Is the Veterans’ Administration a good health care model?

Last week Paul Krugman defended the VHA as a model for national health care policy; Brad DeLong has some critical excerpts.  I am skeptical for a few reasons:

1. It is widely acknowledged that this system did not work well for a long time.  If we are going to cite examples, should we judge them by lifetime performance, or by performance-right-now?  In this case I view the relative efficiency of the now-moment as the exception, and not as a readily available constellation that national policy will replicate.

2. VHA saves a great deal by bargaining down prices of prescription drugs.  If done on a national level, this will cause the supply of such drugs to contract, perhaps significantly. NB: Supply elasticity can be high even with (especially with?) evil, scheming, profit-soaked monopolists.  And don’t forget "current cash-flow" models of investment, which are eagerly invoked by the left in other contexts, such as tax policy.

3. For a variety of reasons (see the excellent comments on Brad’s post), VHA pays doctors much less than usual.  I am more than willing to consider the hypothesis that doctors at the national level earn too much.  But I cannot imagine a healthy process by which a federal single-payer or nationalization plan will bargain down this sum significantly without all hell breaking loose.  Do not forget what neo-Keynesians tell us about the morale effects of nominal wage cuts, much less large real and nominal cuts bundled together.

4. In general, local or restricted health care plans can bargain down prices with less loss of quality and innovation than if that same bargaining were done at the national level.  That follows from the economic theory of high fixed costs and segregated markets.

I do think the VHA warrants further study.  But I would like to see these questions answered before regarding it as a positive model for reform.  Comments are open…


The VHA is braodly similar to hte NHS. A single payer, single provider system. I really don’t see this as a good model as any perusal of the UK papers or blogs on how the NHS actually operates wil lshow. Currently three months before an appointment for radio-therapy after a cancer diagnosis, for example.

Can anyone advocate for the VHA system who has visited VHA hospitals? Yikes! I'm thinking especially of times when I've hit two relatives in one day - one at Vanderbilt and one at the VHA hospital in the same complex of hospitals. NOT a nice contrast.

The VA getting lower price drugs is a good example of price decrimination.

Are you assuming that the US industry with the highest profit margins of any american industry would be destroyed if the government bargins?

Why. The US governement is a major buyer, but it is not a monopoly buyer.

Second, can you cite a single industry that has been destroyed by governemnt buying at such low prices.

I keep here this assertion about govt bargining, but I have never seen one bit of analysis that supports it.
In general, the major critism against govt buying is that the industry is able to capture the process and get the govt to pay too much. So why will the drug industry go in the opposite direction of essentially every other example we have of govt being a major buyer.
Isn't the aerospace industry the best example of govt monopoly buying and that is one of the highest profit margin industries around where the govt waste massive ammounts.

I am not saying you are wrong, I am just asking for a little analysis to support the assertion.

It's my understanding that one reason the VA is so cost efficient is it outsources much of its back-end office work. If that cost is added in, it does no better than other healthcare system.

As someone who spends part of his time at a VA and part at an academic medical center, I can tell you that the VA generally provides very good care. It lacks some infrastructure niceties - the hospitals tend to be older and don't always look as nice, they don't always have the staff to see patients immediately, and care is regionalized - meaning that you have to travel some distance to get certain types of specialty care. Despite all this, satisfaction with VA care is generally very high. This is probably somewhat biased because without the VA, most of the people who are seen there would have no regular source of care (or would pay high out of pocket costs, particularly for drugs).

More interesting than satisfaction, however, is that there is a growing body of research that shows that the VA provides better quality of care than the private sector, particularly for chronic diseases like diabetes and hypertension. This is true despite the fact that the VA has a more severely ill, less educated, and poorer population than most other health systems. In fact, of all measured health systems, only the VA clearly provides better care; other systems all seem to do about the same.

Part of this is due to the centralization of information - the VA electronic medical record system is the best around, you can get information instantly from any prior visit to any VA in the US. The best thing is that you can query the system to get information on how well you're doing on any particular quality measure - you can relatively easily pull blood pressure levels for all those diagnosed with hypertension, for example. That's something that very few systems can do. In addition, the VA is organized in a team structure that provides excellent and systematized support for care of chronic disease. The VA also funds a lot of research on how to best implement quality improvement systems.

