Health Care Spending: The Basics



As a nation, we are spending well over $2 trillion a year on health care. This includes: all of the money that you and I pay out-of pocket to cover co-pays, deductibles and drugs; the dollars that you and I (and our employers) fork over for private insurance; the money Medicare, Medicaid and SCHIP lay out to reimburse doctors, hospitals and patients; the billions taxpayers chip in to fund veterans’ health programs, public hospitals, school programs, and health insurance for government employees as well as the money private charities contribute to health care.

What exactly are we paying for? How much of that money is used to pay the CEOs of drug companies salaries that read like telephone numbers? How much do hospitals eat up?  How much is spent on insurance company ads? How much is used to provide healthcare for the poor?

I’ve decided to do a series of posts spelling out exactly where the money goes. Today, I’m going to start with private insurance.

Many people believe that if we just eliminated the private insurance industry, healthcare would become much more affordable. There is a general sense that the “administrative costs” of private insurance are siphoning off a sizable share of our health care dollars.

There is some truth to that: because we  have  multiple insurers—not to mention so many solo practitioners, small hospitals, clinics, and individuals filing for reimbursement—the paperwork is enormous. If we had only one big insurance company that used just one set of forms we could simplify the paperwork greatly. People who want a “single payer” system, with the government paying all of the bills,  point out that the savings would be enormous.

And we could cut costs even more if, instead of having tens of thousands of health care providers filing for separate reimbursements, doctors, hospitals and clinics joined together into, say, eight our ten large organizations like Kaiser Permanente, each with its own back office.  The doctors would be on salary, so rather than filing for payment for each service they performed, they would receive a monthly check for taking care of their patients, just as they do at Kaiser Permanent or the Mayo Clinic (where doctors are on salary).

In other words, it is not only a fragmented multi-payer insurance industry that generates so much paperwork; on the other side of the transaction a fractured network of separate providers adds to a mind-boggling stack of paper. Unlike most other developed countries, we have turned healthcare into a  cottage industry. This gives us lots of choices: we can select from a Chinese menu of insurance plans and proviers. But it also means higher administrative costs. In this post I would like to focus first on just on how much our huge private insurance industry is costing us. (In a later post, we’ll look at the price we pay for a fee-for-service system of independent providers.)

In our highly competitive health care economy insurers have many
expenses, including advertising, marketing and lobbying. They also must
pay their employees. And, to compete with other insurers for “top
talent,” they feel obliged to pay their executives salaries that some
call “obscene.” In addition, in states where an insurer can charge you
more if you are sick,  health plans spend enormous sums on
“underwriting”—which means figuring out how much more  to charge if you
have a pre-existing condition. They also have to pay bonuses to
eagle-eyed employees who comb through the claims—looking for reasons
not to reimburse you or your doctor. Finally, they have to generate
profits for their shareholders. And investors expect those earnings to
grow, year after year.

How much does all of that cost? Every insurer pays out a certain
percentage of the premiums that it takes in to reimburse patients and
health care providers; what is left over covers expenses and profits.

Wall Street calls the share of premiums that the company is actually
forced to spend on health care its “medical loss ratio.” For example,
in 2006, UnitedHealth Group (UNH) had a “medical loss ratio” of
81.73%;  i.e. for every $1 million of premiums that it took it, in
spent 81.73% (roughly $817,000)  reimbursing claims. This left it with
with a little over $183,000 to cover advertising, marketing,
underwriting, salaries, etc.—plus profits for investors.

If UNH had been less efficient, it would have had a higher medical loss
ratio. In other words, it would have paid out a larger share of its
revenue in reimbursements, and its share price would no doubt have gone
down.  (From Wall Street’s point of view, the goal is for an insurance
company to pay out as little as possible.)

If I became czarina tomorrow, waved my wand, and closed own the entire
private health insurance industry, how much could we save? Would our $2
trillion health care bill fall to $1.8 trillion? $1.6 trillion?

Here is the deeply disappointing answer. All of the money that private
health insurance companies lay out for something other than health
care—that $183,000 out of very million that UNH spreads around paying
for advertising, underwriting, executive bonuses  and lobbying —amounts
to just 4.5 percent of our nation’s $2 trillion dollar annual health
care tab.

