Determined to succeed: Can goal commitment sustain interagency collaboration?

Abstract

Governments have struggled with addressing problems that cross agency boundaries. Since 2012, the New Zealand Government has achieved significant success by holding groups of agencies collectively responsible for achieving intermediate outcome targets (the ‘Results Programme’). The Results Programme has been described as the most important change in how public services are delivered in New Zealand in 20 years. This article uses a mixed methods approach to triangulate 10 features of the Results Programme that appear to contribute to its success. Collaboration literature typically focuses on reducing barriers, often expressed in terms of transaction costs; in contrast, the successes of the Results Programme are explained here as methods for engineering a sense of joint goal commitment, that provides the sustained impetus to succeed despite the barriers encountered.

Keywords

Collaboration goal commitment mixed methods New Zealand

Introduction

Supporting interagency collaboration is a major area of interest for governance literature and practice (Carey and Crammond, 2015). Regardless of how government is divided, some problems will inevitably span the boundaries between agencies (Scott and Boyd, 2016). Addressing such problems has been described as both the ‘holy grail’ (Peters, 1998) and the ‘philosophers stone’ (Jennings and Krane, 1994) of public administration. The barriers to interagency collaboration are generally believed to be high (Altshuler, 2003; Glendinning, 2003; Hudson et al., 1999), and these barriers are frequently expressed in terms of transaction costs (Scott and Bardach, 2019; Thomson and Perry, 2006, Warner and Hefetz, 2008). Efforts to improve interagency collaboration have focussed on reducing transaction costs to make collaborative work easier and to progress more quickly – how to ‘lower the barriers’ to collaboration (Edmondson and Roloff, 2009). This article takes a different approach, and asks whether goal commitment might be used to clear those barriers, or else to smash through them.

From 2012 until 2017, the New Zealand government operated an interagency collaboration programme called the Better Public Services Results Programme (the ‘Results Programme’). The Results Programme combined target setting, public reporting, shared responsibility, and various collaborative governance and management practices. It achieved process success, in that it was held up by public servants as an example of successful collaborative behaviours (Scott and Boyd, 2017). More importantly, it achieved programmatic success (Marsh and McConnell, 2010), in that the Results Programme improved outcomes in all 10 problems to which it was applied (State Services Commission, 2017).

This article considers why the Results Programme was successful, when other interagency collaboration programmes in New Zealand had often failed (Scott and Boyd, 2016). This article uses a mixed methods analysis to triangulate findings from four sources. The analysis identifies features that we interpret as contributing to the success of the Results Programme. These features seem to contribute to a shared sense of commitment felt by participating agencies to achieving the desired goal.

This article includes five parts following this introduction. First, there is a description of the New Zealand context, the Results Programme, and why this should be of scholarly interest. Second, the literature on interagency collaboration and goal commitment are reviewed and summarised. Thirdly, the methodology is described. Fourth, findings are presented. Finally, there is a discussion of how these findings contribute to goal commitment.

The New Zealand context

New Zealand is a parliamentary democracy modelled after the Westminster style of government, with the public sector divided into a large number of single-purpose agencies. Ministers enjoy a high level of autonomy over policy within their warrant and are incentivised to defend their own policy priorities rather than the collective good (Gregory, 2006). New Zealand is regarded as the purest example of ‘New Public Management’ (Pollitt and Bouckaert, 2011) with a focus on single-point accountability (Gorringe, 1994; Gregory, 2006), and this may make collaboration even more difficult (Boston et al., 1996). Public servants have become accustomed to this system, and react to calls to work together by trying to read implicit signals about who will really be held accountable for the success of failure of the nominally joint activity (Schick, 2001; Scott, 2001). While barriers to interagency collaboration have been reported from many jurisdictions (Carey and Crammond, 2015), New Zealand’s New Public Management means these barriers are likely to be raised further.

The results programme

After several failed attempts at encouraging collaboration (Scott and Boyd, 2016), in 2012 the New Zealand Government tried a new approach, which combined lessons from literature on collaboration (Carey and Crammond, 2015), prioritisation (Kohli, 2010), and target setting (Bevan and Hood, 2006). This was described in popular media as the most significant change to how government services are delivered in New Zealand in 20 years (Shelton, 2013). The government chose 10 important and long-standing problems where progress required collaboration between multiple agencies, and set a challenging five-year target for each (listed in Table 1). The targets were declared publicly, with a pledge to report on progress every six months. Small groups of chief executives were collectively answerable to Ministers for progress towards each target. The Results Programme was implemented in a context of significant fiscal constraint, with most participating agencies operating under flat or declining nominal baseline budgets. The Results Programme had no dedicated budget and agencies were expected to make better use of resources already allocated to addressing these 10 problems.

Table 1.

BPS results and targets.

BPS result	Targets
(1) Reduce the number of people who have been on a working age benefit for more than 12 months	Reduce the number of people continuously receiving working age benefits for more than 12 months by 30%
(2) Increase participation in early childhood education.	98% of children starting school will have participated in quality early childhood education
(3) Increase infant immunisation rates and reduce the incidence of rheumatic fever.	Increase infant immunisation rates so that 95% of eight-month olds are fully immunised andReduce the incidence of rheumatic fever by two-thirds.
(4) Reduce the number of assaults on children.	The 10-year rise in children experience physical abuse will be halted and current numbers reduced by 5%.
(5) Increase the proportion of 18-year olds with NCEA level 2 or equivalent qualification.	85% of 18-year olds will have achieved NCEA level 2 or an equivalent qualification.
(6) Increase the proportion of 25- to 34-year olds with advanced trade qualifications, diplomas and degrees (at level 4 or above).	55% of 25- to 34-year olds will have a qualification at level 4 or above.
(7) Reduce the rates of total crime, violent crime and youth crime.	Reduce the crime rate by 15% and reduce the violent crime rate by 20% and reduce the youth crime rate by 5%.
(8) Reduce reoffending.	Reduce the reimprisonment rate by 25%.
(9) New Zealand businesses have a one-stop online shop for all government advice and support they need to run and grow their business.	Business costs from dealing with government will reduce by 25%.
(10) New Zealanders can complete their transactions with the Government easily in a digital environment.	An average of 70% of the most common transactions with government will be completed in a digital environment (up from 29.9% baseline).

After four years, all 10 outcomes were measurably improved both in absolute terms and relative to trend data where available (see State Services Commission, 2017). We consider this evidence of programmatic success (Marsh and McConnell, 2010), but should we? McConnell (2010: 354) describes a programmatic success as the ‘achievement of desired outcomes’. At the time that the analysis presented in this article was conducted (early 2017), the programme was not complete. Later in 2017, there was a general election resulting in a change of government. The programme was discontinued by the incoming government and no final report was issued, although a newspaper article at the time indicated that ‘about two thirds of the targets were met while others fell short’ (New Zealand Herald, 2018). If the ‘desired outcomes’ of the programme were the targets, then the programme was a partial success. If the ‘desired outcomes’ were to improve performance, above baselines and trends, in 10 stubborn crosscutting problems, then the programme was a success in all 10. The authors take the latter interpretation and assume that the targets are a design feature intended to drive performance improvements. We take certain events as suggestive that the government saw the programme this way as well. In 2014, it became apparent that three of the targets would be achieved earlier than anticipated. If the purpose was only to achieve the targets, the government could have responded by claiming victory. Instead the government responded by resetting those three targets to be more difficult (State Services Commission, 2014), suggesting that the true purpose of the programme was to use targets drive performance improvements.

In discussions with public servants involved with the Results Programme, one theme was repeated – progress was never easy (Orange, 2016; Scott and Boyd, 2017a). It involved continued effort to overcome the administrative barriers of a system designed around single-point accountability, siloed working and patch protection. While the Results Programme was a (process and programme) success, it was a hard-won success borne of persistence and dedication.

Interagency collaboration and goal commitment

Interagency collaboration

Working across agency boundaries has become a core part of contemporary arrangement for public service provision (Esteve et al., 2015; O’Leary and Vij, 2012). Despite extensive literature (Carey and Crammond, 2015; Page, 2003), working across agency boundaries remains a fraught and costly endeavour (Bryson et al., 2015; Doberstein, 2016; Minson and Mueller, 2012; Prebble, 2015; Scott, 2015).

‘Coordination’ or ‘collaboration’?

