Abstract
Implementation remains a somewhat neglected dimension within urban planning research after a surge of interest during the 1960s and 1970s in contrast to related cognitive disciplines, particularly healthcare research which has seemingly grappled with the phenomenon with more gusto. Here, we assess the state of urban planning implementation science and compare these efforts with healthcare frameworks explaining implementation to ascertain what new insights can be gleaned on how we investigate the phenomenon.
Urban planning theory composition is the scholarly act of explaining and representing motions of an applied practical discipline (Taylor, 1998). Such theories have been categorised as either substantive or procedural; while the former paints an overarching picture of practice, the latter offer something of a road map for improving practice (Sandercock, 1998). The relationship between theory and practice in urban planning has been somewhat fractious, with practitioners often deriding scholarly insight as too academic and insufficiently applied (Cullingworth and Nadin, 2006; Taylor, 1998). This problem of theoretical representation is perhaps hindered by urban planning practice being a somewhat ill-defined and schizophrenic spatial place making activity which resists definition (we reference it here broadly; encompassing strategy development and ensuing regulatory exercises).
These issues become manifest in studying implementation; a complicated, cluttered, political process 1 of enacting policies, plans and programmes to the extent one scholar laments “to plan is human, to implement is divine” (Gilg, 1996: 15). The problem for scholars being adjudging a plan or policy as successful is an inherently political judgement and often requires theorists to take sides and interact with the contextual conditions they study (Laurian et al., 2004; Talen, 1996; Young et al., 2002). This can include expressing views on manifesto commitments, international treaties, institutional histories, local knowledge and past legacies (Vigar et al., 2000).
Implementation therefore seemingly defies linear and formulaic explanation so insight is subsequently often substantive, with observational, descriptive and normative theories presenting interpretive frameworks which attempt to explain events and actions (Gilg and Kelly, 1997). Landmark urban planning theories pertaining to implementation have notably not advanced since the 1970s and 1980s with the phenomena instead being investigated on the periphery of other conceptual constructs including communication (e.g. Healey, 1997) and integration (e.g. Smith et al., 2014). Other scientific disciplines, including healthcare, political science, public administration and management have vigorously explored implementation with more gusto and a what works trans-disciplinary agenda has emerged which is largely divorced from urban planning research (Davies et al., 2000; What Works Network, 2014). Healthcare scholars in particular have readily debated implementation; one systematic review identified 6000 titles and abstracts of scholarly work on the subject (Greenhalgh et al., 2004) while the Implementation Science peer reviewed journal has published scholarly articles to aid advancement since 2006 (Nilsen, 2015). Unlike urban planning, healthcare is notably more discernable with conspicuous objectives; specifically to improve the human condition by enhancing quality of life and longevity (often at the expense of all other considerations) (Davies et al., 2000; Goldacre, 2008).
This paper reports an investigation to compare implementation theories and ideas urban planning scholars have advanced with similar efforts in healthcare to ascertain what new insights might be gleaned both on the phenomenon itself and how theorists represent it. Healthcare was selected for comparison as scholars here also grapple with thorny issues of putting plans and policies into practice and have amassed a sizable body of work on the subject. We begin by reflecting upon implementation conceptualisation within urban planning research to delineate existing thought and ideas and identify a series of overarching themes. Four recent healthcare implementation conceptual frameworks were then reviewed to synthesise how they differ from urban planning perceptions on implementation and what this tells us about studying the phenomenon. To date, little comparison has been made between the two disciplines work on implementation. As implementation pertains to enacting something in someway, an improved understanding of such processes can potentially identify possibilities for preventing wasteful failures and promoting best practices. This paper is therefore of international interest to anyone investigating the pursuit of sustainable and improved governance.
Development of implementation theory in urban planning
Evaluating the intended ends of a plan remains a disputed element of studying urban planning implementation as (a) what constitutes a successful outcome is an inherently political judgement (Baker and Hincks, 2009; Rydin, 2011), and (b) policies, plans and programmes pursue different types of outcome (Fischer, 2007; Gilg and Kelly, 1997). These divisions lie at the nexus of urban planning theory debates on the purpose of planning; particularly the extent it represents a visionary or regulatory activity.
