[curtainsforus]

It's a matter of economics. The US government wants (has) to rule, not just nuke the country and call it a day. If the whole country's armed, it becomes more costly to have boots on the ground everywhere enforcing the occupation. It at least shifts the equation more in the direction of concessions from the occupying force.

[meheleventyone]

It's the same economics for both sides though. A large-scale, protracted guerilla civil war would be economically devastating for both sides.

[curtainsforus]

There are myriad historical scenarios where large-scale protracted guerilla civil wars would have been economically preferable for the would-be guerillas compared to the alternative (or rather, default).

There's plenty of scenarios where the utility benefits of guerilla warfare (to the guerillas) outweigh the economic costs. To go right to Godwin's law as an extreme example, if every jew in 1930s Germany had engaged in guerilla warfare against the government instead of meekly getting on the trains, the utility benefits to them as a group would have wildly exceeded the economic costs of the hypothetical civil war.

Of course, the best gun is one you never need to fire. A rational alternate-universe government would foresee this, and know that by pursuing policy goal X they would create conditions where guerilla war was utility-positive for group Y, which would hurt the economy and so also the utility for the government. So the government would be (depending on the hypothetical ability of group Y do impede it) incentivised to pick policies that don't leave group Y no recourse but to start shooting.

Mutual armament as common knowledge leads to incentives for both sides to cooperate.