A Bourdieusian Analysis of Good Practice Partnerships: Implications for Private, Voluntary and Independent Early Childcare Leaders

: This article applies the theoretical lens of Bourdieu to explore leadership and the distribution of power between teachers and early childhood practitioners in the private voluntary and independent (PVI) sector in a good practice partnership. It questions the effectiveness of such partnerships in bringing about sustained improvements to practice in this sector. Data were collected in three focus groups with practitioners from 10 settings that were recruited by a consortium of schools to the partnership to take part in a 10-week intervention. The ﬁndings showed that the power imbalance between teachers and early childhood practitioners cancelled out collaborative pedagogical models of leadership intended to underpin the good practice partnership. Furthermore, this power imbalance contributed to the lack of sustainability of the intervention. In conclusion, leadership practices need to be more closely aligned between early childhood practitioners and teachers by drawing on ideas of moral leadership elicited from a symbolic frame. The article has relevance for leaders operating in collaborative contexts where existing power imbalances are evident.


Introduction
This article explores the distribution of power between teachers in a consortium of schools and practitioners from the Private Voluntary and Independent (PVI) childcare sector during a collaborative intervention to improve boys', aged 36-48 months, fine motor skills. The intervention was put in place as a result of a mandatory summative assessment (Early Years Foundation Stage Profile) (EYFSP) of children's learning and development, in a Local Authority (LA) in the North of England, which had revealed an average 18% gap between girls and boys achieving a good level of development (GLD) in literacy, particularly handwriting. The article examines the policy of good practice partnerships which might be understood as networks or collaborations including primary, nursery schools and the PVI sector and are identified by the government to develop childcare professionals and improve the quality of early years provision [1]. Significantly, in these partnerships, teachers in schools have responsibility for leading quality improvement in the PVI sector. This has implications for leadership in the early childhood education and care (ECEC) context with this article considering these within the PVI sector.
There is a paucity of empirical research that explores the implications, particularly the leadership implications, of power imbalances within early childhood partnership work. Whilst it is recognised that collaborative working promotes quality and can enhance teaching practices [2], little consideration has been given to power imbalances in their success. Literatures discussing research practice partnerships, which aim to solve problems through mutual collaborations in United States of America (USA) educational contexts [3,4], have relevance for this research. Wegemar and Renick [4] (p. 12) state that "attention to roles and power and boundaries provides a foundation for developing strategies that may improve the efficacy of educational partnerships". Power differentials are common within such partnerships [5] and are influenced by gender, professional role and educational level [4] (p. 3), all of which have relevance for this research's context. There are barriers to educational partnership work due to differing cultures, norms, values and routines [6,7], which this research explores within a UK early childcare context.

Early Childhood Education and Care in England
The ECEC sector in England is a mix of local authority-controlled and government funded provision (schools and nursery schools), known as the maintained sector, together with privately owned businesses, and voluntary, community-run and social service provision, referred to as the PVI sector [8]. Of significance for this article, the maintained sector is positioned as the provider of early childhood education, the PVI sector as the provider of childcare [9]. Practitioners in the PVI sector are subject to the same regulatory body as those working in the maintained sector and are responsible for delivering the same framework, the Early Years Foundation Stage (EYFS). However, those employed in maintained settings are more likely to be qualified teachers, on a national pay scale [10]. Practitioner qualifications in the PVI sector range from level 3 to degree level and they receive variable levels of pay [11]. These differences in pay and qualifications can lead to teachers having higher status and power than childcare practitioners and, in turn, reinforce the hierarchical separation between education and care in England [9].

