Managing difficulty in social service systems – implications for future system design

Summary

Systems thinking has gained traction in design-thinking in human and social service systems over the past few years.? However, the application of this thinking has not always considered inherent system difficulty in the system design.

The impact and distribution of difficulty has significant implications in the shift towards user-centered design. This shift has occurred following a welcome and increased appreciation that the intended beneficiaries of a service must be considered as part of the system.

Irrespective of the shift in thinking, caution is required to ensure that intended beneficiaries of revised thinking and practice in relation to system design are not disadvantaged by the systems intended to create user advantage and benefit.

It is not sufficient to merely re-design systems around the beneficiary without addressing existing system behaviours, because this risks making beneficiaries worse-off.

Many current systems are characterised by concentrations of power, and the conditions for difficulty to be shifted from well-equipped and resourced system actors to less well-equipped and resourced ones.

The maladaptive behaviours of concentration of power and shifting of difficulty which lead to system failure too often convert into shifting of blame for failure onto beneficiaries.

These existing system behaviours and associated risks are not conducive to the effective development of human-centered design processes and unless they are addressed the desired quantum shift in system performance required to address wicked problems will not be achievable.

Introduction

In recent years systems thinking which focuses on understanding how a system’s outputs are created by the way in which formal and informal system actors and their resources interact and connect has increasingly been applied to human and social service systems. The principles of systems thinking provide a useful way of analysing the systems of social service delivery and understanding how those systems add value to the intended beneficiaries.

More recently in service design of social services there has been a shift towards principles of user-centered design which has included a more explicit recognition that the intended beneficiaries of a service are within the boundaries of a given system, rather than sitting outside it. This recognition that system beneficiaries are part of the system has allowed a greater focus on their role in the system and how to design systems from the beneficiaries’ perspective.

However, this shift whilst noble in its intent risks disadvantaging intended beneficiaries if existing system behaviour is not explicitly addressed and the concentration of power in systems is not shifted closer to beneficiaries and those beneficiaries are not equipped to harness the power for their own benefit.

For the purposes of this paper, I have defined below terms to describe the various actors in a social service system which features commissioning of services by governments, regulatory oversight by a formal body, service delivery by non-government organisations (NGOs) for-profit or otherwise and intended system beneficiaries.

Systems which are the focus of this paper feature formal actors which are generally organisations with power or the capability of control in the given system such as commissioning agencies of government regulatory bodies and contracted service delivery organisations. In addition, informal actors, such as local community organisations which have been enrolled into the system rather than commissioned or contracted into it, and recipients. Recipients (also called clients, consumers, participants etc.) without a controlling capability or a formal relationship are also informal actors, and the presumed beneficiaries of the system outputs.

System Difficulty

When the boundaries of a system are not well-defined, or a system has emerged rather than being designed the system is particularly vulnerable to system difficulty. What is system difficulty? System difficulty is the accumulation of organisational process debt, process misalignment between organisations, IT systems misalignment between organisations, professional misalignment of individuals, lack of role clarity, restricted flows of information or resources and failure to identify what is within the system boundaries. ?

This system difficulty profoundly affects the system’s output and can be so significant that it adds inefficiency to such an extent that the system becomes inoperable without considerable support, usually via the diversion of resources from formal actors. This inoperability also results in the system then being ‘hacked’ through necessity by its actors with work-arounds, short-cuts and informal and unrecorded practices which introduce significant risks to the ability to monitor or report on system activities and outputs. This system hacking also allows system actors to hack the system to their advantage. In some cases, recipients have to hack the system (sometimes with the support of existing system actors or recipient advocates) in order to receive the service the system is supposed to be delivering. Just as in digital systems system, hacking of social service systems makes the system vulnerable to penetration and manipulation by external actors. A common response to system hacking is for new control processes to be introduced by one or more of the formal system actors, which paradoxically adds more process debt and complexity to the system, further reducing its effectiveness and requiring further hacks and work-arounds which increase vulnerabilities and control responses. And so it goes on.

