Risk governance through public sector interactive control systems: The intricacies of turning immeasurable uncertainties into manageable risks

IMPACT

Some public services, like children’s services, are subject to considerable political sensitivity and undergo regular reforms, but little is known of the micro-processes supporting macro-level regulation and policy-making. This article provides insights into this micro-level of inter-organizational relations by investigating risk governance and negotiation employed by English children’s services professionals in their attempt to make sense of the uncertainty in which they operate. Public managers may extract lessons on the role of partnership working in risk identification, measurement and governance. They could also employ the authors’ risk typology (organizational risk, professional risk and risk to clients) to manage and absorb uncertainty beyond, as well as within, organizational boundaries.

ABSTRACT

Drawing on the conceptual distinction between risk and uncertainty, the authors examine the governance of shared risk in public sector partnerships in an area governed by uncertainty. Their case study of a local safeguarding children board (LSCB) reveals differences between risk and uncertainty and ways in which negotiations of LSCB partners’ different risk conceptualizations contributes to uncertainty reduction in children’s services. The article contributes to both to an understanding of risk and uncertainty management in the public sector and to that of public sector inter-organizational management accounting and control.

Keywords:

• Child protection
• children’s services
• management control systems
• risk
• uncertainty

Introduction

While the literature on risk management is somewhat established (Alvarez et al., 2018), organizational mechanisms for dealing with uncertainty are much less clear. In this article, after clarifying the conceptual distinction between the two constructs, we explore how inter-organizational networks navigate the challenges presented by multiple sources of complexity. Specifically, we address how, through interactivity, management control systems (MCS) reduce uncertainty and transform it into risk that can be measured, and therefore managed.

To this end, we draw on Simons’ conceptualization of MCS. Simons (1990; 1991) distinguishes between diagnostic control systems, used for monitoring the achievement of pre-set performance goals, and interactive control systems, aimed at fostering organizational discussion and learning around strategic uncertainties surrounding an organization’s future activity. Consistent with Kominis and Dudau (2012) and Rossing (2013), we recognize that inter-organizational networks are to be viewed as interactive control systems established to address highly uncertain operational environments. Partnership working is increasingly prevalent in the public sector, as the complex social issues that they address require a high degree of collaboration and expertise from multiple professionals and organizations. Over the past decade, public sector partnerships have come to dominate the way in which complex societal needs are addressed, yet our understanding of the ways in which they operate is still incomplete. Given the complexity and unpredictability that underpin many social issues (Head & Alford, 2015), an understanding of how partners collectively manage risk and, perhaps more importantly, how deal with uncertainty, is paramount.

Public sector work is typically complex, often loosely defined, cross-disciplinary meta-strategies (Huxham, 2000), such as ‘crime reduction’, ‘regeneration’, or ‘child protection’. In this study, we looked into one of these governmental ‘meta-strategies’, child protection (also known as ‘safeguarding children’): an area of considerable political sensitivity. Despite concentrated efforts to avoid adverse events through, for example, legislation, policy guidance, and a performance measurement regime which is both intra and inter-organizational in nature, the long chain of serious case reviews in Britain (for example DoH, 1991; Laming, 2003; 2009) suggests that the task of co-ordinating professional and organizational efforts in a highly uncertain and unpredictable environment is a considerable management control challenge.

Our study of inter-organizational control through child protection partnerships suggests that an important aspect of this challenge is the conceptualization and subsequent management of uncertainty. Munro’s, 2011 child protection review, as well as Munro’s subsequent published research (for example Munro, 2019), concur with this view. What we add to Munro’s research is systematic evidence of how uncertainty is understood and broken down through the dialogue in which partnership members engage and through emerging sense-making. Furthermore, in line with previous studies (see Kominis & Dudau, 2012), we find that a persistent problem affecting the effectiveness of inter-organizational networks is the superimposition of diagnostic control systems, effected through a series of key performance indicators and an inspection regime enforcing them and holding organizations accountable. This is particularly problematic in areas of recognized uncertainty which are difficult to tackle due to partial and fragmented information. Interactive control systems, allowing for a dynamic and multi-perspective description of uncertainty, are more appropriate in such settings.

