The organisation of interagency training to safeguard children in England: a case study using realistic evaluation

Background Joint training for interagency working is carried out by Local Safeguarding Children Boards in England to promote effective local working to safeguard and promote the welfare of children. Purpose This paper reports on the findings of the outputs and outcomes of interagency training to safeguard children in eight Local Safeguarding Children Boards. Methods A review of Local Safeguarding Children Board documentation, observations of Local Safeguarding Children Board training sub-group meetings and a series of interviews with training key stakeholders in each Local Safeguarding Children Board were used to assess how partner agencies in the Local Safeguarding Children Boards carried out their statutory responsibilities to organise interagency training. ‘Realistic Evaluation’ was used to evaluate the mechanisms by which a central government mandate produced particular inter-agency training outputs (number of courses, training days) and joint working outcomes (effective partnerships), within particular Local Safeguarding Children Board contexts. Results The ‘mandated partnership’ imposed on Local Safeguarding Children Boards by central government left little choice but for partner agencies to work together to deliver joint training, which in turn affected the dynamics of working partnerships across the various sites. The effectiveness of the training sub group determined the success of the organisation and delivery of training for joint working. Despite having a central mandate, Local Safeguarding Children Boards had heterogeneous funding and training arrangements. These resulted in significant variations in the outputs in terms of the number of courses per ‘children in need’ in the locality and in the cost per course. Conclusions Interagency training which takes account of the context of the Local Safeguarding Children Board is more likely to produce better trained staff, effective partnership working, and lead to better integrated safeguarding children services.


Introduction
Official inquiries into the protection of children from abuse and neglect in England and government policy guidance have consistently advocated that if the professionals concerned with safeguarding children are to work together more effectively they should learn together to work together [1,2]. According to an influential government-commissioned review of the state of child protection by Lord Laming, "…multi-agency training is important in helping professionals understand the respective roles and responsibilities and the procedures of each agency involved in child protection, in developing a joint understanding of assessment and decision-making practices." [3].
Multi-agency or interagency training had been provided under the auspices of Area Child Protection Committees since their establishment following the 1974 inquiry into the death of a child, Maria Colwell, which highlighted a serious lack of coordination among services and professionals. In turn, Local Safeguarding Children Boards were established as part of the government's response to the statutory inquiry into the death of another child, Victoria Climbié [4]. The new concept of 'safeguarding' embraced the narrower concept of 'child protection' as one of its elements but refocused interagency collaboration onto a broader remit of prevention of harm and promotion of well-being [5]. In the 2006 version of Working Together, the government stated: "It is the responsibility of the Local Safeguarding Children Board to ensure that single-agency and interagency training on safeguarding and promoting welfare is provided in order to meet local needs. This covers both the training provided by single agencies to their own staff, and interagency training where staff from more than one agency train together" [2, p. 79].
However, other authors commenting on the expansion of interagency training in this sphere observed that, "The increased investment in training has occurred in the absence of a firmly established evidence base to support the notion that interagency training changes or enhances the behaviour of professionals working together" [6, p. 364]. Their review stressed the size and complexity of the task with which Local Safeguarding Children Boards had been charged. They concluded that "…interagency training is mainly an act of faith albeit one encompassing elements of reason' on the grounds that there was a developing consensus about the requirements for effective inter-professional relationships" [6, p. 372].
In fact there had previously been some research on interagency partnerships for child protection in England [7], which demonstrated the importance of a contextual understanding in the evaluation of multi-agency working over three Area Child Protection Committees, using a combination of the Nuffield Institute's Partnership Assessment Tool [8] with personal interviews. This study concluded that: "…levels of awareness of and commitment to multiagency working are variable within and between [partner agencies], and contingent upon a range of organizational contexts. As such, it should be unsurprising that performance varies, even where there are apparently similar 'technical fix' approaches to joint working" [8, p. 191].
Past research [7] had employed the Realistic Evaluation paradigm [9] to investigate the mechanisms and processes by which training outcomes are achieved. Understanding the organisation of interagency training for joint working is important as what works to produce an effect in one circumstance will not produce it in another [9,10].

