Research Setting the context for using complexity theory in evaluation: boundaries, governance and utilisation

Page created by Rodney Ortiz
 
CONTINUE READING
Evidence & Policy • vol 12 • no 1 • 73–89 • © Policy Press 2016 • #EVPOL
                              Print ISSN 1744 2648 • Online ISSN 1744 2656 • http://dx.doi.org/10.1332/174426415X14298726247211
                              This article is distributed under the terms of the Creative Commons Attribution-NonCommercial 4.0 license
                                                    (http://creativecommons.org/licenses/by-nc/4.0/) which permits adaptation, alteration, reproduction
                                                                                     and distribution for non-commercial use, without further permission
                                                                                     provided the original work is attributed. The derivative works do not
                                                                                                   need to be licensed on the same terms.

                                                                                     research
                           Setting the context for using complexity theory in
                           evaluation: boundaries, governance and utilisation
                                                                        Mat Walton, m.d.walton@massey.ac.nz
                                                                      Massey University, Wellington, New Zealand
IP : 192.168.39.211 On: Tue, 30 Nov 2021 08:23:57

                         Recent literature has usefully explored the application of complexity theory to evaluation. However,
                         there is little discussion of the contextual conditions in applying complexity theory. Drawing upon a
                         single complexity-consistent public health programme evaluation and subsequent policy decisions,
            Copyright The Policy Press

                         this paper considers how programme framing and governance can help or hinder application of
                Delivered by Ingenta

                         complexity theory. A wider framing of the programme and cross-system network governance
                         arrangements are suggested as important components to a complexity theory-informed policy
                         practice.

                                                           key words complexity theory • evaluation utilisation • governance networks

                         Introduction
                         There is a growing literature on the potential and application of complexity theory in
                         a number of areas including: conceptualising and researching public policy (Morçöl,
                         2012; Buijs et al, 2009; Cairney, 2012; Geyer, 2012; Eppel, 2012; Eppel et al, 2011);
                         research utilisation (Lemay and Creso, 2012); and evaluation (Vincent, 2012;Westhorp,
                         2012; Blackman et al, 2013; Patton, 2011;Walton, 2014). Drawing upon a single public
                         health programme evaluation and subsequent policy decisions, this paper considers
                         how programme framing and governance can help or hinder application of complexity
                         theory to public health evaluation and policy.

                         An outline of complexity theory
                         Complexity theory is not one coherent theory or field of research (Keshavarz et
                         al., 2010; Richardson, 2008), although there is a common set of concepts regarding
                         how complex systems behave and evolve over time. While first established in the
                         mathematical and physical sciences (Nicolis and Prigogine, 1989; Kauffman, 1995),
                         over the last 20 years the field has expanded to applied areas including management
                         (Stacey, 2012), health (Plsek and Greenhalgh, 2001), public policy (Morçöl, 2012)

                                                                                                   73
Mat Walton

                          and evaluation (Morell, 2010). Mowles (2014) suggests the most successful translation
                          from natural to social sciences of complexity requires both explicit identification
                          of the strand of complexity theory being applied, and a translation through social
                          theory. Two broad approaches within complexity theory provide either a search for
                          generalisable simple rules that govern complex systems, or understanding complex
                          systems within specific contexts (Byrne and Uprichard, 2012).The discussion below
                          follows the latter approach and is informed by realist complexity thinking and what
                          Byrne refers to as a complexity ‘frame of reference’ (Byrne and Callaghan, 2014).
                             Morçöl suggests that ‘complexity theorists describe emergent holistic systems,
                          whose properties are not reducible to those of their parts.These systems are integrated
                          into and co-evolve with their environments’ (Morçöl, 2001, 112). The idea that the
                          interaction of components within a complex system gives rise to ‘emergent’ properties,
                          which cannot be understood by examining the individual system components, is a
                          fundamental aspect of complexity theory (Goldstein, 1999; Byrne and Callaghan,
                          2014). The process of emergent properties is often described as the whole being
                          greater than the sum of its parts (for example, Cooper and Geyer, 2008). Emergent
IP : 192.168.39.211 On: Tue, 30 Nov 2021 08:23:57

                          phenomena appear as ‘integrated wholes’ (Goldstein, 1999, 49). The interaction
                          between components of a complex system is non-linear, and complex systems self-
                          organise in response to feedback (Morçöl, 2012). Self-organisation and non-linearity
                          mean that impacts of an intervention on emergent phenomena are not dependent on
            Copyright The Policy Press

                          the size of the intervention in terms of resource or effort, making it difficult to predict
                Delivered by Ingenta

                          the type or scale of system adaptations to new feedback (Medd, 2002; Room, 2011).
                             A complex system is open to feedback from the wider environment within which
                          it exists (Gatrell, 2005). Environments may differ between time, social and geographic
                          contexts. It is possible, therefore, that the configuration of seemingly similar systems
                          and their emergent phenomena, can be different across environments (Morçöl, 2012).
                          Conversely, emergent phenomena may be similar across different contexts and system
                          configurations (Richardson, 2008). Systems may maintain stability at an ‘attractor
                          state’ (Mackenzie, 2005), a region towards which a system is pulled (Merry, 1995).
                          A change in nature of emergent phenomena suggests the system has moved to a
                          new attractor state. At the point of change, the range of possible attractor states the
                          system can move to is dependent on the ‘phase’ or ‘state’ space of the system (Byrne
                          and Callaghan, 2014).

                          The complexity challenge for evidence and policy
                          There are a number of implications of a complexity frame of reference for thinking
                          about evidence of and for policy (Sanderson, 2009; Byrne, 2011). One of the
                          most obvious is that evidence is contextual. Context contains both temporal and
                          geographical circumstances, with the history, internal organisation and interaction
                          with other systems all being important. Open to their environments, no intervention
                          operates in isolation, challenging the notion that one intervention can be understood
                          as independent of previous and other concurrent interventions. Stern (2011, ix) notes
                          that ‘[i]n independent spheres it makes little sense that competing administrative
                          units, ministries and agencies each commission their own evaluations of “parts of
                          the whole”’.
                             Drawing upon a realist ontology, consistent with a ‘complex realist’ perspective,
                          Room (2013) summarises a number of challenges for the role of ‘evidence’ in policy:

