One plane pops a hole, no one is seriously hurt, and, as there should be, there’s saturation media coverage, immediate inspections of a portion of the fleet, and deep soul searching about our preventive systems.

But DNA exonerations suggest that we are making serious mistakes — with deep impacts on many lives — all the time — and are not doing much to find of prevent them.

Assuming this is true, why do we allow this to happen, and what should be done?

One answer as to why has to do with the inevitably relativistic nature of the decisions — there are always two sides, so nothing seems right or wrong.  Moreover, precisely because there are two sides, the desire for finality is great.  Just think about the narrow scope of review in most appeals, not really looking for the right result, only the right process.

But maybe the time of acquiescence in such low expectations should be long past.  It is certainly the case that the DNA exonerations have shown that finality and infallibility are not necessarily the same thing.

Some thoughts on ways we might be able to raise our own expectations of the system.

• Measurement. Developing much better ways of tracking and counting outcomes so we can at least know what happens — including follow up for recidivism, returns to court, etc.
• Outcomes. Thinking about acceptable rather than optimum outcomes.  One of the reasons analysis is so hard is that it is hard to decide what the “best” outcome is — particularly cases that involve judicial discretion.  It might be that we need to be more willing to think about reviewing cases as to whether they are outside the acceptable range, rather than the “right” decision.  This standard has been described as the “Oh My God,” standard, and as the “substantial risk of a miscarriage of justice” standard.
• Data Mining. Using new analytic tools that allow one to find underlying patterns in masses of data — are we seeming patterns or changes in child custody decisions, such a differences between states, courts, judges?
• Rethinking Appellate Review. Many states look at only very narrow issues on appeal — and require contemporaneous objection — a practical impossibility for most of the self-represented.  How often does the word “objection” get used in a self-represented hearing?
• Routine Review of Decisions.  If litigant-led, objection-driven appellate review is not providing the safeguards that are needed, maybe we need to think of other (quieter and more efficient) review systems.  Most work in the world is improved by a second set of eyes.   The trick is finding a way to do it that does not further denigrate the role of the judge.
• Interlocutory Appellate Review. As an apellate practitioner in Massachusetts, I was very impressed by the broad availability of immediate interlocutory review in that state.  As a practical matter, almost any lower court decision or ruling can be brought up for (highly discretionary) review by one or other of the two appellate courts.  The world does not end when judges are second guessed — actually it gets better when they learn from mistakes and avoid potentially much more public reversal later on.  Mass statute granting general superintendence power to state supreme court. Mass Appeals Court Guide with statutes.
• Extension of Concept of Fatality Review. Many states now have systems of fatality review, in which, after a domestic violence death, the performance of all the system is looked at to see where the failure of prevention occurred.  Maybe we need to expand this to other areas of social failure.  (One would be where abuse happens after a custody grant.)
• Consideration of Relationship of Caseflow Management to Errors. Individual assignment calendars are usually considered a good thing — and they surely are — as ensuring continuity of understanding by a judge.  But if and when a judge “gets it wrong” then they tend to lock in the error.  Thought should be given to ways to minimize this, and to other ways that caseflow management should go beyond speed to error reduction.

In the end, however, we have to build self-assessment systems that treat every problem as an opportunity for system improvement — not blaming on or another player.  That is the way the airline safety system works.  This is much harder to do in a system such as the legal system, that has historically been about the assignment of blame.