Aims
The overall aim of the research of which this report is a part was to develop an evidence base for interagency training to safeguard children. It was jointly commissioned by the (then) Department for Children Schools and Families and the Department of Health in England as part of a research programme on safeguarding children. In other papers we shall present a series of the evaluations of the learning outcomes of selected interagency training courses. Here we report findings of the first aim of the study, which was to investigate how the agency partners in Local Safeguarding Children Boards carried out their statutory responsibilities to organise interagency training. Couched in terms of Realistic Evaluation, this component of the study examined the mechanisms or processes by which a policy intervention (statutory guidance) produced particular outputs (interagency training for joint working) and outcomes (effective partnerships), within particular organisational contexts (local conditions in the Local Safeguarding Children Boards).

Methods
The study adopted a naturalistic or observational approach. It was designed to investigate how things are, rather than to set up an interagency partnerships to deliver training programme and then to assess the outcomes. The latter approach would not have enabled us to answer our research questions about context, mechanisms and outcomes. Consequently we began by recruiting a group of Local Safeguarding Children Boards willing to participate in the study; crucially, they were not simply the objects of study, but partners in the evaluation.

Analysis of documents
A range of relevant documents was reviewed. These included: annual reports and business plans; training strategies; training programmes; and minutes of the training sub-group (or equivalent) meetings. This documentary evidence provided information about the formal goals and aims of the training programmes, the financial contributions made by partner agencies and the number of courses provided.

Observation
Non-participant observation of meetings of the training sub-groups was used to collect information about the processes through which training needs were discussed and decisions were made about the content of the programme. The project researcher was able to attend meetings in six of the eight sites. Prior to the meetings, he reviewed minutes of the previous two meetings to acquaint himself with the recent issues. Permission was sought and received to take notes of the meeting, including personal observations of group dynamics.

Interviews with stakeholders
Personal and telephone interviews were held with key stakeholders: the agency managers/representatives on the training sub-groups and the training coordinators, who are appointed to organise and deliver train-ing. The aim was to interview at least one member from each partner agency. The topics are summarised in Table 1.
Forty-five agency representatives were approached for an interview and 39 agreed. The remainder were unavailable because of leave/sickness or pressure of work. All eight training co-ordinators were interviewed. Interview schedules were designed by both authors. The interviews and observations and initial data analyses were carried out by DP. Subsequent analyses using the theoretical frameworks were carried out by both authors.

Data collection
During the course of the personal and telephone interviews, answers to questions posed were written on pro forma devised for the study. Hand-written notes were then transcribed into electronic format and tabulated by site. Permission to use anonymised quotes from the interviews was granted by each of the interviewees upon completion of the interview.
A process of emerging themes [11] was used to categorise responses into general statements which reflected the context and process of interagency training and collaboration in each project site. Thus, if there was agreement on a particular topic area, e.g., that the training programme needed more funding, then the general finding would be "training programme requires more funding". If, on the other hand, there was disagreement by some partner agency representatives, this was recorded as "some felt that funding levels are adequate, while others felt that more funding is needed". The same process was applied for each topic/question area and series of interviews. This cumulative process of thematic analysis allowed us both to compare and contrast opinion on interagency training for joint working within sites and also between sites.

Data analyses
A deductive approach was employed for analysing the data. The thematic content analysis of interview data was guided by the Effective Partnership Working Inventory [12]. This inventory is based on substantial research and theory within the field of organisational psychology. It informed the 'mechanism' (or process) component within the Realistic Evaluation framework. The Effective Partnership Working Inventory aims to assess seven dimensions: shared commitment to goals and objectives; Interdependence of outcomes; Role clarity; Cultural congruity; Focus on quality and innovation; True cooperation; and, Interprofessional trust and respect [12, p. 81].