For the economists, an interesting research finding: the evidence suggests that the primary reason that the VA does better is because hospital leaders are provided monetary incentives to improve quality. Areas that are incented improve, those that aren't don't. The care structure and information systems allow this to work (it's hard to improve care if you can't measure it quickly and accurately), but the incentives provide the motivation.

The biggest thing that the VA has going for it is the ability to see it's own gains from preventative medicine. No system in America does that right now, and learning from that is important.

Tyler's point on doctors is an important one, and I think it calls for a movement to shorten the length of medical school. I saw a great graph once which plotted doctors and lawyers in various countries. The correlation was positive and basically a perfect correlation except in the United States, where we have a much higher percentage of lawyers. I imagine it is because of the length and difficulty of law school. We could probably sacrifice small amounts of quality for large amounts in cost gains if it were as easy to get through med school as it is to get through law school.

Tyler's 2nd point argues for prizes for drug advances. There is clearly a conflict between the profit incentive and the legitimate goal of government to provide whatever services it does cheaply. Where possible, government established prizes would be created for drugs, allowing government to purchase new drugs at marginal cost (the size and the goal of the prize could be determined using information markets to get around some of the Hayekian problems with price information).

Marketing costs are a large part of the drug company expenses. A single buyer might save on those. It might even provide an alternative channel for the smaller companies that currently have to go through the big companies to gain access to the market. There is surely some middle gorund between no negotiation and abusive monopsony.

It is unfair to judge the NHS purely as a system because it is also a budget and coverage decision. The budget is small and coverage huge. Newspaper stories are a sample biased towards scandal. It is also worth considering what the system looks like for those who have private insurance too.

I haven't been in a VA hospital for a number of years, but in the 60s and 70s they were hell holes.

Nice post. I wonder if Paul Krugman thought of any of those considerations...hey, didn't he used to be an economist?

On point 1, it's not fair to compare the VA of 20 years ago to today's VA. In 1996, VA dramatically shifted the way it provides health care. They've moved from a large regional hospital-dominated system, to one with smaller health clinics to take care of most outpatient procedures. In the process, they've expanded eligibility, and are now providing health care for many more veterans than before.

The system, as constructed isn't perfect. Funding has been an issue, and VA has attempted to scale back by rationing health care -- cutting certain veterans out of the system, forcing others to wait months, etc. But patients who do manage to get appointments typically receive excellent care.

For what it's worth, on a per patient basis, VA provides care at a cheaper rate than does Medicare.

The one issue that flies under the surface of these issues is eligibility. Any veteran, even one who wasn't wounded or who didn't serve in wartime, is eligible for all health care at VA -- even if they fell off a ladder while painting their house, or caught a cold. The higher-income non-service connected veterans aren't allowed to enroll in the system anymore, but they have been grandfathered in.

A data point, the price of a cup of ciffee is regulated in Italy consumption there in both quality and quantity terms seem pretty strong.

I'm a health economist who's looked at the VA and other military health systems as part of my job(s).

There's one "advantage" the VA has that other government health care systems don't have -- they get to pick there patients, and can limit the number of patients based on their budget. Contrary to popular believe, not every military veteran is in the VA system -- the VA sets eligibility requirements in order to make sure that the number of patients they have is limited to what thay can fit within their budget. In fact, only a minority of former military personnel are in the VA system.

Krugman says, the VA is "highly successful in containing costs, yet provides excellent care." True, but the do it by containing their patient load. Krugman also says something that is flat-out false: "Because it covers all veterans, the system doesn't need to employ legions of administrative staff to check patients' coverage ..."

