The Casebook · Entry 05
Budget versus Actual
The variance report is read as control. The lens asks whether anything closes the loop, or whether the same miss is explained again each period.
The conventional reading
Budget versus actual is the standard instrument of control. It anchors the board packet and the monthly review: variances computed, thresholds flagged, explanations gathered from the managers who own the lines. The ritual is accountability itself, plan against performance, in writing, every period. A function that produces it on time and defends it well is considered managed.
Through the Native Constraint lens
A variance without a consequence is a description. The report measures the gap, and then nothing closes the loop: the explanation is filed, the period rolls, and the same miss returns wearing a new month. Every figure on the page is a lagging indicator, narrated as steering. The practice mistakes producing the report for exercising the control the report was meant to be, and the fluency of the explanations is how the mistake stays hidden.
What it really costs
The explanation economy. Hours spent narrating randomness, because no threshold separates signal from noise, so every wobble earns a paragraph. Tolerances improvised fresh each period, so what gets chased is whoever asked loudest. And the real cost: misses that never convert into changed rules, so the function pays for the same lesson monthly and learns it never. The budget itself decays into a fiction defended quarterly.
Where it puts the burden
On the managers, downstream, every month, re-explaining what the design keeps producing. And on the board, asked to react to numbers describing a period already over, with nothing on the page that leads.
How to see past it
Close the loop or retire the ritual. A variance worth flagging carries a tolerance stated in advance, an owner, and a consequence: a rule changed, a setpoint moved, a decision made and checked next period against the same figure. Below tolerance, silence on purpose, so attention lands only where meaning is.
The report earns its place on the day a miss changes the system that produced it. Until then it is weather reporting, and the forecast never improves.
The variance report is one case. The habit is to read any practice twice: once at its reputation, and once for where the work actually sits, what it costs to keep it there, and whether the system could produce the outcome instead of a person holding the line.