e., there were similar votes for placing the items in two different categories); (b) because of inconsistent loadings on a given factor, several items with high importance ratings needed to be revised for clarity; and (c) 10 items with factor loadings below 0.45 could be omitted because they were redundant and did not have as much diagnostic value as similar items. Related to clarity, a professional editor helped revise items that some participants said were unclear, resulting in a readability index of grade level 11.3.

Confirmatory factor analysis (CFA)

In the next phase, we conducted a CFA to assess the model identified in the EFA, using the 42‐item validation sample. Multivariate outliers of greater than 3 SDs (n = 7) were omitted from the CFA (n = 1692). The CFA was computed using MPlus version 8.11 (Muthén & Muthén, 2023).

Two different methods can be used to model the structure of a measure that is hypothesized to be comprised of a general factor with several highly correlated dimensions: a bifactor model and a second‐order model (Chen et al., 2006). A bifactor model presumes that the general factor (e.g., communities' civic capacity) has direct effects on each item in the scale, with unique variances specific to each dimension beyond the general factor. In this model, we would expect that each item loads on a general CCI factor and one of the six CCI factors. In comparison, a second‐order factor structure better represents the specific dimensions of civic capacity, which in the present case was tested by modeling the CCI as six oblique dimensions such that a general construct—civic capacity—is shared across dimensions rather than being unique to each scale. Bivariate correlations indicate that there is a general factor in the CCI given that the total score is highly correlated with each of six scales, r = 0.86 (Leadership) to r = 0.93 (Collaboration), and the 42 items as a set have a high Cronbach's alpha reliability (α = 0.98).

Given the bivariate correlations in Table 1 as well as results from the EFA, we tested a six‐factor structure of the CCI in which the scales were correlated with each other; i.e., an oblique factor analysis. A bifactor model was also tested in which all 42 items were loaded on a general civic capacity factor as well as one of the six specific factors. Two alternative models were tested, one with a single latent construct of civic capacity encompassing all 42 items, and one in which the CCI dimensions were independent of each other—an orthogonal, six‐factor model.

We used several goodness‐of‐fit statistics to test model fit: the comparative fit index (CFI), where values equal to or greater than 0.95 indicate good fit, and the root mean square error of approximation, where values equal to or less than 0.06 are satisfactory (Hu & Bentler, 1999). To account for model complexity, we also calculated the Bayesian information criterion (BIC), for which comparatively better fitting models are reflected in lower BIC values (Kang et al., 2009).

The first CFA model, with six oblique factors, constrained the 42 CCI items to load onto their corresponding factors, with the factors permitted to correlate with one another. As a result of model specification, the error variances of items 4–6 (diversity in leadership) were allowed to correlate with those of items 12 and 13 (diversity in collaborating entities). The fit statistics for this six‐factor oblique model were acceptable (Table 2). Next, we examined contrasting models: a six‐factor orthogonal model and a one‐factor model. These two models' fit indices were marginal. Finally, the bifactor model was tested in which each item loaded on a general Civic Capacity factor and one of the six orthogonal factors to account for variance unique to each of the six CCI scales. The bifactor model provides the best fit with the data (Table 2), with the χ ² difference between the bifactor model and the oblique six‐factor model being statistically significant. The BIC values indicate that the bifactor model is preferred, although the oblique six‐factor model also is an adequate representation of the data. In practical terms, either a total CCI score or a set of six scale scores in describing a community's civic capacity is supported.

Table 2.

Summary of model fit indices for Civic Capacity Index confirmatory factor analysis.

Model	df	CFI	RMSEA [90% CI]	BIC	∆χ 2 (df)
1. Bifactor model	612	0.98	0.062 [0.054, 0.070]	5624.74
2. Oblique six factors	650	0.94	0.074 [0.069, 0.079]	5881.09	1 versus 2 486.79 (45)***
3. Single factor	653	0.89	0.18 [0.16, 0.20]	7146.63	2 versus 3 1268.54 (6)***
4. Orthogonal six factors	653	0.87	0.21 [0.17, 0.25]	6256.56	2 versus 4 390.05 (6)***

Note: df = degrees of freedom; CFI = Comparative Fit Index; RMSEA = root‐mean‐square error of approximation; CI = confidence interval; BIC = Bayesian information criterion. N = 1692

^*** p < .001.

