<< The Most Transparent DOJ in History | Main | Seeing It Up Close Has an Effect >>

Complexity, the Enemy of Justice

Kent yesterday had an entry in which he noted that a man who shot and killed his wife and her brother twenty one years ago was due to be executed, but was fighting it off with a (typical) blizzard of last-minute procedural motions.  As Kent put it:

So when you are stone cold guilty of a crime for which the death penalty is clearly appropriate, what do your lawyers argue about at the last minute?  Drug expiration dates.  Really.

This got me to thinking:  Why has criminal justice system  --  not just capital punishment, although certainly that too  --  tied itself up with manufactured procedural issues that wander at increasingly huge distances from the central question:  Are we being careful enough to be sure we've got the right guy?

I think it's because we've become entranced with the idea that moral confidence in the system requires perfection or something very close to it, and that perfection requires the kind of microscopic complexity that now stretches on year after year after year.

But this is all wrong.  The quest for perfection is a fool's errand no matter what our punishment scheme is.  For starters, it's unattainable.  And it has reached the point that the complexity it spawns produces more injustice than it averts. Prof. Richard Epstein explains the point brilliantly in the excerpt after the break.  (Simple Rules for a Complex World,  Harvard University Press, 1995; emphasis mine).

The desire for justice in the individual case is one of the strongest forces shaping the operation of legal systems. A great deal of law is made or applied in the context of individual cases, before both judges and administrators. Within that framework, the stated function, as well as sworn duty, of the judge or administrator is to apply the general law to the facts of a dispute in order to reach the correct result. An error in outcome is a miscarriage of justice which undercuts the moral authority of the law and reduces its effectiveness in the control of human behavior. It is therefore understandable that to avoid these injustices considerable resources are devoted to developing procedures that allow for the orderly presentation and evaluation of the relevant evidence.

The demand for justice is often regarded as sacred, as a good beyond price. Unfortunately, in one critical respect the thirst for justice is similar to the demand for any other desirable commodity. The demand for justice is subject to the law of diminishing returns. The initial improvements in the legal system come easily: the law can give all persons notice of the charges that are brought against them; it can give both sides an opportunity to present evidence and to cross-examine the witnesses for the other side. Implementing these common rules of natural justice will lead to major improvements in the legal system, relative to one that contains no procedural safeguards.

But perfect justice demands more than incremental improvements in the operation of the system. It aspires to rooting out error in every individual case. Simple rules do not meet that exacting standard. At best they are only tests; and tests are rules of thumb that work most of the time, but
are known and expected to fail some of the time. To adopt a simple rule therefore is to make an open acknowledgement that the rule cannot be perfect because it will generate an unjust outcome in at least one case, and doubtless in more. Even if we cannot identify which case that is, we know that the simple rule decisively thwarts the goal of perfection.

Complex rules avoid so gross and impolitic an admission up front. If the law can identify enough factors, can indicate the ways in which they should be taken into account, can specify the appropriate burden of proof, and can provide for the exhaustive collection of evidence, then maybe, just maybe, the legal system will reach the heady level of perfection to which it aspires. The upper potential of performance in a complex system is always higher than it is in a simpler system that confesses its weaknesses at the outset, and accepts not merely the possibility, but also the certainty, of bad outcomes and compelling counterexamples.

Yet the gains from seeking perfection are an illusion, for with complexity come the opportunities for gamesmanship that will be part and parcel of social life so long as resources are scarce and individuals are motivated by
self-interest. The relevant comparison between simple and complex rules should be conducted not in the language of aspiration, but in the language of realizable achievement. It is that more humble task which simple rules best discharge, for their relative cost-effectiveness and certainty forestall the vast amounts of intrigue brought into the legal system by the relentless, if naïve, pursuit of perfection. The only question for the legal system is how it will make its errors, not whether it will make them. Simple rules are adopted by people who acknowledge that possibility of error up front, and then seek to minimize it in practice. Complex rules are for those who have an unattainable vision of perfection.


[N]o set of rules will be perfect in its application; indeed, knowing when to quit is one of the driving forces behind a set of simple rules. Nonetheless, even though there are some daunting exceptions, [simple] rules do have the virtue of offering solutions for 90 to 95 percent of all possible situations. Never ask for more from a legal system. The effort to clean up the last 5 percent of cases leads to an unraveling of the legal system insofar as it governs the previous 95 percent. 


I call it paralysis by process. I am sure I stole it from someone else, but I don't remember who.

But Mr. Otis, there are many, many players in the criminal justice system who don't agree with the central question you give. For example here in CA spending time and money on the perception of fairness often appears to take the front seat over actual fairness (i.e. do we have the right guy).

David --

The folks you mention in your second paragraph need to wake up to the fact that, when we go too far to achieve the perception of fairness, both fairness itself AND THE VERY PERCEPTION THEY SEEK will become damaged. People might well think that it's worth five or six or seven years of review to insure we have done things both accurately and fairly, and are seen to be doing so. But after that -- and certainly with the extremes that now exist in California -- people will correctly come to see that (1) as to the killer, it's not fairness but just delay for delay's sake that's going on, and (2) as to the victim's family, it's gratuitous UNFAIRNESS, an unfairness that continues to grow and become more painful as it stretches out to absurd extremes.

Leave a comment

Monthly Archives