Outcomes (vs. outputs)
In applying the Realistic Evaluation framework to the study of interagency training, we differentiated outputs (the training courses provided) from the outcomes, which were defined as 'benefits for staff' in terms of learning outcomes and better working relationships between staff from partner agencies resulting from participation in interagency training. There might also be beneficial outcomes to the organisations involved, e.g., more effective and efficient use of scarce resources and, potentially, for children and families.
Determining the government's intended outcomes for interagency training was straightforward because these were handed down in statutory guidance through Working Together [2]. Interviews and training strategies revealed that a clear consensus among partners, i.e., creating a better-trained workforce to work together to safeguard children and to maintain their well-being, e.g.
"Helps develop our policies and fosters clear standards for training staff who work with children and families" (Agency manager, Site F).
Another manager believed that these outcomes were already being achieved, considering that there was, "Greater staff awareness about child protection issues than when I came into post a few years ago" (Agency manager, Site D).
However, keeping focused on these outcomes could be difficult in a busy policy context and a pressured practice environment, "It is an add-on to what we currently do, a hodgepodge right now" (Agency manager, Site D).
Past research [13] distinguished three different types of outcome that a partnership may be trying to achieve: synergy, transformation and budget enlargement. Inter-agency training has an element of 'synergy' as it brings together partners with different assets and powers to create something which is greater than the sum of its parts. There was also some sense that the work of the training sub-group leads to 'transformation' in that it brings partner agencies together to change the objectives and culture of the organisations, with the direction of change depending on the power of each individual partner. However, this was more often an unstated than stated outcome. Although enlargement of the training budget did result from resources contributed to core funding by partner agencies, unlike in many cases of interagency partnerships there was no additional government money to promote the activity.
There were varying degrees of 'resource' and 'policy' synergy [14]. Agency managers commented repeatedly on the 'added value' from the resources contributed to core funding; they believed that there has been increased effectiveness and to some extent efficiency.
In terms of policy synergy, it was suggested that there had been new perspectives and solutions developed in some of the sites with respect to the courses delivered and special awareness raising sessions, e.g. one of the sites held an away day for Local Safeguarding Children Board members to discuss emerging issues and training needs.
Past research [14], which built on the concept of 'transformation' [13], suggested that different outcomes can occur depending on the power balance between partners. The interviews revealed that there were indeed power imbalances in the training partnerships, particularly between health and social care and other partner agencies, but they do not appear to have either hampered or facilitated the work of the training sub-groups.
"There are good working relations on the training sub-group…different professionals with different backgrounds work well together" (Agency manager, Site B).
There was a clear sense that partner agencies had accepted the need for change and to learn fromand more importantly, about-each other, so 'mutual transformation', whereby all partners change and differences between them begin to reduce, had occurred in varying degrees across the sites.
"One of the successes of the interagency training is the breaking down of prejudices, misperceptions, misunderstanding between partner agencies" (Agency manager, Site A).

Context
When interpreting the findings of the interviews, it is important to bear in mind that safeguarding training had been mandated or 'forced'. It had started from dif-ferent positions in each site and its success was likely to depend on a number of factors, including past working relationships. Relationships in some sites were thought to be very good, "Members of the training sub group work well together… we have a common goal we are working towards" (Agency manager, Site F).
whereas in others they were not so conducive to effective partnership working, e.g.
"I do not feel that there is an equal commitment shared by all members of partner agencies…some people are more active than others…there is a difference between attending and doing" (Agency manager, Site E).

