SAF Moulton2016

bs_bs_banner
The Policy Studies Journal, Vol. 00, No. 00, 2016
The Strategic Action Field Framework for Policy

Implementation Research
Stephanie Moulton and Jodi R. Sandfort
Drawing upon the theories of social skill and strategic action fields (SAFs), this article presents a
SAF Framework for Implementation Research. In the framework, policy implementation systems are
conceptualized as multilevel SAFs that form around a public service intervention. Within this
context, socially skilled actors leverage diverse sources of authority—including but not limited to
political authority—to enable change or stability to a public service intervention. While the
framework has underpinnings in field theory, it is able to encompass multiple theoretical perspectives,
including complexity theories, organizational theories, economic theories, and theories of human
behavior. Importantly, the SAF Framework allows for the integration of results relevant to both
management and policy—change in implementation systems, and change in the behavior or
conditions of an external target group.
KEY WORDS: public policy, public management, policy implementation
Policymakers and scholars have wrestled with the challenges of policy imple-
mentation for more than half a century. These challenges stem from the ambitious
aspirations often embedded in public policy, the diffuse governance of intergovern-
mental relations and service networks, and the tough problems that land in the pub-
lic arena for resolution. Taking place at the intersection of public management and
public policy, implementation often involves changing systems operations and alter-
ing conditions of target groups, both of which are difficult endeavors. As prior schol-
ars have recognized, policy implementation operates through a complex system of
social and political interactions (deLeon & deLeon, 2002; Goggin, Bowman, Lester, &
O’Toole, 1990; May & Jochim, 2013; O’Toole, 1986).1
In this article, we present a theoretical framework that incorporates this complex-
ity, while narrowing in on specific actions that take place at key sites within the imple-
mentation system (Weimer, 2008). We conceptualize these systems as multilevel
strategic action fields (SAFs) that form around a public service intervention. Rather
than starting with a formal policy (Gormley, 1986; Ingram & Schneider, 1990; Matland,
1995; Mazmanian & Sabatier, 1989; Pressman & Wildavsky, 1984), a policy problem
(Elmore, 1980; Hjern & Porter, 1981), or institutional setting (Berman, 1981; Lipsky,
1980; Milward & Provan, 1998; Selznick, 1949), our approach begins with a public
1
0190-292X V
C 2016 Policy Studies Organization
Published by Wiley Periodicals, Inc., 350 Main Street, Malden, MA 02148, USA, and 9600 Garsington Road, Oxford, OX4 2DQ.
2 Policy Studies Journal, 00:00
service intervention—investigating how ideas about creating change in the target pop-
ulation are institutionalized into a set of processes and methods of coordination purpo-
sively intended to bring about that change. Interventions can be introduced through
formal public policy or through the programmatic initiatives of entities such as evalua-
tion firms, nonprofit organizations, local governments, or private funders. Conse-
quently, we define implementation as deliberate, institutionally sanctioned change to a
public service intervention that is legitimated in part by political authority.
In constructing our framework, we draw upon the sociological theories of social
skill and SAFs (Fligstein, 1997, 2001; Fligstein & McAdam, 2011, 2012). These theories
are concerned with how social structures shape societal outcomes, and the role of
human agency in moderating these structures (Bourdieu, 1977; DiMaggio & Powell,
1991; Giddens, 1984; Scott, 1985). In our framework, implementation actors—be they
policymakers, public managers, service providers, funders, or direct service staff—
work within bounded social settings. They employ social skill to interpret and adjust
a public service intervention in ways that build common understanding and recon-
cile competing sources of authority to enable collective action. Different social
dynamics across SAFs lead to variations in how interventions are brought into prac-
tice, which, in turn, can contribute to variations in the outcomes in the system and
for the target population.
We find this foundation useful because of the significance of both social structures
and human agency in determining implementation processes and outcomes. While
other policy scholars have emphasized the role of social structures in policy systems
(e.g., McGinnis, 2011; V. Ostrom, 1980), our approach differs in its ontological starting
point. Unlike rational choice approaches, SAF theory does not presume that social struc-
tures are known or fixed (e.g., the “rules of the game”), leveraged by agents in pursuit
of their individual interests. Rather, rules and resources are produced, reproduced, and
altered by socially skilled actors in relation to collective interests. SAF theory also differs
from institutional theories that downplay the role of agency, focusing on pressures to
conform to norms (e.g., DiMaggio & Powell, 1991). Instead, SAF theory stresses a fun-
damental human need for sociability as the root of social interactions.
In the sections that follow, we describe and illustrate a SAF Framework for Imple-
mentation Research. First, we provide analytical grounding for the framework in the
sociological literature on social systems and SAFs. Next, we present three components
of the framework that extend the SAF approach to inform implementation: the public
service intervention (focus), multilevel SAFs (scale), and legitimization of authority
(driver of change). For each component, we define key elements and offer assumptions
central to informing the relationships between elements. While the framework and
assumptions are articulated generally, we then illustrate the application of the frame-
work to a particular case, the implementation of matched savings programs. Finally,
we conclude by describing areas of inquiry that can be informed by this approach.
Understanding the Complex Social Systems Involved in Implementation
Broadly speaking, social systems are contexts organized by relations among

agents, either individuals or groups, which are shaped over time by both exogenous
Moulton/Sandfort: The Strategic Action Field Framework 3
and endogenous factors and agents’ response to them (Bourdieu, 1977; Fligstein,
1997, 2001; Fligstein & McAdam, 2012; Giddens, 1984; Latour, 2005; Martin, 2003). In
an ambitious effort to advance understandings of social systems and the role of
human agency in determining stability or change, Neil Fligstein and Doug McAdam
(2011, 2012) articulate a theory of “SAFs.” SAFs are mid-level social orders where
collective action takes place around “shared understandings about the purposes of
the field, relationships to others in the field, and the rules governing legitimate action
in the field” (p. 9). The boundaries of SAFs are emergent, developed by participants
as they do shared work, determine their roles and relationships, and craft under-
standing of their goals and rules of acceptability.2
In our framework, the social system surrounding a given public service interven-
tion both constrains and enables what is possible to change in the course of imple-
mentation. For any given public service intervention, there are multiple and often
overlapping SAFs. In each, actors try to understand the intervention, work with
others to develop processes, and make decisions about which rules of conduct are
the most significant. Sometimes the activities undertaken on behalf of the interven-
tion are consistent across fields, shaped by rules or resources that transcend any one
field. For example, formal documents that communicate program requirements oper-
ate as rules that may be shared between SAFs, or mandated staff training might be
required resources introduced across multiple settings. However, the significance of
any rule or resource is not predetermined; actors make decisions about the relative
importance of different external and internal stimuli when deciding what actions
they will undertake. This is in part why implementation systems, when viewed
holistically, are complex and so difficult to control in terms of either process or out-
comes (Martin, 2003).
SAF theory also introduces a distinct assumption about human agency. Humans
seek sociability, engage in meaning making, and construct their own identities
through affiliation with others in groups (Fligstein, 1997, 2001; Fligstein & McAdam,
2012). This contrasts with the ontological position of rational choice that underlies
many approaches to policy studies, in which people are predominately motivated by
self-interest and a desire to both maximize individual gain and minimize the costs of
pursuing those interests (Downs, 1957; Neal, 1988). Instead, drawing from symbolic
interactionism (Goffman, 1959; Mead, 1934), individuals’ agency is directed toward
constructing meaning in relation to others (Fligstein, 2001, 2008; Fligstein & Mc-
Adam, 2012). While agents have individual interests, these interests are defined and
negotiated in relation to others within the system.
Because implementation is fundamentally about change, the theory’s conception
of stability and change is also particularly relevant. An essential factor in this regard
is social skill (Fligstein, 1997, 2001, 2008), defined as “the ability [of an actor] to induce
cooperation by appealing to and helping to create shared meanings and collective
identities” (Fligstein & McAdam, 2012, p. 46). As a central construct in the theory,
social skill highlights what actors do to contribute to the “emergence, maintenance,
and transformation of social orders” (Fligstein & McAdam, 2012, p. 50). While the
social dynamics of coercion, competition, and cooperation all can come into play,
socially skilled actors shape these dynamics using skills and tactics they believe are
appropriate both to the context and their desired outcome. Fligstein’s articulation of
the theory notes several such tactics (Fligstein, 1997, 2001; Fligstein & McAdam,
2012): using formal authority, setting action agendas for others, seizing unexpected
opportunities, framing action options in relation to others’ interests, brokering agree-
ments as a neutral party, trying multiple solutions, networking with those who are
isolated, and telling stories that appeal to others’ identities. Using such tactics,
socially skilled actors can engage others without necessarily relying upon coercion.
Socially skilled actors use knowledge about field dynamics to influence others
to either reproduce the current order or make change. Some work to preserve the
status quo to guarantee stability; in fact, actors benefitting from existing dynamics
within a particular context are rarely motivated to challenge the status quo (Flig-
stein & McAdam, 2012, p. 96). Most field members are committed to prevailing
conditions. But, if an actor understands or frames change in relation to what
already “makes sense” within the setting, social skill can enable the adoption of
incremental changes.
External events can trigger more large-scale changes. A new actor to the field, a
substantial change in a proximate field, or a large-scale event may affect an entire
system. For example, the consequence of hiring a new visionary leader, passage of a
new piece of legislation, or an economic crisis often has implications for many fields.
But, SAF theory stresses that implications of these exogenous forces are not predict-
able. Actors with social skill influence how any of these events will be received and
integrated into a context. They play an important role in interpreting, framing, and
brokering the change.
SAF theory offers some potent analytical insights for the study of policy imple-
mentation. It allows for analysis of SAFs at various levels in a system. It directs schol-
ars to look at what actors actually do to support or impede change (Feldman &
Orlikowski, 2011). It also illuminates that actors’ possible activities are enabled and
constrained by the social dynamics within the field, as well as exogenous factors. These
insights can be applied to the complex dynamics surrounding policy implementation.
Conceptualizing a New Framework for Implementation Research
SAF theory as articulated by Fligstein and McAdam (2012) does not focus on policy
implementation but rather calls attention to the drivers of change in complex social sys-
tems.3 For that reason, while it provides a foundation for a new framework, there are
important enhancements necessary if we are to take these insights and apply them to
the implementation of public service interventions. Our framework also has a narrower
focus than other frameworks of the policy process or institutional analysis. We agree
with Weimer (2008) that a tighter focus around specific actions (in this case, actions
around an intervention) produces research more relevant for policy and practice.
In the section that follows, we define and describe three components that are cen-
tral to our framework: the public service intervention with variation in processes, coor-
dination, and results; multiple levels of strategic action that frame the scale of analysis,
including the actors, resources, and roles at a particular scale; and the activation of
diverse sources of authority through social skill that provides the engine for change or
stability in the public service intervention. Through providing a theoretical framework,