Good Practice Partnerships between Schools and the PVI Sector
Many Local Authorities have ceased to offer free continuing professional development (CPD) to settings [12]. However, the current government supports partnerships between schools and the PVI sector as a way of improving quality and providing CPD for those working in the PVI sector [1]. Such partnerships are not new, an aim of the National Childcare Strategy [13] was to bring together the maintained and PVI sectors as equal partners. Broadhead and Armistead [14] investigated an initiative to establish early education and childcare partnerships between schools and PVI providers and found that sharing of power across such partnerships is problematic. They identified that working conditions, professional identity, and respect for roles and status are key. They also found that effective collaborations require continuity of staff, shared leadership, resources and time [14]. Similarly, findings from research into professional development school partnerships (PDS), which exist in the US and the Netherlands between universities and schools, including pre-kindergarten, suggest that longevity is key to a successful partnership, along with the pre-kindergarten teachers identifying gaps in their practice. Such partnerships can provide leadership opportunities for pre-kindergarten staff involved but research of these partnerships is scant [15].
In England, the Government's review of effective school collaborations [16] found that perceived power imbalances between schools and difficulties in establishing shared objectives were barriers to inter-school collaborations. Furthermore, centrally funded initiatives are not the most effective way to facilitate sustained partnerships, when the funding stops commonly the collaboration also ceases. However, effective school collaborations were found to provide opportunities for staff to take on leadership roles, thus building leadership capacity and providing leadership training. Nevertheless, the same review identified that knowledge of inter-school collaborations is scarce, reaffirming that knowledge of partnerships between schools and PVI settings is scarcer.

This Good Practice Partnership
The Consortium is an alliance of schools across Yorkshire and Lancashire (in northern England) with the expressed aims of raising standards and outcomes for children and supporting continuing professional development of staff. For the purposes of the intervention, The Consortium developed a good practice partnership with ten PVI settings in order to address the stated gender gap in handwriting. Further, they worked with the Local Authority Quality Improvement Team (LAQIT) of a metropolitan borough in West Yorkshire, who were instrumental in identifying participating PVI settings. The intervention was designed and planned by The Consortium and funding was awarded by the National College of Teaching and Leadership (NCTL) in 2015. It was a ten-week programme based around "dough gym", a series of exercises set to music where children use playdough to improve their dexterity.
Twenty participants received training on the intervention, although, initially, this did not include how to assess the children's physical development. The Consortium selected the Peabody Developmental Motor Scales (PDMS) to assess and monitor the boys' fine motor skills, despite participants routinely assessing children's physical development as part of the EYFS. Whilst the PDMS can be effective in studying gender effects on motor performance they also require specific training and can be lengthy to administer [17]. Two teachers from The Consortium were trained in using PDMS but, the first attempts at using it were unsuccessful as the children were reluctant to engage with unfamiliar adults and the assessment was too lengthy to maintain the children's interest. Therefore, an assessment tool, based on the outcomes for physical development in the EYFS, was developed by The Consortium to be undertaken by the participants. They were given resources to support the intervention and provided with money to buy equipment and pay for staff cover whilst they attended six moderation events, led by two teachers from The Consortium.
As Senior Lecturers in ECEC based at a local University, with previous experience of working with young children, the authors of this article were employed by The Consortium. Our role was to elicit the participants' views, experiences and perspectives. This was because The Consortium were concerned about potential differences in status and power between them as teachers and the participants (childcare practitioners), and this could inhibit collaborative working. As The Consortium explained to us, its aim was to work with the participants on the intervention rather than do it to them. The DfE [16] identified that knowledge of inter-school collaborations is scarce, and this suggests that knowledge of collaborations between schools and the PVI sector is scarcer. Therefore, we identified an opportunity to extend understanding of good practice partnerships through a unique exploration of how power was distributed between teachers and childcare practitioners during an intervention, and to understand the implications this might have for leadership in the PVI sector.
To what extent is power shared between teachers in The Consortium and childcare practitioners during the intervention? 2.
How effective was this good practice partnership in bringing about sustained improvements to practice in the participating settings? 3.
What can PVI sector leaders learn when engaging in future collaborations of this nature?