Mechanism/Process
In seven of the eight sites, a training sub group had been established. Here we used the Effective Partnership Working Inventory [12] to assess the seven domains in which these groups were operating effectively as partnerships.
(1) Shared commitment to goals and objectives All members were clear about their own agency's goals with respect to interagency training, i.e., fulfilling the mandate of Working Together [2], i.e.
"…that every partner agency is fully aware of child protection issues and that those in key roles have specialised knowledge" (Agency manager, Site E). and they were also clear about the partnership's goals, i.e.
"ensuring that all staff in partner agencies are fully up to date and ready to carry out their roles in the community" (Agency manager, Site F).
There was widespread belief that the goals of the Local Safeguarding Children Board generally and the interagency training specifically were valuable: it may have been mandated, but members believed in its importance and strategies had been agreed. Consequently there was a, "Lack of conflict between partner agencies…largely due to the fact that the training strategy is framed so carefully" (Training coordinator, Site A).
(2) Interdependence of outcomes All partnership members believed their own agencies' and the partnership's goals with respect to interagency training are interdependent and mutually beneficial, e.g., one manager stated, "Training silos [i.e., independent programmes] are not an effective use of existing resources and do not meet basic tenets of how interagency training should be done" (Agency manager, Site B).
At the simplest level, interagency training requires the participation of staff from different agencies which means that the partners depend on each other to, in the vernacular, 'deliver the bums on seats'. This was not always straightforward and the recruitment of a balanced set of participants was a frequent matter for discussion at the meetings. Recruiting staff from their own agency was one of the tasks specified for the representatives. But, in addition, there was a consensus that the skills and experience brought to, and shared with the partnership were essential to mounting a training programme. Thus many specialist courses, e.g., on adult mental illness and safeguarding, relied on the teaching contributions of health as well as social services staff from the partner agencies; a course presented by an adult psychiatrist is unlikely to give adequate coverage of the issues faced by children's social workers. Further the modelling by the trainers of interagency collaboration by representatives of health and social care was considered an important ingredient in the programme.
(3) Role clarity There were differences in the extent to which the agency managers on the training sub-group understood their own and each others' roles. Many interviewees considered that there was no clear indication of their specific role, "we need to be clearer about what we want partner agencies to do and why" (Agency manager, Site H).
Only two Local Safeguarding Children Boards had thus far developed terms of reference for the training sub-group. However, there was evidence from observations and interviews that all the training subgroups could work constructively to resolve conflicts which arose about status or role. There were evidently some inter-professional conflicts, but these were minimal; respondents were likely to comment that these had more to do with the person than with the partner agency. In general role clarity and role differentiation emerged from working together rather than beginning with a list of roles and responsibilities.

(4) Cultural congruity
Cultural congruity in a partnership can only be achieved if members have first understood the differences between the participants' 'home' organisational cultures [12]. Such understanding was observed in the training group discussion when participants showed tolerance and appreciation of the challenges faced by colleagues in other agencies for example, in recruiting participants and designing courses to take into account their backgrounds and different learning styles. Further, it was apparent that the demands of interagency Observation suggested that training sub-groups were more inclined to share their experiences of good practice than to share learning from errors and mistakes; this was therefore something of a missed opportunity.
In some sites, these discussions were said to take place in annual reviews of their training programmes, whereas for others these were on-going.
There was plenty of evidence of innovation, typically following reviews of existing training programmes leading to the development and implementation of new ones based on training needs analysis or recommendations make by local investigations into child deaths (serious case reviews). Members provided practical support to innovation; this included the co-delivery of new courses, assisting with the development and/or review of a particular course, feeding in new research in a particular field, and peer review of courses delivered. The degree to which this happened in the various sites varied.