we aim to offer common elements that can help researchers specify variables and
generate questions related to implementation (E. Ostrom, 2007, 2011). Within each
component, we provide assumptions that are central to understanding the implemen-
tation of any public service intervention. While the framework has underpinnings in
field theory, it is able to encompass multiple theoretical perspectives, including com-
plexity theories, organizational theories, economic theories, and theories of human
behavior.
Table 1 summarizes the elements of the framework by component and possible
sources of variation. It provides a reference to enable scholars to develop specific
research questions when applying this framework to particular studies of policy
implementation.
Table 1. A Strategic Action Field Framework for Implementation Research
Component Elements Examples of Variation
Program Intervention
Processes of change Degree of complexity as indicated
by routinization, number of steps,
or predictability; targeted change
(e.g., people changing vs. people
processing)
Methods of coordination Degree of reliance on technical
expertise; variation in the
sequencing of tasks (e.g., sequen-
tial, pooled, interdependent); tools
in use with varying coerciveness,
directness, automaticity, and
visibility
Change in system operations Alterations in processes used by
agency (e.g., efficiency, accessibil-
ity), as well as degree of integra-
tion of intervention into every-
day practices (normalization)
Change in target group Alterations in target group experien-
behavior or conditions ces, as well as degree of change
in their behavior or conditions
Scale of Analysis
Policy field (assembly) Types of structures in use, historical
relationships, newness of the field
Organization Degree of intervention alignment
(operationalization) with other program processes and
technologies
Frontlines (enactment) Degree of worker discretion; degree
of engagement with the target
population
Drivers of
Change and Stability
Sources of authority Degree of (perceived) influence
from political authority, economic
authority, norms, beliefs, and
values
Social skills Degree of use of tactics such as
interpreting, framing, brokering,
and bridging
Exogenous shocks Degree of stability or instability;
changes in funding, legislation, or
field actors
Public Service Interventions and Results
The unit of analysis in our framework is one or more SAF that surrounds a
particular intervention. A central tenet of SAF theory is that the boundaries of a
field are not fixed, but rather “shift depending on the definition of the situation
and the issues at stake” (Fligstein & McAdam, 2012, p. 10). For analyzing
implementation dynamics, we suggest that the public service intervention
defines the issue at stake and provides an understanding (often contested) of
the roles carried out by different field actors in response to the issue. This is an
important distinction, as the unit of analysis in our approach is not a singular
policy domain around a policy problem area (Burstein, 1991; May & Jochim,
2013), but rather the fields activated in response to a particular public service
intervention.
It is important to define what we mean by public service intervention. By
intervention, we mean a complex bundle of processes and methods of coordina-
tion intended to transform inputs into outputs (May, 2013; Perrow, 1986). By pub-
lic service intervention, we mean interventions that are intended to benefit the
public that are authorized at least in part by political authority (a term that we
will describe in more detail below). We use the term public service in the demo-
cratic sense to be “in service to the public” (e.g., Denhardt & Denhardt, 2000;
Moynihan & Pandey, 2007; Perry, 2000), not limited to “service”-based
interventions.
In organizational settings, scholars sometimes refer to this complex bundle
as “organizational technology” (e.g., Daft, 2010; Miles, Snow, Meyer, & Coleman,
1978; Orlikowski, 1992; Scott, 2008). Of relevance to our framework, organiza-
tional theorists have proposed a structuration model of technology (DeSanctis &
Poole, 1994; Orlikowski, 1992, 2000; Sandfort, 2003), following the structuration
theory of Giddens (1984). The structuration model of technology acknowledges
the dual nature of technology, as both an independent force that impacts organi-
zational structures and processes, and as a dependent outcome that is shaped
and adapted by human agents engaged in the process of carrying out work. Sim-
ilarly, in our approach, while the public service intervention has independent
attributes that influence implementation dynamics, we also recognize that it is
the product of the SAFs within which it is embedded.
Building from the construct of organizational technology, we delineate two ele-
ments of public service interventions that shape and are shaped by SAFs: (i) the
processes of change that link inputs and outputs to produce anticipated outcomes;
and (ii) the materials, technical skills, and structures that are needed to coordinate
the work performed. Other policy implementation scholars have called attention to
these two elements, but they typically categorize them as attributes of a policy or
domain that exist a priori rather than attributes of an intervention that evolve with
and from strategic action within a field.
First, for any intervention, there are varying logics about how to create change in
the target population. To the extent that there are fewer required steps to transform
inputs into outputs (e.g., fewer links in the causal chain), scholars have asserted that
policies are more likely to be implemented “successfully,” or in line with the design-
er’s intent (Matland, 1995; Mazmanian & Sabatier, 1989; Pressman & Wildavsky,
1984). Some public policy domains, such as utilities, transportation infrastructure,
and national security, deploy routine logics and technology where raw materials can
be seen, measured, and analyzed and outcomes similarly measured. Other policy
domains such as social services or economic development are often less routine and
vary depending on the attributes of the individuals or communities targeted by the
policies, particularly when people are the primary raw materials and the goal is to
change behaviors rather than process individuals through a system (Hasenfeld, 1983;
Sandfort, 2010).
However, even in more “routine” policy areas, complexity in the causal logic
can be introduced through implementation actors. For example, some municipal
utility programs not only collect payments (a simple causal logic) but also seek
to increase the conservation behaviors of consumers (Cromwell et al., 2011).
According to our framework, the relative complexity of municipal collection
practices in one locality compared with another is the result of differences in the
SAF, due in part to varying knowledge and interpretations offered within the
field by implementation actors. Organizational research documents that even in
“simple” interventions, social forces, human interpretation, and actions add
complexity in how the intervention is actually instituted (Orlikowski, 1992; Staf-
ford & Wilson, 2015; Wajcman, 2006). Each public service intervention—simple
or complex—is shaped by a collective understanding about a causal logic within
the SAF in which it is embedded. This is a subtle, yet important distinction from
prior policy implementation frameworks. The complexity of the means-end
chain is not necessarily endemic to the policy domain or problem trying to be
addressed, but rather is reflective of the understandings in the SAFs around a
particular intervention.
The second element of public service interventions includes the raw materials,
technical skills, and structures that coordinate work to transform inputs into outputs
(methods of coordination). Prior implementation scholars have referred to the vary-
ing ability of the policy to “structure” implementation (Matland, 1995; Mazmanian &
Sabatier, 1989). More recently, variation in coordination has been conceptualized as
stemming from different “tools of governance,” defined as “an identifiable method
through which collective action is structured to address a policy problem” (Salamon,
2002, p. 19). Different policy tools—such as grants, contracts, regulations, vouchers,
and tax expenditures—have distinct characteristics, including varying degrees of
coerciveness, directness, automaticity, and visibility (Salamon, 2002). Taken prescrip-
tively, policy tools should be designed and selected a priori to fit the policy problem
to be addressed and maximize outcomes such as effectiveness, efficiency, and equity.
However, in reality, there are trade-offs between outcomes, and particular tools are
often selected for political or institutional reasons rather than economic or substan-
tive rationales.
In our framework, the methods of coordination in use for an intervention evolve
from the dynamics within the SAF—which may or may not be the rationally best
methods to fit the problem. Even within the same policy domain, state and local
actors make adjustments to the “tool” to fit their circumstances (Sandfort, Selden, &
Sowa, 2008; Whitford, 2013). Organizational scholars have long recognized that while
certain structures are better suited for particular tasks or environments (e.g., contin-
gency theory), what is observed in practice is often the structures that are perceived
of as legitimate by actors within the field (DiMaggio & Powell, 1991; Fligstein, 1985;
Scott, 2008).
Attempts to make changes to public service interventions often entail
changes to the intervention’s processes and/or methods of coordination.
Changes that are successful are those that are institutionally sanctioned by the
SAFs within which the intervention is embedded. Over time, processes and coor-
dinating structures become institutionalized, no longer seen as “new” but part of
the way work is accomplished. This is the primary focus of implementation sci-
ence research; exploring the factors that lead to the successful integration of new
evidence-based interventions into standard operating procedures (Damschroder
et al., 2009; Durlak & DuPre, 2008; May 2013; May et al., 2009). While this is one
important outcome of implementation, there are other results that are important
for the implementation of public service interventions. For example, in carrying
out public services, other public values such as efficiency and transparency in
using public investments or the quality and equity in the distribution of resour-
ces to groups targeted by the policy are also important (Lynn, Heinrich, & Hill,
2000; Robichau & Lynn, 2009).
Ultimately, the success of a public service intervention depends upon the extent
to which the intervention achieves the desired change in an external target group. A
well-integrated, institutionalized change that leads to poor outcomes for the target
group is akin to neatly arranged deck chairs on a sinking ship. However, only focus-
ing on target population outcomes often reduces the significance of what it takes to
alter an intervention—viewing system elements as levers that can be raised or low-
ered to produce better outcomes. Implementation scholars have critiqued policy ana-
lysts and program evaluators for ignoring or oversimplifying the processes of
change often necessary to improve target group results (Nilsen, Ståhl, Roback, &
Cairney, 2013; O’Toole, 1986; Robichau & Lynn, 2009). Our application of the SAF
framework provides an approach to incorporate these multidimensional results in
implementation research.
Assumption #1: The starting point for implementation research is a public service
intervention around which collective action occurs, which may result in change to
the system and an external (target) group.
Multilevel Systems
Part of the analytical complexity of implementation comes from the reality that
SAFs rarely exist in isolation but are nested within and related to others. Depart-
ments within large organizations can each operate as distinct fields, sharing some
similarities but possessing unique power, resources, or dominant cultural interpreta-