The contribution we seek to make is to the inter-organizational management control literature (for example Barretta and Busco, 2011; Das and Teng, 2001; Kominis & Dudau, 2012), and to that on management control in the public sector (for example Batac & Carassus, 2009; Kloot, 1997). Neither of these literature streams is well established and neither addresses the mechanisms through which uncertainty is tackled via inter-organizational settings as interactive control systems. Although several studies have considered the impact of uncertainty on the design and use of control systems (see, for example, Chenhall, 2003; Galbraith, 1973), we still lack an understanding of how such systems operate as uncertainty reduction mechanisms. We are going to address this gap here, first, by differentiating between risk and uncertainty and, then, by revealing how risk governance—rather than management—plays a central role in how uncertainty is being tackled through interactive control systems in the public sector. In doing so, we contribute to this Public Money & Management theme by examining the governance of shared risk in the context of inter-organizational networks, and by illustrating how risk management practices help MCSs reduce the uncertainty in which they operate.

Setting the scene: the relationship between risk and uncertainty

The notions of risk and uncertainty are connected through the estimated probability and the potential consequences of an event occurring. A classic study by Knight (2012), originally published in 1921, clarifies the distinction between the two: while a risky situation is one where its probability of occurring and the distribution of its outcomes are known a priori or from statistics, an uncertain situation is one whose probability and outcomes are widely unknown. On the basis of this fundamental distinction, Knight (2012) argues that, in the case of uncertain events, uncertainty is due to inadequate or unmeasurable data. This is important in a management control context where managing uncertainty has different requirements than managing risk (Mousavi & Gigerenzer, 2014; Teece et al., 2016). Moreover, this is relevant to the safeguarding children policy area, which, like other ‘wicked’ societal issues, has causes that are hard to identify and, consequently, solutions which are difficult to implement. It could even be argued that uncertainty appears to be a characteristic of complex issues in contemporary public management (Head & Alford, 2015).

If Knight (2012) distinguished between risk and uncertainty as if they were different concepts used for different realities, Gephart et al. (2009) went a step further to argue that, in fact, the two concepts could bear the same label (risk), defined differently in different eras (early and late modernity) and different schools of thought. While risk in the early modernity is seen in the classical definition of the relation between the probability of an event multiplied by the magnitude of its consequences (for example Slovic, 1992), risk in late modernity (i.e. contemporary society) refers to unquantifiable uncertainties. This position is supported by Miller (2009) who refers to modern versus post-modern conceptions of risk. The former is quantifiable and objective, whereas the latter is subjective and uncertain.

The three perspectives are certainly compatible but, for clarity of semantics and alignment with the MCS literature, we use Knight’s (2012) wording of risk and uncertainty. As such, we perceive risks to be ‘known unknowns’, in the words of the late Donald Rumsfeld, things we know we do not fully know or understand, but with some probability of occurring. Risks have unknown outcomes, but we know what the underlying outcome distribution looks like (Knight, 2012). On the other hand, uncertainties also imply unknown outcomes, but the difference is that, in contrast with risks, we know little (or nothing) either about their existence or about the underlying distribution of their outcomes. Hence, uncertainties are risks that are hard to measure—they essentially are ‘unknown unknowns’.

At the micro-level of professional decision-making and in recognition of ever-growing uncertainty in child protection, Houston and Griffiths (2000) argued for shifting away from an ‘objectivist’ and embracing a ‘subjectivist’ risk assessment paradigm. This essentially entails moving away from managing risk by means of priorly established frameworks such as performance indicators and rigid rules and regulations. What Houston and Griffiths posited, two decades ago, was governing through principles which allowed professionals the space to exercise intuition and autonomous decision-making. This was going to be challenging, as previous studies (for example Dudau et al., 2016; Dudau, 2010; Houston & Griffiths, 2000) showed professionals to be reluctant to depart from the ‘safety’ of the ‘objectivist’ risk paradigm (Houston & Griffiths, 2000), where threats are identified by governments and laid out clearly in policy guidance documents. A middle ground between the objectivist and the subjectivist paradigms could be what Brown and Osborne (2013) call ‘risk governance’, where risk subjectivities lead to ‘objective’ manifestations of risk.

Research on the Knightian concept of uncertainty has waned (Alvarez et al., 2018), with researchers preferring to address the more easily measurable concept of risk; however, little still is understood about risk and uncertainty in public service organizations (Osborne et al., 2020), despite recent calls for research (Teece et al., 2016; Alvarez et al., 2018). Management accounting literature offers useful frameworks to understand uncertainty: MCS used in conditions of high uncertainty (which we explore in more detail in the next section) and the environmental and task uncertainty typology which illuminates the source of uncertainty (Hartmann, 2000).

Environmental uncertainty ‘refers to the degree, or variability, of change that characterizes environmental activities relevant to an organization's operations’ (Ghosh & Lee Willinger, 2012, p. 90). Environmental uncertainty manifests itself when changes to the organization's environment are difficult to predict (Krishnan et al., 2016). It arises from factors external to the organization itself. In the case of the public sector, environmental uncertainty may arise from relationships with stakeholders, as well as policy and regulatory change.

