Collaborative innovation and activation in urban labour markets

Collaborative innovation in public services

Increasing interest in supporting innovation in public services has been driven by a number of factors including: the need to respond to so-called ‘wicked problems’ such as the ageing of the population and resulting increasing demands for more personalised services (Oosterlynck et al., 2019); an acknowledgement of the value of new service models addressing needs poorly served or unresolved by traditional state services (Evers and Ewert, 2015); an interest in more cost-effective provision that avoids unnecessary boundaries between public services; and an enthusiasm for governance models that seek to facilitate collaboration across public, private and third sectors (Lopes and Farias, 2020).

In the field of labour market activation, policymakers have often justified the extension of compulsory work-related activity and welfare conditionality by promising that innovative, personalised services will be made available to address the complex barriers reported by vulnerable groups such as single parents (Gingerbread, 2017). Yet, research from the UK, and other liberal welfare states, suggests that compulsory activation funded by the state and contracted out to mainly for-profit providers has actually produced more standardised services (Fuertes and Lindsay, 2016). Promoting collaborative innovation may therefore offer an opportunity to make good on policymakers’ unfulfilled promises of activation that is personalised and empowers vulnerable jobseekers (Lindsay et al., 2020).

Sorensen and Torfing (2011: 848) have defined collaborative innovation as ‘an intentional process that involves the generation and practical adoption and spread of new and creative ideas, which aim to produce a qualitative change in a specific context’, with collaborative problem-solving across organisational and sectoral boundaries a key feature (Sorensen and Torfing, 2018). For Torfing (2016: 237), the concept of collaborative innovation has particular power in public services given that issues ‘defined and recognised as public tend to invoke a collective ownership of the task of developing a new and innovative solution’, while the ‘wicked problems’ faced by policymakers can only be met by incentivising diverse public stakeholders to share expertise, knowledge, resources and networks. New forms of activity where diverse stakeholders engage in mutual learning, and participate as equals in (and take joint ‘ownership’ of) decision-making on the resourcing and content of public services are seen as the processes at the centre of collaborative innovation (Lindsay et al., 2020; Lopes and Farias, 2020; Torfing, 2016). As noted above, the sought outcome of collaborative innovation – the dependent variable of which we must find evidence if we are to claim that collaborative innovation has made a difference – may range from a break with the aims and design of existing services to establish genuinely new approaches, significant changes affecting ‘who delivers’ and ‘what is delivered’ in services; and/or a (hopefully positive) transformation in service users’ experiences at the frontline (Sorensen and Torfing, 2018).

We have noted that the extant literature identifies appropriate forms of institutional design and leadership as key to facilitating collaborative innovation (Torfing et al., 2020). First, it is argued that collaborative innovation tends to flourish where there is ‘a self-governed participatory arena’ for citizens to engage with service providers, as well as mechanisms for cross-sectoral and multi-agency collaboration ‘that facilitate joint-planning and the exchange of knowledge and ideas’ (Sorensen and Torfing, 2018: 411) and provide clarity on rules of engagement and transparency in decision-making (Torfing et al., 2020). These features seem to connect with a broader literature on the potential value of collaborative governance and funding structures. For our purposes, collaborative governance represents arrangements where multiple public agencies engage with non-state actors in consensus-oriented and collaborative processes to plan and manage public services (Bussu, 2019). Advocates of such collaborative approaches point to positive impacts from governance mechanisms that are: inclusive, tapping the assets and knowhow of stakeholders from diverse communities (Cepiku, 2017); sufficiently open to encourage the joint development and testing of new solutions and support the diffusion of successful innovations (Lopes and Farias, 2020); and consensual but creative, to support and nurture a constructive management of difference through conflict mediation and mutual learning (Sorensen and Torfing, 2015).

Sorensen and Torfing (2018) and Torfing et al. (2020: 597) also see leadership roles as central to ‘producing’ collaborative innovation, arguing that the evidence points to a need for ‘leadership . . . to be re-conceptualised as adaptive activities to bring actors together, create trust, enhance information-sharing, facilitate collaboration, spur mutual learning, manage risks, and track results’. In line with this argument, the collaborative governance literature identifies a complementary role for distributive leadership practices that devolve a degree of decision-making and empower local stakeholders (Beirne, 2017) – within a collaborative space, ‘effective leadership will tend to be facilitative [emerging] from the practices and interactions of different actors, who might lead at different parts of the process’ (Bussu, 2019: 63).

