Barriers to participation in employment litigation: what might make a difference and would it work?

AUT and Victoria University Symposium, Wellington, 22 May 2019 Hon Justice Miller

Tena tātou katoa

When the Chief Judge introduced me as a judge of the Court of Appeal you probably had two thoughts. One was that he knows nothing about employment law. To that I would offer no defence. The other was that he is part of the system that is being examined in this symposium, so it is possible he has been brought along as a specimen to be examined, a living, breathing example of a barrier to participation. To that I would be disposed to offer a modest defence. I think the Court of Appeal has been quite disciplined about its limited role, quite respectful of its duty to defer to the expertise of the Employment Court, and quite good at turning around the cases it does take.

So what do I have to contribute to this symposium? I have spent time thinking about what makes courts work, or not work. I organised the analytical work that led to some case management reforms in the High Court which I will talk about briefly. I designed the Christchurch Earthquake List and ran it initially. I chair a strangely named body, the Judicial Reference Group, which comprises representatives of all benches and senior Ministry of Justice people and deals with courts modernisation projects: electronic filing, case management systems, electronic casebooks for use in court, and so on. Recently I have been thinking about courts governance in NZ. So my perspective is that of a judge, but one who has given some thought to processes and systems.

The very fact that this symposium, the second in a series, is being held is a positive thing. It shows that the institutions of justice in this jurisdiction are willing to subject themselves to critical scrutiny from their end users. Courts have a long history of basing process reforms on the opinions of those at the apex of the system, the judges and the senior bar.¹ These are lawyers, not social scientists, and they do not have perfect visibility of what is happening in the court system. Rarely is there any attempt to measure what is happening now or any plan to monitor the effect of any changes.

It is wise of you to start by accepting that court systems are very complex and no one participant or class of participants knows exactly what is going on. Collective experience will give you a better sense of the problems and allow you to generate some consensus about what needs to happen. You will get better reforms that way. Consensus also matters because reform usually requires some behavioural or cultural change which will not happen unless people buy into it.

The topic is ‘barriers to participation and what might make a difference’. I want to start by asking what it is that the court system exists to do. The answer must inform what it is that needs to be done to improve participation in it, and where the trade-offs can be made.

I have admitted that judges do not know everything that is going on, which makes me an unreliable commentator, and also that reform ought to be informed by hard evidence, which it seems we do not have. From that not entirely secure platform I am going to advance my argument. I argue that the objective of a court system ought to be to make adjudication viable in every case. By adjudication I mean a decision imposed on the parties by a third party; in this jurisdiction, a Judge or an Authority member.

You are thinking we are here precisely because adjudication is not viable and cannot be made so in this jurisdiction. Quite a large proportion of the population experiences employment problems ― 9% according to one NZ survey conducted in 1999² ― and very often those problems are associated with financial stress. The notes from the first symposium indicate that this is a major concern, especially for some communities. Bear with me while I defend my thesis.

People say they come to court for justice. There is a sense in which that is true. A sense of justice usually motivates them; it explains why they think they should win. But the functional reason plaintiffs come is because they want to make another person do something that person would rather not. Defendants participate to avoid being forced to do something. Both parties are there to invoke the state’s power of compulsion.

Ultimately compulsion must follow adjudication. So the availability of adjudication is the engine that drives results. Even those cases that settle long before trial do so under compulsion in the sense that if they do not they will have no choice in the matter; they will face adjudication.

Chief Judge Inglis made this point in the first symposium in this series.³ She cited the UK Supreme Court decision in Unison v Lord Chancellor, in which it was said that people need to know there will be a remedy for them, or against them, should obligations be breached.⁴

That supplies the first reason why we should aim to make adjudication viable. By making it available to parties we bring the power of compulsion to bear. We make the system effective.

The second reason is that the availability of adjudication encourages quality settlements. A quality settlement is one that mirrors what a Judge or Authority member would do. A quality settlement is one that is not distorted by the combination of an imbalance of power between the parties and the inaccessibility of adjudication.

Another way of putting this point is to say that settlements should happen in reasonably close proximity to adjudication. I use proximity here in two senses. The first is temporal; parties should not wait too long. Delay is a cost in itself and a cause of expense. The second is procedural. Procedural barriers to adjudication attenuate pressure to settle and may exacerbate the effects of inequality of arms. I will return to this point in a few minutes. In the meantime, I observe that yours is a jurisdiction in which the legislation itself recognises an inherent inequality of power in employment relationships. If that is so, then I suggest that access to adjudication is more, not less, important than it may be in some other jurisdictions.