Okay, 4.5 percent of $2.2 trillion is “real money.” But  keep in mind
that in 2005, total health care spending in the U.S. grew by 6.9
percent. So the 4.5 percent that I just shaved from our health care
bill by snuffing out the private health insurance industry (and
shifting to a single-payer government plan) would be wiped out by
inflation in one year.

Moreover, a single-payer system also would have administrative
expenses. Not as many, to be sure. The government doesn’t have to
advertise or lobby itself. And it doesn’t pay its executives nearly as
well as the private sector. Nevertheless, the government’s
administrative expenses now equal 2.2 percent of our health care bill.
If we expanded Medicare to cover everyone, those administrative
expenses would grow. So the savings I achieved by wiping out private
insurers could easily fall to just 3.5 percent of our total health care

None of this is an argument for or against a single-payer system. My
goal here is only to illustrate that no single sector of our health
care economy is responsible for the high cost of health care in the
Drug-makers, device-makers, insurers, some hospitals, some
physicians, and some patients…together we all are driving runaway
health care inflation.

The pie chart below shows you how much of the $2 trillion pie goes to
private insurers. In future posts, I’ll color in the other slices.


16 thoughts on “Health Care Spending: The Basics

  1. Are you also counting for compliance costs and for employees working in doctors offices? Many doctors have an employee who’s job it is to ensure patients have the insurance they say they have, check what the insurance covers, fill out billing forms, etc. The clinic I go to serves the students at a very small (<1000 student) graduate school employees 4 such people. Moving to a single payer system would eliminate these positions too.
    There are also other costs to imposed by private insurance companies. For example, the dentist office that i worked at had to undergo a yearly audit to make sure they were collecting the co-pays. there was also a fairly large amount of payments that were refused initially...these problems with insurance companies took a lot of time and money to resolve.

  2. correction:
    the school has less than a 1000 students, the clinic only sees a little over half of them as we have two geographically separate campuses.

  3. I think the best paper that I’ve ever seen on administrative costs in both the healthcare and health insurance industry was written by Ken Thorpe and was published in Health Affairs in 1992. Well worth a read.

  4. With due respect I don’t believe the 4.5% admin cost number.Agree with BK that there are other indirect and hidden costs to both providers and consumers who constantly fight with insurance companies or try to comply with inane/insane rules/regs.
    Whew! It’s bad.
    I DO agree with Maggie that the health care cost issue is a multifactoral problem.
    Dr. Rick Lippin

  5. John, BK and Barry–
    John– thanks for the link to the data.
    BK– As I explain in the post, the administrative costs at the provider’s end are a separate cost that shows up in your doctor’s bill and in your hospital bill, not in your insurance premiums..
    And as I say in the post if there was only one insurer (using one form) those costs would be smaller.
    But even under a public sector “Medicare for all” plan some of the administrative costs would remain. For example, providers would still have to be audited to make sure that they were collecting co-pays –there is a lot of fraud in Medicare and Medicaid.
    There would also be claims that would be refused. (Medicare is beginning to crack down on unncessary treatments and will be doing more of that in the future–they have to or we’ll run out of money.)
    Finally, and most importantly, as I point out in the post, administrative costs are high at the doctor’s end of the transaction because most doctors are charging “fee-for-service” and so they have to submit a claim for every service they provide.
    If they were paid a set amount per patient to take care of that patient for a year, and received a salary that reflected how many patients they cared for, you would not have millions of separate fee-for-serivce claims.
    Medicare is now thinking about switching away from fee-for-service (which creates perverse incentives to over-treat, over-test, etc. in ways that can be hazardous to a patient’s health and very expensive for all of us.)
    The Medicare Advisory Commission is talking about paying doctors (and the hospitals they refer to) for the patients they care for, plus bonuses if they operate more efficiently than their peers (keeping the patients well so that they don’t need to be hospitalized, and if they do need to be hospitalized, avoiding hospital errors and following up with check-ups at the doctor’s office so that they don’t have to be re-hospitalized, etc.
    Barry– I’ll look up that article-
    Thanks, Maggie