From this perspective, delivery managers and street-level practitioners should be empowered through distributive leadership that ‘can help them to focus their attention, search for new ideas and test the most promising ones in the course of daily operations . . . even though they do not have a formal leadership role’ (Sorensen and Torfing, 2015: 157). For distributive leadership to work, there needs to be: delegation of decision-making authority (within agreed boundaries); resources and time made available for stakeholders at relevant levels to take on leadership roles and tasks; and consensus-based processes that provide voice for representatives from different sectors and organisations to collaborate on leading programme development and delivery (Torfing, 2016). An empirical literature in healthcare and education points to potential benefits that can be delivered by distributive leadership, by drawing on the assets and insights of people at different levels and organisations with complementary expertise, sharing the work of leading, and gaining the buy-in and trust of a broader range of stakeholders (Beirne, 2017).

Finally, the broader collaborative innovation literature highlights the importance of practices promoting co-production in delivering new content and ways of working that transform street-level services (Stevens and Agger, 2017). For Cepiku et al. (2020: 55) such practices involve ‘institutional arrangements [including] a participation process’, that is, ‘a well-defined process with a known set of sequential and standardised procedures’ that involves service users and community stakeholders and ensures ‘more equitable relationships’ and ‘improved transparency’ in the co-production of programme content. Thus, co-production means that citizens are empowered to contribute their assets to, and act as equal partners in the design of, innovative solutions within their own services. Indeed, the relationship between citizen-service users and frontline staff is seen as central to the former’s involvement in co-produced collaborative innovation (Sicilia et al., 2016). Street-level workers who have the time, resources and skills to build networks and support participation in local communities have been shown to be key to supporting co-production in urban areas (Sorensen and Torfing, 2018).

Previous studies of activation have rarely specifically referenced this conceptualisation of collaborative innovation, although Lindsay et al. (2020) have reported on the role of boundary spanning frontline workers in supporting such innovative practices in local labour market services, while Andreassen et al. (2020) use the concept to frame a discussion of the varying success of practice-sharing initiatives among welfare providers in Norway. There is also a related social innovation literature that highlights the value of localised activation experiments, including in cities (Lindsay et al., 2018; Evers and Ewert, 2015). More generally, case studies have evaluated how innovative programmes have delivered added value by linking into demand-side economic development strategies and deploying services ‘closer to need’ in urban communities (Orton and Green, 2019).

This is not to say that collaborative local services are a panacea for the challenges of responding to the needs and aspirations of disadvantaged groups and communities. A critical literature notes that localised services – and especially those delivered in partnership with the third sector – are sometimes seen by some policymakers as a means of containing costs and compensating for the ‘downsizing of national public commitment to national social citizenship rights’ (Andreotti and Mingione, 2016: 254). It is also noted that an over-reliance on localised provision risks reinforcing resource and service inequalities within and between urban areas (Sicilia et al., 2016). Nor does localisation always produce collaboration and/or innovation. Local institutional legacies, labour market contexts, social networks and access to financial resources (among other factors) may limit the extent to which different urban stakeholder communities pursue or reject collaboration over other forms of interaction, and prioritise innovation over established ways of working (Andreassen et al., 2020; Jacobsson et al., 2017).

Nevertheless, the idea that collaborative innovation can help stakeholders to arrive at more responsive local solutions and address the needs of communities poorly served by ‘traditional’ welfare services remains an attractive prospect – ‘increasing localisation in terms of capacities and responsibilities’ can create ‘openings for socially innovative actors and networks . . . mobilising new resources and adapting interventions to local specificities’ (Oosterlynck et al., 2019: 186). In relation to the research reported in this article, our argument is that the collaborative innovation outcomes discussed above – a fundamental shift in the content of services and the transformation of ways of working at the frontline – delivered positive benefits for single parents in these urban areas, in terms of a sense of empowerment for most, improved self-efficacy, and access to learning and economic resources for many, and progression into paid work of their choice for some.