Adjudging an intervention to be successful is relative to our disposition and perspective, for policies, plans or programmes might deliver benefits to some and increase hardship to others (Cullingworth and Nadin, 2006). While Wildavsky (1973) notoriously argued an all or nothing assessment in defining successful implementation, policies, plans or programmes will typically achieve some but not all intended outcomes; neither an entire success nor outright failure (Alexander and Faludi, 1989; Talen, 1996). For instance, compromising certain objectives to ensure some degree of delivery is often more desirable than failing to achieve any positive outcomes and thus ought not to be regarded as failure (Clifford and Tewdwr-Jones, 2014). Congruently, some problems defy resolution (wicked problems, Rittel and Webber, 1973) so a successful response will be less negative rather than entirely positive. The subtleties of successful outcomes are also manifest in the politics of competing priorities which can result in optimal strategies for one policy domain being pursued at the expense of such strategy in another (Healey, 2010; Rydin, 2011). To illustrate this point, Smith (2014) describes how a series of trade offs between willing parties is necessary to produce successful spatial policies in local government. Temporal elements also cloud our understanding of successful outcomes as interventions can play out over time to the extent perceived failures can later be viewed as successes (and vice versa) once underpinning assumptions become outdated and discredited (Stead et al., 2003; Taylor, 1998). Policy can similarly persist despite the intended objectives being forgotten (Gilg and Kelly, 1997; Healey, 2010). For example, several light rail schemes in the UK were initially criticised for being delivered late, over budget and not achieving intended levels of return but are now viewed as flourishing examples of integrated transport (Smith et al., 2014).
In respect of outcome types, we differentiate between planning implementation and plan implementation. Plan implementation describes physically achieving construction (or equally prohibiting certain activity) while planning implementation references utilising material (such as other documents) during plan making processes (Alexander and Faludi, 1989; Fischer et al., 2013; Talen, 1996). This distinction is important as well integrated plans (planning implementation) can be poorly delivered (plan implementation), as the unfortunate Merseytram light rail proposal in Liverpool illustrates (Smith et al., 2014).
This complex affinity between plan making and delivery questions rational planning assumptions espousing a linear progression from policy to action (Friedman, 1969; Lindblom, 1959; Rittel and Webber, 1973). Such thinking presumed a sequential series of steps with implementation a specific stage occurring after plan making and falling outside planning’s remit. In essence, it represents a theorist’s ideal world; divorced from the messy, pragmatic conditions of practice. Multifaceted implementation processes therefore expose such standpoints as simplistic representations (Gilg and Kelly, 1997; Rydin, 2011; Young et al., 2002), although discredited rational planning rhetoric is still observed to persist within some planning practices (Cullingworth and Nadin, 2006).
Implementation as fit
Implementation has been beheld to require sufficient fit between agencies and entities utilising and delivering policies, plans and programmes (Fischer et al., 2013; Hull, 2008). Pressman and Wildavsky hinted at fit in their seminal 1973 reflective study “Implementation; how great expectations in Washington are dashed in Oakland” examining how and why federal government policies were not delivered as expected in towns and cities. Their observations implied implementation depended upon delivery chains between different groups of actors which faltered from weak linkages (or implementation deficit, Allmendinger and Thomas, 1998). Likewise, longer and more complex chains (a consequence of competition and the number of actors involved, Riot, 2014) offer greater potential for fragile linkages to buckle under pressure, so a degree of fit is essential between the intervention or mechanism being enacted and the local context of a practiced environment to ensure links are suitably coupled and support forthcoming.
The actors Pressman and Wildavsky (1973) describe reference individuals working in a variety of theatres at different spatial scales for contrasting sectors; for instance, planning policy implementation can require action by actors representing both land use and transport sectors operating at different scales of governance (district and sub-regional) (Smith, 2014; Stead et al., 2003), while others have highlighted interaction and diverging interests between the public and private sectors in respect to housing (Biddulph et al., 2003). Each actor subsequently represents unique interests, objectives and requirements which might prove incompatible with other partners. Tensions are also observed between layers of government, such as misalignment of strategic and local priority’s which can result in disagreements over the suitability of policies and plans (Fischer et al., 2013; Richer and Hasiak, 2014).
Pressman and Wildavsky’s (1973) complex delivery chains also obscure the appropriateness of a policy or plan. Hogwood and Gunn (1984) suggest implementation failures result from “bad execution, bad policy or bad luck” (p. 197). While bad execution and bad luck pertain to delivery chain failures, bad policy represents a separate phenomenon in that policy based on flawed thinking (i.e. a bad policy) never delivers successful outcomes no matter how well implemented; “every policy incorporates a theory of cause and effect (normally unstated in practice) and if the policy fails, it may be the underlying theory that is at fault rather than execution of the policy” (Hogwood and Gunn, 1984: 201). This supposition references past metaphors by theorists equating urban planning to the scientific method, with policy akin to hypothesis in both indicate particular outcomes result from pursuing certain actions/interventions (Magee, 1973; Rittel and Webber, 1973). Such thinking can be discredited as another rational planning theory delusion; in a complex, multifaceted context, policy frequently does not repeat the same outcomes for everyone all the time (Booth, 2011). This fundamental problem also hampers agreements between actors on the veracity of cause and effect patterns as project partners can disagree on the assumptions behind scenario driven modelling exercises testing alternative options (Smith et al., 2014).
