Academy of Management Journal

Advancing Reform: Embedded Activism to Develop Climate

Solutions

Journal: Academy of Management Journal

Manuscript ID AMJ-2019-0769.R3

Manuscript Type: Revision

Qualitative orientation (General) < Qualitative Orientation < Research

Methods, Adaptation/Change < Organization and Management Theory <
Topic Areas, Role of change agents < Organizational Development and
Keywords:
Change < Topic Areas, Environmental Sustainability < Topic Areas, Set-
analytic methods (QCA, etc.) < Analysis < Research Methods, Social
movement theory < Theoretical Perspectives

Activists increasingly seek to influence organizations that also espouse

support for social movement goals, encouraging the use of collaborative
tactics. While there has been growing research on insider activists who
import social movement resources, little is known about how internal
activism might operate through a coordinated and collaborative approach
with external social movement organizations, which we refer to as
embedded activism. Likewise, collaborative activists encounter
organizations with a wide range of prior reform experience, and the
resulting “opportunity structure” for collaboration is not well understood.
Abstract: We investigate how a network of embedded activists can operate to
advance reform efforts across diverse organizations. Our analysis
combines surveys, interviews, and archival records from the
Environmental Defense Fund’s Climate Corps program, which embeds
graduate students in partner organizations to advance climate change
reforms. Embedded activists accomplish this work by matching external
resources with the organizational context in order to generate a fertile
mixture of support and ambiguity and create new solutions. External
resources are especially important for organizations that are at the
extremes of prior issue development.
Page 1 of 63 Academy of Management Journal

1
2
3
4 Advancing Reform:
5 Embedded Activism to Develop Climate Solutions
6
7
8 Todd Schifeling
9 Temple University
10 [email protected]
11
12
13 Sara Soderstrom
14 University of Michigan
15
16 [email protected]
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44 Acknowledgements
45 We thank the AMJ reviewers for their constructive and thought-provoking comments, and especially
46 Senior Editor Jennifer Howard-Grenville for her guidance through the review process. We are also
47 grateful for invaluable feedback from participants of the GRONEN Reading Group, seminars at the Erb
48
Institute, Center for Positive Organizations, Ross Strategy, as well as AOM, ARCS, and GRONEN
49
50 conferences. We thank Panikos Georgallis, Ravi Kudesia, Elisa Alt, Sue Ashford, Maddy Ong, Brent
51 McKnight, and Katrin Heucher for their feedback and guidance. This work was enabled by
52 Environmental Defense Fund Climate Corps leaders, fellows, and hosts, and especially Ellen Shenette.
53 Rob Luzynski, Christina Donovan, and Chantelle Barretto provided excellent research assistance. This
54
55
research was financially supported by the Graham Institute and the Erb Institute at the University of
56 Michigan and EDF.
57
58
59
60
 Academy of Management Journal Page 2 of 63

1
2
3 ADVANCING REFORM:
4
5
EMBEDDED ACTIVISM TO DEVELOP CLIMATE SOLUTIONS
6
7 ABSTRACT
8 Activists increasingly seek to influence organizations that also espouse support for social
9
movement goals, encouraging the use of collaborative tactics. While there has been growing
10
11 research on insider activists who import social movement resources, little is known about how
12 internal activism might operate through a coordinated and collaborative approach with external
13 social movement organizations, which we refer to as embedded activism. Likewise, collaborative
14 activists encounter organizations with a wide range of prior reform experience, and the resulting
15 “opportunity structure” for collaboration is not well understood. We investigate how a network
16 of embedded activists can operate to advance reform efforts across diverse organizations. Our
17
18
analysis combines surveys, interviews, and archival records from the Environmental Defense
19 Fund’s Climate Corps program, which embeds graduate students in partner organizations to
20 advance climate change reforms. Embedded activists accomplish this work by matching external
21 resources with the organizational context in order to generate a fertile mixture of support and
22 ambiguity and create new solutions. External resources are especially important for organizations
23 that are at the extremes of prior issue development.
24
25
26 Social movement scholarship has evolved from studying protests against state actors to
27
28 the use of contentious tactics to challenge recalcitrant private sector organizations (Soule, 2009;
29
30 Baron & Diermeier, 2007). At the same time, there is a growing trend for corporations to
31
32
33 espouse support for goals such as sustainability, social responsibility, and equal rights
34
35 (Dauvergne and Lister, 2013; Dobbin, 2009; McDonnell, King, & Soule, 2015; McDonnell,
36
37 2016). This convergence between the stated goals of progressive social movements and
38
39
40
corporations creates new opportunities for collaboration. Much of this work has taken the form
41
42 of cross sector initiatives that focus on externally-oriented changes like certifications and
43
44 statements of principles (de Bakker, Rasche, & Ponte, 2019; Gray & Purdy, 2018) that involve
45
46
the effort of social movement organizations (SMOs), “organizations that identify their goals with
47
48
49 the preferences of a social movement” (McCarthy & Zald, 1977: 1218). SMOs also have many
50
51 collaborative initiatives to drive internal reforms, but not enough is known about how these
52
53 initiatives operate, as “the collaborative repertoire of modern SMOs continues to be grossly
54
55
56
57
58
59
60
Page 3 of 63 Academy of Management Journal

1
2
3 undertheorized within the study of social movements in markets” (McDonnell, Odziemkowska,
4
5
6 & Pontikes, 2021: 3).
7
8 Research on internal change processes such as internal activism, intrapreneurship, and
9
10 issue selling illuminates how employees bring their social movement affiliations with them into
11
12
13
organizations (Georgallis, 2017; Briscoe & Gupta, 2016; Weber & Waeger, 2017). These
14
15 external identities often inspire the concerns of internal activists who operate as “tempered
16
17 radicals” (Meyerson & Scully, 1995; Bansal, 2003), and they also gain legitimacy by importing
18
19
outside frames and status into their organizations (Gehman, Treviño, & Garud, 2013; Heinze &
20
21
22 Weber, 2016; Kellogg, 2011). More recently, scholars find that social movements infuse internal
23
24 activists with critical resources needed to contest organizations’ practices, including exemplars,
25
26 templates for action, implementation practices, and emotional support (Buchter, 2020; DeJordy,
27
28
29 Scully, Ventresca, & Creed, 2020). This work challenges the common depiction of internal
30
31 activists as operating independently from social movement organizations (SMOs), motivating
32
33 our research into how activists linked to collaborative SMOs push internal reforms across diverse
34
35
36
partner organizations.
37
38 Whereas prior research focuses on internal or external activism (Briscoe & Gupta, 2016),
39
40 collaborative activism that spans these boundaries operates through a form of activism that we
41
42
theorize as embedded. Spanning across SMO and partner organizations enables embedded
43
44
45 activists to develop trust, information sharing, and problem solving (Buchter, 2020; DeJordy et
46
47 al., 2020; Uzzi, 1996). Embedded activists are well positioned to match external resources with
48
49 internal change opportunities at partner organizations. We focus on an important way that
50
51
52 embedded activists can advance reforms by introducing new solutions, a component of issue
53
54 selling (Dutton, 1988). The creation of solutions is important because reform efforts are more
55
56
57
58
59
60
 Academy of Management Journal Page 4 of 63

1
2
3 successful when there are actionable choices available (Alt & Craig, 2016; Dutton et al., 2001),
4
5
6 and seeding an organization with more solutions will tend to increase actions (Cohen, March, &
7
8 Olsen, 1972; Bansal, 2003). Solutions also represent strategic resources that activists can use to
9
10 push for greater action on an issue (Buchter, 2020).
11
12
13
In addition, organizations vary considerably in their prior engagement with activist issues
14
15 and in their openness to activist influence (Briscoe, Chin, & Hambrick, 2014; Crilly et al., 2012;
16
17 Darnall, Henriques, & Sadorsky, 2010; McDonnell, King, & Soule, 2015; Sandhu & Kulik,
18
19
2018). Drawing on proactivity and related research, we theorize how these factors shape a core
20
21
22 tension for embedded activists between organizational support, which provides resources to act,
23
24 and issue ambiguity, which provides opportunities to act. Activists are constrained by the
25
26 organization’s prior issue development (Howard-Grenville, 2007; Andersson & Bateman, 2000;
27
28
29 Bansal, 2003; Wickert & de Bakker, 2018), and they also struggle to amass resources needed to
30
31 formulate effective change proposals, which are scarce for issues that are still emerging at an
32
33 organization (Heinze & Weber, 2016; Howard-Grenville, 2007). Consequently, organizational
34
35
36
support, in terms of positive issue understandings, should act like a fertilizer that boosts the
37
38 growth of change initiatives (Morrison, 2011). On the other hand, support can also reduce
39
40 ambiguity, which deprives activists of the latitude to develop new issue understandings
41
42
(Sonenshein, 2016; Gioia, Nag, & Corley, 2012; Sillince, Jarzabkowski, & Shaw, 2012). As an
43
44
45 organization engages positively with an issue, it can shift from a condition of fertile ambiguity
46
47 (e.g., what should we do about this issue?) to one of constraining routine (e.g., this is how we do
48
49 this work) (Sandhu & Kulik, 2018; Risi & Wickert, 2017). The more organizational support for
50
51
52 an issue, the more elaborate the organizational frameworks around the issue, and the less space
53
54 there may be for activists to develop the issue. Organizational scale also affects this balance as
55
56
57
58
59
60
Page 5 of 63 Academy of Management Journal

1
2
3 larger organizations tend to have more formalized policies for dealing with an issue, but also a
4
5
6 greater range of issue understandings across the organization, while smaller organizations may
7
8 have less resources and expertise on newer issues, but also stronger consensus and vision (Crilly,
9
10 Zollo, & Hansen, 2012; Sandhu & Kulik, 2018; Weber & Waeger, 2017).
11
12
13
Because of this, we would expect embedded activism to operate differently across
14
15 varying “opportunity structures,” which refer to aspects of the organization’s structure and
16
17 culture that enable activist influence (cf., Kellogg, 2011; King, 2008; Schurman, 2004). As the
18
19
prevailing theory of opportunity structures is based on contentious tactics, we develop a novel
20
21
22 framework for collaborative opportunities based on the tension between support and ambiguity.
23
24 We also contend that the external environment is an under-appreciated but crucial element for
25
26 navigating this tension. In developing these arguments, we build on the idea that there is a
27
28
29 “complex interplay between external environment and internal organization” in order to explore
30
31 how embedded activists draw on movement resources to propel changes in particular
32
33 organizational contexts (Crilly et al., 2012: 1431; cf., Weber & Waeger, 2017). Advancing the
34
35
36
growing evidence on the rich interchange between social movements and organizations, we
37
38 investigate how embedded activism can encourage reforms at organizations that have well-
39
40 established programs in an issue area as well as at organizations that are new to the issue.
41
42
We develop this research through studying a prominent program by a collaborative SMO
43
44
45 to promote reforms within organizations, the Environmental Defense Fund’s (EDF) Climate
46
47 Corps. Through this program, EDF has constructed a network to advance climate change reforms
48
49 across business, nonprofit, government, and education sectors, totaling over 764 projects at 394
50
51
52 organizations from 2008-2016 (Delaney, 2017). For each project, EDF sends a graduate student
53
54 fellow to work on energy projects at the host organization and provides extensive support,
55
56
57
58
59
60
 Academy of Management Journal Page 6 of 63

1
2
3 including technical guidance, tactical insights, and motivation. We conceptualize the fellows as
4
5
6 embedded activists who span the boundaries between EDF and its partner organizations.
7
8 Although all the host organizations have enlisted in the program, they also bring considerable
9
10 variance in prior issue development. We take advantage of the diversity of organizations
11
12
13
involved to study how embedded activists integrate external resources from EDF to create new
14
15 solutions for their host organizations, revealing how external practices and internal
16
17 organizational contexts combine to generate outcomes (cf., Crilly et al., 2012; Kellogg, 2009;
18
19
Weber & Waeger, 2017).
20
21
22 To analyze this combination of factors, we develop a fuzzy set qualitative comparative
23
24 analysis (fsQCA) of the creation of new solutions, and we integrate qualitative data to better
25
26 understand the resulting configurations. fsQCA is well-suited to studying how combinations of
27
28
29 conditions (e.g., organizational contexts and external resources) produce an outcome (Fiss, 2011;
30
31 Greckhamer et al., 2018; Ragin, 2008). It also enables us to study equifinality and causal
32
33 asymmetry, where different configurations explain the outcome and its inverse. The results
34
35
36
illuminate how Climate Corps fellows navigate their roles as embedded activists across diverse
37
38 organizational contexts and leverage external resources in different ways to create new solutions.
39
40 The findings contribute to multiple literatures. We develop research on social movements
41
42
and organizations by investigating how embedded activism, a coordinated and collaborative form
43
44
45 of activism, can work to promote internal reforms at diverse organizations. This direction
46
47 transcends existing frameworks that divide activists between insiders and outsiders (Briscoe &
48
49 Gupta, 2016), and integrates research on issue selling and proactivity to better understand how
50
51
52 external social movements use collaboration to influence internal change dynamics. We also
53
54 complement prior research on contentious tactics by theorizing the opportunity structure for
55
56
57
58
59
60
Page 7 of 63 Academy of Management Journal

1
2
3 collaborative activism. In doing so, we highlight a key tension in organizational support for
4
5
6 internal change initiatives, showing how support both enables and constrains these initiatives.
7
8 Further, we uncover the contributions of external ties to managing this tension, which have been
9
10 little considered in relevant research like issue selling (Lauche, 2019). In illuminating a common
11
12
13
problem of change agents – balancing resources to act (from support) with opportunities to act
14
15 (from ambiguity) – we contribute to a growing debate about the role of ambiguity in proactivity
16
17 research more broadly (Grant & Ashford, 2008; Griffin & Grote, 2020; Lim, Tai, Bamberger, &
18
19
Morrison, 2020). Our analyses also extend recent efforts to better situate proactivity dynamics in
20
21
22 organizational contexts (Bindl & Parker, 2017; Hobfoll et al., 2018; Li & Tangirala, 2020). In
23
24 addition, we provide insights into the dynamics that can help organizations make progress on the
25
26 urgent issue of climate change (Ansari, Gray, & Wijen, 2011; Nyberg & Wright, 2020).
27
28
29 EMBEDDED ACTIVISM: SOCIAL MOVEMENT ENGAGEMENT TO ADVANCE
30 INTERNAL REFORM EFFORTS
31
32 Organizational scholars study how social movements influence organizations through
33
34 both external challengers and internal organizational members (Georgallis, 2017; Weber &
35
36
37 Waeger, 2017). In reviewing this literature, Briscoe and Gupta (2016) argue that because of their
38
39 different positions, outsider activists are able to more aggressively challenge targets, while
40
41 insider activists have greater knowledge of the organizational polity and how to drive change,
42
43
44
but are constrained by their dependence on the organization. Reflecting these conditions, insider
45
46 activists “mostly focus on their persuasion and educational efforts,” which include framing,
47
48 internal lobbying, issue selling, and building internal coalitions (ibid: 691; cf., DeCelles,
49
50
Sonenshein, & King, 2019; Soderstrom & Weber, 2019).
51
52
53 While studies of internal activism typically focus on how these change processes play out
54
55 within organizations, keeping external movements in the background, there is also growing
56
57
58
59
60
 Academy of Management Journal Page 8 of 63

1
2
3 evidence of interaction and coordination with external SMOs. External activists can passively
4
5
6 influence this work by creating frames and identities that internal activists later import (Bansal,
7
8 2003; Heinze & Weber, 2016; Kellogg, 2011). Recent research on human rights activism
9
10 documents the active effort by external activists to motivate internal activists and infuse them
11
12
13
with strategic resources to advance reforms (Buchter, 2020; DeJordy et al., 2020). DeJordy and
14
15 coauthors summarize the rich back-and-forth interactions they observe: “Employees and social
16
17 movement organizers may operate at once both inside and outside organizations. As insiders,
18
19
employees experiment with tactics that they may borrow from, or may pass on to, activists in
20
21
22 other organizations...activism shapes the change effort between and through organizations, not
23
24 just in and around them” (2020: 5). Likewise, Buchter (2020) shows how external SMOs
25
26 provided specific implementation practices for LGBTQ reforms that were critical to internal
27
28
29 activists’ tactics: internal activists used negative reactions to these implementation practices
30
31 (either refusals to implement or bigoted responses) to challenge organizational claims of
32
33 supporting equality. These findings point to the need for further research on how SMOs
34
35
36
influence internal activist processes. In particular, we seek to understand how embedded
37
38 activism, which entails collaborative SMOs crossing movement-partner boundaries to influence
39
40 internal reforms, works across diverse organizations.
41
42
Understanding such collaborative activism requires rethinking “opportunity structures,” a
43
44
45 central concept in the literature on social movements and organizations. Previous research on
46
47 corporate opportunity structures for activism presumes contentious tactics, and finds that
48
49 activism will be more successful for targets that are more subject to shaming because of their
50
51
52 prominence and reliance on consumer-facing brands (Bartley & Child, 2014; King, 2008;
53
54 McDonnell & King, 2013). However, different factors are likely to matter with collaborative
55
56
57
58
59
60
Page 9 of 63 Academy of Management Journal