(6) True cooperation
The authors of the Effective Partnership Working Inventory [12] suggest that members should define the requirements for effective partnership working if true cooperation is to ensue. Again, the forced mandate encountered in this safeguarding training context does cast doubt on the extent to which there can be 'true' cooperation between partnership agencies. The observations and interviews confirmed limited evidence of true cooperation. Although the word 'partnership' was uttered during the course of many interviews, no examples were given as to how the training sub-groups went about defining its requirements and expectations in terms of training outputs. There was however an indication that at least one site had taken steps to design integrated policies and working practices for interagency training. In this site, the partner agency training team has worked closely with the Local Safeguarding Children Board training team to ensure that the inhouse training policies and practice closely resembled the Local Safeguarding Children Board's: "Our training strategy is written around Local Safeguarding Children Board training strategy" (Agency manager, Site G).
The extent to which Local Safeguarding Children Boards and the constituent training sub-groups provided training for partnership working at all levels of the partnership varied between sites. In some sites, much effort has been put into inviting senior managers to sit in on the training courses, with mixed success. There was some sense that some managers only paid lip service to the principles of interagency training. Some training and the need to work closely together to meet them had resulted in cultural solidarity.
Some training sub-groups spent time to develop effective processes for working together, although the ways in which they did so varied from site to site. Some used annual away days to strategise about the forthcoming training programme, whereas others build this in to the schedule of training sub-group meetings. However, two sites had not allocated time for fostering these processes; on being questioned about this many agency managers commented that it would be helpful to spend more time to talk. All members of the training subgroup had the interagency training commitments as an add-on to their current job and many were already finding it difficult to keep up to their regular workloads. There was almost unanimous agreement that the members would like to be allotted time for working on interagency training.
"There is commitment but not a lot of time to carry out functions of the training programme" (Training coordinator, Site B).
In most sites, training sub-group meetings were infrequent, generally once every couple of months for two to two and a half hours; the agendas were full of tasks needed to be agreed and assigned, with little time left over for any review of working and interpersonal relationships per se.

(5) Focus on quality and innovation
Interviews and observations revealed that there was great concern for the quality of the training and the extent to which it was meeting the needs of the partner agencies; these aspects were regularly and routinely reviewed. All Local Safeguarding Children Boards used participant feedback forms (known colloquially as 'happy sheets') as the main source of data for quality control; there was otherwise no systematic process in place in any of the sites to ensure the quality of interagency training. One member when asked about quality said that it was "Difficult to know really when we don't measure the effectiveness really well" (Agency manager, Site C). Another insisted that "Evaluation and performance management needs to be improved" (Agency manager, Site F). A third observed that there was, "Not much in addition to course evaluations…we are struggling with this issue…little or no follow-up of users of services, i.e. how do we measure this?…should we base this on referrals, something else?" (Agency manager, Site H).
The observations revealed that most training subgroups engaged in challenging and constructive debate about the programmes. Interviewees felt able openly to question aspects of the interagency training, knowing group; and all acted as intermediaries between the training sub-group and the Local Safeguarding Children Board itself. Training coordinators commissioned training courses and generated revenue through delivering safeguarding courses to non-contributing partner agencies, such as voluntary and private sector organisations. Finally, they played a key role in monitoring and reviewing courses at all levels of delivery.
There are clearly positive aspects and risks with arrangements based so significantly on the efforts of one individual. On the plus side, their personal knowledge and commitment was a significant factor in ensuring that the programme continued. However, this made it vulnerable as was highlighted in Site D during the course of the study when the training coordinator went on extended compassionate leave. The training support group members (one of whom sat on the Local Safeguarding Children Board) pulled together to provide cover and additional administrative staff were employed but it was not easy to replace him. Although this make-shift arrangement did manage to deliver the necessary safeguarding training, this highlighted the vulnerability of the training programme as a whole.

Outputs
Although the organisational structures for interagency training were similar, there were quite large differences in the 'outputs' measured in terms of the numbers and types of courses delivered by each Local Safeguarding Children Board in the training year 2007/8 ( Table 2).
The courses were categorised in terms of their target audiences as defined in government guidance [1, p. 100]. 'Foundation' courses provided a general introduction to safeguarding. Training at level 1 was for staff in 'regular contact' with children aimed at identifying and responding to child abuse. Training at level 2, focused on 'working together' for those who 'worked regularly' with children. Level 3 referred to specialist courses for children who were vulnerable because they were disabled, living with domestic violence or parental mental illness etc.
Differences in the kinds of courses are apparent ( Table  2). Four sites offered free access to foundation level e-learning courses; minutes showed that these were chosen as an inexpensive way of training large numbers of front-line staff. In other sites, foundation training was agreed to be the task of single agency training. Twenty per cent of the foundation courses were for two days. Three Local Safeguarding Children Boards did not offer any of these, having decided that it was too difficult for partner agencies to release staff for all this time and concluding that the same learning outcomes could be achieved in one day. The only evidence to sites have held annual away days for senior managers; some were considered successful but others not so much.
All training sub-groups had developed effective communication processes and these were managed effectively. Thus, training programmes and course content, both in hardcopy and via the web, were efficiently distributed. Lead persons were updated frequently about changes in course dates and pre-requisites. There was also clear and timely distribution of materials for upcoming training sub-group meeting. In short, effective communication was one of the obvious strengths of the training sub-groups. As many commented, they had very good, committed, and effective training coordinators as well as administrative staff, who ensured that the training programme was well delivered, well publicised and well organised.