                                                                      74
Setting the context for using complexity theory in evaluation

                                                    Not only must the evidence for a given policy intervention consider how
                                                    it will work in combination with other policies; it must also differentiate
                                                    according to the order in which these policies are introduced, in the various
                                                    contexts that characterise the policy maker’s domain, and having regard to
                                                    the different timescales of their likely effects. (Room, 2013, 231)

                          Guidance on applying complexity theory to evidence and policy
                          Several approaches to evaluation drawing upon a complexity frame of reference have
                          developed over the last decade, in addition to often sympathetic methodologies from
                          related systems fields (Williams and Hummelbrunner, 2011) and realist evaluation
                          which also offers a framework for working with complexity (Pawson, 2013). In
                          reviewing application of complexity theory to evaluation of development programmes,
                          Vincent (2012) notes two faces of complexity. The first face uses approaches such
                          as Qualitative Comparative Analysis (QCA) and Realist Evaluation for retrospective
                          analysis to identify context, intervention mechanism and outcome combinations.
IP : 192.168.39.211 On: Tue, 30 Nov 2021 08:23:57

                          These approaches highlight both a generative and configuration understanding of
                          causality (Byrne and Callaghan, 2014; Stern et al, 2012). The second face applies
                          complexity theory to action research approaches, highlighting the importance of
                          local adaptations and action, and utilises rapid evaluation-action cycles as a way of
            Copyright The Policy Press

                          managing uncertainty within complex systems.Vincent concludes:
                Delivered by Ingenta

                                                    Having explored the two faces of complexity we can see that is both
                                                    important to attend to local context, diverse perspectives and negotiations
                                                    driving emergent social processes, but also to the wider previously emergent
                                                    social factors that frame and influence current negotiations. (Vincent, 2012,
                                                    26)

                          Two other recent publications provide useful guidance from a realist complexity
                          frame. Westhorp (2012) suggests the use of complexity-consistent theory to help
                          define evaluation questions and appropriate evidence, and provides a list of complexity
                          theory concepts that may act as a complexity checklist against which the evaluation
                          should be able to provide information, listed in Table 1. Buijs et al (2009) provide
                          methodological guidance for policy research from a complexity frame of reference
                          utilising case-based methodologies.The two frameworks suggest that evaluation from
                          a complexity frame requires the following: definition of system boundaries and critical
                          reflection on these; identifying interactions between elements within the system and
                          between co-evolving systems; identifying processes of feedback and how these impact
                          on interactions; identifying change across different scales of the systems from macro to
                          micro; identifying rules and decisions across levels of the system; and understanding
                          temporal ordering of feedback, changes in interactions and emergent phenomena.
                          Trenholm and Ferlie (2013) provide an example of the importance of looking at
                          different scales of a system and their interaction. In studying the TB response system
                          in London, it was found that local innovation was at times limited and other times
                          supported by the macro ‘phase space’ dominated by New Public Management models.
                             A literature considering research utilisation from a complexity lens is also developing.
                          Viewing the policy-making process itself as a complex system highlights the context
                          within which research is conducted and utilised, suggesting that standardised research

                                                                                          75
Mat Walton

                                   Table 1: Checklist for assessing evaluations as complexity-consistent
                                   Checklist for assessing evaluations as complexity-consistent.
                                   Does the evaluation identify …
                                   Constituent elements of the system and drawing conceptual boundaries around the system
                                   Interactions amongst system elements
                                   Local rules that govern interactions and results across system levels from interactions
                                   Forms of feedback that constrain or support change
                                   Initial conditions that affect interactions within the system
                                   Controlling parameters of the system
                                   Interactions between levels of the system

                                   Adapted from Westhorp (2012)

                          strategies and tools for translation will have limited effect (Lemay and Creso, 2012).
                          Within complex policy systems, multiple actors and multiple relationships will be
IP : 192.168.39.211 On: Tue, 30 Nov 2021 08:23:57

                          involved in transferring research knowledge.There will be a divergence in the values
                          and assumptions of these actors that will give primacy to different methodologies,
                          interpretations and outcomes (Smith and Joyce, 2012; Petticrew, 2013). Group
                          and dialectical methods are seen as important for translating a complexity theory
            Copyright The Policy Press

                          understanding of a system into action to change the system (Byrne and Callaghan,
                Delivered by Ingenta

                          2014). Processes for determining policy action across multiple perspectives are
                          required, such as sensemaking (Eppel, 2012; Snowden, 2011) and network governance
                          arrangements (Klijn and Edelenbos, 2013).
                             Network governance can be defined as ‘public policy making and implementation
                          through a web of relationships between government, business and civil society’
                          (Klijn, 2008, 511) for solving complex problems that cannot be solved by a single
                          actor (Klijn and Edelenbos, 2013). Network governance arrangements have been
                          identified as consistent with a complexity frame of reference, with an emphasis on
                          interaction and relationships amongst members of the network leading to ‘emergent’
                          outcomes (Morçöl, 2012; Morçöl, 2014). Governance networks may develop from
                          local level collaborations or be mandated from central government (Bogason and
                          Musso, 2006; Hertting and Vedung, 2012). This paper suggests that governance
                          networks tasked with implementation and evaluation of a cross-agency programme
                          may aid in applying complexity theory to evaluation practice, as well as programme
                          and policy development.
                             In the next section, the complexity frame of reference outlined above will be applied
                          to a case study. The analysis aims to explore implications of a complexity frame of
                          reference for defining programme system boundaries, governance arrangements and
                          associated implications for evaluation utilisation.

                          Case study – New Zealand Fruit In Schools programme and
                          evaluation
                          Fruit in Schools (FiS) was a health promotion programme introduced into New
                          Zealand primary schools in 2006. While not explicitly informed by a complexity
                          frame of reference, the case was chosen because both the programme and evaluation
                          designs have a number of features that are complexity consistent. Examining processes

                                                                                      76
Setting the context for using complexity theory in evaluation

                          of programme and evaluation governance, as well as evaluation utilisation, facilitates a
                          thought experiment of how these processes may be managed if an explicit complexity
                          frame of reference is adopted.
                             In developing this case study, both document analysis and key informant interviews
                          were utilised. Documents included: evaluation reports (Boyd et al, 2009; Boyd and
                          Moss, 2009; Boyd et al, 2007; Dingle et al, 2009); advice to the Minister of Health
                          obtained under the New Zealand Official Information Act (Ministry of Health,
                          2009b; Ministry of Health, 2009a); and media reports. Interviews were undertaken
                          with seven key informants involved in policy making, programme implementation
                          and evaluation of FiS. Drafts of this paper were shared with four participants who
                          indicated an interest, to check both accuracy of description and positioning of the
                          FiS programme and evaluation within contextual background.

