Why Justice Goes Awry
If a morbidity and mortality analysis were done on each miscarriage of justice, we could probably chalk up the results to any number of imperfections—disproportionate access to evidence, disparate advocacy skills, a misunderstood question, a mis-phrased answer, a key document that somehow disappeared and never became part of the evidence, a witness whose distorted memory was persuasively communicated and unjustifiably believed, an ambiguous email that created a false impression, a litigation budget that sank under the weight of crippling discovery, an arbitrary evidentiary ruling, a confusing jury instruction, an unfortunate gap between what someone said and what that person meant, an adjudicator whose hidden biases led to an erroneous credibility assessment or a mistaken legal ruling. The list goes on and on.
In an adversarial system, the search for truth is a battle of narratives. The side with the more sympathetic, more plausible story usually wins, even if the truth belongs elsewhere. Generally, our adversary system favors the better story, not necessarily the truer one. Emotion prevailing over logic.
Fact finders believe the story they want to believe, the one that’s easier to imagine. That becomes their truth. If they don’t feel good about some other story, they won’t believe it, even if it’s the actual truth.
One hallmark of our system is the interplay between passive actors and active ones. Adjudicators are passive; lawyers and witnesses are active. For the most part, adjudicators are spectators, watching the lawyers and witnesses make their presentations until it’s time to reflect on what was presented, decide the case, and announce the winner.
The final decision might or might not be the same as what a faithful application of the law and a fair assessment of the merits would produce when applied to the real facts. It will only be what the adjudicator thinks is the right outcome based on however the adjudicator interprets the evidence that the litigants were able to present, interpretations influenced by the adjudicator’s life experience and personal biases.
Adjudicators will resolve conflicts about the facts in the privacy of their own thoughts. Fairness assessments will be shaped by personal standards and an individual sense of right and wrong.
But how is the adjudicator’s interpretation of the facts also shaped by the adversarial process that develops them? Do aggressive questions that are put to nervous witnesses, and summations that appeal to emotion rather than logic, really deliver objective truth?
Or, despite the good intentions of lawyers who genuinely believe that their questioning is in service of the actual truth, does the process instead show the adjudicator, on occasion, only a distorted version of it? Or two distorted versions? Or a truth represented by only some key facts? Or a truth with makeup on it? Or a truth without important context? Or something far afield from the truth?
If we can rationally accept that, at least on occasion, the adversarial process shows the adjudicator a version of the truth that meaningfully departs from the actual truth, then we must confront this question: How often does that happen?
Could we say with any measure of confidence that the adversary system reliably leads the adjudicator to the actual truth at least 95 percent of the time? Or 75 percent of the time? Less? More?
And if we could measure not just frequency but the degree to which outcomes diverge from the actual truth in any given case, how often would that divergence be greater than, say, 25 percent? Or 50 percent?
All of which leads to this essential question: What are the acceptable metrics of unreliability? How much disparity between outcomes and actual truth should we tolerate before we conclude that the particular truth-finding mechanism is flawed? Or needs refinement or fixing? Or needs to be swapped out altogether in favor of something else?
Unfortunately, we have no way to measure the reliability of the adversary system as a truth delivery system. There is no empirical way to know the degree and frequency of disparity between perceived truth and actual truth—in essence, what we should understand is our adversary system’s margin of error. Is it better or worse than, say, a polygraph machine?
Any evidence we have about this is purely anecdotal. It rests on unreliable sources—disgruntled litigants, or lawyers who contend that, in the cases they lost, the adversary system did not do its job. The winning side always thinks the system worked correctly. The losing side, not so much; perhaps never.
And in a close case, the winning side might think that, but for a twist of fate, the adversary system could have failed them terribly as well.
Presidential Debates
Perhaps we can draw some meaningful opinions about the reliability of adversary systems generally by looking not at the adversary justice system but at a different adversary system: the presidential election system. Like judges and litigators, politicians are also accustomed to the idea that the best way to choose a winner is to put the contenders through an adversary process, not a trial but a debate.
The modern custom began in 1960 with a staid and civil debate between John Kennedy and Richard Nixon. In current versions, these debates are more rough-and-tumble. In primary season, they look like a chaotic circus, with candidates elbowing out each other for precious airtime to elevate their signal above the noise.
In that process, the candidates answer tough questions from journalists or each other, and the decision-makers—voters and pundits—assess those answers not just by their content but by how well they were presented. The process assumes that each candidate, through adversarial questioning, will give answers that prove his or her qualifications and merit. The candidate with the better answers will earn the voters’ trust and will rightfully become the nominee or the president, so the thinking goes.
But that’s not the reality. What the process illuminates well is not what type of president the candidate would be, but something else: how effectively a candidate can answer loaded questions. Or argumentative questions. Or how skillfully a candidate can deliver those kinds of questions, hoping the opponent will give a bad answer or tender up an unflattering sound bite. Or how cleverly a candidate can smear an opponent.
