From 1999 to 2011, Randy Cohen wrote the New York Times column “The Ethicist,” in which he discussed everyday ethical quandries.  (Mr. Cohen was an Emmy-winning writer for David Letterman.  I don't know why he was writing an ethics column.)  In a column dated February 18, 2010, Mr. Cohen fielded the following question:

Q: A friend was caught by police radar going 51 in a 35-miles-per-hour zone.  In front of his children, he admitted that he was speeding but asked if I knew a lawyer to help him fight the ticket.  I think he should accept the consequences, learn from the experience and give his children a lesson in ethics.  He looked at me as if I were from Mars. Shouldn’t he just pay the ticket?

Mr. Cohen provides the following answer:

Even those who think themselves guilty are entitled to their day in court, and there is civic virtue in their exercising this right.  A trial is a way to hold officials accountable for their conduct.  Was the radar gun accurate?  Was the speed zone clearly marked?  Did the police officer behave properly?  And what, given all the circumstances, is an appropriate punishment?  Little of this could be scrutinized if everyone simply paid the ticket.  It would be a court-clogging nightmare if every self-confessed speeder demanded a trial, but it is a fine thing if, now and then, some people do.

Yours is a variation on a question sometimes put to lawyers, public defenders in particular, about the propriety of defending a “guilty” client.  Another answer is that “guilt” in this sense is a legal determination that can only be made in a court of law.

James Boswell, himself a lawyer, once asked his great mentor about the propriety of a lawyer’s “supporting a cause which you know to be bad.”  Dr. Johnson replied: “Sir, you do not know it to be good or bad till the Judge determines it. … An argument which does not convince yourself, may convince the Judge to whom you urge it: and if it does convince him, why, then, Sir, you are wrong, and he is right.”

As a former criminal defense attorney, I am sympathetic to the importance of ‘defending the guilty’; indeed, I defended someone I thought was guilty several times in my practice.  But Mr. Cohen is careless in his analysis here, identifying two reasons to justify his advice — both insufficient. 

• First, he argues that it is not the defendant's responsibility, but the judge’s, to determine whether the defendant is guilty.  (I’ll call this the “role-confusion” argument.)
• Second, he argues that the “system” (i.e., our mechanism for identifying and punishing speeders) must be periodically tested to ensure it is operating properly.  (I’ll call this the “system-testing” argument.)  

Each of those arguments seems fine in their own way and deserve to be invoked in some circumstances… but not in the ones raised by the questioner.  Let me explain.

ROLE-CONFUSION

The “role-confusion” argument – that it is not the defendant’s job to determine guilt, but the judge’s – is inapplicable in this kind of case.  Judges determine guilt only when there is a dispute as to whether the defendant is guilty.  Where there is no dispute, there is no reason to involve the judge to evaluate the accusation.  It is true that there may be cases where the defendant's own evaluation of his guilt is not trust-worthy – and these are the cases which Dr. Johnson refers to in the quotation.  For example, the defendant may be mentally ill, or protecting the true criminal, or unable to correctly evaluate a particularly technical or ambiguous law.  In those situations, the defendant’s own claims of guilt may well be incorrect.

For example, in intellectual property cases, the accused may think that he has deliberately stolen a protected idea, but he may not understand that the idea is actually in the public domain and no longer protected.  (The prosecutor might not understand that either.)  There are many variations on this theme, but they all depend on something missing in Mr. Cohen’s ‘speeding ticket’ example – the reasonable possibility that the defendant is wrong in his self-appraisal of his own guilt.

Speeding is a relatively plain matter.  Combine that fact with our natural tendency to see things in the light most favorable to ourselves, and the accused’s own conclusion that he was guilty seems fairly dispositive.  Absent additional facts, it is difficult to see how, in such a straightforward scenario, he could be wrong.  And therefore, it is difficult to see why he should hesitate to condemn himself, instead of imposing the significant costs of a trial on the judge, prosecutor, police force, support personnel, etc.  Someone with a genuine dispute about his guilt – and there are many defendants like that – needs those resources quite a bit more so than this defendant does.

SYSTEM-TESTING

The other argument – “system-testing” – fails for similar reasons.  The system certainly does need to be tested, and there is a civic virtue in ensuring that systemic failures are identified.  However, the entire purpose of system-testing is to identify actual systemic failures, which means that system-testing should occur in cases where there is more likely to have been a fault in the system.  All things being equal, we should be looking for systemic failures in the places where they are more likely to exist.

Now obviously there’s no point in testing the system only where we predict a failure will be found — our predictions may be just as fallible as the system itself.  But that does not mean that system-testers should be oblivious to how likely system failure is in a given case.  And a crucial marker of such a potential failure is the parties’ disagreement as to the defendant's guilt.  It is much more likely that testing the system will find a failure (such as a miscalibrated radar gun, a dishonest policeman, etc.) where the policeman and accused disagree whether he was speeding, than where they agree.

In Mr. Cohen’s example, there is nothing to make us think the system would benefit by fully trying this case.  If the defendant honestly believed he was not speeding, that would give us some reason to doubt the radar gun’s calibration, the policeman's honesty, and a host of other things.  But absent such a belief, the judicial resources would be better deployed testing the system in a circumstance where there is a genuine dispute.  Given the plentiful number of contested speeding tickets, the system will not go untested, and this trial would only have a marginal system-testing effect.  Better to test the system where there is some indication of possible failure.

Mr. Cohen should see this himself.  He notes that not every speeding ticket should be tried, but does not explain why this one should be.  If we shouldn't exercise discretion in determining which admittedly-guilty defendants to try, then why not try every case, whether or not the defendant pleads guilty?  Why not prevent people from pleading guilty?  We don’t do that, for the obvious reason that trying cases uses very valuable social resources that are badly needed in cases where there are real disputes.  The purpose of trial is to adjudicate between conflicting evaluations of guilt, something missing in this case.  It is self-indulgent and wasteful for this defendant to plead not guilty.  

If you wish to discuss this post with me, I'd welcome receiving an email from you.  Please email me at language.on.holiday@gmail.com.  

P.S.  It is curious that Mr. Cohen fails to discuss the ethical ramifications of pleading ‘not guilty’ even though the man believed he was guilty.  Numerous issues distinguish entering a plea from providing perjured testimony (or even a regular ol’ lie or fib or what-have-you), and as noted above, I agree there are circumstances where it is appropriate to plead ‘not guilty’ notwithstanding your belief to the contrary.  Yet it does seem it should give “The Ethicist” pause that he’s advised the accused to stand before a court of law, and say something he thinks is untrue.  At least a mention?