CCI‐SF short form

We created a 12‐item short form of the CCI because program evaluations often use a battery of brief, valid outcome measures. We reasoned that the CCI‐SF should (a) draw from the content of all six scales, (b) differentiate between low‐ versus high‐capacity communities, and (c) correlate with the validation scales, particularly social capital and community resilience. Although the effect sizes for community differences on all 42 items were at least moderate (ƞ ² > 0.056), 12 items were medium/large (ƞ ² > 0.10), F(53,1920) > 2.94, p < .0001, suggesting that they had particularly good diagnostic value, that is, a high item discrimination index (Boateng et al., 2018). These 12 items were significantly correlated with community cohesion (median r = 0.47) as well as community resilience (median r = 0.57) and well‐being (median r = 0.47) and accounted for significant variance beyond the validation scales in community resilience and well‐being (Table 3). In fact, the CCI‐SF was as predictive of community resilience and well‐being as the full 42‐item CCI, which is unsurprising given that the CCI‐SF and 42‐item total score are correlated r = 0.97. Finally, the CCI‐SF is a highly reliable scale, α = 0.94. Thus, the short form has good diagnostic properties, is valid, and represents the full scale's item content well but not as comprehensively as the 42‐item scale.

Table 3.

Civic Capacity Index (CCI) and validation scales predict community resilience and well‐being.

Variable	Community resilience			Community well‐being
	B	SE B	β	B	SE B	β
			Six CCI Scale Model
Step 1
Constant	18.62***	0.91		33.80***	0.91
CCI Leadership	0.01	0.03	0.02	0.11***	0.03	0.12***
CCI Inclusion	0.05	0.03	0.04	0.04	0.03	0.05
CCI Institutions	0.14***	0.03	0.14***	0.09***	0.03	0.11***
CCI Civic Culture	0.42***	0.04	0.41***	0.24***	0.04	0.28***
CCI Coalitions	0.05	0.03	0.04	0.00	0.04	0.00
CCI Collaboration	0.17***	0.03	0.17***	0.10**	0.03	0.11**
R adj 2			0.59***			0.39***
Step 2
Constant	−6.86***	1.08		14.85***	1.35
CCI Leadership	−0.04	0.02	−0.04	0.08**	0.03	0.09**
CCI Inclusion	0.17***	0.02	0.17	0.11***	0.03	0.13***
CCI Institutions	−0.16***	0.03	−0.17***	−0.11***	0.03	−0.14***
CCI Civic Culture	0.23***	0.13	0.23***	0.11***	0.03	0.13***
CCI Coalitions	0.05*	0.03	0.05*	0.02	0.03	0.02
CCI Collaboration	−0.02	0.03	0.02	−0.02	0.03	−0.03
Trust in Governance	0.49***	0.03	0.44***	0.32***	3	0.35***
Social Control	0.24***	0.02	0.25***	0.15***	0.02	0.18***
Social Cohesion	0.09***	0.02	0.07***	0.06*	0.03	0.05*
Community Connected	0.11***	0.02	0.10***	0.13***	0.02	0.13***
R adj 2			0.78***			0.52***
			Model Using CCI‐SF
Constant	−8.80***	1.08		14.34***	1.32
CCI‐SF	0.30***	0.02	0.28***	0.21***	0.02	0.24***
Trust in Governance	0.36***	0.02	0.32***	0.23***	0.02	0.24***
Social Control	0.23***	0.01	0.24***	0.13***	0.02	0.16***
Social Cohesion	0.13***	0.02	0.11***	0.09***	0.03	0.09***
Community Connected	0.14***	0.01	0.11***	0.15***	0.02	0.15***
R adj 2			77***			0.51***

Note: CCI‐SF = Civic Capacity Index‐Short Form

^* p < .05

^** p < .01

^*** p < .001

Convergent validity

Overall, the six CCI scales' correlations with the validation measures (Table 1) are moderate (r ≥ 0.27) to large (r ≥ 0.50) in size. As expected, the CCI Inclusion scale is strongly correlated with the Social Justice scale. Also, the CCI Civic Culture scale correlates significantly with the three related measures of social capital—Social Control, Social Cohesion, and Community Connectedness—in most cases more strongly so than the other CCI scales.