Implementation direction
Pressman and Wildavsky’s (1973) work remained wedded to a rational view of implementation in that (a) policy was perceived hierarchical and synonymous with top-down government so local actors remit extended only so far as delivering centralised objectives (Barrett and Fudge, 1981), and (b) policy is still thought to cause rather than result from actions (Vigar et al., 2000). While this criticism might indicate bottom-up governance is more suited to initiating implementation, questions persist on the extent such actions are isolated and independent (Allmendinger and Thomas, 1998; Smith, 2014). To investigate bottom-up governance constraints, urban planning scholars observe linkages between planning and policy mechanisms with contextual environments rather than accentuate models and frameworks which translate knowledge into action based on naturalistic presumptions of causality.
Contexts and mechanisms
Some urban planning theorists frame synergy between contexts and mechanisms as ontology; the role of structure and agency in achieving outcomes, particularly how local authority actors perform their duties in a strategic policy vacuum (Biddulph et al., 2003; Smith, 2014; Vigar et al., 2000). While structure denotes formal and informal entities shaping actions (such as national policy or professional practices and values), agency references how individuals interpret and respond to the world, potentially acting on their own recognisance (Smith, 2014). Structure and agency are not so much alternatives but represent interlinked constructs sustained by social relations (Vigar et al., 2000), thereby highlighting threads between context, mechanism and outcome are intricate and multifaceted.
Contexts and mechanisms have also been explored through communication and discourse theories. In particular, Healey (1997, 2010) espoused a form of communicative rationality proposing a set of principles underpinning successful communication in community engagement based on the works of German philosopher Habermas. Gilg and Kelly (1997) similarly frame implementation as a negotiation or bargaining process between different interest groups in their reflections on rural planning and development regulation, but advise such activities implicate the dynamics of power and conflict to achieve consensus, control and compliance rather than pertain to open discourses exchanging ideas.
Availability and access to the resources needed for implementation is another aspect of context and mechanism, for a partner can wield considerable ascendancy by controlling admission and supply to vital elements, as observed in studies of infrastructure and local development framework delivery (Baker and Hincks, 2009; Clifford and Tewdwr-Jones, 2014). Resources can be defined widely, including finance, skills and specialist expertise plus the tools and legislative remit required for implementation (Fischer et al., 2013; Stead et al., 2003).
More recently, urban regime theory from the United States examined implementation through a wider lens of power and change execution. City/regional governance is represented here as partnerships between public sector capacity and nongovernmental (often commercial) resources to deliver a range of policy goals including regeneration and inward investment. Access and control of power is identified as a key facet of instigating transformative action (Mossberger and Stoker, 2001; Stoker and Mossberger, 1994).
Discourse can also challenge the balance of power by empowering communities and enfranchising bottom up decision making against authoritarian elites (Clifford and Tewdwr-Jones, 2014). The United States advocacy planning movement focused on facilitation and representation as a means for civil society advocates to deliver aspirations of disenfranchised communities (Davidoff, 1965). In this case, their efforts became problematic as the demands of disaffected New York communities proved subjective and challenging for civil society advocates to interpret and relate to (Taylor, 1998). The advocacy planning movement thus highlights the challenge of translating theoretical ideals to the disordered and indistinct practice environment.
Decline of urban planning implementation research
Urban planning scholars have offered various explanatory implementation theories which attempt to enlighten our understanding of planning and plan implementation (Clifford and Tewdwr-Jones, 2014). While considerable interest was expressed in such insight during the 1970s and 1980s in Anglophone countries this preoccupation has waned (deLeon and deLeon, 2002) possibly for several factors;
Theorists might believe they understood the problem of implementation and moved onto devising normative solutions, such as collaborative discourse (Healey, 1997) or integration (Fischer et al., 2013), The top-down policy hierarchy represented by established implementation theory’s fell out of vogue (Clifford and Tewdwr-Jones, 2014), Implementation investigation inevitably explores pessimistic constraints which fell foul of a more avant-garde, positive, place making scholarly agenda (Rydin, 2003), The highly contextual subject matter was adverse to generalised theory development so lacked appeal in increasingly positivistic, technocratic sciences (deLeon and deLeon, 2002), The perpetual difficulty in defining the scope, aim and purpose of urban planning (Sandercock, 1998) perhaps restricted debate due to uncertainty in expounding on successful delivery, Neo-liberal doctrines favoured by western governments presumes markets are the only effective way of delivering policy (Clifford and Tewdwr-Jones, 2014) thereby reducing appetite for insight into implementation and what else might work, and Planning practitioners still frequently persist implementation afflicts developers and regulators rather than planners while shunning scholarly insight so are seemingly disinterested in demanding fresh understanding (Cullingworth and Nadin, 2006).
Implementation perception within urban planning research is subsequently somewhat spurned and neglected despite associated practical problems remaining resolute.
Recurring elements within planning implementation
Summation of urban planning theorists implementation ideas within urban planning research.