The policy context

Local activation targeting vulnerable groups like single parents provides a particularly interesting context for a discussion of the potential value of collaborative innovation. In liberal welfare states like the UK, new public management (NPM)-dominated governance has seen largely standardised, contracted-out activation imposed across urban areas with little recognition of the specific challenges of different cities, local communities or particular user groups (Lindsay et al., 2018). The activation policy agenda pursued by successive governments since the mid-1990s has seen the DWP contract out national programmes to large, mainly for-profit ‘prime contractor’ organisations. Policymakers have claimed that payment-by-results contracts incentivise prime contractors to design innovative and personalised provision (Fuertes et al., 2020). At the time of our research, the DWP used this model to fund all main activation programmes targeting unemployed groups, including single parents (with the national ‘Work Programme’ – since re-branded as the ‘Health and Work Programme’ – absorbing most resources).

Critics have countered that the contractualism, focus on cost reduction and target culture created by the DWP has led to a standardised, ‘work-first’ approach where providers have little discretion to tailor or shape their services (Bennett, 2016). There is evidence that strict contractual oversight and a performance management regime that demands job entries as a condition of financial survival have produced ‘structural rigidities that inhibit the innovation process’ (Casebourne and Coleman, 2012: 28). These trends have been accentuated with the establishment of the Health and Work Programme, which draws on significantly smaller budgets to offer an even narrower range of services – ‘a very major diminution of UK active labour market policy’ (Orton and Green, 2019: 4). While (at the time of the research) the UK Government held all reserved powers over compulsory activation, in Scotland a parallel network of local employability and training provision has emerged, which since the mid-2000s has been formalised under ‘Local Employability Partnerships’ that are coterminous with Scotland’s 32 local government areas. Such partnerships co-ordinate non-mandatory activation normally funded by the devolved Scottish Government and local government, with many services delivered by further education colleges or the third sector. Such local services have been commended as providing tailored provision for vulnerable jobseekers, filling the gaps left by standardised work-first activation; but there is also a sense of a disconnected multi-level governance (MLG) whereby there is limited co-ordination across local/Scotland-level and UK-level policy and programme design (Fuertes et al., 2020).

The other main component of the UK’s approach to activation involves increasingly stringent welfare conditionality, which requires working-age benefit claimants to engage in a range of work-related activity (including participation in welfare-to-work programmes) and follow other strict conditions or face financial sanctions. While single parents were not initially a key target group for stricter benefit rules, successive governments have overseen ‘incremental inroads into a comparatively light work-related conditionality regime’ (Whitworth and Griggs, 2013: 126). Mandatory work-focused interviews with advisers based at the public employment service, Jobcentre Plus, were introduced from 2001 and have since been extended and made more frequent. The introduction of ‘Lone Parent Obligations’ in 2008 effectively rendered single parents subject to similar levels of conditionality and compulsion as all other job seekers (Campbell et al., 2016). Grassroots organisations have pointed to the negative financial and wellbeing impacts of welfare conditionality on single parents and their children (Gingerbread, 2017).

The research context

Our research discusses the experiences of single parents and stakeholders involved in the MIW programme. MIW was a programme of support for single parents with complex needs, defined as those: with disabilities, or caring for someone with disabilities; with a large family (three or more children); residing in depressed labour markets; living with complex personal circumstances; with little work experience; or unemployed for more than two years. The aim was to help single parents toward or into paid employment, based on voluntary participation. Services included a range of learning and employability-building activities, support in accessing (and some transitional funding for) childcare, debt and benefits advice, health and wellbeing provision, and (crucially) intensive, one-to-one support from a local keyworker.

MIW was supported by the Big Lottery Fund in Scotland (hereafter ‘The Fund’), a non-departmental public body responsible for distributing 40% of all funds raised for good causes by the UK’s National Lottery; it supports community projects and has a specific focus on engaging service users with multiple and complex needs. MIW was delivered between 2013 and 2017 in five local government areas. We carried out research in all five areas, and have reported our general findings elsewhere (Batty et al., 2017). We restrict our discussion in this article to MIW’s operation in Scotland’s two main cities: Edinburgh and Glasgow.