Power and Status in ECEC
There are well documented disparities in the status, pay and conditions between practitioners working the PVI sector and those working in schools [9,18], with Moss [19] (p. 354) describing the PVI workforce as "scandalously poorly paid". This is still the case as outlined by Bonetti [20] in a recent review of the workforce in England. Selbie et al. [11] explain that there is a longstanding separation between education and childcare in England, where the institutionalised care of young children has low status and pay, as it is perceived by society to be an extension of maternal duties. The Organisation for Economic Cooperation and Development (OECD) [21] reported that across Europe low pay, status and poor working conditions remained a concern for those working with the youngest children and that practitioners associated with delivering early education, such as pre-school teachers, had higher qualifications than childcare practitioners. Moss [19] argued that the split between education and care persists in policy, clearly signaling the difference in status afforded to education over childcare. This has not changed, the current EYFS [22] sets out different requirements for levels of qualification for those working in the PVI sector than those in the maintained sector. PVI providers are required to employ one member of staff qualified to level 3, whilst the maintained sector must employ a qualified teacher at level 6.
The cumulative effect of working in a sector with low pay and status means that childcare practitioners are often relegated to positions of low power, while teachers have potentially greater power in both education and wider society. Here power is understood through Bourdieu's [23] concept of capital (economic, social, cultural and symbolic) and is employed as an analytical tool to explore the good practice partnership as a site to consolidate or challenge the power imbalance. Andrew [24] (p. 307) argues that cultural capital and "professional privilege" are more accessible to teachers, with degree level qualifications, despite them doing the same work as childcare practitioners. The intersections of gender and class are equally significant for understanding these disparities in power as 97% of childcare in the PVI sector is undertaken by women [25] and is largely the domain of working-class women [24,26]. Andrew [24] explains that class tensions are explicit between degree qualified staff, teachers, and those with less credentials, such as childcare practitioners. Furthermore, Taggart [27] notes, the positioning of childcare as the work of working-class women leaves the workforce open to exploitation by society and government as they are required to assume high levels of responsibility for low pay. Andrew concurs suggesting that, as teachers and childcare practitioners exist in the field with different amounts of power, this makes them "amenable to exploitation" [24] (p. 307). Drawing on Bourdieu's [28] work we argue that the constraints of gender, class and access to capital becomes embodied and repeated in the everyday practices of individuals and groups. Bourdieu [28] (p. 52) explains the embodiment of dispositions and classificatory systems as "habitus", and argues that this becomes internalised as "second nature" [28] (p. 56) with the result that individuals carry with them, at all times, their present and past positions in the social structure, in short, "knowing one's place" [23] (p. 82). Therefore, we argue, the persistent divide between education and childcare, in conjunction with issues of class and gender, means that practitioners are positioned to have lower power than teachers.

Leadership in ECEC
The article draws on Bolman and Deal's [7] Symbolic Frame as a lens to analyse leadership when power imbalances exist in partnership work. This Frame considers that systems, policies, routines, myths and rituals help leaders and practitioners make sense of chaotic and ambiguous environments, such as the establishment of new partnerships. Underpinning suppositions to the Symbolic Frame that are relevant to this analysis include [7] (pp. 241-242): • What is most important is not what happens but what it means. • Activity and meaning are loosely coupled; events and actions have multiple interpretations as people experience situations differently. • Events and processes are often more important for what they express or signal than for their intent or outcomes.
These suppositions complement Bourdieu's [23] concept of capital outlined above and notions of power when considering the practitioners and partnership work, particularly in relation to individuals' interpretations of meaning to themselves. Further, the Symbolic Frame has relevance for leadership in situations where individual commitment and motivation are essential to success; there is ambiguity and uncertainty, conflict and scare resources are apparent; and there is "bottom up" working [7] (p. 303), all of which are pertinent to this study's context.
Effective leadership practices concomitant with the Symbolic Frame includes inspiring others; energy and passion, and the effective communication of that to others; visibility; and use of values as a base for building a cohesive culture [7] (pp. 320-321). This resonates with the ECEC context as Hallet [29] notes that leaders should be a positive role model and draw on their passion and enthusiasm when driving change in this context. Practitioners in ECEC commonly care deeply for and about young children and leaders can draw on this shared value to build a culture with cohesiveness and meaning, whilst communicating a vision of the organisation's capabilities and mission [7] (p.321). When considering leadership practices, these points suggest an attending to others, both practitioners and children, which echoes conceptions of moral leadership and an ethic of care [30,31], whilst maintaining the key facets of the Symbolic Frame.