On the contrary, it does not cover all veterans, and it does have "legions [pun intended?] of administrative staff to check patients' coverage." They have an entire web site ( )devoted just to eligibility which states, in part "All Veterans are Potentially Eligible" (emphasis mine). There is an eight-level system of "priority" detailed at . It has categories like, "Veterans with service-connected disabilities rated 30% or 40% disabling" (priority 2) and "Veterans who agree to pay specified copay with income and/or net worth above VA Income Threshold and income below the Geographic Means Test Threshold" (priority 7 -- which has FOUR "subpriorities," only two of which are currently in use.

Does Paul Krugman really believe they can determine eligibility under such complete rules with fewer administrative staff than it takes another health system just to look at someone's ID card and take down their policy number?

I can confirm what "PolicyDoc" above said about their electronic medical record system. It's state-of-the art, and they actually do use it to save money. For example, if they get a deal on some drug (say, Nexium), they can switch nearly everyone on therapeutic equivalents (say, Protonix) within 90 days. All they do is send a message to every doctor who prescribed Protonix, given them a list of patients to whom they've prescribed it, and ask for approval to switch them to Nexium. It's not mandatory, since everyone knows that two "therapeutic equivalents" are not really equivalent for ALL patients -- but they say that about 95% of the patients switch.

Joel W. is also right about preventative care -- in the private sector, the "churn" of people in and out of health insurance companies makes it so that if an insurance company spends on preventative care, by the time the patient is around long enough to avoid a disease as a result, they are quite likely to be insured with another company. So you have a muted verion of the classic "tragedy of the commons" -- you pay for preventative care, and some other company benefits from lower costs.

Think about it: If it weren't for the "churn," health insurance companies wouldn't just cover preventative care -- they'd require it. And they'd probably even require (say) blood tests to make sure you're taking your preventive drugs (like blood pressure medicine). (Assuming preventative care is actually cost-effective, of course.)

Robin Hanson's proposal for "time-consistent health insurance" might be able to solve that problem if people couldn't shop around too much. You'd have to get around the fact that insurance companies would want to give discounts to people whose previous insurance companies mandated preventative care.

Jokes at the VA

Why are bullets better than a VA nurse†¦

You can fire a bullet,
A bullet only kills once,
A bullet can draw blood,

Let me assure you that the VA is a great example of the potential for nationalized health care; however this is not an endorsement, but a warning.

While the physicians at our VA are the same staff physicians at the University Hospital, the healthcare at the VA is far worse despite their far larger budget per patient. This is due to the gross negligence of the rest of the staff, and the bureaucracy which arrests all effort by physicians to care for patients. No one I know who has been involved with a VA either as a patient or provider would ever want a nationalized health care system if the VA were to be the model.

My thoughts on healthcare:

1) Medical professionals are paid more than they should be, and the best way to reduce the costs of employing them is to improve medical school and residency. The educational value of "cramming" for a medical exam is negligible. Professors are in many cases incompetent teachers payed very well - why? I propose letting students vote with their feet, with easy information sharing about classes and firing profs that can't attract students, with entrance exams and evaluation during residency to determine whether students actually learned something, rather than relying solely on a certificate/diploma, which is stupid. Then students would be forced to take classes that taught them what they needed, and would avoid unnecessary work and bad teachers. Textbooks should further be open-source collaborations among teaching professionals and researchers in many cases - the current textbook system is a racket. Much of this applies to the university system as a whole. Residency is very grueling, pushing a lot of people away from and plus I can tell you firsthand that it sucks to be a patient who needs more care to fix something a sleep-deprived resident screwed up. This needs to be ratcheted down. Another thing is that there is too little competition among medical schools - perhaps the government should kick-start things by starting a couple new ones. I know they accept many incompetent people and reject many competent ones, and lack of competition leads to inferior quality...

2) Medical malpractice is expensive. I propose setting up a couple special arbitration organizations that would be competent, unlike judges and juries in most cases, to determine what really constitutes medical negligence under some set of circumstances. I also propose government offering medical malpractice insurance more or less at cost, as current malpractice insurance providers are providing a bad deal (collusion?).