tions. Similarly, fields can span organizations. Fligstein and McAdam (2012) explic-
itly note that SAFs are both vertically nested and often hierarchically connected to
other fields. These vertical and horizontal ties introduce resource dependencies and
different sources of authority.
There are often numerous fields engaged around any particular public ser-
vice intervention. We follow the lead of other scholars and embrace a heuristic
understanding of these systems as multilevel (M. Hill & Hupe, 2014; Kiser &
Ostrom, 1982; Lynn et al., 2000; E. Ostrom, 2007; Robichau and Lynn, 2009). Such
multiple governance frameworks direct analytical attention to distinct responsi-
bilities in constitution, direction, and operation played at different levels in a sys-
tem (M. Hill & Hupe, 2014). In our framework, the public service intervention is
shaped by decisions made at different levels of strategic action. Each level oper-
ates as its own SAF, in which actors possess varying power, play particular roles,
and share a broad interpretive frame that guides action within the field at that
level. Similar to other multilevel frameworks, rules in use at a higher level may
constrain and enable the rules and action at a nested lower level (E. Ostrom,
Cox, & Schlager, 2014). In our framework, we differentiate policy fields, organi-
zation, and frontline levels within these systems.
Policy fields are bounded networks among organizations carrying out a sub-
stantive policy or program in particular places (Stone & Sandfort, 2009).4 In any
geographic area, there is a potential pool of organizations including public agen-
cies, industry or other advocacy groups, philanthropic funders, membership
associations, and service providers, which might participate in implementation
because of their expertise or interests. Resources and understandings flow
between agencies through horizontal networks or top-down intergovernmental
relationships. In our conceptualization, policy fields are the macrolevel SAF
within an implementation system.
At this level of action, decisions are made that affect the assembly of the public
service intervention. Ideas are considered that inform the logic about how to create
change in the target population, identifying processes that are feasible or desirable. It
is often at this level that evidence-based practices and results of policy analyses are
disseminated and debated. Resources may be mobilized when there is a perceived
gap between current field capacity and desired outcomes. For example, field actors
may secure additional funding for training around a generally accepted best practice.
At this level, the tools to coordinate joint action and the delivery of an intervention
are also negotiated, such as choices to structure delivery through direct grants or
competitive contracts, or as vouchers provided directly to the target population.
Operational decisions about the intervention are made at the organizational
level. Organizations are the mezzo level of implementation systems—the institu-
tional link between the policy field and frontlines of the implementation system,
and actors within organizations make important decisions that cross levels. At
the overlap with the policy field level, actors within organizations take action to
authorize key aspects of the processes and coordinate the intervention. The
ideals shaped in the policy field about elements such as performance criteria or
accountability rules need to be reconciled and integrated into the SAF dynamics
within implementing organizations. Sometimes, state or regional public agencies
provide authoritative guidance; however, private foundations, nongovernmental
organizations, and even direct providers can take such actions to reconcile field-
level parameters and operational realities.
At the overlap with the frontlines of the system, organizations operationalize
how the intervention will be carried out on a day-to-day basis. They develop proce-
dures for engaging program participants, ensuring standards of care or quality, and
documenting activities involved in carrying out the intervention. The current prac-
tices and expertise within the service organization directly shape how a public serv-
ice intervention is operationalized. If the organization’s current practices align with
the intervention’s logic of change, requisite technical skills and structures, implemen-
tation is easier. A corollary of this is that organizational actors often make changes to
the intervention to increase alignment with their existing assumptions and systems.
Making changes to the intervention that do not align with current practices often
requires new knowledge, resources, and willingness to be generated from within the
organization or the larger policy field.
The final level in our framework is the frontlines, where the implementation system
interacts directly with the target group, be they students, energy consumers, or citi-
zens. Sometimes, the frontlines involve face-to-face interactions between staff members
and the target group. Other times, the system and target group interact through virtual
means, such as online registration or call centers. The frontlines are where the public
service intervention is enacted for and with target group members, who are often,
themselves, “key problem solvers” in determining successful outcomes for the policy
change (Cohen & Moffit, 2009). It is the microlevel of the implementation system.
As SAFs, the frontlines are shaped by the larger organizational context within
which they operate. In fact, frontlines are sometimes subsumed into the operational
level of governance. However, for public service interventions, we think it is impor-
tant to recognize the distinct field dynamics that emerge between frontline workers
and target population participants. Frontline workers share ideas, assess options,
and commiserate about the inadequacies of directives to respond to particular situa-
tions (Maynard-Moody & Musheno, 2003; Sandfort, 1999). While theories about
street-level bureaucracy suggest generalizable constraints within frontline public sys-
tems (Lipsky, 1980; Maynard-Moody & Portillo, 2010), conceptualizing these imple-
mentation sites as SAFs suggests that such constraints can be mediated by the
structures and social skills demonstrated by actors within that field or proximate
fields. With regard to intervention processes, frontline workers often hold construc-
tions of the target population and beliefs about the efficacy of a particular interven-
tion that influence the degree to which they engage with the intervention or simply
follow procedures formulaically (Soss, 2005). While this distinction is more important
for interventions focusing on changing individual behavior, research documents dif-
ferences in outcomes by frontline workers even in more routine interventions.
Taken together, the three levels of action are not dissimilar to other policy and
governance frameworks. However, our framework places emphasis on the field
dynamics at each level that affect the public service intervention.
Assumption #2: SAFs at multiple levels of a complex implementation system con-

strain and enable action around the public service intervention, thereby causing
variation in enacted practices.
Activating Legitimate Authority to Shape Implementation Practice
Implementation scholars have long recognized the importance of formal and

informal rules and resources that govern participant interactions (Kiser & Ostrom,
1982; E. Ostrom, 2007; Sabatier & Jenkins-Smith, 1993; Weible, Sabatier, & McQueen,
2009). However, rules and resources do not operate as absolutes. To be followed,
they have to be backed by authority, which derives from the perceived legitimacy of
the entity or person issuing the directive (Blau & Scott, 1962; Fligstein, 2001). Further,
there are multiple, often conflicting sources of authority in any field. Elinor Ostrom’s
(2011) notion of “polycentricity” reflects this notion, as she highlights that authorities
from different jurisdictions (what we could call “fields”) interact to shape the rules
and norms that guide behavior. Within the SAF framework offered here, it is the role
of actors with social skill to activate, resolve, and interpret signals from diverse sour-
ces of authority to create collective understanding for action around the public serv-
ice intervention.
We identify four sources of authority that are common in public implementa-
tion systems. First, public service interventions draw upon some source of
“political authority.” Political authority defines what is legally permissible,
establishes mandated processes, stipulates which public institutions are vested
with legal responsibilities, and delineates hierarchical governance relationships
between fields. This type of authority may involve negotiations between
branches of government, as implementation activities may be beholden to both
legislative oversight and administrative regulations. It may also involve the
courts to interpret what is permissible for a given intervention. Political author-
ity is also demonstrated when citizen preferences are reflected through delibera-
tive processes (Denhardt & Denhardt, 2000); however, the extent to which such
processes are legitimate varies by field.
There are other important sources of authority that can legitimate interactions
around a public service intervention. In the last 40 years, economic authority that
derives from market conditions is particularly potent in policy implementation
(Bozeman, 2007; Dahl & Lindblom, 1953; Salamon, 2002). Economic authority stems
from competition and reliance on market mechanisms to coordinate action. When
public agencies incentivize behavior through contracts with performance criteria or
use vouchers that tap market mechanisms, they draw upon this form of authority to
develop rules for interaction. Within implementation systems, economic authority is
often quite visible, creating incentives tied to specific processes or outcomes. Yet, sig-
nals often conflict, and consequences are not inevitable but shaped by actors within
a particular context.
Unlike the mandates imposed through political authority or the incentives intro-
duced through economic authority, there are other informal sources of authority
(Khademian, 2002). Norms associated with professional expertise may be a particu-

