The Delegation Is the Decision

Legislation is precise about who holds a power. It names an office, and the naming is not casual: the choice of which official is accountable is often the most negotiated part of a bill.

The provision governing delegation then permits that official to hand the power to a subordinate, and the handing is a document almost nobody outside the building reads.

This is necessary. A named officer cannot personally perform thousands of statutory functions, and a government in which they tried would do nothing at all.

It also means the operative question is never what the statute says about who decides. It is what the delegation order says, and those are different documents with different visibility.

The codified regulations carry many delegations, which makes them findable. Many others sit in internal orders that are public only in the sense of being disclosable on request.

There is a version of this that goes further, where a power is not delegated to a person but effectively to a process.

The level to which a power is delegated is the whole substance. A decision reserved to the head of an agency is made a handful of times a year with deliberation.

A decision assigned to whoever is handling the matter, applying a manual, has been moved out of the hands of any identifiable official altogether.

The same decision delegated four levels down is made hundreds of times a year, quickly, by people applying a practice rather than exercising judgement.

That is efficient and it removes the last thing a challenge needs, which is somebody whose judgement can be examined rather than a procedure that was followed.

Neither is wrong. But they produce different governments, and the choice between them is made administratively without a hearing, a notice or a vote.

The manual then becomes the operative law, and manuals are revised without notice, without comment, and frequently without a published date.

The register publishes the significant ones, and reading a run of them is one of the few ways to watch the internal shape of an agency change.

Redelegation is where it becomes hard to follow. A power delegated once can frequently be delegated again, and the chain is rarely documented in one place.

Establishing who actually held the authority to sign a given decision can require assembling three or four orders issued years apart, which is a genuine obstacle to review.

Courts handle this badly, and understandably so. A challenge to authority requires the challenger to trace the chain, and the chain is held by the party being challenged.

The general provisions that govern how such documents are published set the framework, and their standard for what must be published is narrower than the range of what actually matters.

There is a further asymmetry in what delegation does to reversal. A decision made at the top can be revisited by the person who made it, and they are findable.

A decision made four levels down is defended by the whole chain above it, because reversing it implies the delegation was wrong, and defending it costs nobody anything.

So delegation quietly changes the cost of error correction, which is a more consequential property than the cost of making the decision in the first place.

The people who understand this best are the practitioners who appear before agencies regularly. They know which official's signature a matter needs and how to get it in front of them.

That knowledge is worth money and is not written down anywhere a member of the public could find, which makes it another quiet advantage of the well-advised over everyone else.

None of this argues for less delegation. The alternative is paralysis, and the officials named in statutes could not do a tenth of what is assigned to them.

It argues that a delegation order is a governing document and should be treated as one: published in a single register, indexed by the power it moves, current as at a date.

That would cost very little and would let a citizen answer the question a statute appears to answer and does not, which is who is actually deciding this.