Unlike environmental uncertainty, task uncertainty is an inherent job characteristic (Hartmann, 2005). Task uncertainty is defined by Galbraith (1973, pp. 36–37) as ‘the difference between the amount of information and knowledge required to perform the task and the amount of information already possessed’. Task uncertainty is directly related to the information and knowledge available to the person who is executing the task—as information increases, uncertainty decreases (Daft & Lengel, 1986). Environmental and task uncertainty are challenging at best of times, but arguably particularly so in inter-organizational contexts. MCS literature can help illuminate these issues.

MCSs and the management of risk and uncertainty

MCSs are organizational information seeking and gathering, accountability and feedback mechanisms designed to help the organization adapt timely to changes in its substantive environment (Lowe, 1971). Such changes, both structural and behavioural, are meant to reconcile and correct discrepancies between organizational strategic objectives and actual results measured and evaluated against these objectives (Lowe, 1971). This definition of MCS implies a view of the organization consistent with the one proposed by Ewusi-Mensah (1981, p. 314) that is a ‘living purposeful’ or ‘adaptively rational’ system whose survival largely depends on its ability to interact successfully with the surrounding environment and adjust accordingly to its dynamic change. From this evolutionary perspective, MCSs are the means through which information and incentives are provided to decision-makers within organizations so that they can adapt in a timely way to changes in the existing environmental conditions and thereby remain viable (Emery & Trist, 1965; Terreberry, 1968).

Simons (1990; 1991) distinguishes between two main ways in which decision-makers within organizations use MCS (Tessier & Otley, 2012) to respond to the uncertainty in their operating environment. MCS can be used diagnostically to monitor the extent to which an organization is on track to meet pre-set goals and take corrective action in case outcomes are not aligned to intended plans. MCS can also be used interactively to stimulate information gathering and organizational dialogue around strategic uncertainties, that is, ‘contingencies that could provide threats or opportunities as circumstances change’ (Simons, 1991, p. 50). Simons (1991) identifies strategic uncertainties (such as changes in customer/client preferences, partner/competitor actions, new technologies, and new government regulations) as unknown factors which need to be identified and monitored continuously to ensure the survival of the organization. These uncertainties represent threats to the organization’s current strategy, as well as opportunities for its future strategy; it is therefore imperative they are assessed and understood, if the organization is to remain viable in the face of its dynamic environment.

Interactive control systems—or any control systems used interactively—have an attention-focusing function and are meant to facilitate the generation and analysis of information about strategic uncertainties and their potential outcomes. These interactive control systems are information systems that organizations use to stir constructive dialogue among organizational members from different hierarchical levels (Simons, 1994), so that an understanding of the strategic uncertainties can be achieved, and their impact can be evaluated. As such, interactive control systems presuppose intensive, personal involvement of managerial staff and face-to-face communication (Tessier & Otley, 2012). An organization can use any MCS interactively, by shifting the emphasis away from the monitoring of the factors known to determine the success of the current strategy towards the discussion of currently unknown factors which may affect the development and success of future strategies (Widener, 2007; Tessier & Otley, 2012; Kominis & Dudau, 2012).

Simons (1994) outlined his concept of the interactive use of MCS as part of the ‘levers of control’ framework, which scholars found to be a fruitful lens through which to give focus on the strategic response to uncertainties (for example Tuomela, 2005; Widener, 2007). Specifically, some scholars (Kominis & Dudau, 2012; Rossing, 2013) adapted his concept of interactive MCS to inter-organizational settings. This newfound application of Simons’ framework has expanded it conceptually from its original function to now examining the facilitation of dialogue between members of different organizations. In a public sector context, specifically, Kominis and Dudau (2012) apply Simons’ framework to inter-organizational controls, providing evidence that the perceived level of task and environment uncertainty drives the use of a more interactive model of control.

Management control systems literature proposes that interactive control systems are more likely to be introduced when organizations face higher uncertainty (Kominis & Dudau, 2012; Dudau et al., 2020; Corduneanu & Lebec, 2020). It is therefore surprising that the empirical research on the use of interactive MCS in inter-organizational settings is scarce—particularly given the multiple possible sources of uncertainty in these contexts, especially in the public sector (Head & Alford, 2015). In addition, while the role of MCSs in uncertain environments seems recognized, we have yet to understand how they go about reducing uncertainty. Our study investigated that through an empirical case study of an established child protection partnership.

Methodology

Our empirical evidence comes from a case study of an English Local Safeguarding Children Board (LSCB), in ‘Brigam’, in the North West of England. LSCBs are mandatory partnerships introduced in 2006 (through the Children’s Act 2004) bringing together mid management level professionals involved in child protection at the local government level, under a statutory duty to collaborate under the leadership of the children’s services authority.