Methods
The research approach was interpretive within a qualitative methodological paradigm with the purpose of being able to explore the opinions and experiences of practitioners participating in the intervention [32]. King and Horrocks [33] argue that interpretive research enables people to share their experiences whilst the researcher searches to uncover meaning from what is shared.

Focus Groups
Focus groups were chosen to collect data from twenty participants (two from each setting) in order to establish a collaborative approach, develop trust and generate thinking through the interactions of the group [34]. They took place in a centrally located school selected for convenience of travel. In the first focus group the participants were split so that each group had a maximum of ten participants. Unfortunately, subsequent focus groups included all the participants as other commitments, including afterschool care, meant some could not attend. Therefore, to avoid losing participants it was decided to include them all in one sitting. Although focus groups are suggested to be useful for eliciting large amounts of information from small groups there can be limitations. It can be difficult to distinguish an individual view from the group and some can be influenced by others [34]. The group size in this study could have increased the effects of these limitations; however, we were sensitive to these constraints and offered every member of the group an opportunity to contribute. Participants were encouraged not only to answer our questions, but to talk to each other and relate their experiences to those who shared a common frame of reference, whilst focusing on the intervention [35,36]. By adopting this approach and asking probing questions we were able to extract meaning, considered to be a strength of the focus group approach [37]. Prior to the first focus group, participants were met to establish rapport [38]. We adopted an informal, friendly and open approach, sharing our experiences as former early years practitioners to establish trust and minimize power imbalances between researchers and participants [38]. At a second meeting we explained the chosen method of data collection and agreed how the focus groups would operate while gaining consent from all participants. The focus groups were recorded and transcribed by the researchers.

Interviews
Once we had completed the focus groups and written the final report for The Consortium, we took the opportunity to collect additional data to investigate if the intervention had continued within settings. This was to ensure better insight into its sustainability. We consequently interviewed two members of the LAQIT, with responsibility for quality in the participating settings, and a participant who had been involved in the intervention. We carried out semi-structured interviews to gain individual thoughts and experiences resulting in a better understanding of their distinct involvement during and beyond completion of the intervention [38].

Validity and Reliability
To minimise researcher bias, we transcribed the audio recorded data separately, then came together to discuss and share our initial thoughts. We discussed and analysed direct quotes from participants, considered by Chandler, Ansty and Ross [32] as a way to acknowledge the voices of participants. Critical discussion took place between us, and this strengthened the validity and reliability of the research. By using "investigator triangulation" [39] (p. 228) we were able to question each other's opinions and thoughts during and after the research phase [39].