1
2
3 tactics because the mechanisms of influence differ: whereas contentious activists threaten losses
4
5
6 to targets, such as reputational damage and operational disruption, collaborative activists offer
7
8 rewards to their partners, such as reputational enhancement and operational gains (den Hond &
9
10 de Bakker, 2007). For embedded activism and other collaborative approaches, we argue that the
11
12
13
opportunity structure concerns the partner organization’s prior history of engagement with an
14
15 issue and its accumulation of issue-specific resources, which shape the terrain for activists to use
16
17 persuasion to expand reforms.
18
19
We integrate research on organizational proactivity and issue selling to theorize how
20
21
22 embedded activism operates across varying organizational contexts. The proactivity literature
23
24 studies processes where organizational members are “taking control to make things happen
25
26 rather than watching things happen. It involves aspiring and striving to bring about change in the
27
28
29 environment and/or oneself to achieve a different future” (Parker, Bindl, & Strauss 2010: 828).
30
31 These processes include issue selling, which refers to “individuals’ behaviors that are directed
32
33 toward affecting others’ attention to and understanding of issues” (Dutton & Ashford, 1993:
34
35
36
398). Issue selling matches a collaborative approach to influence organizations through
37
38 persuasion rather than disruption (Briscoe & Gupta, 2016). Persuasive tactics like issue selling
39
40 are meaning-making practices that are deeply contextualized within organizations (Alt & Craig,
41
42
2016; Howard-Grenville, 2007; Sonenshein, 2016; Wickert & de Bakker, 2018). Existing
43
44
45 organizational schemas determine how actions are understood and evaluated: attempts at issue
46
47 selling must “enact the schemas of recipients so they attend to and act on new issues” (Howard-
48
49 Grenville, 2007: 572). This context structures the opportunities for collaborative approaches to
50
51
52 activism, particularly in terms of the partner organization’s balance of support and ambiguity.
53
54
55
56
57
58
59
60
 Academy of Management Journal Page 10 of 63

1
2
3 A key challenge for issue selling work is the development of positive issue
4
5
6 understandings within an organization. Issue selling requires expertise in an issue to make an
7
8 effective case for the issue meriting attention and action (Dutton et al., 2001; Howard-Grenville,
9
10 2007; Sonenshein, DeCelles, & Dutton, 2014; Wickert & de Bakker, 2018). It is not enough to
11
12
13
be passionate about the topic; fitting the issue with the organizational context involves mastering
14
15 the details about how desired reforms relate to the organization’s practices and priorities (Bansal,
16
17 2003). The development of solutions is especially important: “many social issue selling
18
19
initiatives fail because of the lack of a viable solution (Bansal, 2003; Howard-Grenville, 2007),
20
21
22 showing the importance of theorizing on how solutions may be crafted to effectively promote
23
24 selling success” (Alt & Craig, 2016: 795). However, the bottom-up nature of issue selling
25
26 implies that this knowledge is often not available within the organization and must be somehow
27
28
29 created or sourced. Howard-Grenville (2007: 574) summarizes the challenge that can exist even
30
31 in an organization with top-level issue endorsement: “The dilemma for issue sellers is how to
32
33 advance [their] issues in an organizational context in which the dominant meanings and norms
34
35
36
may blind others to the issues, their consequences, and the value in addressing them.” These
37
38 disconnects may lead to perceptions of an unsupportive organizational context, which
39
40 discourages issue selling (Ashford, Rothbard, Piderit, & Dutton, 1998; Morrison, 2011).
41
42
However, organizational support, in the form of positive issue understandings, can also
43
44
45 hinder issue selling because it reduces ambiguity about how to develop an issue, which restricts
46
47 creativity. We adopt Weick’s definition of ambiguity as “‘an on-going stream that supports
48
49 several different meanings at the same time,’ when ‘multiple...explanations are plausible’”
50
51
52 (1995: 91, 134; quoted in Lingo & O’Mahony, 2010: 52). When organizations develop
53
54 commitments, frameworks, and other positive issue understandings, they also define the issue
55
56
57
58
59
60
Page 11 of 63 Academy of Management Journal

1
2
3 and the appropriate schemas for dealing with it, reducing ambiguity. In contrast, when an
4
5
6 organization has little prior experience with an issue, there are many different plausible
7
8 directions for issue development (Deken et al., 2018; Kannan-Narasimhan & Lawrence, 2018).
9
10 In these contexts, embedded activists are less constrained by organizational expectations,
11
12
13
consistent with Grant and Ashford’s analysis of ambiguity as “a weak situation rather than a
14
15 strong situation. In weak situations, individuals feel less pressure from the environment to think,
16
17 feel, and act in a prescribed manner” (2008: 15). This lack of constraint can empower activists,
18
19
as Sonenshein theorizes with the related concept of equivocality: “under conditions of high
20
21
22 equivocality, social change agents have a higher level of interpretive discretion to construct the
23
24 meaning of an issue in strategic and/or variable ways that best support their pursuit of change”
25
26 (2016: 356; cf., Gioia et al., 2012; Sonenshein, 2014). Likewise, an emerging trend in the
27
28
29 broader literature on proactivity suggests that as a source of uncertainty, ambiguity is valuable
30
31 because it stimulates curiosity, exploration, and innovation (Griffin & Grote, 2020).
32
33 While issue selling initiatives struggle in the absence of prior issue development, such
34
35
36
development can also have the unintended consequences of limiting their creativity. For
37
38 example, pushing reforms to reduce a company’s scope three carbon emissions, which stem from
39
40 distant causes like the actions of end users, will only be intelligible to managers that are already
41
42
well-versed on the climate issue. Alternatively, organizations that have built up these
43
44
45 understandings may restrict innovative approaches, as when they direct actions to conform to
46
47 established protocols. Consequently, there is a tension between support, which creates resources
48
49 to act, and ambiguity, which provides opportunities to act. In addition, organizational scale
50
51
52 affects this tension by influencing the flexibility and distribution of existing issue
53
54 understandings. For example, larger organizations are likely to have more formalized routines
55
56
57
58
59
60
 Academy of Management Journal Page 12 of 63

1
2
3 and policies, entrenching understandings (Sandhu & Kulik, 2018), but also a greater variety of
4
5
6 understandings across a more dispersed workforce (Crilly et al., 2012; Weber & Waeger, 2017),
7
8 while smaller organizations may have fewer resources and expertise around newer issues, but
9
10 also stronger consensus and vision.
11
12
13
In seeking to advance reforms, embedded activists can draw on external movement
14
15 resources in different ways to manage these opportunity structures. External collaborators are a
16
17 vital source of new ideas and perspectives in many other contexts (Deken et al., 2018; Lauche,
18
19
2019; Strike & Rerup, 2016), but how do embedded activists use movement resources to catalyze
20
21
22 existing issue understandings within varying organizational contexts? How do these resources
23
24 combine with the balance of support and ambiguity? In exploring these questions, we aim to
25
26 generate a theory of embedded activism that considers how organizational and environmental
27
28
29 contingencies combine to advance reform (Crilly et al., 2012; Weber & Waeger, 2017).
30
31 EMPIRICAL CONTEXT
32
33 In order to investigate embedded activism, as a coordinated and collaborative form of
34
35 activism, we study EDF Climate Corps. EDF is one of the most prominent collaborative SMOs
36
37 in the U.S., with a focus on environmental issues. Through the Climate Corps program, EDF
38
39
40 sends graduate student fellows inside partner organizations to help translate the unwieldy and
41
42 often politicized issue of climate change into actionable responses. EDF acts as a broker as it
43
44 recruits host organizations and negotiates the initial scope of fellowship work with them, and
45
46
then, through the fellow, facilitates the transfer of information and expertise from external
47
48
49 experts and EDF staff. The program director summarizes its operations: “trained environmental
50
51 change agents are embedded inside organizations, working from the inside out to identify
52
53 solutions and build the business case for investment in energy efficiency” (Delaney, 2017: 131).
54
55
56 From 2008-2016, nearly 400 organizations have enlisted over 750 EDF Climate Corps fellows in
57
58
59
60
Page 13 of 63 Academy of Management Journal

1
2
3 ten-week projects to design customized solutions to challenges involving energy efficiency,
4
5
6 renewable energy, and energy strategy. Through this work, “EDF sees a future where fellows
7
8 continue to evolve, grow, and expand their project work...accelerating the pace of change in the
9
10 energy sector” (Delaney, 2017: 143).
11
12
13
The structure of this program allows unique insights into how embedded activism
14
15 operates across diverse organizational contexts. Although all host organizations already have
16
17 some interest in energy reforms, there is considerable variance in prior issue development. Hosts
18
19
range from multinational corporations with aggressive climate goals and sustainability
20
21
22 departments to single unit establishments launching their first initiatives. These comments from
23
24 an EDF manager capture this variance in prior issue development:
25
26 For example, one of my hosts is very, you know, sustainability is not part of their business. They
27 know that they need to do something with it and they need to report, they have goals that they
28 need to report. But otherwise, I don’t think there’s a ton of buy-in. So I definitely do think that
29
makes it more difficult for the fellow trying to gather information and just explain what the heck
30
they’re doing there because people are like “Wait, what? Why are you collecting energy data?
31
32
Why are you trying to come up with this model for tracking this?” It’s just not something that is
33 in the culture as opposed to a company that’s been doing sustainability for years and has publicly
34 stated goals and has ten people on their energy team.
35
36 Consistent with the goals of Climate Corps to expand reform ambitions and our focus on
37
38 the proactivity of fellows as embedded activists, we aim to understand the creation of new
39
40
solutions within each project that go beyond the original project plans. The proactive
41
42
43 development of solutions is central to the type of issue selling tactics that Climate Corps uses to
44
45 propel change at partner organizations (Dutton et al., 2001; Howard-Grenville, 2007). EDF
46
47 managers consistently remarked on how high-achieving fellows uncovered hidden problems and
48
49
50 new opportunities. One EDF manager noted that by being proactive, fellows are “discovering
51
52 something that was overlooked by [hosts] in the first place and bring even more value that way.”
53
54 Reflecting their support for reforms, host supervisors also viewed fellow proactivity as “going
55
56
57
58
59
60
 Academy of Management Journal Page 14 of 63

1
2
3 beyond” rather than “mission drift.” Fellows internalized the pursuit of new solutions for the
4
5
6 hosts, as reflected in their goals for their fellowship: “move the company forward on a culture
7
8 shift…to a view that integrates sustainability in business strategy,” “to convince [my host] to
9
10 expand its energy efficiency and sustainability initiatives,” and to “help create tangible change
11
12
13
by helping implement concrete energy-saving projects, as well as pushing for organizational
14
15 changes that open the door for future progress.”
16
17 Organizations are more likely to implement reforms when actionable solutions are
18
19
available (Alt & Craig, 2016; Dutton, 1988; Bansal, 2003). In a host survey conducted 15
20
21
22 months after the 2016 fellowships completed, EDF found that 93% of the respondents (N=68
23
24 hosts, 31 hosts did not respond) said they used or implemented some or all of the
25
26 recommendations made by their fellow. In open-ended responses, multiple hosts articulated how
27
28
29 the fellows’ work helped to advance their climate efforts. One noted, “It truly drove operational
30
31 changes in a majority of the organization’s properties.” Another reflected, “The 2016 Fellow’s
32
33 work and final report were very well received within the organization and instrumental in the
34
35
36
concepts being considered on a large scale.” Non-implementation is commonly due to lack of
37
38 capital, organizational discord, or departure of the “project champion” (Delaney, 2017: 137).
39
40 Solutions also become strategic resources that activists can use to push for greater
41
42
reforms (Buchter, 2020). By seeding organizations with new solutions, EDF fellows enrich the
43
44
45 organizational context with new ideas, enhancing the potential for reform and the scope for
46
47 future action. As a specific example, one EDF manager reported:
48
49 Our first year working with [a company], we hired a fellow to do basic energy efficiency, just
50 start looking at what investments they could make in lighting, what investments they could make
51 primarily in cooling. The fellow did that and did a really great analysis, but what he really
52
identified is that nothing was systematized...So his recommendation was ok, here are some quick
53
and easy things that you could fix, but really the bigger, more attractable problem was how do
54
55
you get a handle on management, how do you start to build this as an internal unit, as an approach
56 to doing this effectively. And, ultimately, how do you save a company money by doing those. He
57
58
59
60
Page 15 of 63 Academy of Management Journal

1
2
3 wasn’t tasked with doing that, he just saw it as an opportunity, and since then we’ve had fellows
4 that have taken that and run with it at [the company] and they’re looking more at a portfolio
5 approach to investing in energy management.
6
7
8
9 Next, we explain our methods for investigating how fellows match movement resources with
10
11 organizational contexts to develop new solutions.
12
13 METHODS
14
15 Data
16
17 We studied the 2016 cohort of Climate Corps, which encompassed 126 fellows at 99 host
18
19 organizations. Reflecting the growth and diversity of the program, hosts included a diverse array
20
21 of sectors, with 66 companies, 20 government agencies, and the rest nonprofits and higher
22
23
24 education organizations. Our data consisted of surveys, interviews, and observations, as well as
25
26 archival data from EDF, including the initial project proposals from host organizations and the
27
28 final reports from fellows. These data enabled us to compare how the project evolved in response
29
30
to the fellow’s interactions with the organization and the Climate Corps network. As we explain
31
32
33 below, we derived our measure of solution creation from this comparison.
34
35 To gather data on movement resources and fellows’ experiences, we conducted a survey
36
37 of the 126 fellows in the 2016 Climate Corps program in three waves at the beginning, middle,
38
39
40 and end of their fellowship. These surveys captured on-going processes over the course of the
41
42 projects, minimizing the bias that emerges in reflective interviews. Each survey consisted of 20
43
44 close-ended questions, repeated across panels, that asked about resource use and evaluation of
45
46
47 the experience. There were also open-ended questions that asked for more details about
48
49 experiences, tactics, and outcomes (see Supplemental Materials for survey questions). In order to
50
51 maximize our sample, we included all 94 fellows (75%) who completed at least one survey. We
52
53
ensured that there were no significant differences in the variables used in our analysis due to
54
55
56 response rate: T tests comparing respondents and non-respondents were null and F tests from
57
58
59
60
 Academy of Management Journal Page 16 of 63

1
2
3 ANOVA models comparing each response level were also null.1 Missing data reduced our
4
5
6 sample to 86 fellows. Of the 86 cases, 22 worked at organizations that hosted multiple fellows in
7
8 2016, which could possibly have altered the relevant project dynamics. However, a T test
9
10 indicated no significant difference between these groups in solution creation (p-value of 0.68).
11
12
13
There were three pairs of fellows who worked on related projects at their host organizations.
14
15 Conservatively, we dropped these six cases to remove any influence from overlapping projects,
16
17 leaving a final sample of 80 fellows. Our findings were nearly identical when including all 86
18
19
fellows, and the two largest groups of paths were also evident in a subset analysis of the 64 solo
20
21
22 fellows (see Table 3 in Supplemental Materials).
23
24 To gain a more textured understanding of the program’s dynamics, we also conducted 55
25
26 interviews across program participants: 31 Climate Corps Fellows from the 2016 cohort (19
27
28
29 shortly after the completion of their fellowship in August-October 2016 and 12 in February
30
31 20202), as well as 12 alumni of the program (May-June 2016), 9 EDF engagement managers who
32
33 worked as liaisons between the fellows and the host organizations (May-June 2016), and 3 host
34
35
36
supervisors (2017). The interviews were semi-structured and lasted approximately 45 minutes
37
38 (see interview guide in Supplemental Materials). During the interviews we asked about
39
40 individual motivations, interests, tactics, and roles in engaging with climate change initiatives.
41
42
We were particularly interested in how fellows utilized EDF resources, as well as the roadblocks
43
44
45 and successes they experienced. We recorded and transcribed the interviews for analysis in
46
47 NVivo. Table 1 provides an overview of the data sources.
48
49
50
51
52
53 1 There was a marginal F test (p < 0.096) for external search (see measures section), which reflected
54 moderately greater search for fellows with fewer than three survey responses.
55 2 The interviews in 2020 were completed by a research assistant who was blinded to the theorizing and
56 QCA analysis to minimize risk of bias.
57
58
59
60
Page 17 of 63 Academy of Management Journal