The core LSCB agencies are social services (children’s services, i.e. social care and adult social services), education, the health authorities, an organization offering employment advice to young people, the Child and Family Court Advisory and Support Service, local probation boards, youth offending teams, local prisons and the police. In addition to these, there are the statutory lead member for children’s services (a local councillor who holds political accountability for children’s services including education and social care) and a lay member (a member of the local community with personal experience in this area), as well as voluntary and community sector organizations which may choose to get involved but are not under a statutory obligation to do so. The partnership encompasses a variety of professionals with different backgrounds and responsibilities. To ensure co-ordination, the interaction among network members is led by the Children Services Authority (the director of which holds professional accountability for service quality). LSCBs have been conceptualized as interactive control systems (Kominis & Dudau, 2012) in that they are formal, yet flexible, structures through which intelligence about the child protection environment is gathered. The ‘controller’ is the children’s services authority and the lead member and they respond dynamically to intelligence coming from middle managers in each of the organizational members who sit in the LSCB and make explicit to the other board members the subtle changes in the environment which are likely to affect their work and the LSCB mission.

The in-depth single case study design (Yin, 2003) selected is appropriate given the relatively early stage of the management control literature on public sector inter-organizational relations (Eisenhardt, 1989) and highly suitable to gaining understanding of context rich data (Otley & Berry, 1994) on risk and uncertainty in public sector partnerships. Our evidence comes from nine in-depth semi-structured interviews with LSCB partners. Our interviewees were an independent chair, the director of children’s services, the assistant director of children’s services, the business manager (professionally a social worker, having an administrative leadership position in the board), the prison governor, the lead member, a lay member a hospital trust representative and the police representative on the Brigam LSCB.

Our interview protocol was wide in scope, aiming primarily at an understanding of the role of the LSCBs and of how they tackle child protection, with all the uncertainties surrounding this policy area, coupled with the political sensitivity it triggers. Our interviewees were encouraged to discuss issues relevant to our research using their own terms and vocabulary. When required, the interviewer asked for elucidations on or illustrations of the issues under discussion (Byrne, 2004). Interviews were audio-recorded and lasted between 30 and 120 minutes.

Our data analysis entailed a thematic analysis of interview transcripts. Specifically, we adapted the approach to thematic analysis proposed by Braun and Clarke (2006). We gained familiarity with the data by transcribing our audio-recordings verbatim and by reading and annotating the resulting transcripts. We generated an initial set of codes from our theoretical underpinnings and complemented it with additional codes emerging from interview materials. Given the possibility that our interviewees could refer to ‘risk’ or ‘uncertainty’ almost interchangeably in colloquial speech, distinguishing between these two constructs proved critical for our analysis. For this reason, we considered references to issues of ‘controllability’ and ‘measurability’ in interviews as salient theoretical features allowing us to discriminate between the constructs. Codes were subsequently aggregated into themes via a process of progressive refinement and review. The presentation of our findings is organized around a set of themes resulting from this iterative process.

Findings

Our case study on inter-organizational management control allowed us some insights into the way environmental and task uncertainty are perceived in the policy area of child protection. We explored the ways in which this uncertainty is conceptualized and understood, and examined the risk governance mechanisms that enable the management of this uncertainty.

Environmental and task uncertainty

Official reports into child protection explicitly recognize uncertainty and the need to learn to work with it:

Child protection work involves working with uncertainty: we cannot know for sure what is going on in families; we cannot be sure that improvements in family circumstances will last. In fact, many of the problems in current practice seem to arise from the defensive ways in which professionals are expected to manage that uncertainty in their everyday job. For some, following rules and being compliant can appear less risky than carrying the personal responsibility for exercising judgment (Munro. 2011, p. 6).

Our interviewees, however, do not think the uncertainty in their work environment is necessarily inherent to their line of work, nor that it goes hand in hand with environmental complexity in the policy area. Rather, they think that perceptions of uncertainty are heightened by political sensitivity around systemic failure and the subsequent volatility that, in the words of one of our interviewees, ‘can bring down governments’. Perhaps by way of consequence, governments are thought to react quickly to failure in this area, introducing more changes than other policy areas have seen:

[This policy area is] more liable to change than others, because government panics about safeguarding issues. (Police representative.)

Yet public servants appear not to be equipped to adapt quickly to policy change—indeed, many are conservative in their decision-making and action taking:

But what we’re not that good at, I don’t think, is amalgamating what those themes [emerging from serious case reviews] are and then responding, and I don’t mean this locally particularly, but nationally, it’s difficult to identify what the key things [to do] are. (Assistant director of children’s services.)