larly potent form of authority in particular fields. Expressed through professional
codes, evidence-based practices, or industry standards, norms can create binding
expectations of what needs to be considered during implementation activities. They
are demonstrated when powerful field actors issue white papers or offer training
programs, or when social scientists attempt to entreat others to take up their
research-based intervention. Heather C. Hill (2003) refers to these influences as
“implementation resources,” vesting authority in nonstate actors to frame viable
interventions and desirable results. Implementation resources communicate the
norms and best practices relevant to a core intervention, attempting to narrow the
range of appropriate responses to the task at hand. Rather than coercive pressures
that stem from formal authority, the terminology “norms” implies that the ideas are
often promulgated through imitation, referred to as mimetic isomorphism (DiMag-
gio & Powell, 1991). While this normative pressure might exist, it is not guaranteed
to shape the core intervention in any particular context.
A final source of authority is the beliefs and values shared by individuals and
groups. Scholars document how shared beliefs and values reduce uncertainty, oper-
ating as a cognitive framework that provides a means of sensemaking, often in light
of what is understood about the past (Khademian, 2002; Weick, 1995; Yanow, 1996).
For example, moral categorizations of both workers and the target population may
shape how an intervention is implemented (Mettler, 2007; Schneider & Ingram, 1997;
Soss, 2005). Shared beliefs about the efficacy of an intervention may shape whether
or not actors integrate new activities into daily practices (Damschroder et al., 2009;
May & Finch, 2009).
The four sources of authority described here are meant to be illustrative; the
manifestations of authority depend upon the definitions within a particular situation,
consistent with other applications of complex systems theory.5
Assumption #3: Within each setting, there are various potential sources of author-
ity that can provide field actors with a rationale for the practices they enact.
The significance of any particular authority source depends on how it is

interpreted within a particular context. In some contexts, a mandate associated
with a funding source is understood to provide cause for staff to abandon profes-
sional norms about service quality. In another context, that same mandate moti-
vates mangers to seek supplemental funding or develop an advocacy strategy to
push back against the directive. Rather than having objective significance,
authority sources are better thought of as latent; they can either be activated or
squandered by field actors to motivate collective action (Feldman & Quick,
2009). Actors involved in implementation at any level in a system are actively
engaged in cocreating the conditions within the SAFs where implementation
activities get carried out.
So what are the mechanisms for this cocreation? How are authority sources acti-
vated or squandered? Our framework stresses the significance of social skill as
providing an explanation for why change or stability surrounds the implementation

of a particular intervention (May, 2013). Building upon the assumptions about
human sociability discussed earlier, social skill acts as a “lubricant.” Fligstein and
McAdam write (2012, p. 178) “Rules and resources exist and are building blocks. But
without the social skill of people who interpret these rules and mobilize these resour-
ces, the rules and resources do not matter all that much.” Said another way, legiti-
mate authority is not externally determined. Rather it is endogenous to a particular
field within the implementation system.
Socially skilled actors implicitly recognize their own agency when deciding
how to respond to what others might understand to be compulsory. They recog-
nize that often, competing claims to authority are alive within a field. Because
people have different vantage points about what should occur during implemen-
tation, they often inconsistently respond to competing legitimacy claims. In these
situations, socially skilled actors use their understanding of people in the context
to develop an agreement about a plausible way forward. They influence others’
understanding by offering their interpretation of events, frame action options,
and set an action agenda by engaging others and appealing to their interests.
They can act as a neutral party and broker agreements between people who share
distinct understandings of the tasks at hand.
Socially skilled actors can also serve as boundary spanners, bringing ideas and
authority from one level of the system (e.g., the frontlines) to other levels in the sys-
tem (e.g., the policy field) or transferring ideas and authority between proximate
fields (e.g., two organizations within the same policy field). Direct experience at the
operational levels of the system may provide legitimacy in policy field discussions.
And, the ability to interpret signals from the policy field for organizational actors
may provide a source of legitimacy in operational discussions. Socially skilled imple-
menters may bring resources and authority from other contexts to help reconcile
competing authority claims within their field of action.
There are also instances where exogenous shocks alter the sources of authority
salient within an implementation system. Perhaps, there is an economic crisis or
another event that fundamentally changes peoples’ understanding of the intent or
desired outcomes for an intervention. Our framework stresses that the consequences
of even large shocks are not inevitable. Implementation requires reconciling contro-
versies and brokering understandings about the logic of change and the coordination
required to carry out an intervention. This is a social process that is generalizable
across SAFs, those at the policy field, organizational, and frontline levels in an imple-
mentation system. It is also dynamic, requiring effort to either maintain the status
quo or make changes consistent with the new authority claim. This leads us to the
fourth assumption of our framework:
Assumption #4: Field actors utilize social skill to re-frame what is perceived to be
legitimate practice for public service interventions, leveraging existing authority
sources in new ways, facilitating the creation of additional sources of authority, or
capitalizing on exogenous shocks to create new authority for action.
Figure 1. Framework of Strategic Action Fields in Complex Implementation Systems

Key: P, political authority at different levels in the system; E, economic authority at different levels in
the system; N, norms at different levels in the system; B, beliefs and values (mechanisms of culture),
at different levels in the system.
Putting It All Together
The three components of the framework offer a way to conceptualize change

and stability to a public service intervention as it occurs in a multilevel implementa-
tion system. Figure 1 provides an illustration of this framework. SAFs operate at the
policy field, organization, and frontline levels, with the public service intervention as
the core. The SAFs at each level are the units of analysis for implementation research.
At each level, sources of authority activated by socially skilled actors create bounda-
ries around the legitimate actions that are deemed possible in relation to the inter-
vention, but the boundaries themselves are dynamic.
Actions taken around the public service intervention shape both the logic of
change adopted and the materials and structures employed to coordinate activities,
and this has causal significance in determining what results. As we have noted, there
are dual results of implementation activities: changes to system operations, and
changes to the behavior or conditions of an external target group. These changes can
be observed as objects of scholarly inquiry. And as more is uncovered about whether
and under what conditions awareness of results helps anchor socially skilled action
in implementation systems, such findings can help improve implementation of
public service interventions across policy domains and settings. To better illustrate
how this application can occur, we briefly consider an illustrative case.
Applying the Framework: An Illustrative Case
Thus far, we have presented the SAF framework independent of context to allow
it to be generalizable to a variety of public service interventions. However, to demon-
strate how the framework can be applied to formulate research questions, it is useful
to illustrate its application in a specific case; in this instance, the Individual Develop-
ment Account (IDA) matched savings program. We provide brief background about
the case, and then apply the three components of the framework to the case, high-
lighting how the elements and assumptions of the framework are realized in the
case, and how these elements can help structure research questions.
Case Background
As income-support programs fell out of favor in the United States during the
1990s, scholars began to realize that poverty was not merely caused by lack of earn-
ings but also by limited assets (McKernan & Sherraden, 2008; Shapiro & Wolff, 2001;
Sherraden, 1991, 2000). Because most policies encouraging asset development oper-
ate through the tax system, there was growing bipartisan concern that the regressive
nature of asset-based policies would have negative consequences for lower income
populations (Sherraden, 2000).6 Like most social problems, many interventions have
been enacted at the national, state, and local levels to boost household assets, ranging
from changes to means-tested criteria for public benefit programs, to goal directed
savings incentives, to financial education programs. The IDA program is one
example.
IDAs are a public service intervention designed to offer program participants a
financial match (often more than 1:1) for savings deposited with a financial institu-
tion. After a minimum amount of time, the asset can be used to defray costs associ-
ated with higher education, home purchase, or the start-up of a small business.
While federal and state legislation provides some direct support for IDAs,7 national
surveys of program operators suggest that nearly half of the programs also receive
funding from private sources (Corporation for Enterprise Development [CFED],
2014). Policy researchers at the CFED, a national nonprofit organization supporting
the program, estimate that over 1,100 sites have operated IDAs, opening over
100,000 accounts for income-eligible participants (CFED, 2014).
Focus: Public Service Intervention
Applying the SAF framework to the IDA case, we begin by unpacking the public
service intervention and the potential for variation. Certain aspects are relatively clear:
matched savings accounts (the intervention) are made available to qualified low and
moderate income households (the target population) with the purpose of facilitating
asset development through home purchase, business start-up, or higher education