Look, for example, at how the Democratic candidates in the early debates this year tripped over themselves to show how aggressively they could attack each other or grab airtime. The debates had the feel of a verbal free-for-all or boxing match. The morning-after headlines were not pretty: “Bloomberg rivals pounce on stop-and-frisk.” “Things turn ugly between Klobuchar and Buttigieg.” “A whiff of desperation in the Democratic pile-on.” “Democrats hurl stinging attacks across the stage.”
For those who were hoping that the debates would show the Democratic field at its best, what they got was a Democratic field at its worst. A triumph of short-term tactics over longer-term strategy. Of self-interest over collective public good.
Of course, questions in a political debate are more loaded and leading than those at a trial, if such a thing is possible, and the rules of engagement more ad hoc. But these two adversarial processes—trials and presidential candidate selection—share something significant. Their principal utility is in how they separate those with superior tenacity or verbal tactical skills from the rest of the pack, though that doesn’t necessarily identify who deserves to win.
Presidential debates seek to reveal who will be the best president. Trials seek to reveal the truth. Instead, what they both do well is test how skillfully someone can deliver self-serving answers in a high-pressure setting while protecting or polishing one’s image. People with superior skills in that department typically come out on top, regardless of whether that’s merited.
These adversarial processes also have in common that they do not reward actual truth but rather the perception of truth. Generally, the candidate or litigant prevails whose overall presentation is more appealing, whose story is more desirable and easier to imagine even if that’s not necessarily the way it really is or was.
The list of commonalities goes on. Protagonists lose when they’re hit with a cheap shot. Personal flaws and verbal gaffes take on outsized importance. The adversary process magnifies them. They lead to loss of stature and suggest impaired credibility or diminished trustworthiness. Clever zingers displace reasoned analysis. Trivialities and sensational sideshows obscure relevant facts. Likable personalities attract a decision maker’s favorable attention, often unfairly. Candidates or litigants with an unpolished edge or clumsy presentation will falter, even if they might be worthier on the merits.
This isn’t to say that the adversarial justice system is necessarily bad. Or that a polygraph machine would be better or more reliable. Cross-examination often exposes faulty memories and falsehoods. Arguments from advocates often clarify confusing facts and help lead decision makers to fair outcomes. But the playing field typically gives the advantage to the more skillful interrogator, to the better witness, and to the superior orator, not necessarily to the side armed with the actual truth.
A New System
We seldom question whether the justice system as we know it is designed to produce the objectively right outcome. After all, we, the current generation of lawyers and judges, inherited this justice system; we didn’t invent it or build it. We took it as we found it.
Few can pinpoint when it really began or who conceived or created it. Like life itself, it just evolved. It developed organically, based on how disputants over the ages thought it should be shaped or changed to serve what they thought would produce fair and true outcomes. They worked with what was handed down to them, just as we do. They tinkered with it in small steps. Now, as then, it’s still a work in progress.
But if we were writing on a clean slate, unencumbered by history, would this be what we would design for ourselves? Or might we fashion something else, like a system in which the adjudicator is not a passive observer of a contest between dueling adversaries but more like a detective, actively engaged in asking the questions and investigating?
In such a system, the disputants would still have the chance to present their evidence and stories to presumptively neutral adjudicators. But the adjudicators would take the investigative lead—they would ask the first questions, requesting additional information as needed to fill in gaps or resolve inconsistencies.
Fact-finding could be a more iterative and transparent process, a give-and-take between adjudicators, lawyers, and witnesses, in a more collaborative journey aimed at uncovering the actual truth, rather than at developing a reasonable facsimile of it. Disputants and their counsel would have a good sense of what was troubling the adjudicators in real time. They could address those concerns by offering additional evidence or by having a conversation.
To be sure, adjudicators would need to resolve credibility issues, but they would have as much of a role in that effort as the lawyers, each asking questions, conducting cross-examination, and developing evidence, instead of being limited as adjudicators to hearing only whatever the disputants chose to present.
This is not a totally unfamiliar model. Arbitrations sometimes do this. Specially tasked commissions do this as well. So do inquests. Corporate internal investigations use many elements of this.
But whether this model on a broader platform would be feasible, scalable, affordable, or desirable, and whether it would be safe and effective at the retail level, are hard to know. Could it be implemented successfully? Would stakeholders have confidence in it?
To move from what we have now to this or any other new model would require years of study, honest debate, thoughtful planning, some experimentation, and careful execution. And it would require overcoming our innate resistance to change.
So at least for the rest of our lifetimes, we seem destined to live with our well-known adversary system. Despite its imperfections, we’re content to assume it’s sufficiently reliable, although there’s no hard evidence to show how reliable it is.
We could take comfort in clichés like “let’s not let the perfect become the enemy of the good,” or “the devil we know is better than the devil we don’t.” On a macro level, we’re content to accept that there will always be some margin of error, if the margin of error is low enough. But “low enough” has yet to be defined, and it doesn’t seem that anyone is trying to figure this out.
On a micro level, though, no error is acceptable. For the victim who received no compensation because of the system’s failures or for the blameless defendant who suffered a major loss because the adversary system delivered the wrong result, we might justifiably wish there had been a better way.
Comments about On Reconsideration may be sent to [email protected].