This particular quotation also highlights that the environmental uncertainty of a fluid policy field does not translate easily into manageable tasks.

Another issue contributing to the perception of environmental uncertainty is the linear, black-and-white findings of serious case reviews—official inquiries into child protection failures from which there are lessons to be learned:

Sometimes these reviews don’t help, you know … public inquiries and things like that, because they’re very black and white, aren’t they [they highlight the consequences] and that is damaging, I think, because everyone becomes frightened then and because they’re frightened, they become more risk averse, so then … I think they over-exaggerate every issue. (Police representative.)

The discrepancy between the uncertainty felt in one’s work on child protection and the simple way in which it is conceptualized in serious case reviews can generate fear that uncertainty is perceived rather than real and that what might appear as uncertainty is in fact low competence levels. With this comes risk aversion accompanied by administrative burden. In exaggerating the issues behind the cases they encounter and the consequences of their decisions, professionals come to feel the burden on decision-making despite being aware of the meta-task of their mission and what they are supposed to achieve. The paralysis can occur when faced with the uncertainty of ‘how’ to achieve their joint mission:

I think there’s kind of meta certainty about what you’ve got to try and do, but the how of, how you go about doing it, isn’t prescribed at all … So the uncertainty is about how you achieve it rather than about what you achieve, (Director of children’s services.)

This is essentially about the way environmental uncertainty in reflected in task uncertainty, that is, the level of task programmability and process standardization that professionals experience in the way that they do their job, and the procedures that are (or not) in place to guide them in doing so. In their work, professionals have learned to work with policy, as well as with guidelines and procedures, requiring different degrees of compliance. They seek to match real life cases with the pathways dictated by these work procedures and guidelines. This can be a challenging task, as often the match is imperfect and substantial discretion is required to ‘fill in the blanks’, to navigate through the ‘grey’ areas not specifically foreseen by legislation or guidelines:

[Policies and guidelines] are talking about black and white and there’s no grey, but [actually] there’s a mass of grey in between. Actually what you’ve really got is a very thin white line and a very thin black line and a massive grey in the middle, and that’s where really from a police perspective, we’re operating in that grey. Yes, we’ve got the law which clearly says black and white but when we’re talking about procedure and process … that’s where you end up in a lot of this grey area. (Police representative.)

Police is not the only agency operating in that ‘grey zone’—health and social services are also mentioned in this light:

Health is quite a complex area. You have providers of health visitors, district nurses, drop-in centres, you have providers of critical care. Not everybody is aware what each of them provide, even now because it’s quite a complex arrangement, unnecessarily so sometimes you think. (LSCB business manager.)

Social services … will have all their processes … around safeguarding but the difficulty is that quite often they’re dealing with people with complex problems … So they need to have a process as a guideline but they effectively need to step off either way, and … when … they come and beat you up because you’ve not done something by the process line, it’s because quite often the process is too rigid. (Police representative.)

The task becomes all the more complex when the disparate lines of action for the different agencies and professions are integrated into a jigsaw of collective and shared responsibility in a policy environment. Seen from the side-lines, for example by a partnership member who is not in any of the professions and organizations represented at the board, the bare anxiety of coupling the different sets of organizational controls into effective inter-organizational ones removes the professionals’ ability to tackle the problems encountered in their work environment, heightening their complexity—essentially creating a situation of self-generated (task) uncertainty:

I sometimes get the feeling that they’ve been trained in a particular way to say, right, in this circumstance, these are the tasks and rules that you follow, and whether that is ingrained deeply in them, that actually they feel uncertainty where somebody like me, who’s looking from the outside, wouldn’t see it, so I might look and go, well, that’s just common sense, that’s just the sensible thing to do, but they would see it far more as, yes, but what about X, Y and Z, and I should be following this because this threshold hasn’t been reached and I have to do this, this and this. So I suppose I would see the uncertainty more as being kind of self-generated by people. (Lay member.)

Some of the professionals recognize this phenomenon, too:

At the moment, each of the partners will have their own set of policies and procedures and you effectively need to almost rip all of those up and form new [joint] policies and procedures that all partners buy into, and this is where the rub comes about the sharing of information because at the moment we’ve all got our own policies and procedures on sharing of information and if they’re that one degree out, as you get further down the line, you’re further apart, aren’t you? (Police representative.)

What these interviewees suggest is that some of the uncertainty experienced by professionals during the board exchanges on inter-professional and inter-organizational dependencies is due to organizational and professional baggage. Specifically, professionals viewing situations through the lens of formal policies and procedures can render them unable to see simplicity in real-life problems, and thus fail to offer common-sense solutions to them.