Ethics
Ethical approval was granted through the University's mechanisms and participant informed consent was obtained. The participants were anonymized to ensure that they could not be identified by The Consortium during the intervention or subsequently, and, to protect the identity of all involved in the good practice partnership, a pseudonym has been used.
During the focus group meetings and interviews it was essential that authentic voices of participants were recorded in an attempt to produce "uncontaminated theoretical interpretation" [40] (p. 715). We represented the participants' voices through direct quotes taken from the focus group meetings and interviews, but as St. Pierre and Jackson argue [40] (p. 716) "words can never retain presence", therefore, we acknowledge that it was not possible through the written word to capture the nuances of body language, tone of voice and facial expression, which were present during data collection. Although, it is argued that unequal power relationships between researchers and participants are unavoidable within the realm of qualitative social research [41,42], we attempted to reduce the power imbalance by holding the interviews and focus groups at locations convenient to the participants [41], and ensured the participants had opportunity to provide feedback on the transcriptions. Nevertheless, power returned to us during analysis of the data as we chose what to include in our reports to The Consortium; therefore, making choices for the participants [32]. As researchers we were cognisant that we were employed by The Consortium but we had nothing to gain from omitting participant perspectives and, because of our former experiences as practitioners and our current roles educating practitioners, we were keen to ensure the participants' voices were heard.
Not only did we need to consider power relationships and ethical practice with the participants but also with The Consortium. In line with BERA's ethical guidelines [43], we made our research available to The Consortium and the participants for scrutiny, and both The Consortium and the participants were invited to a dissemination event at the University. However, as May [44] suggests, problems can lie outside of the researchers' control when a sponsor is leading a project and we had little say in the dissemination of reports with third parties.

Analysis and Discussion
Key features of the intervention were used as a chronological framework to organise and analyse the data:

•
Bid and initial design • Training and resources • Assessment of children and invention implementation • Moderation events and meetings with researchers.
Having done this, Bourdieu's [23] concepts of capital and Bolman and Deal's [7] Symbolic Frame were then applied to explore the distribution of power at each key feature and make recommendations for leadership practices in this context.

Bid and Initial Intervention Design
The participants were not part of the initial bid and design of the intervention although this can, arguably, be explained by the short time period allowed for the completion and submission of such bids. However, by not including the PVI settings from the outset, The Consortium had control of the intervention thus reinforcing existing differences in power and clearly signaling The Consortium as the more powerful partner in the good practice partnership.
Nevertheless, analysis of the focus group data suggested that being involved in the intervention engendered positive feelings in the participants. One described herself as "feeling lucky, privileged to be involved" while another described herself as a "bit empowered". Although this comment suggests this practitioner felt an increase in power by being selected to take part, neither her nor the other participants questioned their involvement in the intervention. This was despite practice in the PVI settings being positioned as the problem, even though the EYFSP data informing the intervention was collected when the children had spent almost a year in school. The participants were grateful to be involved describing themselves as "lucky", "thankful" and the intervention "as a great opportunity for us". Seemingly, the participants accepted the negative judgement of their practice inferred in the design of the intervention. This could indicate, as Bourdieu [28] (p. 56) explains, that for the participants and the teachers involved in this partnership, the dispositions associated with the traditional separation between education and childcare had become embodied and internalised as "second nature", with the result that the respective positioning of their practice was not questioned. This could also be understood as embedded routines and systems as part of these participants' symbolic frames.
The Symbolic Frame [7] offers an alternative explanation of the participants' gratitude at their involvement as expressed by the practitioner interviewed six months after the end of the intervention. She said: Teachers teach and we teach but just in a different way on a level that's appropriate for the children . . . It showcased what we do does have purpose, and it's not just coming in and having a bit of childcare.
The participant appeared to be aware of the discourse of childcare which undermines their skills as practitioners and results in their work being undervalued. However, involvement in the intervention for this participant signaled that her practice, although different from teaching, had value and by showcasing it there was an opportunity to improve her access to capital and the status of ECEC work.