3) Preventative care is systemically underrated. This is partly a lack of education about what helps and what that's worth. This is partly basic human psychology. In terms of insurance covering preventative care, in a rational market the consumer would go for a health care plan that provided preventative care, and then, if the consumer switched providers (saying people should "just stick with one" is stupid, because encouraging this leads to people getting stuck with a health insurance provider that progressively screws them more and more) the new provider would see that preventative care and give lower prices. This last part not happening could be partly fixed with a well-implemented central health history database, which would provide abstracted data to insurance providers for pricing purposes. The first part - well, I can say educate the population about medicine better, but in practice, such an effort would fail. I have some thoughts on such education, but this comment is long enough already.

cb - Nobody was producing flu vaccines partly because the pharmaceutical industry, despite being able to do so profitably, decided to make a point about not pushing it around, and partly because doing so would have entailed some practice changes that companies are too stuck-up to do. This can not be solved by doing as it asks - only by taking away its power and motivation to do that. Part of this is corporate governance, which I have some thoughts on too, but this comment is long enough already. Anyway..... Q: What's the best way to deal with a politico-economic conundrum? A: Make it irrelevant.

It is widely acknowledged that this system did not work well for a long time. If we are going to cite examples, should we judge them by lifetime performance, or by performance-right-now? In this case I view the relative efficiency of the now-moment as the exception, and not as a readily available constellation that national policy will replicate.

Why is it the exception? if nothing had changed at the VA current good performance could be an aberration. But it appears that the VA system has undergone major changes. Isn't it more reasonable to say that this is an example of a system that solved many of its problems and now works much better as a result, and to conclude that this is not an accident but a result of intelligent management? If you disagree, how about giving reasons?

People should be allowed to go naked in certain recreational areas only - Or specify those areas and you have another persuasive speech topic mature world