Each of the CCI scales, as well as the validation measures, are correlated with Community Resilience and Community Well‐Being (Table 1), with effect sizes that are large. We used stepwise multiple regression to assess how much variance in community resilience and well‐being could be explained by the CCI and validation scales as a set. Multivariate outliers (n = 8) of greater than 3 SDs were 90% winsorized such that extreme values were set to 5% or 95% of the normal distribution. Income and education were entered as control variables but were trimmed from the model because they did not account for unique variance in either community resilience or well‐being with the validity scales entered in Step 2 (Table 3). All but two predictor variables contributed significantly to the R ² _adj = 0.78 of the variance in Community Resilience, F(10,1830) = 650.61, p < .0001, with the exceptions being Leadership and Collaboration (Table 3). With Community Well‐Being as the outcome variable, all but two of the predictors accounted for R ² _adj = 0.52 of the variance F(10,1830) = 201.34, p < .0001; the exceptions were Coalitions and Collaboration. For both outcome variables, Civic Culture and Trust in Governance were particularly robust predictors, with aspects of social capital—particularly Social Control—explaining unique variance in both outcome variables. The variance explained in each outcome variable would be considered very large, given that an R ² _adj of 0.26 is large.

To assess whether the measures of neighborhood social capital and community resilience were conceptually distinct from the processes measured by the CCI, we conducted a factor analysis, with oblique rotation, of all scale scores. The first step was to determine if there was common method bias. To assess this, we conducted Harman's Single‐Factor Test that involves an exploratory factor analysis of all items in the scales listed in Table 1, with the number of factors restricted to one and without rotation. The amount of variance explained in the single factor was 38%, or less than the 50% cutoff that suggests there is significant common method variance. However, the validity of this approach has been questioned (Howard et al., 2024); among the remedies recommended are improved item clarity and anonymous responses, both of which we used.

The factor analysis of all scales showed that the six CCI scales loaded on the first factor and all but one of the validation scales loaded on a second factor, with Trust in Government cross loading on both factors. However, the two factors were correlated r = 0.68, which is consistent with the bivariate correlations in Table 1 as well as the unique contributions of the CCI‐SF and validity measures in predicting Community Resilience and Community Well‐Being (Table 3).

Construct validity

Group differences

To assess measurement invariance, we conducted exploratory analyses that revealed no meaningful differences (i.e., small Cohen's d effect sizes of <0.10) on the CCI scales by age, education, income level, or gender. However, Latine individuals, as compared to self‐identified non‐Hispanic Whites, rated their communities as higher on each CCI scale except Institutions, t(1938) > 3.38, p < .001, with effect sizes that were small (Cohen's d = 0.28) to medium (d = 0.45).

As hypothesized, significant but small group differences emerged on the indicators of involvement in community leadership activities. First, those in paid leadership positions, versus those who were not, rated their communities as higher on each CCI scale, t(1762) > 3.78, p < .001, with effect sizes that were small, Cohen's d = 0.19 to 0.37. Similarly, those in unpaid (informal) leadership positions, versus those who were not, rated their communities as higher on each CCI scale, t(1762) > 2.07, p < .02, with effect sizes that again were small, d = 0.10 to 0.22. Finally, those who were involved in community collaborations, versus those who were not, were significantly higher on Institutions, Civic Culture, and Coalitions, t(1762) > 2.07, p < .02, although the effect sizes were small, d < 0.19.

Relation to community health

Multiple regression analyses were computed to determine if the CCI‐SF is related to indicators of community health. First, all variables were centered, then the CCI‐SF was entered as the sole predictor. As shown in Table 4, the CCI‐SF is associated with each of the indicators of community health that are moderate in size, the exception being that the CCI‐SF is unrelated to crime rates.

Table 4.

Civic Capacity Index and county‐level indicators of community health.