(change in target group behavior). However, other aspects of the intervention are open
for debate. For example, programs differ in the amount of match provided to induce
participants to save money, ranging from 1:1, 2:1, 3:1, 6:1, or even 7:1. The processes of
creating change differ in the additional services or “steps” required, beyond the
matched savings account, including additional education and counseling, restrictions
on eligible withdrawals, and penalties for ineligible withdrawals. Methods of coordina-
tion differ in their automaticity—some programs rely heavily on the existing infra-
structure of financial institutions to coordinate savings and withdrawals, while others
rely on nonprofit partners to collect deposits or manage withdrawals in partnership
with financial institutions. Researchers have documented that these variations in the
IDA intervention can impact both program viability and participant outcomes (Grin-
stein-Weiss et al., 2013; Mills et al., 2008; Schreiner, Clancy, & Sherraden, 2002).
In our framework, these elements of variation become the focus of research
investigations, comprising dependent and/or independent variables to be explained
and explored. Questions can be framed to understand variation in the different
attributes of the intervention. Under what conditions are certain processes (e.g.,
higher or lower match rates) employed? Which conditions give rise to the adoption
of certain methods of coordination, such as more or less automatic payment struc-
tures or reliance on a particular nonprofit network as service providers? Questions
can also be framed to investigate results, including how the matched savings pro-
gram is integrated into a policy field or organization, or the consequences of pro-
gram components on participant savings or financial literacy. In this way, attributes
of the intervention are dependent variables that are affected by the SAF, but are also
independent variables that influence participant outcomes. Interpretive research
using qualitative methods would likely focus on understanding how an intervention
develops, focusing attention on it as a dependent variable. Quantitative, quasi-
experimental designs could be employed to disentangle the effects of the system and
the resulting intervention on target population outcomes.
Scale: Multilevel SAFs
The second component of the framework calls attention to the multilevel SAFs
that surround an intervention. Like many policy areas, the selection of IDA processes
(such as the match rate) are often not based on empirical evaluations of which
savings-match rate yields the most desirable outcome. Rather, implementers make
this decision based on what they consider legitimate or feasible within the context at
the policy field, organization, and frontlines of the system.
In the IDA case, the constellation of actors in the policy field and the methods
employed to facilitate coordination vary substantially by state. For example, about
one-third of the states have a centralized network structure with a public agency dis-
bursing funds to local nonprofit organizations (CFED, 2014).8 In some states,
nonprofit-led working groups hold regular meetings and disseminate best practices
to facilitate coordination. In other states, no formal state-wide coordination or
funding structures exist, and individual programs seek funding directly from field
organizations, foundations, or individual donors. These distinctions in field actors,
resource distributions, and their collective decision processes reflect differences in
what is understood to be a legitimate resolution of options when assembling the
intervention in a particular place.
Even when states have a central agency authorizing or coordinating the program,
there is considerable variation at the organizational level that can affect IDA interven-
tions. The core mission and focus of the central organizations often varies, with some
housed in state housing and community development agencies, others in state human
service agencies, and still others in nonstate community development corporations or
community action agencies (CFED, 2014). Similarly, there is considerable variation in
the service providers offering IDA programs. For example, in Oregon, the 10 agencies
operating IDA programs include refugee and minority group organizations, commu-
nity development agencies, and a housing developer. Differences in missions and rou-
tine technologies employed by the organizations (e.g., providing public benefits or
educating citizens) lead to variations in adaptations of the IDA intervention to fit the
ongoing work of the organization; for example, housing organizations might be more
likely to emphasize the homeownership goals of the IDA program than other goals.
The enactment of the program at the frontline level can substantially influence how
the program is experienced by the target population and subsequently what results.
For example, frontline workers may adopt their own internal processes for approving
or denying savings withdrawal requests from participants; some may require in-per-
son consultations, while others may approve requests taken over the phone or by
email. These variations may be due to coping strategies to deal with heavy workloads,
lessons learned from professional trainings, or long-held beliefs about the target group.
This component of the framework calls attention to the scale of analysis.
Researchers may be particularly interested in actions taking place around the inter-
vention at a particular level. For example, researchers may compare states as distinct
policy fields, identifying how differing historical relationships, resources, and social
skills of actors influence the resulting IDA intervention. At another scale of analysis,
researchers might compare the primary mission of service organizations (e.g., hous-
ing, employment assistance, poverty alleviation) and explore how this influences the
integration of the program processes into day-to-day operations. At yet another
scale, a researcher might focus on frontline conditions and explore how workers at
this level use particular resources to inform their own work with clients. Research
can also explicitly investigate the nested levels of systems through multilevel model-
ing. Regardless of the scale of analysis, researchers should acknowledge how the
constraints and decisions of the proximate fields shape the implementation options
at other levels, with an awareness of the complete implementation system.
Stability and Change: Authority and Social Skill
The final component of the framework explicates field elements that contribute
to the stability of and change of the intervention—namely, the activation of diverse
sources of authority through social skill. In the IDA case, all four sources of authority
are activated at different times and scales of analysis. In terms of political authority,
some states passed laws that, among other things, specified the match rate and reve-
nue sources. And, government agencies often oversaw the development of rules to
guide IDA implementation. When political support for grant funding was low,
socially skilled actors often leveraged more politically palatable economic incentives
through individual and corporate tax credits to supplement revenue for IDA pro-
grams (Portland State, 2013).
Aside from more formal sources of authority, the initial spread of the IDA pro-
gram reflects the influence that shared norms can have within an implementation sys-
tem. The Center for Social Development (CSD) at Washington University, where
IDA’s were first conceptualized, is a leading source of technical assistance and evalu-
ation in the county. Socially skilled actors within CSD framed strategies and
equipped state-level actors with tools to bring their programs to scale, leading to
some degree of standardization of programs across fields. Conversely, differing
beliefs about the causes of participant poverty and their ability to save contributes to
variation in the intervention across sites. Where lack of savings is perceived to be
due to poor budgeting or financial knowledge, an emphasis on financial education is
emphasized. When the lack of savings is seen as caused by lack of income, a greater
emphasis may be placed on job training and income support programs.
Researchers can analyze drivers of change and stability as independent variables
that may contribute to variations in the intervention. In practice, socially skilled
actors at various levels mobilize authority sources to either introduce or block
change. To what extent are regulations or normative pressures (e.g., from CSD) acti-
vated to introduce new practices or prevent change to existing interventions? These
drivers can also be analyzed as dependent variables that are shaped by field dynam-
ics. Researchers particularly interested in the social process of SAFs might explore
how and under what conditions particular social skills are applied. This likely
requires a more in-depth, qualitative approach.
Using the Strategic Action Framework for Implementation Research
Through the illustrative case, we show how the SAF framework could be
applied to inform investigations around a particular intervention. It suggests ways
the framework can give rise to new directions for social science research and new
possibilities to inform practice (O’Toole, 2004). While the context often drives the
specific research questions of interest, the SAF framework is well equipped to frame
general areas of inquiry.
First, the SAF framework provides a new way to make sense of the variation
observed in a particular intervention across sites of implementation. Why do we see
an intervention being adopted in different ways across settings (e.g., states, counties,
local organizations) in order to create the “same” desired change in the behaviors or
conditions of the target group? The framework offered here would direct researchers
to document the existing social structures (rules and resources in use) in the imple-
mentation system, and the strategies used by socially skilled actors to introduce
change or reinforce the status quo, including the mobilization of particular sources
of authority or exogenous shocks to the system.9
Second, a concern of great practical significance is the overall capability of imple-
mentation systems to deliver desired change. For example, a new “evidence-based
practice” regarding an intervention has been identified, but it is unclear the extent to
which implementation sites are ready and able to adopt the new practice. How do
historical and current conditions within the SAF influence the time it takes to inte-
grate new or modified interventions into standard operating procedures? What char-
acteristics of a SAF influence the likelihood that actors will consult research-based
evidence regarding the intervention? Our framework directs researchers to consider
current processes of creating change and methods of coordination in use at various
levels of the implementation system, and the extent to which the new practices devi-
ate from the existing practices. Further, latent sources of authority that may be mobi-
lized to legitimate (or thwart) new practices may be identified.
Third, this theoretical framework emphasizes the importance of individual
agency to affect implementation results. The exertion of social skill as a means for
reconciling inherent ambiguities of policy implementation is rather unique to the
SAF framework. For this reason, there are many unanswered questions about the
use of social skill that can be addressed by further inquiry. What factors in the SAF
affect the types of skills exerted by actors to frame problems, shape understandings,
and direct others’ actions in implementing public service interventions? Does this
skill set depend upon the level in the system, familiarity with the sources of author-
ity that are conventionally recognized, or the significance of exogenous shocks?
When actors move across fields, how do they analyze new settings and adjust their
skills and knowledge of relevant authority sources? Does intervention complexity
shape which authority sources are relevant?
The nature of these questions implies that research informed by this frame-
work will likely employ mixed methods (Small, 2011). Investigators should
design their studies in relation to their questions, context, and data availability.
Data can come in different formats—survey responses, field notes, administra-
tive data, audio files, or transcriptions. The means of collecting this data also
vary, from email to mail surveys, single to group observation, single source or
merged administrative records, focus groups, or interviews. Triangulation of
data types and collection methods improves validity and enables more confi-
dence in drawing conclusions from the analysis. And, the analysis procedures—
the means of making sense of the information—are similarly varied. Hierarchical
linear modeling, agent-based modeling, or inductive coding with analytical
memoing all might be appropriate analytical procedures.
Because of the pragmatic concerns of implementation research, it is also quite
possible that fruitful investigations can be developed collaboratively with practi-
tioners, as is often the case in organizational science (e.g., Luscher & Lewis, 2008;
Van de Ven, 2007). Given the significance paid to localized knowledge of a SAF in
this framework, studies that combine the desire to deepen knowledge about imple-
mentation with the applied aim to explain variation in practice—including the
integration of public service interventions into systems and changes in target group
conditions—will be particularly fruitful.
Conclusions
Because policy and program implementation happens in complex systems, a

