• Facts on Medicare administrative costs

    Holman Jenkins wrote in the Wall Street Journal last night,

    Many on the left tell us the solution is Medicare-for-All, because Medicare is so much more efficient than private insurers, spending a mere 2% on overhead compared to 20% or higher for private plans. […]

    This requires overlooking a lot. Even if overhead-to-medical spending were the right measure, much of Medicare’s overhead is hidden on the books of other agencies, including Health and Human Services, which provides management, and the IRS, which handles revenues.

    Sounds convincing, doesn’t it? But I am not convinced Jenkins has considered all the facts. They’re out there. For example,

    [A] portion of IRS costs are allocated to Medicare’s overhead by OACT [CMS’s Office of the Actuary]. […] [Also,] (1) the Social Security Administration, not the IRS, calculates and collects Part B premiums for the vast majority of Medicare enrollees, and the Railroad Retirement Board does so for former railroad workers; and (2) a portion of the SSA’s and the railroad board’s costs are allocated to Medicare’s overhead by OACT. […] OACT does include the cost of claims processing, which is done by what used to be called “carriers” and “intermediaries” and are now called “Medicare administrative contractors.” […]

    The federal agencies for which Treasury collects expenditure data, and which are therefore included in the trustees’ reports on Medicare administrative spending, include the Treasury Department, the IRS, the SSA, CMS, the Department of Health and Human Services, the Medicare Payment Advisory Commission, the Area Agency on Aging, the Department of Justice, the Federal Bureau of Investigation, and the Railroad Retirement Board (see the appendix). In addition, the appendix lists “quality improvement organizations,” which are private-sector organizations with which CMS contracts. The appendix also indicates that payments by CMS to insurance companies that process claims for Medicare’s original fee-for- service program are included in the trustees’ definition, as are the cost of buildings that house CMS staff and the cost of the numerous demonstration projects Congress requires CMS to conduct.

    This is not just buried in an academic journal article. You’ll find it freely available in my post. I’ve written more on this and other issues relevant to Jenkins’ piece. If you’re interested, read his (Google the title to get it ungated) and then these:



    • It requires willful stupidity to think administrative costs wouldn’t be lower. Has Holman Jenkins ever seen a billing office in a doctors office? Is it so hard to imagine one?

    • I note that he provided no numbers to accompany his claims, so I will translate. He said “much of”, not “most of”. That means less than half (else he would have said “much”). Therefore, we can conclude that whatever these unknown overheads are, even if he didn’t deceive in some other way, their total is less than twice 2%, or 4%.

    • Don’t forget that administrative costs should be measured per beneficiary. When using this measure, Medicare has higher administrative costs than private insurance:


      Also, this entire idea that comparing administrative costs is the same as comparing cost efficiency is absurd. Administrative costs can increase efficiency. Low administrative costs could just signal underinvestment in management and fraud detection among other services.

      • I help pay for Medicare. I pay a lot of taxes. I have private insurance, which I choose for my corporation. Why should I care what spending is per beneficiary? Most beneficiaries, especially in private plans spend almost nothing on health care. 50% of people account for 3% of health care spending. What I want to know is how much I am spending as a percentage of total spending. I know that there will be admin costs for all care. What percentage of what I spend for care, private, or taxes, goes to pay for that administration? (You can run numbers yourself if you want. Assume a private population and Medicare population each with 100 patients. Assume the private group has 10 sick and 90 healthy folks. Assume Medicare has 90 sick and 10 healthy people.


        • Obviously Medicare spends more per beneficiary. It is my
          sense that while this measure is far from perfect, it is a more
          accurate reflection of administrative costs than as % of total
          spending. It seems, at least to me, that the amount of
          beneficiaries would matter more than the total amount of spending.
          Nothing you have said does much to persuade me otherwise. Still, I
          don’t want to get stuck on this point because my larger point is
          that low administrative costs don’t indicate efficiency

      • Measuring costs per beneficiary is a silly measure especially here where you are comparing over 65s to under 65s. Is there any question that the average 79 year old sees a doctor more frequently than the average 40 year old?

        Has anyone looked at the overhead cost on the other side of the ledger? Do providers spend more on billing private insurance than Medicare?

      • That analysis is interesting but he undermines his own point inside it. He says, “Medicare beneficiaries are by definition elderly, disabled, or patients with end-stage renal disease.”

        So far so good.

