Systems Architects vs. Service Designers
The most impactful policymaking is not service design
In my previous post (I realised I forgot to link to the report I reviewed!), I briefly touched on a distinction between two types of policymakers: system architects and service designers. In this post I want to flesh out this distinction, and show why I think that most attempts to understand the role of policymakers focus too heavily on the service designer role. The service design approach has value, but has too much influence over attempts to reform policymaking.
Service Designers
The name service designer is pretty self-explanatory: they are people who design services for a certain group of users. The wikipedia entry for service design says “Service design is the activity of planning and arranging people, infrastructure, communication and material components of a service in order to improve its quality, and the interaction between the service provider and its users.” Glad we cleared that up.
There are service designers in many industries and sectors, including those who refer to themselves as service designers and those who don’t. Someone has carefully designed the service I receive as a writer on substack, as well as you as a reader to make it incredibly easy-to-use for both of us. Even in commodity products, there are service designers - shopping at the supermarket is buying products, not services, but the supermarket company has carefully designed the shopping experience to suit your needs and thus maximise the profits they can make from you.
The service design approach is characterised by:
An expectation that the solution will be providing a service (yes this is almost tautological, but it is worth making explicit).
A central focus on service-users, identifying who they are and what their needs are.
Co-design of the service with users.
Thinking about policy as an ongoing process of refining or redesigning a service.
The adoption of design principles developed for the tech sector, and a close relationship with digital experts in service design and delivery
A lot of the advice and best practice on policymaking in the civil service is based on a service design philosophy. Wikipedia’s entry on service design actually has a section on service design in the UK Government, which is all about the Government Digital Service and Behavioural Insights Team. The Policy Profession’s official blog is called Public Policy Design. Most policy labs in the civil service seem to be populated with people with job titles like “service designer” and “creative technologist”.
Before I continue, I want to stop and appreciate the service design approach. This post is arguing against the dominance of this approach, but I think it can be very effective when used for the right purposes. I’m not going to discuss its strengths in much detail, you can find that extensively elsewhere from people far better qualified to talk about it than me. There are many examples of policy makers who have been helped by following a service design methodology and the civil service is better off for having it play a major part in policy innovation.
Why the focus on service design?
The official policy profession standards are a good counterpoint to user-centred design proselytising. “User-Centred Design, Digital and Behavioural Insights” are just one of eighteen skills an effective policy professional is expected to have. The service designer obviously needs most of the other skills too, but why so much focus from policy profession thought leaders on this one particular skill set?
To be clear, I am not claiming that this distinction is not recognised, or that I am original in my ideas. The idea for this post was my own, but in researching for it, I found a very similar treatment of the topic from Andrea Siodmok, former Deputy Director for the Policy Lab. Her team also helped put together a systems toolkit for policymaking. I’m not claiming that policy experts ignore the importance of systems thinking, or its distinction from design. Nor is the distinction a clearly defined one; there is overlap between the two types of thinking, and each is somewhat of a fuzzily defined tendency rather than a clearly articulated set of standards. I just think that policy innovators over-emphasise service design in their practice.*
I think one reason for this overemphasis is too much influence of the Government Digital Service on innovation in policy. Clearly, modernising government should involve a hefty dose of digital expertise. Digital enthusiasts are often the most exciting leaders in business and government because tech culture is so optimistic and innovative. Many (most?) of my favourite management books have come from tech company leaders. But we shouldn’t get caught up in this enthusiasm and let tech experts think that they know how to do policy. Policymakers should be leading the digital experts, rather than following them and copying their ways of doing things.
I suspect another reason is borrowing from wider public sector management. Most of the public sector is not involved in government policy, but rather the more humdrum aspects of public service delivery, such as schools, hospitals, local government and job centres. Managers and leaders in these roles are also much less restricted in what they can say publicly than civil service policy leaders who work in a more political environment. Therefore, it is to be expected that the discourse on public sector management is dominated by people in service delivery. Most civil servants are not thought leaders themselves. They rely, perhaps too much, on a thriving discourse of external experts to shape how they think about what the government should do. So if public discourse says that user-centred service design is the best practice, it would take a very confident civil servant to suggest pursuing an alternative model. It would also require the civil service to largely invent the approach themselves, rather than borrowing from best practice elsewhere. There may be international comparators, but other national governments likely suffer from similar problems.
System architects: an alternative
The alternative system architect approach is less about designing every aspect of the service and more about arranging the rules and institutions of society in a way that moves it towards a certain desirable goal. This is typically because the government doesn’t have the power to directly control every aspect of the system. This might be a capability thing, but more likely it is because providing a service to solve the problem is either an ineffective solution, or undesirable, e.g. because it infringes on people’s freedoms.
The systems architect approach is characterised by:
Achieving high-level social goals, with the end customer being every member of the public, or a large portion of them
Finding the right point to intervene in complex systems with large numbers of actors, both institutions and individuals
Limited ability of government to control certain aspects of the system
Developing rules and regulations, or reorganising money and resources.