Healthcare implementation research
While urban planning scholars have neglected implementation, healthcare theorists have extensively investigated the phenomenon 2 (Nutley et al., 2003). Within healthcare, interventions comprise measures 3 for advancing patient care by reducing mortality and prolonging longevity (Davies et al., 2000). Implementation is subsequently an established and mature research topic within this international field and provides a useful counterpoint to urban planning scholars efforts at representation, especially as both traditions espouse normative theories of practice (Goldacre, 2008; Sandercock, 1998) and question rational and linear assumptions of knowledge transferring into action (Nutley et al., 2003).
Healthcare implementation frameworks examined.
We selected four recent frameworks published in leading healthcare academic journals (Box 1) with aid from two respected international healthcare implementation scholars. The review process involved first establishing the means of construction (including methodology and foundation) before describing the notions and ideas in each framework. Our synthesis was achieved by taking themes established in Table 1 and asking, “this framework is the same except that…” We then reflected upon how and why differences occur to adjudge the implications for the study of implementation.
Promoting Action on Research Implementation in Health Services (PARIHS) framework
The PARIHS framework presents a multidimensional view of the complex factors for implementing evidence based practice (Rycroft-Malone et al., 2004). Developed in the late 1990s (although not then called PARIHS) by practiced scholars drawing on their experiences of practice improvement and guideline implementation through the use of case studies (Kitson et al., 1998), subsequent refinement was achieved through empiric testing in a variety of healthcare settings, notably an investigation of practice development in a clinical case study (Rycroft-Malone et al., 2004). A 2010 review identified 24 published studies employing the framework as a means of analysis and a basis for intervention design (Helfrich et al., 2010).
This framework proposes; “successful implementation is dependent upon the nature of the evidence being used, the quality of context, and, the type of facilitation required to enable a successful change process” (Rycroft-Malone et al., 2004: 914), although successful implementation remains undefined.
Evidence is considered; “socially and historically constructed” (Rycroft-Malone et al., 2004: 921) and goes beyond conventional ideas (within healthcare) of evidence equating to scientific research. Instead; “the constituents of ‘evidence’ should encompass other forms of information that is derived from evaluating practice” (Rycroft-Malone et al., 2004: 920) thereby referencing a broad range of knowledge, including scientific research, clinical experience, patient experience and local intelligence. This construct appears translatable to urban planning scholars observations suggesting knowledge from a diverse array of evidential sources is employed to support and inform decision making (Cullingworth and Nadin, 2006; Vigar et al., 2000). Within PARIHS, evidential quality is assessed in terms of methodological rigor/astuteness, acknowledgement of weaknesses and uncertainties, its relevance and the value offered. 4 PARIHS also recognises evidence forms only part of a decision making processes which is commensurate with urban planning theorists indicating evidence informs debates and the politics of decision making (Rydin, 2011).
Context describes the setting where implementation occurs; “getting evidence into practice involves more than identifying high quality research evidence. The idea of relevance, organizational ‘fit’ and adequate resources highlight a dependence on an organization’s political and contextual agenda and investment” (Rycroft-Malone et al., 2004: 921). This construct appears familiar to urban planning scholars who understand planning within a social, political and economic context (Rydin, 2003). PARIHS suggests this construct includes the receptivity of a context to change, organisational/professional cultures and values, leadership style and the role of evaluation in offering feedback and fostering a learning organisation. These factors seem more restricted in scope to urban planning scholars observations (i.e. omitting overt political, spatial and temporal contexts), perhaps because PARIHS concerns implementing evidence based practices rather than representing holism.
Facilitation references expediting successful implementation; “Facilitators have the potential to work with individuals and teams to articulate these issues [of evidence and context], and enable the development and implementation of strategies that acknowledge and incorporate these factors” (Rycroft-Malone et al., 2004: 921). This construct also appears translatable to urban planning theorist’s observations of individuals expressing knowledge to enact transformation through communicative discourse (Healey, 1997). The framework also explores the extent actors perform tasks for others or educate to enable which is perhaps a new element to urban planning investigation; although urban regime theory explores empowerment (Mossberger and Stoker, 2001) and advocacy approaches consider provisioning (Davidoff, 1965).
The PARIHS framework offers a set of conditions or a nucleus of a theory for understanding implementation. They are expansively defined within the framework so further empirical study can refine these conditions in offering deeper insights for a particular example; the overall strength therefore is perhaps more than the sum of the parts. The difficulty with this approach is an overall lack of prescription; by avoiding generalised assumptions the framework condenses many complex and multifaceted components together to the point they become incidental. Likewise, the failure to define successful implementation creates ambiguity for scholars although the framework is perhaps being idealistic in presenting what implementation ought to be rather than how it is.