The third reason for aiming to make adjudication viable is that if it is viable some cases will reach hearing and result in judgments or determinations. That is something to celebrate, provided there are not too few, or too many. Dame Hazel Genn said that adjudication is essential if civil justice is to perform its function of reinforcing values and practices.⁵ Speaking in 2007, she made a point that seems prescient in the world we now occupy, in which disposals have become the only thing that matters to courts administrators. She said that the state’s responsibility for civil justice is being shrugged off on the basis that civil justice is a purely private matter, nothing more than a service offered to litigants.⁶

In fact, access to justice is a civilising force, and the product of the civil justice system is a public good.⁷ Judgments allow other people to order their affairs according to law. In my opinion the real measure of the health of a judicial system is the proportion of cases that it succeeds in getting to hearing. If enough do so, we get the precedents we want, and we can take some comfort that trial is in fact viable in many cases; in short, that the system is working.

The absence of trials is a genuine problem in some jurisdictions. In 1962 11.5 per cent of civil cases in Federal Court got to trial in the US. By 2002 that had fallen to 1.8 per cent.⁸ In case you are wondering, there was a very similar picture in crime, where just 3.3 per cent of cases reached trial.⁹ Those figures come from a book called The Death of the American Trial by Robert P Burns.

In England a similar though less dramatic pattern has been observed. There was between 1990 and 2000 a dramatic decline in the absolute number of cases going to trial in the Queen’s Bench Division, from about 3200 per annum to just over 500.¹⁰ The proportion of all civil proceedings reaching trial had fallen from 1.5 to 0.5 per cent.¹¹ (That data is old now, of course, and I am not entirely sure why the figure is as low as 0.5 per cent. Perhaps it has something to do with the prevalence of personal injury claims or perhaps it includes basic debt recovery claims.)

If making adjudication viable is the objective of a court system, the questions for today can be framed in this way: how can adjudication be made viable? What are the barriers to adjudication, and how can they be reduced?

I have five suggestions. By way of overview, they collectively address two problems. One is the problem of proportionality ― that is, the need to ensure that processes are proportional to what is at stake. The other is the problem of certainty of adjudication ― that is, the need to ensure that the power of compulsion is in fact available and will be exercised, absent settlement, on a date certain.

The first of the five suggestions is that you must decide what adjudication entails. Does it mean the Rolls-Royce model, to use the traditional metaphor, with discovery and oral evidence and cross-examination? Or does it mean a hearing on the papers, or some kind of inquisitorial process? You want a range of options, a menu from which you can choose. There may be limits to what you can do here without the aid of the legislature, limits to what you can achieve without the parties’ agreement. But you should first see what is possible within the existing framework.

In the court system, generally, there is enormous potential for improvement, for evidence-based reform, when it comes to this question of what adjudication ought to entail in particular cases or types of case. I spoke before about the limited perspective of those who have traditionally been in charge of procedural reform. Judges as a collective have not found it easy to think beyond the instant case to the unmet need for adjudication that is out there in the community.

Courts know that many parties cannot afford very much process. They know that as a result some cases will never get to hearing, and many others will never enter the system at all. But they do not know the measure of this unmet need and their processes do not encourage them to take it into account. They have traditionally failed to recognise the incentive effects of their behaviour on people who are in the system already, awaiting trial, or thinking about going to court. When faced with concerns about expense, they have sometimes preferred to emphasise that process rights, like discovery and oral evidence with cross-examination, exist to guarantee the effective exercise of substantive rights, and the Rolls-Royce approach is the best way that the courts know of to get the right answer every time.

By contrast, summary process always carries with it an enhanced risk of error. To adopt it is to accept that we will get the answer wrong in a minority of cases. It is to say that some errors are an acceptable price for making adjudication accessible to people who would otherwise be denied it.

Only in recent years have courts begun to engage with the problem of how to strike the balance. In 2009 the High Court of Australia delivered a judgment, Aon Risk Services Australia Ltd v Australian National University, in which the Court recognised a public interest in the efficient despatch of court business;¹² put another way, the Court held that when deciding what to do in the instant case a judge may take into account the needs of all litigants, present and prospective. That is a good principle. It has been accepted in New Zealand.¹³ The High Court Rules 2016 now recognise that discovery, which is often blamed for the cost of litigation, must be proportional to what is at stake. But we have not gotten terribly far. We still struggle with the idea that there are cases in which the interests of justice require that process rights be curtailed to some extent. We still lack ways to gauge where the balance ought to be struck. We do not have a full suite of procedural tools to fashion an appropriate form of adjudication for every case.