1
2
3 -------------------------------
4
5
Insert Table 1 about here
6
7
8
-------------------------------
9
10 Qualitative Comparative Analysis
11 We analyzed our data with fuzzy set Qualitative Comparative Analysis (fsQCA) to
12
13
14 investigate how organizational contexts and movement resources combined to generate new
15
16 solutions, using the fellow as the unit of analysis. This method excels at analyzing causal
17
18 processes that include combinations of conditions (conjunctural causation) and that can develop
19
20
21
through multiple distinct paths (equifinality) (Fiss, 2011; Ragin, 2008). These types of processes
22
23 were likely to characterize embedded activism, where a variety of internal and external factors
24
25 jointly affect outcomes, and different approaches can be effective. While conjunctural causation
26
27 and equifinality are difficult for regression analysis to handle, fsQCA is well-equipped for them.3
28
29
30 The fsQCA method requires that variables are measured as fuzzy sets that range from
31
32 zero to one, where zero means a case is completely out of the set, one means it is completely in,
33
34 and 0.5 is the crossover point where there is maximum uncertainty about set membership. For
35
36
37 most variables, we used the standard direct calibration method, which uses a logistic function to
38
39 fit the data between thresholds for completely out, completely in, and crucially, the crossover
40
41 point (Ragin, 2008; Schneider & Wagemann, 2012). To select the thresholds, we generally relied
42
43
44
on contextual knowledge due to the novel way we operationalized theoretical concepts to fit our
45
46
47
48
49
50 3 In regression analysis, conjunctural causation requires complex interaction modeling and can increase
51 multicollinearity. In contrast, QCA addresses conjunctural causation by analyzing combinations of causal
52 conditions rather than individual variables. Also, equifinality hinders regression analysis because if the
53 outcome occurs when a predictor is low, there is a weaker relationship between the predictor and the
54 outcome (assuming a positive relationship). However, with QCA this situation is instead evidence that
55 there are other pathways that lead to the outcome, supporting the discovery of multiple distinct causal
56 paths. (Contradictory evidence in QCA would be when the predictor is high but the outcome is low).
57
58
59
60
 Academy of Management Journal Page 18 of 63

1
2
3 empirical context. After calibration, QCA then involves an analysis of set relations to identify
4
5
6 conditions that are sufficient and/or necessary for an outcome.
7
8 Our primary analysis sought to identify combinations of conditions that were sufficient to
9
10 generate new solutions. Following the set-theoretic logic of QCA, these combinations cause the
11
12
13
outcome in the sense that if the combinations occur, then the outcome will also occur. To
14
15 identify these sufficiency relationships, we used the QCA module in R with the following steps
16
17 (Dusa, 2017). First, we constructed a truth table to organize the data into rows for each possible
18
19
combination of the explanatory conditions (see Table 1 in Supplemental Materials). We used the
20
21
22 standard consistency threshold of 0.8 to identify positive rows, which are combinations that are
23
24 subsets of the outcome (Greckhamer et al., 2018; Schneider & Wagemann, 2012). Following
25
26 established practices, we included rows with only one case and excluded rows with true
27
28
29 contradictions, which were cases that belonged to the row but were not members of the outcome
30
31 (ibid). Second, we followed the “enhanced standard analysis” procedures for treating logical
32
33 remainders, which were possible combinations that lacked observations, and using Boolean
34
35
36
algebra to generate a logical reduction of the included rows (Schneider & Wagemann, 2012;
37
38 Dusa, 2017). Accordingly, we excluded rows with theoretical and logical contradictions, such as
39
40 rows that were consistent with the outcome and its converse. We then compared the
41
42
“parsimonious” solution, which accepts all remainders that contribute to simplifying the solution
43
44
45 term, to the “intermediate” solution, which includes the subset of simplifying remainders that are
46
47 also consistent with theoretical expectations. Third, following established guidelines, we
48
49 included all paths from the solution term with a consistency above 0.8 and a proportional
50
51
52 reduction in inconsistency (PRI) above 0.7 (Greckhamer et al., 2018).
53
54
55
56
57
58
59
60
Page 19 of 63 Academy of Management Journal

1
2
3 Following QCA best practices (Greckhamer et al., 2018; Schneider & Wagemann, 2012),
4
5
6 we also conducted additional analyses. As part of a full QCA, we first checked for the existence
7
8 of any necessary conditions (i.e., if solution creation is present, then this condition must also be
9
10 present), but there were none. We also verified the robustness of the set relations with a higher
11
12
13
consistency threshold (0.85) for including truth table rows in the minimization step. QCA also
14
15 supports asymmetric causality and we report analyses of the inverse of the outcome, i.e., low
16
17 solution creation.
18
19
Measures
20
21 Outcome: High solution creation. We operationalized the creation of new solutions as
22
23
additions to the project scope from the initial proposal, which the host organization designed, to
24
25
26 the final project summary, which the fellow authored and the host approved for public
27
28 dissemination. Because these summaries were shared publicly, hosts were cautious about what
29
30 was included as the summaries could be perceived as commitments for future efforts. Through
31
32
33 additions of new solutions, fellows encouraged their host organizations to become more
34
35 ambitious in their climate change reforms. For example, one project summary reported: “Pushing
36
37 beyond the project’s initial scope, [the fellow] designed an innovative framework for Employee
38
39
40 Crowd Financing for Public Buildings energy efficiency projects.” We coded both the initial plan
41
42 and the final report to identify discrete reforms, such as lighting upgrades or renewable energy
43
44 plans, and then extracted the ones that were added (i.e., were not part of the original proposal).4
45
46
High solution creation represented the proactive work of fellows to advance climate change
47
48
49 reforms at their host organizations. Since we cannot know the entire scope of organizational
50
51
52
53
54
55 4 We also included proposal specificity as a key condition in our analysis in part to incorporate its
56 influence on solution creation.
57
58
59
60
 Academy of Management Journal Page 20 of 63

1
2
3 knowledge ex ante, these ideas were new in the limited sense that they were not part of the
4
5
6 original project plan.
7
8 We coded the project data based on a modified version of the U.S. Department of
9
10 Energy’s Assessment Recommendation Code. This framework was created to track specific
11
12
13
recommendations made in an energy audit program, and has been used in prior research on
14
15 organizational change (Dowell & Muthulingam, 2017). Two research assistants coded the project
16
17 data. They initially jointly coded projects from a previous year of the Climate Corps program.
18
19
Then once they established consistent coding, they moved to coding the 2016 projects. They
20
21
22 independently coded 20 projects at a time, with 5 projects that overlapped. The overlapped
23
24 projects were used to ensure coding consistency over time. The research assistants flagged any
25
26 questions and met with the authors weekly to resolve any open questions. The authors reviewed
27
28
29 the coding to ensure alignment. Using this coding procedure, we found the average fellow
30
31 created 2.6 new solutions that the organization could implement to address climate change
32
33 beyond those specified in the project proposal. The median outcome was two new solutions,
34
35
36
while eight percent did not break any new ground and five percent added five or more new
37
38 solutions (see Figure 1). The prevalence of new solutions also indicated the normativeness and
39
40 centrality of solution creation to Climate Corps.5
41
42
43
44
45
46
47
5 While our focus on the development of new ideas reflected our theoretical interest in fellow proactivity,
48 there were also averages of 2.9 ideas in the project plan, 1.1 of these completed, and 1.8 dropped. These
49 numbers further indicated the importance of fellows’ ability to proactively adapt beyond the initial plan, a
50 point that EDF staff members confirmed in interviews. We also verified that the numbers of ideas
51 planned, proposed, and completed did not determine the number of ideas added. Chi-squared tests
52 between ideas added and the other items were all null. There was a marginal Chi-squared score for ideas
53 added and dropped (p < 0.1) but this disappears after removing a single outlier that added seven ideas and
54 dropped five. Note that by defining set membership with substantive thresholds rather than with variation
55 around a mean, QCA mutes the influence of outliers. Likewise, the percentages of ideas completed and
56 dropped were unrelated to ideas added, based on null F tests from ANOVA regressions.
57
58
59
60
Page 21 of 63 Academy of Management Journal

1
2
3 -------------------------------
4
5
Insert Figure 1 about here
6
7
8
-------------------------------
9
10 We calibrated the set of projects with high solution creation using the thresholds of zero
11
12 for completely out, four for completely in, and 1.5 for the crossover point. We chose the
13
14 crossover threshold to reflect the difference between tagging on a single new solution versus
15
16
making more extensive contributions. Making multiple additions suggests a more consistent and
17
18
19 developed ability to drive solution creation. Although the outer thresholds have less influence on
20
21 QCA results, it is important to explain the logic behind these decisions as well. For the upper
22
23 anchor, the difference between four versus three additional solutions reflected a qualitative step
24
25
26 up in terms of making numerous versus multiple additions. The lower anchor of zero additions
27
28 was self-explanatory. Some examples help to better explain these choices. In a case that was
29
30 fully out, the fellow stuck to the plan to assist with carbon reporting and did not create any new
31
32
33 solutions. In a case that was close to fully in, the fellow expanded on an initial plan to improve
34
35 employee engagement and track progress towards sustainability goals with several new solutions
36
37 around lighting, HVAC systems, facility layouts, and solar installations. In a case that was more
38
39
in than out, the fellow expanded on a plan to conduct analyses of the company’s carbon footprint
40
41
42 and climate risks from suppliers by developing an implementation strategy for supplier
43
44 engagement and training materials for suppliers.
45
46 Contributing Conditions: Organizational Context. We included four conditions to
47
48
49 analyze the balance of organizational support and ambiguity concerning how the fellow should
50
51 develop his/her project: new initiative, program newcomer, open-ended project, and small
52
53 organizational scale. These conditions connect to theorizing in the proactivity literature:
54
55
56 “Ambiguity may be present in a variety of weak situations, including unclear job and role
57
58
59
60
 Academy of Management Journal Page 22 of 63

1
2
3 prescriptions, vague task instructions, or unspecified interpersonal expectations, routines, and
4
5
6 standard operating procedures” (Grant & Ashford, 2008: 15). If the organization has less
7
8 experience with the energy issue and with the Climate Corps program, as well as vague plans for
9
10 the fellow, then there is more ambiguity about the definition of the climate issue and how to
11
12
13
proceed. In addition, organizational scale can both decrease ambiguity through greater
14
15 formalization and increase it through greater diversity and fragmentation of issue understandings.
16
17 First, we examined whether the project was a new initiative for the host organization,
18
19
meaning that the project did not build on prior resources specific to energy issues. This
20
21
22 organizational history sets the context for embedded activism. Prior issue engagement can
23
24 provide resources and schemas for further actions (Bansal, 2003; Howard-Grenville, 2007),
25
26 while also limiting the space for creative action to develop an organization’s issue understanding
27
28
29 (Sonenshein, 2014, 2016). As an example of the latter point, one EDF manager commented on
30
31 the opportunities of working with organizations that were new to the climate issue: “it’s a
32
33 profound project when they say, ‘we think we’re ready to set a goal, we see our peers setting
34
35
36
goals, we know it’s probably important, we don’t know how to do it, let’s hire a Climate Corps
37
38 fellow to help us start to scope that out. Those are some of my favorite projects because the
39
40 fellow has this opportunity to really change the course for a company.” Likewise, another
41
42
engagement manager identified how prior issue development can limit a fellow’s scope of
43
44
45 action:
46
47 [An experienced organization] has already decided on a lot of things. One of my fellows was
48 working with a city and she said, “They have this carbon reduction goal but I don’t think that
49 should be their goal. I think that they should change their goal.” And she actually suggested that
50 they revise their goal. But she was like, “should I do that? They’ve already set this goal and
51 should I be telling them they should change it?” So that’s a challenge that can come up too that
52 [the fellow] thinks there’s something [the organization] should be doing differently but [it] might
53 be a little set in its ways.
54
55
56
57
58
59
60
Page 23 of 63 Academy of Management Journal

1
2
3 We measured the extent to which the project was a novel initiative for the host
4
5
6 organization by coding the project plans and reports for binary mentions of climate change
7
8 commitments (e.g., this project will help/helped us meet our energy goal), prior work (e.g., the
9
10 fellow will continue/continued to develop our climate initiative), and other resources (e.g., the
11
12
13
fellow will work/worked with our sustainability team and energy database). Including
14
15 information from both the proposal and report increased the recall and relevance of
16
17 organizational resources. Also, the nature of these resources ensured that they predated the
18
19
fellows’ efforts: i.e., even the most effective projects did not lead to new climate goals and staff
20
21
22 within the ten-week time period of the fellowship. We established an inventory of organizational
23
24 resources and used this to develop a coding dictionary (see Table 2 in Supplemental Materials).
25
26 Two research assistants followed a similar coding approach as for solution creation. We met and
27
28
29 reviewed progress weekly to ensure coding consistency and accuracy. We created an additive
30
31 index of these codes and multiplied by negative one to reverse the scale, which ranged from zero
32
33 to negative six when each component was present in both the plan and the report.
34
35
36
We then calibrated the index into a fuzzy set using the direct method and thresholds of -1
37
38 for completely in, -5 for completely out, and -2.5 for the crossover point. We chose -2.5 to
39
40 demarcate cases with lesser variety and consistency in organizational preparation. To achieve a
41
42
score greater than -2.5, at most there was only a single consistent component or two inconsistent
43
44
45 components. Below the threshold, at least all three components were inconsistently present (in
46
47 one document only) or one component was consistently present (in both documents) plus there
48
49 was an additional inconsistent component. At the upper threshold of -1, the case contained at
50
51
52 most one inconsistent component. At the lower threshold of -5, all three components were
53
54 present and at least two were consistent.
55
56
57
58
59
60
 Academy of Management Journal Page 24 of 63

1
2
3 Second, we included the degree to which the organization was a program newcomer. An
4
5
6 organizational history of involvement in the Climate Corps program would likely provide
7
8 support for the fellow while restricting ambiguity about expectations for the fellow’s work. From
9
10 a support perspective, repeated engagements seed an organization with schemas and resources,
11
12
13
such as analytic tools, datasets, and engaged supporters, that may enable fellows to push the
14
15 organization even further (Howard-Grenville, 2007; Wickert & de Bakker, 2018). An EDF
16
17 manager reported an example of this dynamic: after the first fellow at one company
18
19
recommended developing an energy management system in 2014, “the 2015 fellow took that,
20
21
22 and ran with it, started drafting an RFP, was working with local vendors, scoped the system out,
23
24 and now here in 2016 they’ve implemented the recommended system. Now their fellow in 2016
25
26 is starting to use that data to make decisions around where to go next, and that’s largely tied to
27
28
29 that first project in 2014.” However, such prior activity also potentially restricts what fellows
30
31 should do and how the climate issue should be addressed, as the prior fellow’s tool can become a
32
33 fixed schema that limits further creative action (Sonenshein, 2014).6
34
35
36
After reverse coding the number of prior fellows at an organization, we calibrated the set
37
38 of program newcomers using the direct method and thresholds of 0 previous engagements for
39
40 completely in, -5 for completely out, and -0.99 for the crossover point. The crossover threshold
41
42
captured the transition from some program experience to none. We chose a higher number of
43
44
45 engagements (reverse-coded) for the lower boundary to distinguish the organizations with
46
47 extensive Climate Corps experience from the organizations with intermediate experience.
48
49
50
51
52
53 6 Even if fellows worked in different parts of an organization, there will be stronger expectations for the
54 later fellow to the extent that communication occurs across the divisions, which was likely unless
55 participation by multiple parts of an organization in Climate Corps was a coincidence. In addition, later
56 fellows received guidance from prior fellows through EDF channels about how to act in the organization.
57
58
59
60
Page 25 of 63 Academy of Management Journal

