The Count Is a Procedure

Most of a democratic transfer of power is informal. Concessions, transitions and handovers rest on convention rather than on law.

One step is not informal, and it is the step at which state results become a national result.

The counting provision governs it, and for most of its life nobody read it, because a procedure that has never failed looks like a formality.

The original text was drafted after a disputed election and was, by common agreement among people who had to work with it, close to unintelligible.

It used terms it did not define, described a process in sentences of extraordinary length, and left several questions with no clear answer at all.

The state certification deadline is the part of the rewrite doing the most quiet work.

That was tolerable for a long time because the underlying results were not close enough for the ambiguity to matter.

Fixing a date by which a state must identify its result removes the pressure that ambiguity had created, because a deadline converts a dispute into a matter of fact.

The enacted history shows the rewrite, and the changes are worth reading as a list of the assumptions that turned out to be unsafe.

It also transfers weight onto state processes, which vary far more than the federal step and receive far less attention.

The threshold for raising an objection was raised substantially, so a single member could no longer force the process into a prolonged dispute.

A federal procedure can only be as reliable as the state determinations feeding it, and hardening one end of that chain makes the other end the weak point.

The role of the presiding officer was stated explicitly as ministerial, resolving a question the original text left to inference.

That is not a criticism of the reform, which was plainly right. It is where anyone worried about the next failure should now be looking.

A single conclusive slate from each state was specified, with a defined process for identifying it, closing the gap that competing submissions had opened.

The proceedings contain the debate over those changes, and the striking feature is how much of it concerned what the old text had actually meant.

A legislature spending days establishing the meaning of its own procedure is the clearest possible evidence that the procedure needed rewriting.

The deeper lesson is about where fragility sits in a system, and it is not where people look for it.

Attention concentrates on the contested parts of an election: the campaign, the count, the litigation, all of which are adversarial by design and therefore hardened.

The ceremonial step attracted no attention precisely because it had never been contested, so it was never stress-tested and never improved.

The administration data covers everything up to certification and stops there, which is where the federal procedure begins.

That gap in the record is itself informative. The counting step is administered by a legislature rather than by election officials, so it sits outside election administration entirely.

What remains unresolved is the harder question the rewrite did not attempt.

The procedure now says clearly what happens when a state sends one slate. It says much less about what happens if a state fails to send anything at all.

That scenario requires a state institution to fail rather than an individual to misbehave, which is less likely and would be far harder to correct.

The general principle is worth stating plainly: the parts of a system nobody argues about are the parts nobody has tested, and they are where the next failure will be.

The comparison with other democracies is instructive and rarely made in these terms.

Most systems hand the certifying step to an electoral commission or a court, precisely so that the body declaring the result is not the body the result creates.