(7) Interprofessional trust and respect
Between the core members of the training subgroups themselves, interprofessional trust and respect were readily observable and this was frequently asserted in the interviews. Outside these core partners, trust was not always present, as one interviewee observed, "Some people are more active than others…there is a difference between attending and doing" (Agency manager, Site E).
From the interviews, it was not clear that ordinary members of the training sub-group had a good understanding of the roles and responsibilities of the members of the Local Safeguarding Children Board itself. The Boards carried out their functions through a number of sub-groups and committees, many of which had an influence on the development and financing of interagency training. The chairs of the training sub-groups were represented on the Board and so they did understand, but they did not seem to have passed this on.

The key role of the training co-ordinator
The importance of the training coordinator in the organisation and delivery of interagency training cannot be understated; this was the key person responsible for designing, delivering and coordinating the programme. In seven of the eight sites training coordinators they delivered training themselves. They were responsible for drafting and reviewing the training strategy and the training budget. They promoted the training and, with the help of their administrators, circulated the training programme to partner agencies. Some also managed administrative staff and provided training and support for the trainers. In two sites (Site F and Site G) the training coordinator was the chair of the training sub- In order to assess volume of the output (training provision) in relation to children's needs, the numbers of courses and the days of training provided were divided by the local authorities' own estimates of the numbers of children in need (all vulnerable children, including those in need of child protection). The figures for 2009 (the nearest available) were published by Department for Children Schools and Families [15]. These analyses show that the number of courses per children in need varied from one per 54 children (Site C, a county council) to one per 156 in Site H (a London Borough), which was providing roughly two-thirds number of courses of the average for the sites. Taking into account the length of the courses offered, Site C was providing nearly twice as many course days per child in need than the average, as well as the largest number of courses and course days. This is attributable to a strong and active training subgroup and a particularly energetic training coordinator who was supported by two assistants.
The output of training courses is likely to be influenced by the size of the financial contributions made available to the training subgroup by the partner agencies. As shown in Table 2, Site C had the second highest budget provided by the partner agencies, more than twice that of Site B, which had a similar number of children in need; in this respect it was very similar to Site F, but that Local Safeguarding Children Board was provid-ing only two-thirds of the course days offered by Site C. Consequently, the cost per course day (ratio C:D) was over £1000 more in Site F compared to Site C. Course days were also good 'value for financial contribution' in sites A, B and E. However, sites G and H were considerably less cost efficient. Site G had easily the highest proportion of children in need and the largest budget, but also spent close to 50% above the average per training course day. Cost efficiency estimates, however, do not take account that in many of these the partner agencies made significant contributions 'in kind' through members of their staff who acted as trainers at no charge to the Board. So although Site G may appear less cost efficient relative to the other sites, this is large part explained by the fact that the training budget is based on annual 'subscriptions' paid by partner agencies, which were ear-marked for training and as such already included various in-kind contributions found in some of the other sites. Site H, on the other hand, was the only Local Safeguarding Children Board which did not have a dedicated training subgroup. There was relatively little involvement of agency representatives, the training coordinator was not supported by an active group of colleagues and the Local Safeguarding Children Board had taken the decision to commission all its training from external trainers, which was more expensive than in-house training.
In terms of partner agency contributions per child in need (ratio B:D), estimates varied from a low of £29.30 per child in need in Site A to a high of £62 per child in need in Site C. However, there was no simple correlation between these contributions and the numbers of children in need. This is not surprising because, in *Source: [15].