While some authors use the term ‘collaboration’ in a general sense to describe a range of practices for interagency working (see for example, Axelsson and Axelsson, 2006), other authors use collaboration to refer to one category of practice among others like ‘cooperation’ and ‘coordination’ (Eppel et al., 2008; Gadja, 2004; Gregson et al., 1992; Howarth and Morrison, 2007; Huxham and Macdonald, 1992; Keast et al., 2007; Marrett, 1971; Sadoff and Grey, 2005; Thomson and Perry, 2006). Several of the earlier references in this article describe practices that they call coordination (Jennings and Krane, 1994; Peters, 1998), though it would be inaccurate to suggest that clear boundaries exist (Thomson and Perry, 2006). There is a general sense in the literature that, when compared to cooperation or coordination, collaboration involves greater ‘depth of interaction, integration, commitment, and complexity’ (Thomson and Perry, 2006: 23), but unfortunately these terms are not used precisely or consistently. Gray (1989) suggests that collaboration is distinguished in part because it involves integration over a longer period, with cooperation and coordination being perhaps early developmental stages. Keast et al. (2007) suggest that collaboration is distinguished by the actions that are performed together; collaboration involves shared working, rather than merely shared information or shared planning. Eppel et al. (2008) distinguish collaboration by its responsibilities; collaboration involves shared responsibility for a common goal, with other forms of interagency working involving cooperating or coordinating in areas of common interest and overlap while maintaining separate responsibilities. It is this latter distinction that is most relevant for this article, and we define interagency collaboration as the practice whereby administrative divisions of government take shared responsibility for achieving a common goal. The methods employed in achieving this outcome might include sharing information, jointly developing plans and strategies to better align activities (sometimes called coordination) or working together to jointly develop and deliver solutions (sometimes called collaboration), or a combination.

A related term, popularised by the UK’s Blair government, is ‘joined-up government’, which is also used inconsistently and ambiguously in the literature (Carey and Crammond, 2015), and refers to a range of practices (Ling, 2002). Ling (2002) presents examples of ‘joined-up government’ from different jurisdictions, includes practices characterised as both ‘coordination’ (p. 628) and ‘collaboration’ (p. 620, 621), and suggests collaboration as a means to achieve coordination (p. 629).

The field of interagency collaboration is sometimes included within broader collaborative literature alongside ‘cross-sectoral collaboration’ (Bryson et al., 2015). In cross-sectoral collaboration, the public sector is one actor among others. While some of the practices in interagency collaboration may be similar to those in cross-sectoral collaboration (such as the use of governance structures and processes), there may also be important differences. Interagency collaboration is frequently imposed from above (by politicians), whereas cross-sectoral collaboration more regularly involves the recognition by participants of overlapping or aligned interests or capabilities (Provan and Kenis, 2008), and the recognition of collaborative opportunity (Bryson et al., 2015).

Context and generalisability

Carey and Crammond (2015) provide a recent synthesis of empirical evidence specific to collaboration in an interagency context, and identified the following as features found to promote effective collaboration: political support, leadership at all levels, a combination of top-down and bottom-up approaches, decentralised control, and designing for context. While some ‘instruments’ (management structures and processes) were seen as useful in some contexts, there was little commonality across multiple studies. It is therefore with some trepidation that we consider which contextual or contingent factors (Donaldson, 2001) may explain the findings of this New Zealand case.

One defining characteristic of the Results Programme was that the problems selected were those that had been resistant to past attempts at remedy. They were stubborn problems (Scott and Boyd, 2017), and nothing about their context in 2011 suggested that they represented a particular ‘collaborative opportunity’ (Bryson et al., 2015). Nonetheless, they were selected as priorities by the Government of the day. The global financial crisis meant that there were no additional resources available. The New Zealand context was one of individual accountability without a strong history of collaborating successfully. For these reasons, we consider this a case study not of how to make interagency collaboration easy, but how to succeed when it is hard.

Costs and barriers

Interagency collaboration is hard for many reasons. Rather than the efficiency of a hierarchical chain of command, interagency collaboration is an ongoing negotiation (Thomson and Perry, 2006) with significant coordination costs (Van Huyck et al., 1990). Decisions taken by an individual may not be supported by others, and so decision-making often occurs by consensus (Ren et al., 2005), which can be very slow. Goals may be imperfectly aligned, and agencies may act opportunistically to pursue their own goals (Laan et al., 2011). Working together may mean more optimal use of resources for production, but the transaction costs tend to be much higher (Krueger and McGuire, 2005).

Goal commitment

A commonly considered construct for explaining task performance is ‘goal commitment’ (Klein et al., 2013), based on a body of psychology literature known as ‘goal setting theory’ (Locke and Latham, 1990). Goal commitment is defined by Klein et al. (2013: 67) as ‘a volitional psychological bond reflecting dedication to, and responsibility for, a particular target’.

Goal commitment is well established as a contributor to performance in management literature from the private sector (for example, Erez and Judge, 2001; Schweitzer et al., 2004), with a small but growing consideration in public administration literature (for example, Bronkhorst et al., 2015; Favero and Bullock, 2015; Jung and Ritz, 2014; Latham et al., 2008; Pedersen, 2016; Rutherford and Meier, 2015). While important caveats and limitations to the measurement of goal commitment remain, there is general agreement on the positive relationship between goal commitment and performance (Klein et al., 1999). This is explained by the finding that, compared to less goal-committed individuals, highly goal-committed individual exerts greater effort and persistence towards goal attainment (Seijts and Latham, 2000); and are more likely to develop task-related strategies (Earley et al., 1992). Studies on the antecedents of goal commitment identify these as the expectancy of goal attainment and the attractiveness of goal attainment (Hollenbeck and Klein, 1987; Klein and Wright, 1994). Individuals were more highly goal-committed when the goal was perceived as difficult, but with a sharp drop in commitment if the goal was perceived as impossible. Expectancy was higher in cases of high self-esteem, and in the presence of social comparison where others had been successful. The attractiveness of goal attainment was in turn influenced by the anticipated reward structure, when there was a credible expectation of consequences (rewards and sanctions) associated with goal achievement. In the public sector, Wright (2007) linked the attractiveness of goal attainment to concepts of mission valence (the public servants’ perception of the social contribution of the goal) and public service motivation (the desire to serve the public interest).

Goal commitment literature is primarily concerned with individual performance. In those cases where team performance is considered (Aube and Rousseau, 2005; Hoegl and Parboteeah, 2006), these are teams within the same organisation, rather than spanning multiple organisations or agencies. Interagency collaboration has historically been considered unsuitable for goal setting because of ‘the problem of many hands’ (Thompson, 1980), or the inability to disentangle the contributions made by individuals to the observed collaborative outcomes. Several recent cases, such as Virginia Performs, GMAP Washington, Scotland Performs, Victoria Ahead (see Kohli, 2010), and now New Zealand’s Results Programme, suggest that interagency goals may be more useful than previously considered.

Methodology

The Results Programme has been evaluated separately using four evaluation methods, each described below and available online. The combination of these evaluations through a mixed methods approach is likely to result in more adequate explanations than any one method alone (Creswell, 2013). This section describes the use of triangulation and complementarity to analyse the results of these evaluations, and limitations of this approach.

Evaluation methods

The first evaluation was published in a peer-reviewed journal (Scott and Bardach, 2019), and featured a comparative study (Yin, 1981). The relative progress towards each of the 10 goals was attributed to differences between them (for example, target type, relationship history, and governance settings). This method had limitations that may reduce both the reliability and completeness of the findings: the method relied on source documents written by public servants involved in the Results Programme, and their decisions to include or exclude material represent unavoidable selection bias (Heckman, 1990); the methodology only analyses features that differ between the 10 goals, and cannot draw findings on the importance of the many areas in which work towards the 10 goals was alike; and the analysis assumes that success is equivalent to the degree of progress towards the goal, which does not allow for the possibility that the goals may differ in their degree of challenge.

The second evaluation (Scott and Boyd, 2016a) compared programme features against process success, drawn from the normative ‘collective impact’ framework (Kania and Kramer, 2011; Preskill et al., 2014). The use of process measures reduced confounding errors associated with the relative difficulty of individual goals, but is limited by the assumption that the process measures in the Collective Impact framework were valid. The Collective Impact framework was originally designed for use in cross-sector collaboration between organisations in the government, corporate, and community sectors but was adapted for government use with help from the original authors (Hallie Preskill, 26 August 2015, personal communication; Mark Kramer, 25 August 2015, personal communication).

The third evaluation (Scott and Boyd, 2016b) used a multiple case study method to explore individual success stories (George and Bennett, 2005). Public servants involved in the Results Programme identified 89 cases that they believed best illustrated the success of the programme, and described these in short vignettes of 200–800 words (State Services Commission, 2016). The evaluation analysed these case vignettes through exploratory coding (Manning and Cullum-Swan, 1994) and identified a number of recurring themes about government practices that were shown to make a difference. The main strength of this method is that it performs analysis at a more granular level (with 89 rather than 10 elements). However, this method introduces a new type of selection bias, as public servants submit case vignettes that conform to their perspective on success and include only elements that they believe are pertinent to the story.

The fourth evaluation (Scott and Boyd, 2016c) explored behavioural change (Simons, 2013). A literature search was conducted of government documents and media reporting, and analysed thematically (Fereday and Muir-Cochrane, 2006), looking for descriptions of how behaviours had changed in response to the Results Programme and specifically the introduction of performance targets. This provided an additional point of comparison, of behaviours before and during the programme.