                          The setting
                          New Zealand has a population of 4.4 million, of which nearly 15% are Maori
IP : 192.168.39.211 On: Tue, 30 Nov 2021 08:23:57

                          (indigenous) and 77% are of European or ‘other’ descent. Two significant groups in
                          this ‘other’ category include those who identify as Asian or Pacific Islands ethnicities
                          (Statistics New Zealand, 2013). A unitary Westminster-style Parliament sees Cabinet
                          Ministers able to quickly implement policy and programme changes, often following
            Copyright The Policy Press

                          election of new government administrations on a three-year election cycle (Blank
                Delivered by Ingenta

                          and Burau, 2010).
                             In the early 2000s two high-profile government strategies were developed to address
                          key issues of health and nutrition. The Healthy Eating – Healthy Action (HEHA)
                          strategy (Ministry of Health, 2003) linked aims around obesity prevention, nutrition
                          and physical activity; and the Cancer Control Strategy (Cancer Control Taskforce,
                          2005) saw adequate nutrition as an important preventative of cancer. These broad
                          and ambitious strategies provided the framework for many specific programmes and
                          projects, including a number based within schools and targeting children (Walton et
                          al, 2010; Cushman and Clelland, 2012).

                          The policy and the programme
                          In 2005 a high-profile Fruit in Schools initiative for schools serving low socioeconomic
                          communities was announced as part of New Zealand’s Cancer Control action plan
                          (Cancer Control Taskforce, 2005).The FiS programme provided a piece of fruit per day
                          for each Year 1–8 child in the targeted schools.The programme also aimed to increase
                          the focus on health promotion in schools so was supported by a Health Promoting
                          Schools (HPS) framework, coordinators and associated teacher release and training.
                          The programme is described more fully in Box 1. The free fruit of FiS was seen by
                          both the government officials and the programme’s evaluators as a “carrot” (IF 3) or
                          a “foot in the door to a suite of health promotion activities” (IF 23), as schools had
                          to “sign up to this wider programme” (IF 23) of Health Promoting Schools.

                                                                                 77
Mat Walton

                                                    Box 1: The Fruit in Schools Programme

                                                    Targeted schools: low socioeconomic profile primary, contributing, and intermediate
                                                    schools (Years 1–8). FiS was not mandatory – schools were invited to take part.

                                                    Intervention:

                                                    •    Part 1 – one piece of fruit per school day per student

                                                    • Part 2 - Health Promoting Schools programme targeting four key cancer-prevention
                                                    areas: nutrition, physical activity, SunSmart (melanoma prevention) and smokefree.

                                                    • Programme supports: 25 dedicated FiS / HPS advisors; each school lead-teacher could
                                                    access funded release time for training and planning

                                                    Timeline: FiS expanded quickly through several phases, with each phase extending the
IP : 192.168.39.211 On: Tue, 30 Nov 2021 08:23:57

                                                    number of schools eligible for the programme. Phase One began in October 2005; Phase
                                                    Three a year later, by which time there about 268 primary schools in the scheme. Later
                                                    Phases (Four and Five) saw an expanded budget and the inclusion of slightly higher
                                                    socio-economic status (SES) profile schools and intermediate schools (Years 7 and 8) in
            Copyright The Policy Press

                                                    the FiS programme.
                Delivered by Ingenta

                          While the Ministry of Health was the lead agency and funder of the Fruit in Schools
                          programme, it worked closely with an External Reference Group that included
                          policy and practice representatives of the Ministry of Education, SPARC (Sport and
                          Recreation New Zealand, now Sport New Zealand) and non-government agencies
                          (NGOs) that provide programmes in schools targeted at the four health areas of
                          FiS. The group received regular reports from the evaluation and acted as a problem
                          solving group to support implementation and align initiatives. The FiS evaluation
                          report states that:

                                                        … a restructuring of the group occurred in late 2007 – causing some concern
                                                        among national agency representatives who saw it as a forum for shared
                                                        decision-making and accountability. (Boyd et al, 2009, 54)

                          The phased roll-outs of the programme took place against a moving backdrop of
                          related policies and activities. Several other high-profile nutrition and physical activity
                          initiatives were implemented within schools or targeted towards children over the
                          FiS implementation period. Some, but not all, child nutrition-focused interventions
                          over this period were evaluated through separate processes, and by separate agencies
                          (Education Review Office, 2009; Pledger et al, 2012; McLean et al, 2009).

                          The evaluation
                          An independent team, comprised of education and health researchers, began a mixed
                          method evaluation in late 2005. In the early stages, 2005–06, the evaluation was

                                                                                                 78
Setting the context for using complexity theory in evaluation

                          concerned with ‘process and formative evaluation’ intended to help stakeholders
                          improve the programme (Boyd et al, 2009, 4). Between 2006 and 2008 the focus
                          shifted to impact evaluation.
                             Evaluation methods included: surveys to students and teachers at most of the
                          schools at the start of 2006 and then at the end of each of the three years, 2006, 2007
                          and 2008; questionnaires to lead teachers in each school; 12 school case studies, and
                          interviews and online surveys with agency partners and stakeholders (Boyd et al,
                          2009, 10). Student questionnaires tracked a cohort of students, from Year 4 in 2006 to
                          their Year 6 in 2008, to consider changes in attitudes and behaviours. A comparison
                          group of schools was also selected.
                             Informed in theory and methodology by HPS, in part the evaluation design was
                          a pragmatic response to political positioning and tight implementation timelines.
                          Constant movement in the policy context created unforeseen programme changes.
                          Notably, most schools initially in the comparison group of schools not receiving fruit
                          subsequently received fruit in an added Phase Three of the programme.This left only
                          seven schools from the original comparison group of 34 schools in the evaluation
IP : 192.168.39.211 On: Tue, 30 Nov 2021 08:23:57

                          study (Boyd et al, 2009).
                             Final main evaluation findings were that students in FiS schools showed a slowing
                          in the expected decline in healthy attitudes and behaviours observed as students get
                          older. FiS students had maintained positive attitudes to, and awareness of, healthy
            Copyright The Policy Press

                          behaviours; increased their intake of healthy foods and increased physical activity and
                Delivered by Ingenta

                          sunsmart practices. They had positive views about school and took an active role in
                          promoting their school as a ‘healthy school’.The schools gave students health-related
                          leadership opportunities, increased their involvement with the programme’s agency
                          partners, and engaged with other health and wellbeing initiatives. This all raised the
                          value placed on health promotion in the schools and created a ‘protective climate’
                          for students.

