Suppose the national defense of the United States were relegated to the private sector. Instead of the publicly funded Army, Navy, Air Force, and Marines, the country would be defended by private militias funded mainly by insurance companies. In the event of foreign attack on U.S. soil, the militias would defend those citizens in the affected areas who'd paid defense insurance premiums through their places of work (or, if self-employed, as individuals).
The best-armed troops would defend the wealthiest and most hawkish segments of the population, who would have paid the highest premiums.
The less-wealthy and more dovish customers who'd chosen a less-generous policy would likewise be defended against attack, but they could expect to pay heavily out of pocket because their insurance would only cover costs for weapons and manpower above a fairly high deductible. The doves' militias might or might not call in air support, knowing the insurance company would pay for it only in the most dire circumstances—difficult to calibrate as bombs are dropping all around you. Or perhaps these troops would belong to defense maintenance organizations (DMOs) that blended defense and insurance functions. If so, the soldiers would be required to follow strict protocols that would likely forbid not only air support but also the use of tanks.
Poor people and outright pacifists would buy no defense insurance at all, and therefore would end up being saddled with ruinously large debts to private militias they'd chosen hastily after the invasion was under way. Alternatively, these individuals might simply say the hell with it, wave a white flag, and surrender.
The advantage of a market-based national defense is obvious: Every citizen would receive an individualized amount of military protection, based on the value each of us placed on defending the homeland. Those who were particularly fretful about another 9/11 would pay a lot; those inclined to dwell on the rarity of foreign attack (two from the air and one by land in all of U.S. history) might prefer to use that money to buy a pottery wheel and a kiln for the basement. The freedom to choose is what this country is all about.
But the drawbacks of a privatized defense would weigh quite a bit heavier. The imposition of costs would bear no relationship to the buyers' ability to pay. Instead, cost would be driven by geography. People who lived in places that made particularly tempting targets, like Washington, D.C., would pay high premiums, and people who lived in places that had previously been attacked—Baltimore (War of 1812), Pearl Harbor (World War II), New York City (9/11)—would pay the highest premiums of all. Insurance companies would be reluctant to write policies for anyone living in these places and instead would seek out customers whose environs made unlikely military targets. Manhattan might be poorly prepared to repel enemy attack, but Middlebury, Vt., would be impregnable.
The overall cost of military defense would skyrocket even more than it does in our real-world, thoroughly socialistic system of national defense. This cost explosion would be partly due to welcome improvements as defense technology became able to kill many more people than previously thought possible. But costs would also be driven up by the enormous time and effort that defense insurers and militias would spend quarreling over who would pay the bills. (Assuming the militias were paid on a fee-for-service basis, rather than put on salary, soldiers would have an incentive to drive up costs further by putting deployment on a hair trigger.) Market purists would blame all cost problems on the inherently inflationary nature of third-party insurance payments and would argue for increasing the portion of national defense paid out of pocket in order to discipline buyers and pressure the private militias to get their expenses in line.
Bottom line: A market approach to national defense would give us a lousy national defense. That might be hard for many to perceive, since military technology, fueled by generous federal research grants to the more prestigious militias (and sometimes to less-prestigious militias via congressional earmark) might dazzle the world, and foreign potentates might flock to our shores to make use of our hardware and know-how. But too many Americans would be left undefended or underdefended, and too many militias and insurance companies would devote so much money and manpower to cost-shifting battles that they would lack sufficient funds to extend coverage or maintain existing coverage at an affordable level. The percentage of the defense-insured population would decline steadily, and people who purchased defense insurance would find that it covered a dwindling range of contingencies. The United States would end up more vulnerable to attack than Canada, Western Europe, or Japan.
But even that wouldn't be the most exasperating thing about having a market-based military. The most exasperating thing would be that, among the Democratic candidates for president, only Dennis Kucinich would be willing to state outright that protecting lives is the business not of markets, but of governments.