So that was my first suggestion; decide what adjudication ought to entail in your jurisdiction and in the differing types of case that you see. Recognise that there will be room for creativity within your existing processes.

The second suggestion follows from the first. It is triage, at the beginning. Get the people into court, find out what are the issues, which ones might need adjudicating, who are the witnesses, what process might need to be followed.

Triage must be performed by a person with authority and knowledge. It is an exercise in intensive case management. One of the Earthquake List’s main features was that one judge would do all the case management. Initially that was me. Through familiarity with the cases the judge got better at issue identification and could be more pro-active about managing cases. The judge could identify issues for separate decision, which is an efficient form of adjudication. The judge could also set the scene for settlement; for example, the conferences very often established that the parties were far apart because they had not agreed on what was broken about the house; in those cases the judge might direct that their quantity surveyors meet on site and report back. The usual result was that the difference between them vanished and the case settled. The judge could also manage the cases as a body, deciding which cases were a priority for hearing because they presented some issue of principle that was common to a number of cases. I believe this worked well. It led to the introduction of a leaky building list in the High Court at Auckland.

The third suggestion is that the Court or Authority should set the hearing date after the first case management conference and choose a date that is as early as possible. Why do I say that? The empirical work we did in the High Court in 2011 showed that most cases settle against the trial date.¹⁴ They begin to settle about the time that the parties are needing to incur the costs of preparing evidence. Case management should exploit that dynamic. We were not previously monitoring behaviour in the pre-trial window at all. When we did, and understood how cases were behaving, we were able to increase the rate at which the lists were overbooked― by that I mean booking more cases than we had capacity to try on a given day, much as the airlines overbook seats ― and that reduced the time to trial for a standard case by three months.¹⁵ We were able to tell the profession that they could reliably get a short cause ― less than 5 days’ hearing time ― to trial within 12 months from filing.¹⁶

I am not suggesting that a court should focus on hearing dates to the exclusion of mediation. Issues conferences are an excellent opportunity to tease out settlement options, and the court should be alert to that. In the study I was involved with we found that about 30% of cases settle soon after filing and before the first case management conference.¹⁷ That is a significant percentage. So filing alone creates an incentive to settle and the case management conference can help by identifying what really divides the parties. I’ve just explained that in the Earthquake List it very often led quite directly to settlement. Mediation should be on tap to take advantage of those opportunities. Sometimes the court can set it in train at the case management conference. It can adjourn the case management conference for mediation if that seems a good idea. But as a general rule it should not delay setting down.

My fourth suggestion follows from the third. It is that before making any changes you should do your best to find out what is actually happening. How are cases behaving and how does that behaviour relate to your processes? Reform should be based on evidence so far as possible. What I have told you is based on what I learned in the High Court, from the limited work we did there, and from running the Earthquake List. Those lessons may hold good for your tribunals, but of course they may not. I expect that Bridgette Toy-Cronin, who authored the Wheels of Justice report and is speaking today, may have more to say about that.¹⁸

My final suggestion is this: confront the myths that are out there about your jurisdiction. Perception has a way of becoming reality. When we did the empirical work I spoke of in the High Court there had been a lot of bad press. There was much talk in the pages of the NZLJ about the death spiral of civil litigation, said to be the result of interminable delays in getting cases to trial. The claims were wrong. By international standards the High Court did a good job, and still does. It gets almost 10% of ordinary civil proceedings to trial.¹⁹ Unbelievably, we did not know that figure until we did the empirical work I spoke of. You can debate whether 10% is the Goldilocks number; neither too many nor too few. What cannot be denied is that it is a better number than the miniscule percentage that reach trial in some comparable jurisdictions. We were able to show the evidence to the profession, and the death spiral talk stopped.