Training and Allocation of Resources
Effective leadership within the symbolic frame includes inspiring others and creating a shared vision [7], and through the training and allocation of resources provided, this was achieved by The Consortium. Training included attendance at two conferences and the opportunity to meet Alistair Bryce-Clegg, a consultant specialising in early childhood education. The participants described themselves as "inspired" and "having their eyes opened". One said, "it was fascinating learning all about the shoulder and how that works". The participants appreciated the opportunity to extend their knowledge and this echoes research in New Zealand [45] which found that the most effective CPD for practitioners in ECEC occurred when the programme disseminated new knowledge. However, the same research identified that CPD worked best when the practitioners had opportunities to contribute their ideas, which was not a feature of this intervention. Participating settings also received "dough-related" resources and additional money to spend on provision to support children's physical development. This money for resources was appreciated because, as one participant explained, "as a charity we don't have spare pennies so it was really good to say I'm going to spend this on fine motor skills for my children".
Through the symbolic lens, the training and allocation of resources can be framed as effective leadership, whereby the practitioners, settings and children benefit. However, it is well documented that PVI sector funding has been insufficient to cover costs and is below that of maintained schools [46]. This exacerbates the education and care divide, reinforcing the view that professional development opportunities tend to be available for teachers rather than childcare practitioners [21]. The lack of financial resources to support professional development in PVI settings limits practitioners' access to cultural and educational capital. Whilst they did access elements of such capital through the training and allocation of resources, nevertheless, The Consortium controlled all resources and, through their superior economic power, continued to perpetuate the power imbalance between education and care, teachers and practitioners, thus reinforcing "powerlessness" in the practitioners [47] (p. 193).

Assessment and Intervention Implementation
The design and use of the assessment tool for measuring the child's physical development was integral to the intervention as a way of demonstrating its effectiveness. It was evident from the intervention's design that, initially, the teachers saw themselves as being better placed to carry out the assessments as they excluded the participants from the choice and use of the original assessment tool and its subsequent modification-another potential example of the teachers' "professional privilege" [24] (p. 311) and greater power. However, power shifted to the participants when responsibility for the assessments transferred, and this had symbolic value to the participants who all described feeling increased confidence as one commented, "I took my baseline, I was dead impressed, it improves your confidence". This reported increase in confidence is noteworthy because, as Mukherji and Dryden [48] state, childcare practitioners continually observe, assess, record and plan for children's learning, they have considerable knowledge and expertise in assessing children's development. Yet despite this expertise the participants appeared to need validation from the teachers. This validation, or making visible the participants' skills and expertise in assessing children's development, is an aspect of effective leadership within the symbolic frame, and appeared a powerful driver for change and implementation of the intervention.
The participants repeatedly referred to increased feelings of confidence as they carried out the intervention. Significantly, these feelings allowed them to share their knowledge and expertise with other colleagues in the setting. They took responsibility for leading practice as one explained "it improves your confidence, we're able to cascade so now all the staff are involved, more confident and they are taking note". This cascading of expertise might be understood as the practitioners developing social and cultural capital, and also their leadership skills. The participants were able to take on informal leadership roles in the setting and this concurs with findings from research into professional development of prekindergarten staff involved in a PDS in the USA [15]. Setting leaders could make sense of these positive developments and encourage other practitioners by sharing and celebrating this success [7] (p. 321). The participants also had responsibility for choosing the boys taking part in the intervention and they noted its positive impact on those selected. This, in turn, enhanced the participants' levels of confidence with one stating, "it's improving their independence skills, it's special, you can see the difference in four weeks, and it's a boost".
These views from the participants suggest they experienced improvements to their levels of knowledge, practice and confidence from those aspects of the intervention they were able to control in the setting, including the boys' assessment, implementation of the intervention and the training of colleagues. Dyer and Firth [49] explain that confidence matters when working with young children. It contributes to the development of social and cultural capital akin to the "professional privilege" [24] (p. 307) of teachers and could support the development of practitioners' cohesiveness and the fostering of shared values.
Cultural capital might also be accrued through recognition of the practitioner's expertise by a teacher. This was articulated by one participant who stated "when X (a teacher) asked if she could come and observe, so you are passing the benefits onto their setting . . . you feel privileged". Andrew's [24] (p. 311) findings are pertinent here, in that, childcare practitioners have a "desire for recognition" and are aware that their profession and skills need to be more widely acknowledged. This resonates with effective leadership as described within the symbolic frame, which highlights the strong signal that such actions convey about the value of a person's work. The Frame's suppositions include consideration of the importance of meaning to the individual and what events express or signal as well, with the recognition of expertise appearing significant to practitioners in this collaboration.
The data suggested that there was at least a temporary shift in power from the teachers to the participants during implementation, which appeared to be associated with increased levels of confidence described by all participants. However, significantly, this was not apparent when they referred to aspects of the intervention not occurring in the setting. This was illustrated at the early stages of the implementation during the first focus group, one participant remarked, "it would be really helpful if The Consortium could share everyone's contact details, to support each other, share best practice and check our understanding". The rest of the group agreed this would be useful, but The Consortium did not act on the suggestion. It may be that The Consortium assumed that this could be left to the participants as it is common practice in professional networks to share contact details, thus building social and cultural capital. Whilst the participants recognised the benefits of creating a network amongst themselves, it appears that they felt the power to do so resided with The Consortium. Six months after the intervention ended one practitioner commented: It's a shame we didn't share details and keep in touch . . . it was nice to share ideas and speak to each other . . . I think we would have continued doing that.
Such networks are a form of social and cultural capital as they can bestow advantage or provide access to power for those included [50]. Cultural capital is significant here as it is symbolic of status and power [23] and, what might appear to be the small act of the Consortium and the participants not exchanging contact details, has a larger symbolic significance as it reproduces the relative social positions in the field of teachers and participants [23].