                          What happened after the evaluation?
                          The centre-right government elected in November 2008 immediately ended several
                          child and youth focused nutrition and physical activity programmes. FiS also came
                          under review (Treasury, 2010). The Minister of Health made an interim decision to
                          continue the programme until after the evaluation report was completed (Ministry
                          of Health, 2009a).
                             In reporting the FiS evaluation findings to the Minister of Health, the Ministry of
                          Health used almost verbatim the summary provided in the Evaluators’ report (Boyd
                          et al, 2009).The Ministry advised that free fruit was ‘well-received in these high needs
                          school communities and was recognised as a key enabler in raising the importance
                          placed on health and wellbeing’. It went on to say that : ‘… benefits cannot solely be
                          attributed to the FiS programme’ (Ministry of Health, 2009b). While the evaluators
                          emphasised that it was the support and activity surrounding FiS, and not the piece
                          of fruit in itself which was making the difference, this was minimised in reporting
                          to the Minister. In line with the incoming government’s focus on ‘frontline services
                          rather than back-room bureaucracies’ (English, 5 December 2008) the Minister of
                          Health reframed the wider HPS component as ‘administrative’. Later in 2009 it was
                          announced that the Government would ‘continue the Fruit in Schools programme,
                          but is cutting the administration component’ (Ryall, 29 October 2009). The cut

                                                                                 79
Mat Walton

                          ‘administration component’ included: FiS co-ordinator and staff salaries; liaison
                          staff from District Health Boards; teacher-release time; meetings and other support
                          activities of the HPS approach. The FiS programme continues to deliver free fruit
                          to eligible schools.
                             Both the timing of reports to the Minister of Health, and comments from some
                          key informants, suggest that evaluation findings may have had some bearing on the
                          decision to carry on funding the fruit provision part of the FiS programme. Having
                          said this, the decision to stop funding for teacher release, professional development
                          and coordinator support is counter to headline findings of the FiS evaluation. Key
                          informants suggested that popularity and symbolic significance of the fruit provision
                          were amongst reasons for keeping this, while the other FiS programme components
                          did not have such support. The decision regarding the FiS programme was taken by
                          the Minister of Health alone.

                          Fruit in Schools through a complexity frame of reference
IP : 192.168.39.211 On: Tue, 30 Nov 2021 08:23:57

                          The FiS programme and evaluation were not explicitly informed by a complexity
                          frame of reference. In fact, many of the developments in the application of complexity
                          theory to policy and evaluation have been published since the programme and
                          evaluation establishment in 2005. However, using the seven criteria for complexity
            Copyright The Policy Press

                          consistent theory identified by Westhorp (2012), a brief summary of the FiS evaluation
                Delivered by Ingenta

                          against these criteria is provided. The first criterion is that elements of the system
                          are identified and boundaries drawn, while the second criterion is that interactions
                          between system elements are identified. The evaluation explicitly drew upon HPS
                          and related socioecological systems theory to outline a FiS ‘system’. The evaluation
                          framework identified different levels of the system under study including schools,
                          regional agencies, national agencies and policies and programmes operating across these
                          levels of organisation. At the school level the HPS framework focused attention on
                          interaction between school policies and practices, curriculum, students, staff, parents
                          and the wider community. Survey, interview and case study data were collected across
                          the levels of system and from different perspectives within schools.
                             The third criterion is that local rules are identified that govern interactions. The
                          impact of rules on school practices were identified in several areas. A particular focus
                          was the impact of other programmes and administrative regulations that schools
                          operate under.These could act as both facilitators and barriers to activity in the four
                          FiS health areas. The fourth criterion is that forms of feedback are identified. The
                          evaluation design was cognisant of feedback both within and between levels of the FiS
                          system. The evaluation design looked for interaction between student level changes,
                          school level changes and inter-agency relationship changes. Data triangulation was
                          used to consider the interactions between data at these different levels of the system.
                             The fifth criterion is that initial conditions that affect how the system operates are
                          identified. Data analysis had a focus on understanding FiS system trajectories, although
                          this term is not used by the evaluators. Identifying changes from initial conditions
                          was a focus of analysis, described by the evaluators as:

                                                                     80
Setting the context for using complexity theory in evaluation

                          •                         Comparing the patterns over time for FiS and comparison students
                          •                         Data triangulation (that is, looking at what the qualitative and quantitative data
                                                    from different stakeholders told us about key aspects of student, school and
                                                    interagency practice)
                          •                         Comparing changes over time to the expected patterns documented in key
                                                    literature
                          •                         Comparing the patterns of changes over time for different stakeholder groups
                                                    (that is, comparing the responses of stakeholders in different phases of FiS). (Boyd
                                                    et al, 2009, 11)

                          Whilst an understanding of trajectories for school and inter-agency practices was
                          achieved, an explicit complexity frame of reference would likely see these trajectories
                          fully explored and featured in key findings included in the executive summary.
                             Identifying interactions between levels of the system and control parameters that
                          determine whether a system radically changes pattern are the two final criteria. As
                          described above, data was gathered from school, local partner agency and national
IP : 192.168.39.211 On: Tue, 30 Nov 2021 08:23:57

                          partner agency perspectives, with interactions between these levels considered. The
                          concept of controlling parameters may be one area where a socioecological and HPS
                          view does not explicitly address concerns from a complexity frame of reference.
                             In summary, the FiS evaluation appears consistent in many ways with a complexity
            Copyright The Policy Press

                          frame of reference.The evaluation has something to say against six of seven complexity
                Delivered by Ingenta

                          theory criteria identified by Westhorp (2012). Multiple methods were used to gain
                          insight into changes in practices at different levels within the system, from the
                          perspective of multiple stakeholders. Selecting the FiS case is not to suggest it is
                          unique in being consistent with complexity theory. In fact, it is likely that many
                          evaluations drawing upon socioecological theory in public health would likewise be
                          consistent in several areas.
                             While the evaluation itself is consistent with the emphasis a complexity theory
                          frame provides on evidence, the case highlights contextual issues that limit the
                          use of the evaluation findings. These contextual issues are also considered from a
                          complexity theory perspective below. The first is the framing of the programme.
                          The second is the programme governance. It is suggested that an approach informed
                          by complexity theory and network governance would allow for fuller utilisation of
                          evaluation findings.