1
2
3 In addition to new initiative and program newcomer, which reflected organizational
4
5
6 history, open-ended project captured the level of specificity in the initial project plans. While
7
8 detailed plans could indicate greater organizational support, they could also stifle fellow
9
10 creativity. In contrast, fellows might benefit from the greater latitude available in more open-
11
12
13
ended projects (Deken et al., 2018; Kannan-Narasimhan & Lawrence, 2018; Sonenshein, 2014).
14
15 If managers have only a general sense of how an issue should be addressed, then embedded
16
17 activists will be more empowered to create new solutions.
18
19
We developed this measure from the level of detail in the initial project proposal. To
20
21
22 capture this info, we utilized the hierarchical design of the coding framework that we applied to
23
24 code the distinct energy reform ideas. This framework had four nested levels of specificity. For
25
26 example, the code for installing occupancy lighting sensors was at the most detailed level (level
27
28
29 4) and was beneath three nested levels: building energy use (level 1), lighting (level 2), and
30
31 lighting controls (level 3). After reverse coding, we calibrated the set of open-ended projects
32
33 using the direct method and thresholds of average specificity in the project proposal of -1 for
34
35
36
completely in, -4 for completely out, and -2.9 for the crossover point. The crossover threshold
37
38 distinguished projects about general problem areas from those that indicated types of solutions
39
40 (e.g., lighting versus lighting controls). At the upper threshold of -1, projects were maximally
41
42
general (e.g., building energy use), while at the lower threshold, they were maximally specific
43
44
45 (e.g., occupancy lighting sensors).7
46
47
48
49 7 An example of a plan that was close to fully out: “The fellow will assist in three main focus areas. The
50 first is compiling and accessing project data for 179D Tax Credit Program. Secondly the fellow will
51 evaluate energy savings potential associated with installation of a Wireless Energy Management System.
52 Finally, the fellow will look at renewables, specifically the installation of a solar wall on mid-rise
53 buildings.” An example that was more in than out: “The energy program could use assistance with
54 formalizing a process for identifying the best energy conservation projects, implementation of the energy
55 projects, and documenting the savings. The energy program also needs assistance with energy data
56 management such as the current benchmarking efforts, energy monitoring, and future dashboards.”
57
58
59
60
 Academy of Management Journal Page 26 of 63

1
2
3 Our last contextual condition captured organizational scale. Organizational scale can both
4
5
6 inhibit and enhance ambiguity. Larger organizations should have more elaborated procedures
7
8 and frameworks that reduce ambiguity and the opportunities for proactivity (Hannan & Freeman,
9
10 1984; Morrison, 2011: 386; Sandhu & Kulik, 2018). However, larger organizations can also
11
12
13
include more diverse perspectives as well as communication barriers, which increase ambiguity
14
15 (Sonenshein, 2016; Baumann-Pauly, Wickert, Spence, & Scherer, 2013; Weber & Waeger,
16
17 2017). Organizational scale can also alter the contributions of movement resources. For example,
18
19
one EDF manager suggested that large organizations heightened the value of an internal focus
20
21
22 and reduced the contribution of external expertise:
23
24 A lot of the companies they are in are big. And so they need to be able to gather a lot of
25 information from a lot of different people quickly. That can be really challenging to navigate an
26 organizational structure that you don’t know anything about – that you’ve just been thrown in. I
27 think people who are really good at figuring out who they need to talk to and what information
28 they need as quickly as possible is really, really valuable. Having those interpersonal skills
29 sometimes can be a lot more valuable than having more content knowledge.8
30
31
We constructed the set of small organizational scale by combining information on the
32
33
34 host organizations’ size and scope that we collected from organizational websites. While we
35
36 would prefer to directly measure formality and centralization, size and scope were our best
37
38 available proxies. Integrating size and scope suggested more than three distinctions, and so we
39
40
41 used Ragin’s (2008) indirect calibration method to recognize finer gradations. Our five
42
43 thresholds distinguished completely out (5,000 or more employees and national operations),
44
45 mostly out (same size and metropolitan operations), more out than in (500 or more employees
46
47
48
and national operations), more in than out (same size and metropolitan operations), and mostly in
49
50 (fewer than 500 employees and local operations) from completely in.
51
52
53
54
55 8The temporary status of the organizational members in our sample heightened the challenges from
56 organizational scale of communication and navigating protocols.
57
58
59
60
Page 27 of 63 Academy of Management Journal

1
2
3 Contributing Conditions: Movement Resources. In addition to these measures of the
4
5
6 organizational context, we included two conditions to analyze the use of movement resources:
7
8 external search and external focus. We utilized our survey data about exploration and advice-
9
10 seeking behaviors to construct two distinct sets around engagement with the EDF network. We
11
12
13
averaged the data for fellows with multiple surveys and verified that there were no significant
14
15 differences between groups with different response rates (all F tests > 0.096). These data enabled
16
17 us to characterize the extent to which fellows activated their external connections through EDF.
18
19
Our first measure captured the intensity of external search for resources. Fellows reported
20
21
22 how often they used each of seven categories of EDF resources in the last month (other fellows,
23
24 engagement manager, resource library, consultants, webinars, LinkedIn group, and mentors). We
25
26 calibrated the set of high external search with the direct method and thresholds of 20 for
27
28
29 completely in, 5 for completely out, and 10.1 for the crossover point. We chose this crossover
30
31 point because it corresponded to about 2.5 connections with EDF a week, indicating external
32
33 search activity on more workdays than not. At the upper threshold, fellows sought external
34
35
36
resources nearly every workday, while at the lower threshold, it was closer to once a week. For
37
38 example, in a case that was close to fully in, the fellow reported: “I used [the EDF network] a lot:
39
40 the LinkedIn group, I posted I don’t know how many questions on there, maybe 20 of them. So it
41
42
was really great because people were pretty willing to respond, and even if they didn’t know the
43
44
45 answer then they would direct you to what resources they knew could be helpful to you.” While
46
47 a fellow in a case that was close to fully out commented that she used EDF to “network after
48
49 work, mostly” and instead “worked with different staff within the organization to meet my
50
51
52 goals.”
53
54
55
56
57
58
59
60
 Academy of Management Journal Page 28 of 63

1
2
3 Second, we calculated the ratio of advice-seeking from the Climate Corps network to
4
5
6 advice-seeking from within the host organization, based on separate questions from the survey.
7
8 From these data, we calibrated the set of external focus with the direct method and thresholds of
9
10 1.25 for completely in, 0.5 for completely out, and 0.99 for the crossover point. We chose the
11
12
13
crossover threshold so that fellows in this set were at least as likely to seek advice from their
14
15 external connection as from their host organization, even though their direct supervisor was in
16
17 the host organization.9 At the upper threshold, external advice was used substantially more, while
18
19
at the lower threshold, internal advice predominated. The spacing between these thresholds was
20
21
22 uneven because we would expect fellows to seek advice from their supervisor and other
23
24 organizational members. Comparing the two conditions, external search captured whether a
25
26 fellow intensively sought to incorporate movement resources, while external focus indicated
27
28
29 whether the fellow was aligned towards external versus internal guidance.
30
31 Although more conditions were possible for measuring organizational contexts and
32
33 external resources, these six were the most meaningful for investigating embedded activism, and
34
35
36
QCA requires parsimony because there are 2n possible combinations with n conditions. Table 2
37
38 provides summary information for each condition.
39
40 -------------------------------
41
42 Insert Table 2 about here
43
44 -------------------------------
45
46 Qualitative Interview and Survey Analysis
47
48 In order to better understand why particular combinations of conditions produced high
49
50 solution creation, we triangulated QCA results with the qualitative data from interviews and
51
52
53
54 9 While both the supervisor and the fellow affected the extent of internal advice seeking, we only
55 observed the fellow-side of the relationship. From our interviews, it was typical for supervisors to have
56 little time to engage with fellows as this was not the supervisors’ primary jobs.
57
58
59
60
Page 29 of 63 Academy of Management Journal

1
2
3 open-ended survey questions. It is important to note that the datasets for each analysis were
4
5
6 distinct – the interviews and open-ended survey responses were not used to determine any of the
7
8 conditions analyzed in QCA. This triangulation of different data sources and analyses
9
10 strengthened confidence in our groupings of distinct pathways (Patton, 2015; Stake, 1995), and
11
12
13
provided a lens into the underlying mechanisms that connected the conditions in each pathway.
14
15 Our approach to coding the interviews and qualitative survey responses was analytic
16
17 abduction (Peirce, 1955): we iterated between reviews of pre-existing theoretical constructs and
18
19
multiple readings of interview transcripts and open-ended survey responses (Snow, Morrill, &
20
21
22 Anderson, 2003). This analytic procedure included analyst triangulation: we used multiple
23
24 analysts, including two research assistants who were blinded to the QCA analysis, to minimize
25
26 biases (Miles & Huberman, 1994; Patton, 2015). Our analysis consisted of the following stages.
27
28
29 First, we used NVivo to code the interviews and survey responses for instances of
30
31 resource use. Importantly, this coding was completed by the two research assistants following a
32
33 similar coding strategy as for solution creation. However, in order to minimize bias, the weekly
34
35
36
progress meetings to ensure coding consistency and accuracy occurred only with the second
37
38 author who was blinded to the QCA analysis outcomes. Second, after the coding was complete,
39
40 we grouped the fellows by each pathway that emerged in the QCA analysis and reviewed all
41
42
examples of resource use to check alignment with the pathways from QCA. Third, we
43
44
45 constructed narrative examples of the fellows’ experiences from the interview and survey
46
47 qualitative data, focusing on conditions that were important in each QCA pathway and
48
49 interdependencies among the conditions (Heinze & Weber, 2016). Through evaluating not just
50
51
52 the conditions in isolation, but a narrative at the individual level, we integrated the qualitative
53
54 data to illustrate the QCA pathways and investigate underlying mechanisms. We compared
55
56
57
58
59
60
 Academy of Management Journal Page 30 of 63

1
2
3 approaches to using movement resources across the QCA pathways, looking for similarities and
4
5
6 differences in the underlying mechanisms. This allowed us to develop a more nuanced
7
8 understanding of how fellows utilized external resources differently across varying levels of
9
10 contextual ambiguity. Finally, in a follow-up interview with a key informant at EDF, we also
11
12
13
were able to confirm that our research design and findings were consistent with EDF’s objectives
14
15 and understandings of the Climate Corps program.
16
17 RESULTS
18
19 The six columns in Table 3 each document a pathway to high solution creation. The table
20
21 provides evidence of equifinality in that multiple distinct combinations of conditions are
22
23
sufficient to generate the outcome. Note that filled circles indicate the presence of a condition
24
25
26 and circles with “X” indicate its absence, while blank spaces indicate that the condition is
27
28 irrelevant to the configuration because some cases include it and some do not. Large circles are
29
30 core conditions from the parsimonious solution, while small circles are peripheral conditions
31
32
33 from the intermediate solution, which makes restrictions based on theoretical expectations (Fiss,
34
35 2011). We used the expectations that preparation, prior fellows, and external search would
36
37 contribute to solution creation, and we note when we are discussing peripheral conditions that
38
39
40 rely on these expectations. Consistency and coverage statistics are at the bottom of the table.
41
42 Consistency reports the strength of the configuration’s subset relationship with solution creation,
43
44 which is akin to a measure of statistical significance of a sufficiency relationship between the
45
46
configuration and the outcome. Providing data on effect sizes, raw coverage summarizes the
47
48
49 percentage of case evidence that fits a configuration, while unique coverage refers to case
50
51 evidence that only fits with that configuration. Overall, the solution shows strong consistency
52
53 (0.828) and coverage (0.65). Out of 80 cases, 49 of them are in positive rows in the truth table
54
55
56 (see Table 1 in Supplemental Materials), and 34 of these are in the solution.
57
58
59
60
Page 31 of 63 Academy of Management Journal

1
2
3 In order to interpret a QCA table, it is helpful to identify a key condition or small number
4
5
6 of conditions, and then examine the configurations around those conditions. Reflecting our
7
8 interest in how fellows combine movement resources with different levels of contextual
9
10 ambiguity, we organize our analysis by the level of ambiguity in each pathway, focusing on the
11
12
13
three conditions that clearly denote greater ambiguity (new initiative, program newcomer, and
14
15 open-ended project): low where there are only conditions that reduce ambiguity, moderate where
16
17 there is a mix of conditions increasing and decreasing ambiguity, and high where there are only
18
19
augmenting conditions. Consistent with our view of organizational scale as having the potential
20
21
22 to both restrict and augment ambiguity, there are also variants of the low and moderate
23
24 ambiguity groups where small organizational is present or absent.
25
26 By grouping similar pathways in terms of contextual ambiguity, we identify three
27
28
29 different ways that fellows match movement resources with organizational contexts to generate
30
31 high solution creation. In pathways with moderate ambiguity, these resources are either absent or
32
33 irrelevant. In contrast, all the pathways with either low or high ambiguity include at least one
34
35
36
type of movement resource. To understand this pattern, we examine each group of pathways in
37
38 detail, and then present an analysis of how fellows’ use of movement resources differs in
39
40 contexts of low versus high ambiguity. We also enrich our analysis of each group with
41
42
qualitative data from the survey responses and interviews (interviews from around half the cases
43
44
45 in each group of pathways: 10 in low ambiguity, 6 in moderate, and 3 in high).10
46
47
48
49
50
51
52
53
54 10 Interview coverage for the individual paths: Path 1 (8/12), Path 2 (1/5), Path 3 (1/5), Path 4 (5/7), Path
55 5 (1/3), and Path 6 (3/8). Open-ended, qualitative survey coverage for the individual paths: Path 1 (11/12),
56 Path 2 (3/5), Path 3 (5/5), Path 4 (7/7), Path 5 (3/3), and Path 6 (5/8).
57
58
59
60
 Academy of Management Journal Page 32 of 63

1
2
3 -------------------------------
4
5
Insert Table 3 about here
6
7
8
-------------------------------
9
10 Low Ambiguity: Catalyzing Established Issue Understandings
11 The largest grouping contains three pathways that feature organizational conditions that
12
13
14 reduce ambiguity. Each potential source of organizational support that decreases ambiguity is
15
16 present in one path: strongly-defined project plans (Path 1), extensive preparation (Path 2), and
17
18 prior fellows (Path 3). Organizational scale is also part of each pathway in varying states,
19
20
21
reflecting its cross-cutting influence on ambiguity: the organizations in Paths 1 and 3 are small,
22
23 while those in Path 4 are large.
24
25 Each pathway in this group also includes some movement resources. Confronted with
26
27 well-developed issues, through either preparation, program experience, or detailed plans,
28
29
30 successful solution creation here entails leveraging external connections to catalyze established
31
32 frameworks and introduce new perspectives into the organization. A supervisor at a host
33
34 organization from this group identified this dynamic as a motivation for participating in Climate
35
36
37 Corps:
38
39 We bring [the fellows] in on projects that have been long-standing with the company. It’s really
40 helpful for us to get a new perspective on some of these projects because you can fall into that
41 hole of doing things just because it’s the way they’ve always been done. Sometimes getting a
42 fresh perspective is really motivating to look at a project in a new light.
43
44 In Path 1, fellows utilize external search to manage an absence of project openness. This
45
46
path is especially remarkable as detailed oversight should preclude creative acts like solution
47
48
49 creation (Sonenshein, 2014). Henry was an exemplar of Path 1.11 His initial project plan was
50
51 very specific: “The Climate Corps fellow will focus on two discrete tasks. First, they will help
52
53
54
55
56 11 All fellow names are pseudonyms.
57
58
59
60
Page 33 of 63 Academy of Management Journal