  6. Dr Rick —
    You write: “With due respect I don’t believe the 4.5% admin cost number.”
    I really don’t know what to say to you. No one has ever disputed that number. (It’s in a huge pie graph at the begining of my book which has been out for 1 1/2 years and reviewed in medical journals including “Health Affairs”; it’s a number than many others have cited.
    Finally, anyone who can do 4th grade arithmetic can calculate the number.
    You simply add up all of the premiums that insurance companies take in, then subtract how much they pay out to reimburse health care providers and patients (this is all public information because they are publicly traded companies.)
    Also, keep in mind that insurers have no reason to exaggerate how much they pay out–quite the opposite, their shareholders will buy their stock only if they pay out less than other insurers.
    Finally, after you’ve added up all the premiums, subtracted how much they pay out for medical care, you are left with what they keep for themselves (to pay salaries, pay for advertising, underwriing, profits for their shareholders etc.)
    You then divide $2 trillion dollars by the amount they keep for themselves, and you find it equal 4.5% of the $2 trillion.
    As I explain in the post there are also administrative costs at the other end–in the doctor’s office and in the hospital where providers are filing claims.
    Those costs could be reduced somewhat(but not eliminated) if we had single-payer–one insurer using one form.
    But even government health care programs like Medicare and Medicaid have administrative costs eating up roughly 2.5% of the $2 trillion we spend on healthcare.
    Medicare has to hire people to look at the claims, watch out for fraud, pay the claims etc.
    And if Medicare was covering everyone those government administrative costs would be roughly twice as high as they are now (the government now pays roughly half of the $2 trillion bill).
    The way to really reduce administrative costs at the provider’s end would be if instead of paying doctors “fee for service” (which means they file millions of claims for individual services–and are tempted to over-treat) we paid doctors a lump sum for taking care of a certain number of patients for a year–plus a bonus if they did a good job of managing their care and keeping them well.
    In other words, doctors would be on salary plus bonus. Doctors at the Mayo Clinic are on salary. Doctors at the Cleveland Clinic and Kaiser are on salary. And their costs are much lower.
    The Mayo Clinic can take care of Medicare patients for half of what it costs UCLA Hospital (where many doctors are in private practice) to take care of very similar patients . . .(see

  7. PNHP’s number is at $350 Billion per year, which I believe more in the 10-18% range…?
    That just the excess from the for-profit insurance industry and does NOT include other admin savings from rationalizing hospital administration, regional bulk purchasing, etc. The 31% is the total excess admin overhead system-wide compared to Canada. About 1/2 I believe (~15) due to for-profit fragmented insurance side, and 1/2 due to other admin overhead.
    Definitely more then 4.5%!!!

  8. Rick/Steve/BK——Maggie is absolutely right on the 4.5% (no one outside of the hard-core single-payer crowd takes the PNHP figure seriously). As to Medicare’s admin costs, she is probably being too conservative on the potential increase. Ben Zycher at the Manhattan Institute estimates 6%.

  9. Dr. Steve B–
    Please see my comment to Dr. Rick Lippin.
    The numbers are all publicly available and the math is really easy.
    You might suspect that the insurers have in some way fudged the numbers –except it is not in their interest to exaggerate how much they are spending on healthcare or to downplay how much they are keeping for themselves and their sharehodlers.
    Their shareholders want the insurers to spend as little as possible on reimbursements and keep as much as possible in profits.
    So the numbers are honest; the math simply requires adding, subtracting and dividing.
    As it happens, I just talked to Dr. Matt –who is part of PNHP (Physicians for a National Health Plan)–and as he noted, they are not economists. That’s why he’s asking me to speak on a panel.

  10. Dr. Steve B–
    Please see my comment to Dr. Rick Lippin.
    The numbers are all publicly available and the math is really easy.
    You might suspect that the insurers have in some way fudged the numbers –except it is not in their interest to exaggerate how much they are spending on healthcare or to downplay how much they are keeping for themselves and their sharehodlers.
    Their shareholders want the insurers to spend as little as possible on reimbursements and keep as much as possible in profits.
    So the numbers are honest; the math simply requires adding, subtracting and dividing.
    As it happens, I just talked to Dr. Matt –who is part of PNHP (Physicians for a National Health Plan)–and as he noted, they are not economists. That’s why he’s asking me to speak on a panel.

