At this late date, with the last of the Abscam trials behind us, it is still deficult to evaluate the FBI performance. Abscam remains a pyrotechnic display, an unprecedented event, something that could not have happened under J. Edgar Hoover. Hoover sought favorable treatment for his bureau from Congress. He courted some members, intimidated others and worried nearly all with his formidable powers. But, absent extraordinary circumstances, FBI agents in those days didn't ferre out crooked congressmen. Hoover knew where he wanted to draw the line, and he drew it in concrete.

His successor, William Webster, has drawn the line in a very different place. Whether he has chosen the right spot is my question. By the traditional yardstick, he deserves much credit: All the Abscam defendants who have been tried have been convicted. Although the Arab impersonations give the investigations a bizarre cast, and the supervision of informants at times appears lax, criminal investigations, it must be understood, are messy, unpredictable affairs, seldom running to a predetermined course. The fact is that Webster did what Hoover would not have done, and he did it reasonably well.

But there is more to the matter. Viewing Abscam as a whole leaves one with the uneasy feeling that the bureau got the wrong guys. Not "wrong" in the sense that innocent men were convicted -- these men all seem guily -- but rather "wrong" in that Webster may have expended the bureau's already diluted authority and prestige on an effort that simply wasn't worth it.

In American policing, the decision to initiate an investigation is the unusual one. Because there are so many more opportunities to apprehend violators than resources permit, underenforcement of the laws is the central characteristic of our system. And this is nowhere more true than with regard to consensual crimes. As to such offenses as drug dealing, gambling and prostitution, the law enforcement goal is not to banish every instance of illegality. It is more modest -- to keep things from getting out of hand. The aim is deterrence, not prohibition. No police agency could deploy enough personnel to locate the point when added manpower no longer produced a commensurate increase in arrests. That would stretch agency resources (and public tolerance) to the breaking point.

A bribery case calls for great investment in time and personnel. This is because bribery is a consensual matter; there is no complainant. Both parties are willing participants: One wants to sell, the other to buy, and neither wants the police to know about it. If the police do learn, it is because they have managed to find out what the parties wanted kept secret. And they have done so by means that are peculiarly susceptible to abuse: undercover agents, paid informants, electronic surveillance. As the Abscam investigations illustrate, there are no nice or easy ways to make bribery cases.

In this setting, investigations must be sharply focused. The goal is to get the drug importer rather than the street vendor, the lottery banker rather than the numbers runner, to strike for the jugular whenever possible. Certain offenses -- marijuana usage on campus, for example -- are so common that the police can effortlessly accumulate arrests; they refrain from doing so because of their obligation to look into more serious matters. If making corruption cases on Capitol Hill is like making pot cases on campus (as the recklesness of the Abscam defendants, and much popular opinion, suggest), then it is unclear whether FBI Director Webster has served us well. We need to know if the men arrested were more blameworthy than their colleagues who were not subjected to FBI enticements.

Taking $50,000 for letting someone into the country is hardly the kind of stuff that threatens the national welfare. The Abscam defendants sold their office, but only in a technical sense did they sell out their constituents.

But are there others who do peddle themselves on issues that really matter -- vital matters of defense or energy or the environment? And what does the FBI know of them?

Abscam seems to have neither improved our understanding of official misbehavior nor whetted our appetite to learn more. Nobody wants to see these men get off, but neither is there a public cry for digging deeper, for finding out more. It can't be that all the remaining 528 elected representatives are, as it is sometimes said, "a bunch of crooks," but neither can it be confidently asserted that the Abscam defendants differ from other members, except perhaps in their carelessness and lack of sophistication. In short, whether Abscam-type transactions are rare or routine remains a matter of conjecture.

As to the future, it would be surprising if the bureau were to do it gain. Abscam appears to be a one-time thing that punished a few thieves without providing the country -- or the FBI -- with a better map of the scope of official corruption or ways of coping with it.

It is not enough.