Part of this exercise of confronting adverse perceptions involves thinking about what information you publish. Traditionally courts published their judgments and that was that. They published nothing about the great majority of cases, those that came to court but were resolved short of trial. We are a little better now, but only a little. Bare statistics about disposals are publicly available.  These tell us something about productivity.  So, for example, the published data tells me that the Employment Court received 194 cases in 2018 and disposed of 185, from which can be derived the Court’s clearance ratio.²⁰ The Ministry of Justice Annual Report, which the Ministry uses as evidence of performance when justifying its budget bids, focuses heavily on disposals.²¹ The measure of performance used for the Employment Court is the clearance ratio and the percentage of cases on hand that are fewer than 12 months old (76% last year, if you are interested).

Productivity matters, but mainly to those whose job it is to worry about the efficient use of public money. If we can, we ought to publish data that allows prospective litigants to better predict what will happen with their case, and when. The information that may help them is; what sort of cases are coming to court, what issues do they raise, how are those issues being resolved, and when are they being resolved. This is information that tells the reader something about what courts are actually achieving. It is information that intermediaries, such as lawyers who practise in the field, can examine and use to advise potential litigants. It can reduce somewhat the uncertainty in going to court. It also addresses information asymmetries than can arise in jurisdictions where you get repeat players in the litigation game.

Repeat players were a feature of the earthquake cases. The plaintiffs were one-off, but the insurers were repeat players and as such likely to be better informed. We put together a simple spreadsheet which was published periodically, identifying the cases and who the lawyers were, listing the significant issues, and recording the milestones, including settlement. This was aimed directly at the information asymmetry problem. Potential plaintiffs with claims against a given insurer were able to see that it was defending cases that raised a given issue and they were able to see whether those cases settled, and when they settled relative to filing date. They could talk to other plaintiffs. The publication of this information ought to have eased some of the friction in the system. It turned out, unexpectedly, that many of the plaintiffs used the same lawyer, so it may be there was less need for the spreadsheet than I assumed at the outset. I offer it only as an example of an attempt to give potential litigants better information on which to base decisions about coming to court.

I am not saying that you will succeed in making adjudication viable in every single case. Plainly, courts cannot achieve that objective on their own. Some cases will never be viable, absent a funding mechanism to ensure that representation is available. But it is nonetheless the right objective, in my opinion.

Having ventured that far, I am now going to stick my neck out. I am going to comment from my outsider perspective on two features of dispute resolution in the employment sector.

The first and to me remarkable thing is that you can get two first instance adjudications, one in the Employment Relations Authority and if you do not like that, another one in the Employment Court. I know that the Authority process is more inquisitorial and the Court more adversarial. I do not want to be taken to suggest that one is better than the other. It seems to me that there is much to be said for both models. But from an access to justice perspective, there really ought to be one merits adjudication and one appeal, period. It seems extravagant to permit any single case to be the subject of two first instance adjudications. That possibility is unlikely to appeal to prospective litigants. Returning to the point I made earlier, there may room for a triage mechanism in which the decision about which form of adjudication will be offered is made on a case-by-case basis.

The second thing is that, as I understand it, mediation is practically compulsory in your jurisdiction. Mediation is an excellent thing. It can settle disputes quickly and efficiently, on a basis that achieves value for both sides and even preserves relationships. I argue nonetheless that courts and tribunals should not make it compulsory. For one thing, it conveys the message that there are no principles worth standing on; nothing that should not be bargained away. It is no part of a court’s function to convey that message. For another, to insist on mediation is to create a process barrier to adjudication, and that is bad for the reason I mentioned earlier. It would be especially so if you have to undergo mediation before you can be assigned a hearing date.

To me the insistence on mediation raises a question about the quality of settlements. We have not looked into that. We tend to assume that if the parties settled the outcome must have been good for both of them. That may be true, but it need not be. My own experience of settlement processes at Judicial Settlement Conferences was that the uncertainty of litigation, the risks of delay, and above all the costs of trial, were always major factors. If the costs and delays of adjudication play a large part in mediations in this jurisdiction, that would point to a potential issue with the quality of settlements. The further away in time, the more remote the hearing, the larger looms the problem of cost and the better the position occupied by the stronger party.

Of course, the stronger party will save costs by settling early too, and perhaps more if, as may well be the case, their representatives are more expensive that the plaintiff’s. For that reason some economists would tell us that costs are neutral, since a dollar is a dollar, and if a rich person can save a dollar by settling she will behave rationally and do just that. This is to overlook the fact that the burden of costs may be much greater for the party that can least afford to pay them. For them the opportunity cost of money spent on litigation may be very high. It is also to overlook the opportunity for strategic behaviour. If the stronger party knows that the weaker party may not be able to go the distance, then it is rational to behave in ways that increase the weaker party’s costs.