number of knotty theoretical and practical problems have hindered the development
of scientific knowledge. The SAF Framework for Implementation Research provides
an approach to move inquiry forward. Rather than broadly describing policy proc-
esses, institutional design or public management, our framework narrows in on the
actions and results that surround a particular public service intervention. While the
theory of SAFs is often applied to understand field evolution and development more
broadly, our framework unpacks the field dynamics around the processes and coor-
dination of an intervention. Such a focus is particularly well suited for making sense
of the variations observed in implementation across settings, assessing readiness for
replication of evidence-based interventions, and leveraging the roles of skillful actors
in implementation improvement.
Additionally, this framework conceptualizes formal policy in a slightly dif-
ferent way than other approaches to policy implementation. In our approach,
formal policies are only one source of political authority that shapes implemen-
tation dynamics alongside other sources of authority. Actors within the field
interpret the signals from these authority sources to legitimate field interac-
tions. Sometimes, formal policy is the impetus that leads actors to initiate or
alter a public service intervention, as envisioned by a linear logic of governance
model (Robichau & Lynn, 2009). However, other times, signals from multiple
sources of authority conflict during implementation of a particular interven-
tion, creating controversies that need to be resolved within the implementation
system. No one source of authority is completely deterministic, but rather vari-
ous sources are interpreted and framed by socially skilled actors to legitimate
certain decision options for the collective.
Finally, whereas much of prior implementation research has focused on identify-
ing universal constraints, this framework acknowledges there are unique circumstan-
ces present when implementing an intervention in a particular context. Importantly,
though, it moves beyond the particularities of context to describe a generalizable
social process. Both social structures and human agency are at play in shaping imple-
mentation dynamics and results. These dynamics can be observed and analyzed
across the various settings and the multiple scales present in an implementation sys-
tem. Importantly, the SAF Framework allows scholars to focus on the results relevant
to both management and policy—change in implementation systems, and change the
behavior or conditions of an external target group.
Stephanie Moulton is an associate professor at the John Glenn College of Public

