An economic collapse could, conceivably, lead to a civil war, though a civil war is not inevitable, nor even a very likely result of that collapse. Riots would make sense, but not war. Were it to happen, however, it probably would create the sort of sides that Mr. Bracken suggests.
There is some merit to the general idea that a war might have a distinct ethnic/socioeconomic cast, but the details are what matter, and they're not spelled out. What's important there, however, is the concentration of populations. If one side is defined by poor/color, those people will be concentrated in urban areas.
As usual, the actual contours of such a war would be defined by logistics; in particular, food, water, resources, and manufacturing capabilities.
Manufacturing is in urban areas; but the people who would be fighting on the poor/color side would not be employed in those industries, and the infrastructure that feeds them would break down in any case. The ability to wage war would be limited to supplies on hand. Urban areas would quickly be starved out due to the disruption of food and water supplies.
The idea of government control is glossed over -- but the implication that it would support the poor/color side is not remotely convincing.
The real response of government and, more importantly, the military, would be driven by the circumstances of how this supposed war breaks out. If it were to happen at all, it would almost certainly begin in the form of riots on the poor/color side of the equation; and in that case, the government (including the military infrastructure) would come down against the poor/color side.
But really, it's not a convincing scenario in the first place.
I think another big consideration is electrical power. The NIMBYs have force the power plants out of the cities, into the rural areas, and now they can be cut off. I think that’s going to be huge.