Withheld is the needed for treatment, revealing medical evidence! In 2008 the U.S. Congress still has not corrected the 1987 U.S. Supreme Court DOD human “to harm† experimentation issue! [5] This is the from 1944, 1994 U.S. Senate Report’s now 64 years of: A. "During the last 50 years, hundreds of thousands of military personnel have been involved in human experimentation and other intentional exposures conducted by the Department of Defense (DOD), often without a servicemember's knowledge or consent." INTRODUCTION, 1st. Paragraph, 1st. Sentence., B. "most Americans would agree that the use of soldiers as unwitting guinea pigs in experiments that were designed to harm them, at least temporarily, is not ethical." INTRODUCTION, 3d. paragraph, in part the last sentence. And C. "The findings and conclusions contained in this report are those of the majority staff and do not necessarily reflect the views of the members of the Committee on Veterans' Affairs." Chairman. FORWARD, 2nd. paragraph, last sentence. [8]
Each Executive Branch (DOD & CIA) Project completes the R&D process. The prior lessons learned are reviewed. The then Scope of Work defines what the experiment is "designed" to do. The how, where, when and who is identified. The conducted RESEARCHED cause and effects are closely followed. From the results are DEVELOPED safe production, use, the needed for treatment and protection, e.g., the DOD manufacturers handling of hazardous materials such as Depleted Uranium, Agent Orange, the biological agents of Project 112 [9] and the jet-engine noise levels of Project 7210 [2]. All is in the Executive Branch record! Under the cover of national security the revealing treatment evidence: 1. Is not cause identified in a subject’s Medical History, so that they never the wiser become. The deceived victim’s "to harm" effects are not Medical History recorded, therefore not addressed! Prevented is any follow up by independent civilian and Department of Veterans Affairs (VA) physicians. And 2. For veterans’ the resulting experiment specific injuries are not in the VA "schedule of ratings for disabilities"!
From 1953 the U.S. Senate "to harm" lessons learned were in direct disobedience of the DOD Secretary's TOP SECRET, right to say no order. [4] Then known by the Secretary's of all Services, Joint Chiefs of Staff, and their R&D Board. During the U.S. Senate’s reported 50 years, most of the veteran "to harm" service records were destroyed in a 1973 National Personnel Records Center (NPRC) fire. Those that survived had all witnesses censored by Congress’s 1974 Privacy Act! The "Veterans Right to Know Act" was proposed in the 2005 and H.R. 4259 [109th] 2006 Congress. It never became law.
The U.S. Supreme Court 1987 STANLEY decision [5] extends the coverage of their 1950 FERES Case that a death by a military barracks fire was an "incident to service" [1]. STANLEY treats the 1958 DOD "experiments that were designed to harm" disobedience of the 1953 order [4] as also an "incident to service"! The next year was the U.S. Congress’s few 1988 Veterans’ Judicial Review Act. Established was the Veteran’s Legislative severely restricted, Article I Court. "The court may not review the schedule of ratings for disabilities or the policies underlying the schedule.", i.e., the "to harm" Research and Development (R&D) experimental effects and their causes! The Veterans Court Chief Judge's no teeth statement with his noted VA ignoring of the Court’s decisions! [7] The Secretary of the VA was given FINAL DECISION authority on these issues. [10] Included is the power of NO APPEAL to this LEGISLATIVE Veterans Court or to the independent U.S. Judicial Branch Courts. If allowed an APPEAL, it is not part of the record at the Article I Veterans Court. It’s also missing at the next appeal at the U.S. Judicial Branch Article III, Court of Appeals for the Federal Circuit.
The DOD needed for treatment policy revealing evidence is withheld during the VA "disabilities" procedure, e.g., as were the VA’s received original records and this veteran’s 1952-1958 retained duplicates. They survived the 1973 NPRC fire and Congress’s witness censoring 1974 Privacy Act. Evidence that resulted in the to-date VA overlooked "MPerR PERMANENT" "SURGEON HQ ARRC JUN 25' 58 MEDICALLY DISQUALIFIED FOR MILITARY SERVICE"!
A withheld example is through the events after 1991. Acting on the advice of HMO physicians the veteran returned to the VA. In 1994 a VA Criminal Investigator noted as misplaced the original VA records. In 1995 a VA Hearing Officer “certified† the veterans retained copies as part of the disability file. In 1999 the VA ENT Chief stated "the Veterans signs and symptoms of Meniere's Disease clearly are documented in his" [USAF 1952-1956 “certified†] "service record" with his "A STRESS REACTION MAY PRECIPITATE AN EPISODE (OF MENIERES DISEASE) AND CYCLES MAY REPEAT ENDLESSLY" and "THERE IS NO RELIABLE TREATMENT". It is these since service “stress reaction† attacks that resulted in the loss of a third generation medical practice, business failures and unemployment. In 2005 the VA rediscovered their original records! In 2006 VA locally awarded was a 100% disability. Cited was the Social Security Administrative Law Judge 1996 early "disability benefits solely as a result of his service connected condition." This is for the 50 years later continuing consequences of an unprotected 1952-1956 PROJECT 7210, J47 jet-engine from a 60 decibel (dB.) normal hearing base, 158 dB. to over 176 dB. Noise level.[2] An American Medical Association (A.M.A.) 87,381 to over 699,051 sound pressure multiple (X).[3] The prior to 1949 developed ear muffs, flight line noise protection was not available. Not known is the, A.M.A. for each 6 dB. increase sound pressure doubles, beyond 699,051X unprotected afterburner over pressurization!
There now is no 64 years later "Veterans Right to Know". After they complete Honorable Service despite the efforts of some [8] Congress has not given back to veterans these rights. Revealed would be the few’s corrupt for the greater good of all. As accomplished by the end justifies the "designed to harm" means. All carried out under the cover of our nation’s wars! A few key members in Congress, have dishonored those that serve. Prevented is injury treatment and correction! Lost are those prior to service rights that convicted rapists and murderers keep! [6]
[1] U.S. Supreme Court, Feres v. United States, 340 U.S. 135, 146 (1950).
[2] USAF PROJECT 7210 "A COMPILATION OF TURBOJET NOISE DATA", BOLT BERANEK & NEWMAN, INC. CAMBRIDGE 38, MA. Sound pressure levels for all jet-engines in-service. Conducted at Wright Patterson Air Force Base (WPAFB) DAYTON, OHIO in 1952. 1954 logged in as the 401st report for that year published as REPORT 54-401 July 1956.
[3] American Medical Association (A.M.A.) Family Medical Guide Third Edition pages 364-366 with the for each 6 dB. increase sound pressure doubles and the 60 dB. Normal hearing noise level.
[4] DOD Secretary's 26 February 1953 NO non-consensual, human experiment’s Memo pages 343-345. George J. Annas and Michael A. Grodin, "The Nazi Doctors and the Nuremberg Code; Human Rights in Human Experimentation" (New York: Oxford University Press, 1992). In REFERENCE [8] as NOTES 72, 168 & 169.
[5] U.S. SUPREME COURT, JUNE 25, 1987, U.S. V. STANLEY, 107 S. CT. 3054 (VOLUME 483 U.S., SECTION 669, PAGES 699 TO 710). In REFERENCE [8] cited in NOTE 169.
[6] U.S. State Dept., "U.S. Report under the International Covenant on Civil and Political Rights July 1994, Art. 7".
[7] CHIEF JUDGE FRANK Q. NEBEKER, STATE OF THE COURT FOR PRESENTATION TO THE UNITED STATES COURT OF VETERANS APPEALS THIRD JUDICIAL CONFERENCE OCTOBER 17-18, 1994. In the Veterans Appeals Reporter. www.firebase. net/state_of_court_brief.htm Annual Judicial Conference Transcript.
[8] U.S. Senate December 8, 1994 REPORT 103-97 "Is Military Research Hazardous to Veterans' Health? Lessons Spanning Half a Century." Hearings Before the U.S. Senate Committee on Veterans' Affairs, 103rd Congress 2nd Session. With NOTES 1 to 170. COMMITTEE PRINT - S. Print. 103-97.
[9] "Project 112 (Including Project SHAD) Home" chemical and biological experiments; www.
[10] United States Code (USC) Title 38, 511. Decisions of the Secretary; finality.

