There is a nice passage early on in Amnon Rubinstein’s Jurisdiction and Illegality (Oxford, Clarendon, 1965). Responding to a claim of Hans Kelsen that voidness is a monolithic concept and arguing instead for a distinction between unlawful decisions that are void and those that are merely voidable, Rubinstein writes (at p. 6):
Even if we take the extreme case of a completely and manifestly unauthorized act, such as an arrest warrant issued by a private person, the validity of that warrant may have to be determined in an action for false imprisonment. By brushing aside the preposterous suggestion that such a warrant could be relied upon by the defendant, the court will in effect decide that the so-called warrant is a piece of scrap-paper. In this sense the warrant may be said to be voidable or annullable. But can it be classified under the same category as a warrant issued by an authorized magistrate which, because of some irregularity, can be set aside in appropriate proceedings? Certainly not.
In Rubinstein’s extreme case, the warrant is void, in the latter case, the warrant is merely voidable. But there are many more rounds one can play in Rubinstein’s conceptual game. What about the ‘authorized magistrate’ whose term has run out? Is that a “completely and manifestly unauthorized act”? Does it matter if the magistrate’s term ran out a year ago? Would the same result be reached if the magistrate’s five-year (say) term was later determined to have run out because (say) Leap Years have an extra day? What if the magistrate is an expert decision-maker to whom deference is owed and took the view, on a reasonable interpretation of the relevant statute, that his term was not yet up? What if the appointing authority committed some error in the process of appointing the magistrate or renewing the magistrate’s term? In short, why stop at two categories — void and voidable — instead of developing an infinite number?
There may, of course, be pragmatic reasons for limiting the number of categories. But that is precisely my point: pragmatism and policy are vital tools for administrative lawyers. Conceptual analysis is tremendous fun in the seminar room, but it has tended to cause administrative lawyers tremendous difficulties in practice. One need only mention the law (if one can call it that) of “jurisdiction”, an elegant conceptual construct that can really only be understood by reference to the pragmatic, policy-based considerations that have influenced judicial applications of “jurisdiction”.
In my view, the leading cases on voidness and voidability reveal that values, rather than conceptual analysis, provide the structure of judicial review of administrative action. I contend (here and with refinements here) that the values that underpin administrative law are the rule of law (a concern for individual dignity and autonomy), good administration (a concern for the efficient and effective attainment of statutory objectives), democracy (a concern to give effect to the duly-enacted wishes of elected representatives) and separation of powers (a concern to maintain distinct roles for distinct institutions).
Consider Boddington v British Transport Police  2 AC 143. Here, Boddington had been convicted of smoking on a train, in breach of a bye-law. At his trial before a stipendiary magistrate, Boddington sought to argue that the bye-law was invalid. Whether Boddington could raise the invalidity of the bye-law was ultimately determined by the House of Lords.
To be sure, there are some important nods to conceptualism in the speeches. Lord Irvine of Lairg LC said, for instance, that it would be “a fundamental departure from the rule of law if an individual were liable to conviction for contravention of some rule which is itself liable to be set aside by a court as unlawful”. An invalid bye-law is a nullity and cannot ground a successful prosecution. Permitting this, Lord Steyn said, would involve “an injustice which cannot be tolerated in our criminal justice system”.
Equally, however, the Lord Chancellor warned that it would be necessary “in every case…to examine the particular statutory context to determine whether a court hearing a criminal or civil case has jurisdiction to rule on a defence based upon arguments of invalidity of subordinate legislation or an administrative act under it. There are situations in which Parliament may legislate to preclude such challenges being made, in the interest, for example, of promoting certainty about the legitimacy of administrative acts on which the public may have to rely”.
This suggests that values can influence the classification of a bye-law as void or voidable (or, to eschew the language of conceptualism, whether an invalid bye-law can nonetheless have continuing effects): here, good administration (a concern to avoid chaos) and democracy (a concern to give effect to provisions passed by Parliament) are seen to play a role. Lord Browne-Wilkinson certainly argued that good administration could be relevant: “I am far from satisfied that an ultra vires act is incapable of having any legal consequence during the period between the doing of that act and the recognition of its invalidity by the court. During that period people will have regulated their lives on the basis that the act is valid. The subsequent recognition of its invalidity cannot rewrite history as to all the other matters done in the meantime in reliance on its validity”. As Lord Slynn of Hadley put it, “[t]he unscrambling may produce more serious difficulties than the invalidity”.
It is true that the Lord Chancellor went on to insist upon “the strength of the presumption against a construction which would prevent an individual being able to vindicate his rights in court proceedings in which he is involved”. But is this not simply to say that, in a criminal prosecution where, in Lord Steyn’s words, “where the liberty of the subject is at stake”, that the rule of law value weighs more heavily? Indeed, might one not think that the Lord Chancellor and Lord Steyn’s earlier nods to conceptualism were also motivated by the violence to individual liberty that would be done by permitting convictions for breaches of invalid bye-laws?
This is not the place for a detailed review of the literature or the cases. Boddington is often held up as an example of the force (and judicial acceptance) of conceptualism. On closer inspection, however, it seems to be a case that is as much about values as about the metaphysics of voidness and voidability.