Variable	Child abuse		Crime rate		Poor health		Suicide
	B	SE B	B	SE B	B	SE B	B	SE B
Step 1: CCI‐SF
Constant	−0.19	0.13	0.14	0.13	−0.06	0.13	0.01	0.14
CCI‐SF	−0.42**	0.13	−0.04	0.14	−0.37**	0.13	−0.37**	0.14
R adj 2	0.16***		0.01		0.12**		0.12**
Step 2: CCI‐SF x SES
Constant	−0.01**	0.14	0.21	0.16	0.11	0.11	0.09	0.13
CCI‐SF	−0.17	0.15	−0.08	0.16	0.02	0.12	−0.33*	0.17
SES	−0.47***	0.13	−0.07	0.15	−0.69***	0.11	−0.19	0.16
CCI‐SF x.SES	−0.02	0.12	−0.08	0.14	0.01	0.10	0.17	0.15
R adj 2	0.29***		0.05		0.44***		0.13**
Step 2: CCI‐SF x % minority
Constant	−0.18	0.13	0.13	0.20	−0.07	0.11	0.02	0.13
CCI‐SF	−0.40**	0.14	−0.02	0.14	−0.31**	0.11	−0.40**	0.14
minority (%)	−0.20	0.16	0.31**	0.15	0.50***	0.12	−0.26	0.16
CCI‐SF x minority	−0.01	0.16	0.15	0.16	−0.01	0.13	−0.02	0.17
R adj 2	0.16***		0.10***		0.37***		0.15**

Note: CCI‐SF = Civic Capacity Index‐Short Form; SES = socioeconomic status. All variables were centered.

^* p < .05

^** p < .01

^*** p < .001.

Next, we assessed whether the CCI‐SF is still predictive when the two social address variables of socioeconomic status (SES) and percent of the county population identifying as ethnic minorities were entered into the regression equations. Two predictors related to social capital—social control and community connectedness—originally were included in the regression analyses but were trimmed from the models because they did not account for unique variance in any of the four outcome variables. We also tested whether communities' civic capacity moderates the association between SES (or ethnic profile) and community health and well‐being. This hypothesis is baed on the assumptions that (a) social address variables such as SES (Western et al., 2012) and ethnic profile (Berger & Sarnyai, 2015) are associated with public health outcomes via proximal processes such as community resources and norms (Bronfenbrenner & Morris, 2006) or discrimination and family stress (Merçon et al., 2020); and (b) that civic capacity mitigates the effects of negative distal processes such as low SES and discrimination against ethnic minority groups.

In Step 2 of the regression analyses that focus on SES (Table 4), SES was strongly, inversely related to child maltreatment and poor health and moderately so to suicide rates but not to counties' crime rate. In all regression equations except suicide rates, the CCI‐SF was no longer predictive of the outcome variables. The test of whether civic capacity moderated the effects of SES was not significant for the four outcome variables.

A different pattern of associations emerged with minority status as a predictor as compared to those involving SES (Table 4). First, civic capacity was significantly related to all outcome variables except crime rate, with beta values that were similar to those obtained with CCI‐SF scores as the sole predictor. Second, minority status was predictive of two of the outcome variables (crime and poor health), and there was no evidence of moderation.

Prediction of community resilience

The subset of our sample who completed a 1‐year follow‐up was used to assess the stability of scores and whether CCI scores predict later ratings of community resilience. As shown in Table 5, the stability of CCI scores was relatively high despite the aftermath of the COVID pandemic, ranging from 0.67 to 0.77. Inclusion was the only CCI scale to show a significant change with time, an increase that was a small effect size, Cohen's d = 0.14. Trust in Governance decreased significantly, whereas Community Resilience increased significantly, with small effect sizes of d = 0.11 and d = 0.16, respectively.

Table 5.

Stability and change in communities' civic capacity, trust in governance, resilience, and well‐being.