From uncertainty to risk management

To start with, our interviewees seem to conceptually distinguish uncertainty from risk. They referred to the two in different contexts and prompted by different interview questions. LSCB professionals’ understanding of these issues seemed to converge on uncertainty being external to them (mainly environmental uncertainty). Then, the challenges encountered in acting upon that uncertainty is arguably the realm of task uncertainty, whereas the provisions in place to enable action is seen as risk management. The intersection between the realization of real-life complexity and the need to take action, which may or may not match organizational guidelines (i.e. ‘objectified’ risk), is well illustrated in this quotation:

You’re trying to get to the end, whatever the problem might be, and [there is a] path that you take to get there. Now, if you start to face problems on that path, which your policies and procedures don’t necessarily address, they don’t give you the answers, you’ve very much got to rely on trust, that you’re going to do the right thing to try and get to there. Now sometimes we end up doing the wrong thing but for the right reasons. (Police representative.)

Therefore, when there is misalignment between the uncertain environment and the professionals’ standard responses (arguably themselves the result of objectivist risk management processes), there is a need for flexibility in re-interpreting these responses and applying them in different ways than perhaps originally intended (i.e. engaging subjectively with risk). Interpersonal trust seems to be instrumental to such flexibility:

You’ve got to have flexibility to make things work especially in a partnership. You’ve got to have that flexibility, otherwise it just doesn’t work because there are too many barriers clashing … Trust is absolutely vital. You’ve got to trust. (Police representative.)

Further down the line of flexibility comes a reflection on risk and negotiation of risk assumptions across the board membership:

I think sometimes there’s a bit of a kind of … people will not go all of the way because if you go all of the way and you take full responsibility, you’re bringing risk to yourself and to your organization. (Lay member.)

All the different agencies are expected to bring their files to be shared on this day and everybody to go through and look, did we have the same information at the same time? Did we not? Was there pieces of information missing that weren’t shared? Why was that? Why did we not share that, if you had it why did we not know about it? … So it’s little things like that that you need to know; just little things that make a huge difference. (Lead member.)

If risk is considered when making sense of uncertainty, it is important to see how LSCB members transition from uncertainty to risk. One of the interviewees offered an example of it happening when uncertainty becomes certain when something goes wrong (for example when a child dies) and when risk factors are identified a posteriori:

There may be new, there may be different presentations and each case is different and it may be that there are added factors and added layers to getting to what the key issues are … So, it’s often where you don’t look for it that child protection and serious safeguarding issues are found … you can’t be in the job of managing risk and expect things not to go wrong. So, when things go wrong, I will identify and work to look at … what is it and why is it that things have gone wrong and how can we learn from it and try and address what we’re doing going forward to make sure that we reduce the risk of that happening again … that’s what we do, we manage risk, so to think that we won’t get things wrong is … naïve. (Director of children’s services.)

At an intra-organizational—rather than inter-organizational—level, LSCB members seem to face uncertainty, learn from it when failings occur and turn those lessons into risk management processes. But we are yet to see how risk is managed through interactive control systems, our child protection partnerships. To address that, it is important to start by establishing what is understood by ‘risk’ in our LSCB. Three conceptual directions emerged: risk to clients, risk to organizations (risk as organizational liability) and risk to the professional involved (risk as personal liability). The first of the three occurred only once in our interviews, but in powerful terms:

I was dealing with a lot of very serious issues where it was threats to life … and you had to make snap decisions about what actually you were going to do and the consequences of getting it wrong were someone seriously injured or even worse, and the problem comes there, and that’s the real risk. (Police representative.)

The ‘real’, underlying, risk to clients is often overshadowed by the anxiety caused by considering organizational and personal liability when doing ‘the wrong thing for the right reason’—the ‘wrong’ being stepping away from the black and white lines or legislation and guidelines and into the ‘grey’ of ambiguous solutions to complex problems. Both types of liability are illustrated well by the following quotation:

There’s a duty to co-operate, but co-operating, I think, in many people’s minds means talking to each other, listening to each other, but it doesn’t necessarily mean saying, okay, well, I’ll take responsibility for this part even though actually it might fall within your organization … people will not go all of the way because if you go all of the way and you take full responsibility, you’re bringing risk to yourself and to your organization. (Lay member.)