                          Programme framing and governance
                          In complex issues, it is likely that action will be required across traditional agency
                          boundaries, and between government and non-government (NGO) agencies
                          (Australian Public Service Commission, 2007). The FiS programme was in effect a
                          collaboration between the Ministry of Health, Ministry of Education, SPARC, and
                          NGOs at national level, and their delivery agents (district health boards, schools,
                          regional offices) at local levels. However, the political governance arrangements placed
                          political decision-making responsibility with the Minister of Health. Drawing upon
                          ideas from both complexity theory and network governance, two related issues are
                          identified. First, expanding the boundaries drawn around the FiS programme and
                          evaluation. Second, the role of a governance network in utilising the evaluation
                          findings.

                                                                                          81
Mat Walton

                          Programme framing, defining system boundaries
                          A key step in many systems and complexity methods is actively defining the boundaries
                          of the system under study (Byrne and Callaghan, 2014; Midgley, 2000; Williams and
                          Hummelbrunner, 2011), with recognition that boundaries are constructed around
                          open and interacting systems. The FiS evaluation did consider how FiS was affected
                          by other nutrition- focused programmes and activities within schools. Even though
                          the focus was FiS, as one informant stated “… the other initiatives sort of blended
                          in with Fruit in Schools because in some ways they are aligned, but it’s very hard
                          to tease out what was what …” (IF 3). One implication of a complexity frame of
                          reference is that policy and programme designs need to be aware of interactions.
                          The final evaluation report does suggest that FiS should be better aligned with HPS
                          and the school curriculum. Also, that partner agency capacity to support schools is
                          reviewed (Boyd et al, 2009). In making a decision on FiS the Minister of Health
                          requested a report on existing HPS capacity to support the FiS programme. This
                          report was light on detail and did not fully explore the implications of removing the
IP : 192.168.39.211 On: Tue, 30 Nov 2021 08:23:57

                          FiS coordinator workforce (Ministry of Health, 2009b). No mention was made of
                          partner agency capacity or curriculum alignment, although arguably these are both
                          outside the Minister of Health’s remit.
                            A focus on programme interactions suggests a wider system boundary is required.
            Copyright The Policy Press

                          The FiS is clearly both a health promotion and school system intervention. It has
                Delivered by Ingenta

                          many interacting programmes. This was recognised by establishing the inter-agency
                          groups. Here it is suggested that network governance theory offers a set of ideas
                          complementary to complexity theory that can act to implement a complexity
                          theory perspective. A cross-agency network governance group could be delegated
                          responsibility for looking across the broader area of school health programmes
                          including FiS. This would legitimise an interaction focus in the evaluation, and
                          recommendations to Ministers could consider complementary changes across
                          programmes.

                          Network governance – a process for managing interaction
                          A key mechanism for governance networks to produce solutions to complex problems
                          lies in the ability to bring multiple perspectives and knowledge into a deliberative
                          decision-making process (Klijn and Edelenbos, 2013; Hertting and Vedung, 2012).
                          Meadows, from a systems perspective, identified ‘the ability to transcend paradigms’
                          as the most effective leverage point to transform systems (Meadows, 1999, 19).
                          Identifying competing paradigms of actors within a system is required to understand
                          action within that system and the value placed on certain outcomes.
                             Because complex systems contain vertical as well as horizontal interactions, a
                          network governance group should consist of cross-agency and cross-system level
                          (local, regional, national) actors. Within this network process, some complexity-
                          informed principles should be considered.The focus would not be insular on a single
                          programme, but consider a wider view of the systems within which the programme
                          operates and interaction with both system context and other programmes. Complexity
                          concepts, such as phase shift, attractor states and system trajectories provide a meta-
                          theory within which more specific discipline and subject matter assumptions can be
                          considered (Westhorp, 2012).

                                                                    82
Setting the context for using complexity theory in evaluation

                             Network governance arrangements provide a process to consider evaluation findings
                          and make decisions across the two faces of complexity identified by Vincent (2012),
                          iterative programme development and broader policy setting decisions. In the FiS case,
                          a network governance group would require delegated authority to consider evaluation
                          findings and make tactical and programme development decisions. Participants
                          involved with the FiS case note that while in place, the external reference group were
                          able to make decisions to problem solve during implementation.
                             A network governance group could also be tasked with making policy setting
                          recommendations to relevant Ministers, for example programme termination or
                          expansion. In the FiS case, recommendations came from only the Ministry of Health
                          to the Minister of Health. Recommendations to government Ministers from a network
                          governance group would of course represent a variety of perspectives from across the
                          system. Collective decision making by multiple Ministers with a stake in the ‘system’
                          under study would be a logical extension of network governance arrangements.
                          Related to the programme framing and central importance of interactions within
                          the school-health ‘system’, the remit of a network governance group would need
IP : 192.168.39.211 On: Tue, 30 Nov 2021 08:23:57

                          to be wider than only the FiS programme. Policy setting recommendations would
                          be across multiple programmes within health and education ministerial portfolios.

                          Evaluation utilisation
            Copyright The Policy Press
                Delivered by Ingenta

                          The previous discussion has outlined how complexity theory and network governance
                          frames could inform programme boundary definitions and governance arrangements.
                          With reference to the FiS case study, the implications of these frames for evaluation
                          utilisation are considered.
                             The Minister of Health received a report on the FiS evaluation that largely reflected
                          the final evaluation report executive summary. Much of the context-sensitive analysis
                          contained in the evaluation reports was lost within the executive summary. The
                          importance of initial conditions and divergent trajectories of FiS schools were not
                          included when describing key outcomes and mechanisms. If explicitly reflecting
                          a complexity frame, this information would likely drive the findings summary. Its
                          absence may reflect the challenge of fitting in the complexity of programme and
                          context interactions within a summary document. Pawson and Manzano-Santaella
                          (2012) note that realist exploration inside the ‘black box’ of programmes is difficult
                          within a tight word limit.A review of the degree to which Complex Adaptive Systems
                          (CAS) elements are incorporated in health promotion evaluations also suggested that
                          publishing restrictions may limit discussion of CAS perspectives (Kania et al, 2013).
                             A network governance group could provide a forum where detailed examination
                          of findings is possible, and implications of findings for practice, rather than findings
                          themselves, forms the basis of a summary. Byrne and Callaghan (2014) discuss dialogical
                          processes as needed to unify views of systems under study, including views generated
                          through research with those of people living and working in the systems under study.
                          Sanderson (2009), drawing upon complexity theory and Dewian pragmatism also
                          recognises the need to engage those ‘doing’ within the system in a move towards
                          ‘practical rationality’. The network governance group could be delegated authority
                          to make programme adaptation and development decisions and policy setting
                          recommendations (for example, termination or expansion) to Ministers.