General theory of implementation
May’s (2013) general theory attempts to provide foundation for understanding, designing, predicting and evaluating implementation by offering researchers and practitioners a set of conceptual tools. The general theory combines elements of normalization process theory, sociological theories of social systems and social cognitive theories from classical psychology. It is based on a meta-analysis and uses the theory ladder climbing notion to construct an all encompassing high level or grand theory by combining findings from more diminutive attributes (Nilsen, 2015), and was tested on a worked example of clinical guideline implementation. The central premise suggests social mechanisms activated and procured by individual performances are responsible for determining implementation processes; “implementation expresses ‘agency,’ and should be understood and evaluated against the problem of how human agents take action in conditions of complexity and constraint” (May, 2013: 2). There appears affinity between these propositions and ontological ideas of structure and agency urban planning scholars use to explain transformation. May offers a conceptualisation based on three independent propositions (Capability, Capacity and Potential) feeding into a fourth (Contribution).
Capability resonates with notions of fit and integration by referencing an individual agent’s ability to introduce a complex intervention based on the potential fit between an intervention and work context; “If workability and integration cannot be sustained, then the embeddedness of the complex intervention will be threatened as the capacity of agents to employ it is confounded” (May, 2013: 5).
Capacity alludes to requirements for resources and communication by considering absorption of a complex intervention within a social system and connects with the availability of social-structural resources (including social networks) to individuals; “The incorporation of a complex intervention within a social system depends on agents’ capacity to cooperate and coordinate their actions” (May, 2013: 6).
Potential describes translating capacity into collective action; “If potential cannot be sustained, then the embeddedness of the complex intervention will be threatened as agents’ commitments are withdrawn” (May, 2013: 8). May delineates this as dependent on individual potential to enact complex interventions using socio-cognitive resources, corresponding to shared commitments, intentions attitudes and beliefs in favour of change. The danger being a lack of ambition or inertia can become prevalent if implementation failure is feared.
Contribution recounts the perpetual input individuals provide to ensure continuity of an intervention over time and space; “If contribution cannot be sustained, then the embeddedness of the complex intervention will be threatened as agents’ efforts diminish” (May, 2013: 9), thereby intimating implementation is a continual process rather than a singular event.
The general theory perceives implementation as an inherently individual led activity for people orchestrate and sustain change in organisations. Congruent ontological notions feature in some urban planning implementation studies seeking to represent transformative change (Biddulph et al., 2003; Gilg and Kelly, 1997) and also planning in a vacuum where the state withdraws (Smith, 2014; Vigar et al., 2000). There are also interesting parallels with the nuances behind urban regime theory for both perceive individual actors responding to changes by forming regimes offering empowerment and facilitating activity (Stoker and Mossberger, 1994). The categories of capability, capacity and potential are thus quite appropriate in defining what investigators should observe in the challenges actors face when enacting change and asserting their dominance, although from a methodology standpoint the perpetual element of contribution is troubling as it implies implementation is a never ending process so scholars only examine brief snapshots. May’s assertion individuals assume the primary role in implementation processes seemingly overlooks several aspects pertaining to implementation which is a more nuanced phenomenon, as Smith (2014) indicates actors agency is still shaped by underlying structures including norms and values instilled from professional training, past legacies and decision support instruments.
Framing the conceptualisation as a general theory is also debateable since it is based on positivistic assumptions a phenomenon always behaves in the same manner. Such a proposition seems inappropriate for interpreting complex, multidimensional phenomena where the same mechanisms can produce very different outcomes based on contextual conditions (Booth, 2011; Nilsen, 2015). Above all, this theory appears somewhat sanitised and unreflective of the indistinct, practical realities of urban planning implementation.
Consolidated framework for implementation research (CFIR)
This framework resulted from an academic literature review of published, peer reviewed studies concerning implementation and innovation
5
in health service delivery and organisation with a view to integrate and consolidate a range of emerging theories, thereby providing an overarching typology for establishing what works, where and why (Damschroder et al., 2009). It aims to achieve consistency in language and terminology, promote comparison research and stimulate new theories on implementation while providing a road map for implementation which sees far into the distance. A systematic review found 26 articles using CFIR, mostly to evaluate implementation but some employed the framework to design interventions (Kirk et al., 2016). The framework (Figure 1) is based on five interactive domains; the intervention, inner and outer setting, individuals involved and the implementation process.
The Consolidated Framework for Implementation Research (CFIR).
The intervention references characteristics of the implemented article, particularly their adaptability. CFIR suggests interventions comprise a core impervious to change and an adaptable periphery permitting modification to suit particular settings without destroying integrity. The degree of compatibility or fit between an intervention and inner setting is an essential consideration, with poor fit occurring where resistance is encountered (left side of Figure 1). These ideas are comparable to observed notions of compromise, negotiation and bartering with stakeholders in urban planning practice to secure desired outcomes without losing sight of core policy intentions (Biddulph et al., 2003).