  11. Maggie– You are right (roughly) about the overall savings to the system in insurance company overhead but you’re wrong to claim that this would be the total savings from a single payer system. As you note, while insurance company overhead averages 20%, or about 17% more than Medicare, private insurance pays only 35-37% of total health care costs, so eliminating it would save about 5-6% of total system spending (Lewin estimated this at 5.3% in their study of the California Kuehl bill published January 2005).
    However, as several commentators have noted, there are numerous other savings that single payer would permit: Using Lewin’s calculations again, the savings in hospital billing and other administrative costs would be 1.9% of total system costs, savings in physician administration costs would reduce costs by 3.6%, bulk purchasing of drugs, etc., by 2.8%, and the primary care emphasis that would be possible in a unified, organized system, another 2.2%, for a total system savings of 15.8%.
    This is pretty close to the figure of 15% used by PNHP, which is drawn from both this kind of analysis and by comparison with the Canadian system.
    As Lewin and many others (including the GAO) have shown, these savings are sufficient to cover all the uninsured and to eliminate copays and deductibles, without spending any more than we are now spending.
    The point that is often ignored is that a single payer system would not only replace a costly private insurance system with a much more efficient public one, but it would eliminate the need for many other administrative costs and would permit savings in many other areas.
    It would also, of course, provide planning and budgeting tools that could be used to contain costs going forward into the future. No other plan now being discussed has any way of arresting the continuing escalation of health care costs.

  12. Dec 07 6:49–
    (You forgot to leave your name!)
    You are right that private insurers cover 35-37% of total health care bills; the govt’ covers about half. (The rest comes from the out-of-pocket payments patients make, and charities.)
    While covering half of our bill, the govt’s administrative costs equal just 2.5% of the total $2 trillion national bill. While covering a little over a third of that bill, private insurers generate costs of 4.5%. So private insurers’ overhead is clearly higher.
    But if the government took over that third that private insurance now covers, one could expect the government’s administrative costs might rise to equal another 1% of $2 trillion, for a total of 3.5%.
    But meanwhile we would save the entire 4.5 percent of $2 trillion that represents the private insurers’s costs.
    We wind up with a net savings of 3.5% (4.5% minus an addition 1% that govt would spend while picking up more of the bill) I know Lewin wants to say its over 5%–I’m afraid Lewin has its own biases. (As I’ve explained in comments on this thread before, the math is pretty easy and the numberw are all public.)
    But let’s pretend you save over 5% of $2 trillion in insurer’s administrative costs if you switch to single payer. Your savings are still wiped out by one years inflation as the cost of health care products and servcies continues to climb over 6% a year, year after year.
    And the rest of the numbers you cite have little or nothing to do with whether or not we have a single-payer system.
    For example a single-payer system does not suddenly give the government the right to negotiate for lower drug prices (which is how you get savings by buying drugs in bulk). Right law, the law explictly says that gov’t can’t do that.
    Medicare has the size and the clout to get significant discounts–if Congress would let it negotiate. We don’t need single-payer for that to happen. In fact, I’m pretty sure that we can get the law changed on that point in the 1st year of the next president’s administration. Single-payer has nothing to do with it.
    And once Medicare negotiates for discounts, private insurers will insist on the same discounts, just as they do in other countires that have a hybrid private sector/ public sector system.
    There is also no reason to assume that a single-payer system guarantees more preventive care.
    Other developed countries offer far more preventive not because they have a single-payer system (Most–Do Not. Offhand, the U.K. is the only country that I can think of that has a true single payer system)
    You get more preventive care in those other countries because they have more primary care docs and fewer specialists. They do this in various ways a) by limiting the number of medical students who can choose to specailize; b) by refusing to pay specialists the high salaries that we do, and c) by limiting the number of surgeries that specialists perform in a given year.
    So patients see fewer specailists, spend more time with primary care docs, get more preventive care, and outcomes are better. This is also true in the U.S., in states like Minnesota, where there are many more primary care docs and many fewer specialists than in Southern California. After adjusting for differences in race, sex, age and overall health of the population, it turns out that care in Minnesota costs half as much–and outcomes, patient satisfaction and health are better. So it could happen here, without single-payer. We just need to limit the supply of specialists and make more primary care available and affordable.
    In these other countires preventive care is often free (no co-pay or dedcutible) which encorages people to go for preventive care. And in contrast to the U.S. where we haver a shortage of priamry care docs and an excess of specailsts–you can easily find someone to provide preventive care.
    Meanwhile, these other countires refuse to cover some of the most sophisticated procedures, tests, treatments and medication we do here–or they pay doctors and hospitals very little for these treatments–becuase there isn’t enough medical evidence to show that benefits outweigh the risks.
    We, on the other hand, tend ot pay for anything that someone wants to sell, assuming that if it’s newer, it must be better. And we don’t regulate prices.
    Again, countries like Germany, France, Switzerland don’t have “single-payer” but they do regulate both private sector and public sector care.
    As for savings in doctors and hospital’s adminstrative costs, again single payer is not necessary to cut those costs.
    You could simply insist that all insurance companies use the same forms for a particular DRG (diagnosis–since hospitals are paid by diagnosis) or service (since docs are mostly paid fee for service.)
    This would mean requiring insurance companies to simplify their insurance, so that there are not so many differences between policies. (This is what Switzerland has done in its regulation of its private insurers and it works very well. )
    Finally, the big difference between the cost and quality of health care in the U.S. and in other developed countries comes in the fact that we squander one out of every three health care dollars on unncessary treatments, tests and procedures, as well as over-priced cutting-edge drugs and devices because Americans have been trained to believe that more care is always better care and that newer is always better.
    This overtreatment is bad for our health–which is a major reason why quality of care is poorer here. (Or at least not nearly as good as it should be given that we are spending twice as much.)
    High administrative costs are not hurting our health.
    Needless angioplasties, MRIs etc. are.
    We have three decades of medical evidence proving how wasteful our system is in this regard. I’ve written about this here–
    Please read it; I think you’ll find it interesting.