1
2
3 the [host] apply for the subsequent round of [type of] funding. Second, using [a specific] team’s
4
5
6 behavioral findings, the fellow will assist with energy planning and setting key milestones so that
7
8 the agency can continue energy efforts until energy personnel are funded.” Despite this detailed
9
10 guidance, Henry added four solutions to the project, including a strategic framework for
11
12
13
prioritizing buildings energy saving opportunities and a model that applied the strategic
14
15 framework to the host’s data. He mobilized external resources in multiple ways, including
16
17 tapping into expertise from the field and social support to help overcome obstacles. These
18
19
distinct uses of movement resources are each reflected in the interview quotes below:
20
21
22 [expertise and tools] I looked into [the EDF] database, and found out the top 10 most relevant
23 fellows who did something similar to me. So I reached out to them and set up an interview, and I
24 asked them about their project...to figure out what kind of challenges I should expect, and how
25 did they overcome them. That was probably one of the most useful things that I did.
26
27 [social support] There were a lot of fellows. We were hanging out a lot together, so we were
28 talking about our projects all the time, like this is what I did, these are the challenges I
29 have…And I got a lot of support.12
30
31 Faced with a very specific project plan, Henry imported new tools and approaches to
32
33
34 create new solutions. The external connection enabled him to find new directions for an already
35
36 elaborated issue. For example, one of his key expansions was the development of a spreadsheet
37
38 tool, which integrates data the host already had collected, to analyze all the buildings they
39
40
41
oversee and prioritize focus:
42
43 When I analyzed all of that data on a city level, especially because such an understanding didn’t
44 exist before, when I presented that to my supervisor, my supervisor was totally on board, so she
45 was seeing the progress...I presented to other people and they were very excited to see that such a
46 tool exists, and I was telling them very politely that you’re focusing on something wrong, like
47 focusing on the small agencies: their budget is $200,000 per year and we have an institution that
48 they use a million dollars of electricity per month. I was like, forget those small ones, we have to
49 focus on these big ones.
50
51
52
53
54 12This fellow met with other fellows in the same geographic area who were working with different hosts.
55 Many fellows mentioned remote connections with each other, so co-location wasn’t necessary for this
56 support. Across the sample, there are null differences in the use of movement resources by city or region.
57
58
59
60
 Academy of Management Journal Page 34 of 63

1
2
3 The tool was effective in part because Henry integrated what he learned from external search
4
5
6 with the data the host had in place, bringing in a new perspective to the host that helped them
7
8 advance their organizational energy goals. Fellows in this path noted the small scale enabled
9
10 them to understand the organization as Henry commented, “I set up individual meetings with
11
12
13
everyone.” However, external search was important in part due to limited internal resources, as
14
15 reflected in the following, “I am working for a very small [organization] and trying to develop
16
17 their energy efficiency program when they are strapped for resources.”
18
19
In Path 2, fellows balance the extensive issue infrastructure implied by a large well-
20
21
22 prepared organization with both extensive external search and an external focus in advice-
23
24 seeking. Fellows again draw on the external connection to catalyze established initiatives. For
25
26 example, Clara worked on identifying key climate risk mitigation opportunities in the supply
27
28
29 chain and added three solutions, including sustainability training videos and webinars for
30
31 suppliers, as well as a resource sheet to support their sustainability evaluations. She mobilized
32
33 the EDF network to build out this program:
34
35
I needed some help understanding how other companies treat supplier engagement efforts,
36
37
especially around ratifying and enforcing compliance with supplier code of conducts. I reached
38 out first to my engagement manager, who was able to connect me with former fellows that had
39 worked on supply chain projects. One of those fellows further connected me with someone in
40 their company who works on the supplier code of conduct and I was able to get an informational
41 interview.
42
43 This use of movement resources helped Clara design custom supplier tools for improving their
44
45 CSR scores, which led to tangible progress towards fulfilling the company’s established
46
47
48
initiatives: “The last week I was there, we found out that three of the suppliers reassessing had
49
50 scored the necessary points in the CSR survey to help [the host organization] meet their target.
51
52 These increased scores were accomplished by one-on-one conversations, resource finding, and a
53
54
webinar co-hosted with the survey company.”
55
56
57
58
59
60
Page 35 of 63 Academy of Management Journal

1
2
3 In the third configuration in this group, Path 3, a history of previous fellows reduces
4
5
6 ambiguity. Interestingly, this path features an external focus with an absence of external search.
7
8 Fellows in this path consistently noted that they reached out to EDF engagement managers and
9
10 previous fellows for advice, for example: “my EDF supervisor provided ideas on how to follow
11
12
13
up projects” (Brittany) and “[It helped] talking to other fellows who are at similar organizations
14
15 or working on similar projects” (Michael). Michael’s reflection highlighted how the small
16
17 organizational scale was related to this external advice-seeking, “Currently at the host
18
19
organization there are limited resources available to help me complete my projects. There are
20
21
22 essentially two people who work on sustainability at the organization and both are very busy. I
23
24 have essentially been relying upon previous fellows and EDF resources in order to complete my
25
26 fellowship.” However, none of the fellows in this pathway mentioned tapping into other potential
27
28
29 EDF resources such as webinars, consultants, LinkedIn, or the online library, and thus were not
30
31 in the high external search set. This mixed pattern reflects a connection between prior fellows
32
33 and external advice-seeking: consulting with prior fellows provides high value assistance,
34
35
36
diminishing the need to engage in extensive additional external search.
37
38 The process of leveraging their external connection to catalyze established issues is
39
40 illustrated through the experience of Carlos, who worked at a city’s transportation agency and
41
42
was tasked with assessing energy use for lighting. He added six solutions, including a plan for
43
44
45 alternative financing options for energy-efficiency projects, an employee engagement survey to
46
47 gauge knowledge of sustainability projects, and a toolkit to highlight the organization’s
48
49 sustainability progress. This work built on prior efforts by fellows, who collected and aggregated
50
51
52 energy data, and also included quick impact, low investment reforms like LED lighting and
53
54 occupancy sensors. Carlos reported:
55
56
57
58
59
60
 Academy of Management Journal Page 36 of 63

1
2
3 The biggest success I experienced with my fellowship was a project that reviewed energy-
4 efficiency and water conservation measures at the agency’s proposed regional transportation
5 center. [And] my greatest source of support for my work was my engagement manager as well as
6
prior EDF fellows at my host organization. My engagement manager was able to listen to my
7
ideas and connect me to other resources. For me, being able to talk to her on a regular basis was
8
9
very helpful.
10
11 Moderate Ambiguity: Navigating Organizational Fit
12
The next largest group contains two paths that combine some conditions that restrict
13
14
15 ambiguity with others that enhance it. There are also again variants with either small or large
16
17 organizational scale. In Path 4, fellows engage with large well-prepared organizations,
18
19 suggesting issue elaboration, but that are also program newcomers, indicating less established
20
21
22 expectations for the fellow’s role. Strongly-defined project plans is also a peripheral condition.
23
24 Path 5 features small organizations that have extensive preparation and prior engagements with
25
26 Climate Corps – two factors that provide an elaborated issue environment – but also open-ended
27
28
29
project plans. Interestingly, in both of these paths, fellows emphasize internal advice-seeking
30
31 over leveraging their external connection. In this mixed environment of internal support and
32
33 ambiguities, fellows succeed through cultivating allies and ownership for sustainability issues
34
35
within the organization. Ambiguities about their role or issue make it crucial for them to attend
36
37
38 to how their initiatives fit with the organizational values and priorities.
39
40 The following example illustrates Path 4, where the fellow relied on internal advice to
41
42 navigate the moderate ambiguity in the organizational context and create solutions. Melanie was
43
44
45 a fellow for a manufacturing company. Her objective was to shape and execute a renewable
46
47 energy strategy. Over the course of her fellowship, she added five new solutions to her project,
48
49 including auditing utility data, conducting a resource use risk analysis, and surveying employees.
50
51
52
While the host had a dedicated group focused on sustainability, the lack of experience with the
53
54 new role of a Climate Corps fellow created ambiguity about expectations and responsibilities.
55
56 Melanie reflected on this ambiguity as a source of tension:
57
58
59
60
Page 37 of 63 Academy of Management Journal

1
2
3 It’s a very hierarchical structure and there’s certain unwritten rules about what you do and don’t
4 do. I didn’t feel very prepared for that and that’s one of the suggestions I gave to my supervisor:
5 how do you function in this multinational corporate environment, particularly as a [fellow] who
6
doesn’t have any power.
7
8
9 The large organizational scale often was experienced by Melanie as another challenge. Fellows
10
11 in this path noted difficulties in figuring out who the key stakeholders were, how to connect with
12
13 them, and where to locate critical information, or if it existed at all. As Melanie noted:
14
15
There are over 100 manufacturing facilities and 26,000 employees, so it’s quite a big operation...I
16
need to collaborate with many different segments of the business and it can be difficult to get the
17
18
information that was needed about the company.
19
20 This combination of support and ambiguity elevated the importance of navigating the
21
22 internal context. Throughout her fellowship, Melanie looked internally for advice, noting that she
23
24
reached out to someone at the host organization for advice at least once a day. The advice helped
25
26
27 in both developing content for recommendations and navigating the organization. With respect to
28
29 content for her work, she connected with plant managers internationally to understand how
30
31 different strategies may work across the host’s footprint. She also reached out frequently to the
32
33
34 legal team and HR “because a lot of the sustainability questions deal with employees, attraction,
35
36 retention, and program issues like training opportunities, education.” With respect to navigating
37
38 the organization, she also used internal advice to manage some more difficult relationships: “I
39
40
41
also was having issues with my supervisor so I talked with someone who had worked with them
42
43 before on the best way to handle it.”
44
45 An internal focus also helped fellows to navigate organizational history around an issue.
46
47
For example, as Melanie learned about the company’s significant history and experience around
48
49
50 sustainability from her supervisor, she began to revise her approach to the project and frame her
51
52 work as aligned with organizational values:
53
54 It’s a culture where many people would not think climate change is a thing, especially not caused
55 by people, so that whole underlying issue. But stewardship was something that was important to
56
57
58
59
60
 Academy of Management Journal Page 38 of 63

1
2
3 see, at least for the operations executives. So if we talk in terms of stewardship, taking care of
4 what you have, being efficient, not wasting, doing the right thing, that was a lot more effective.
5 And I talked not very directly about climate change in that way just because of the culture of
6
business.
7
8
9 Overall, she focused on networking within the host organization to get data and information that
10
11 helped her shape different approaches to addressing challenges and expanding her work.
12
13 Pathway 5 was similar to 4 because of the moderate ambiguity experienced by fellows.
14
15
16
Even though the hosts in this configuration had significant preparation and engagement with
17
18 Climate Corps, the project proposals themselves were ambiguous. One fellow commented that he
19
20 “described these challenges [with the focus of his project proposal] to my engagement manager
21
22
and Climate Corps alumni” who all said “this is not uncommon.” Facing challenges with
23
24
25 defining projects in the context of prior initiatives, the fellows in this pathway also emphasized
26
27 internal advice – “continuous interaction with my supervisor” and “turning to my colleagues [at
28
29 the host]” – to help shape their projects and decide which solutions to prioritize and add to their
30
31
32 project scope. One of the fellows directly noted how a former fellow, now employed full time at
33
34 the host, provided support: “One of the three people I reported to was a former EDF Climate
35
36 Corps fellow, and throughout the fellowship was a great resource for both work plan guidance,
37
38
39
as well as strategy advice.” The role of small organizational scale is reflected in the Fellows’
40
41 comments that they knew where within the organization to go for data or advice. One
42
43 appreciated that he is “understanding the role of different verticals and getting data from various
44
45
departments.”
46
47
48 In many ways, the moderate ambiguity context of these two pathways represents the
49
50 typical conditions in issue selling research. Issue selling efforts encounter a mix of supports from
51
52 allies, aligned projects, executive endorsements, etc., as well as ambiguous terrain from newly
53
54
55 developed roles, limited prior issue development, and integration (e.g., Bansal, 2003; Howard-
56
57
58
59
60
Page 39 of 63 Academy of Management Journal

1
2
3 Grenville, 2007; Wickert & de Bakker, 2018). In this context, prior research suggests that
4
5
6 successful issue selling will focus on tactics of navigating the organizational polity, implicitly
7
8 assuming that the “moves that matter” play out within the organization (Dutton et al., 2001). Our
9
10 data enable us to situate this dynamic as particular to organizational contexts with moderate
11
12
13
ambiguity.
14
15 High Ambiguity: Building Legitimacy
16
17 The last path stands alone as a possible third group with extensive ambiguity. Path 6
18
19 includes large-scale organizations that were program newcomers and, as a peripheral condition,
20
21 had open-ended projects. Although large scale may imply greater guidance, organizational scale
22
23
can also make internal communication and understanding more fragmented (Baumann-Pauly et
24
25
26 al., 2013). Where the low ambiguity group features unmitigated conditions that increase issue
27
28 elaboration, Path 6 occupies the opposite end of the spectrum with contextual conditions that all
29
30 enhance ambiguity. Remarkably, the configuration in this group also includes the same direction
31
32
33 of resource use as the low ambiguity group. In Path 6, fellows engage in extensive external
34
35 search. At the same time, the absence of an external focus is a peripheral condition in this path.
36
37 Confronted with little support and direction, external search again becomes a key input to
38
39
40 solution creation. However, rather than serving to catalyze an already-elaborated issue, here the
41
42 practice provides missing guidance and expertise. The fellow, Steven, illustrates this dynamic.
43
44 Steven worked at a national health care provider that tasked him with developing a sustainability
45
46
strategy. He expanded the project with six new solutions including lighting upgrades, a service
47
48
49 procedure redesign for energy efficiency, and employee engagement programs. There were no
50
51 prior fellows and high project ambiguity in the initial proposal. Steven recognized the extensive
52
53 ambiguity of the fellowship, as he reflected below:
54
55
56
57
58
59
60
 Academy of Management Journal Page 40 of 63

1
2
3 I think going into the fellowship, we really didn’t know what my role was going to be...I had the
4 description from EDF and what the [host] had provided but besides that you can go to the [host’s]
5 website and you can’t even find the word sustainability. It was really kind of ambiguous walking
6
into the situation...I knew a little bit of what to expect but really not the full spectrum. And so we
7
knew that I was going to start at least with a macro approach and help them organize really what
8
9
they had done, the progress they had made, and then start identifying areas of opportunity...I think
10 there was a lot of skepticism with me coming into the [organization] in the first place simply
11 because they just hadn’t had anyone with such a specific background in sustainability or energy
12 coming into the organization.
13
14 Within this path, fellows often experienced the large organizational scale as a source of
15
16 greater ambiguity. Similar to Path 4, fellows here referred to the difficulty of figuring out how to
17
18
19
navigate larger organizations and understand who key stakeholders were. Steven noted this: “The
20
21 biggest challenge was navigating an organization of this scale in such a short period of time.
22
23 There were so many stakeholders for every project I wanted to implement, and it required
24
25 numerous meetings to obtain approval from all groups to proceed with projects.”
26
27
28 Steven mobilized a number of external resources to help navigate the ambiguity and
29
30 move forward in addressing the sustainability challenges. He exhibited similar approaches to
31
32 external search as others in the pathway, using EDF materials to learn about different tools and
33
34
35 experts that could be utilized, and connecting to others for social support. The following quotes
36
37 provide examples of each:
38
39 [expertise and tools] The webinars were really helpful. I really didn’t know too much
40 about the lighting as a service opportunity and through the webinar I was able to connect
41
42
with a lighting vendor who really helped me explain the concept to the [host] to see if it
43 were something they would be interested in pursuing.
44
45 [exemplars] I really enjoyed reading some of the past sample reports, final reports, that
46 [EDF] provided. That type of information really helped me see how other fellows had
47 framed their findings and their thoughts.
48
49
50
[social support] The network was also really helpful just from sort of a “I’m out here on
51 an island. How are things going over in your neck of the woods? Are you experiencing
52 similar pushbacks, trials, and tribulations that I’m experiencing, and how did you push
53 forward?”
54
55
56
57
58
59
60
Page 41 of 63 Academy of Management Journal

1
2
3 Steven used these external resources to support solution creation throughout his fellowship.
4
5
6 Movement resources helped him set expectations for himself. For example, he noted that the
7
8 exemplars of previous work that EDF made available helped him decide what areas to focus on,
9
10 what areas may have more opportunity than he realized, and where he might have done as much
11
12
13
as he could: “I could utilize [this information] to say okay I feel like maybe I’m tapped out in
14
15 one area, what’s another thing for me to look at.”
16
17 Importantly, he also used the external knowledge and experts to build legitimacy for his
18
19
recommendations within the host organization. The following quote shows one example of this
20
21
22 type of external resource use:
23
24 At the beginning of my fellowship, the organization was skeptical of [LED lighting’s]
25 price and sophistication. As such, I reached out to several lighting vendors and had them
26 bring representatives from lighting manufacturers to present and demo their LED lighting
27
28
products. My host organization is currently piloting the technology in several fixtures
29 throughout campus.
30
31 This mode of external resource use – to legitimize new ideas within the organization – is distinct
32
33 to conditions of high ambiguity, as we did not observe it in conditions of low ambiguity. After
34
35
36
bringing in external resources to show the host the feasibility of recommendations, Steven then
37
38 focused internally – navigating the funding and approval processes within his host. He shared
39
40 one example below:
41
42
I set up a call [with a high status industry peer from another geography] and had my
43
44 manager, the unit head, and his assistant sit on that call. We had an hour and half discussion
45 about lighting retrofits and operating room strategies and programs that he had implemented
46 and how he had achieved success. And once we had that call, I saw something change in my
47 manager and he really got excited about it. It was after that that he really became supportive
48 and we started talking about okay where are some realistic opportunities that you want to
49 tackle given the funding we have available and the process for getting projects pushed
50
51
through [the host organization].
52
53
54
55
56
57
58
59
60
 Academy of Management Journal Page 42 of 63