System Simplicity

Whilst the simplest system is the goal, system simplicity is not the opposite of system difficulty. What is system simplicity? System simplicity is the simplest viable (where viability includes controls and governance which minimise risk to all actors) system which efficiently delivers outputs to the intended recipients. System compliance (governance, regulation and privacy) are all features of a system which are intended to keep system actors and recipients safe and can’t be dispensed with. The important thing is to make those features as straightforward as possible, without diluting their effectiveness. For example, making demonstrating compliance with system safeguards as straightforward as possible to increase compliance and the measurability of compliance enhances the safety of system actors and avoids obscuring or disguising risk to system actors.

In any system (particularly one delivering social services) a certain amount of difficulty is inherent, human lives are complex and an individual client has free will and autonomy, clients as well as staff choose to not follow instructions about how to engage with the system, even when these instructions are clear. In addition, the very nature of social services is that they are delivered by different organisations with different purposes and different processes which may not be compatible with other systems and where the prospect of inter-organisational compatibility is low. This is not to say that ??such services should be delivered from one department of government or one NGO, since large bureaucratic agencies frequently exhibit all of the shortcomings associated with system difficulty.

All systems feature a certain amount of difficulty which needs to be managed in order that the outputs of that system are delivered as efficiently as possible. I have observed, particularly in government and non-government delivery of social services a tendency to shift difficulty from one system actor to another in a way which is advantageous to the system controllers (e.g. commissioning agencies) or managers and disadvantageous to other participants in the system. This can happen because the boundaries of the full system are not adequately understood, or because of a hierarchy which allows system controllers to arrange matters for their own advantage at the expense of others. The location of power in the system has a strong correlation to how difficulty is distributed. In general terms formal actors such as system controllers or managers have more power than other system participants and use that power to their advantage, but not always to the advantage of the whole system and sometimes to its cost.

System mapping

To understand system difficulty, it is necessary to understand the system and the first stage in understanding a system is to map it. This is a step which is prone to inbuilt limitations, especially when existing system controllers or formal actors pre-determine the scope of the mapping, or the intended outcome.

The goal should be to identify all the formal and informal system actors, even unintended ones not considered at the original design phase, or which have been added or co-opted into the system over time as a precursor to designing the simplest viable system to deliver the intended outputs. Once the simplest viable system has been identified the task should be to agree which system actor has what role in the system, and what the key relationships are between the actors. If, for example, the relationship is primarily one of exchange of data (such as between a contractor and a regulator) compatibility of data type and rules of data management must also be addressed.

Formal and informal system actors and awareness of the system

Formal actors know that they are in a system which has a series of relationships with other system actors. These relationships may be highly regulated and managed, and indeed the way the system is configured may be part of that regulation and management. In short, these relationships are formal and confer status in the system.

This status also confers power. Whilst formal actors may (and frequently do) argue about how power is distributed, it appears self-evident that actors which are bound to the system in any commissioning, contracted or regulatory way share the majority of system power between them.

Informal actors including direct or indirect recipients of the system output have little formal power. They may have informal power which arises because of system hacking, but such power is not endorsed by formal actors. In addition, formal actors will move quickly, either collectively or individually, to eliminate informal power or capture it to return it to the formal actors.

Due to incomplete system visibility by informal actors they are frequently unaware of the extent or complexity of the system they are part of.? The diagram below illustrates examples of the relationships of different system actors and their status in a simple system. In this case formal actors are in the blue boxes and the formal relationships are in green. The informal activity is identified by the figures with dotted-line boundaries.

In system mapping too often the system managers and deliverers of the service fail to identify that recipients of the service are part of the system not just a recipient of the system’s intended outputs. By doing this any initiatives aimed at increasing efficiency or productivity of the system can result in the shifting of difficulty away from the formal to the informal parts of the system. Shifting difficulty over the fence, however, does not shift difficulty out of the system, it merely redistributes it from one part to another. In fact, shifting difficulty from formal, skilled/professional parts of the system to informal actors is more likely to add overall inefficiency in the given system given that skilled professionals (formal actors) can reasonably be expected to be better resourced and equipped to manage system difficulty than informal actors.