Triangulation and complementarity

This paper uses a mixed methods approach to improving the reliability of the above evaluations and the depth of substantive understanding of the phenomena being studied (Greene, 2007). Mixed method research combines qualitative and quantitative data, methods, and/or paradigms from related studies (Teddlie and Tashakkori, 2009). This article uses mixed methods for two purposes: triangulation and complementarity. Triangulation seeks to corroborate findings from different methods to overcome the limitations of each of the preceding evaluations, in order to improve the validity of inferences about constructs and the strength of the warrant for inquiry results (Campbell and Fiske, 1959). Complementarity seeks to elaborate or illustrate findings from one method using findings from another method, or to develop a more comprehensive understanding of the construct(s) being studied (Rossman and Wilson, 1985). This article begins by using triangulation to identify the best-supported findings from the previous evaluations, and then uses complementarity to explain and elaborate on each of the best-supported findings, explored further below.

The four evaluations are good candidates for mixed methods study because they represent reasonably diverse methods for understanding the same phenomenon from different perspectives (Greene, 2007). The four evaluations use a mix of developmental, formative and summative evaluation (Patton, 2011), normative and positive reference (Jensen, 1983), and qualitative and quantitative data (Teddlie and Tashakkori, 2009), as shown in Table 2. It is difficult to objectively weight the relative status or value of each evaluation, hence they have been weighted equally for the purpose of analysis.

Table 2.

Summary of evaluation methods.

	Development stage	Data type	Data source	Reference
Scott and Bardach (2019)	Formative, summative	Qualitative and quantitative	Mostly internal (documents written by public servants)	Comparison between 10 case elements
Scott and Boyd (2016a)	Developmental, formative, summative	Qualitative	Mix internal/external	Comparison to (normative) collective impact framework
Scott and Boyd (2016b)	Summative	Qualitative and quantitative	Internal (accounts by public servants)	Descriptive of case vignettes
Scott and Boyd (2016c)	Formative	Qualitative	Mostly external (media reporting)	Comparison between past and present behaviour

Triangulation

Each of the four evaluations describes aspects of the Results Programme that may have contributed to its success. These findings were grouped into three categories: strongly supportive, weakly supportive and opposing. The selection of these three categories was largely opportunistic, as two of the four evaluations were already structured in this manner. In the case of the other two evaluations, findings were assessed as strong if they were described in the evaluation reports as supported by strong evidence, or evidence that suggested a strong effect size. Findings were assessed as weakly supportive if they had mixed evidence (supporting/opposing) with the majority supportive, if there was weak supportive evidence only, if the evidence came from only a small number of cases, or if the evidence suggested a small effect size. As the findings of each evaluation were mostly qualitative, there was some subjectivity in the assessments. The research designs of the original evaluations were such that only two evaluations reported findings that were associated with less success (described as ‘opposing’).

The findings from the four evaluations were then compared, to look for features that appeared in more than one evaluation method. Findings that were consistently supported or opposed by more than one method are more reliable than those identified by one method only (Greene, 2007). Some features are supported by one method and opposed by another, or described by only one evaluation method; these cannot be corroborated using the research design, and are excluded from the findings. The results of the triangulation method are shown in ‘Findings’ section of this article.

Complementarity

Each of the findings that were corroborated using the triangulation method above were then analysed further to provide descriptive accounts of these findings. Each evaluation report was re-read to find any descriptive information relating to the corroborated findings, and a summary of this analysis forms ‘Actor selection’ to ‘Goal monitoring’ sections of this article.

Limitations

This study is limited in completeness, external effects and the extent to which findings can be generalised or applied in other contexts, as discussed below. Triangulation and complementarity are used to compare multiple descriptions of the same phenomena (Creswell, 2013). Therefore it is only useful when that phenomenon is described in multiple evaluations, and any features identified in only one evaluation method are ignored. Some of these features identified in single evaluations may represent real success factors in the Results Programme that may be important in trying to replicate the approach elsewhere. Triangulation is therefore likely to be more reliable than the individual evaluations and produce less spurious associations, but the findings that can be corroborated from multiple evaluations are likely to be less complete and omit several real associations.

While the four evaluations are diverse in their methods, they focus their study on the Results Programme itself and ignore broader questions of performance in the New Zealand public service. The Results Programme has been successful in improving performance in all 10 cases, but it is unknown whether this improved the performance of the New Zealand public service as a whole. None of the contributing evaluations are able to assess any changes in performance in areas outside the Results Programme. It is plausible that the Results Programme created focus on a small number of problems and improved performance in those areas, but at the expense of performance in other areas.

Each evaluation uses publicly available data, and may therefore miss other important sources of information. In a research synthesis, Carey and Crammond (2015) describe the importance of informal, and often subversive, managerial craft (as described in Bardach, 1998). Such elements are typically important in New Zealand (Lips et al., 2011), but are unlikely to be recorded in the source documents used in this article (Scott and Bardach, 2019). This is one limitation to the methods used and an opportunity for further research.

This article relies on secondary data, in the form of the prior evaluations. These evaluations were not conducted for the purpose of exploring goal commitment, and the subsequent link to goal commitment drawn in this article was generated inductively. As with all inductive reasoning, this has the advantage that the original (primary) data collection and analysis was not conducted with a particular goal in mind, and this reduces one potential source of bias. The disadvantage is that the prior evaluations, by not looking specifically for data relating to goal commitment, may have ignored potentially useful data that would have led to a deeper understanding of the phenomena. At least one of the authors of this article was involved in each of the prior evaluations, and this too has advantages and disadvantages. Prior involvement means that the authors had a good understanding of the case and methods used in these evaluations, and had access to the primary data as reference when this was useful for verifying or expanding upon a particular inference from the evaluation reports. One disadvantage of prior involvement is that any biases or assumptions made by the authors in the prior evaluations may be perpetuated in this article. The use of secondary sources, the inductive approach, and prior involvement of authors, are each therefore further sources of limitations to this study.

Each of the 10 cases in the Results Programme shares many of the same contextual factors: the same time period (2012–2017), the same Government (the fifth National Government), and the same jurisdiction (New Zealand) with its somewhat unusual public management system. Attempts to implement the recommendations from this article in other jurisdictions should be cognisant of the impacts of contextual factors, and potential implementers of this approach are encouraged to contact the authors to discuss how the Results Programme may be adapted to fit their problem setting.

This article earlier considered the difficulty and ambiguity involved in terms like ‘coordination’ and ‘collaboration’, and these ambiguities create difficulties in grounding the findings within the broader literature. The case study describes a collaboration in the sense of joint working with shared responsibility for a common goal. It describes a range of activities performed by agencies in service to these goals, including some that could be described as both coordination and collaboration.

Findings

The four previous evaluations included a total of 45 findings; 10 of these were supported by more than one evaluation, listed in Table 3 and discussed below. For the purpose of discussion, these are grouped into three themes: actor selection, goal selection and goal monitoring. Table 3 demonstrates the value of the mixed methods approach; no single evaluation had identified all 10 features described in this study, and each evaluation included several associations that could not be corroborated.

Table 3.

Key findings of previous evaluations.

	Scott and Bardach (2019)	Scott and Boyd (2016a)	Scott and Boyd (2016b)	Scott and Boyd (2016c)
Actor selection
Maximum of three agencies	Strong	NA	Strong	Strong
Building on existing relationships	Strong	NA	NA	Strong
Cascading levels of governance	Strong	Strong	NA	NA
Goal selection
Selecting a small number of problems	NA	Strong	NA	Strong
Public commitment and transparent reporting	NA	Strong	NA	Strong
Specifying results at an intermediate-outcome level	Weak	NA	NA	Strong
Alignment between results and targets	Weak	NA	NA	Weak
Goal monitoring
Aligned accountability systems	Strong	NA	NA	Strong
Fast feedback and learning	Weak	Strong	Strong	Weak
Sharing administrative data	NA	Strong	Strong	NA

Strong: this feature was strongly supported as contributing to the success of the BPS Results approach; Weak: this feature was weakly supported as contributing to the BPS Results approach; NA: not applicable, this feature was not described in this evaluation.

Actor selection

Three of the features identified through more than one evaluation method related to the selection of responsible parties or actors:

Maximum of three agencies

Building on existing relationships

Cascading levels of governance

Achieving each target involved contributions from many agencies, but the most successful approaches limited core decision-making to two or three agencies, utilising tiers of involvement when necessary. In some cases, more than three agencies had an interest in the outcome, but greater progress was made when two or three critical partners formed a core group, and involved others on an as-needed basis for information sharing and coordination. Network theory notes the tension between efficiency and inclusion (Provan and Kenis, 2008), with the assumption that coordination (transaction) costs of a group rise as membership increases (Van Huyck et al., 1990; Weber, 2000). An alternate explanation is that felt responsibility and individual effort diminish as group size increases, as observed experimentally in the Ringelmann effect (Ingham et al., 1974).