1
2
3 Comparing the Use of Movement Resources across Low and High Ambiguity Groups
4
5 All the Climate Corps fellows had access to the same portfolio of potential external
6
7 resources – other fellows, EDF analysis tools, experts, webinars, EDF engagement managers,
8
9
10 and the broader EDF network. Yet, fellows drew on those resources differently. Moreover, how
11
12 fellows put resources into use to contribute to solution creation varied across contextual
13
14 conditions. Table 4 summarizes the different strategies for utilizing movement resources in
15
16
17 organizational contexts with low versus high ambiguity, with representative examples.
18
19 -------------------------------
20
21 Insert Table 4 about here
22
23 -------------------------------
24
25 Nuanced approaches to resource use emerge through comparisons across the two
26
27 pathway groups. Fellows in both low and high ambiguity contexts used the external EDF
28
29
30 network for social support, leveraging movement resources to find approaches past roadblocks
31
32 and to receive feedback on new strategies. Notably, this pattern is distinct from fellows under
33
34 conditions of moderate ambiguity, who were more likely to discuss the EDF network in ways
35
36
37 that were disconnected from their fellowships, most often as a job search resource rather than as
38
39 a resource for their project work.
40
41 Moreover, fellows used the same type of external resources differently based on the
42
43
44
context. Fellows in contexts of low or high ambiguity integrated knowledge and tools from
45
46 external networks in order to highlight new opportunities, translating what was available
47
48 externally for their host organizations. For fellows in low ambiguity contexts, this resource move
49
50 was sufficient to help catalyze new actions. With significant experience around climate, energy,
51
52
53 and sustainability, these organizations already had the necessary skills internally to move
54
55 forward. On the other hand, for fellows in high ambiguity contexts, the hosts had a limited
56
57
58
59
60
Page 43 of 63 Academy of Management Journal

1
2
3 history of engagement and gaps in internal skills for addressing the issue. Fellows needed to also
4
5
6 develop external resources to fill this skills gap, enabling their organizations to engage more.
7
8 Another difference emerged as fellows were adapting exemplars from other organizations
9
10 to their hosts. In low ambiguity contexts, fellows mobilized exemplars to reinvigorate existing
11
12
13
programs, as they used templates or comparison cases to amplify programs already in place and
14
15 catalyze more robust action. In high ambiguity contexts, however, fellows used exemplars to
16
17 show the feasibility of new climate or energy ideas, selecting comparisons that would help them
18
19
legitimate their proposed solutions and build support for ideas that had not been considered
20
21
22 previously. These comparisons provide evidence that it is not simply whether fellows emphasize
23
24 movement resources, or which potential resources are incorporated, but how fellows enact
25
26 resources in alignment with contextual ambiguity that leads to solution creation.
27
28
29 Low Solution Creation: Exploring the Inverse Outcome
30 A strength of QCA is that it supports investigation of asymmetric causality between an
31
32
33 outcome and its inverse. However, using the standard QCA procedures, there are no consistent
34
35 pathways for the inverse of our main outcome, low solution creation. As Fiss comments on a
36
37 parallel null QCA finding: “there are many ways to be nonperforming here, but no consistent
38
39
40 pattern” (2011: 410). We then reanalyzed the data with “the minimum recommended threshold”
41
42 (ibid: 403) of 0.75 consistency, which produced three consistent paths that include nine cases out
43
44 of seventeen in positive rows in the truth table. Table 5 reports the results, and we add depth to
45
46
our interpretation with data from short answer survey responses from each case as well as one
47
48
49 interview in Path B. Although we have lower confidence in these findings, they provide
50
51 suggestive evidence.
52
53 The three paths include each of our ambiguity groupings: low in Path A with experienced
54
55
56 hosts and strongly-defined project plans, moderate in Path B with these two conditions but also
57
58
59
60
 Academy of Management Journal Page 44 of 63

1
2
3 new initiatives, and high in Path C with new initiatives, program newcomers (peripheral
4
5
6 condition), and open-ended projects. Organizational scale is also a part of each solution, with
7
8 large organizations in Paths A and B (peripheral condition in B) and small organizations in Path
9
10 C. The most striking result is that all three paths include the absence of external resources,
11
12
13
elevating the importance of these conditions for the outcomes of embedded activism. As Shreya
14
15 from Path A noted, “The (host) team and their experience is the greatest resource for me.”
16
17 Another fellow in this pathway, Ashley, commented on her lack of connection to the EDF
18
19
network, “It doesn’t feel like a very connected network. I met people for a week and now they
20
21
22 are scattered all over the place. I don’t feel connected to them.”13
23
24 Paths A and C combine these resource conditions with contexts that decrease and
25
26 increase ambiguity respectively. These paths mirror our main results where the presence of
27
28
29 external resources generated high solution creation. Interestingly, Path B features moderate
30
31 ambiguity, suggesting that large organizations (as a peripheral condition) with prior fellows but
32
33 new initiatives limit solution creation, regardless of movement resources. This may be in part
34
35
36
because of challenges for the fellow in navigating organizations in this situation, as evident in
37
38 Lisa’s experience. She reflected, “I’ve felt frustrated and discouraged because my project is
39
40 strategy oriented and I’ve felt that my direct report as well as my peers are not – they are instead
41
42
highly focused on products. It has felt like a mismatch at times and I often feel that I am
43
44
45 speaking a different language and coming to energy management from an opposite approach of
46
47
48
49
50
51
52 13 The different levels of engagement with the EDF network largely reflect individual differences, as all
53 fellows attended the same training together and were encouraged to engage with smaller networks by
54 region and project type. There are null differences in the two movement resources by gender and race (T
55 tests, race coded as white vs. other), as well as by school, host organization, industry, city, and region
56 (correlation p values, using counts for current year and all previous years).
57
58
59
60
Page 45 of 63 Academy of Management Journal

1
2
3 those that I am working alongside and learning from.” While the significance of these results is
4
5
6 borderline, they reinforce the importance of movement resources.
7
8 -------------------------------
9
10 Insert Table 5 about here
11
12 -------------------------------
13
14 DISCUSSION
15
16 To integrate our empirical findings at a more abstract level, we present a model in Figure
17
18 2 of how the opportunity structure for collaborative activism combines with embedded activist
19
20
21
resource strategies to advance reforms in organizations. When the opportunity structure of an
22
23 organization has low ambiguity, embedded activists face a situation of significant prior issue
24
25 development and understanding. Through importing external resources, these activists are able to
26
27 catalyze established efforts with the introduction of new potential solutions. As ambiguity
28
29
30 increases to a moderate level, embedded activists turn internally, selling the issue further within
31
32 the organization, and expanding the potential solutions considered. When ambiguity is high,
33
34 embedded activists combine external search for knowledge and exemplars with an internal focus
35
36
37 on legitimizing the efforts to prompt consideration of additional solutions within the
38
39 organization. By matching opportunity structure with resource strategies, embedded activists can
40
41 succeed in advancing social movement goals across diverse organizations.
42
43
44 -------------------------------
45
46 Insert Figure 2 about here
47
48 -------------------------------
49
50
Social Movements and Organizations
51
52 While the focus of research on social movements and organizations has been on
53
54
55
contentious tactics, corporations increasingly espouse progressive goals and SMOs have shifted
56
57
58
59
60
 Academy of Management Journal Page 46 of 63

1
2
3 towards collaboration (Odziemkowska, 2020). In a comprehensive study of interactions between
4
5
6 large environmental SMOs and Fortune 500 firms from 2007-2012, there were about twice as
7
8 many collaborative interactions as contentious ones (McDonnell et al., 2021). Further, research
9
10 that does consider SMO-firm collaboration tends to focus on external-facing efforts such as
11
12
13
certifications and other forms of multi-sectoral partnerships (de Bakker et al., 2019; Gray &
14
15 Purdy, 2018). Our analysis of EDF’s Climate Corps program sheds light on how SMOs use
16
17 collaboration to drive internal change at partner organizations. This study also advances an
18
19
emerging trend in studies of insider activism to highlight the rich connections between insiders
20
21
22 and external social movements (Buchter, 2020; DeJordy et al., 2020). Insider activists are not
23
24 always self-organizing with passive movement influence, but they may also be directly and
25
26 actively connected to external SMOs.
27
28
29 Our investigation into how collaborative SMOs coordinate change efforts inside of
30
31 partner organizations leads to several theoretical innovations. First, we introduce the concept of
32
33 embedded activism to characterize activists that are neither insiders or outsiders, but cross
34
35
36
movement-partner boundaries (cf., DeJordy et al., 2020). This role provides a mix of the
37
38 complementary strengths and weaknesses of insider and outsider activists (Briscoe & Gupta,
39
40 2016), as embedded activists combine substantial knowledge of organizational opportunity
41
42
structures with access to a wider range of movement resources. We integrate research on issue
43
44
45 selling and proactivity more broadly to characterize how embedded activism works (Dutton et
46
47 al., 2001; Grant & Ashford, 2008). These literatures provide insight into persuasion-based tactics
48
49 in varying organizational contexts. In particular, we highlight a key tension between
50
51
52 organizational support, which provides activists with resources to act, and issue ambiguity,
53
54 which provides opportunities to act. Embedded activists are able to draw on movement resources
55
56
57
58
59
60
Page 47 of 63 Academy of Management Journal

1
2
3 to manage this tension. We find that successful activists match these external resources to their
4
5
6 organizational contexts in order to build a fertile mixture of support and ambiguity.
7
8 Second, we develop a revised theory of opportunity structures to fit with the collaborative
9
10 tactics we observe. The prevailing theory is based on contentious tactics and holds that
11
12
13
opportunities for activist influence are greater for firms that are more vulnerable to disruption,
14
15 due to factors like prominence and reliance on consumer reputation (Bartley & Child, 2014;
16
17 King, 2008). As collaborative activism relies on persuasion rather than disruption (den Hond &
18
19
de Bakker, 2007), we integrate issue selling research to theorize the organization’s existing issue
20
21
22 understandings as the key factor shaping activist opportunities. In order to persuade partner
23
24 organizations, activists must engage with their current frameworks for interpreting and managing
25
26 an issue (Howard-Grenville, 2007). Our research demonstrates how the organization’s history of
27
28
29 engagement with the issue, accumulation of issue-specific resources, and organizational scale all
30
31 affect the development and distribution of these understandings, and ultimately the terrain for
32
33 collaborative activism. Organizations with extensive prior issue development offer many
34
35
36
resources for activists to engage with, but little space to deviate from established routines and
37
38 issue understandings; alternatively, organizations that are new to an issue present activists with a
39
40 clean slate, but without a foundation of established issue understandings and legitimacy.
41
42
Consistent with the criticisms of political opportunity structure theory for being overly structural
43
44
45 (Goodwin & Jasper, 2004), this theory of collaborative opportunity structures suggests a
46
47 landscape that activists must interpret and navigate, rather than a direct and linear influence on
48
49 activist outcomes.
50
51
52 Proactivity and Issue Selling
53 Our research also contributes to the literature on proactivity (Grant & Ashford, 2008;
54
55
56 Parker & Bindl, 2017), especially concerning the role of ambiguity and the related construct of
57
58
59
60
 Academy of Management Journal Page 48 of 63

1
2
3 uncertainty. Grant & Ashford (2008) theorize that ambiguity and uncertainty motivate proactive
4
5
6 behavior in order to obtain greater clarity. However, if uncertainty also stimulates image
7
8 concerns, as when employees are embarrassed to not understand a work task, then proactivity
9
10 may decline (Lim et al., 2020). More recently, Griffin and Grote (2020) attack the idea that
11
12
13
individuals and organizations seek to eliminate uncertainty. They argue that uncertainty is also
14
15 desirable because it nurtures curiosity and innovation. Our findings support Griffin & Grote’s
16
17 positive interpretation, and also suggest a way to reconcile inconsistent results: proactivity
18
19
benefits from a balance of support and ambiguity, which together provide resources and
20
21
22 opportunities for creative action. While an argument for balance may seem trivial, it complicates
23
24 prevailing understandings of a linear relationship between organizational support and proactivity
25
26 (Ashford et al., 1998; Morrison, 2011). The more an organization supports an issue, the more
27
28
29 routinized its frameworks for understanding the issue, and the less ambiguity and opportunities
30
31 for creative action (cf., Risi & Wickert, 2017; Sandhu & Kulik, 2018).
32
33 We further advance the literature on issue selling by investigating the influences of
34
35
36
organizational support and external resources. Scholars have long recommended that
37
38 organizations should do more to bolster issue selling (Ashford et al., 1998; Dutton et al., 2001),
39
40 but how this actually works in practice remains unknown. Our unique data in terms of a network
41
42
of organizations working to accelerate an issue’s development, as well as the variety of
43
44
45 organizations involved, enable us to shed light on this topic. Although prior research strongly
46
47 suggests that issue selling should thrive in organizations with greater support (Ashford et al.,
48
49 1998; Dutton et al., 2001; Bansal, 2003), our findings point to a more nuanced tension between
50
51
52 support and ambiguity (Sonenshein, 2014, 2016; Gioia et al., 2012), which external resources
53
54 can help manage. Similar to previous work on issue selling (Bansal, 2003; Howard-Grenville,
55
56
57
58
59
60
Page 49 of 63 Academy of Management Journal

1
2
3 2007), in organizational contexts with moderate ambiguity, we see that internal moves matter.
4
5
6 However, in more extreme cases of ambiguity – low or high – crossing organizational
7
8 boundaries to integrate external resources becomes important. In conditions featuring low
9
10 ambiguity, the main challenge is how to work with the previously built-up issue understandings
11
12
13
at the organization. External resources help move the issue from “it’s already set” to “here are
14
15 new possibilities.” In conditions featuring high ambiguity, excessive uncertainty threatens to
16
17 prevent issue development. In this context, external resources help to “unfreeze” and narrow the
18
19
cognitive load by tightening the issue space and building legitimacy and capacity for change. As
20
21
22 a result, issue selling initiatives are able to move from “there’s so much to do, where do I even
23
24 start” to action. Although previously little studied, the use of external resources is likely
25
26 important because issue selling, as an upwards influence process, implies that organizations lack
27
28
29 issue expertise and other assets (Howard-Grenville, 2007), which external supporters may be
30
31 well positioned to provide. Overall, we suggest that the repertoire of successful moves varies
32
33 with the organizational context, rather than being “one size fits all.”
34
35
36
Practical Implications for Climate Change
37 In addition, this work adds to the growing body of research around corporate
38
39
40 sustainability and climate change. We provide practical implications for organizational leaders
41
42 working to address climate change. Previous research on social issue selling, like climate change,
43
44 highlights the importance of external identities for motivating employees to promote change
45
46
(Bansal, 2003; Meyerson & Scully, 1995; Ong & Ashford, 2016). We bring attention to the
47
48
49 importance of these external ties not just for motivation, but also for knowledge, tools, and
50
51 expertise that can be used to catalyze change in specific organizational contexts.
52
53 For practitioners, this highlights the importance of collaboration with SMOs, professional
54
55
56 associations, and alumni networks to empower change agents within organizations to act on
57
58
59
60
 Academy of Management Journal Page 50 of 63

