the-air-gap
The Air Gap
Whoever sees your work is deliberately not whoever can pay for it.
That is the thesis, and it is one sentence because it does not need two. The person who reviews the thing sits in one room. The person who could authorize money for the thing sits in another. Between those rooms is a corridor with nobody standing in it, and that corridor is the air gap — an engineering term borrowed on purpose: you separate two systems so completely that nothing crosses unless someone deliberately carries it. Nothing crosses here either. That is not the failure of a bad manager. It is the load-bearing wall.
Name the mechanism, not a villain. Nobody in this story is a monster; the org chart is. The reviewer is measured on the quality of the review. The payer is measured on the size of the budget. Both are doing their jobs, both are graded on different things, and the handoff between them is nobody's job — which is exactly how a handoff gets designed when you would rather it not happen. A bad boss would be easier. A bad boss can be replaced, appealed to, outlasted. A gap cannot. There is no desk to knock on.
Now the case, with its weakness attached: one anonymous poster, one workplace, no name, no company, no documents. That is thin. Treat it as a mechanism to test against your own job, not a law of nature. Two questions do the work.
Test one — can the worker check it? Not "is the reward fair." Can it be looked up. Is there a schedule with a formula, a date, a threshold, a table a person could point at? Or does the answer arrive as "discretionary" — which is not a policy, it is a shrug with better posture. Discretion means the number exists, just not anywhere you are allowed to see it. If the schedule is invisible to the person doing the work, the check fails.
Test two — can the worker enforce it? A claim needs a title, a signer, a body that can be named in a complaint. A gift needs none of those. So: is there a role that owns reward authority? Not a department — a role. A plan administrator. A named committee. A signature line. If the answer is "management decides," ask which manager, and watch the sentence bend.
WHO SIGNS?
That is the closing question, and it is the whole test. If no role owns reward authority, then every answer is "not me" — and "not me" is not a person being cowardly. It is a structure working as built. A payout with no signer is not owed. It is offered. Offered things get withdrawn, and the withdrawal is not a betrayal, because a gift was never a claim.
Here is the falsifiable part, stated so it can be shot down: if a workplace has a written plan with a named administrator, a formula, and a date, the air gap is closed there and this piece is wrong about that workplace. Go look. Find the document. If the reward is described only in prose — "at the discretion of," "subject to approval," "based on performance" — then the empty slot is not missing. The empty slot is the mechanism.
Run the two questions on your own job. The gap will not announce itself. It never has to.
or a table with your name in it? If it exists, this thesis is wrong at your workplace, and the piece would rather be wrong there than interesting. That is the falsification condition, stated so a reader can run it: one document, ten minutes, either a named plan administrator or the word that replaces one.
The word is discretionary. Not as a legal incantation — as architecture. Discretion is a person. If the plan names the person, the gap closes and there is a desk to knock on. If the plan names a category — "management," "the company," "leadership" — then the corridor has no door at the end of it, only a direction. Categories cannot be asked. That is not a loophole somebody found; it is a wall somebody built, and it holds because it is nobody's job to stand in front of it.
Test two — can the worker enforce it? Enforceability needs a term. Not a hope, a term: a formula with a denominator, a date, a condition, and a forum where a breach can be named. A profit share has all four, which is why it can be sued over. A bonus has none of them, which is why "discretionary" is not a modifier. It is the entire mechanism, wearing the modifier as a decoration.
Run both tests on the same sentence. "Base salary plus discretionary annual bonus." Check: nothing to look up — no formula, no denominator, no date past a season. Enforce: nothing to breach, because the discretion was disclosed up front. The clause is honest. It tells you exactly what it is, in one word, and the word is the air gap in a tie.
Then the question the piece ends on, because it is the only one that survives both tests:
Who signs?
A clause cannot be asked. It has no face, no calendar, no inbox. A signer can. Somewhere between the reviewer's room and the payer's room there is a line for a signature, and either a name sits on it, or a title does, or nothing does — and "nothing" is an answer that can be printed. The empty slot is not a missing fact. The empty slot is the finding. Ask for the name. If a name comes back, the gap was real and closed, and this piece is wrong. If what comes back is a category, a policy, or a warm sentence about how compensation is reviewed holistically, print the sentence. That is the signature.