Collaboration appeared more successful when built upon trusting relationships between contributing parties, ideally with a track record of succeeding together even if on smaller projects. Thomson and Perry (2006) explain that trust reduces transactions costs by reducing the need for monitoring. An additional explanation is that where agencies value the relationship, they feel a psychological contract to deliver on their promises to each other (Ring and Van de Ven, 1994), raising commitment (see ‘Goal commitment in collaboration literature’ section for further discussion of commitment, commitments and relationships).

Agencies worked together with cascading governance (Kania and Kramer, 2011) at ministerial, chief executive, policy, operational and secretariat level. New Zealand public servants had been socialised in a system of single-point accountability (Gregory, 2006) and therefore reportedly looked for signals as to where that single point lay. The most successful governance arrangements appeared to go to great lengths to reinforce the idea of shared and equal responsibility. One example was in the resourcing of secretariats, where successful groups jointly resourced a secretariat including staff from each agency. In contrast, where a secretariat group was resourced by a single agency, this appeared to act as a signal to other agencies that they would be held less responsible for achieving the goal, resulting in those agencies feeling less committed.

Goal selection

Four features that were identified by more than one evaluation method related to the selection of goals:

Selecting a small number of problems

Public commitment and transparent reporting

Specifying results at an intermediate-outcome level

Alignment between results and targets

While it seems unlikely that 10 is a ‘magic’ number, limiting the approach to a small number of problems appears important in maintaining the attention of senior leaders. Several authors report an inverse relationship between the number of targets and the effectiveness of a performance regime (Barber et al., 2010; Boyne and Gould-Williams, 2003; Chun and Rainey, 2005; Weiss and Piderit, 1999). This may be because having a small number of problems raises the relative perceived importance of each problem, raising the stakes (or ‘attainment attractiveness’ – Klein and Wright, 1994) for success or failure in each goal.

A notable component of the Results Programme was the Government’s promise to regular and transparent reporting over five years. The New Zealand public sector has a high degree of churn, with many programmes cancelled or altered before their completion (Norman and Gill, 2011). When it is not clear that a programme will persist long enough for leaders to be held responsible for success or failure, this potentially acts to limit the felt commitment to achieving the desired outcome. Publicly setting and reporting on target progress and achievement creates a sense of commitment by increasing a programme’s potential exit costs (Cels et al., 2012). Additionally, public reporting of performance against a target can be a strong incentive to perform (Bevan and Fasolo, 2013), as public servants are typically strongly motivated by reputation – that is, being perceived as doing a good job (LeGrand, 2007). These reputational consequences may increase the extrinsic motivation towards goal attainment.

Great care was taken to choose goals that had robust synecdoche with the problem targeted and yet relatively strong attribution to the changes in service delivery. The New Zealand government describes these as ‘intermediate outcome’ measures (Baehler, 2003). By contrast, long-term outcomes, typically with poor attribution, are described as ‘end-outcomes’. Progress measures, output measures and end-outcome measures have each been criticised as poor methods of managing collaborative performance (Carey and Harris, 2016; Scott and Boyd, 2016). The use of intermediate-outcome measures in the Results Programme may have supported goal commitment by balancing the meaningfulness of the goal and the instrumentality of the actors in achieving that goal.

Participants in the Results Programme distinguish between a result (the outcome sought), a target (the degree of change) and the measure (how that change will be assessed). Results, targets and measures tend to be conflated in other target literature (for example, Smith, 2009). Separating these three components also creates risks – where there was poor fit or alignment particularly between the result and the target this drove confusion and distraction. Goal ambiguity tends to lessen feelings of commitment (Chun and Rainey, 2005).

Goal monitoring

Some goals are qualitative and subjective in nature. In contrast, the Results Programme set quantitative goals where there was little subjectivity in whether the goal had been achieved. Triangulation corroborated a final three features that related to the monitoring of the achievement of these quantitative goals:

Aligned accountability systems

Fast feedback and learning

Sharing administrative data

Public servants are required to pursue multiple competing objectives. The Results Programme aligned goals across several accountability systems, with the same goals and actors included in reports to Cabinet, agency planning documents presented to Parliament, and in the assessment of performance by the State Services Commission (Scott, 2016). This consistency likely further increases the attractiveness of goal attainment.

The Results Programme has been characterised as data driven (Jensen et al., 2014). The most ‘successful’ of the 10 goals involved measurement systems that minimised the delay between actions and observed effects. These delays can occur in two places: between the action and its impact on societal outcomes (‘performance lag’ in Figure 1); and between the change in outcome and its measurement (‘measurement lag’ in Figure 1). Measurement delays can cause many problems: cancelling successful initiatives because they haven’t yet produced measurement results; continuing with unsuccessful initiatives because their impact is not yet known; and demotivating public servants and ministers because the impacts of recent actions cannot be seen. One interpretation of the data is that low delay allowed more adaptive management (Carey and Harris, 2016) because public servants could see what was working and make a change (‘adaptive response’ in Figure 1). Another interpretation is that up-to-date data allowed the pressure of goal attainment to be felt more acutely; all parties (public servants, ministers, and the public) had accurate and timely information on progress towards the goal, and this supported the accountability systems described above.

Figure 1.

Delays between action and measurement.

The most successful of the 10 cases created a shared measurement system that allowed the use of data to support decision-making. This made it easier for all parties to understand both what was happening and the underlying causes. In public administration, where causal relationships can be indirect and less easily identified (Wilson, 1989), this deeper data analysis may be necessary to better target services. Access to this data also made performance more transparent, and therefore may have contributed to a greater sense of felt responsibility.

Discussion

The Results Programme consists of 10 previously intractable problems, and in each case outcomes have improved above baselines and trend data (where available) without the addition of new resources. The section above describes 10 features of the Results Programme that may have contributed to its effectiveness.

Some of these features can be found in other literature on interagency collaboration, either in empirical studies (see Carey and Crammond, 2015 for an overview), or theoretical frameworks (see in particular Ansell and Gash, 2008; Bryson et al., 2015; Emerson et al., 2012, Thomson and Perry, 2006), where they are described as ‘reduc(ing) complexity and transaction costs’ (Thomson and Perry, 2006: 28). Interagency collaboration is generally thought to be difficult, and efforts to identify how the barriers to collaboration can be lowered are commendable.

In contrast, in this article we explain these same features and others, not as making interagency collaboration easier, but instead as mechanisms for maintaining persistence and determination when collaboration is hard, through goal commitment. Some examples of how the same feature can be explained using different theories are shown in Table 4. The goal commitment theory was more consistent with the experiences of public servants, who reported that collaborating was hard. In a memoire on working in the Results Programme, Deputy State Services Commissioner Ryan Orange commented: ‘All of the Results teams met significant opposition to breaking from previously established approaches to governance, decision-making, funding, partnerships and delivery’ and ‘the passion of the leaders and teams for improving the lives of New Zealanders was essential in sustaining their efforts to overcome resistance to working differently’ (Orange, 2016: 17).

Table 4.

Observed features of successful collaboration can be explained as reducing transaction costs or increasing goal commitment.

	Reducing transaction costs	Increasing goal commitment
Selecting a small number of goals	Reduces monitoring costs	Increases attainment attractiveness of each goal
Building on existing relationships	Reduces monitoring and coordination costs	Magnifies the importance of psychological contracts
Maximum of three agencies	Reduces coordination costs	Increases the instrumentality of each actor

In this section, we propose a framework for creating and sustaining goal commitment in interagency collaboration, as an explanation of the effectiveness of New Zealand’s Results Programme. We begin by exploring how concepts similar to goal commitment have been used in collaboration literature before, and then attempt to construct a novel framework to accommodate the findings of this article.

Goal commitment in collaboration literature

The concept of commitment is not new to collaboration literature, though different authors have used the term in slightly different ways. Thomson and Perry (2006) describe a multi-dimensional framework that involves the cyclical and reciprocal relationship between negotiation, commitment, implementation and assessment (based on Ring and Van de Ven’s, 1994 ‘process framework of collaboration’). Thomson and Perry explain that these processes work together to reduce complexity and transaction costs between agencies. It is apparent from a general reading of the article however, that Thomson and Perry (2006) intend ‘commitment’ to stand for the ‘commitments’ or promises that agencies make to each other (for example ‘commitment to future action’, p. 22). This same sense of commitment as promises is used by other authors, such as Imperial (2005), and differs from the sense of being committed to achieve a target or goal. Indeed, the parliamentary system of government in New Zealand meant that public servants were limited in the explicit ‘commitments’ that they could make to each other because these might be overruled by the priorities of individual ministers. Instead, they could only continually renegotiate provisional agreements and adapt to changing circumstances, motivated by a sense of commitment to a shared goal.

Ansell and Gash (2008) describe the importance of commitment to collaborative process, and propose that collaborative will be more successful if the participating parties feel obliged to adhere to the agreed governance mechanism. This gets closer to the ideas discussed in this article, but still describes a slightly different phenomenon. Between 2004 and 2012, New Zealand public servants were committed to working together (see ‘the sector approach’, Scott and Boyd, 2016) but with mixed results. The Results Programme differed because it was no longer sufficient only to work together, but instead public managers had to be resourceful and creative to reach the goal to which they felt committed.