  13. Maggie–
    You are right that Medicare, and the insurance companies, could behave efficiently and could monitor and fund medical practice in a coordinated fashion — if they were willing to submit to the kind of regulation that the European governments impose on their insurance funds and private insurance companies. But a hundred years’ experience with regulation in the US doesn’t give one much hope that this can be achieved any time soon. Discussing how these could be achieved, in practice, in a single payer system has the virtues of both explicating what is possible and, at th e same time, providing a model — expansion of Medicare to cover everyone and improvement of its benefits — of how it could be achieved. Talking about is a possible, from the Wennberg data, without providing a practical model for how to achieve it, doesn’t really move us very far along toward real reform of the health care system.

  14. Len–
    Actually MedPac (the independent commission that advises Congress on Medicare spending) already is translating the DArtmouth research into practical models– as are Congressmen including Senator Clinton’s people.
    See recent Health Affairs articles in the most recent anniversary issue– two by Wennberg one
    by Lopert and Marily Moon titled “Toward a Rational Value-Based Drug Benefit for Medicare.”
    As Shannon Brownlee point out in her post on this blog today, Medicare already has pilot projects going. . . .A lot of Medicare reform takes place “under the radar” —
    I’ll be writing more specifically about Medicare reforms in coming posts.
    Single-payer can’t possiblly work unless we reform Medicare and use it as the model. We just can’t afford single payer if it is as wasteful and unregulated as Medicare–we can’t cover the 15% of the population not covered, raise Mediciad fees paid to providers to what they shoudl be (at least equal to Medicare fees) and keep up with 6% annual inflation.
    None of this is theoretical–it is all begininng to happen now. Few people read the MedPac reports. But Congress is going to have to make a choice this year between slashing the fees that it pays all doctors by 10% (which it won’t do) or acting on some of the Dartmouth reserach in terms of covering unncessary over-priced products and procedures.
    As to “a hundred years of regulation in the U.S.” not working–this is a myth that began in the 1980s, with the neo-cons and the Reagan administration.
    Prior to the 1980s, regulation of the utilities worked very well. (Deregulation brought us Enron.)
    REgulation of the airline industry worked fairly well (they were in far better shape than they are today.)
    FDA regulation has worked well during those periods of time when we had a good FDA chairman and the agency had funding.
    Airline safety regulation has worked well–much better than healthcare safety regulation–because the airlines took it seriously.

  15. Maggie, when you give the five percent figure for administrative and advertising costs, it does not look to me like you are including the huge burden of administrative costs that each and every office bears. Hours of red tape, faxing, phoning, and letters written to try to get things covered.
    Multiply the salaries of the people who are purely there to cope with insurance companies by the number hospitals, offices, radiology centers, labs, etc, and I cannot believe it would only be 5% of costs.
    Single payer would not eliminate the need for navigation of benefits, but removing the “gate-keeper” role of insurance and things would get easier and less expensive.
    Why does our health care cost so much more than the Canadian or English systems; not only insurance, but insurance is a big part, and I don’t think the difference is only 5%.
    What do you think?