At The Fund’s request, MIW was delivered through area-based partnerships led by third sector organisations working in collaboration with providers in the public sector, with local government and/or its delivery agencies as a key partner in all areas. Targets for numbers of single parents to be engaged were agreed between The Fund and partnerships, but the programme eschewed the payment-by-results model that dominates contracted-out activation in the UK. Accordingly, partnerships and providers were not incentivised to achieve job entries as quickly as possible in order to secure their financial survival, but rather to engage the agreed number of single parents (or more) and support them toward appropriate employment or other positive learning or personal development outcomes. A similar governance and funding model has been deployed by The Fund in other programmes targeting vulnerable unemployed groups, for example, ‘Talent Match’ which seeks to support disadvantaged young people, with many of the positive features (and challenges) identified below also reported in these cases (Wells et al., 2018).

MIW engaged with 3115 single parents over four years. Of all participants across the programme, 422 were engaged through the Edinburgh MIW team, compared with a much larger group of 1475 participants in Glasgow. Almost all participants surveyed after six months reported positive learning or employability-development outcomes, while 30% had entered paid employment. There was a significant difference in job entry rates reported for MIW participants in our two cities: Edinburgh (35%) and Glasgow (20%). This is likely to reflect labour market conditions in the two cities (see below) and the scale and complexity of the user group engaged in Glasgow: while there were many similarities in the profile of MIW users, according to our survey evidence reported elsewhere (Batty et al., 2017), those in Glasgow were significantly more likely to be unqualified, to have never had a job, and to report low levels of ‘job search self-efficacy’ – a self-belief in their capacity to return to work.

MIW’s overall job entry rate of 30% mirrored exactly that achieved for single parents participating in the UK Government Work Programme at the time of our research (Gingerbread, 2017), and was slightly higher than the 28% achieved by a previous Scotland-wide programme targeting single parents (Graham and McQuaid, 2014). More specific comparison of the extent and range of impacts delivered by the different programmes is difficult – for example, official data for the Work Programme provide little by way of analysis of learning or health benefits, but it is important to note that, at any rate, broader ‘human capital development’ and wellbeing outcomes have rarely been an explicit priority for such work-first activation programmes (Fuertes and Lindsay, 2016). Indeed, there is evidence that experiences of compulsion and welfare conditionality have impacted negatively on the wellbeing of some vulnerable jobseekers, including single parents (Campbell et al., 2016) – the potential for such negative impacts is not explored by evaluations of national activation programmes.

MIW was relatively generously funded, with the cost-per-job outcome of £7424 (€8740) in 2017. The (generally positive) discussion of the impact of the programme below therefore needs to be understood not as evidence that localised programmes can deliver activation ‘on the cheap’, but rather that well-resourced, locally responsive provision can deliver value-for-money (Oosterlynck et al., 2019). That said, the cost-per-job outcome reported above was similar to that for the UK Government’s Flexible New Deal of £7495 (€8824) in 2017. (The UK Government’s Flexible New Deal was the predecessor to the Work Programme, for which single parents were one target group. Specific Work Programme figures for this client group are not available) (Batty et al., 2017).

The urban context

The urban contexts discussed below are also somewhat distinctive. Glasgow is Scotland’s largest city, with a population of approximately 599,000. It is a major centre of economic activity, particularly in services, but continues to experience challenges associated with post-industrialisation. Indeed, Glasgow consistently reports among the highest levels of worklessness of any Scottish city (24% of all households, compared with only 15% in Edinburgh and a Scotland average of 17%). Economic activity rates are approximately 69%, significantly lower than the Scotland-wide and Edinburgh rates of 77% (Office for National Statistics (ONS), 2020). At the time of the research, Glasgow’s ILO unemployment rate was 5.6%, low by recent historical standards but above the Scottish average (4.1%). Glasgow reports the highest proportion of single-parent households in Scotland and more than half of single parents are not in employment.

Edinburgh is Scotland’s capital, with a population of approximately 482,000. While there are pockets of unemployment and poverty, the city benefits from one of Scotland’s strongest local labour markets, with business services and hospitality and tourism among its key sectors. At the time of the research, unemployment stood at 3.2%, below the Scottish average (see Table 1). Single-parent households as a proportion of all family households were near the Scottish average of 27% (and well below the Glasgow figure of 40%). Like Glasgow, single parent unemployment was higher than the Scottish average (48.5% compared with 41.9%).