Moderation Events and Meeting with the Researchers
The power imbalance between the teachers and participants continued to play out in the moderation meetings, perhaps unsurprisingly as these were held in the lead school and facilitated by a teacher, symbolic of the existing hierarchy within the partnership. When power imbalances exist in the arrangement and running of meetings, decisions around place, space, agenda setting and chairing should all be considered by leaders to help reduce the symbolic reinforcement of existing hierarchies. This was demonstrated by data from the second focus group when the participants reported to us that they were confused about some of the dough gym exercises and worried that they might be doing them incorrectly. One said "we didn't know what a lasso was", with another stating "we didn't know if the windmill was two arms or one". The participants suggested to us that a video of the moves would be useful, but they had not discussed this with The Consortium at moderation meetings. It was left to us to request this in our feedback. The Consortium responded by producing a DVD, which the participants' reported had "clarified things". Interestingly, the power imbalance between participants and us as researchers appeared to be less than that between the teachers and participants.
Even though the participants were positive in their response to the intervention, they stated they were unlikely to continue with the programme in its current form, the interviewees from the LAQIT confirmed: Most of the settings have continued to offer dough gym activities a few times a year though none of them are baselining or tracking using the specific criteria from the project.
The participant interviewed six months after the intervention also regretted that opportunities had been missed to extend the intervention into other settings. She commented: At one point they [The Consortium] were saying that we'd be a point of contact for them [settings] to show them how to do it and I was quite looking forward to that, and that would have been quite nice, to network with other settings and passing on your knowledge . . . but it just never seemed to come off. It's disappointing because you were the expert.
The participant saw herself as an "expert" in the intervention but once again constrained by a lack of access to social and cultural capital she did not consider taking the lead on contacting other settings and sharing her expertise. This was potentially an opportunity for a setting lead to energise this situation by continuing to celebrate success while building on a shared value around the care of young children [7,29].