What this quotation also suggests is, again, the dichotomy between uncertainty and risk. Understanding the unknown unknowns associated with uncertainty, intelligence gathering is key and this is what is achieved through talking and listening to partners in partnerships like the LSCBs. When these unknowns become known, through the connectivity between the different viewpoints in inter-organizational settings, they become actionable, which becomes the realm of risk and risk management. Whether or not the risks considered are to clients, organizations or individuals themselves, this risk management culture seems to push decision-makers to:

 … over-exaggerate every issue. They make it more serious than … it necessarily is, so you end up doing twice, three times, four times as much work. They almost start to try and look down every single offshoot for what are the risks and what are the problems and then come up with a solution for all of those problems. When we’ve not even gone down that route. Because they’re risk averse, they’re trying to account for every eventuality, so if there are five different routes that you could go with something, they will try and account for all five routes, just in case. (Police representative.)

This depicts attempts to deal with uncertainty by ‘over-contracting’ through risk identification, measurement and management. In this sense, risk aversion of the type described by our interviewee does not seem to be sustainable in the long term in an inter-organizational partnership setting. If nothing else, it casts a shadow of doubt over the compatibility of the three risk management process types pertaining to risk to clients, to organizations and to employees.

Risk management or risk governance?

If LSCB representatives observe the uncertainty in the child protection policy context, communicate and gather intelligence to the point of risk identification, measurement and management, the issue remaining is exploring how this happens and the answer which emerges is that it is through risk negotiation. One example of risk negotiation is in terms of risk thresholds in the board:

What you could see when people were discussing it is that there was still a kind of pushback to the council to take the lead on doing it. But everybody round that table needed it and it was [supposed to be] both ways. [Brigam] Council didn’t necessarily … want to fully let it go, it was their responsibility, but people didn’t want to fully take it off them, because they didn’t want to take the responsibility. (Lay member.)

Hence, while from a risk to client perspective, more than one organizational representative felt they could contribute to the work, the risk to organizations mindset appeared to over-rule that sense of responsibility towards the clients, in the process of risk negotiations among the organizations. The picture of risk governance in inter-organizational settings is further enriched by an example of how the use of jargon indicating different risk thresholds for different organizations. The example offered by one of our interviewees was that of ‘high risk’:

To the police, high risk missing from home means you drop everything else and all you’re focusing on is that trying to find that one individual person, whether it’s an adult, whether it’s a child, it doesn’t matter. If they’re deemed high risk, missing from home, it’s because there is an immediate threat to their life, so it might be somebody with mental health issues who’s intimated that they’re going to go and kill themselves. That would be high risk. (Police representative.)

So when one of the partner agencies, typically health or social services, contacts the police using the ‘high risk’ label for one of their children, the assumption, on the side of the police, is that there is immediate threat to life. In reality, however, this is not necessarily the meaning held by their partners. What their partners often mean is ‘very serious situation’, whether or not the threat is to one’s life. They may be using this particular jargon in negligence, ignorance, or ‘to build in a defence line [in case] something goes horrible’, to be content that they alerted the police with the highest possible urgency so they did the absolute best they could. This last explanation taps onto the blame culture, which often becomes apparent in the aftermath of adverse events and which tends to affect social work disproportionately, due to their statutory duties which other professionals do not share in the eye of the law:

There’s a triangle of agencies who get the beating every time something goes wrong. It’s the police, it’s the health, it’s social services, so that triangle of services, they’re the ones who always come under scrutiny when something’s gone wrong … we’re very much process driven, police and health. Social services, well … they will have all their processes … around safeguarding but the difficulty is that quite often they’re dealing with people with complex problems … So they need to have a process as a guideline but they effectively need to step off either way … because quite often the process is too rigid. (Police representative.)

This could be one of the problems faced by LSCBs—as interactive control systems intended to deal with uncertainty—namely, the existing culture of accountability, which requires specific processes and procedures in a context if high uncertainty. Also:

I sometimes think that … people use policies and procedures as a prevention of doing something as well. So the sharing of information, so because someone said under data protection you can’t share personal information, it’s against their data protection, or it’s against their human rights, they’ll immediately latch on to that as being an excuse why they’re then not sharing information and so they’re going on a very fine narrow line. (Police representative.)

This could be interpreted as a consequence of the lingering culture of accountability and managing by exception. In other words, this control culture could be the reason why LSCBs do not fully fulfil their potential and purpose as uncertainty reduction mechanisms. If people perceive that they are under constant monitoring and control, they are more likely to stick to the procedure/rule and not attempt to ‘go the extra mile’.

Discussion

Our study revealed how environmental uncertainty in child protection, partly responsible for the task uncertainty experienced by LSCB members, prompts risk negotiation on three levels: risk to clients, organizational risk and individual risk, organizational risk being the main concern in those negotiations, yet the risk to clients being most mentioned in the negotiations. In our case study ‘Brigam’ LSCB, the aim of risk negotiation was to reduce uncertainty.