OPEN IN VIEWER

Table 1.

Single parents in case study cities.

	Single-parent households as % of all households with children	Single parents not in employment as % of all single parents
Edinburgh	27.0	48.5
Glasgow	40.4	50.4
Scotland	27.6	41.9

Source: NHS Health Scotland (2016).

Andreotti and Mingione (2016) note that localised activation and welfare provision is most likely to deliver innovation where local authorities have the scale, resources and access to services and professional expertise to respond to demand from heterogeneous user groups. There was an implicit acknowledgement of the need for scale, resources and infrastructure within the funding of the MIW programme – the areas selected to test the initiative were Scotland’s five largest local authorities; and this research focuses on the two largest areas (and Scotland’s two main cities). As we will see below, the availability of (and sometimes gaps in access to) service infrastructures and professional expertise were important to understanding the successes and limitations of the programme.

The two cities also featured somewhat different approaches to the delivery of MIW. In Edinburgh, a gradual process of partnership formation saw the emergence of a programme that sought to engage single parents city-wide, but was based in four localities reporting high levels of poverty. Keyworkers were recruited to a crucial street-level engagement role, providing intensive support for single parents and connecting up (and offering signposting to) a range of different service offers. Keyworkers were based in local community centres and third-sector-led childcare facilities, but specifically not in welfare-to-work agencies. In order to expand the choice of services open to single parents, a range of partner agencies were funded to provide tailored provision, including One Parent Families Scotland (OPFS), a third sector organisation run by (and offering advice for) single parents. Other partners included the local voluntary services council, a representative body that helped to co-ordinate, and signpost service users to, appropriate third sector providers.

The approach in Glasgow was more clearly built on an existing local employability infrastructure in the city – it was suggested that the size of the target group in Glasgow (which has Scotland’s largest population of single-parent families), and the existence of Local Employability Partnership agencies that had provided similar services in the past both to single parents and other disadvantaged groups, meant that MIW could be delivered most efficiently through existing networks. The local authority’s arm’s-length provider of employability services played a prominent role in co-leading the programme and delivering employability-building and keyworker support. Other partners included a third sector provider of vocational training, a local charity providing lifelong learning in one of the city’s disadvantaged communities, a childcare provider and OPFS. The programme operated city-wide through five delivery points and four community-based engagement hubs run by OPFS.

In both cities, keyworkers played an important role. These street-level workers were recruited on the basis of evidence of their capacity to provide intensive support and commitment to user empowerment. Some keyworkers had experience in delivering mainstream activation services, but others had a background working in urban development and anti-poverty projects; many were themselves single parents and/or had connections to the communities that they served.

These city-based services co-existed alongside: Local Employability Partnership provision (see above); the UK Government’s welfare conditionality and activation regime, delivered through Jobcentre Plus and the Work Programme; and other UK government-funded programmes to promote economic development (as part of ‘City Deals’). As noted above, the MLG of economic and labour market policy in the UK, as in other welfare states, has been acknowledged as a potential source of tension, and this provides an additional theme for our analysis (Kazepov et al., 2019). While MLG assumes a more ‘relaxed vertical governance’ where ‘cities and regions are increasingly expected to more self-reliant’ (Pierre, 2019: 104), we will see below that there were significant challenges in managing fundamental differences of governance and philosophy between local MIW partnerships that sought to co-produce innovative services that responded to specific opportunities and challenges in urban communities, and a national government that remained committed to NPM priorities that privilege standardisation and cost containment alongside a policy focus on work-first activation.

Collaborative governance and distributive leadership in local activation

In terms of the institutional design features posited as producing collaborative innovation, our research found consistent evidence of effective collaborative governance arrangements in both cities. First, both projects benefited from collaborative engagement with the funder. As noted above, The Fund is a non-departmental public body, and therefore enjoys a degree of independence from the direct control of central government. As a result, rather than seeking to impose a pre-ordained programme design from the top down, The Fund’s policy and funding managers embarked on an extensive programme of stakeholder engagement (including with single parents) prior to taking any decisions on the governance, resourcing or content of the programme. Drawing on insights from these discussions, The Fund and its partners in both cities agreed to eschew the payment-by-results contracting that has driven the standardisation of activation in the UK and other liberal welfare states (Fuertes et al., 2020). Instead, it was agreed that partnerships, co-led by local government or their partners and third sector organisations, would gradually bring together – and agree a modus operandi with – relevant local stakeholders. The Fund provided up-front grant funding. As partnerships were formed, roles, responsibilities and resource allocations were agreed by consensus and formalised in relatively straightforward partnership agreements.

