An initiative's budget, staff, and attention are cut within the budget owner's authority, cycle after cycle, until it ends without a cancellation event its defenders could rally around.
Borderline: whether this move is justified depends on the situation, and only the person in it can judge.
An example
The product innovation lab at a toy manufacturer came with a history: the previous chief executive, now on the board, had founded it, and in six years it had produced two licensing deals and no hit. The new chief executive judged the money belonged in the core brands. In his first budget he cut the lab's funding 15 percent under the company-wide efficiency target and left two vacancies unfilled. He moved its standing slot on the executive calendar to quarterly. When the lab's director asked him directly whether the lab had a future, he told her it had a year to show a product a core brand would launch, and he told the board the same at its next meeting. In the second budget the lab's funding fell another 20 percent. By year end its best engineer had moved to the core brands, and the lab closed in the spring with a short note to staff.
Why it works
Ending a program outright has a price: a named decision, backlash from defenders, a confrontation with its champion. Routing the outcome through ordinary budget cycles spreads that price thin, which makes ending it easier. As the program shrinks, its results decline, which makes keeping it look less worthwhile and supports the next cut, and defending it becomes personally costly, since attaching your name to a shrinking program signals poor judgment. It works for a budget owner with authority over the line items and headcount, when no powerful patron contests each cut.
The thinking behind it
Thomas C. Schelling Schelling's Arms and Influence (1966) describes salami tactics: a line crossed in slices, each too small to justify a response. Cutting a program cycle by cycle works the same way on its defenders. No single reduction is large enough to fight, so the program ends without the event that would have drawn opposition. Defenders who lay every cut on one page reassemble the slices into a decision that someone has to own.
John M. Darley Darley and Latané's Bystander Intervention in Emergencies: Diffusion of Responsibility (1968) found that people who believed others were also present were less likely to help, because each felt a smaller share of the obligation. A program declining through routine budget cycles has many people who could defend it and no single moment that demands it. Asking for a yes or no by a set date concentrates the responsibility on one decision-maker again.
When it fits
Budget owners who have concluded that a program should end, and who expect an outright cancellation to cost more in conflict than the program is worth, are the people it suits. You need authority over the program's budget and staffing across several cycles.
How to use it well
Settle the end state, and tell your own manager and the board.
- Make each cut within your authority, and give its real reason in the plan.
- Tell the program's leader honestly what the program would need to show to survive, and by when.
- Move staff toward work with a future before the program closes.
- When asked directly whether the program is ending, say what you have decided.
Close it with a plain note when the time comes. A slow decline does not need a secret to work.
What it costs and where it backfires
It takes several cycles, and the program's staff spend them in uncertainty, which costs morale and often your best people. Defenders who see the pattern can force the question in a forum that demands an answer. The line is the answer you give: denying the plan when asked directly, or letting the program's people believe in a future you have ruled out, conceals a decision from the people it affects most, and they will learn it eventually.
When it is used on you
Watching a program you care about shrink cycle by cycle, lay the cuts side by side on one page, present them as a single decision, and ask who made it. Ask for a yes or no on funding by a set date, and find a patron or an outside constituency that makes the decline visible.