"In general, the major critism against govt buying is that the industry is able to capture the process and get the govt to pay too much.......I am just asking for a little analysis to support the assertion." (Posted by: spencer at Feb 3, 2006)
I'm glad Spencer mentioned the VA (Veterans Administration). He brings up valid arguments. However, consider the VA bulk drug pricing package negotiated with the drug companies. Even though a generic, compare the $129.00 Zocor prescription seniors can pay at local drug stores for 30 pill, a 30-day supply (K Mart Zocor 30-day supply $5), to that of a veterans VA prescription. That same exact supply of Zocor, this costs the veteran the maximum copay of $8.00 for a 30-day supply.
Now, should the veteran's prescription call for 30-day supplies be increased to 60, 90, or 200 pills, the cost remains at $8.00. Here's the reality check on the cost of drugs? When the VA set the 30-day copay cost to the veteran, the cost for drug medication, was not even considered! (Federal Register 12/6/01, then $7). Now ask yourself, are seniors getting ripped off when having to pay for 30-day supplies, $129 or more for a variety of drugs, and when a store for many, like K Mart, is not easily available to them? When these bulk purchase prices were negotiated, and set with the VA, can you imagine the celebrating going on both by the VA and the drug suppliers? And the drug companies swelling with pride on the profit realized. Both sides jumping up and down. Were they happy? You bet!

However, now the rest of the story. In my piece posted here, April 19, 2008, I mention this. Millions of veterans' are charged $16, double the copay cost, for, using the Board of Veterans' Appeals court description, of an "actual dispensed" 30-day $8 medication supply. Why is this? Because, the prescription requires that a pill supply be split. Add up the one month split pill supply for the "actual dispensed" $8 supply over-charges by the well over 1.1 million prescriptions. Veterans $8 copay, as mentioned, of 30-day supplies can be anywhere from 30,60,90, or whatever. But if the veteran is required to split these pill supplies, the VA doubles, or over-charges, in the copay cost.

Often we forget the little guy, the SMB, in our discussions of the comings and goings of the Internet marketing industry. Sure there are times like this when a report surfaces talking about their issues and concerns but, for the most part, we like to talk about big brands and how they do the Internet marketing thing well or not so well.

Comments for this post are closed