For decades the routine ran the same way. A company that wanted a shareholder's proposal kept off its own ballot wrote to the Securities and Exchange Commission, laid out why the proposal fell outside Rule 14a-8, and waited for a staff attorney to write back saying the agency would not object. The letters ran long. They stacked reason on reason, because nobody could be certain which argument the particular attorney reading it that month would accept, and a fourth reason cost nothing to include.

That stopped in stages. Last November the Division of Corporation Finance said it would no longer weigh the merits of most of these requests. For the season that followed, a company could still file its notice and, if it represented that it had a reasonable basis for exclusion, get back a letter saying the staff would not object, with no view on whether the arguments were any good.1 On August 14 the Division went the rest of the way and stopped responding altogether, including to the one narrow category it had kept and to those no-objection letters.2 Companies still file a notice saying what they are leaving off and on what grounds. Nobody at the agency writes back.

The notices changed across that season, once nobody was grading the arguments. A law firm that advises public companies on these filings, and therefore has an interest in how the season gets read, reviewed the notices in early July and counted roughly 135 companies excluding about 165 proposals, about 95 percent of them carrying that representation. More than 80 percent of the notices rested on a single basis for exclusion. Less than 5 percent cited three or more. The same review observes that the old no-action requests "tended to include as many arguments as possible."3

What the fourth argument was doing

Nothing in the underlying law moved. The bases for exclusion in Rule 14a-8 are the ones they have been for years. What moved was the identity of the reader, and with it the price of a weak argument. When a staff attorney was going to weigh the letter and issue an informal view, an extra claim worked as free insurance against a reviewer whose application of published standards varied case by case. Once the staff stopped weighing, the reader who could still hurt the company was a proponent willing to sue, and six proponents sued six companies over exclusions that season.3 A thin argument sitting beside a strong one hands that reader something to pull on. The counts improved because padding started to cost something.

Count the reasons in the last decision memo that crossed your desk. The number measures how well the writer thinks they can predict the person who has to approve it.

There is a second finding in that review, and it cuts against what most people expected would happen. With the referee gone, companies did not start excluding everything they plausibly could. The firm's read was that absent a straightforward and noncontroversial basis, companies put proposals into their proxy materials even where they had reasonable arguments to keep them out, weighing litigation, investor reaction and reputation, among other risks, against proposals that were headed for low support anyway.3 Taking away the reviewer's judgment seems to have made companies more careful.

Every company runs its own no-action process

You have a version of this somewhere in your building. It might be the capital request listing nine benefits, or the exception memo citing four precedents for a single deviation, or the reorganization deck carrying a market rationale, a cost rationale, a talent rationale and a customer rationale for a decision that turned on one of them, and everybody in the room knows which one. Rigor usually gets the credit for the length. The cause is that the approving body's criteria move as its membership rotates, and the author is covering objections they have no way to predict.

I have spent enough years on the delivery side of large organizations to recognize the shape of a document written for an unpredictable reader. It runs long. Somewhere inside it, usually third, sits the actual reason.

What surprises people is how durable the habit is. The steering committee that demanded nine benefits gets dissolved in a reorganization, and the nine-benefit template stays in the shared drive, because by then it is simply what a serious proposal looks like around here. The mechanism is the one I described in August, when a discriminatory hiring rule survived the deletion of the policy that carried it. Removing the reviewer does not remove the behavior the reviewer taught.

Which gives you a diagnostic you can run this week without asking anyone a question. Argument count is a reading of your own governance taken from the outside. If your people routinely bring five reasons for one decision, they are telling you how legible your approval criteria are, and the hedge is aimed at you.

The thing everyone organized around was never a rule

In explaining itself, the August statement quoted a 1976 release stating that no response or other action by the Commission or its staff is required on these filings.2 The apparatus that both sides built their calendars around for the fifty years since was an informal courtesy that hardened into an expectation, and then into something people treated as law. I would bet there is one of those sitting in your own approval chain right now, with nobody able to name the decision that created it. That is the same problem I wrote about earlier this month from the opposite direction, in a federal program that ran for years because no budget line ever forced anyone to defend it.

So here is the question I would put to a leadership team this quarter. What is the one criterion that will actually decide the next significant request that comes to you, and have you ever said it out loud to the people who have to write that request? If they could predict you, what would they stop writing?