	Time 2										Mean (SD)
	1.	2.	3.	4.	5.	6.	7.	8.	9.	10.	Time1	Time2	p
Time 1
Civic Capacity Index
1. Leadership	.70	.63	.62	.61	.55	.57	.66	.42	.54	.40	49.03 (18.79)	49.68 (16.54)	.33
2. Inclusion	.65	.72	.62	.63	.54	.56	.66	.45	.44	.41	41.20 (20.05)	43.18 (17.78)	.004
3. Institutions	.62	.57	.74	.66	.58	.63	.72	.53	.63	.51	53.24 (20.82)	50.28 (17.68)	.16
4. Civic Culture	.68	.64	.70	.75	.63	.66	.75	.49	.64	.49	47.89 (19.93)	47.13 (17.08)	.25
5. Coalitions	.60	.56	.60	.65	.67	.67	.67	.42	.67	.44	48.23 (19.08)	48.36 (16.98)	.86
6. Collaboration	.61	.59	.65	.68	.62	.68	.71	.49	.61	.48	47.05 (20.24)	47.69 (17.26)	.39
7. CCI‐SF	.69	.65	.73	.72	.64	.68	.77	.54	.67	.52	47.96 (19.51)	48.14 (16.98)	.39
8. Trust	.54	.47	.64	.60	.52	.58	.64	.53	.66	.50	55.45 (19.37)	53.23 (21.38)	.02
9. Comm. Resilience	.56	.50	.63	.62	.51	.55	.64	.52	.76	.56	59.29 (20.81)	61.49 (18.18)	.001
10. Comm. Well‐Being	.52	.44	.58	.58	.47	.51	.58	.49	.61	.60	61.52 (17.10)	60.40 (16.90)	.13

Note: CCI‐SF = CCI‐Short Form. Time2‐Time1 mean lag = 1.23 years. Highlighted correlations are stability coefficients. All correlations: p < .001. N = 428.

To examine the effects of civic capacity on later community resilience, we conducted a 3‐level multilevel model (MLM) using SPSS Statistics Version 29. The MLM permits estimation of the variance associated with repeated measures over time (level 1) within respondents (level 2) and within counties (level 3) based on the specification of fixed‐effect variables entered into the model (Raudenbush & Bryk, 2002). We estimated the association between (a) baseline civic capacity on the CCI‐SF, change in civic capacity over 1 year, and baseline community resilience, and (b) their two‐way and three‐way interaction effects, on (c) Time 2 community resilience. Fixed effects’ significance was determined by Wald tests' p values.

With all main and interaction effects entered into the model, initial level 3 (county) differences were no longer significant predictors of Time 2 community resilience, Wald Z = 1.09, nor were any of the two‐way and three‐way interactions. However, all three of the baseline predictors were significant: community resilience (stability across time), t(414) = 10.57, p < .001; baseline civic capacity on the CCI‐SF, t(414) = 7.90, p < .001; and change on the CCI‐SF over 1 year, t(414) = 8.56, p < .001. A multiple regression showed that 66% of the variance in Time 2 community resilience was explained by the three predictors, F(3,421) = 273.33, p < .001, with 57% attributable to stability in community resilience and 9% to civic capacity.

DISCUSSION

Our work on community‐driven change and civic capacity generated information about the characteristics and practices of communities that successfully work together to address challenges like the COVID‐19 pandemic. These qualities help illuminate how different communities navigate challenges and leverage their strengths to advance resilience and well‐being. Communities with a high civic capacity for solving tough problems through community‐driven change share several attributes. First, many people from all parts of the community exercise leadership for the common good. Second, people in these communities are willing to address the ethical dimensions of civic challenges, such as fairness and social justice, in inclusive, constructive ways. Third, authority figures, as well as public, private, and nonprofit institutions, are committed to working together. Fourth, these communities coalesce quickly around civic challenges. This sense of collective agency encourages civic engagement through the creation of coalitions that can plan and act together in pursuit of the commonweal.

The CCI measures the presence or absence of these attributes. With some qualifications discussed in the limitations section, the data support the psychometrics of the CCI. In terms of reliability, alpha coefficients for each scale justify their use in making decisions about a community's policies and practices, and their 1‐year stability coefficients are high enough to make them useful in predicting future civic capacity and community resilience. Related to municipal planning, the CCI provides information about dynamic processes that are correlated with various public health indicators, albeit not as predictive as, but more malleable than, socio‐economic status. Thus, as a diagnostic tool, the CCI can help civic actors make better decisions based on a credible assessment of a community's civic capacity to respond to challenges.