The emphasis on collaboration and consensus-building (in defining sought outcomes and designing new responses) ensured strong buy-in from all partners. The Fund’s requirement that both cities defined a co-leadership role for the third sector meant that grassroots organisations run both by and for single parents were included. Their local knowledge ensured a voice for single parents as new programme content was developed, and added to the quality and credibility of MIW. Key stakeholders emphasised the value of a collaborative approach to funding and governance that rejected the top-down control associated with NPM and work-first activation.

The lesson is learn as you go. Don’t do a rigid client journey. MIW as a concept was allowed to naturally evolve based on lone parents, their feedback, what they want to be doing. It was not focused on outcomes, but they have happened anyway. The lesson is let go. (Edinburgh lead partner manager, year 4)

Our stakeholder interviewees also identified a number of challenges associated with the approach to collaborative governance that emerged in both cities. For example, the gradual process and pace of partnership formation was a source of frustration among some stakeholders keen to get services operating on the ground. Another challenge to collaborative governance lies in perceived power inequalities between actors, and especially the danger that public sector bodies dominate resource allocation (Cepiku, 2017). In both cases, the keyworkers, who played a central role in delivering MIW, were employed by arm’s-length organisations core-funded by local government, and there was some concern that MIW could become ‘municipalised’ (i.e. too closely aligned with existing local government services and priorities). Where there was a reliance on existing employability infrastructures this raised concerns – that were to some extent justified – that established local government services were being re-packaged, limiting the scope for including and resourcing other partners.

Finally, while key stakeholders’ evidence – and our own evaluation of the match between service users’ needs and MIW provision – suggested that both partnerships were inclusive in engaging delivery partners and tapping a range of expertise, there were some challenges in connecting with the work of other public agencies. For example, mental health problems quickly emerged as a recurring theme in discussions with single parents, and, while keyworkers were able to signpost service users to community mental health services, it proved more problematic to establish a consistent health and wellbeing component within MIW. While overall public health spending in Scotland has seen modest real-terms increases in recent years, the National Health Service has acknowledged that increasing demand for mental health services has not always been met with sufficient recruitment for community-based services, including in disadvantaged urban communities (NHS Health Scotland, 2015). These gaps in local provision were reflected in our interviews.

There was also evidence of effective distributive leadership in both cases. Distributive leadership typically involves the sharing of expertise and decision-making on resource allocation among service providers, and a collaborative approach to the leadership and co-ordination of services so that users experience a ‘seamless’ offer. In Glasgow, keyworker support, engagement and advice services provided by OPFS were co-located in employability service hubs. In Edinburgh, a deliberate attempt to co-locate keyworkers in community and childcare facilities improved MIW’s capacity to reach single parents not being served by mainstream provision. The inclusive nature of programme leadership meant that many single parents engaged with multiple MIW partners or other signposting options, taking in debt and money advice, vocational training, mental health services, personal development and access to childcare.

I think it’s probably the flexibility and the time to focus on people, and to focus on their needs and be adaptable to do it at their pace, and respond to the things that need dealing with at the time . . . I can see the impact of it. It pulls together so many different things . . . (Edinburgh keyworker, year 3)

Two features consistent across both cities were: the work undertaken by MIW partners (and supported by collaborative governance and management arrangements) to minimise ‘handovers’ and co-lead a joined-up, multi-stakeholder service; and the crucial role played by keyworkers – new frontline leadership roles that delivered innovation by connecting up different services and signposting options and providing a trusted, consistent point-of-contact. The resulting joined-up, personalised approach marked a significant break with the standardised and partial services that single parents had experienced under work-first.