The inner setting portrays context where implementation processes occur, like an organisation and include internal structural, political and cultural attributes. The outer setting describes the wider economic, political and social contextual aspects in which organisations exist. The authors suggest the two settings are somewhat co-dependent; “the line between inner and outer setting is not always clear and the interface is dynamic and sometimes precarious” (Damschroder et al., 2009: 5), and seemingly allude to similar notions of multiple partners in urban planning implementation study.
Individuals involved references actors participating in implementation processes who are implicit with an intervention. CFIR proposes some individuals express authority and magnetism while purveying organizational and professional practices, norms and values using familiar ontological notions to urban planning scholars; “Individuals have agency; they make choices and can wield power and influence on others with predictable or unpredictable consequences for implementation” (Damschroder et al., 2009: 5).
Implementation process alludes to changes required for an intervention to be used by individuals and within organisations. CFIR advocates these processes as a series of related but non-sequential sub-procedures which may occur simultaneously at multiple levels within an organisation; “These sub-processes may be formally planned or spontaneous; conscious or subconscious; linear or nonlinear, but ideally are all aimed in the same general direction: effective implementation” (Damschroder et al., 2009: 6). These notions seem congruent with integration in urban planning investigation although the description here fails to encapsulate the often fraught nature of such activity.
CFIR actively combines multiple theoretical frameworks employed in past empirical study to present an underlying model on the basis complex multifaceted phenomenon like implementation cannot be understood comprehensively by a single theory. Urban planning theorists have also attempted connecting theories in their previous endeavours, although these efforts are somewhat less ambitious in comparison to CFIR which tries to capture how individuals operate within structures and are influenced by other structures to construct interventions which meet local needs and address wider objectives in a complex, multidimensional environment. CFIR thereby ranges from governance to individual communication and dialogue.
The difficulty with utilizing CFIR is perhaps how much weight to apportion to the different theories it covers; while some will be pivotal, others are incidental. Like PARIHS, CFIR is primarily substantive rather than procedural and so seemingly overlooks several practiced aspects, including the provenance of an intervention and its subsequent relevance and acceptability. It is equally unclear on processes for achieving fit encompassing negotiation and communication together with the inevitable intrusion of power and politics. In this respect CFIR offers an overview which is somewhat hazy on the more contentious implementation elements.
CFIR’s preference for a snug fit between mechanism and context is another debatable point for urban planning scholars suggest this not always practicable due to resource constraints and conflicting agendas meaning a less than ideal intervention is sometimes the best which can be achieved (Biddulph et al., 2003; Clifford and Tewdwr-Jones, 2014). This is perhaps indicative of the pragmatic practices urban planning scholars observe and again suggests CFIR’s assumptions are possibly too idealistic.
Diffusion of innovations in service organizations
This landmark unifying framework resulted from a systematic review of innovation in service delivery and organisation within the healthcare sector (Greenhalgh et al., 2004; Nilsen, 2015). Combining findings from 495 sources, it describes how innovations are accepted within healthcare services via diffusion and dissemination factors (Figure 2), drawing inspiration from classically inspired marketing economics and rural sociology theories on modernising farming practices (Rogers, 2003). A meta-narrative approach was adopted to achieve a synthesis which involves telling the story of research in a particular discipline. The central premise implies interventions are adopted (or rejected) by individuals at different rates through diverging responses and actions transcending established scientific fields. The framework is subsequently constructed upon ontological ideas of structure and agency in elucidating how and why change occurs, and is intended to illuminate rather than provide answers. It has four underlying elements; user system, resource system, change agency and knowledge purveyors which exist within an outer context (Figure 2), and was tested on four examples (telemedicine, integrated care pathways, General Practitioner fund holding, and electronic patient record keeping).
Diffusion of Innovations in Service Organizations Framework.
The resource system represents processes whereby knowledge purveyors develop and disseminate innovations to change agents who affect transformation by diffusing the innovation within a practice user system. The outer context describes external influences determining whether an innovation is adopted and sustained, including wider inter-organizational collaborations and networks. Strong linkages are required between the different components.
The framework insinuates individuals are a key component in adopting innovations; People are not passive recipients of innovations. Rather (and to a greater or lesser extent in different persons), they seek innovations, experiment with them, evaluate them, find (or fail to find) meaning in them, develop feelings (positive or negative) about them, challenge them, worry about them, complain about them, “work around” them, gain experience with them, modify them to fit particular tasks, and try to improve or redesign them—often through dialogue with other users. (Greenhalgh et al., 2004: 598)
The alternative influences determining how an innovation permeates through an organisation are thought to exist on a continuum. Here, diffusion represents an end where unplanned, informal and decentralised exertion spreads an innovation, while planned, formal and centralized animation persists in active dissemination at the opposite extreme.
These aspects of the framework present several analogies to urban planning implementation research, particularly the influence of external contexts on decision making and the power of individuals in shaping agendas. It is also interesting to reflect upon these implications for how innovations become institutionalised with professional practices (March and Olsen, 1989).
