The Department of Veterans Affairs is really hard to staff.

Its Veterans Health Administration (VHA) has 30,000 vacant clinical positions. Eligibility-claims processers are in such short supply, there remains a waiting list 75,000 veterans long. Appeals of eligibility denials have a backlog of more than 300,000 and take an average of 2.5 years to resolve. The VA even lacks undersecretaries to supervise those areas.

To top it all off, President Trump ousted former Secretary David Shulkin over differences about whether to pay for veterans to receive care from private providers, and his pick to succeed Shulkin—Rear Adm. Ronny Jackson, M.D.—recently withdrew his name from consideration over allegations of on-duty drunkenness, harassing female coworkers, and such and such.

This ongoing soap opera, however, keeps anyone from asking the right questions or proposing the right reforms.

Shortages and waiting lists at the VA are hardly surprising. Its health care system, the VHA, is the United States’ version of the U.K.’s single-payer National Health Service. It is an entirely socialist enterprise, where the government owns the means of production (hospitals, clinics, CT scanners, bedpans), employs the workers, decides how much everyone gets paid, and generally chooses how to allocate all those capital and human resources.

In other words, it’s a system without true prices—and that’s why it doesn’t work.

The purpose of the price mechanism is to get the right amount of stuff to the right place at the right time. Instead of using prices to allocate resources within the VA, Congress relies on its own wisdom and that of VA bureaucrats. So it’s no surprise that thousands of veterans die waiting for eligibility-appeals rulings, and others kill themselves while waiting for mental health services.

As is the case in Britain, Canada, and all other countries where government runs health care, in some regions the VA has too much stuff and not enough patients. In other areas, it has too many patients and not enough stuff.

Some complain the problem is not misallocation of resources but garden-variety underfunding. It’s hard for the VA to hire doctors when it pays so much less than the “private” sector, where doctors’ main source of income is, well, other government programs that pay more. Perhaps government should overpay VA clinicians as much as it overpays non-VA clinicians.

It’s impossible, absent a price mechanism, to know whether VA salaries (prices) and overall spending are too low, but there is reason to think they might be. Sub-optimal access to care may be a durable political equilibrium in a government insurance program structured like the VA.

Congress faces strong political incentives to renege on the commitment it makes to take care of active-duty personnel if they suffer a service-related illness or injury. At the moment Congress issues that promise, it costs Congress nothing. By the time the bill comes due, most veterans don’t need the VA, and those who do have nowhere else to go. Veterans who depend on the VA are a small and captive clientele who cannot compete with the political power of other interest groups (defense contractors, the elderly) in the battle for ever-larger shares of federal spending.

Trump dismissed Shulkin, we are led to believe, because Shulkin opposes increasing access to care by having the VA pay private providers to care for veterans in areas where the VA’s waits for care are severe. Shulkin derides the idea as “privatization.” It’s nothing of the sort.

Privatization is when government transfers ownership of a resource from itself to private individuals. It’s what many former Soviet-bloc countries did after the Berlin Wall fell.

By contrast, Trump wants the VA to work like Medicare—i.e., to have government write checks to private doctors and hospitals, instead of government doctors and hospitals. It’s not privatization when you remodel the VA on Medicare, any more than it is privatization when Medicare switches from writing checks to private doctors and hospitals to writing checks to private insurance companies.

True privatization would look very different from what Trump proposes. It could even create a new political equilibrium that provides better care to veterans and prevents so many soldiers and sailors from ending up dead, injured, and sick in the first place.

Privatization would transfer ownership of the VA’s physical capital (land, structures, gizmos) to private citizens—ideally, to the people the VA exists to serve: veterans. Privatizing the VA would constitute a hefty transfer of wealth to veterans that would be a large and welcome step toward making good on Congress’ promise to care for them.

Shulkin also warns privatization would dismantle the VHA, but that too is incorrect. It would transform the VHA from the largest government-run integrated health system in the United States to the largest private integrated health system in the United States. Veterans could continue to receive care from the same specialists as before, but through a system that veterans themselves own, operate, and choose. If veterans-cum-shareholders so choose, that system could also treat non-veterans, injecting much-needed competition into every health care market in the country.

Current VA enrollees would still need Congress to underwrite, Medicare-style, the care they receive through that system, or whatever system they choose. But further reforms would allow active-duty personnel to receive future veterans benefits from entities who face incentives to keep rather than renege on those commitments. Indeed, such reforms could make Congress and the president less likely to engage in unnecessary wars, thereby reducing the risk that active-duty personnel would need veterans benefits in the first place.

Instead of making an implicit, unfunded promise of veterans benefits, Congress should immediately increase military pay enough to allow active-duty personnel to purchase a standard package of life, health, and disability benefits from private insurance companies at actuarially fair rates. Private veterans-benefits insurance would begin paying claims the moment military personnel leave active duty. Since active-duty personnel could choose their veterans-benefits insurers, those who treat veterans the way Congress has would soon find themselves put out of business by those who keep their promises. Like a privatized VHA, some of the insurance companies could be owned and operated by veterans, or by financial institutions on which active-duty personnel already rely.

The most dramatic change, however, would be in the incentives the president and Congress face. As noted above, Congress currently makes an implicit, unfunded promise to care for veterans. Pushing the cost of veterans benefits into the future artificially hides the cost of current military spending—particularly when war increases those costs.

Forcing Congress to pre-fund veterans benefits with an immediate bump in military pay would force Congress to strike a better-informed balance between military spending and other priorities. And since the bump in pay would be tied to the cost of actuarially fairveterans-benefits insurance premiums, those premiums and military pay would automatically rise at the moment Congress or the president commits troops to battle. Congress would have to grapple with the additional veterans-benefits costs war creates at the moment it (or the president) chooses between war or, say, diplomacy. Forcing Congress to give up more butter to use its guns will cause Congress to enter fewer unnecessary wars, and to exit them faster.

Many veterans groups understandably oppose false “privatization” efforts out of fear that they could result in fewer resources for the system on which many veterans are utterly dependent. They might have a harder time opposing actual privatization, which would transfer massive amounts of wealth to veterans, as well as save soldiers’ lives.

This article appeared on Reason on April 29, 2018.