Context
In the case of interagency training for joint working, we encountered a situation or 'context' wherein we have a 'forced partnership' placed on Local Safeguarding Children Boards by a central government mandate, in other words, there is no other choice but for partner agencies to work together. Moreover, mandatory interagency training for joint working was taking place without any additional funds from central government. This mandated collaboration has affected the outcome and outputs of interagency training for joint working [17]. Some partner agencies felt that the costs borne of interagency training are too high and that they have been forced into relationships that appear suboptimal to them, yet potentially beneficial to the Local Safeguarding Children Board as a whole. This finding supports literature on organisational collaboration which states that networks that have been mandated by central governmental have different dynamics than those that come together by mutual consent [17]. However, even within a strong and well-established culture of mandated collaboration, interagency training regimes varied considerably across Local Safeguarding Children Boards. These findings are consistent with a body of research evidence on some of the negative consequences of 'mandated partnerships' in the public sector [18][19][20][21].
Agency managers displayed optimist, pessimist and realist approaches to interagency training for joint working [22]. That partner agencies were forced to collaborate as a direct result of government policy represented the 'realist' approach, while others looked at this as an opportunity to change organisational policy and practice to safeguard children and enhance their well-being, i.e. taking the 'optimist' approach. There were very few managers across the sites who took a wholly 'pessimist' approach to interagency training for joint working [17].

Mechanisms or 'process'
The training sub-group is the main vehicle for ensuring that the outcomes of interagency training were achieved. Application of the Effective Partnership Working Inventory [12] showed that shared commitment to goals and objectives and interdependence of outcomes among partner agencies were the strongest domains, followed by role clarity and cultural congruity. Focus on quality and innovation, true cooperation, and interprofessional trust and respect were the weakest domains in the inventory [12].
The success (or failure) of the training sub-groups in organising and delivering interagency training for joint working (the 'mechanism') was affected by past work-the absence of a government funding formula, Local Safeguarding Children Boards are left to negotiate the contributions from partners and in a recent survey, over half of Local Safeguarding Children Board chairs considered these insufficient [16, p. 34-36]. Anticipated cuts in partners budgets (June 2010) highlight the vulnerability of the system: a third of the 57 chairs in the recent survey ranked training first or second in importance when asked about the effects of an inadequate budget [16, p. 35].

Organisational outcomes
Organisational outcomes varied in terms of synergy, transformation and budget enlargement in this research [13]. Interagency training displayed elements of synergy as it brought together partner agencies with different assets and institutional powers to implement an interagency training programme which no one agency could effectively provide on its own. Organisational transformation was occurring; albeit on a very small scale through joint working of members on the training sub-groups, some of which were effected by power imbalances. Budget enlargement did occur via the pooling of several sources of training funds, but that this was more distributive between partner agencies as no additional monies were made available by central government. Instead, partners pooled their own staff's time and skills to develop and, in most cases, deliver the programmes.
Sites showed varying degrees of 'resource' and 'policy' synergy [14]. Agency managers felt that interagency training for joint working had increased the effectiveness of joint training as well as the efficiency in terms of the efficient use of available funding resources. Some sites have introduced new training courses and information sharing sessions, all of which have created some policy synergies between partner agencies. In short, these synergies have led to clearer standards for training staff who work with children and families and increased staff awareness about child protection issues, both of which have had positive consequences for children's services in the Local Safeguarding Children Boards participating in the evaluation. Although there have been some synergies, the extent to which real transformation has occurred is much less evident [14]. Partner agencies have accepted the need for change and to learn from other partner agencies, and a lot of ground has been made in terms of the breaking down of prejudices, misperceptions, misunderstandings, and power imbalances between partner agencies. ing relationships between partner agencies, and that between Area Child Protection Committees (the forerunners to Local Safeguarding Children Boards) and partner agencies. Where there was a history of positive working relationships prior to the formation of the Local Safeguarding Children Board, then it was more likely that interagency training working under the Local Safeguarding Children Board continued as before. This too holds true for joint working relationships that did not share a common vision-and shared commitment-to safeguarding children and were not characterised by trust and reciprocity, and where autonomy was lacking [23]. As past research has pointed out [12], it is important to have interdependence of outcomes, which-in the case of interagency training for joint working-we do. There were some challenges in the mechanisms in terms of agency managers not knowing exactly what their role was on the training sub-group and organisational culture clashes, but these were the exception not the rule and did not appear to impact on the effectiveness of the training sub-groups in organising and delivering training [12].
A training sub-group made up of agency managers 'forced' or 'mandated' to work together was successful because the responsibilities set down to them by central government act as a unifying force, which acted as an impetus for overcoming some of the challenges and barriers encountered in the organisation of interagency training for joint working in the past. Vested interests, lack of organisational and role clarity, inter-professional antagonisms, and funding conflicts-although present-did not take centre stage as each partner agency was responsible for contributing funding and obliged to collaborate under Government legislation. Moreover, having clear objectives handed down by Government in terms of training content and participation in interagency training has resulted in less training duplication and a reduction in conflicting interagency training for joint working 'agendas'.