        Then he says, ” Naturally, Medicare beneficiaries need, on average, more health care services than those who are privately insured. Yet the bulk of administrative costs are incurred on a fixed program-level or a per-beneficiary basis.”

        I don’t get why there is a “yet” in the middle of that. If everyone needs more healthcare, of course there is going to be more costs on a per beneficiary basis. For Medicare, nearly 100% of their patients are dealing with the medical system ALL THE TIME. Contrast that with a private insurer who have several beneficiaries who see the doctor once a year, don’t need prescriptions, and have healthy family members.

        The fair comparison to make would be to isolate a set of patients from the private insurance market who have similar health problems to Medicare patients and ask how much of their care is used up in administrative costs.

        It may be lower, but arguing that as a fixed dollar amount private admin costs are lower therefore private insurance is more efficient ignores the reality that administration is a lot cheaper when most of your patients aren’t sick.

        Finally, you can take single payer systems outside of the country which are vastly cheaper administratively on both a percentage and actual dollar basis.

        • Looking at the administrative costs as a share of total costs and claiming that that proves Medicare has lower administrative costs is also a misleading and incomplete comparison.

          In an ideal world, we would have the kind of comparison you mention, but we don’t. Medicare advantage, by the way, does not qualify as a true apples to apples comparison for a number of reasons.

          Anyways, the main point I am trying to make is that low administrative costs could actually indicate less cost efficiency not more.

      • Per beneficiary? Well yes, because Medicare enrollees are by definition older and generally sicker than the rest of the population so it’s not hard to imagine higher per beneficiary administrative costs (especially since we’re talking fee-for-service here).

        And yes, I agree that administrative costs do not equal cost efficiency. But we as Americans love starving the governmental beast, right? Which is why we often criticize inefficiency in the same breath as promoting cuts in vital services.

        • I doubt many of the Americans on here are fans of “starving the beast”, but I am puzzled as to why you think cutting funding for programs somehow decreases cost efficiency. I understand opposing cuts to programs, but, if anything, cuts should increase cost efficiency.

      • logic fail
        as your heritage foundation study notes, medicaid patients tend to be older and sicker, and require a lot more care
        to be hypothetical:
        a private plan that doesn’t accept people with pre existing conditions, and has a largely young and healthy pool; their cost will be low, cause those people don’t use a lot of health care

        a medicaid plan that has old sick people, moving from short term nursing homes to ERs to physical therapy to …

        of course your costs are gonna be higher !!!!!!
        I mean, I look at older family members; the amount of care they get, it must be an administrative nightmare keeping track of all the different doctors, hospitals, aids, ambulances, etc etc etc

        so, come back when you ahve a fair apples to apples study

    • He trots out the whole “fraud prevention” argument, too. That’s always amused me, because fraud prevention is usually done on a cost-effective basis – i.e. “does spending X amount of money preventing Y amount of fraud make sense?”. If Medicare still costs significantly less in administrative costs even factoring in some degree of fraud unreported, then why bother?

      • Medicare doesn’t cost significantly less. Indeed, per beneficiary costs are a good deal higher. This is an mostly because Medicare has sicker and older patients, so I’m not arguing that that is an indictment of Medicare.

        Medicare may well be more cost efficient than private insurance. I personally don’t think it is, but I could be wrong on that. However, the administrative costs argument will not convince anyone who actually understands what administrative costs are. These people realize that lower administrative costs do not equal more cost efficiency.

    • Comparing private administration costs to Medicare administrative costs is NOT comparing apples to apples. It is more like comparing apples to nuts.

      John adds a very important point. It is less expensive to administer one claim of $750 than three claims of $250 each. The degree of illness really doesn’t matter nor does the age. It is how many claims are filed and for the same amount of money the private insurer has to administer around 3 times as many claims (based upon average cost comparison per capita Medicare vs private).

      But there are more reasons. I wonder if any of these reasons were taken into account.

      1)Part of the costs insurers have are taxes imposed by governments that have little to do with their efficiency.
      2)Medicare doesn’t count its $60Billion in annual fraud
      3)The costs of the FBI and other agencies sent in to monitor fraud.
      4)Promotion of Medicare
      5)Cost of capital
      6)Rent and Utilities that are not paid for by Medicare yet exist.
      7)Salaries of legislators, staff and all the surrounding costs etc. not included in Medicare’s filings. Consider what the costs are for all of Congress (and down the line) and then consider the portion of those costs that were due to Congressional action on Medicare and Medicare related items.
      8)Expenses Medicare turns over to the providers that insurers must add to administrative costs.