Affairs, The Ohio State University. Her research interests include public policy
implementation, evaluation, and management, with a specialization in housing
and consumer finance policies. Her recent publications appear in journals such as
Journal of Policy Analysis and Management, Journal of Urban Economics, Journal of Pub-
lic Administration Research and Theory, and the Journal of Money, Credit and Banking.
Jodi R. Sandfort is a professor at the Humphrey School of Public Affairs, Univer-
sity of Minnesota. Her research interests include policy and program implementa-
tion and program evaluation, with a specialization in government–nonprofit
relations and social welfare policy. Her recent publications appear in journals such
as Critical Policy Studies, Journal of Public Administration Research and Theory, and the
Journal of Policy Analysis and Management.
Notes
We would like to thank Scott Allard, Peter Hupe, Jennifer Mosley, Edella Schlager, Chris Weible, Andy
Whitford, Craig Smith, and three anonymous reviewers for their helpful comments on earlier drafts of
this article. This paper elaborates ideas at the foundation of our book Effective Implementation in Practice:
Integrating Public Policy and Management (Jossey-Bass, 2015). Order of authorship is alphabetical, as all of
the ideas, and limitations, come from our dynamic collaboration.
1. There is also a stream of complexity theory in public policy that documents policy processes taking
place within complex systems (Anderies & Janssen, 2013; Colander & Kupers, 2014).
2. There are a relatively large number of peer review articles published since 2010 that utilize this theory
(see e.g., Abrutyn, 2012; Ozen & Ozen, 2011). However, virtually all focus applications to societal-
level systems rather than the operational-level application taken here, which focuses on dynamics
within and between SAFs in pursuit of an institutionally sanctioned change.
3. For example, in their book, Fligstein and McAdam (2012) apply SAF theory to help explain how
changes to the mortgage finance system in the United States contributed to the economic crisis of
2008. Their focus is on creation and disruption of SAFs at an abstract level, looking at how the
mortgage-lending crisis destabilized proximate fields. While certainly relevant to public policy, this
account does not provide insight into the more detailed activities involved in policy implementation.
In our application, we would narrow in on a specific intervention, such as mortgages originated to
lower income households, and seek to understand how specific dimensions of the intervention, such
as the terms of the mortgage and underwriting processes, are shaped by the dynamics within the field
and subsequently impact outcomes.
4. This construct resembles the idea of political scientists who focused on local “policy subsystems”
(Milward & Wamsley, 1984) or “localized implementation structures” (Hjern & Porter, 1981). Yet,
because of our training in organizational science and institutional theory, we prefer the terminology
of “policy field” with its distinct theoretical foundation.
5. Others note that in studying complex systems, it is often preferable to articulate general mechanisms
that can be more precisely specified in particular contexts (Colander & Kupers, 2014). This theory ben-
efits from constructs that are loosely framed, so that more specific definition can be undertaken in spe-
cific applications, as we will illustrate below with our IDA example.
6. Policy that operates through the tax system is more regressive than social insurance or other means-
tested policies because low-income people do not have sufficient income to benefit as dramatically
from reductions in tax liability or tax credits unless they are made refundable.
7. The Assets for Independence Act of 1998 provides direct federal funding to support IDAs, and other
programs including the Community Service Block Grant, Office of Refugee Resettlement financing,
and Temporary Assistance for Needy Families, created other federal funding alternatives during the
1990s.
8. At one time, 43 state governments had enacted legislation or administered such programs, although
only 15 states had state-led programs operating in 2013 (CFED, 2014).
9. An example of such a study is Taylor, Rees, and Damm’s (2014) account of implementation of a U.K.
work policy at the macro level.
References
Abrutyn, Seth. 2012. “Toward a Theory of Institutional Ecology: The Dynamics of Macro Structural
Space.” Review of European Studies 4 (5): 167–81.
Anderies, John M., and Marco A. Janssen. 2013. “Robustness of Social-Ecological Systems: Implications
for Public Policy.” Policy Studies Journal 41 (3): 514–37.
Berman, P. 1981. “Educational Change: An Implementation Paradigm.” In Improving Schools: Using What
We Know, ed. R. Lehming, and M. Kane. Beverly Hills, CA: Sage, 253–86.
Blau, Peter Michael, and W. Richard Scott. 1962. Formal organizations: A comparative approach. Stanford,
CA: Stanford University Press.
Bourdieu, Pierre. 1977. Outline of a Theory of Practice. Cambridge: Cambridge University Press.
Bozeman, Barry. 2007. Public Values and Public Interest: Counterbalancing Economic Individualism. Washing-
ton, DC: Georgetown University Press.
Burstein, Paul. 1991. “Policy Domains: Organization, Culture and Policy Outcomes.” Annual Review of
Sociology 17 (1): 327–50.
Corporation for Enterprise Development (CFED). 2014. Resource Guide: State IDA Program Support.
Assets and Opportunity Scorecard. http://assetsandopportunity.org/scorecard/rg_StateIDAProg-
ramSupport2014_f.pdf. Accessed January 18, 2016.
Cohen, David K., and Susan Moffit. 2009. The Ordeal of Equality: Did Federal Regulation Fix the Schools?
Cambridge, MA: Harvard University Press.
Colander, David, and Roland Kupers. 2014. Complexity and the Art of Public Policy: Soliving Society’s Prob-
lems from the Bottom Up. Princeton, NJ: Princeton University Press.
Cromwell, J. E., III, R. D. Colton, S. J. Rubin, C. N. Herrick, J. Mobley, K. Reinhardt, and R. Wilson. 2011.
Best Practices in Customer Payment Assistance Programs [Online]. Denver, CO: Water Research Founda-
tion. http://www.waterrf.org/publicreportlibrary/4004.pdf. Accessed January 18, 2016.
Daft, Richard L. 2010. Organization Theory and Design. Mason, OH: Cengage Learning.
Dahl, R. A., and C. E. Lindblom. 1953. Politics, Economics and Welfare: Planning and Politico-Economic Sys-
tems, Resolved into Basic Processes. New York: Harper & Brothers.
Damschroder, Laura J., David C. Aron, Rosalind E. Keith, Susan R. Kirsh, Jeffery A. Alexander, and Julie
C. Lowery. 2009. “Fostering Implementation of Health Services Research Findings into Practice: A
Consolidated Framework for Advancing Implementation Science.” Implementation Science 4: 50.
deLeon, Peter, and Linda deLeon. 2002. “What Ever Happened to Policy Implementation? An Alternative
Approach.” Journal of Public Administration Research and Theory 12 (4): 467–92.
Denhardt, Robert B., and Janet Vinzant Denhardt. 2000. “The New Public Service: Serving Rather than
Steering.” Public Administration Review 60 (6): 549–59.
DeSanctis, Gerardine, and Marshall Scott Poole. 1994. “Capturing the Complexity in Advanced Technol-
ogy Use: Adaptive Structuration Theory.” Organizational Science 5 (2): 121–47.
DiMaggio, P., and W. Powell. 1991. The New Institutionalism in Organizational Analysis. Chicago: Univer-
sity of Chicago Press.
Downs, Anthony. 1957. “An Economic Theory of Political Action in a Democracy.” The Journal of Political
Economy 65 (2): 135–150.
Durlak, Joseph A., and Emily P. DuPre. 2008. “Implementation Matters: A Review of Research on the
Influence of Implementation on Program Outcomes and the Factors Affecting Implementation.”
American Journal of Community Psychology 41 (3–4): 327–50.
Elmore, Richard F. 1980. “Backward Mapping: Implementation Research and Policy Decisions.” Political
Science Quarterly 94 (4): 601–16.
Feldman, Martha S., and Wanda J. Orlikowksi. 2011. “Theorizing Practice and Practicing Theory.” Orga-
nization Science 22 (5): 1240–53.
Feldman, Martha S., and Kathryn Quick. 2009. “Generating Resources and Energizing Frameworks
through Inclusive Public Management.” International Public Management Journal 12 (2): 137–71.
Fligstein, Neil. 1985. “The Spread of the Multidivisional Form.” American Sociological Review 50: 377–91.
———. 1997. “Social Skill and Institutional Theory.” American Behavioral Scientist 40: 397–405.
———. 2001. “Social Skill and the Theory of Fields.” Sociological Theory 19 (2):105–25.
———. 2008. “Fields, Power and Social Skill: A Critical Analysis of the New Institutionalism.” Interna-
tional Public Management Review 9: 227–52.
Fligstein, Neil, and Doug McAdam. 2011. “Toward a General Theory of Strategic Action Fields.” Sociolog-
ical Theory 29 (1): 1–26.
———. 2012. A Theory of Fields. Oxford, UK: Oxford University Press.
Giddens, Anthony. 1984. The Constitution of Society: Outline of a Theory of Structuration. Berkeley, CA: Uni-
versity of California Press.
Goffman, Erving. 1959. The Presentation of Self in Everyday Life. New York: Anchor Books.
Goggin, Malcolm L., Ann O’M. Bowman, James P. Lester, and Laurence J. O’Toole. 1990. “Studying the
Dynamics of Public Policy Implementation: A Third-Generation Approach.” In Implementation and
the Policy Process: Opening Up the Black Box, ed. by Dennis J. Palumbo, and Donald J. Calista.
New York: Greenwood Press. 181–97.
Gormley, William T., Jr. 1986. “Regulatory Issue Networks in a Federal System.” Polity 18 (4): 595–620.
Grinstein-Weiss, Michal, Michael Sherraden, William G. Gale, William M. Rohe, Mark Schreiner, and
Clinton Key. 2013. “Long-Term Impacts of Individual Development Accounts on Homeownership
among Baseline Renters: Follow-Up Evidence from a Randomized Experiment.” American Economic
Journal: Economic Policy 5 (1): 122–45.
Hasenfeld, Yeheskel. 1983. Human Service Organizations. Englewood Cliffs, NJ: Prentice Hall.
Hill, Heather C. 2003. “Understanding Implementation: Street-Level Bureaucrats’ Resources for Reform.”
Journal of Public Administration Research and Theory 13 (3): 265–82.
Hill, Michael, and Peter Hupe. 2014. Implementing Public Policy: An Introduction to the Study of Operational
Governance, 3rd ed. Thousand Oaks, CA: Sage Publications.
Hjern, Benny, and David O. Porter. 1981. “Implementation Structures: A New Unit of Administrative
Analysis.” Organization Studies 2 (3964): 211–27.
Ingram, Helen, and Anne Schneider. 1990. “Improving Implementation through Framing Smarter Stat-
utes.” Journal of Public Policy 10 (1): 67–88.
Khademian, Anne. 2002. Working with Culture: The Way the Job Gets Done in Public Programs. Washington,
DC: Congressional Quarterly Press.
Kiser, Larry L., and Elinor Ostrom. 1982. “The Three Worlds of Action: A Meta-Theoretical Synthesis of
Institutional Approaches.” In Strategies of Political Inquiry, ed. Elinor Ostrom. Beverly Hills, CA:
Sage.
Latour, Bruno. 2005. Reassembling the Social: An Introduction to Actor-Network Theory. Cambridge, UK:
Oxford University Press.
Lipsky, Michael. 1980. Street-Level Bureaucracy: Dilemmas of the Individual in Public Services. New York, NY:
Russel Sage Foundation.
Luscher, Lotte S., and Marianne W. Lewis. 2008. “Organizational Change and Managerial Sensemaking:
Working through Paradox.” Academy of Management Journal 51: 221–40.
Lynn, Laurence, E., Jr., Carolyn J. Heinrich, and Carolyn J. Hill. 2000. “Studying Governance and Public
Management: Challenges and Prospects.” Journal of Public Administration Research and Theory 10 (2):
233–62.
Martin, John Levi. 2003. “What Is Field Theory?” American Journal of Sociology 109 (1): 1–49.
Matland, Richard E. 1995. “Synthesizing the Implementation Literature: The Ambiguity-Conflict Model
of Policy Implementation.” Journal of Public Administration Research & Theory 5 (2): 145–74.
May, Carl, and Tracy Finch. 2009. “Implementing, Embedding, and Integrating Practices: An Outline of
Normalization Process Theory.” Sociology 43 (3): 535–54.
May, Carl. 2013. “Towards a General Theory of Implementation.” Implementation Science 8 (1): 1–14.
May, Carl R., Frances Mair, Tracy Finch, Anne MacFarlane, Christopher Dowrick, Shaun Treweek, Tim
Rapley, et al. 2009. “Development of a Theory of Implementation and Integration: Normalization
Process Theory.” Implementation Science 4 (29): 1–9.
May, Peter J., and Ashley E. Jochim. 2013. “Policy Regime Perspective: Policies, Politics, and Governing.”
Policy Studies Journal 41 (3): 426–52.
Maynard-Moody, Steven, and Michael Musheno. 2003. Cops, Teachers, Counselors: Stories from the Front
Lines of Public Services. Ann Arbor: University of Michigan Press.
Maynard-Moody, Steven, and Shannon Portillo. 2010. “Street-Level Bureaucracy Theory.” In Oxford
Handbook of American Bureaucracy, ed. Robert Durant. Oxford, UK: Oxford University Press, 252–77.
Mazmanian, Daniel A., and Paul A. Sabatier. 1989. Implementation and Public Policy. Lanham, MD: Univer-
sity Press of America.
McGinnis, Michael D. 2011. “An Introduction to IAD and the Language of the Ostrom Workshop: A Sim-
ple Guide to a Complex Framework.” Policy Studies Journal 39 (1):169–83.
McKernan, Signe-Mary, and Michael Sherraden, eds. 2008. Asset Building and Low Income Families. Wash-
ington, DC: Urban Institute Press.
Mead, George Herbert. 1934. Mind, Self and Society. Chicago: University of Chicago Press.
Mettler, Suzanne. 2007. Soldiers to Citizens: The G.I. Bill and the Making of the Greatest Generation.
New York: Oxford University Press.
Miles, Raymond E., Charles C. Snow, Alan D. Meyer, and Henry J. Coleman. 1978. “Organizational Strat-
egy, Structure, and Process.” Academy of Management Review 3 (3): 546–62.
Mills, Gregory, William G. Gale, Rhiannon Patterson, Gary V. Engelhardt, Michael D. Eriksen, and Emil
Apostolov. 2008. “Effects of Individual Development Accounts on Asset Purchases and Saving
Behavior: Evidence from a Controlled Experiment.” Journal of Public Economics 92: 1509–30.
Milward, H. Brinton, and Keith G. Provan. 1998. “Principles for Controlling Agents: The Political Econ-
omy of Network Structure.” Journal of Public Administration Research and Theory 8 (2): 203–22.
Milward, H. Brinton, and Gary L. Wamsley. 1984. “Policy Subsystems, Networks, and the Tools of Public
Management.” In Public Policy Formation, ed. Thomas A. J. Koonen, and Kenneth Hanf. Boulder,
CO: Westview Press, 105–27.
Moynihan, Donald P., and Sanjay K. Pandey. “The Role of Organizations in Fostering Public Service
Motivation.” Public Administration Review 67 (1): 40–53.
Neal, Patrick. 1988. “Hobbes and Rational Choice Theory.” Western Political Quarterly 41 (4): 635–52.
Nilsen, Per, Christian Ståhl, Kerstin Roback, and Paul Cairney. 2013. “Never the Twain Shall Meet?—A
Comparison of Implementation Science and Policy Implementation Research.” Implementation Sci-
ence 8 (63): 1–12.
Orlikowski, Wanda J. 1992. “The Duality of Technology: Rethinking the Concept of Technology in Organ-
izations.” Organization Science 3 (3): 399–427.
———. 2000. “Using Technology and Constituting Structures: A Practice Lens for Studying Technology
in Organizations.” Organization Science 11 (4): 404–28.
Ostrom, Elinor. 2007. “Institutional Rational Choice: An Assessment of the Institutional Analysis and
Development Framework.” In Theories of the Policy Process, ed. Paul Sabatier. Boulder, CO: Westview
Press.
———. 2011. “Background on the Institutional Analysis and Development Framework.” Policy Studies
Journal 39 (1): 7–27.
Ostrom, Elinor, Michael Cox, and Edella Schlager. 2014. “An Assessment of the Institutional Analysis
and Development Framework and Introduction of the Social-Ecological Systems Framework.” In
Theories of the Policy Process, ed. Paul A. Sabatier, and Christopher Weible. Boulder, CO: Westview
Press.
Ostrom, Vincent. 1980. “Artisanship and Artifact.” Public Administration Review 40 (4): 309–17.
O’Toole, Laurence J. 1986. “Policy Recommendations for Multi-Actor Implementation: An Assessment of

