Ruminations on Judicial Error

I've recently read two excellent books by American surgeon Atul Gawande on performance improvement: The Checklist Manifesto, and Better. Both these books are mesmerizing, and it struck me as I was reading that they have much that is applicable to the judiciary. 

Ditto this blog post on medical error by Michel Accad, San Francisco cardiologist:

Notice that I said "the judiciary." Not lawyers, or the law. 

Talk of legal reform often refers to the "legal system" as if the courts, law firms, and the judiciary were all one integrated whole. But they are far from that. Like all complex systems, the legal system has several component organizations, each of which operates with a distinct set of forces, opportunities, internal dynamics, and constraints, and each of which has its own needs. Above all, each has its own incentive structure.

One of those incentives is power. Power is an integral part of all systems, and understanding power gradients is crucial for understanding how systems operate and how to change them. And power is - unironically stated - a powerful motivator.

In a well-designed system, power is matched with accountability, corollary vulnerability, or some other trade-off, usually related to bearing risk. Where this is not the case, as in the judiciary, internal standards of practice or other checks and balances must exist if a system is not to be arbitrary, exploitative, or otherwise self-serving. Those internal checks and balances are, sadly, eroding in all public systems for a variety of reasons, and in the judiciary no less so than elsewhere.

While private systems have their own challenges, they are not eroding in the same way as public systems are due to the capacity of user response to shape private systems. Immunity to user response is an unfortunate feature of public systems that is all too seldom factored into analysis of their performance. Users, incidentally, may be internal or external, direct or indirect, and may exert their influence collectively or individually, at the same time or episodically.

The process of user response to system decline was beautifully analyzed by Albert O. Hirschman in 1970, with his book Exit, Voice, or Loyalty: Responses to Decline in Firms, Organizations, and States. The best brief explanation I have found of Hirschman's meisterwerk is by his fellow economist Rajiv Sethi of Barnard College: To take just the opening thought from the work, as Sethi explains, the issue is the correction of "repairable lapses" in organizational performance: "firms and other organizations are "permanently and randomly subject to decline and decay, that is, to a gradual loss of rationality, efficiency, and surplus-producing energy no matter how well the institutional framework within which they function is designed."

What Hirschman talks about, in short, is errors. Errors that beget other errors, and alter the culture of an organization toward the norming of errors.

Reversing organizational decline can be framed as being about managing response to error, evoking the kind of writing that Gawande and Accad have done about medicine. 

But here's the thing in the judiciary: apparently, judges do not make errors. So perverse is the mindset about judicial errors that not only do judges not make errors, but also, if a judge makes an error, it becomes a non-error by virtue of having been made by a judge. 

Now, you may think that judges acknowledge error all the time, given that the appeal process is all about identifying and correcting errors. And a good thing too, as far as it goes. But that's precisely the problem. The appeals process only catches those errors that are appealed; it only corrects those that are deemed to reach the threshold of the standard of review, and finally, and worst of all, the appeal courts are not immune from making errors themselves. And at the final level of appeal, the error becomes law, virtually untouchable not just by the litigants involved or the public, but by the elected officials who are supposed to be the supreme law makers.  

So we have this conversation going on within law, worldwide really in legal circles, as to how to improve the legal system's performance. Generally this goes by the name of "legal reform" or "access to justice." It can involve court rules review, alternate business systems and billing practices, changes in self-governance, easing access for self-represented litigants, and many other things. Several members of the judiciary are quite active in this legal reform conversation, as individuals, and sometimes as members of their courts or judicial organizations. As it is readily apparent that the interest in legal reform that these members of the judiciary bring to the table is deeply sincere, I admire their initiatives. 

Only this: 80% of the attention is being paid to 20% of the problem - because no one, not even the judges, talks about uncorrected, unresolved judicial error, which is 80% of the problem.

Actually, it might be more accurate to say that 80% of the need for legal system reform is created by the actions of the judiciary - not every one of which is an error. But by virtue of such principles and doctrines as judicial independence, fully 80% of the issues that should be discussed under a legal reform rubric are declared off-limits.

This is not an unusual feature of organization or system change efforts. What usually happens when complex system change is undertaken is that the easy changes get tackled first. This seems logical, of course; why would you not pick the low-hanging fruit? The answer is simple: to actually change a system, you have to change its most powerful elements, not its least powerful elements. Otherwise, the dance of change becomes repetitive and ultimately useless. 

Here's how it goes. The easy changes are things like fiddling with rules, altering departmental reporting procedures, editing forms, and inevitably, forming committees. And writing reports. And however much these things might feel good, by definition they cannot trickle up the power gradient - that is the whole point of power, to control other people and not to be controlled by them. And because they do not trickle up, real, fundamental change never happens. Whether the change agents in the organization realize this or not is immaterial; what is material is that they have no choice but to do the easy things, and then to do them again. Judicial decision-making doesn't change after rules review, so you redraft the forms. Then you run some seminars; do a staff retreat maybe. Hold a conference. Still no change? You do another rules review. After all, if you are an internal employee, you have to earn your keep. Or if you are an external consultant, you have to produce something for your fee.

This is not to say that some change in system culture is not eventually achieved, and that there is never any change in judicial actions. The fact that there is some change is evidenced by the advent of the "activist judiciary." That is not the same thing, however, as reducing errors. It may simply mean that different, but equally untouchable, errors are made.

In effect, if you pick the low hanging fruit, you never bother to buy the ladder. And if you don't buy the ladder, you can never get the fruit at the top. And to take that metaphor to its extreme: it eventually rots.