You get to build a relationship with that one person instead of it being different people going through paperwork . . . it definitely helped having just one person . . . and obviously the more you go and see them, the more and more you tell them things, so having the same one really does benefit you. (Glasgow service user, year 2)

There were some initial challenges in Glasgow, where effort was required to clarify the roles of – and ensure that there was no duplication of effort across – keyworkers and other partners offering specialist advice and support. These tensions were less apparent in the Edinburgh case, where the location of keyworkers in agencies and communities where there was little by way of an existing employability offer to single parents meant that their work was clearly distinctive.

Distributive leadership requires the empowerment of frontline employees and managers to arrive at shared solutions in collaboration with a range of stakeholders (Torfing, 2016). MIW managers reported substantial autonomy in co-leading the development of service content in response to single parents’ needs, and to ‘stop doing things’ that were not working. We therefore concur with previous evidence of how collaborative governance, flexible funding and distributive leadership ‘enables local [activation] partnerships to abandon approaches which are not working and to implement alternatives’ (Orton and Green, 2019: 8). Similarly, keyworkers suggested that they felt listened to, and that feedback from the frontline regularly influenced changing programme development.

I think this is one of the best jobs I’ve ever had . . . I’ve had a lot of experience in the past . . . Quite often people will ask my opinion, and that’s quite nice for me, because I am like: ‘Oh, someone’s listening.’ Quite a lot of the things we have done is because I have said ‘Maybe we could try that’ . . . (Edinburgh keyworker, year 3)

Relatively high levels of resourcing meant that keyworkers’ caseload numbers were low, which created the time and space to listen and respond to service users, participate in shared learning across MIW partners, and engage in the ‘individual reflection’ that is important to collaborative innovation (Torfing, 2016: 162). The tenor of the support offered to single parents also marked a clear break with mainstream activation. Much UK government activation focuses on compelling claimants to enter employment quickly, with an explicit aim of reducing welfare claiming, whereas MIW keyworkers and money advice experts offered support in maximising the financial benefit that single parents were able to claim from the welfare system. Previous studies in the same areas have noted similar tensions between UK government interventions that seek to increase welfare conditionality and the ‘income maximisation’ work of local actors determined to ensure that vulnerable groups claim the benefits to which they are entitled (Bennett, 2016). Thus, while such income maximisation work marks the sort of break with the past/divergence from dominant policy priorities that is seen as a key outcome of collaborative innovation, there are clearly challenges where such fundamentally differing logics underly the approaches of local/urban and national government stakeholders in the MLG of activation (Fuertes et al., 2020). These same fundamental differences limited opportunities for collaboration and information-sharing between local MIW partnerships and Work Programme providers, undercutting any scope to develop complementary services.

Co-production for innovation with urban service users

Finally, our research found clear evidence of management and frontline practices promoting co-production among service users, keyworkers and service delivery partners. Interviews with keyworkers and other stakeholders were characterised by a consistent focus on how MIW was designed to empower single parents to make choices, define their own objectives and shape their journeys toward work.

It is really client-led. From day one when we are getting to know someone and understand, ‘What is it you would really like help with?’ We will develop an action plan that is based on what they have told. That will be modified if later . . . things come out later on sometimes. We don’t have a designated programme of ‘day one you do this, day two you do this’. (Edinburgh key worker, year 1)

The flexible funding and collaborative governance and leadership practices described above – alongside management and workplace practices that provided the time, space and autonomy for co-production – appear to have freed keyworkers and service users to co-produce innovative and personalised employability services. Keyworkers spoke of enjoying considerable scope in responding to service users’ interests, and described how single parents had shaped programme content. A number of specific institutional features and workplace practices were described as enabling user co-production, amounting to the well-defined ‘participation process’ advocated by Cepiku et al. (2020) above: the absence of prescriptive programme content imposed from the top down; the resulting, explicit autonomy for local managers and keyworkers to co-design and continuously adapt services in collaboration with users; and the extensive use of feedback loops and focus groups to capture single parents’ priorities and views on existing provision.

Crucially, every interview that we conducted with single parents – some of whom had made progress toward or into work – while others continued to face substantial barriers – produced positive reflections on co-production. Single parents spoke of the distinctive experience of MIW in offering a genuine range of activity, and how keyworkers emphasised the importance of taking control and making choices.