Conclusions
The aim of the research was to explore the distribution of power between teachers and childcare practitioners in a good practice partnership. Through the unique use of Bourdieu's work in this context we propose that the good practice partnership was not a site for power sharing. From the outset the teachers had greater power and this remained largely unchallenged. As Bourdieu [28] explained, it appeared that each knew their place. This is a small-scale study of a short-term intervention, nevertheless, it has offered insight into the effect of power imbalances on such partnerships as a means of improving practice in PVI settings. In this case, we argue that the power imbalance contributed to the lack of sustainability of the intervention and therefore it generated, at best, short term improvements to practice. Therefore, if the English Government is to increasingly rely on good practice partnerships as a way of improving practice in the PVI sector, then, as West [51] suggests, open debate is required about inequalities between the sectors and the workforces. This research highlights the need for greater awareness and appreciation of the skills and expertise of practitioners in the PVI sector by teachers, government and society. In the short term we argue that sustained improvements to practice could be supported through good practice partnerships if practitioners from the PVI sector were treated as equal partners and involved in all stages of an intervention. Longer term, we argue for a paradigm shift in approaches to improving quality from short term collaborations to sustained partnerships which equally value the skills and expertise of both sectors' workforces. This would require social, political and financial commitment.
As stated already, for most of the intervention power was retained by teachers in The Consortium and this might, in part, be explained by their limited understanding of the childcare practitioners' skills and expertise, thus undermining their intention to do the intervention with the participants and not to them. This chimes with the findings of Broadhead and Armistead [14] who argued that clear insight into the roles and responsibilities of the other sector supported such collaborations. It appears that the "professional privilege" [24] (p. 311) and power of the teachers was maintained in this good practice partnership. Nevertheless, participants reported improvements to practice and improved levels of confidence for both them and the wider staff teams in the settings. For a small number of participants, some increase in access to social and cultural capital was acquired. The data suggests that the intervention was sustained in some settings, but in a reduced and modified format. Perhaps this was not surprising in this short-term good practice partnership because as West [51] found, sustained collaborative arrangements required shared values, shared leadership, and reciprocity, which were not established as part of this intervention.
Bolman and Deal's [7] Symbolic Frame was adopted to explore potential leadership practices that could have supported this partnership work, with suggestions for future practices identified. The power imbalances could also be understood as embedded routines and systems as part of these participants' symbolic frames and understanding of these might support leaders in making sense of potentially chaotic and ambiguous partnerships [7] (p. 236). This has provided an understanding of the rituals involved when power imbalances are evident and we tentatively suggest implications for practice that could help leaders when dealing with colleagues who are engaged in similar collaborative work. Leaders could draw on an ethic of care and moral leadership to unify practitioners under a shared value of care for young children. Setting leaders within the PVI sector should take opportunities to celebrate success and use that success to energise their teams through the duration of partnerships. This can continue to build confidence in their staff whilst allowing other colleagues in the setting to see positive interactions with those who might be viewed as having a higher power status. Whilst this Symbolic Frame [7] appears relevant to support the analysis of this sector and context, it is questionable whether its suggested leadership practices would be successful when such power imbalances exist. Leaders within the PVI sector could also "know their place" when supporting practitioners in such partnerships and the suggested leadership practices might still be ineffective. This suggests a potential theoretical limitation of this construct that requires ongoing research around the influence of power imbalances within collaborations. However, using two significant theoretical lenses from Bourdieu and Bolman and Deal has provided a deep understanding of the Symbolic Frame in this context to identify opportunities to rethink leadership practice.
The introduction argued a paucity of literature exploring power imbalances within early childhood partnership work and the findings make tentative steps to provide evidence for this gap whilst acknowledging the limitations of small-scale research from a shortterm, UK intervention [34]. This article has noted a growing literature base considering effective educational partnership work [3][4][5][6]15], however this research offers an in-depth, theoretically informed analysis in a narrow context. We suggest that more research is needed into how good practice partnerships, between schools and the PVI sector, can be developed and sustained. Further, both national and international perspectives should be considered through a number of small and large-scale methodological approaches. As Peleman et al. [52] state, because of austerity resources are limited for improving practice in many European countries; therefore, policy makers require empirical evidence into the dynamics of good practice partnerships that have worked effectively. Such empirical findings could be generalised to a number of educational partnership contexts.  Data Availability Statement: The datasets used/or analysed during the current study are available from the corresponding author on reasonable request.