Who should judge the judges? Does anyone evaluate their work? Because society wants its judges to be independent, they are given long terms, even lifetime tenure, to insulate them from pressures that might influence their decisions. Occasionally, this system protects jurists who are incompetent, biased or lazy. So what is the best way to correct these problems while preserving the independence of the judiciary?

In some states, bar associations conduct polls; often the media make periodic assessments. These can range from pop lists of the "Ten Best and Ten Worst Judges" to more thoughtful and orderly reviews. In the District of Columbia, the bar regularly solicits comments from its members on the performance of local judges. The information is tabulated and sent confidentially to the judge and to the administrative judge of the court on which he sits. In order to assist all states wishing to pursue this kind of formal evaluation process, a special committee of the American Bar Association has developed guidelines that will be considered by the association's house of delegates later this month.

The premise of the guidelines is that judicial evaluation is primarily for the purpose of self-improvement, though it is also useful in helping chief judges make assignments, in developing continuing judicial education programs and, in some cases, in assisting those who decide whether to retain or reappoint judges in office. The evaluation process is not to be confused with the disciplinary process that addresses misconduct. It should involve a periodic assessment of all judges, and it should be under the supervision of the highest court of the state.

The guidelines suggest criteria: a good judge should be a person of high integrity who knows and understands the law and whose actions are free of bias, impatience and discourtesy. He should be a good manager who writes well and communicates effectively. Performance in these areas should be assessed by a committee or agency created by judicial authorites that would gather information from lawyers, other judges, public records, court personnel, litigants and other sources.

Some difficult questions must be resolved by each jurisdiction. How can the confidentiality of sources be protected if judges want an opportunity to respond to criticism? When should evaluations be made available to those outside the judiciary? What special needs for this information arise in states where judges are elected, and under what conditions should reports be provided to voters? Circumstances in 51 different jurisdictions will have to be considered in creating state programs, but the drafters of the ABA guidelines have done well in focusing attention on this subject and providing a useful framework for further discussion.