We documented convergent validity with data showing the CCI scales to be correlated with processes that define civic capacity: trust in government (median r = .64), social justice (median r = .60), and community resilience (median r = .68). We had theoretical reasons to believe that CCI scores might be correlated with measures of social cohesion (i.e., collective efficacy and community connectedness; median r = .43) because of their association with community resilience, although these measures do not provide information about the complex decision‐making processes involved in building civic capacity. A factor analysis of all scale scores revealed that the CCI scales and the validation measures loaded on separate factors, with Trust in Government straddling both, yet the two factors overlapped substantially. This pattern of associations among variables explains why the CCI‐SF and validity measures make unique contributions to predicting Community Resilience and Community Well‐Being. Convergent validity of the CCI and, to some extent, discriminant validity, were supported by these correlations (for criteria, see Ziegler, Booth, & Bensch, 2013).

One purpose of this study was to test whether the factor structure of the CCI aligns with the concept mapping results from the development of the survey (Chrislip et al., 2023). The confirmatory factor analysis and the concept mapping process we conducted support the conception of civic capacity as a coherent, singular construct with different facets. Progress emerges from the interplay of these facets in particular situations or challenges. Each facet of the CCI represents a necessary but not sufficient aspect of civic capacity:

• The Leadership scale taps into many people exercising leadership in synergistic ways in support of collective action (Chrislip & O'Malley, 2013; Lasker & Weiss, 2003). This contrasts with the customary decision‐making deference to political authorities and influential powerbrokers coupled with limited roles for community members and tightly circumscribed processes for public participation (Briggs, 2008).

• Inclusion of diverse perspectives, particularly those who typically are excluded from power, results in community members being more able to confront issues of discrimination and social justice while addressing presenting issues in constructive ways (Hicks, 2020; Lasker & Guidry, 2009). This perspective on community‐driven change entails shared power between community members and decision makers, and a greater focus on equitable outcomes for all.

• To make community‐driven change work, Institutions are committed to working with other sectors and community members as equal partners. They provide knowledge and resources to support collective action and invest in building the civic capacity of their communities. Institutions operating in isolation from one another fragment efforts to address civic challenges and preclude finding common ground (Putnam & Garrett, 2020).

• Civic Culture describes the social networks, norms, and attitudes—the willingness to work together—in communities that support community‐driven change. A multitude of social networks facilitate partnerships and civic engagement. This may explain why this scale was a particularly robust predictor of community resilience and well‐being.

• In community‐driven change, the formation of Coalitions to address civic challenges depends on the ability to recognize collective challenges and the willingness to address them in a collaborative way. Through broad inclusion, the capacity to work across factions, and a steady focus on the greater good, community coalitions help create the collective credibility to exert influence over key institutions, add legitimacy to public decisions, and share accountability for outcomes (Briggs, 2008; Lee & Ospina, 2022; Stone et al., 2001).

• Collaboration relies on inclusion and fair, trustworthy processes informed by a synthesis of expertise and lived experience to address civic challenges (Briggs, 2008; Hicks, 2020). It reflects a choice to disrupt existing ways of doing business; e.g., deference to political authorities with a limited role for community members—and to create a context where problems can be solved collectively (Briggs; Stone et al., 2001).

Uses of the CCI

The CCI helps civic actors leverage current civic capacity, identify gaps, build future resilience, and evaluate the impact of community‐driven change. The CCI helps agencies and organizations understand the ability of community members to engage in community‐driven change. It may guide civic leadership programs for what needs to be done to build civic capacity and collective efficacy. Additionally, the CCI assessment provides a starting place for building resilience and civic capacity. It can also serve as a catalyst for structured community dialog. By making collaborative processes transparent, the assessment helps community members reflect on their experience in local collaboration, identify gaps in inclusion or coordinated efforts, and prioritize where to invest for future collaboration and civic leadership building.

