Online courts will operate in what has been called a problem-solving way. To begin with, they must have seemed frustratingly slow. The very first users were asked to fill in forms on their home computers, print them and then send them in by post. When the forms arrived, court staff would re-key the information they contained.
This was not as daft as it sounds. Developers can learn a great deal from the errors made by early users of software. The programmers then refine the forms until people can understand them and complete them with ease. Once that’s been achieved, the information keyed in by court users is fed straight into the court system.
All forms were designed using a standard template, adapted from the government’s gov.uk website. The design had first been used on early online sites, such as those allowing defendants to make a plea online1 and offering help with court fees.2 However, it is surely wrong in principle for court sites to be branded “GOV.UK” and hosted on a government website. In a democracy, the courts must be seen to be independent of the government.
When I put this point to HMCTS early in 2018, I was told that no users had objected. But I suspect that was a failure of consultation rather than a sign of approval: there is no evidence that anybody was asked to consider the advantages and disadvantages of close links to the government.
The importance of judicial independence was accepted by the government when tribunals were moved from government departments and brought into HMCTS. It was accepted when the UK Supreme Court was taken off the government website. I have long maintained that the same should happen to the online court. In June 2018, I was told that the issue was “under active consideration”.
That said, the government website is very well designed. It manages to be dignified and authoritative without appearing intimidating. It works well on all platforms.
No doubt the “GOV.UK” brand is cemented into the website at a very deep level of coding. But the courts website can still be rebranded and given a different web address. It just needs somebody in government to authorise it and fund the extra work involved.
Before we knew what the developers were up to, I tried to imagine how an online court would work. I guessed that it would need to find out pretty early on whether a user was bringing or defending a claim. So it might ask you whether there’s something you are complaining about. If that’s so, it will ask you what it is. There’ll be a series of suggestions. Maybe an item you ordered online hasn’t arrived. Or perhaps something you bought has stopped working. Maybe you fell off your bike because there was a pothole in the road.
On the other hand, somebody may be complaining about you. Perhaps there’s a dispute over whether you have paid a utility bill. Maybe you’ve already had an email from the court about this. Or perhaps the person who bought your car from you has found something wrong with it. Again, you’ll be selecting from a series of options.
At this stage, the online court would try to establish whether there really was any dispute between you and the other side. It may just be a question of allowing a debtor more time to pay. If there is a dispute, the website will alert you to alternative ways of resolving it, such as mediation. The software should also give you basic legal guidance and alert you to sources of free or affordable advice. At some point, the online court will collect a fee from you.
Each question the online court asks you will depend on the response you have just given. So if you tell it you are claiming compensation for an accident, it will ask where the accident took place and who you think was to blame. If you are suing over a faulty product, it will ask how long ago you bought it and perhaps help you to search online for the shop or website that sold it to you. If you say you sent it back, it will ask when.
The online court won’t ask you whether your claim is for tort or breach of contract. It can work that out for itself. But it will want to know how long ago the problem happened. If that date falls outside the limitation period for your kind of case, the online court will tell you that the claim can’t go ahead. The “help” button should tell you why.
It wouldn’t stop there, of course. If you bought something, there should be a receipt. If that’s available online, the online court should show you how to copy and upload the appropriate web page. If it’s a paper receipt, you’ll be shown how to photograph it with your smartphone. The online court will invite you to take pictures of the damaged goods or dangerous pothole. There might even be video evidence that you can upload.
That’s as far as I got with my thought experiment. Let’s move on to what we are now expecting.
In due course, the software will have assembled a series of answers that summarise the user’s position — a logical sequence that can be read online by the other party to the claim and, in due course, by the judge. That’s stage 1 of the process: helping court users to explain their claim or defence clearly enough to be understood by their opponent and to be resolved by the court.
But it will go further than that, according to Sir Terence Etherton, the master of the rolls. “It will assist individuals to find the right sources of legal advice and help in order to enable them to consider whether they have a viable legal dispute or whether a more appropriate means of complaint or redress is available, such as a relevant ombudsman scheme… In that way, this also is a form of preventive justice, made accessible to all through the changes it makes.”3
Stage 2 will involve a mix of conciliation and case management, conducted partly online and partly by telephone — although probably not face-to-face. This will be done by case officers, “court administrators exercising judicial functions under the supervision of the judiciary and independently of the government”.4 Parties will have the right to ask for their decisions to be reconsidered by a judge.
The online court will be the first in the country to have a built-in mediation process, we were told.5 Its aim will be to resolve the case; or at least to narrow the issues. “This is a significant departure from the court’s existing role as it will require a court officer actively to engage the parties in mediation and conciliation processes,” the master of the rolls said in June 2017. “Case officers will promote the best resolution method for each case: mediation, online alternative dispute resolution, early neutral evaluation (that is, an evaluation of the legal strength of the claim, probably carried out by a district judge) or proceeding straight to a full trial.”6
Unless the case has settled, it will then proceed to a decision: stage 3. The court might have enough information to decide the case without a hearing. But if the judge thought the two sides should argue it out, there could be a telephone conference call. Perhaps the judge and the parties would be able to see and hear each other on Skype or FaceTime. And, if appropriate, there might be a traditional face-to-face hearing — but with strict time limits. For the first time — in theory, at least — judges would be required to do their own legal research.
Writing to Lord Justice Briggs with its comments on his interim report, the Civil Justice Council’s online dispute resolution advisory group said in March 2016 that “as soon as possible, in all cases, whether progressed through the traditional system or via some online court, actions should be started online”. It argued that the development of an online court should be seen as an evolutionary process, refined in the light of experience. It could also be developed in the future to take account of emerging technologies “such as social networking, artificial intelligence, and predictive analytics”.
Perhaps counterintuitively, the advisory group recommended that stage 3 of the online court — the judgment stage — should be delivered first. One reason for this is that allowing judges to decide cases without a traditional hearing would immediately relieve pressure on the system. Another is that it would be easier to design and implement stage 3 than the other stages. Professor Susskind’s advisory group argued that stage 3 should be followed by stage 2 — the mix of conciliation and case management — and then by stage 1, the decision-tree process (formally described as a suite of “diagnostic rule-based expert systems”).
This approach seems very wise. But it was not followed. It looked as if HMCTS wanted to grasp the low-hanging fruit first.