Within the user context, the framework proposes a change sequence (Figure 2);
System antecedents for innovations represents how some organisational features (both structural and cultural) help successfully assimilate an innovation. As different organisations provide contrasting arenas for realising change, no universal or generalised organisational features for achieving successful implementation are thought to exist. System readiness for innovation denotes whether an organisation is amenable to change. Organisations may be unwilling to assimilate a particular innovation but receptive to changes in general. Implementation and routinization of an innovation references maturing from merely considering an innovation to indoctrinating it into systems and routines. This is believed to occur in non-linear processes including multiple shocks, setbacks and unanticipated events. Assimilation by the user system considers the adoption setting of an innovation, such as a team, department or organisation. Within these units, alterations are made to existing structures and working practices including formal decision making procedures, evaluation periods and devices for sustaining change (e.g. learning and development). Individual adoption of an innovation therefore represents just one component of assimilating a complex innovation within an organisation. The final stage represents consequences of implementation processes, experience of which informs the first stage again through a feedback loop.
The diffusion of innovations in service organizations framework offers a comprehensive assessment of implementation processes. This is perhaps inevitable given the nature of systematic review; it examines and combines all relevant studies to expand the range of their findings. The difficulty is utilizing such techniques on phenomenon which are highly contextual and unique (like implementation) creates frameworks which are overly descriptive and can lack relevance.
As per the other reviewed frameworks, this example combines several theories to try and represent the phenomenon. It particularly integrates ontological aspects of transformation with organisational alteration and the legitimisation of change. Above all it demonstrates these processes are somewhat iterative and complex, involving trial, reflection and adaptation by individuals in addition to creating linkages and synergies between disparate parts.
There are perhaps inevitably several questionable assumptions in this framework; particularly the inclusion of a rational, linear model of knowledge to action in Figure 2 while describing something more iterative in the text. Furthermore, change is assumed to be initiated by actors rather than a response to wider initiatives and externalities. This also leads to a propensity to perhaps undervalue the role of structures in shaping individuals actions such as professional training and ethics. Like the other healthcare frameworks, it is substantive and overlooks disordered practiced aspects including power and politics in transformation. This framework additionally focuses on organisations despite implementation potentially occurring outside of such formal entities, such as in communities. We also question whether a unified theory as presented here is possible given the multitude of competing and contrasting ideas within urban planning pertaining to the nature and role of the State.
The diffusion framework is subsequently somewhat haphazard. It catalogues thinking on individual and organisational change but simultaneously does not question the veracity of change while avoiding contentious issues and presenting potentially contradictory reasoning. These issues are perhaps indicative of attempting to draw general conclusions on a highly contextual phenomenon from a series of disparate studies.
Summary of findings
Summary of findings.
= Considered ✓? = Partially considered
One fresh element warranting further consideration is the use of diffusion. Individuals becoming aware, empowered and delivering transformative change to systems and structures (in possible akin to emancipation) represents a recurring element in all these frameworks, perhaps intimating implementation is as much a personal as an institutional matter. Such reasoning offers an alternative to paradigm shifts and the discrediting of old orthodoxy as a means of transformation (Sandercock, 1998), and helps explain how institutions are bastions of stability and arenas of change (March and Olsen, 1989).
The weaker symmetry develops from some of these substantive frameworks trying to guide practical endeavours at implementing interventions. Rather than presenting implementation as it is they seem to espouse an ideal approach or view to promote best practice and so become divorced from the pragmatic realities of delivery. Aspects such as power and politics are therefore conspicuously absent despite pervading practice in terms of service provision and resource availability (Goldacre, 2008). Being cynical, we question the fundamental, idealistic healthcare assertion of striving to improve the human condition rather than the realism of doing what it can with the resources available, and in this respect healthcare is perhaps not far removed from the pragmatic urban planning world.
The frameworks likewise presume optimal implementation is desirable and possible in contrast to urban planning theorists observations that sub-optimal interventions are the best one can achieve in political processes of negotiation and compromise (Clifford and Tewdwr-Jones, 2014; Healey, 2010). While dismissing rational models of knowledge to action, healthcare scholars still seemingly presume evidence and associated change is unambiguous, positive and improves practice which perhaps blinds investigators to nuisances of pragmatism and politics in implementation processes.
Discussion
Studying implementation involves attributing meaning to practice phenomena. As such incidents are contextually rooted, results are situation specific and do not lend themselves to generalisation: what happens in one case may not reoccur elsewhere (Booth, 2011). This reasoning creates fundamental issues for scholars seeking to offer overarching insight which goes beyond an individual study, as methodologies based on combining or pooling results from disparate studies (e.g. systematic review or meta analysis), or which measure performance in relation to another (e.g. cost benefit analysis) are seemingly inappropriate as they meld together a series of conflicting observations of disparate contextual practices.