      There are many more items one can add if they think about things, but one must recognize that what is being compared is Medicare’s administrative costs less their hidden administrative costs. There are no hidden administrative costs where private insurers are concerned.

      • please provide documentation of your claim
        “It is less expensive to administer one claim of $750 than three claims of $250 each. ”

        Or, more precisely, that it is less expensive to administer one claim of 50,000 dollars, over 6 months, for multiple providers at multiple sites, then 100 office visits at 250 a pop…

        I just saw this 1st hand with a relative; in the nursing home, then ambulance to the ER, then inpatient, then nursing home, then….
        and this has been going on for two or three months..a huge trail of paper, from people at multiple vendors.etc etc
        but if you have some data, I am open to change my view

    • The IRS bearing the costs of “handling revenue” collection
      is some sort of underhanded false accounting — that is efficiency!
      The IRS is already handling revenue; CMS doesn’t need its own
      revenue collection arm, unlike private insurers.

      • Fair point. But, I’ll just add that the benefits of competition and market incentives far outweigh the costs of profits, advertising, and decentralized revenue collection.

        I’ll also add that private insurance premiums do not create the deadweight loss that taxation creates.

        • Do they? 2-3% overhead vs ~30% overhead? I’m no economist but even if it’s a “value judgment” then 97-3 sounds immensely better than 70-30, when the benefit is some vague notion of “competition.”

    • Medicare also contributes to the cost of medical education:
      recently stated as $8 Billion for post-graduate medical education,
      i.e., residents. Since this contribution does not reimburse the
      expense of any direct healthcare encounter, should it be considered
      as an administrative expense? Also, medicare reimburses the medical
      schools at a higher rate for health care offered by its employed
      physicians. Is this extra-reimbusement acknowledged in any of the
      medicare accounting statements? And, if its not an administrative
      expense, should Medicare have any involvement in medical education
      support at all? Finally, the failure of Medicare to expand its
      financial contribution to post-graduate medical education
      represents yet another insult to anyone considering a future
      professional commitment to life as a physician! The level of
      paradigm paralysis is profoundly pervasive for our nation’s entire
      healthcare industry.

    • Many on the left tell us the solution is
      How a medicaid for all. Allow anyone
      to get medicaid at medicaid reimbursement rates and see what
      happens. Most people might stick with private insurance. With more
      enrollees Medicaid would have more pricing power. You could phase
      out Medicare and put the retired people on medicaid along with
      Government employees. Then eliminate tax deductability of
      insurance. I think you could do in such a way that government
      spending on health care would not increase. But the politicians
      would never do that because old people vote.

    • There is dead weight loss to taxation. Some info on
      Medicare funding for the commentors above: Medicare, which provides health care to about
      50 million elderly or disabled Americans, is financed through a
      combination of funding streams: a Medicare payroll tax; general
      revenue (mostly from federal income taxes); premiums paid by
      Medicare users; and a tax on Social Security benefits and state
      payments toward the prescription drug benefit. We’ll go a little
      deeper in the weeds. The program is financed through two separate
      trust fund accounts: the Hospital Insurance (HI) Trust Fund and the
      Supplementary Medical Insurance (SMI) Trust Fund. Medicare payroll
      taxes go to the HI trust fund, which is primarily used to pay for
      inpatient hospital stays (otherwise known as Medicare Part A). The
      SMI trust fund is used to pay for physicians visits, outpatient
      care and prescription drug benefits (Medicare Part B and Part D).
      The SMI trust fund is funded primarily (about 75 percent) through
      general revenue, with most of the rest coming from patient
      premiums. Let’s start with the first half of Pawlenty’s statement,
      that “only about 50 percent” of Medicare is paid for by either
      premiums or payroll taxes. According to the 2011 Annual Report of
      the Boards of Trustees of the Federal Hospital Insurance and the
      Federal Supplementary Medical Insurance Trust Funds, Medicare
      expenditures came to $523 billion in 2010. Here’s a breakdown of
      the revenue side of the equation: • Payroll taxes, $182 billion •
      Premiums, $61 billion • Taxation of benefits, $13.8 billion •
      General revenue, $205 billion • Transfers from states for
      prescription drug benefits, $4 billion If you break out premiums
      and payroll taxes, Pawlenty is correct that they covered about half
      the cost of Medicare in 2010.
      Also, you can only tax
      people so much, so all taxes go into one bucket. The separate SS
      and medicare taxes exist only to fool the voters into thinking they
      paid for their benefit thus making it more politically difficult to
      change the programs.