Thinking about policy as a few big one-off changes that are hard to reverse
I saw an experienced policymaker recently say on Twitter that policymakers have only three main tools at their disposal: regulation, taxation and subsidies. Note the entire absence of service delivery in this list - this person either missed it out, or just doesn’t consider that policy. Given this person worked in the energy and climate space, they were mostly correct. Short of nationalising not just energy infrastructure but significant parts of the economy, the government isn’t going to hold all the levers in this area (and economic central planning doesn’t have the best track record where it is possible). Instead, we can intervene at certain parts of the energy system. The government might wish to design a certain user-centred experience for e.g. decarbonising your home, but it doesn’t have the power to design each part of the process itself. There’s definitely benefit to thinking about the ‘end users’ of climate policies, but it isn’t going to tell you which part of the system is the right leverage point.
Systems architecture is already a discipline within digital & IT. When I first chose the phrase, I didn’t intend any analogy between my usage and the existing usage within IT. I don’t really know what a digital systems architect does. The phrase was either buried in the back of my brain or I independently came up with it. But having done the barest of research, I actually think there might be a useful analogy. A digital system architect fuses business and technology strategy to achieve certain goals or requirements. More concretely, they are responsible for ensuring an organisation has the right software, hardware, tools and processes or user interactions to support the overall organisation or team’s goals.
The analogy isn’t perfect. The IT systems architect in theory has control over the processes within their organisation. Whereas the government is limited in its ability to effect change on society in areas outside of direct service provision. The IT systems architect is more closely integrated with service designers than the policy systems architect, because in IT you can not only map but directly control all user interactions in the system.
If ‘system architect’ feels either too analytical or technocratic to you, then try an alternative framing of ‘stewardship’. Instead of technologists, economists and data scientists, this conjures images of ecologists or even religious theories of humans as God’s stewards. To me they seem like two sides of the same coin. They are both not about controlling, managing or dominating, but instead about guiding and maintaining. One just takes a more analytical lens than the other to understanding the problem. The two meet in the middle for me in the work of Elinor Ostrom on managing public commons.
I’d argue that not only is system architecture a major part of policymaking, but the people involved in it are the most influential policymakers. The system architects are those who allocate billions of pounds of spend, who draft Bills and set regulations that impact entire sectors. The team who design, for example, the probation service can have a big impact on the performance of that service for assisting offenders and protecting the public. But if I was tasked with reducing the numbers of reoffenders, I wouldn’t start by reforming the probation service, or rather it would form a small part of the necessary reforms. Instead you might look at things like decriminalising certain minor offenses like drug possession (legislation). Or redirecting funding to youth centres (spending). (Disclosure: I have zero expertise on crime or justice policy and am broadly parroting what liberal wonks seem to suggest).
In fact, even in the public services sections of the civil service, most policymakers do at least as much system architecture and service design. Much of the service design is effectively outsourced to private contractors or arms-length delivery videos. DHSC does not do much health service design, that’s the job of the NHS. There are a lot more commissioning managers in the NHS than there are policy professionals in DHSC, but the latter have more power on an individual level. That’s because they do things like control which spending goes where, or define statutory duties that all NHS trusts/CCGs have to follow.
So if the civil service is going to professionalise policymaking, a focus on service design can be helpful. But improvements in the system architect role of policymakers is likely to pay bigger dividends.
Going back to the report from my last post, the system architect obviously overlaps a lot with their “Government Business Expert” policy specialism. Especially the last responsibility listed for that specialism:
“Coordinate policies and services with a system-wide perspective (understand the range of government interests, identify synergies and make connections)”
It even has the word system! Professionalising this as a policy specialism would go a long way to meeting my critique of existing policy profession innovation. The one thing this maybe misses is the sense of ownership of a system. This requires both a deep subject matter expertise and the ability to break out of siloes and see the bigger picture. I suppose it should be obvious that the best policymaker would combine elements of each of the 3 specialisms listed in the report. But by formally separating them, there would be a risk that the most powerful decision-makers become even more generalist and divorced from understanding of the systems they control.
Having explained the system architect approach, I think there’s another reason why a lot of policymaking innovation focuses instead on service design. Ministers get far more involved with the activities I’ve described as system architecture. They may show some interest in service design, but typically are more content to leave it to managers. Many argue they should show more interest in policy delivery and service design, but it’s also partially a sensible allocation of their scarce time and attention to the activities with the largest impact. The reason this may turn policy innovators off from trying to professionalise this side of policy is that it’s much harder to do so without thinking about the role of ministers. This makes everything orders of magnitude more controversial and difficult to change. But this doesn’t mean we should shy away from it. We may find that ministers welcome a bit more rigour from their officials.
*I actually discovered Andrea’s blogs after I had drafted most of this post. I am now annoyed by how much my thinking overlaps with her posts (I’m not original after all), but also much more confident in the conclusions I’ve drawn because I have expert corroboration. It also annoys me that people within government have developed these tools, but they are poorly integrated into the actual practice of policymaking. Professionalising UK policymaking is a path littered with many reports and few, incremental actions.