IT'S AN OLD and complex police question. Can or should the performance of police officers be graded on the basis of numerical ticket or arrest "quotas"? Without "goals," "objectives" or other measurements, will officers be as efficient as they can be? Here in Washington, a new police department "performance evaluation" program scheduled to begin today contains some numbers that trouble diligent members of the force -- of all ranks.

According to one section, a patrol officer who over a four-week period fails to achieve one criminal arrest, two traffic or noncriminal arrests, 50 parking tickets and 10 moving or pedestrian violations will be considered "unsatisfactory." An officer who produces at least two criminal arrests, six traffic or noncriminal arrests, 150 parking tickets, and 50 moving and pedestrian violations rates as "outstanding."

On the surface, these are not outlandish measures. An officer who regularly fails to accomplish the tasks in that first list probably has a problem. The officer who regularly meets the higher numbers may be doing a whale of a job. But should there be pressures on all officers to get their numbers up -- no matter how? What might that take away from other important policing tasks that are not so readily measured?

Should a patrol officer who spends time handling family disputes arrest somebody -- everybody -- quickly and file a report, or spend a little time to mediate? How about the officer at the end of a four-week span who "needs" another 75 parking tickets or five more traffic arrests. Should he keep an eye out for red-light runners, or spend a shift ticketing cars that are parked a foot too far from the curb at 3 a.m.?

Goals -- meaning some indications that laws are being enforced -- are worth setting forth for police commanders and their patrol officers. But as one expert in police work notes, "if you demand numbers, people will produce them -- one way or another. It forces officers to produce numbers irrespective of need."

Except for this section, the new police program is an improvement over the current system, which is based on more subjective criteria -- loyalty and initiative. Capt. Robert Scanlon points to other useful job-related standards in the new program, including an officer's behavior at crime scenes and the quality of follow-up reports.

There is leeway for discretion, too. The various categories are to be graded and then weighted according to importance, with an average assigned annually. Evaluations will not be used for promotion or salary purposes, but two consecutive unsatisfactory evaluations could lead to dismissal. Capt. Scanlon notes that the criteria can be adjusted -- and therein lies the key. Good sense and thoughtful administration can go a long way to balance productivity with effective policing.