[This is an excellent example of the interweaving of ideas about institutional competence and the notion of "purpose" to build an argument about the bounds of judicial deference.]

. . . But I would urge that there is a presumption . . . that a rule-making or order-making agency with a specialized jurisdiction is meant to have some discretion, some policy-making function.

It is at this point that we come to the heart of the difficulty. Discretion, as we have said, is not self-defining; it does not arise parthenogenetically from "broad" phrases. Its contour is determined by the courts, which must define its scope and its limit. Put in terms of judicial review: why do courts in some cases pronounce the law and in some not? Put in terms of administrative power: why are agencies in some cases permitted to make law and in some not? The answer, I submit, should run primarily and presumptively in terms of clear statutory purpose, or as Professor Fuller would say, "the intention of the statute." The phrase "clear purpose" has certain doubtful connotations. The word "clear" suggests a clarity, an objectivity of application which quite evidently is not present in these situations. What it means very simply - if simplicity has not by this time been quite frightened away - is that where the judges are themselves convinced that certain reading, or application, of the statute is the correct - or the only faithful - reading or application, they should intervene and so declare. Where the result of their study leaves them without a definite preference, they can and often should abstain if the agency's preference is "reasonable."

I am aware that an answer in terms of statutory purpose will appear to be either a crashing platitude or a resounding rationalization for results otherwise determined. But I am convinced that it is a valid, if obvious, initial premise and that it seems a useful function to demand that a court make explicit reference to it. The first function of the criterion is cautionary. It is meant to assert the responsibility of the judge where an administrative action is in question to bring the action under his own scrutiny in terms of statutory purpose. If it is agreed that this is his duty, it may be argued that there is no occasion to insist on something so trite. But I have the impression that in recent years some judges - quite distinguished judges - when faced with an administrative action to which they can attribute expertness, fail to focus the issue of statutory meaning. . . . The second function of my suggested formulation is more ambitious. It is intended to suggest a presumptive limit of judicial power and in terms of that limit to offer an analysis in which the cases can be discussed in terms of a rational and consistent attitude. When we say "discussed in terms of" statutory purpose, we have in mind a concept of statutory purpose which takes account of the fact that the legislature in realizing its purposes has chosen to work through an administrative agency, and so (presumptively, as we have said) to confer on it some policy-making function. This discretion should normally be permitted to function short of the point where the court is convinced that the purpose of the statute is contradicted. The court, we have said, should test each exercise of power in terms of statutory purpose. But in a great many cases the court will grant that any one or two or more proposed answers is consistent with the statute. In such a situation its function is complete when it decides that the administrative answer is reasonable, unless the presumption yields to certain considerations considered below. The agency is in as good and, because of its specialization, presumptively in a better position to make the choice. There is, furthermore, a value in this recognition of administrative autonomy; it may invigorate the sense of responsibility, stimulate initiative, and encourage resourcefulness.

...
 

It will no doubt be urged against a test in terms of purpose that it gives no predictive clue to the likely result in a new case because there is no rule for the ascertainment of purpose. That is indeed true, as it is of all statutory interpretation. The test channels the search for an answer but the answer ultimately depends on the appreciation of the particular judge. Furthermore, though the purpose test be the primary or basic criterion there are additional considerations which determine its application; some of these considerations are implied in the test of statutory purpose, some operate outside of it. Among them are (1) the degree to which the framing of a rule appears to depend on expertise, (2) the clarity with which a rule can be made to emerge and be given a stable form and content, (3) the importance of the rule in the statutory and administrative scheme, (4) the possible psychological advantage of judicial as compared with administrative pronouncement, and (5) the role of the court as the guardian of the integrity of the legal system. Thus, there will be cases where though the rule to be laid down or the decision to be made does not clearly emerge from a study of the statute, nevertheless the court will - indeed should - take upon itself the power and responsibility for decision. . . .