The diagram above illustrates what can happen when difficulty is shifted around a system by formal actors (blue) onto informal actors (orange). At each stage the formal actors seek to reduce or minimise their own difficulty by shifting it to another part of the system without adequate consideration of the effectiveness of the whole system. This behaviour frequently shifts difficulty to the parts of the system least well equipped to deal with it.

Impact on recipients of shifting difficulty

This shifting of difficulty around, not out of the system and on to the least well-equipped and resourced system actors conspires to reduce overall system efficiency and degrade the quality of the system outputs. Since the system outputs/services are received by the recipient, as the last receiver, the recipient is then vulnerable to the criticism that they are failing, rather than that the system as a whole is failing. Victim-blaming is commonly used to relieve the formal actors in a system of responsibility for the system’s shortcomings and failures. ?In these cases, the flow of difficulty provides a conduit for the flow of blame for failure.

In cases where a system is malfunctioning (such as the NDIS) it is notable that much of the public discourse has been about rorts and fraud undertaken by recipients (participants in NDIS language) and recipient sourced service providers (who have penetrated the system due to inadequate controls, see reference to system hacking above). This focus obscures the difficulty experienced by legitimate participants accessing the packages they are (or might be or were previously) entitled to. This emphasis on the apparent mendacity of the recipient and the actual fraud committed by the system’s bad actors shifts responsibility for failure away from the system’s formal actors and is used to evade legitimate critiques of system design.

This failure of system design and governance can be argued to have come about, in part, as a result of shifting ‘choice’ on to recipients, but not equipping them to manage the choices available to them and relying too heavily on trust as a risk control mechanism. These system design failures materialise in the recipient, and it is then convenient to be able to blame the recipient for the failure of the system.

System behaviours and the implications for human-centered design

Existing systems (as illustrated above) are characterised by the vesting of significant power and control with the formal actors and limited power and control in the informal actors. These system characteristics do not lend themselves easily to a shift to user-centred design because frequently it is the formal actors who conceive and develop user-centred design on behalf of recipients, a design practice which reinforces the concentration of power in the system.

When the formal actors in a given system approach the idea of ‘putting the client at the centre’ the result can be something that looks like the diagram below:

In this example the various formal actors (in blue) in the system are collaborating (in green) with the intention of allowing the recipient to make the most of the system’s resources and receive the right outputs and achieve the best outcomes as a result.

This well-intentioned design creates the conditions for another problem to emerge where the aspiration to work together is vulnerable to a system hack. In this case the hack being that formal actors conspire (tacitly or otherwise) to move difficulty onto the recipient.

In fact, it may be that a poorly designed ‘client centred’ approach makes recipients worse off. As well-intentioned collaboration amongst formal actors increases so does their ability to conspire to transfer difficulty to the participant in a more sophisticated way.

From the recipient’s perspective they are now surrounded by the formal actors in the system paradoxically are more vulnerable to having system difficulty shifted onto them as follows:

In Fig 4 above the formal actors are collaborating but in doing so are shifting system difficulty directly to the recipient (in red).

Collaboration, conspiracy and complicity.

The conditions for collaboration are also the conditions that allow formal actors, irrespective of what they aspired to do, to conspire to shift difficulty away from themselves onto recipients. When we consider that shifting system difficulty onto informal actors leads to system failure, the question is, at what point have formal actors become complicit in the system failure? Formal actors are, as I have described, the professional government, regulators and NGOs in the system. They are the system designers and controllers. Is it acceptable for these actors to advance a plausible deniability argument that they were unaware the system was (or had) failed? More profoundly (because admissions of system failure do occur) how can formal system actors admit to system failure without accepting that they were complicit in the failure and take responsibility for the resolution rather than blaming the recipient?

The Robodebt Royal Commission identified multiple process and system failures and yet no-one has acknowledged complicity in the failures and, thus far it seems, no-one has taken responsibility.