She never made me do anything that I didn’t want to do. It’s like she’ll talk me through things . . . but if there is a goal that I want to set it has to come from me . . . She’s more there to support us to feel supported, she doesn’t tell us to do something. It’s like we’re making the decisions and the changes and she’s just sort of being that person to lean on. (Edinburgh service user, year 4)

Survey and interview data confirmed that single parents did indeed take up a wide range of learning, wellbeing and employability-development opportunities. It is important to note that it was the collaborative governance, distributive leadership and co-production context described above that was crucial to facilitating the multi-agency services that provided personalised options for single parents. Up-front, grant-based funding, and a consensus-based approach to resource allocation and programme design prepared the ground for ‘stable and enduring processes of interaction, collaboration and creative problem-solving’ (Torfing, 2016: 148) that in turn allowed for the emergence of new ways of working and personalised solutions and services.

These positive stories of the benefits of personalised services concur with international evaluation evidence – for example, Levy-Yeyati et al.’s (2019) systematic review of activation programmes noted consistent evidence of positive impacts from personalised, client-group-specific support. Such personalised approaches also contrast sharply with UK government-funded compulsory activation. The UK Government’s (2014: 129) own evaluation of the Work Programme concluded that there was limited evidence of ‘substantive personalisation in the sense of delivering customised support services to individual participants, tailored to specific needs . . . for these participants, a standardised service was deemed sufficient and appropriate’. Research with single parents participating in both Work Programme and Jobcentre services has found that many felt that their specific needs were largely disregarded (Gingerbread, 2017), and that some reported experiences of intimidation and/or humiliation (Lindsay et al., 2019). MIW’s rejection of standardised, compulsory activation in favour of personalised, co-produced support would again appear to represent exactly the sort of distinctive break with dominant policy and service logics (to the benefit of service users) that is sought by advocates of collaborative innovation.

The urban contexts for these case studies both limited and facilitated the potential for benefits from collaborative innovation. While both service users and keyworkers placed an emphasis on the former making choices, clearly these choices were constrained by the labour market and public services context of both cities. Thus, although single parents expressed satisfaction in choosing the pace and direction of their job search, the employment outcomes achieved by most involved entry-level work in sectors such as retail and social care (the sort of low-paying service work that is key to labour demand in both cities). Significantly lower job entry rates in Glasgow were most likely a function of the weaker labour market in that urban area – as noted above, persistently high rates of worklessness in the city suggest that those facing multiple barriers are likely to continue to struggle to find employment (Bennett, 2016). Official data also suggest that a larger proportion of the jobs that are available in Glasgow are low-paid and low-skilled – for example, approximately 14% of employee jobs are in ‘elementary occupations’, double the rate in Edinburgh and much higher than the Scotland average of 11% (ONS, 2020). While it might be argued that such entry-level positions offer a first step back into work for some people, these jobs may also make demands (in terms of shifts and locations) that are difficult for single parents (Whitworth and Griggs, 2013).

Similarly, while we have highlighted that single parents chose to engage with a range of provision, signposting options were limited by the availability (or absence) of complementary local services. Local government in both cities has seen funding for services such as adult education and housing support come under pressure, and the fragility of this provision (as well as gaps in some healthcare services) sometimes constrained single parents’ choices. Crucially, the weakness of affordable and accessible childcare provision in both cities also meant that single parents’ aspirations were framed by their ability to balance work and/or learning with caring.

However, it can still be argued that both urban contexts also provided fertile ground for collaborative innovation in activation. Innovation outcomes – whether in the field of urban development or employability – are likely to be shaped by the ‘organising capacity’ within different cities (Van Den Berg et al., 2003; Van Winden, 2008) – that is, the extent to which institutional arrangements and resources and extant networks support (or hinder) opportunities for cross-sectoral and inter-agency integration. In these urban areas, the proximity of local stakeholders, which in some cases meant that there were well-established networks of trust, or at least that partnerships could be readily established; the presence of a critical mass of people to make a range of services viable; the availability of physical infrastructure resources to (co)locate services; and the profusion of service employment opportunities that lacked in some elements of job quality but were accessible and offered a degree of working-time flexibility – all of these distinctly urban characteristics in these cities contributed to an organising capacity that prepared the ground for effective collaborative innovation.