I repeat that I am not saying mediation is a bad thing. Nor am I suggesting that it is not working in practice in the employment jurisdiction. I see that the mediators are well regarded. They get high approval ratings. I see from some of the papers that they settle 76% of cases and in 2018 they settled just short of 9000 cases, compared to the 750 determinations of the Employment Relations Authority and the 180 disposals of the Court.²² So they are getting through a lot of business. But these are measures of quantity not quality. The mediators are working within the system as it is. Within that system they serve as a complement to adjudication. Is it possible that they might get better results if adjudication was more accessible?

Debates about forcing people into mediation were held in the High Court too. At one time the judges and associate judges ran a lot of settlement conferences, which are a species of mediation. I did some myself. This behaviour reflected what was going on in other jurisdictions, notably in England following the Woolf reforms.²³ The thinking was that most cases settle and a Judicial Settlement Conference facilitates early settlement, so they should be encouraged.

So far, so good. But there was an added thread to it: hearing time was precious and the court should only assign it to those cases that were considered likely to reach trial. That was problematic. It gave judges an incentive to press parties to settle long before the shadow of trial loomed over them, out of of an administrative concern that we could not accommodate all the cases that would otherwise reach trial.

I am not sure whether that sort of thinking has been a factor in the processes adopted in this jurisdiction. If so, I suggest that some reflection is warranted. I make two points. The first is that the assumption that more cases will reach trial if the parties are not forced into mediation is not necessarily correct. The High Court essentially got out of the mediation business some years ago. One reason was that it is too easy for a judge to convey a view, perhaps inadvertently, about the likely outcome. Judges and associate judges were put to work deciding cases rather than conducting Judicial Settlement Conferences. The Court was not swamped with trials as a result. It was and is under pressure, but it coped. (I should say that very recently the associate judges have gone back to doing more Judicial Settlement Conferences, but only because they have capacity to do so.)

The second point is that if you adopt a practice of setting cases down early you do need to pay close attention to your scheduling practices. Traditionally the High Court judges allowed counsel to estimate trial duration and left it to the registry to determine the rate of overbooking. Each case went to the end of the list, so the trial date counsel were given was a function of the number of hearing days assigned to all the cases ahead of it in the queue, the number of judges available to hear cases then, and a loading or overbooking rate. If you just keep stacking them up then your waiting times will get longer. Cases will be delayed by cases ahead of them in the queue that will settle. You do not want that. So you need to insist that a judge, rather than counsel, decides how much time will be assigned to each case. Judge time is a scarce resource and the court should ration it. And you need to schedule with a good understanding of what proportion of cases settle in the pre-trial window and when they settle relative to trial. With that knowledge you can overbook fixtures and you can monitor and manage cases in the pre-trial window so as to minimise the risk that come the day you will be unable to produce a judge to try the case.

Of course, the High Court has a lot of judges, and it also runs a national roster. This means that it can overload by a factor of six. The Employment Relations Authority has 16 members and the Employment Court is smaller again. I am not suggesting that it is possible to overload by anything like that amount in this jurisdiction. I am simply discouraging any assumption that the Authority and Court will be hopelessly swamped if the parties are not required to mediate before they get a hearing date. Before reaching that conclusion you should take a close look at practices for setting down and for case management in the pre-hearing window.

This brings me to my last topic, about which I have only a little to say. I know that Court or Authority-imposed costs are a big concern for you. As I understand it, potential claimants worry that costs would be ruinous and so they never begin proceedings. This emerged as an issue at the first symposium.

As I understand it the Employment Court costs regime has since 2016 been based on the schedules in the High Court Rules.²⁴ Those who established the costs regime in the High Court aimed for a middle ground between the English full-costs model and the US no-costs model.²⁵ The idea was that the costs award would make a reasonable contribution to the actual costs of the winning party. This is a very Kiwi approach, designed to strike a practical balance.

They may have got the balance right for the High Court, though I am not confident about that. Assuming they did, the High Court model need not be right for the Employment Court or the Employment Relations Authority. I understand that in England the rule in employment cases is no costs ― that is, a different rule from that used in the ordinary civil jurisdiction. If so, the reasoning that led to the High Court regime would seem not to apply in this jurisdiction. Beyond that, it would be unwise for me to venture an opinion. A policy decision is needed. It would be a good thing if it were informed by evidence about the impact of costs in practice.