Implementation theories can therefore be primarily described as mid-range. They imply a common nucleus or thread behind phenomena like implementation which offers only partial explanation (so is of a mid-range, neither entirely generalised or context specific) (Pawson and Tilley, 1997). Such suppositions are open to being tested and refined through refutation on new empiric examples (Popper, 1963) and therefore evolve.
These issues also extend to the use of theory in that we champion explanatory or substantive accounts rather than advocate panacea or prescribe sequences or conditions for achieving perfect delivery. This is a key weakness to some arguing theory ought to inform practice development and advance successful implementation (Helfrich et al., 2010; Nilsen, 2015), yet arguing for particular outcomes (when successful implementation is subjective) reduces scholars to taking sides in political activities, thereby becoming lobbyists.
Conjoined theories
Conjoined theories are required to explain complex and intertwined activity like implementation which transcends the range of an individual theory. In pursuing such goals, scholars borrow (or grave rob) theories and frameworks from other cognitive disciplines and classical domains to fill gaps in interpretations (Allmendinger and Thomas, 1998). Such practices can encourage problems concerning context stripping, the preservation of discredited theories and the fusing of conflicting, incompatible knowledge. Yet the alternative of employing grounded theory is not ideal either for it can be isolationist and context specific, thereby revealing little of an overarching phenomenon.
Going forward, we identify several dilemmas in conjoining theories:
Which theories to select from the myriad of options available, and how to preserve the contextual factors shaping these ideas? What type of theories to produce? Are explanatory theories sufficient to our audience and how can we achieve impacts without being prescriptive? How to apportion appropriate weighting to different theories? Identifying which properties of a theory are critical or merely incidental? How might we resolve ambiguities and contradictions between theories, and How might we test and refine our theories further given the unique contextual conditions inherent in a case study restrict generalisation?
There is seemingly no correct approach to conjoin theories and scholars instead rely on iterative and disjointed processes and acumen.
A future for implementation research?
Implementation has something of a chequered past with urban planning theorists and was written off as an endeavour for representing political activity involving compromise and negotiation. Yet it does seemingly have a future for the problems to which it pertains remain pertinent, especially in the current practice climate where resources are scarce and failures need preventing. Policy, plan and programme production can require considerable investment in terms of financial and political capital, so fear of failure can prove profound and damaging to promoters in terms of reputational harm (Smith et al., 2014). By suggesting how and why such failures might occur, implementation research can help reassure reflective decision makers into pursuing ambitious schemes for transformation rather than succumb to the fear of failure and inevitable slide towards safe, incremental initiatives (May, 2013; Smith et al., 2011; Sykes et al., 2014). While scholars cannot definitively say what works, they can identify what does not in a given instance which is of equal value in achieving good environments.
Research on implementation is illustrative and highlights to practitioners where danger exists and vigilance required via diffusion processes. Here, scholars advise, preach and lecture interested parties on the implications of their endeavours rather than present panacea to implementation problems in political arenas. Practitioners similarly need to be competent in their responsibility to reflect and translate these ideas into practice, perhaps utilizing the ideas on individual adaptation described in the diffusion of innovations in service organizations framework.
Conclusion
Implementation study is essentially observational, with theorists attempting to interpret meaning on a complex practical phenomenon at the nexus of praxis. Which elements in the relationships between contexts, mechanisms and outcomes scholars focus on remains a persistent debate, with theory serving as something of an aide memoire or nucleus of an idea to be shaped and hardened through empiric enquiry. The highly contextual nature of implementation makes generalisation inappropriate for such enquiry with theorists instead justifying their findings through constant process of refutation and refinement so subsequent theories evolve rather than remain static. As there is a degree of intrinsic, skilled judgement here, more work on unlocking how scholars go about representing implementation possibly using ethnographic or anthropological techniques could benefit our understanding of the phenomenon further.
Such a standpoint leaves scholars as interested observers rather than expert advisers who present panacea to fix implementation woes, which potentially limits the appeal of such insight to practiced disciplines like urban planning. How we remediate such a situation is ill-defined and requires practitioners to be reflective and adapt at translating theoretical ideas and principles into transformative actions via diffusion processes.
There are clear differences in perception on certain aspects of implementation research by healthcare and urban planning scholars, primarily concerning employing theory and influencing practice which can result in certain pertinent aspects being overlooked. Our review suggests that rather than reject trans-disciplinary comparison as incomparable analogy, a juxtaposition of alternative vistas can aid the identification of fresh perspectives and possibilities which at the very least ought to be contemplated.
Footnotes
Acknowledgements
I am grateful to Professor Jo Rycroft-Malone and Dr Christopher Burton of the School of Healthcare Sciences, Prifysgol Bangor University (UK) for helping identify the four conceptual frameworks considered in this paper.
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.