1
2
3 climate change. These connections should provide the social support of peers facing similar
4
5
6 challenges, as well as opportunities to learn approaches for overcoming these difficulties. They
7
8 should also convey specific advice, mentoring, and technical support that fit the organizational
9
10 context in which the climate change advocate is situated. Given that the grand challenge of
11
12
13
climate change requires significant expertise and interdisciplinary knowledge, and is
14
15 continuously evolving, the spread of these skill-based external resources may be critical for
16
17 amplifying efforts. Our research suggests that leaders within organizations should develop
18
19
connections externally to make these potential resources available for internal issue champions,
20
21
22 particularly when their organizations are either new to the issue or entrenched in an established
23
24 approach. It also suggests that external issue advocates that connect with climate change activists
25
26 inside of organizations should strategically develop programs for providing these diverse types
27
28
29 of resources.
30
31 Limitations and Future Directions
32
33 The EDF Climate Corps program is a prominent example of how SMOs can use
34
35 collaborative tactics to advance internal reforms at partner organizations through embedded
36
37 activism. While collaboration takes many forms, our findings should generalize to initiatives that
38
39
40 cross movement-partner boundaries and seek internal changes, as opposed to external-facing
41
42 projects. Although Climate Corps is remarkable in its scale and sophistication, we believe that it
43
44 is not unusual for SMOs and NGOs more broadly to seek to empower change agents in partner
45
46
organizations (e.g., Human Rights Campaign, Aspen Institute, LeanIn.org, World Business
47
48
49 Alliance for Global Health, Green Chemistry conferences, Integrative Medicine associations, and
50
51 other professional association and union training programs; cf., DeJordy et al., 2020; Heinze &
52
53 Weber, 2016; Howard-Grenville, Nelson, Earle, Haack, & Young, 2017; Mayer, Ong,
54
55
56 Sonenshein, & Ashford, 2019). There should also be considerable demand among internal
57
58
59
60
Page 51 of 63 Academy of Management Journal

1
2
3 activists for external support (Buchter, 2020; DeJordy et al., 2020), which is in line with findings
4
5
6 on external information-seeking in issue selling initiatives (Andersson & Bateman, 2000). At the
7
8 same time, organizations will likely be reticent to admit activists, constraining the development
9
10 of embedded activism. The boundary work that enables this form of activism merits greater
11
12
13
study. For example, it may help to use intermediary parties that can act as brokers, as in Climate
14
15 Corp’s recruitment of graduate students (cf., Kaplan, Milde, & Cowan, 2016; Kellogg, 2014).
16
17 As an initial study of how embedded activism operates, we took an exploratory approach
18
19
in this paper. Future studies could apply other methods to provide further evidence for or against
20
21
22 the framework we develop. For example, field experiments could randomize the use of external
23
24 resources. Our expectations are that external resources would be more effective in driving
25
26 solution creation when there is high or low ambiguity about how to act on the issue, and that
27
28
29 external resources would have less of an effect when there is a moderate level of ambiguity.
30
31 Although QCA is consistent with our expectations about equifinality and configurational
32
33 causality, it also puts tight limits on the number of conditions that can be considered. There are
34
35
36
likely additional features of the organizational context, such as goal alignment, openness to
37
38 change, and progressive values (e.g., belief in anthropogenic climate change) that would affect
39
40 the dynamics. Such analyses could help clarify the role of organizational scale, which is
41
42
consistently influential across paths but in different ways.
43
44
45 More broadly, there are many unanswered questions about how collaborative activism
46
47 works, which are growing in importance as SMOs increasingly adopt such tactics (McDonnell et
48
49 al., 2021; Odziemkowska, 2020). We offer one model for how such activism works across
50
51
52 varying partner organizations, generating a revised theory of opportunity structures for
53
54 collaborative influence. Future research should chart the wider variety of collaborative tactics
55
56
57
58
59
60
 Academy of Management Journal Page 52 of 63

1
2
3 and investigate further how heterogeneity in partner organizations influences the success of these
4
5
6 tactics.
7
8 This study also points to opportunities for issue selling research. Variation in the impact
9
10 of issue selling moves across organizations is an open area in this literature as well. Our findings
11
12
13
about the contribution of external resources also indicate that more research should examine the
14
15 permeability of organizational boundaries and how cross-boundary processes affect issue selling
16
17 (Lauche, 2019). We also identify a nuanced role of organizational support in terms of positive
18
19
issue understandings, which suggests that further research is needed to understand how
20
21
22 organizations can best empower issue selling. In addition, following our recognition of issue
23
24 selling practices in Climate Corps, future work could explore under what conditions temporary
25
26 organizational members, like embedded activists in our case or even consultants and other
27
28
29 professionals, take on roles similar to issue sellers.
30
31 More broadly, issue selling research could consider the varying stages and dimensions of
32
33 reform. Our focus on the creation of new solutions matches our theoretical interest in the
34
35
36
fellows’ proactivity as well as data availability, but it would also be valuable to investigate the
37
38 links between reform ideas and implementation (cf., Buchter, 2020). As a comparison, social
39
40 movement scholars have found that activists pressuring states have a greater influence on agenda
41
42
setting than on legislative passage (Amenta et al., 2010). In addition, this research opens up
43
44
45 questions about how to compare reforms: while we focus on scope, the depth of reforms in terms
46
47 of resources committed and impacts realized should also be considered. Likewise, social
48
49 movement scholars caution that success can reflect modest ambitions, and that consequences
50
51
52 continue to unfold in unintended and indirect ways (Tilly, 1999). The “finality” of issue selling is
53
54 itself at best ambiguous as issue sellers continue to push further reforms in an ongoing process.
55
56
57
58
59
60
Page 53 of 63 Academy of Management Journal

1
2
3 Greater attention to such nuances in conceptualizing and measuring outcomes would advance the
4
5
6 issue selling literature.
7
8 CONCLUSION
9
10 Overall, this research illuminates how collaborative social movement organizations can
11
12 propel changes inside of partner organizations by embedding activists who work to catalyze
13
14 internal reforms. We highlight the joint importance of organizational contexts and external
15
16
17 resources for shaping organizational change efforts (Weber & Waeger, 2017). The picture that
18
19 emerges is one of cultivated and resourceful agency spanning movement-partner boundaries,
20
21 rather than isolated campaigners, constrained by a lack of organizational support, or contentious
22
23
outsiders. Our empirical context is particularly relevant for this topic. Mitigating climate change
24
25
26 requires organizations to effectively incorporate diffuse external concerns into significant
27
28 reforms. Supporting and empowering embedded activists to champion the cause within
29
30 organizations appears promising and deserves greater support and study.
31
32
33
34
35 REFERENCES
36
37 Alt, E., & Craig, J. B. 2016. Selling Issues with Solutions: Igniting Social Intrapreneurship in
38 for-Profit Organizations. Journal of Management Studies, 53(5): 794–820.
39 Amenta, E., Caren, N., Chiarello, E., & Su, Y. 2010. The Political Consequences of Social
40 Movements. Annual Review of Sociology, 36(1): 287–307.
41 Andersson, L. M., & Bateman, T. S. 2000. Individual Environmental Initiative: Championing
42
Natural Environmental Issues in U.S. Business Organizations. Academy of Management
43
44 Journal, 43(4): 548–570.
45 Ansari, S., Gray, B., & Wijen, F. (2011). Fiddling while the ice melts? How organizational
46 scholars can take a more active role in the climate change debate. Strategic
47 Organization, 9(1), 70-76.
48 Ashford, S. J., Rothbard, N. P., Piderit, S. K., & Dutton, J. E. 1998. Out on a Limb: The Role of
49 Context and Impression Management in Selling Gender-Equity Issues. Administrative
50
51
Science Quarterly, 43(1): 23–57.
52 Bansal, P. 2003. From Issues to Actions: The Importance of Individual Concerns and
53 Organizational Values in Responding to Natural Environmental Issues. Organization
54 Science, 14(5): 510–527.
55
56
57
58
59
60
 Academy of Management Journal Page 54 of 63

1
2
3 Baron, D. P., & Diermeier, D. 2007. Strategic Activism and Nonmarket Strategy. Journal of
4
5
Economics & Management Strategy, 16(3): 599–634.
6 Bartley, T., & Child, C. 2014. Shaming the Corporation: The Social Production of Targets and
7 the Anti-Sweatshop Movement. American Sociological Review, 79(4): 653–679.
8 Baumann-Pauly, D., Wickert, C., Spence, L. J., & Scherer, A. G. 2013. Organizing Corporate
9 Social Responsibility in Small and Large Firms: Size Matters. Journal of Business Ethics,
10 115(4): 693–705.
11
Bindl, U. K., & Parker, S. K. 2017. New Perspectives and Directions for Understanding
12
13
Proactivity in Organizations. In S. K. Parker & U. K. Bindl (Eds.), Proactivity at Work:
14 Making Things Happen in Organizations: 559–584. New York: Routledge.
15 Briscoe, F., Chin, M. K., & Hambrick, D. C. 2014. CEO Ideology as an Element of the
16 Corporate Opportunity Structure for Social Activists. Academy of Management Journal,
17 57(6): 1786–1809.
18 Briscoe, F., & Gupta, A. 2016. Social Activism in and Around Organizations. The Academy of
19
Management Annals, 10(1): 671–727.
20
21 Cohen, M. D., March, J. G., & Olsen, J. P. 1972. A Garbage Can Model of Organizational
22 Choice. Administrative Science Quarterly, 17(1): 1–25.
23 Crilly, D., Zollo, M., & Hansen, M. T. 2012. Faking It or Muddling Through? Understanding
24 Decoupling in Response to Stakeholder Pressures. Academy of Management Journal,
25 55(6): 1429–1448.
26 Darnall, N., Henriques, I., & Sadorsky, P. 2010. Adopting Proactive Environmental Strategy:
27
28
The Influence of Stakeholders and Firm Size. Journal of Management Studies, 47(6):
29 1072–1094.
30 Dauvergne, P., & Lister, J. 2013. Eco-Business: A Big-Brand Takeover of Sustainability.
31 Cambridge, Mass.: MIT Press.
32 de Bakker, F. G. A., Rasche, A., & Ponte, S. 2019. Multi-Stakeholder Initiatives on
33 Sustainability: A Cross-Disciplinary Review and Research Agenda for Business Ethics.
34
Business Ethics Quarterly, 29(3): 343–383.
35
36
DeCelles, K. A., Sonenshein, S., & King, B. G. 2019. Examining Anger’s Immobilizing Effect
37 on Institutional Insiders’ Action Intentions in Social Movements. Administrative Science
38 Quarterly, 0001839219879646.
39 DeJordy, R., Scully, M., Ventresca, M. J., & Creed, W. E. D. 2020. Inhabited Ecosystems:
40 Propelling Transformative Social Change Between and Through Organizations.
41 Administrative Science Quarterly, 0001839219899613.
42
Deken, F., Berends, H., Gemser, G., & Lauche, K. 2018. Strategizing and the Initiation of
43
44 Interorganizational Collaboration through Prospective Resourcing. Academy of
45 Management Journal, 61(5): 1920–1950.
46 Delaney, L. 2017. Embedding Environmental Advocates: EDF Climate Corps. Managing for
47 Social Impact Innovations in Responsible Enterprise: 129–143. Springer.
48 Dobbin, F. 2009. Inventing Equal Opportunity. Princeton, N.J.: Princeton University Press.
49 Dowell, G. W. S., & Muthulingam, S. 2017. Will firms go green if it pays? The impact of
50
51
disruption, cost, and external factors on the adoption of environmental initiatives. Strategic
52 Management Journal, 38(6): 1287–1304.
53 Dusa, A. 2017. QCA.
54
55
56
57
58
59
60
Page 55 of 63 Academy of Management Journal

1
2
3 Dutton, J. E. 1988. Understanding Strategic Agenda Building and Its Implications for Managing
4
5
Change. In L. R. Pondy, R. J. Boland Jr., & H. Thomas (Eds.), Managing Ambiguity and
6 Change: 127–144. New York: Wiley.
7 Dutton, J. E., & Ashford, S. J. 1993. Selling Issues to Top Management. The Academy of
8 Management Review, 18(3): 397–428.
9 Dutton, J. E., Ashford, S. J., O’Neill, R. M., & Lawrence, K. A. 2001. Moves that Matter: Issue
10 Selling and Organizational Change. Academy of Management Journal, 44(4): 716–736.
11
Fiss, P. C. 2011. Building Better Causal Theories: A Fuzzy Set Approach to Typologies in
12
13
Organization Research. Academy of Management Journal, 54(2): 393–420.
14 Gehman, J., Treviño, L. K., & Garud, R. 2013. Values Work: A Process Study of the Emergence
15 and Performance of Organizational Values Practices. Academy of Management Journal,
16 56(1): 84–112.
17 Georgallis, P. 2017. The Link Between Social Movements and Corporate Social Initiatives:
18 Toward a Multi-level Theory. Journal of Business Ethics, 142(4): 735–751.
19
Gioia, D. A., Nag, R., & Corley, K. G. 2012. Visionary Ambiguity and Strategic Change: The
20
21 Virtue of Vagueness in Launching Major Organizational Change. Journal of Management
22 Inquiry, 21(4): 364–375.
23 Goodwin, J., & Jasper, J. M. 2004. Caught in a winding, snarling vine: The structural bias of
24 political process theory. In J. Goodwin & J. M. Jasper (Eds.), Rethinking social movements:
25 Structure, meaning, and emotion: 3–30. Lanham: Rowman & Littlefield.
26 Grant, A. M., & Ashford, S. J. 2008. The dynamics of proactivity at work. Research in
27
28
Organizational Behavior, 28: 3–34.
29 Greckhamer, T., Furnari, S., Fiss, P. C., & Aguilera, R. V. 2018. Studying configurations with
30 qualitative comparative analysis: Best practices in strategy and organization research.
31 Strategic Organization, 16(4): 482–495.
32 Griffin, M. A., & Grote, G. 2020. When Is More Uncertainty Better? A Model of Uncertainty
33 Regulation and Effectiveness. Academy of Management Review, 45: 745–765.
34
Hannan, M. T., & Freeman, J. 1984. Structural Inertia and Organizational Change. American
35
36
Sociological Review, 49(2): 149–164.
37 Heinze, K. L., & Weber, K. 2016. Toward Organizational Pluralism: Institutional
38 Intrapreneurship in Integrative Medicine. Organization Science, 27(1): 157–172.
39 Hobfoll, S. E., Halbesleben, J., Neveu, J.-P., & Westman, M. 2018. Conservation of Resources in
40 the Organizational Context: The Reality of Resources and Their Consequences. Annual
41 Review of Organizational Psychology and Organizational Behavior, 5(1): 103–128.
42
Howard-Grenville, J. A. 2007. Developing issue-selling effectiveness over time: Issue selling as
43
44 resourcing. Organization Science, 18(4): 560–577.
45 Howard-Grenville, J., Buckle, S. J., Hoskins, B. J., & George, G. 2014. Climate Change and
46 Management. Academy of Management Journal, 57(3): 615–623.
47 Howard-Grenville, J., Nelson, A. J., Earle, A. G., Haack, J. A., & Young, D. M. 2017. “If
48 Chemists Don’t Do It, Who Is Going To?” Peer-driven Occupational Change and the
49 Emergence of Green Chemistry. Administrative Science Quarterly, 62(3): 524–560.
50
51
Kannan-Narasimhan, R. (Priya), & Lawrence, B. S. 2018. How innovators reframe resources in
52 the strategy-making process to gain innovation adoption. Strategic Management Journal,
53 39(3): 720–758.
54
55
56
57
58
59
60
 Academy of Management Journal Page 56 of 63

