Holding conquered ground
A Socratic walk-through of holding conquered ground — reasoned out one step at a time, not lectured.
The question we started with
THE QUESTION #Why is beating an army so much cheaper than governing the territory it was defending?
A campaign that destroys a country's army can be over in weeks, with a force smaller than the one it defeated. Then the same force finds itself unable to hold what it took, and the bill for holding it dwarfs the bill for taking it. If the hard part is beating the professionals, why is the easy part the one that breaks armies?
Reasoning it through
REASONING #Ask first what makes an army beatable. An army has to concentrate to be useful. It masses, it occupies positions, it moves in formations that need roads and fuel, it communicates. Every one of those necessities makes it findable — and once found, it can be struck by a force that need only be superior at the point of contact, not everywhere. Better still, the contest has an end condition that the enemy supplies: when his organisation stops functioning, the campaign is over, whatever remains of his manpower.
Now ask what the occupier is asked to do next, and notice the shape of the task has changed completely. There is no opposing organisation to break. The requirement is presence — continuous, distributed, everywhere at once — because what an occupier must produce is not a defeated enemy but an ordinary person's expectation that tomorrow will be safe. That expectation is destroyed by the incident that happens, not by the ninety-nine that were prevented. Security has to hold at every point simultaneously, which means the requirement scales with population, area and road network, and not at all with the army that was defeated.
Put a number on that scaling, since the arithmetic is the argument. Planning ratios for this kind of work are usually written per head of population, and the figure most often quoted is around twenty security personnel per thousand inhabitants — recalled, and genuinely contested, since it was drawn from a small and non-random set of cases and its defenders and critics both have institutional stakes. Take it only as arithmetic: at that rate, a territory of twenty-five million people implies half a million troops. Compare that with the force needed to defeat its field army, and the two numbers do not live on the same page. Nothing in the first calculation refers to the enemy at all.
There is a second reversal, and it is the one that undoes the occupier's remaining advantage. The defeated side need not disband. It can disperse — and dispersal is not a weaker version of massing, it is a change of category. Firepower, manoeuvre and air superiority are all advantages in finding and destroying concentrations. Against people who do not concentrate, they have nothing to bite on. The occupier's strengths were bought for a contest that is no longer being offered.
Then there is the matter of time. A battle is an event with a cost. Governing is a rate: a cost per month with no natural terminus, and the total is set by duration, which the other side largely chooses. That converts a finite expenditure into an open-ended one, and open-ended expenditures are decided in the occupier's own capital rather than on the ground.
Underneath all of it sits something an army cannot manufacture. Ordinary policing works overwhelmingly through voluntary compliance and through information that residents volunteer — who is new in the street, whose vehicle that is. An occupier can supply force but not consent, and without consent the intelligence that makes policing efficient dries up, which pushes it back toward the manpower-intensive method the arithmetic above already priced.
Is that a mechanism or a mood? It predicts something specific: occupation cost should track the population and terrain governed, not the difficulty of the campaign that preceded it. What would refute it: densely populated territories taken easily proving cheap to hold, and thinly populated ones taken after hard fighting proving expensive. That pattern would mean the cost was residual enemy strength after all, rather than the different shape of the second task.
The analogy
THE ANALOGY #Think of a locksmith who can open any door in a city in an afternoon — and is then made responsible for keeping every door in it locked, every night, indefinitely. The first job scales with his skill and ends when the last door opens. The second scales with the number of doors and never ends, and no amount of additional skill substitutes for the number of people needed to stand near them.
doors have no opinions, whereas a population decides whether to help or hinder, so the number of doors needing watching is itself a function of how the occupier behaves — something the analogy makes look fixed when it is the most movable quantity in the whole problem.
Clarifying the model
THE MODEL #The misreading to avoid is that occupation is hard because resistance is fierce. Often it is not, at least at first; the cost is structural and it arrives before anyone shoots. Distributing a force thinly across a country to produce continuous presence is expensive even if nobody opposes it, because presence is a quantity of person-hours per square kilometre and cannot be concentrated the way combat power can.
Two refinements connect the steps. First, the per-capita planning figure is a symptom, not a law — what it encodes is that the denominator changed from "enemy formations" to "people," and any ratio with that denominator will be large. Second, this is not simply the argument that a weaker side wins by exhausting a stronger side's will. That describes the resister's strategy; this describes why the occupier's cost curve changes shape the day the campaign ends, even against no strategy at all.
A picture of it
THE PICTURE #How to readThe line is nothing but arithmetic on one recalled and contested planning ratio — twenty security personnel per thousand inhabitants — so read the slope, not any single point. Trace across from a population and up to the troops it implies, and note how quickly the figure passes the size of any force a state would commit to a campaign. The instructive part is what does not appear on either axis: the defeated army, its size, and its quality are absent from the calculation entirely.
What became clearer
WHAT CLEARED #Beating an army is cheap because the enemy has to present himself: he concentrates, becomes findable, and supplies an end condition. Holding his ground is expensive because nothing presents itself — the requirement becomes continuous presence measured against the governed rather than the defeated, the opponent can change category by dispersing, and a finite event becomes an open-ended rate. The two tasks are priced by different quantities, which is why success at the first tells you almost nothing about the bill for the second.
Where to go next
ONWARD #- Why local security forces are the standard exit route from this arithmetic, and why standing them up is slower than equipping them.
- How consent lowers the manpower requirement by restoring the flow of information that makes policing efficient.
Key terms
TERMS #| Term | What it means |
|---|---|
| Force density | the number of security personnel per unit of population or area, the usual planning currency for occupation. |
| Culminating point | the moment an advance can no longer sustain itself, the same limit read from the campaign's side. |
| Stabilisation | the tasks of maintaining order and basic services after organised resistance ends. |
| Dispersal | abandoning concentration so as to present no target, which removes the value of an opponent's firepower. |
Every term the collection defines is gathered in the glossary.