Much of the collaborative governance literature notes the importance of trusted relationships (Ansell and Gash, 2008; Huxham and Vangen, 2000). This too can be related to the word ‘commitment’, in the sense of being feeling a volitional bond or felt obligation to an other person or persons as necessary or implied in the nature of the interpersonal relationship. Such a sense of obligation may be specific (the promises discussed above) or tacit and general, in the sense of feeling obliged to supporting each other, which Ansell and Gash (2008) relate to the importance of commitment to collaborative process. The nature of much of the data used in this article makes it difficult to comment on the relationships within the Results Programme; the primary sources in three of the four evaluations were written by public servants, with an understandable reluctance to comment on specific relationships in publicly available documents. The relationship between goal commitment and interpersonal or relationship commitment would be an interesting area for future study, and may potentially augment the framework proposed later in this article.

Huxham and Vangen (2013) identifies several themes that may influence collaborative success, and notes that practitioners frequently identify ‘commitment and determination’ as an important factor, but then omits this from further discussion. Without further description it is hard to be certain, but we suspect that this concept of ‘commitment and determination’ might be similar to the ‘commitment to goal achievement’ described in this article. Goal commitment has been described elsewhere as having two dimensions: intensity and duration (Nutt, 1983; Scott, 2014, 2018; Scott et al., 2015), and these two dimensions may be what was intended by the dual name ‘commitment and determination’. Similarly, Provan and Milward (2001: 416) mention the importance of ‘commitment to network goals’ as a determinant of network success, without providing guidance on how this may be engineered or supported.

If the collaboration literature is not particularly forthcoming on the concept of commitment to goal achievement, perhaps there are lessons from related concepts, like ‘motivation’ – the desire or willingness to do something. Motivation has been an area of considerable interest in collaboration literature (see Bresnen and Marshall, 2000; Burkhardt, 2001; Gazley, 2014; Mulroy and Shay, 1998). However, in these articles, motivation is seen as an initial condition – that is, individuals or organisations being motivated to begin to collaborate with others. Past attempts at collaboration in New Zealand have begun with great enthusiasm only to fade after a couple of years (see Baehler, 2003; Jensen et al., 2014). The Results Programme has differed in that commitment has been sustained over time.

While constructs related to goal commitment can be found in collaboration literature, none are sufficient for describing the sustained determination and perseverance that characterised the Results Programme.

A framework for creating and sustaining goal commitment in interagency collaboration

We interpret the case study as an example of goal commitment being used to achieve and sustain interagency collaboration, and note above that goal commitment does not appear sufficiently explained in the interagency collaboration literature. This section proposes a theoretical framework for understanding goal commitment in an interagency collaboration context that relates goal commitment theory, mission valence theory, and the data from the case study (as recorded in prior evaluations).

The ‘Findings’ section of this article identifies and describes 10 features of the Results Programme that multiple prior evaluations found contributed to its success. Even assuming that each of these 10 features are indeed truly contributive, it is likely that some are specific to the context of the case study, or that these features are specific examples of how a more general principle can be expressed. Therefore, in this section we try to group these 10 further, and relate these to goal commitment theory, in the hope that at this more generic level it is possible to identify antecedents to goal commitment that have more generalised relevance.

The findings were earlier grouped into three themes: actor selection, goal selection and goal monitoring (‘Actor selection’, ‘Goal selection’, ‘Goal monitoring’ sections, respectively, of this article). These groupings were subjective, but appeared to the authors to represent reasonable thematic categories for the 10 findings.

In the goal commitment literature, antecedents to goal commitment are described as having two primary components, consistent with expectancy theory (Vroom, 1964): expectation of goal attainment and attractiveness of goal attainment (Klein and Wright, 1994). Thus as a basis for generating a theoretical framework, the authors were able to combine the three groups of findings in this article (actor selection, goal selection, and goal monitoring), with the two primary components of expectancy theory (expectation and attractiveness), which made possible a three-by-two grid (see Table 5). Then, the findings were reexamined to determine if any of the contributive features of the Results Programme could be related to categories in Table 5, and further (below), whether these could be related to relevant theory identified earlier. We propose this framework as six principles for programme design to build and sustain goal commitment to interagency collaboration.

Table 5.

A framework for creating and sustaining goal commitment in interagency collaboration.

	Actor selection	Goal selection	Goal monitoring
Expectation	Instrumentality of actors	Likelihood of attainment	Likelihood of programme completion
Attractiveness	Distribution of responsibility	Intrinsic motivation	Extrinsic motivation

Instrumentality of actors

In goal commitment literature, attainment of the goal is something that is usually assumed to be the direct result of the performance of the individual (see review article by Latham et al., 2008). In public administration, the direct activities of public servants are frequently only indirectly related to the intended social outcome (see the production model of public performance, Van Dooren, 2006). In the Results Programme, the goals were selected at an ‘intermediate-outcome’ level, where the achievement of the goal was at least partially attributable to the actions of those public servants held responsible.

Likelihood of attainment

The likelihood of goal attainment is recognised as a contributor to goal commitment, with effort increasing as goals become more difficult and then dropping sharply if attainment becomes very unlikely (Klein and Wright, 1994). The goals in the Results Programme were selected such that there was an expectation that the level of effort of public servants would contribute to whether the targets would be achieved. For most of the goals, the achievement of the goal was in question, possible but not inevitable, for the duration of the Results Programme.

Likelihood of programme completion

Experimental studies on goal commitment assume the completion of the activity (Klein and Wright, 1994). In the public sector, this cannot be assumed. Complex social problems may require sustained attention over multiple years. Programmes that span many years run the risk that Government priorities will change and goals set now may not be relevant in five years time. In the Results Programme, the Government effectively tied its own hands, by publicly announcing the targets and promising to report every six months. More generally, this can be considered as an example of engineered exit costs (Cels et al., 2012).

Distribution of responsibility

While the goals frequently examined in goal commitment literature are individual goals (Locke and Latham, 2002), public administration goals often involve the problem of many hands (Thompson, 1980). Goal commitment diminishes as the number of responsible parties increases (Ingham et al., 1974), and New Public Management typically emphasises single-point accountability, where blame inevitably falls to one person and diminishes goal commitment of others. In the Results Programme, small groups of public managers’ were held collectively responsible for the success of the collaboration. The groups needed to be small, so that each member retained a high-level of responsibility. But they also needed to be designed to promote equal shared responsibility so that each member had an incentive to remain committed to the target.

Intrinsic motivation

Instrumentality and expectancy (both described above) are two of the three components of mission valence, along with ‘meaningfulness’ (Wright and Pandey, 2011). Rainey and Steinbauer (1999) describe mission valence as a public servant’s perception of the attractiveness of a mission. In this context, meaningfulness refers to the perception of the goal as being valuable. The targets in the Results Programme were meaningful, in that they related to problems that were important to New Zealanders, were priorities of government, and were framed at an intermediate-outcome level such that they could be seen as intrinsically valuable (in contrast to output or activity measures). Target regimes elsewhere have been criticised for decreasing intrinsic motivation (Deci, 1971; Eisenberger et al., 1999; Fisher, 1978), though this did not appear to be the case in the Results Programme. New Zealand public servants report that the targets have had a ‘galvanising’ effect on public servants and served as a ‘rallying call’ (both excerpts from interviews conducted as part of Scott and Boyd, 2017). Orange (2016: 17) reported ‘the Results spoke to the key drivers of why many of the people working on the Results had become public servants in the first place.’ The motivation of public servants differs from that in the private sector (Houston, 2000), with a larger role for public service motivation (Perry, 1996). Thus the attractiveness of goal attainment must be viewed both in the context of the extrinsic motivations traditionally assumed in goal commitment literature, and also intrinsic motivation of public servants to meaningfully serve the community.

Extrinsic motivation

Extrinsic motivation, too, may work differently in the public sector, with a typically lesser role for performance-related pay (Scott, 2016). Instead, public reputation tends to be a very strong motivator for public servants (Bevan and Fasolo, 2013). The Results Programme featured easy to understand public reporting that provided reputational rewards and sanctions for public servants. Prospect theory suggests that a fear of reputational loss is more powerful than the potential for reputational gain (Tversky and Kahneman, 1991). However, a system based purely on fear of loss of reputation may not be sustainable (LeGrand, 2007). The Results Programme featured a combination of quantitative and qualitative reporting, including ‘success stories’ that may have contributed to ongoing enthusiasm and self-esteem (two determinants of goal commitment in the private sector; Klein and Wright, 1994). Additionally, the small number of targets (10) increased the relative importance of success of failure in each target.