                                                                                 83
Mat Walton

                             To be consistent with a complexity frame of reference, when making programme
                          development decisions or policy-setting recommendations to Ministers, the network
                          governance group would seek to clarify the types of outcomes emerging from the
                          system, and whether the system has undergone any phase shift. For FiS, if it seemed
                          that school systems (or those on certain trajectories at least) had moved into a
                          situation where health promoting actions were the norm, then it could be reasoned
                          that some programmes or aspects of them could be terminated while maintaining
                          the health and education outcomes of importance. Rather than a blanket approach,
                          recommendations could consider where this may not be the case, the role of initial
                          conditions, and configurations of system attributes that appear to be driving any
                          divergence of outcomes between school system cases.

                          Conclusion
                          Drawing upon complexity theory-related literature and the FiS case study of
                          a complexity consistent programme and evaluation, this paper has argued that
IP : 192.168.39.211 On: Tue, 30 Nov 2021 08:23:57

                          programme framing and governance arrangements can provide the mechanisms for
                          applying complexity theory within evaluation, programme and policy processes.
                             The FiS evaluation provided findings of use from a complexity theory perspective.
                          A nuanced understanding of the role of free fruit and HPS processes within schools
            Copyright The Policy Press

                          was achieved. While an explicit complexity consistent method would likely provide
                Delivered by Ingenta

                          more detailed understanding of programme and context configurations, the FiS
                          evaluation did identify several interaction effects with other programmes and
                          ‘emergent’ outcomes.The FiS evaluation showed that sympathetic system perspectives,
                          in this case socioecological theory utilising a mixed-method design, can provide
                          complexity-consistent results. This was also demonstrated in evaluating the Welsh
                          healthy school schemes (Rothwell et al, 2010). Less complexity-consistent was the
                          framing, governance and evaluation utilisation. It is suggested that network governance
                          arrangements can be a key vehicle for delivering complexity-consistent evaluation
                          framing and utilisation.
                             From a complexity frame, defining the boundaries of the system under study
                          has implications for legitimacy of evaluation questions, what data are collected and
                          which programme interactions are legitimate areas for evaluation and policy action
                          (Byrne and Uprichard, 2012; Midgley, 2000). Governance networks bring together
                          a variety of different problem definitions, preferred solutions and positions from
                          which to value outcomes (Hertting and Vedung, 2012). By explicitly bringing these
                          perspectives together both ideas for, and implementation of, solutions for complex
                          and ‘wicked’ problems may be achieved (Ferlie et al, 2011; Australian Public Service
                          Commission, 2007).
                             The case study highlights different processes required to bring together the two faces
                          of complexity identified by Vincent (2012). First, evaluation activity should inform
                          ongoing incremental programme developments. Non-linear interactions within
                          complex systems and ‘emergent’ system properties imply that unintended impacts
                          will occur (Eppel, 2012). One task of a programme governance network would be
                          to help steer systems in a desired direction through refinements across programmes
                          and actors. This requires timely reporting of evaluation findings and flexibility to
                          focus on emerging outcomes.

                                                                     84
Setting the context for using complexity theory in evaluation

                             The other use of evaluation is to inform broader policy settings and decisions
                          regarding overall programme continuation, adaptation or termination. Here again
                          it is suggested that the governance network is best placed to fully consider a
                          context-sensitive evaluation, detailing types of system trajectories and interactions.
                          Recommendations from the governance network would be made to government
                          ministers across relevant portfolios for collective decision making.
                             While this paper has focused on application of complexity theory, it is likely that
                          the analysis and proposed prescriptions would also apply to the practice and influence
                          of systems and realist approaches more widely.
                             There are some obvious limitations to the complexity-framed scenario outlined
                          here. First, devolving decisions to networks, and sharing political responsibility, requires
                          political will and trust from government ministers. In New Zealand, recent changes
                          to public service organisation to foster more cross-government working in complex
                          areas suggests a recognition that more networked approaches are required (Duncan
                          and Chapman, 2012). It is yet to be seen how consistent with a complexity frame these
                          efforts will be in practice. The study by Trenholm and Ferlie (2013) is a reminder of
IP : 192.168.39.211 On: Tue, 30 Nov 2021 08:23:57

                          how the success or failure of localised programmes can be influenced by established
                          traditions at a more macro level. Even if more complexity-consistent programmes and
                          evaluations are commissioned, the influence they have on wider political and policy
                          systems will likely be limited if reductionist and managerial frames are dominant at
            Copyright The Policy Press

                          the macro level. A number of challenges in the actual FiS evaluation resulted from
                Delivered by Ingenta

                          implementation time pressures, as well as changes in programme management and
                          political priorities. A complexity-framed, network-governance approach requires
                          sufficient time for genuine and ongoing stakeholder involvement and dialogue. The
                          second area of limitation is around the established assumptions of evidence and action
                          held by policy practitioners and within specific disciplines.
                             Sanderson (2009) notes that the dominant response to complexity is to increase
                          technical sophistication of methods, rather than adopt new ways of thinking.Within
                          the case study, informants identified a tension regarding competing assumptions of
                          ‘good’ evidence between health and education disciplines. The complexity frame
                          suggested here implies new ways of thinking that challenge several assumptions of
                          evidence based policy making (EBPM) as often practised (Pawson, 2013). To apply
                          complexity theory, from evaluation design through to governance and policy design,
                          will require buy-in across the policy system including politicians, evaluation and policy
                          practitioners, and those implementing programmes across agencies.The ‘phase space’
                          of the policy system needs to allow for complexity theory informed policy practice.

                          Acknowledgements
                          Thank you to participants for generously giving time for this research; to Angelique
                          Praat, Sally Boyd, Nan Wehipeihana; Jackie Cumming and Jenny Neale for commenting
                          on earlier drafts; Marie Russell for contributing to participant interviews and document
                          review. This research was supported by the Marsden Fund Council from Government
                          funding, administered by the Royal Society of New Zealand.