Power and difficulty

Another feature in system behaviour to be wary of is the shifting of difficulty disguised as re-distribution of power in the system. Remember that shifting tasks from formal to informal actors can increase overall system difficulty, something which is especially the case when tasks are shifted to actors in the system who are unaccustomed to holding the power to make decisions associated with the tasks they have been given.

Take the apparently simple concept of user choice, a principle asked for by system recipients and the means to achieve it designed by system managers from their own perspectives. If you are used to making choices in your personal and professional life, or if those choices are governed by clearly defined frameworks, or if you have received specific training, you may fail to take account of them when designing choice for others, particularly people who are not accustomed or skilled in exercising it who have become so disempowered and system dependent that they don’t want choice or the burden of choice.

This example illustrates how the scale of difficulty is magnified by moving it from formal to informal actors in the system, because informal actors are less adept at undertaking tasks they are unaccustomed to, or not trained for. As a result, overall system difficulty increases at the same time as difficulty for formal actors decreases.

In such cases it is necessary for formal actors to quantify difficulty and the magnifying effect of transfer of tasks to informal actors as a component of system design. This does not mean that user choice for example can never be transferred to informal actors, but that sufficient effort should be put into equipping the receiving actors with the ability to exercise their new responsibilities.

Relinquishing power as a precursor to client-centred design

System designers and controllers can be reluctant to shift their power to other parts of the system, let alone to informal actors. What is frequently not understood or adequately considered in moving to a ‘client centred’ aspiration is the loss of power that formal actors must experience in order for recipients to have more power in the system. The formal actors in the system have strong vested interests in knowing exactly where the power is, which actor holds it and how they will exercise it. This is what enables formal actors to shift blame of system failure to the less powerful parts of the system.? For true client-centred approaches to system design to work, the first thing that has to happen is for the formal actors to accept:

• System power will be shifted away from formal actors to informal ones.
• The new system may be more difficult (initially) for the formal actors to manage.
• Responsibility for system failure and performance will remain with formal actors.

A conceptual depiction of a recipient designed system

In contrast to Figs 2 and 3 above Fig 5 below depicts that the formal actors (in blue) are proximate to the recipient but not overwhelming them, at any given time they can be engaged or disengaged by the recipient in response to need.

In addition, the size of the input from the formal actors is controlled by the recipient (green arrows), because the recipient has the power to do that. The whole system remains responsible for the outputs to the recipient, with an emphasis on the responsibility of the formal actors to manage and minimise system difficulty and take responsibility for system failure.

It can be seen that in order to reach this state existing hierarchical systems need to be dismantled and reconfigured from a different starting point (that of the recipient) not merely tweaked by existing system controllers to the assumed benefit of recipients.

However, if existing system behaviours such as holding onto power, shifting difficulty to recipients, conspiring in failed systems and victim blaming are not addressed, the conditions are not right for truly human-centered design processes to deliver the quantum shift in system performance required to address so called wicked problems.

Conclusion

Difficulty is inherent in systems and in order that the system is as efficient as possible it should be held and managed by the parts of the system best equipped to manage it. The role of the formal actors is to ameliorate or reduce overall system difficulty through collaboration. This collaboration can be active (working together on an area of difficulty which may not be mutual, but which if resolved would reduce system difficulty) or passive (for example by using compatible digital platforms to conduct business). The collaboration may result in the loss, or transfer of power from one part of the system to another, but if overall system difficulty is reduced this is a positive outcome.

Only when formal system actors are reconciled to the following will the conditions be right for client centred systems to work:

• Shifting difficulty onto another part of the system is not the same as reducing system difficulty,
• Power must be ceded directly to other system actors or dispersed more widely in the system,
• All system actors must be equipped to exercise their power,
• Performance is a whole system measure.

This indicates that shifting to client-centred design and improving system performance will not be achieved by merely adapting existing power-based systems and their behaviours. What is required is the dissolution of hierarchy, control and the concentration of power among formal actors to the detriment of informal actors or recipients. Only when this is achieved as part of the client-centred design process will the conditions exist for new systems to emerge, replacing outmoded systems which will create the possibility of significantly improved system outcomes which benefit the recipient.