I have watched education reform for 20 years, and studied its 150 year history quite thoroughly. Ah, you may not have thought that education reform had a 150 year history, even if you knew the system itself is that old. But the fact is that public education systems have been the subject of constant reform campaigns, some successful, some not. The instructive value of studying that history is observing who was successful at generating change, and why. Invariably, successful change campaigns have come from organizations that control how teachers teach, and there are only two organizations that have that power: the university education faculties that have a monopoly on teacher training, and the unions that have the power to communicate with teachers throughout their careers. Education reform that opposes the will of these two organizations has not ever succeeded even if it came from the apparent controllers of the system itself, and never will. And the tragedy is that these two organizations profit (ie grow) from teacher, and thus student and community, failure, not from success. Thus, the successful campaigns were not those that changed the system for the better. In 150 years of education reform efforts, no one inside or outside the system has been able to address and comprehensively deal with 80% of the problem, thus public education systems remain an (apparent) failure.

The lessons from education reform that can be applied to legal reform are thus both simple, and unexpected. The first lesson is not to look where you think the answer is, but really to examine the system for the source of power, which may also not be what you expect. There is a vast belief in law in the value of pro bono work, for example. The entire premise of the Access to Justice movement sometimes seems to be that if only the government would fund a lawyer for everyone, everything would be fine. But the self-represented litigant phenomenon has shown that access to lawyers is not the problem: you can get in front of a judge without one. The flaw in the pro bono ointment is that it is not lawyers that litigants want to patronize, any more than people who seek medical attention come to hospitals to patronize nurses. The ultimate goal of litigation is to see a judge. And the problem, as every report done on the SRL experience will tell you, is what happens when you get in front of a judge. 

It is not unique to the SRL phenomenon that things go wrong in courtrooms. With or without a lawyer, the brutal reality is that judges make mistakes, and those mistakes frequently remain uncorrected. SRLs are simply, I believe, making that reality so evident that public confidence in the courts can no longer survive the continued failure of the judiciary itself to tackle this issue. It is the flip side of the transition to an equality-minded world that the myth of judicial infallibility cannot convincingly be sustained. 

If the legal system is to reform in the direction of improved access to justice, then the frequency with which justice is actually perceived to have been dispensed will have to increase.

One must acknowledge, of course, that the litigation process does not serve only individual outcomes, but the development of the law, and even unjust outcomes can serve this larger purpose. In fact, the development of the law and the importance of the rule of law are used to uphold the untouchability of judges.

But this reasoning does not withstand scrutiny, for one simple reason: errors are the single best learning opportunities that individuals, organizations, or systems can have. This means that better catching of errors will create better judging and better law.

Surgery provides a perfect analogy. Surgeons learn how to do things right, but as Gawande's writing makes obvious, mistakes get made. As in aviation, a model he uses in the Checklist book, understanding the errors goes a long way toward innovation in terms of preventative strategies and vigilance. So, while no one undergoes surgery to experience an adverse outcome or to have a surgeon make a mistake, these adverse outcomes or mistakes contribute significantly to the development of surgical technique.

But only if they are caught, and responded to in a systematic way. In surgery, this happens automatically: it is impossible to hide error. In judging, hiding error that is not put through the appeal process is the order of the day. Thus, the learning potential of all those errors is lost to the system: it cannot be a learning opportunity for judges individually or collectively.

Chances are, then, that reducing uncorrected judicial error will improve the quality of law. And even if the law does not substantively go in a different direction, increased attention to unrecognized error may well increase the pace at which the law now evolves.

In pursuit of system improvement, then, the key question is, who really controls what you want to change in the system? That is where the change efforts need to go (or where they need to start). And because that is where the real seat of power is, it will be the most heavily guarded and defended, the most hostile to change, and the last place anyone who wants to change a system will want to go, because it will be career suicide to go there. 

If the judiciary chooses to continue to circle the wagons and point its own reform efforts at the peripheral 20% of issues rather than at the area inside the wagons, of course the sky will not fall. The only cost of delaying change is the combined cost of the institutional decay - ie, the cost that uncorrected errors extract from those who suffer them; the cost of the useless window-dressing reform efforts, and the accrued vulnerability of the whole system as a result of pent up need for change. But as education reform shows, these costs can be turned to good account, serving the subject system. The budget of public school systems and their offshoots is likely about five times what the actual cost of simply educating children would be. Circling the wagons, rallying the troops, and disguising failure are lucrative lines of work, employing thousands of people with high status. The whole economy benefits enormously from school failure, and that is not only in compensatory services, repetition, or the cost of illiteracy and innumeracy in the work world. In the US, tragically, where the school-to-prison pipeline is well-established, incarceration is one of the largest industries in many states. In short, the education system is not a failure. It is a brilliant success. Just, not at doing what it is actually constituted to do.

In the same way, in a trend already observable throughout law, a failure to reform judicial practice will breed system growth. All systems; all organizations, have but one primal urge, and that is to grow. There is an unfortunate tendency, even in law, to grant a mantle of beneficence to systems that are deemed "non-profit." But there is no such thing as non-profit - there is only profit that can be taken out as cash by the owners, and profit that cannot be taken out. In the latter case, profit presents as system growth. 

Perhaps it is because the urge to grow (vs. to profit) is not acknowledged or understood that it is so hard to ferret out, isolate, examine, and address. Why it so often swims around under the surface of organizations, never named, never acknowledged, never understood. And why its interdependence with system failure - or error - cannot be addressed in system design. 

Fortunately, error is not the only route to system growth, and system growth is not inherently a bad thing. But a system should grow as a result of success, not of failure. And in order to differentiate the two, and to structure incentives for system behaviour accordingly, the existence of errors must be acknowledged.

Which reminds me of another great book, this one a much lighter read than all of the foregoing:

(edited July 27, 2016)