                                                                                 85
Mat Walton

                          References
                          Australian Public Service Commission, 2007, Tackling wicked problems: A public policy
                            perspective, Canberra: Australian Public Service Commission
                          Blackman, T, Wistow, J, Byrne, D, 2013, Using qualitative comparative analysis to
                            understand complex policy problems, Evaluation 19, 126–40
                          Blank, RH, Burau,V, 2010, Comparative health policy, Basingstoke: Palgrave Macmillan
                          Bogason, P, Musso, JA, 2006, The democratic prospects of network governance,
                            American Review of Public Administration 36, 3–18
                          Boyd, S, Dingle, R, Campbell R, King, J, Corter, A, 2007, Taking a bite of the apple:The
                            implementation of Fruit in Schools,Wellington: New Zealand Council for Educational
                            Research (NZCER)
                          Boyd, S, Dingle, R, Hodgen, E, Kind, J, Moss, M, 2009, The changing face of Fruit
                            in Schools: 2009 overview report, Wellington: New Zealand NZCER and Health
                            Outcomes International
                          Boyd, S, Moss, M, 2009, The changing face of Fruit in Schools: The 2008 case studies,
                            Wellington: New Zealand NZCER and Health Outcomes International, 98
IP : 192.168.39.211 On: Tue, 30 Nov 2021 08:23:57

                          Buijs, J-M, Eshuis, J, Byrne, D, 2009, Approaches to researching complexity in public
                            management, in Teisman, G, Buuren,AV, Gerrits L (eds), Managing complex governance
                            systems, New York, NY: Routledge, 37–55
                          Byrne, D, 2011, Applying social science: The role of social research in politics, policy and
            Copyright The Policy Press

                            practice, Bristol: Policy Press
                Delivered by Ingenta

                          Byrne, D, Callaghan, G, 2014, Complexity theory and the social sciences: The state of the
                            art, Oxford: Routledge
                          Byrne, D, Uprichard, E, 2012, Useful complex causality, in Kincaid, H (ed), The Oxford
                            Handbook of Philosophy of Social Science, Oxford: Oxford University Press, 109–29
                          Cairney, P, 2012, Complexity theory in political science and public policy, Political
                            Studies Review 10, 346–58
                          Cancer Control Taskforce, 2005, The New Zealand Cancer Control Strategy: Action Plan
                            2005–10, Wellington: Ministry of Health
                          Cooper, H, Geyer, R, 2008, Using ‘complexity’ for improving educational research
                            in health care, Social Science and Medicine 67, 177–82
                          Cushman, P, Clelland, T, 2012, Addressing health issues in New Zealand schools,
                            International Journal of Health Promotion and Education 50, 159–68
                          Dingle, R, Hodgen, E, Boyd, S, Shapleski, J. Kind, J, Moss, M, 2009, The changing
                            face of Fruit in Schools: Technical report, Wellington: NZCER and Health Outcomes
                            International
                          Duncan, G, Chapman, J, 2012, Better public services? Public management and the
                            New Zealand model, Public Policy 7, 151–66
                          Education Review Office, 2009, Schools’ progress towards meeting national administration
                            guideline 5 on food and nutrition: Part 2, Wellington: Education Review Office
                          English, B, 2008, Crown accounts show need for economic action – Press release 5
                            December, www.beehive.govt.nz/release/crown–accounts–show–need–economic–
                            action
                          Eppel, E, 2012,What does it take to make surprises less surprising? The contribution
                            of complexity theory to anticipation in public management, Public Management
                            Review 14, 881–902
                          Eppel, E, Matheson,A,Walton, M, 2011,Applying complexity theory to New Zealand
                            public policy: Principles for practice, Policy Quarterly 7, 48–55

                                                                      86
Setting the context for using complexity theory in evaluation

                          Ferlie, E, Fitzgerald, L, McGivern, G, Dopson, S, Bennett, C, 2011, Public policy
                            networks and ‘wicked problems’:A nascent solution? Public Administration 89, 307–24
                          Gatrell,AC, 2005, Complexity theory and geographies of health:A critical assessment,
                            Social Science and Medicine 60, 2661–71
                          Geyer, R, 2012, Can complexity move UK policy beyond ‘evidence–based policy
                            making’ and the ‘audit culture’? Applying a ‘complexity cascade’ to education and
                            health policy, Political Studies 60, 20–43
                          Goldstein J, 1999, Emergence as a construct: History and issues, Emergence 1, 49–72
                          Hertting, N,Vedung, E, 2012, Purposes and criteria in network governance evaluation:
                            How far does standard evaluation vocabulary take us? Evaluation 18, 27–46
                          Kania, A, Patel, AB, Roy, A,Yelland, GS, Nguyen, DTK,Verhoef, MJ, 2013, Capturing
                            the complexity of evaluations of health promotion interventions:A scoping review,
                            Canadian Journal of Program Evaluation 27, 65–91
                          Kauffman, S, 1995, At home in the universe:The search for the laws of self-organization and
                            complexity, New York, NY: Oxford University Press
                          Keshavarz, N, Nutbeam, D, Rowling, L, Khavarpour, F, 2010, Schools as social complex
IP : 192.168.39.211 On: Tue, 30 Nov 2021 08:23:57

                            adaptive systems:A new way to understand the challenges of introducing the health
                            promoting schools concept, Social Science and Medicine 70, 1467–74
                          Klijn, EH, 2008, Governance and governance networks in Europe: An assessment of
                            ten years of research on the theme, Public Management Review 10, 505–25
            Copyright The Policy Press

                          Klijn, EH, Edelenbos, J, 2013, The influence of democratic legitimacy on outcomes
                Delivered by Ingenta