the Field.” Journal of Public Policy 6 (2): 181–210.
———. 2004. “The Theory — Practice Issue in Policy Implementation Research.” Public Administration 82
(2): 309–29.
Ozen, Hayriye, and Sukru Ozen. 2011. “Interactions in and between Strategic Action Fields: A Compara-
tive Analysis of Two Environmental Conflicts in Gold-Mining Fields in Turkey.” Organization &
Environment 24 (4): 343–63.
Perrow, Charles. 1986. “Economic Theories of Organization.” Theory and Society 15 (1): 11–45.
Perry, James L. 2000. “Bringing Society In: Toward a Theory of Public-Service Motivation.” Journal of Pub-
lic Administration Research and Theory 10 (2): 471–88.
Portland State University. 2013. Oregon’s Individual Development Account Initiative: Building Financial Resil-
ience Across Oregon. Evaluation Report. http://oregonidainitiative.org/wp-content/uploads/2013/
10/IDA-REPORT-2013_10-25-13-reduced-size1.pdf. Accessed January 19, 2016.
Pressman, Jeffrey L., and Aaron B. Wildavsky. 1984. Implementation: How Great Expectations in Washington
Are Dashed in Oakland or Why It’s Amazing That Federal Programs Work at All, 3rd ed. Berkeley: Uni-
versity of California Press.
Robichau, Robbie Waters, and Laurence E. Lynn, Jr. 2009. “The Implementation of Public Policy: Still the
Missing Link.” Policy Studies Journal 37 (1): 21–36.
Sabatier, Paul A., and Hank C. Jenkins-Smith. 1993. Policy Change and Learning: An Advocacy Coalition
Approach, ed. Paul A. Sabatier, and Hank C. Jenkins-Smith. Boulder, CO: Westview Press.
Salamon, Lester. 2002. The Tools of Government: A Guide to the New Governance. New York: Oxford Univer-
sity Press.
Sandfort, Jodi. 1999. “The Structural Impediments to Human Service Collaboration: The Case of Front-
Line Welfare Reform Implementation.” Social Service Review 73 (3): 314–39.
———. 2003. “Exploring the Structuration of Technology within Human Service Organizations.” Admin-
istration & Society 34 (6): 605–31.
———. 2010.”Human Service Organizational Technology.” In Human Services as Complex Organizations,
2nd ed., ed. Yeheskel Hasenfeld. New York: Sage, 269–90.
Sandfort, Jodi, Sally Coleman Selden, and Jessica Sowa. 2008. “Do the Tools Used by Government Influ-
ence Organizational Performance? An Examination of Early Childhood Education Policy
Implementation.” American Review of Public Administration 38 (4): 412–38.
Schneider, Anne Larason, and Helen Ingram. 1997. Policy Design for Democracy. Lawrence: University of
Kansas.
Schreiner, Mark, Margaret Clancy, and Michael Sherraden. 2002. Final Report: Saving Performance in the
American Dream Demonstration: A National Demonstration of Individual Development Accounts. St. Louis,
MO: Center for Social Development.
Scott, W. Richard. 1985. “Systems within Systems.” American Behavorial Scientist 28: 601–18.
———. 2008. Institutions and Organizations: Ideas and Interests, 3rd ed. Los Angeles, CA: Sage Publications.
Selznick, Phillip. 1949. TVA and the Grass Roots: A Study of Politics and Organization. Berkeley: University
of California Press.
Shapiro, Thomas and Edward Wolff, eds. 2001. Assets for the Poor: The Benefits of Spreading Asset Owner-
ship. New York: Russell Sage Foundation.
Sherraden, Michael. 1991. Assets and the Poor: A New American Welfare Policy. Armonk, NY: M. E. Sharpe.
———. 2000. “From Research to Policy: Lessons from the Individual Development Accounts.” Journal of
Consumer Affairs 34 (2): 159–81.
Small, Mario. 2011. “How to Conduct a Mixed Methods Study: Recent Trends in a Rapidly Growing Lit-
erature.” Annual Review of Sociology 37 (1): 57–86.
Soss, Joe. 2005. Making Clients and Citizens: Welfare Policy as a Source of Status, Belief and Action. Albany:
State University of New York Press.
Stafford, Benjamin, and Elizabeth Wilson. 2015. Winds of Change: Socio-Technical Processes Shaping Electrical
Policy Implementation. Minneapolis: Humphrey School of Public Affairs, University of Minnesota.
Stone, Melissa, and Jodi R. Sandfort. 2009. “Building a Policy Fields Framework to Inform Research in
Nonprofit Organizations.” Nonprofit and Voluntary Sector Quarterly 38 (6): 1054–75.
Taylor, Rebecca, James Rees, and Christopher Damm. 2014. “UK Employment Services: Understanding
Provider Strategies in a Dynamic Strategic Action Field.” Policy & Politics, doi:10.1332/
030557314X14079275800414
Van de Ven, Andrew. 2007. Engaged Scholarship: A Guide for Organizational and Social Research. New York:
Oxford University Press.
Wajcman, Judy. 2006. “New Connections: Social Studies of Science and Technology and Studies of
Work.” Work, Employment & Society 20 (4): 773–86.
Weible, Chris, Paul Sabatier, and Kelly McQueen. 2009. “Themes and Variations: Taking Stock of the
Advocacy Coalition Framework.” Policy Studies Journal 37 (1): 121–40.
Weick, Karl E. 1995. Sensemaking in Organizations. Thousand Oaks, CA: Sage Publications.
Weimer, David L. 2008. “Theories of and in the Policy Process.” Policy Studies Journal 36 (4): 489–95.
Whitford, Andrew. 2013. “Information and Uncertainty in Policy Implementation: Evidence from the
Implementation of EPA Waivers.” Journal of Public Administration Research and Theory 24: 267–88.
Yanow, Dvora. 1996. How Does a Policy Mean? Interpreting Policy and Organizational Actions. Washington,
DC: Georgetown University Press.

SAF Moulton2016

Uploaded by

Copyright:

Available Formats

SAF Moulton2016

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

SAF Moulton2016

Uploaded by

Copyright:

Available Formats

bs_bs_banner

The Policy Studies Journal, Vol. 00, No. 00, 2016

The Strategic Action Field Framework for Policy

Understanding the Complex Social Systems Involved in Implementation

Broadly speaking, social systems are contexts organized by relations among

Conceptualizing a New Framework for Implementation Research

stability in the public service intervention. Through providing a theoretical framework,

Component Elements Examples of Variation

Public Service Interventions and Results

similarities but possessing unique power, resources, or dominant cultural interpreta-

Assumption #2: SAFs at multiple levels of a complex implementation system con-

Activating Legitimate Authority to Shape Implementation Practice

Implementation scholars have long recognized the importance of formal and

(Khademian, 2002). Norms associated with professional expertise may be a particu-

The significance of any particular authority source depends on how it is

providing an explanation for why change or stability surrounds the implementation

Figure 1. Framework of Strategic Action Fields in Complex Implementation Systems

Putting It All Together

The three components of the framework offer a way to conceptualize change

Applying the Framework: An Illustrative Case

Focus: Public Service Intervention

asset development through home purchase, business start-up, or higher education

Scale: Multilevel SAFs

Stability and Change: Authority and Social Skill

Using the Strategic Action Framework for Implementation Research

Because policy and program implementation happens in complex systems, a

Stephanie Moulton is an associate professor at the John Glenn College of Public

O’Toole, Laurence J. 1986. “Policy Recommendations for Multi-Actor Implementation: An Assessment of

You might also like