We have used the CCI as an instructional tool for building civic capacity. In one urban neighborhood with large SES gaps, substance abuse problems, and a history of low civic engagement, we used the CCI data to scaffold a culture of collaboration. The CCI provided initial data about civic strengths and challenges in the community and helped identify potential areas for investment. Three community conversations were co‐led by a university‐based facilitator, a City Council member, and a neighborhood organizer. Participants were drawn from a nonprofit collaborative that the City Council member hosts and expanded by the neighborhood group. As with the data we collected from the counties across the state, the neighborhood's scores for inclusion were low. Some residents struggled with this but the conversation provided an opportunity to build a commitment to improve how the community works collaboratively to address shared challenges. This was especially important for people of color and lower‐income residents who had felt excluded.

Future directions

Measurement typically is a means to an end, not the end itself. Even so, measurement is essential to learning and adaptation. “We have to know what is going on around us, how our actions impact others, how the environment is changing, how we're changing” if we are to adapt or grow (Wheatley & Kellner‐Rogers, 1999, p. 2). This statement encapsulates the utilitarian value of the CCI for advancing future research and applications that enhance community resilience and well‐being.

For example, one urgent research initiative involves using the CCI to examine how multiple communities responded to the COVID‐19 crisis. Cross‐national research shows that higher confidence in local institutions and civic engagement are related to lower COVID mortality rates, whereas inequalities, such as those measured by the CCI's Inclusion scale, correlate with higher mortality rates (Elgar et al., 2020). In addition, civic engagement and government transparency build public trust (Schmidthuber et al., 2021), and this reservoir of trust likely contributes to whether public health guidelines will be followed. Given that the CCI accounts for substantial variance in community resilience, its data may be able to help determine if these observations are coincidental or reflect a deeper, causal relationship.

Although there is solid evidence that the CCI provides accurate information as far as it goes, other elements essential to a well‐functioning, resilient community remain. The literature and our own data indicate that, in addition to the CCI scales, the sense of collective efficacy and belonging provide important information that helps to predict community resilience. Other adaptive capacities include available resources, particularly information and communication infrastructures, and the resourcefulness to identify problems and establish priorities (Boston et al., 2024). If we are to fully account for community resilience, information about access to human and material resources is needed to complement the dynamic processes of civic capacity.

There is much to be learned about effective ways to build civic capacity. For instance, what level of civic capacity does it take to actively sustain the process of community‐driven change over long periods of time? The CCI is decidedly pluralistic, given that 80% of the items refer to variants of “working together,” “networks,” and collaboration. These pluralistic dynamics are consistent with the bonding and bridging aspects of social capital as foundations of community resilience and well‐being (McCrea et al., 2016), as is multisector collaboration (Kulig et al., 2013). But what is the optimal mix of stakeholders to create sufficient credibility and influence to act? How essential and how broad is the diversity of perspectives and resources needed to achieve success? How engaged and resourceful do stakeholders need to be to make progress?

Network analyses (Varda & Sprong, 2020) can help answer some of these questions. Such analyses respond to Pares and colleagues' (2017) call to examine the relations among residents and also account for the multifaceted relations among public, private, and nonprofit entities that transcend social and political organizations. For example, Perkins et al. (2002) recommended that communities create network‐bridging opportunities that require a foundation of broad‐based community organizations capable of engaging in sometimes contentious interchanges with other institutions. This helps protect against an overemphasis on social cohesion that can undermine engagement in constructive conversations about conflicting values, perspectives, and plans.

Limitations and pushbacks

Several study limitations should be noted as far as the generalizability and validity of our results. First, our sample is not representative of local populations. This was by design, given that a criterion for enrollment in the study was that participants be at least moderately familiar with local governance so they could provide more accurate information about their community's civic capacity and features of community resilience. Relatedly, although we developed exhaustive contact lists for each of the state's counties, less than 5% of those we contacted completed surveys. Replication studies are needed to test the generalizability of our findings.