Implications for practice

While other studies describe ideal conditions for collaboration (see ‘collaborative opportunity’, Bryson et al., 2015), this article describes how collaboration was sustained and ultimately successful without these ideal conditions. Interagency collaboration is frequently imposed upon public servants at times and circumstances not of their choosing: political announcements, crises, or increased media attention might all put pressure on public servants to act to solve a problem that cannot be solved by agencies acting separately. In these cases, governments lack the luxury of waiting for collaborative opportunities to arise. This article points to how public servants may design solutions for solving those persistent and ‘wicked’ problems (Head and Alford, 2015) that cross agency boundaries and don’t have easy answers.

This article contributes to public administration practice through implications of three categorical types (see Table 6). The first and most notable implication for practice is to consider this case study as a proof of concept – evidence that sustained commitment and determination to achieving a shared goal is possible, even when the ‘problem of many hands’ (Thompson, 1980) suggests otherwise. The second category of implication is to consider the features and artefacts of the case study – the 10 findings identified by more than one evaluation method – may be treated as exemplars and potential candidates for policy transfer (Dolowitz and Marsh, 2000). The third category of implication is to consider the framework developed in this article to be an initial and tentative theory, with principles for designing context-specific programmes to build and sustain commitment to shared interagency goals. In this last regard, a single case study is not sufficient for confirming a generalisable theory, and would benefit from additional study as addressed below.

Table 6.

Implications for practice, in three types.

Implication type	Contribution	Example practice implication
(1) Case study as proof of concept	Goal commitment can help interagency collaboration succeed even when transaction costs are high	Programmes based on goal commitment should be considered for solving interagency problems
(2) Case study as exemplar design	The Results Programme had 10 features that multiple evaluations identified as contributing to its success	Goal commitment can be supported by public reporting that increases exit costs
(3) Case study as inspiration for generalisable theory	A theoretical framework describes six general antecedents to interagency goal commitment.	Goal commitment is generally supported by increasing the expectation of programme continuation and completion.

Implications for research and theory

This study identifies 10 features that appear to have contributed to the success of a large government programme in New Zealand. New Zealand has a strong focus on accountability (Gregory, 2006; Scott and Macaulay, forthcoming) and a history of abandoned reforms and cancelled programmes (Scott and Boyd, 2016) that shape the expectations of public servants, and likely influence how they react to the various design aspects of the Results Programme. Hence the importance and relevance of each these features is likely to be context dependent, or even specific to the case described. However, each feature can be explained as contributing to goal commitment, consistent with the experiences of public servants involved in the Results Programme, who report that they remained committed to achieving the goals despite high transaction costs. The expectancy theory of goal commitment (Klein and Wright, 1994) and various public administration literature on public service motivation (Perry, 1996) and mission valence (Wright and Pandey, 2011) provide relevant associations and constructs that have been used to develop a more theoretically grounded framework.

This framework requires further testing in different contexts, and such research could take the form of new case studies, comparative research, or secondary research on past data. As noted in the methodology section, this article relies on secondary data from earlier evaluations. The relationship between the case study and the concept of goal commitment was generated inductively, and the original evaluations were not conducted from the perspective of exploring or testing goal commitment or its antecedents. This suggests an opportunity to conduct similar re-examinations of other case data through the lens of goal commitment. For example, the various prioritisation schemes (for example, Virginia Performs, GMAP Washington, Scotland Performs, Victoria Ahead; Kohli, 2010), or target/goal schemes (for example, the US federal Cross Agency Priority Goals; Putansu, 2015) may provide additional data on which to understand the goal commitment construct. In each case, the goal would be to explore whether the framework has descriptive, explanatory, or predictive power, and whether it may be refined further.

One limitation in drawing conclusions across the collaboration literature lies in imprecision of the language used, as discussed earlier. While we characterise the case study as collaboration, in the sense of shared responsibility for a common goal, we acknowledge that some of the actions in service of those goals would elsewhere be described as cooperation or coordination, in that some solutions involve actions that were jointly conceived but separately delivered. The lack of precision (across multiple dimensions) of these terms makes it difficult to conduct like-comparisons in the form of meta-analyses or evidence syntheses. The field would benefit from more consistent use of more precise language that can distinguish between different types of action and different types of responsibilities.

This article broadly aims to contribute to our understanding of interagency collaboration. So many pages of academic journals have been filled with discussions of collaboration that we can no longer refer to it as a ‘black box’ (Thomson and Perry, 2006), and yet goal commitment remains an unexplored aspect. Ideally, research and theory would inform practice that worked on both ‘ends’ of the collaboration problem: how to improve goal commitment and how to reduce barriers. Public servants would be more committed to succeeding, and the barriers they faced would be lower, though some contexts would inevitable warrant more focus on one or the other. Refining our understanding of the antecedents and consequences of goal commitment in an interagency collaboration context would have numerous benefits, from better predicting, understanding and designing for collaborative success, to discriminatorily matching programme designs to different contexts.

In tackling this underexplored aspect of interagency collaboration, this article aims to ‘zig’ where other articles ‘zag’; while other authors explore how to lower the barriers to collaboration and make it easier, goal commitment is proposed here as an explanation for how effort can be created and sustained where interagency collaboration is hard.

Footnotes

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.

ORCID iD

Rodney James Scott

References

Altshuler

(2003) From barriers to successful collaboration: Public schools and child welfare working together. Social Work 48(1): 52–63.

Ansell

Gash

(2008) Collaborative governance in theory and practice. Journal of Public Administration Research and Theory 18(4): 543–571.

Aube

Rousseau

(2005) Team goal commitment and team effectiveness: The role of task interdependence and supportive behaviors. Group Dynamics: Theory, Research, and Practice 9(3): 189.

Axelsson

(2006) Integration and collaboration in public health – A conceptual framework. The International Journal of Health Planning and Management 21(1): 75–88.

Baehler

(2003) Managing for outcomes: Accountability and trust. Australian Journal of Public Administration 62(4): 23–34.

Barber

Moffit

Kihn

(2010) Deliverology 101: A Field Guide for Educational Leaders. Thousand Oaks, CA: Corwin Press.

Bardach

(1998) Getting Agencies to Work Together: The Practice and Theory of Managerial Craftsmanship. Washington, DC: Brookings Institution Press.

Bevan

Fasolo

(2013) Models of governance of public services: Empirical and behavioural analysis of ‘econs’ and ‘humans’. In: Oliver

(ed.) Behavioural Public Policy. Cambridge: Cambridge University Press, pp. 38–62.

Bevan

Hood

(2006) What’s measured is what matters: Targets and gaming in the English public health care system. Public Administration 84(3): 517–538.

10.

Boston

Martin

Pallot

, et al. (1996) Public Management: The New Zealand Model. Auckland: Oxford University Press.

11.

Boyne

Gould-Williams

(2003) Planning and performance in public organizations: An empirical analysis. Public Management Review 5: 115–132.

12.

Bresnen

Marshall

(2000) Motivation, commitment and the use of incentives in partnerships and alliances. Construction Management and Economics 18(5): 587–598.

13.

Bronkhorst

Steijn

Vermeeren

(2015) Transformational leadership, goal setting, and work motivation: The case of a Dutch municipality. Review of Public Personnel Administration 35(2): 124–145.

14.

Bryson

Crosby

Stone

(2015) Designing and implementing cross‐sector collaborations: Needed and challenging. Public Administration Review 75(5): 647–663.

15.

Burkhardt

(2001) Thoughts on motivational problems in networks. In: Mandell

(ed.) Getting Results through Collaboration: Networks and Network Structures for Public Policy and Management. Westport, CT: ABC-CLIO, pp. 239–241.

16.

Campbell

Fiske

(1959) Convergent and discriminant validation by the multitrait-multimethod matrix. Psychological Bulletin 56: 81–105.

17.

Carey

Crammond

(2015) What works in joined-up government? An evidence synthesis. International Journal of Public Administration 38(13–14): 1020–1029.

18.

Carey

Harris

(2016) Developing management practices to support joined-up governance. Australian Journal of Public Administration 75(1): 112–118

19.

Cels

de Jong

Nauta

(2012) Agents of Change: Strategy and Tactics for Social Innovation. Washington, DC: Brookings Institution Press.

20.

Chun

Rainey

(2005) Goal ambiguity and organizational performance in US federal agencies. Journal of Public Administration Research and Theory 15(4): 529–557.

21.

Creswell

(2013) Research Design: Qualitative, Quantitative, and Mixed Methods Approaches. New York, NY: Sage,

22.

Deci

(1971) Effects of externally mediated rewards on intrinsic motivation. Journal of Personality and Social Psychology 18(1): 105.

23.

Doberstein

(2016) Designing collaborative governance decision-making in search of a ‘collaborative advantage’. Public Management Review 18(6): 819–841.

24.

Dolowitz

Marsh

(2000) Learning from abroad: The role of policy transfer in contemporary policy‐making. Governance 13(1): 5–23.

25.

Donaldson

(2001) The Contingency Theory of Organizations. Thousand Oaks, CA: Sage.

26.

Earley

Shalley

Northcraft

(1992) I think i can, i think i can … processing time and strategy effects of goal acceptance/rejection decisions. Organizational Behavior and Human Decision Processes 53(1): 1–13.

27.

