Why digital government needs a clearer relationship with the private sector to scale
The next step in digital government maturity is a rational relationship with the private sector
The Ministry of Justice talked publicly this week about its form builder tool, sparking debate about the prospects it could develop into a generic, reusable capability for departments and other public bodies to build high quality web forms. It’s easy to imagine how this could accelerate and scale digital government following in the footsteps of GOV.UK Notify and, to a lesser extent, GOV.UK Pay. Most of government is forms, most of the time, as Matt Edgar once observed.
Development of a generic tool would likely produce an equal and opposite reaction. Some projects would use the forms solution because it has the GOV.UK association. Others wouldn’t yet because they’d consider it inferior to existing solutions. Incumbent providers of (Granicus, Jadu, IEG4, Netcall to name but some) will be cornered into a fight or flight strategy. Both camps would criticise the other’s choices. And after several years, patient marketing and investment in spurts, a core part of government technology infrastructure may emerge. And the world (ok, bits of the UK public sector) might have better looking, more usable forms.
It doesn’t have to be this way. But digital government hasn’t yet reached a rational relationship with the private sector. For most of the last 10 years there have been two ecosystems. The GDS-backed ‘insource’ approach which also actively seeks to reduce spend on traditional vendors, and the ‘buy it from traditional vendors’ default from every other part of government. When you pit yourselves against the existing market, every time you need their help it becomes an embarrassing failure. If you don’t have a viable alternative to ‘buy it’, you’ll never have the solutions you need or the market influence to achieve it.
I used to advocate a ‘glo-cal’ hybrid approach: the best of multinational vendors (say, AWS or Salesforce) and the best design and research capabilities that local SMEs had to offer. It’s a way of breaking out of the jaws of the national providers of bespoke sector-specific solutions operating in a shrinking market with underwhelming roadmaps. But Microsoft’s recent price rises show how hard it is to avoid vendor lock-in for essential tools (this isn’t just about Office v GSuite but Windows, SQL databases and servers too) and the ad hoc use of local SMEs won’t scale to deliver core solutions over a five year strategy.
Wardley Mapping offers a more strategic, coherent way through this dilemma. By analysing the maturity and importance of each technical component, organisations can determine how to rent components that should be purchased as invisible commodities and build important, bespoke components. However, that requires the market to be sufficiently mature that it offers components that can be reliably integrated with custom components. Mostly, in the UK public sector, the market isn’t offering that.
Potentially, forms provide a new way forward. But to succeed, we would need to not start with forms but instead think bigger and act smaller. A better form builder normally generates more forms. Mostly we need fewer. Most of government is forms most of the time only because they’re the consequence of data silos. No citizen woke-up wanting to fill in a form. Instead, to access a public service the citizens is asked to provide information back to government that it already holds elsewhere.
So we need to start with data. What do we need to know and where does that information currently reside? No forms product can improve government services without better managed data. The forms for vulnerable people shielding from COVID and the track and trace ‘system’ demonstrate this.
Once a form has data, it triggers a workflow – authorising, tracking, reporting, usually. Forms products without workflows reduce us to government by spreadsheet. The ‘register to vote’ service on GOV.UK which only emails local authorities rather than actually registering voters demonstrates this. The more custom the workflow, the less useful the form builder. Richard Pope’s platform government playbook draws attention to the need to start at the right level of abstraction for reasons like this.
Government could instead tackle these challenges. Rather than spending time and money recreating the technical features, government could instead support the market to evolve – so it can either become a commodity supplier or specialise in helping organisations define the custom, visible components they need. This might involve government unblocking data sharing constraints, establishing technical criteria for APIs (building on Rosalie Marshall’s work), data standards for form inputs and design patterns for approvals and verifications. Smaller, discrete technical components might then still be needed – citizen single-sign on, for example. But public bodies could then choose from a range of products developed to national standards, depending on the user needs, level of software support or just plain old price.
There’s an example of an alternative approach emerging in London government, thanks to the work of Peter Kemp and team at the GLA. The redesign of the London Development Database has seen the GLA work with providers of local authority planning software to define the data that planning authorities need to provide and how that will flow to the GLA. It appeared as though some vendors were more open to change than others, some planning authorities more amenable than others.
Mature digital government will scale beyond services, departments, public bodies and national borders. It will necessarily be a shared ecosystem of free, open source, and commercial solutions. Done in the right way, it can deliver better public services. But it needs a definition of working in the open which includes a rational role for the private sector.