1
2
3 Kaplan, S., Milde, J., & Cowan, R. S. 2016. Symbiont Practices in Boundary Spanning: Bridging
4
5
the Cognitive and Political Divides in Interdisciplinary Research. Academy of Management
6 Journal, 60(4): 1387–1414.
7 Kellogg, K. C. 2009. Operating Room: Relational Spaces and Microinstitutional Change in
8 Surgery. American Journal of Sociology, 115(3): 657–711.
9 Kellogg, K. C. 2011. Hot Lights and Cold Steel: Cultural and Political Toolkits for Practice
10 Change in Surgery. Organization Science, 22(2): 482–502.
11
Kellogg, K. C. 2014. Brokerage Professions and Implementing Reform in an Age of Experts.
12
13
American Sociological Review, 79(5): 912–941.
14 King, B. G. 2008. A Political Mediation Model of Corporate Response to Social Movement
15 Activism. Administrative Science Quarterly, 53(3): 395–421.
16 Lauche, K. 2019. Insider Activists Pursuing an Agenda for Change: Selling the Need for
17 Collaboration. In J. Sydow & H. Berends (Eds.), Managing Inter-organizational
18 Collaborations: Process Views, vol. 64: 119–138. Emerald Publishing Limited.
19
Li, A. N., & Tangirala, S. 2020. How Voice Emerges and Develops in Newly Formed
20
21 Supervisor-Employee Dyads. Academy of Management Journal.
22 https://doi.org/10.5465/amj.2018.0961.
23 Lim, J. H., Tai, K., Bamberger, P. A., & Morrison, E. W. 2020. Soliciting Resources From
24 Others: An Integrative Review. Academy of Management Annals, 14: 122–159.
25 Lingo, E. L., & O’Mahony, S. 2010. Nexus Work: Brokerage on Creative Projects.
26 Administrative Science Quarterly, 55(1): 47–81.
27
28
Mayer, D. M., Ong, M., Sonenshein, S., & Ashford, S. J. 2019. The money or the morals? When
29 moral language is more effective for selling social issues. Journal of Applied Psychology,
30 104(8): 1058–1076.
31 McCarthy, J. D., & Zald, M. N. 1977. Resource mobilization and social movements: A partial
32 theory. American Journal of Sociology, 82(6), 1212-1241.
33 McDonnell, M.-H. 2016. Radical Repertoires: The Incidence and Impact of Corporate-Sponsored
34
Social Activism. Organization Science, 27(1): 53–71.
35
36
McDonnell, M.-H., & King, B. 2013. Keeping up Appearances: Reputational Threat and
37 Impression Management after Social Movement Boycotts. Administrative Science
38 Quarterly, 58(3): 387–419.
39 McDonnell, M.-H., King, B. G., & Soule, S. A. 2015. A Dynamic Process Model of Private
40 Politics: Activist Targeting and Corporate Receptivity to Social Challenges. American
41 Sociological Review, 80(3): 654–678.
42
McDonnell, M.-H., Odziemkowska, K., & Pontikes, E. 2021. Bad Company: Shifts in Social
43
44 Activists’ Tactics and Resources After Industry Crises. Organization Science.
45 https://doi.org/10.1287/orsc.2020.1410.
46 Meyerson, D. E., & Scully, M. A. 1995. Tempered Radicalism and the Politics of Ambivalence
47 and Change. Organization Science, 6(5): 585–600.
48 Miles, M. B., & Huberman, A. M. 1994. Qualitative data analysis: An expanded sourcebook.
49 Qualitative data analysis: An expanded sourcebook. Thousand Oaks: Sage Publications.
50
51
Morrison, E. W. 2011. Employee Voice Behavior: Integration and Directions for Future
52 Research. Academy of Management Annals, 5(1): 373–412.
53 Nyberg, D., & Wright, C. (2020). Climate-proofing management research. Academy of
54 Management Perspectives, 10.5465/amp.2018.0183
55
56
57
58
59
60
Page 57 of 63 Academy of Management Journal

1
2
3 Odziemkowska, K. 2020. Frenemies: When Firms and Activists Collaborate. Academy of
4
5
Management Proceedings, 2020(1): 11805.
6 Ong, M., & Ashford, S. J. 2016. Issue-selling: Proactive efforts toward organizational change. In
7 S. Parker & U. K. Bindl (Eds.), Proactivity at work: Making things happen in
8 organizations. London: Routledge.
9 Parker, S. K., & Bindl, U. K. 2017. Proactivity at Work: A Big Picture Perspective on a
10 Construct that Matters. In S. K. Parker & U. K. Bindl (Eds.), Proactivity at Work: Making
11
Things Happen in Organizations: 1–20. New York: Routledge.
12
13
Parker, S. K., Bindl, U. K., & Strauss, K. 2010. Making Things Happen: A Model of Proactive
14 Motivation. Journal of Management, 36(4): 827–856.
15 Patton, M. Q. 2015. Qualitative research & evaluation methods: Integrating theory and practice.
16 Qualitative research & evaluation methods: Integrating theory and practice. Thousand
17 Oaks: Sage Publications.
18 Peirce, C. S. 1955. Logic as Semiotic: The Theory of Signs. In J. Buchler (Ed.), Philosophical
19
Writings of Peirce: 98–119. New York: Dover Publications.
20
21 Ragin, C. C. 2008. Redesigning social inquiry: Fuzzy sets and beyond. Chicago: University of
22 Chicago Press.
23 Risi, D., & Wickert, C. 2017. Reconsidering the ‘Symmetry’ Between Institutionalization and
24 Professionalization: The Case of Corporate Social Responsibility Managers. Journal of
25 Management Studies, 54(5): 613–646.
26 Sandhu, S., & Kulik, C. T. 2018. Shaping and Being Shaped: How Organizational Structure and
27
28
Managerial Discretion Co-evolve in New Managerial Roles. Administrative Science
29 Quarterly, 64(3): 619–658.
30 Schneider, C. Q., & Wagemann, C. 2012. Set-theoretic methods for the social sciences.
31 Cambridge: Cambridge University Press.
32 Schurman, R. 2004. Fighting “Frankenfoods”: Industry Opportunity Structures and the Efficacy
33 of the Anti-Biotech Movement in Western Europe. Social Problems, 51(2): 243–268.
34
Sillince, J., Jarzabkowski, P., & Shaw, D. 2012. Shaping Strategic Action Through the
35
36
Rhetorical Construction and Exploitation of Ambiguity. Organization Science, 23(3): 630–
37 650.
38 Snow, D. A., Morrill, C., & Anderson, L. 2003. Elaborating analytic ethnography: Linking
39 fieldwork and theory. Ethnography, 4(2): 181–200.
40 Soderstrom, S. B., & Weber, K. 2019. Organizational Structure from Interaction: Evidence from
41 Corporate Sustainability Efforts. Administrative Science Quarterly, 0001839219836670.
42
Sonenshein, S. 2014. How Organizations Foster the Creative Use of Resources. Academy of
43
44 Management Journal, 57(3): 814–848.
45 Sonenshein, S. 2016. How Corporations Overcome Issue Illegitimacy and Issue Equivocality to
46 Address Social Welfare: The Role of the Social Change Agent. Academy of Management
47 Review, 41(2): 349–366.
48 Sonenshein, S., DeCelles, K. A., & Dutton, J. E. 2014. It’s Not Easy Being Green: The Role of
49 Self-Evaluations in Explaining Support of Environmental Issues. Academy of Management
50
51
Journal, 57(1): 7–37.
52 Soule, S. A. 2009. Contention and Corporate Social Responsibility. New York: Cambridge
53 University Press.
54 Stake, R. E. 1995. The art of case study research. The art of case study research. Thousand
55 Oaks: Sage Publications.
56
57
58
59
60
 Academy of Management Journal Page 58 of 63

1
2
3 Strike, V. M., & Rerup, C. 2016. Mediated Sensemaking. Academy of Management Journal,
4
5
59(3): 880–905.
6 Tilly, C. 1999. From Interactions to Outcomes in Social Movements. In M. Giugni, D. McAdam,
7 & C. Tilly (Eds.), How Social Movements Matter: 253–270. Minneapolis: University of
8 Minnesota Press.
9 Uzzi, B. 1996. The Sources and Consequences of Embeddedness for the Economic Performance
10 of Organizations: The Network Effect. American Sociological Review, 61(4): 674–698.
11
Weber, K., & Waeger, D. 2017. Organizations As Polities: An Open Systems Perspective.
12
13
Academy of Management Annals, 11(2): 886–918.
14 Weick, K. E. 1995. Sensemaking in Organizations, vol. 3. New York, NY: Sage.
15 Wickert, C., & de Bakker, F. G. A. 2018. Pitching for Social Change: Toward a Relational
16 Approach to Selling and Buying Social Issues. Academy of Management Discoveries, 4(1):
17 50–73.
18
19
20
21
22
23
24
25
26
27
28
Figure 1. Added Solutions
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Page 59 of 63 Academy of Management Journal

1
2
3 Figure 2. Combining Opportunity Structures with Resource Strategies to Advance Reform
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27 Table 1. Overview of Data Sources
28
29 Data Quantity Analytical Purpose
30
31 Archival Data
32 Project Proposal 118 (1/fellow) Analyze initial project scope established for the fellow,
33 whether or not there were new initiatives, and how open-
34 ended the plan was
35
Final Project Reports 118 (1/fellow) Analyze solution creation in publicly available end-of-
36
fellowship commitments by fellow and host
37
38 Historical Climate 620 Establish program newcomers
39 Corps engagements
40 Surveys
41
42 Fellow responses (3 94 (completed at Close-ended questions establish movement resource use:
43 waves of survey) least one survey) intensity of external search and external focus of advice-
44 seeking
45 Open-ended questions coded for use of movement
46 resources by fellows
47
Interviews
48
49 Fellows 31 Analyze resource use within projects
50
51 EDF Engagement 9 Background and context; inform QCA calibration
52 Managers
53 Host supervisors 3 Background and context; inform QCA calibration
54
55
Alumni 12 Background and context
56
57
58
59
60
 Academy of Management Journal Page 60 of 63

1
2
3 Table 2. Fuzzy Set Membership Calibrations
4
5
Fully Out Crossover Fully In Std.
6 Condition Definition Mean
(0) point (1) Dev.
7
Dependent Variable
8
9 High solution Additions to the project scope from 0 1.5 4 2.6 1.6
10 creation the initial proposal to the final project
11 summary.
12 Organizational Context
13 New initiative Absence in project documents of -5 -2.5 -1 -3.5 1.3
14 organizational resources specific to
15 energy issues. Measurement is the
16 inverse of the sum of binary mentions
17 across the proposal and report of
commitments (e.g., this project will
18
help us meet our energy goal), prior
19
work (e.g., the fellow will continue to
20
develop our climate initiative), and
21 other resources (e.g., the fellow will
22 work with our sustainability team and
23 energy database).
24 Program Inverse of the number of previous -5 -1 0 -1.7 2.3
25 newcomer engagements between the host and the
26 Climate Corps program.
27 Open-ended Absence of detailed plans in project -4 -2.9 -1 -2.9 0.8
28 projects proposal, as indicated by the inverse
29 of the average level in the hierarchical
30 coding scheme, which goes from level
31 1 (least specific; e.g., building energy
32 use) to level 4 (most specific; e.g.,
33 occupancy lighting sensors).
34 Small org. Small organizational scale in terms of >10,000 1,000 <250 n/a n/a
35 scale size (number of employees) and scope employees, employees, employees,
36 (geographical spread of operations). international broader one city
37 operations metropolitan
38 area
39 Use of Movement Resources
40 External Engagement with external EDF 5 10.1 20 13.1 10.3
41 search resources, measured as use of each of
42 seven categories of Climate Corps
43 resources in the last month (other
44 fellows, engagement manager,
45 resource library, consultants,
46 webinars, LinkedIn group, and
mentors).
47
External focus Ratio of advice-seeking from the 0.5 0.99 1.25 0.8 0.3
48
Climate Corps network to advice-
49
seeking from within the host
50 organization.
51
52
53
54
55
56
57
58
59
60
Page 61 of 63 Academy of Management Journal

Table 3. Configurations of Causal Conditions Leading to High Solution Creationa

1
2 Low Ambiguity: Catalyzing Established Issue Moderate Ambiguity: Navigating High Ambiguity:
3 Understandings Organizational Fit Building Legitimacy
4
5
1 2 3 4 5 6
6 Organizational Contextb
7
8 New initiative   
9
10 Program newcomer    
11
12 Open-ended project    
13
14 Small org. scale      
15
Movement Resources
16
17 External search    
18
19 External focus     
20
21
22
Raw coverage 0.242 0.210 0.129 0.316 0.151 0.359
23 Unique coverage 0.060 0.037 0.031 0.026 0.033 0.064
24
25 Consistency 0.861 0.860 0.878 0.848 0.914 0.841
26
27 Number of cases 12 5 5 7 3 8
28
29 Overall solution consistency 0.828
30
31 Overall solution coverage 0.65
32 a Following Fiss (2011), black circles indicate the presence of a condition and circles with “X” indicate its absence. Large circles indicate core conditions; small
33
ones, peripheral conditions. Blank spaces indicate conditions that are irrelevant to the configuration. Overall scores are for parsimonious solution.
34 b The first three organizational context conditions (new initiative, program newcomer, and open-ended project) all reflect high ambiguity when present and low
35 ambiguity when absent. Small organizational scale, however, has a mixed relationship to ambiguity.
36
37
38
39
40
41
42
43
44
45
46
47
 Academy of Management Journal Page 62 of 63

Table 4. Overview of Strategies for Utilizing Movement Resources across Organizational Contexts with Low/High Ambiguity
1
2 Resource Type Low Ambiguity: Catalyzing Established Issue Understandings High Ambiguity: Building Legitimacy
3 Social support Overcome obstacles Overcome obstacles
4 Part of it was just talking through the challenges of, you know, Other fellows who worked on similar projects and I exchanged
5 sometimes you run into a dead end. And it was really helpful to be industry insights and updates often. We learned from each other's
6 able to talk to other people who were going through something host organizations. - Long
7
similar and find out that, okay yeah, if you follow this path much
8
further you might not get anything out of it and so, maybe you
9
switch it up and change direction and look somewhere else. - Jeff
10
11
12 Expertise and Highlight new opportunities Highlight new opportunities
13 tools Climate Corps provided a lot of contacts...We talked to lighting I used [the EDF financial tool] as a starting point and over the
14 retrofit companies who had a lot of experience with retrofitting course I modified it as per need...that was the first time I was doing
15 street lights and street lamps because that’s a really big opportunity that thorough financial modeling at the same time as the calculation
16
for energy savings for the city. It brought us a few different cities on how much energy they might be saving and in respect to how
17
who had gotten different grants and funding that had implemented much it cost. - Vivek
18
programs that [the city] was striving. It was helpful to have some of
19
20
that basic knowledge and understanding so we could translate that Provide needed knowledge and skills
21 into how [the city] might utilize some of their strategies. - Kaitlyn The [sponsor] was I would say decently knowledgeable about the
22 technology, but not down to the engineering and the technical side
23 of it. How does that [technology] actually work in the building?
24 They were not aware of it, and what are the limitations of using that
25 kind of technology...It's those kinds of technical details that they
26 were relying on us to tell them. - Vivek
27
28 Exemplars Reinvigorate existing programs Show feasibility of ideas
29
One thing that I drew on a lot was looking at other sustainability Interviewed industry subject matter experts to design and launch an
30
plans and particularly other sustainable building guidelines that had RFP for an energy study. Used RFP responses to show alternate
31
32
previously been developed in the [same] sector...there are a lot of primary power sources, including environmentally friendly
33 organizations that had already developed this kind of document and alternatives, that improve reliability. – Cathy
34 published it on their website. And so I spent a lot of time just
35 reviewing, basically doing a literature review of what else was out
36 there, and looking particularly at peer [organizations to my host].
37 That was a major resource that I drew on for the green buildings
38 guidelines...I went around figuring out what sort of work I could
39 leave behind that would be useful to a number of different
40 stakeholders [to my host]. - Jacob
41 Note: distinct resource uses indicated in bold.
42
43
44
45
46
47
Page 63 of 63 Academy of Management Journal

1
2
3 Table 5. Configurations of Causal Conditions Leading to Low Solution Creation (Inverse
4
5 Outcome)a
6 A B C
7
8 Organizational Contextb
9
10 New initiative  
11
12 Program newcomer   
13
14 Open-ended project   
15
Small org. scale   
16
17 Movement Resources
18
19 External search   
20
21 External focus  
22
23 Raw coverage 0.316 0.231 0.152
24
25 Unique coverage 0.080 0.013 0.051
26
27 Consistency 0.784 0.781 0.789
28
29 Number of cases 7 2 1
30
31
32 Overall solution consistency 0.753
33 Overall solution coverage 0.383
34
a Following Fiss (2011), black circles indicate the presence of a condition and circles
35
36 with “X” indicate its absence. Blank spaces indicate conditions that are irrelevant to
37 the configuration. Overall scores are for parsimonious solution.
b The first three organizational context conditions (new initiative, program newcomer,
38
39 and open-ended project) all reflect high ambiguity when present and low ambiguity
40 when absent. Small organizational scale, however, has a mixed relationship to
ambiguity.
41
42
43
44
45 Todd Schifeling is an assistant professor of management at Temple University, Fox School of
46 Business. He received a PhD in Sociology at the University of Michigan, where he was also a
47
postdoctoral fellow at the Erb Institute. His research examines organizational responses to social
48
49 movements and cultural change.
50
51 Sara Soderstrom is an associate professor in organizational studies and program in the
52 environment at University of Michigan. Her research advances understanding of how micro-level
53 interpersonal interactions influence macro-level outcomes, and how macro-level forces
54 shape the meaning people create around emergent sustainability issues. She received her PhD
55
56
from Northwestern University.
57
58
59
60