Common method variance, where bias may arise because both the predictor (civic capacity) and outcome (community resilience) are measured using the same self‐report method and source, can inflate correlations between measures of different constructs (Podsakoff et al., 2024). This issue was at least partially remediated by (a) having a substantial subsample of participants complete the same surveys a year later, with most stability coefficients being robust and the CCI predictive of later community resilence; (b) testing the association between counties' average CCI score and various measures of public health pulled from online sources, and (c) a Common Method Bias score of 38% that was below the 50% that indicates significant common variance is present. Podsakoff et al. also noted that the validity of a questionnaire is likely to be compromised when respondents have limited experience with the survey's topics. As noted earlier, very few of our respondents lacked leadership experience or knowledge. Among the other methods Podsakoff et al. recommended to minimize common method variance that we employed are protecting respondents' anonymity; varying the response options, including item reversals; and minimizing task difficulty with clear and concise language.

We should emphasize that a foundation of civic capacity is the inclusion of diverse voices at the decision‐making table. Some community members—a quite small but vocal minority of our original sample (<1%)—took umbrage at this emphasis, asserting (sometimes with inflammatory language) that it has a liberal bias against rural areas, while others were gratified that these issues were being raised. We responded by adhering to the advice of Bogenschneider et al. (2019, p. 144):

The peril is that dialog will work only if researchers can differentiate themselves from being viewed as just another interest group. To epitomize the contributions that research can make to policymaking, scholars must stay true to the science and avoid actions and attitudes that could provoke prevailing perceptions of liberal bias. The power of nonpartisanship.

To address this credibility issue, initiatives using the CCI should be convened by diverse stakeholders who understand the purpose and uses of the instrument, thus ensuring a representative sample that produces credible results. We also need to work closely with convenors to help them create a communications plan to support the implementation of the CCI and to define the appropriate audience for the survey and subsequent community conversations.

CONCLUSIONS

We began this article by asking what form of governance helps communities solve challenging problems while building their capacity to address future challenges—their resilience and well‐being. The CCI data from our sample unequivocally show that most communities are struggling to address civic challenges like the pandemic. Few communities are “maintaining” much less “thriving,” based on a widely used measure of well‐being. In today's contentious context, responding to civic challenges has become even more intractable. But the CCI data from our surveys—along with insights from our expert panel and our experience from collaborative work with communities—reveal two convincing answers to what governance processes are especially potent: the necessity of including diverse voices and perspectives to address collective challenges, and the presence of a vibrant civic culture that encourages and supports collaboration. In contrast, communities that lack these features may find themselves caught in a vicious cycle of decline, as Putnam (1993) found in his research on social capital in modern Italy.

Through our work on understanding the critical role of civic capacity, we have learned what it takes to respond to civic challenges in more productive ways. We know that civic challenges are both adaptive and collective. They require collaborative problem ‐solving and mutual learning to create effective responses. We know that communities that have the capacity to work together in inclusive and constructive ways make more progress on these kinds of challenges and thus are more resilient in facing future challenges.

This prompts the question of how to systematically build and mobilize civic capacity to address these challenges. This work includes expanding the conception of civic engagement from a narrow focus on civic participation, education, and advocacy to collaborative problem solving driven by the community itself. It means developing collective leadership capacity where many people in the community exercise leadership in service of the broader good. It points to building a civic culture that connects institutions, sectors, and people with the relationships and skills to work together. And it means creating intermediary organizations with the resources and expertise to support collaborative civic engagement.

Our experience, coupled with years of insights from scholars and practitioners, helped us understand that those communities that make more progress marshal their collective resources to solve their own problems. As equal partners, governments, institutions, and community members work together to make public decisions as inclusive and accessible as possible. Civic work, in this sense, is collective work that is done by coalitions that reflect the diversity of the community and the nature of the challenges. It relies on the leadership activities of many people to elevate and change the way issues are addressed and to mobilize the energy to act together. With enough civic capacity, communities can solve tough problems while building a civic culture that enhances resilience and well‐being and deepens democracy.

MacPhee, D. , Chrislip, D. , & Schmitt, P. (2026). Becoming resilient: Community‐driven change and the Civic Capacity Index. American Journal of Community Psychology, 78, 277–294. 10.1002/ajcp.70100

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.