Future of court reform - early thoughts

On 14 May 2024, the Master of the Rolls released this speech on The Future of Courts.

I spoke on the same panel after The Right Honourable Sir Geoffrey Vos and others.

Court reform today

This is my understanding of where Court reform stands today:

The court reform programme is out of money.

They won’t be building a pre-court funnel” to help people navigate that space for two reasons: firstly, many of those user journeys are non-linear. People don’t want a pre-court funnel - it would be great to see the research that concluded this published. Secondly, they ran out of money.

They want to let the pre-court market of mediators deal with mediation… and then, if a resolution between parties isn’t found… they post the information they have into a court management system via API. Seamless. (Or, ideally, seamful.)

Public money will be spent supporting mediation suppliers for the problems the market fails to reach (read: to sufficiently incentivise the market). And defining standards and protocols they want/hope/force everyone to use.

They have set up an Online Procedure Rule Committee (OPRC) to run this (a group of judges who ratify how you can run things). It has two subcommittees, rules and technology. I assume one group agrees on policies and the other agrees on API documentation.

The amount of money they have to do this is unclear.

Let’s not obsess about that for now because, frankly and pragmatically, the plan makes sense. Money will turn up.

Getting it done

There are some gaps in policy and operations, which I will talk about. But first, to remain optimistic, here is the outline of a plan and some risky assumptions to test. I spoke to some people involved in Open Banking.

Following the principles I outlined in my talk, give a small team the power and autonomy to make good decisions. Don’t surround them with armies of Programme Management Offices and people desperately wanting to announce something.


  1. Build a team of 15 people, half of whom are technologists.
  2. Allow them to work for 6 months.

The hard bits are:

  • How do you identify and authenticate users?
  • What data is sent over, and in what structure so it’s relevant to the court process?
  • What case management system does the data go into?

I assume the Digital Justice System (DJS) case management system is impressive after 10 years of the transformation programme.  


  1. There are many hundreds of small mediation suppliers. Small firms are unlikely to run their tech - it’s a very human process.
  2. There are only a handful of suppliers of the underlying technology.
  3. So, this is fundamentally a question of getting someone like Thomson Reuters to build a pipe into a system probably run by CGI.
  4. That means calling the suppliers and asking for the price as there is no alternative.
  5. See my previous post for my thoughts on market capture. For now, let’s move on.

Unanswered questions

How will the market of mediation suppliers pass information between each other?

And how will that litigant journey be signposted?

I’m guessing the answer to complicated human matters that sit in the liminal space between policy, advice, and support is—as usual—“Pro-bono lawyers and Citizens Advice will do that.”

How can the poor state of tooling in the mediation space be improved?

I assume it’s a pretty varied picture from a user experience perspective. Maybe the OPRC could informally publish some kind of service standard and drive it.

Many of the negative experiences of digital services in the wider market mirror the issues I discussed on the day. We have been referring to this as the unit cost problem”. Small organisations with low volumes of users find it tricky to achieve a good digital service. Tooling will help to overcome this, much like Squarespace helped in the private sector.

How will public money be diverted to civil issues the market fails to support?

The market has its incentives. Private health care is great at fixing ski injuries for a reason. Which civil cases are important but don’t have a clear business model? Those will need public money.

It would be good to know the forum that makes those choices and how they have mapped current need/provision.

As a fellow panel member argued, socially excluded groups have low trust in technology. We need to actively work against that. Opening up the design process will help. I’m sure there’s lots to learn from DWP on doing this.

How can we make sure that AI or automation is thoughtfully used?

Automation is clearly going to play a role in this. It was said as much, but, it’s important to avoid issues like the Home Office found with their supposedly innocuous use of facial recognition.

These tools are powerful. Automation often drives conformity - it’s just easier (read: cheaper) to agree with the machine. But in doing so, we embed any norms further into the state.

Maybe this is too important for automation to take hold? Maybe slow is a feature? If Liberty’s official position on Facial Recognition is to ban it, maybe we should really heed the power of inbuilt discrimination.

If we are to use this technology there should be an open register on who has reviewed the training data for bias. And an operating principle that an automated agent’s outcomes are only a helper. The legal authority rests with the Judiciary.

How can we slow, reverse or avoid market capture?

We find ourselves in this situation because the public sector is pretty wasteful at doing these things. I observed that on the ground, and in my talk said:

Major systems integrators and technology providers have essentially captured the public sector. With all the connotations that capture involves. I was explicitly in a role to try to stop that from happening. To generate a creative tension that drove quality.

This was very helpful, given that major sections of the civil justice system are built and run by some of the world’s largest IT and business consulting firms.

CGI are still the major provider, they have just renewed the contract. Elsewhere major consulting firms are winning large swathes of contracts, probably to do strategic work.

Does the Ministry of Justice, HMCTS and the Judiciary have a sufficiently insourced team to avoid consultancy capture? Why not spend the margin on insourcing? We said this 10 years ago. If not, avoid unnecessary deliverables management by signing a time and materials contract.

Better still, create a dynamic ecosystem of suppliers. Yes, it introduces complexity into your contract management environment. But it’s worth it. From my talk again:

Getting the right people in is hard enough. Procurement and hiring are flawed processes in government, so I would focus there.

Reform this process, bring specialist buyers in, increase the political power of technicians. Avoid the trap of aiming for fairness, aim for defensibility. Exactly as they did in the Covid vaccine roll out. It shouldn’t take the Prime Minister and a global pandemic to experiment with delivery models:

20. The establishment—following the suggestion of Sir Patrick Vallance—of the Vaccine Taskforce outside of the Department of Health and Social Care, and comprising a portfolio of experienced individuals from industry, healthcare, science and Government was vital to its success, as was the bold, authoritative leadership of Kate Bingham. The Government was right to act to accelerate the delivery of institutions like the Vaccines Manufacturing Innovation Centre proposed in the Industrial Strategy, and to have invested further in manufacturing capacity.

How can the Treasury, Cabinet Office, NAO and ONS help?

Reform is difficult. Court reform’s next larger context” is the Treasury, Cabinet Office, National Audit Office and Office for National Statistics, so reform must happen in that context.

It would challenge current Treasury orthodoxy, but recent research publications point at ways forward:

  • Public Digital and Nesta published The Radical How, which presents a way to build a mission-oriented government that delivers better outcomes, reduces risk, saves money, and rebuilds trust
  • Tony Blair Institute for Global Change and Faculty published Governing in the Age of AI: A New Model to Transform the State, which advocates that NAO urgently review its value-for-money evaluation approach in a way that would increase a department’s tolerance for risk of failure.”
  • Sarah Gold, Dr. Natalie Byrom and Rachel Coldicutt published People first, always, which focussed on what a new government must do to make the future possible - I was delighted to contribute to it.

Eliot Fineberg

May 24, 2024