Limitations
When interpreting the findings of this research, it is important to keep in mind that the evaluation teamthrough its inclusion criteria-chose Local Safeguarding Children Boards that would be in a position to contribute meaningfully to the study. The recruitment of sites with what could be deemed well-functioning interagency training programmes was intended to enhance the chances of observing effective partnership working and the identification of good practice from which other Local Safeguarding Children Boards may learn. Findings from an evaluation which included sites without a well-established structure for the commissioning and provision of training (e.g. training sub-committee and training co-ordinator/s) or a substantial programme of interagency training courses might have resulted in different findings.

Conclusion
This paper looked at the mechanism or 'process' (Local Safeguarding Children Boards, training subgroups and training coordinators) by which a policy intervention (statutory guidance) went on to produce particular outputs (interagency training for joint working) and outcomes (effective joint working), within particular organisational contexts (local conditions in the Local Safeguarding Children Boards). Idiosyncratic mechanisms and processes characterised by local needs and organisational capacity had resulted in quite disparate funding and training regimes across the various sites. No two Local Safeguarding Children Boards displayed identical partner agency membership on training sub-groups, contributions (fiscal or inkind), commitment to, and active involvement in the organisation of interagency training. The study also found that the context of interagency training for joint working varied across the sites in terms of reactions to mandated collaboration and approaches to dealing with this. We were, therefore, not surprised to find that the outputs and outcomes of interagency training for joint working were experienced differently by agency managers and training coordinators in different local circumstances.
The fundamental question asked in Realistic Evaluation is "What works for whom in what circumstances?" [10]. Working Together [2] has played a key role in the mechanism and outputs of interagency training for joint working. There are also strong signs that it has fostered effective joint working through interagency training, which has the potential to create a bettertrained workforce to work together to safeguard children and to maintain their well-being. However, the generic guidelines offered in these policies have not resulted in homogenous training organisation mechanisms across the various sites. Interagency training for joint working has triggered different organisational responses, producing different partnership outcomes, which very much depended on the particular circumstances of the site. The effectiveness of the outputs and outcomes of interagency training was contingent on the context in which it was introduced. What worked to produce effective organisation and delivery of interagency training in one site did not produce it uniformly across all sites. In short, there is no 'one-size-fits-all' approach to effective interagency training. However, appropriately developed interagency training mechanisms (processes) which take account of the contextual circumstances of the Local Safeguarding Children Board should produce positive outcomes for partner agencies. Further papers from this research (in preparation) will report the learning outcomes for participants on the courses in terms of their attitudes towards working together, their self-efficacy in relation to locally integrated child protection/safeguarding services and their knowledge of substantive issues.