                            in governance networks, Administration and Society 45, 627–50
                          Lemay, MA, Sá, C, 2012, Complexity sciences: Towards an alternative approach to
                            understanding the use of academic research, Evidence & Policy 8, 473–94
                          Mackenzie, A, 2005, The problem of the attractor: A singular generality between
                            sciences and social theory, Theory Culture Society 22, 45–65
                          McLean, RM, Hoek, JA, Buckley, S, Croxson, B, Cumming, J, Ehau, TH, Tanuvasa,
                            AF, Johnston, M, Mann, JI, Schofield, G, 2009, Healthy eating – healthy action:
                            Evaluating New Zealand’s obesity prevention strategy, BMC Public Health 9, 452
                          Meadows, D, 1999, Leverage points places to intervene in a system, Hartland, VT:
                            Sustainability Institute
                          Medd,W, 2002, Complexity and the social world, Social Research Methodology 5, 71–81
                          Merry, U, 1995, Coping with uncertainty: Insights from the new sciences of chaos, self-
                            organization, and complexity, Westport, CT: Praeger
                          Midgley, G, 2000, Systemic intervention, philosophy, methodology, and practice, New York:
                            Kluwer Academic / Planum Publishers
                          Ministry of Health, 2003, Healthy eating – healthy action oranga kai – oranga pumau: A
                            strategic framework, Wellington: Ministry of Health
                          Ministry of Health, 2009a, Health Report 20090889, 21 May, Wellington: Ministry
                            of Health
                          Ministry of Health, 2009b, Health Report 20091752, 9 October,Wellington: Ministry
                            of Health
                          Morçöl, G, 2001, What is complexity science? Postmodernist or postpositivist?
                            Emergence 3, 104–19
                          Morçöl, G, 2012, A complexity theory for public policy, New York, NY: Routledge
                          Morçöl, G, 2014, Complex governance networks: An assessment of advances and
                            prospects Complexity, Governance and Networks 1, 5–16

                                                                                 87
Mat Walton

                          Morell, JA, 2010, Evaluation in the face of uncertainty, anticipating surprise and responding
                            to the inevitable, New York, NY: Guilford Press
                          Mowles, C, 2014, Complex, but not quite complex enough: The turn to the
                            complexity sciences in evaluation scholarship, Evaluation 20, 160–75
                          Nicolis, G, Prigogine, I, 1989, Exploring complexity: An introduction, New York, NY:
                            WH Freeman
                          Patton, MQ, 2011, Developmental evaluation: Applying complexity concepts to enhance
                            innovation and use, New York, NY: Guilford Press
                          Pawson, R, 2013, The science of evaluation, London: Sage
                          Pawson, R, Manzano-Santaella, A, 2012, A realist diagnostic workshop, Evaluation
                            18, 176–91
                          Petticrew, M, 2013, Public health evaluation: Epistemological challenges to evidence
                            production and use, Evidence & Policy 9, 87–95
                          Pledger, M, McDonald, J, Cumming, J, 2012, Increases in support structures for healthy
                            eating especially in low decile schools in New Zealand, Australian and New Zealand
                            Journal of Public Health 33, 543–9
IP : 192.168.39.211 On: Tue, 30 Nov 2021 08:23:57

                          Plsek, PE, Greenhalgh, T, 2001, Complexity science: The challenge of complexity in
                            health care, BMJ 323, 625–8
                          Richardson, KA, 2008, On the limits of bottom-up computer simulation: Towards
                            a nonlinear modeling culture, in Dennard, L, Richardson, KA, Morçöl, G (eds),
            Copyright The Policy Press

                            Complexity and policy analysis, Goodyear, AZ: ISCE Publishing, 339–54
                Delivered by Ingenta

                          Room, G, 2011, Complexity, institutions and public policy, Cheltenham: Edward Elgar
                            Publishing
                          Room, G, 2013, Evidence for agile policy makers: The contribution of transformative
                            realism, Evidence & Policy 9, 225–44
                          Rothwell, H, Shepherd, M, Murphy, S, Burgess, S, Townsend, N, Pimm, C, 2010,
                            Implementing a socialecological model of health in Wales, Health Education 110,
                            471–89
                          Ryall, T, 29 October 2009, Fruit in Schools future confirmed, Wellington: Minister
                            of Health
                          Sanderson, I, 2009, Intelligent policy making for a complex world: Pragmatism,
                            evidence and learning, Political Studies 57, 699–719
                          Smith, KE, Joyce, KE, 2012, Capturing complex realities: Understanding efforts to
                            achieve evidence-based policy and practice in public health, Evidence & Policy 8,
                            57–78
                          Snowden, D, 2011, Naturalizing sensemaking, in Mosier, K, Fischer, M (eds), Informed
                            by knowledge: Expert performance in complex situations, New York, NY: Psychology
                            Press, 223–4
                          Stacey R, 2012, Tools and techniques of leadership and management: Meeting the challenge
                            of complexity, Oxford: Routledge
                          Statistics New Zealand, 2013, New Zealand Official Yearbook 2012, www.stats.govt.
                            nz/yearbook2012
                          Stern, E, 2011, Foreword, in Forss, K, Marra, M, Schwartz, R (eds), Evaluating the
                            complex: Attribution, contribution and beyond, New Brunswick, NJ: Transaction
                            Publishers, vii–xi
                          Stern, E, Stame, N, Mayne, J, Forss, K, Davies, R, Befani, B, 2012, Broadening the range
                            of designs and methods for impact evaluations, London: Department for International
                            Development

                                                                       88
Setting the context for using complexity theory in evaluation

                          Treasury, 2010, Annex 6, Vote health expenditure and other reviews: Line-by-line
                            review of Ministry DE (Departmental Expenditure) and NDE (Non Departmental
                            Expenditure),Wellington: The Treasury New Zealand
                          Trenholm, S, Ferlie, E, 2013, Using complexity theory to analyse the organisational
                            response to resurgent tuberculosis across London, Social Science and Medicine 93,
                            229–37
                          Vincent, R, 2012, Insights from complexity theory for evaluation of development
                            action: Recognising the two faces of complexity, London, PANOS / IKM
                            (Information and Knowledge Management): Emergent Research Programme, 1–47
                          Walton, M, 2014, Applying complexity theory: A review to inform evaluation design,
                            Evaluation and Program Planning 45, 119–26
                          Walton, M, Waiti, J, Signal, L, Thomson, G, 2010, Identifying barriers to promoting
                            healthy nutrition in New Zealand primary schools, Health Education Journal 69,
                            84–94
                          Westhorp, G, 2012, Using complexity-consistent theory for evaluating complex
                            systems, Evaluation 18, 405–20
IP : 192.168.39.211 On: Tue, 30 Nov 2021 08:23:57

                          Williams, B. Hummelbrunner, R, 2011, Systems concepts in action: A practitioner’s toolkit,
                            Stanford, CA: Stanford University Press
            Copyright The Policy Press
                Delivered by Ingenta

                                                                                 89
You can also read