The Spare

The spare in a royal family has four properties, and a hospital ward reproduces every one of them.

The first is that the value is entirely in availability rather than activity. Nobody judges a spare by output, because output is not the point. The second is that the contingency is unlikely on any given day and certain across enough days. The third is that the cost is continuous, visible and easy to total, while the benefit is occasional, invisible and impossible to attribute, because it consists of things that did not happen. And the fourth, which is the one Harry’s title captures, is that the arrangement is understood by everybody as a ranking. The spare is defined by not being the one that matters.

Put those four together and you have a structural prediction: the spare will be resented as a cost, will be cut in any serious cost round, will be missed only in retrospect, and the person occupying the position will know exactly how they are regarded. All four happen on wards, routinely, and almost nobody involved recognises the pattern because nobody has given it a name.

So this page gives it one. The Spare is the capacity a rota holds against variability. What follows is how much of it is needed, which is arithmetic, and what happens when it is removed, which is not.

Before any argument about the spare, there is an argument about the base, and it is usually lost by accident.

A whole-time equivalent on a 37.5 hour contract is paid for a little under two thousand hours a year, and that is not what the ward receives. Annual leave and bank holidays are contractual. Sickness is real and has a long run rate. Mandatory training is a condition of registration. Maternity, carers leave and the rest are what a workforce of actual people produces.

Take those off and the number arriving on the roster is materially smaller than the number in the budget. This uplift is not the spare. It is the cost of employing human beings rather than machines, and leaving it out of a plan does not save it. It guarantees the shortfall arrives later as overtime, agency or an empty slot.

There are 8,760 hours in a year. A post that must be filled at every one of them cannot be covered by two people, or three, or four. Divide the hours in the year by the hours one person is genuinely available and the answer is close to six.

It is worth sitting with, because it explains a category of argument that otherwise looks like obstruction. When a director of nursing says that adding one nurse to every shift costs more than five posts, that is not a negotiating position. It is division.

It also explains why small movements in the deductions matter so much. A percentage point on sickness, or five more days of mandatory training, moves the establishment across a whole hospital by a number of posts nobody budgeted for and everybody will be asked to absorb.

Now assume the establishment is right and every post is filled. The ward is still not safely staffed, because the work does not arrive in equal quantities.

Demand varies, and roughly half of any distribution sits above its own middle. Staff exactly to the average and the ward is short whenever the shift is busier than typical, which is most of the shifts anybody actually remembers. Covering the great majority of shifts requires a margin above the average, and the size of that margin is calculable rather than negotiable.

That margin is the Spare. Note what the chart does not show, because it is the whole problem: there is no bar for the value of being adequately staffed on a bad night. The cost of the spare is a number. Its benefit is a set of events that did not occur.

Here is the confusion this whole page exists to correct, and it is committed most often by people quoting the method this firm practises.

Waste is work that consumes resource and produces nothing anybody wants. The classic seven are overproduction, waiting, transport, over-processing, inventory, motion and defects. Read that list carefully: not one of them is capacity held deliberately against variability. Waste is activity that should not be happening. The spare is availability that is doing its job by being available.

The test to tell them apart takes one question. Remove it, and see who absorbs the difference. Remove genuine waste and nobody is worse off, which is what makes it waste. Remove the spare and somebody absorbs it every time, and if you cannot see who, that is not evidence it was waste. It is evidence that the person absorbing it is doing so silently, which is the condition this study has been describing from the first chapter.

Ohno’s point about buffers is routinely inverted by people who have read a summary. The image is a lake with rocks under it: lower the water and you discover the rocks. The instruction is to remove the rocks and then lower the water. It is not to lower the water and hope the hull holds. In a hospital the rocks are variability in arrivals, in length of stay, in staff availability and in how long anything actually takes. Until those are genuinely reduced, deleting the buffer does not expose a problem to be solved. It transfers the problem to patients and to staff, where nobody is measuring.

So the honest version of the method, which is the one this firm will defend in front of a board, has two halves and they must be done in that order. Attack the waste, relentlessly, because it is free money and it makes the spare smaller. And until the variability is actually gone, size the spare deliberately, name it, fund it, and defend it. A programme that does the second half first is not Lean. It is a cost cut wearing its vocabulary.

Everything above treats the spare as a quantity. On a real ward it is a person, with a name, standing in a corridor being looked at by somebody wondering what they are for.

They have titles. The floater, the runner, the supervisory ward manager who is counted in the numbers and therefore is not really supervisory, the bank nurse who does not know where anything is kept, the second on nights. What they have in common is that on a good shift they have no allocated patients, and on a bad one they are the reason it did not become a serious incident.

And they know how they are regarded. Being structurally designated as surplus is not a neutral experience, which is precisely the thing Harry’s title conveys and which no workforce plan has ever contained a line for. Staff who are repeatedly told, by rosters and by budgets rather than by anybody’s words, that they are the part that is not strictly needed, behave the way anybody would. They disengage, or they work twice as hard to prove otherwise, and both of those cost the organisation something it will book under a different heading.

The role pages make the same point from the other side. A ward manager answers for the shift and controls almost nothing that shapes it, and what they do all day is absorb the difference between the plan and the day. When the funded spare is removed, the ward manager silently becomes it. That is not a promotion and it is not recorded anywhere, and it is the single clearest example on this site of the goodwill buffer replacing a budget line.

Which is why staffing levels belong with the people pages and not in a finance annex. The argument is about arithmetic, and the thing being argued over is a person who will read the outcome as a statement about their worth.

Understaffing is not a stable state, because the loop is short and it feeds itself.

Remove the spare and the remaining staff absorb the variability. Absorption becomes exhaustion, exhaustion becomes sickness, and sickness is one of the deductions in the first chart, so available hours fall and the establishment required to deliver the same roster rises. Some people leave. Vacancies are covered at agency rates by staff who do not know the ward, which raises the load on the ones who stayed and raises the error rate for a while. Each turn makes the next turn more likely and more expensive.

None of it appears as a single decision. It appears in the accounts as a workforce problem, one or two years after a staffing decision that was recorded at the time as a saving, which is the misattributed cost from the entity model running at full speed.

Meridian is not defending an existing establishment. It is choosing one, which is a rare and much better position, and it changes what this arithmetic is for.

The uplift and the coverage factor belong in the first draft of a business case rather than being discovered during recruitment, because they are the difference between a plausible staffing line and a fictional one. The spare has to be a named and defended figure with a stated service level, because anything left as a residual will be removed in the first cost round by somebody who genuinely cannot see what it was doing. And in a market short of clinicians the establishment is not only a cost, it is a constraint on whether the hospital can open at all, which is where this page and the siting question become one conversation.

It is also why running the whole thing first is worth more here than in a mature market. An established hospital can correct an establishment over several years. A new hospital that opens understaffed in a short labour market starts the loop above in its first week, and will be fighting it during exactly the period the group needs to prove the asset works.