Our research makes two key contributions to the literature on uncertainty. First, we extend the work of Chenhall (2003) and Galbraith (1973), who considered the impact of uncertainty on the design and use of control systems but did not address the issue of how such systems go about reducing/absorbing uncertainty. Second, we contribute to enhancing the limited understanding of the links between environmental and task uncertainty, a less researched topic in the area of uncertainty (Kominis & Dudau, 2012). Our LSCB interviewees elude to a process of turning task uncertainty into risk governance.

The study also contributes to the limited inter-organizational management control literature (for example Barretta and Busco, 2011, Das and Teng, 2001; Kominis & Dudau, 2012; Bukh & Svanholt, 2020), which had not conceptualized inter-organizational settings as uncertainty-reducing mechanisms, although some did regard partnerships as interactive control systems (for example Kominis & Dudau, 2012). To this literature area, we bring evidence of the mechanisms through which partners appear, through dialogue with others, to identify known risks (transitioning from being known to others in the network to being known to themselves) and negotiating ownership of it (whose risk is it, the clients’, the organization or a personal liability) with the other LSCB partners, in order to, then, manage it. We did not find evidence of inter-organizational risk management per se, unless we concede that management encompasses the notion of governance, where identified risks get shared through negotiation among the partnership members. Indeed, risk governance appears to be the central mechanism through which interactive MCSs fulfil their uncertainty reduction role.

By illustrating how risk management practices link to public sector interactive control systems, we contribute to a third strand of literature: that on risk management in public sector organizations (for example Brown & Osborne, 2013; Osborne et al., 2020; Warren, 2019) and on risk management through management accounting (for example Rana et al., 2019a, 2019b). We build on the argument suggesting that risk cannot and should not always be minimised (Brown & Osborne, 2013) but that it can be shared in shared services (Rana et al., 2019b). We then extend these theoretical arguments by illustrating with empirical data and by offering an understanding of the processes through which risk management processes break down emergent uncertainty.

Conclusion

The domain in which public sector organizations operate is inevitably complex (Munro, 2011, p. 2019), making partnership working ubiquitous. In this article, we have examined a child protection partnership—the Brigam LSCB—as an interactive system of management control characterised by the facilitation of dialogue between multiple parties, enabling the diffusion of knowledge and ideas (Simons, 1990).

The use of a case study allows for a detailed account of the functioning of an interactive management control system from multiple perspectives within that system. However, as this is a single-case study, caution should be exercised when generalising from our findings. Rather, our conclusions are tied to the context in which they were found, and more studies of other policy and organizational contexts are needed to support and extend our claims.

Our findings detail the task and environmental uncertainty faced by Brigam LSCB and the way these are perceived by professionals within it, suggesting that interactive control systems are an effective method of dealing with the uncertainty in a public sector setting. We unravel the mechanisms through which interactive systems of management deal with uncertainty by turning the unknown into manageable risk. Building from past research, a key contribution of this article is suggesting how interactive MCS in public sector settings reduce uncertainty: by turning it into risk, through risk governance.

There are several practical implications of these findings. To start with, in policy areas reliant on partnership working, yet with little effectiveness of these collaborations due to stubbornly constant poor outcomes (for example child abuse and child homicide rates in the policy area LSCBs operate), it is useful to see how inter-organizational structures work and how they derive their concepts of risk from sheer uncertainty. It may be that the focus is placed on the ‘wrong’ risk dimensions, whereas others may be overlooked. Indeed, our study provides a useful typology of risk (risk to clients, professional/personal risk and organizational risk) which may be useful to risk management assessors in such policy areas, whether at an intra- or an inter-organizational level. We also provide a hierarchy of these risks, with the risk to clients being the most useful driver of risk negotiations around the other two risk types, despite risk to organizations being a real concern for partnership members. Good practice might be to only ‘objectify’ risk to clients, leaving the other two to subjective interpretation, so that they do not drive operational work. This may be very different from how children safeguarding organizations work, many of their policies and procedures having an organizational, rather than client-specific, focus. While this is understandable, a balance can and should be aimed for across the three risk dimensions, with the one to clients leading the metrics. Munro’s, 2011 comprehensive child protection review suggests that this is a plausible and much needed direction in child protection work.

Acknowledgements

This study was supported by the Chartered Institute of Management Accountants and the Adam Smith Research Foundation. We would also like to take the opportunity to thank the anonymous reviewers and the guest editors of this Public Money & Management theme.

Disclosure statement

No potential conflict of interest was reported by the author(s).