In truth, all regulatory agencies waver between their civics-textbook mission to protect some form of the public interest and a far more unfortunate function of forming a gauzy illusion or protection across conduct the public would not tolerate if it did not think that it was being protected by the regulators. The essential element in any regulatory system that reassures more than it regulates is that it have an immense capacity for self-delusion. Let me describe some of the elements of that capacity and leave it to you to decide whether they look familiar.

The agency's role must be heavily reactive and defined largely in terms of cases brought to it by those whom it regulates. It is helpful in this context if the relevant legislature and press have a strong tendency to evaluate the agency largely in terms of permits issued and rate cases processed.

The agency's budget must be a tiny fraction (less than 1 percent) of the gross revenues of those whom it regulated. As a general rule, the total salaries of the commissioners should not equal much more than half of the salary of the chief executive of the largest regulated entity.

Regulation should be on an "audit" basis, examining only a small fraction of the total number of accounts or plant designs or operating practices of the regulated entities.

The system should deal with its critics more or less the way the Tar Baby dealt with B'rer Rabbit. It should have an almost infinite capacity for repressive tolerance, the extending of exquisite procedural courtesy to participants who are never, in fact, allowed to get their hands on anything vital. This can be expected to frustrate critics to a point at which they become obsessed or shrill or demagogic or a little crazy. Then, of course, their arguments are more easily dismissed as obsessed, shrill, demagogic or crazy, especially if others who had these characteristics all along have at some point been attracted into the fray.

Almost a corollary of the preceding point, intervenor funding of any substantial sort is to be avoided like the plague. Discovery and crossexamination must, of course, be allowed on some occasions, but the occasions can be limited and the examining boards constantly reminded of the need to keep the issues narrowly defined and the proceedings moving along. Embarassing questions from other forums can sometimes be avoided by preemptive legislation, such as that which denies the states any role in setting the standards for the radioactive exposures that their citizens will face.

The enforcement system must be a relatively benign one, with only the most distant possibility of a truly severe penalty. Infractions or testimony that tend to mislead the regulatory agency in ways short of outright perjury are dealt with, if at all, under a lenient point system or with fines that bear only a trivial comparison to annual revenues unless the violator is unlikely enough to be quite small.

Though it is not essential, it is very helpful to have the system reinforced by courts that defer copiously to agency expertise in the face of challenges from citizens groups, but that bow with equal fervor to management prerogatives in the face of challenges to the agency from the regulated entities. Should a commissioner be overly rambunctious in publicly questioning the system as it works in particular instances he can (as Chairman Pertschuk of the Federal Trade Commission recently was) be ordered to disqualify himself from further proceedings, though I think one would search a long time to find a case in which a commission had been disqualified for a speech in glowing praise of a regulated technology.

Research programs that would actually test the validity of the underpinnings and hypotheses of the regulatory system are rarely, if ever, undertaken. This is ostensibly because they would be too costly. Policy-evaluation offices are avoided or hamstrung or told what not to evaluate.

It can also be expected that the regulated industires will utter frequent expressions of anguish at the toughness of the regulatory regime under which they labor. In this, they do run the risks of the boy who cried wolf in the event that they are ever faced with a regulatory injustice so great that it outweighs the advantages of the system.

The semantics that surround such a process are obviously critical to maintaining the illusions. Accidents become incidents; citizen become protest groups; ambiguous situations are held to offer "no evidence" of wrongdoing. Transfer payments from consumers or from subsidiaries to parents are labeled "taxes" for rate-setting purposes even when none of it actually goes to the government. The efficiencies of the free-enterprise system are constantly invoked despite the absence of one of its essential elements -- vigorous competition.