Edmondson

Roloff

(2009) Overcoming barriers to collaboration: Psychological safety and learning in diverse teams. Team Effectiveness in Complex Organizations: Cross-disciplinary Perspectives and Approaches 34: 183–208.

28.

Eisenberger

Rhoades

Cameron

(1999) Does pay for performance increase or decrease perceived self-determination and intrinsic motivation? Journal of Personality and Social Psychology 77(5): 1026.

29.

Emerson

Nabatchi

Balogh

(2012) An integrative framework for collaborative governance. Journal of Public Administration Research and Theory 22(1): 1–29.

30.

Eppel

Gill

Lips

, et al. (2008) Better connected services for kiwis: A discussion document for managers and front-line staff on better joining up the horizontal and the vertical. Institute for Government and Policy Studies, New Zealand.

31.

Erez

Judge

(2001) Relationship of core self-evaluations on goal setting, motivation, and performance. Journal of Applied Psychology 86(6): 1270–1279.

32.

Esteve

van Witteloostuijn

Boyne

(2015) The effects of public service motivation on collaborative behavior: Evidence from three experimental games. International Public Management Journal 18(2): 171–189.

33.

Favero

Bullock

(2015) How (not) to solve the problem: An evaluation of scholarly responses to common source bias. Journal of Public Administration Research and Theory 25(1): 285–308.

34.

Fereday

Muir-Cochrane

(2006) Demonstrating rigor using thematic analysis: A hybrid approach of inductive and deductive coding and theme development. International Journal of Qualitative Methods 5(1): 80–92.

35.

Fisher

(1978) The effects of personal control, competence, and extrinsic reward systems on intrinsic motivation. Organizational Behavior and Human Performance 21(3): 273–288.

36.

Gadja

(2004) Utilizing collaboration theory to evaluate strategic alliances. American Journal of Evaluation 25(1): 65–77.

37.

Gazley

(2014) Intersectoral collaboration and the motivation to collaborate: Toward an integrated theory. In: Big Ideas in Collaborative Public Management. London: Routledge, pp. 46–64.

38.

George

Bennett

(2005) Case Studies and Theory Development in the Social Sciences. Cambridge: MIT Press.

39.

Glendinning

(2003) Breaking down barriers: Integrating health and care services for older people in England. Health Policy 65: 139–151.

40.

Gorringe

(1994) Commitment perspectives on public sector governance. New Zealand Government, Wellington.

41.

Gray

(1989) Collaborating: Finding Common Ground for Multiparty Problems. San Francisco, CA: Jossey-Bass.

42.

Greene

(2007) Mixed methods in social inquiry. San Francisco, CA: Jossey-Bass.

43.

Gregory

(2006) Theoretical faith and practical works: De-autonomising and joining up in the New Zealand state sector. In: Christensen

Lægreid

(eds) Autonomy and Regulation – Coping with Agencies in the Modern State. Cheltenham: Edward Elgar.

44.

Gregson

Cartlidge

Bond

(1992) Development of a measure of professional collaboration in primary health care. Journal of Epidemiology and Community Health 46: 48–53.

45.

Head

Alford

(2015) Wicked problems: Implications for public policy and management. Administration & Society 47(6): 711–739.

46.

Heckman

(1990) Varieties of selection bias. The American Economic Review 80(2): 313–318.

47.

Hoegl

Parboteeah

(2006) Team goal commitment in innovative projects. International Journal of Innovation Management 10(03): 299–324.

48.

Hollenbeck

Klein

(1987) Goal commitment and the goal setting process: Problems, prospects and proposals for future research. Journal of Applied Psychology 72: 212–220.

49.

Houston

(2000) Public-service motivation: A multivariate test. Journal of Public Administration Research and Theory 10(4): 713–728.

50.

Howarth

Morrison

(2007) Collaboration, integration and change in children's services: Critical issues and key ingredients. Child Abuse and Neglect 31(1): 55–69.

51.

Hudson

Hardy

Henwood

, et al. (1999) In pursuit of inter-agency collaboration in the public sector: What is the contribution of theory and research? Public Management an International Journal of Research and Theory 1(2): 235–260.

52.

Huxham

Macdonald

(1992) Introducing collaborative advantage: Achieving interorganizational effectiveness through meta-strategy. Management Decision 30(3): 50–56.

53.

Huxham

Vangen

(2000) Leadership in the shaping and implementation of collaboration agendas: How things happen in a (not quite) joined-up world. Academy of Management Journal 43(6): 1159–1175.

54.

Huxham C and Vangen S (2013) Managing to collaborate: The theory and practice of collaborative advantage. New York: Routledge.

55.

Imperial

(2005) Using collaboration as a governance strategy: Lessons from six watershed management programs. Administration & Society 37(3): 281–320.

56.

Ingham

Levinger

Graves

, et al. (1974) The Ringelmann effect: Studies of group size and group performance. Journal of Experimental Social Psychology 10(4): 371–384.

57.

Jennings

Krane

(1994) Coordination and welfare reform: The quest for the philosopher’s stone. Public Administration Review 54: 341–348.

58.

Jensen

Scott

Slocombe

, et al. (2014) The management and organisational challenges of more joined-up government: New Zealand’s Better Public Services reforms. State Sector Performance Hub, Working paper 2014-1. DOI: 10.13140/RG.2.1.3115.3680

59.

Jensen

(1983) Organisational theory and methodology. The Accounting Review 58(2): 319–339.

60.

Jung

Ritz

(2014) Goal management, management reform, and affective organizational commitment in the public sector. International Public Management Journal 17(4): 463–492.

61.

Kania

Kramer

(2011) Collective impact. Stanford Social Innovation Review 9(1): 36–41.

62.

Keast

Brown

Mandell

(2007) Getting the right mix: Unpacking integration meanings and strategies. International Public Management Journal 10(1): 9–33.

63.

Klein

Cooper

Monahan

(2013) Goal commitment. In: Locke

Latham

(eds) New Developments in Goal Setting and Performance. New York, NY: Routledge, pp. 65–89.

64.

Klein

Wesson

Hollenbeck

, et al. (1999) Goal commitment and the goal-setting process: Conceptual clarification and empirical synthesis, Journal of Applied Psychology 84(6): 885–896.

65.

Klein

Wright

(1994) Antecedents of goal commitment: An empirical examination of personal and situational factors. Journal of Applied Social Psychology 24(2): 95–114.

66.

Kohli

(2010) Golden Goals for Government Performance: Five Case Studies on How to Establish Goals to Achieve Results. Washington, DC: Centre for American Progress.

67.

Krueger

McGuire

(2005) A Transaction Costs Explanation of Interlocal Government Collaboration. Working Group on Interlocal Services Cooperation. Paper 21.

68.

Laan

Voordijk

Dewulf

(2011) Reducing opportunistic behaviour through a project alliance. International Journal of Managing Projects in Business 4(4): 660–679.

69.

Latham

Borgogni

Petitta

(2008) Goal setting and performance management in the public sector. International Public Management Journal 11(4): 385–403.

70.

LeGrand

(2007) Delivering public services through choice and competition: The other invisible hand. Princeton: Princeton University Press.

71.

Ling

(2002) Delivering joined–up government in the UK: Dimensions, issues and problems. Public Administration 80(4): 615–642.

72.

Lips

AMB

O’Neill

Eppel

(2011) Cross-agency collaboration in New Zealand: An empirical study of information sharing practices, enablers and barriers in managing for shared social outcomes. International Journal of Public Administration 34: 255–266.

73.

Locke

Latham

(1990) A Theory of Goal Setting and Task Performance. Upper Saddle River, NJ: Prentice Hall.

74.

Locke

Latham

(2002) Building a practically useful theory of goal setting and task motivation: A 35-year odyssey. American Psychologist 57(9): 705.

75.

McConnell

(2010) Policy success, policy failure and grey areas in-between. Journal of Public Policy 30(3): 345–362.

76.

Manning

Cullum-Swan

(1994) Narrative, content, and semiotic analysis. In: Manning

Cullum-Swan

(eds) Handbook of Qualitative Research. Thousand Oaks, CA: Sage, pp. 463–477.

77.

Marrett

(1971) On the specification of interorganizational dimensions. Sociology and Social Research 56: 83–89.

78.

Marsh

McConnell

(2010) Towards a framework for establishing policy success. Public Administration 88(2): 564–583.

79.

Minson

Mueller

(2012) The cost of collaboration why joint decision making exacerbates rejection of outside information. Psychological Science 23(3): 219–224.

80.

Mulroy

Shay

(1998) Motivation and reward in nonprofit interorganizational collaboration in low-income neighborhoods. Administration in Social Work 22(4): 1–17.

81.

New Zealand Herald (2018) PM Jacinda Ardern: Merit in target system, but Labour will set its own, 23 January. Available at: https://www.nzherald.co.nz/nz/news/article.cfm?c_id=1&objectid=11980625&ref=rss (accessed 10 October 2019).

82.

Norman

Gill

(2011) Restructuring – An over-used lever for change in New Zealand’s state sector? Working paper 2:6, Institute for Government and Policy Studies, New Zealand.

83.

Nutt PC (1983) Implementation approaches for project planning. Academy of Management Review 8(4): 600–611.

84.

O’Leary

Vij

(2012) Collaborative public management: Where have we been and where are we going? The American Review of Public Administration 42(5): 507–522.

85.

Orange

(2016) New public passion: Reflections from New Zealand on public service reform. Discussion Paper 08(01). UNDP Global Centre for Public Service Excellence, Singapore.

86.

Page

(2003) Entrepreneurial strategies for managing interagency collaboration. Journal of Public Administration Research and Theory 13(3): 311–339.

87.

Patton

(2011) Developmental Evaluation: Applying Complexity Concepts to enhance Innovation and Use. New York, NY: Guilford Press.

88.

Pedersen

(2016) A ‘Heart of Goal’ and the will to succeed: goal commitment and task performance among teachers in public schools. Public Administration 94(1): 75–88.

89.

Perry

(1996) Measuring public service motivation: An assessment of construct reliability and validity. Journal of Public Administration Research and Theory 6(1): 5–22.

90.

Peters

(1998) Managing horizontal government: The politics of co-ordination. Public Administration 76(2): 295–311.

91.

Pollitt

Bouckaert

(2011) Public Management Reform: A Comparative Analysis: New Public Management, Governance and the Neo-weberian State. Oxford: Oxford University Press.

92.

Prebble

(2015) Public value and limits to collaboration. International Journal of Public Administration 38(7): 473–485.

93.

Preskill

Parkhurst

Juster

(2014) Guide to evaluating collective impact. Collective Impact Forum. Available at: http://www.fsg.org/publications/guide-evaluating-collective-impact (accessed 2 February 2020).

94.

Provan

Kenis

(2008) Modes of network governance: Structure, management, and effectiveness. Journal of Public Administration Research and Theory 18(2): 229–252.

95.

Provan

Milward

(2001) Do networks really work? A framework for evaluating public‐sector organizational networks. Public Administration Review 61(4): 414–423.

96.

Putansu

(2015) Cross agency priority goals in the US government: Can directed collaboration be a stepping stone toward politic-centered performance? Policy and Society 34(1): 25–35.

97.

Rainey

Steinbauer

(1999) Galloping elephants: Developing elements of a theory of effective government organizations. Journal of Public Administration Research and Theory 9(1): 1–32.

98.

Ren

Beard

Atkins

(2005) A survey of consensus problems in multi-agent coordination. In: Proceedings of the 2005 American control conference, IEEE, 2005, Portland Oregon, 8 June 2005, pp. 1859–1864.

99.

Ring

Van de Ven

(1994) Development processes of cooperative interorganizational relationships. Academy of Management Review 19(1): 90–118.

100.

Rossman

Wilson

(1985) Numbers and words: Combining quantitative and qualitative methods in a single large-scale evaluation study. Evaluation Review 9: 627–643.

101.

Rutherford

Meier

(2015) Managerial goals in a performance-driven system: Theory and empirical tests in higher education. Public Administration 93(1): 17–33.

102.

Sadoff

Grey

(2005) Cooperation on international rivers: A continuum for securing and sharing benefits. Water International 30(4): 420–427.

103.

Schick

(1996) The spirit of reform: Managing the New Zealand state sector in a time of change. State Services Commission of New Zealand, New Zealand.

104.

Schweitzer

Ordoñez

Douma

(2004) Goal setting as a motivator of unethical behavior. Academy of Management Journal 47(3): 422–432.

105.

Scott

(2001) Public Sector Management in New Zealand: Lessons and Challenges. Wellington: Australian National University.

106.

Scott

(2014) Group model building and mental model change. Doctoral Thesis, University of Queensland, Australia.

107.

Scott

(2016) The performance relationship between department chief executives and the State Services Commission in New Zealand. State Sector Performance Hub, working paper 2016-3. DOI: 10.13140/RG.2.1.4299.4162/1

108.

Scott

(2018) Group Model Building: Using Systems Dynamics to Achieve Enduring Agreement. Melbourne: Springer.

109.

Scott RJ and Macaulay M (forthcoming)Making sense of New Zealand’s ‘spirit of service’: social identity and the civil service. Public Money and Management.

110.

Scott

Bardach

(2019) A comparison of management adaptations for joined‐up government: Lessons from New Zealand. Australian Journal of Public Administration 78(2): 191–212.

111.

Scott RJ and Boyd R (2016) Joined-up for what? Response to Carey and Harris on joined-up governance practices. The Australian Journal of Public Administration 76(1): 138–144.

112.

Scott

Boyd

(2016a) Collective impact in the Public Sector: the New Zealand Results approach. State Sector Performance Hub. Working paper 2016-1. DOI: 10.13140/RG.2.1.2839.5929

113.

Scott

Boyd

(2016b) Case studies in collaborating for better public services. State Sector Performance Hub. Working paper 2016-2. DOI: 10.13140/RG.2.1.3232.8081

114.

Scott

Boyd

(2016c) Results, targets and measures to drive collaboration: Lessons from the New Zealand Better Public Services reforms. In: Butcher

Gilchrist

(eds) The Three Sector Solution: Delivering Public Policy in Collaboration with Not-for-profits and Business. Canberra: Australian National University Press, pp. 235–257.

115.

Scott

Boyd

(2017) Interagency Performance Targets: A Case Study of New Zealand’s Results Programme. Washington, DC: IBM.

116.

Scott

Cavana

Cameron

(2015) Interpersonal success factors for strategy implementation: A case study using group model building. Journal of the Operational Research Society 66(6): 1023–1034.

117.

Scott

(2015) Does collaboration make any difference? Linking collaborative governance to environmental outcomes. Journal of Policy Analysis and Management 34(3): 537–566.

118.

Seijts

Latham

(2000) The construct of goal commitment: Measurement and relationships with task performance. In: Goffin

Helmes

(eds) Problems and Solutions in Human Assessment. Norwell, MA: Kluwer Academic Publishers, pp. 315–332.

119.

Shelton

(2013) Al Morrison Leaving Conservation Dept, joining State Services Commission. Wellington: Wellington Scoop. Available at: http://wellington.scoop.co.nz/?p=54803 (accessed 2 February 2020).

120.

Simons

(2013) Levers of Organization Design: How Managers Use Accountability Systems for Greater Performance and Commitment. Boston, MA: Harvard Business Press.

121.

Smith

(2009) Performance Measurement for Health System Improvement: Experiences, Challenges and Prospects. Cambridge: Cambridge University Press.

122.

State Services Commission (2016) Better Public Services: Case studies. Available at: https://www.ssc.govt.nz/bps-case-studies (accessed 2 February 2020).

123.

State Services Commission (2014) Cabinet Paper: Better Public Services: Refresh of Targets and Measures. New Zealand Government, New Zealand. CAB (14): 588.

124.

State Services Commission (2017) Cabinet paper: Better public services results refresh. New Zealand Government, New Zealand. SEC (17) 093.

125.

Teddlie

Tashakkori

(2009) Foundations of Mixed Methods Research: Integrating Quantitative and Qualitative Approaches in the Social and Behavioral Sciences. Thousand Oaks, CA: Sage.

126.

Thompson

(1980) Moral responsibility of public officials: The problem of many hands. American Political Science Review 74(4): 905–916.

127.

Thomson

Perry

(2006) Collaboration processes: Inside the black box. Public Administration Review 66(1): 20–32.

128.

Tversky

Kahneman

(1991) Loss aversion in riskless choice: A reference-dependent model. The Quarterly Journal of Economics 106(4): 1039–1061.

129.

Van Dooren

(2006) Performance measurement in the Flemish public sector: A supply and demand approach. Unpublished Doctoral Thesis, KUL, Belgium.

130.

Van Huyck

Battalio

Beil

(1990) Tacit coordination games, strategic uncertainty and coordination failure. The American Economic Review 80(1): 234–248.

131.

Vroom

(1964) Work and Motivation. New York, NY: Wiley.

132.

Warner

Hefetz

(2008) Managing markets for public service: The role of mixed public–private delivery of city services. Public Administration Review 68(1): 155–166.

133.

Weber

(2000) Interdependence in laboratory groups and organizations. Doctoral Thesis, California Institute of Technology, USA.

134.

Weiss

Piderit

(1999) The value of mission statements in public agencies. Journal of Public Administration Research and Theory 9: 193–226.

135.

Wilson

(1989) Bureaucracy: What Government Agencies Do and Why They Do It. New York, NY: Basic Books.

136.

Wright

(2007) Public service and motivation: Does mission matter? Public Administration Review 67(1): 54–64.

137.

Wright

Pandey

(2011) Public organizations and mission valence: When does mission matter? Administration & Society 43(1): 22–44.

138.

Yin

(1981) The case study crisis: Some answers. Administrative Science Quarterly 26(1): 58–65.