    • just think, if we did have medicaid for all, we wouldn’t have to waste time on this stuff, all we would have to do would be to find ways to make care more efficient.
      I mean, what would you do with all your spare time ?
      (this goes to all the posts on the relative merits of GOP plans for national tax credit/exchange care.)

    • Holman Jenkins is correct to point out that many on the left suggest adopting a Medicare-For-All system. Some view this as a panacea of sorts, as it will simultaneously reduce the numbers of uninsured and bend the cost curve. Whether Medicare-For-All would indeed be able to accomplish the dual goals of improved access and cost containment is an empirical question that is highly unlikely to ever be answered. What can be said is that Medicare as currently constituted suffers from a multitude of issues that prevent it from being a system worth emulating from an efficiency perspective. To illustrate, I will focus on Medicare’s procurement policies regarding drugs and biologics. First, the peculiar mixture of market and administrative pricing, such as Medicare Part B’s Average Sales Price + 6%, is not conducive to minimizing cost. In fact, such formula’s often introduce perverse incentives which lead to higher, not lower prices of physician-administered drugs. Second, Medicare’s explicit prohibition of considering costs in its deliberations on whether new medical technologies are deemed “reasonable and necessary” and therefore reimbursable would appear to be inconsistent with the objective of efficient purchasing. Third, Medicare’s fixed co-payments, such as Part B’s 20% co-insurance for all drugs and vaccines, bear no relationship to the products’ value. In other words, a patient pays 20% regardless of the product’s price or its effectiveness.

      From the left’s perspective a VA-For-All system would be a more suitable choice than Medicare-For-All. The VA’s system of hospitals, doctors, and pharmacies resembles more closely the single payer systems of, say, England and Australia, than Medicare. Moreover, the VA operates under a rather hard budget constraint (something conspicuously absent with Medicare), adopts a monopsonist position in terms of pricing negotiations with product manufacturers, and manages a national formulary. Nevertheless, be careful what you wish for. Should the U.S. ever implement a VA-For-All system of health care (even less likely than the remote possibility of Medicare-For-All), patients would encounter problems similar to the ones, say, NHS beneficiaries are faced with: Rationing (particularly of newer, more expensive medical technologies), under-utilization of diagnostics and therapeutics, and chronically under-funded hospitals and clinics.

    • Hi folks. You seem to be an intelligent group. Would you please comment on this:
      The 2011 Medicare report shows that the total monthly outlay for covering all beneficiaries averages $939.60 per month. Here’s from the official report:

      In 2011, Medicare covered 48.7 million people: 40.4 million aged 65
      and older, and 8.3 million disabled. About 25 percent of these
      beneficiaries have chosen to enroll in Part C private health plans that
      contract with Medicare to provide Part A and Part B health services.
      Total expenditures in 2011 were $549.1 billion. Total income was
      $530.0 billion, which consisted of $514.8 billion in non-interest
      income and $15.2 billion in interest earnings. Assets held in special
      issue U.S. Treasury securities decreased to $324.9 billion

      Does this mean that with all the sick and disabled people, end of life, fraud and abuse, etc., that Medicare costs less per month than a silver health plan on the exchange for a 60 year old before subsidy?

      Does this also mean that it makes sense to cover low income 60-64 year olds in Medicare, rather than give subsidies on the exchange? After all, what would be the monthly cost through Medicare for those not at end of life?

      • Maybe Medicare should cover 60 year olds, but $949 is more than the cost of a Silver policy for a 60 year old in my area, and I live in one of the most expensive areas. I can’t find an area with a premium greater than $949 for a 60 year old.

        • Cardinal,
          Thanks for the reply.
          The Health Net premium is $929. But what would a premium be for a 75 year old? $3,000?
          My point is that if Medicare were available for 60 year olds, the total premium would be much less.
          Right now, Medicare part A costs $441 per month without work credits. Add that to parts B and D and you still come out around $600. And adding a supplement still puts you at less than the $929 for a 60 year old.
          Why not have voluntary part A for all with an open market for all supplemental plans. That would keep a free market, while doing away with most medical bankruptcies.
          And you could charge enough for the 18-64 year olds to cover the shortfall for the older folks, and still remain very competitive in the marketplace.

          • I see. Sorry I didn’t understand the full argument you were making. You are saying, then, that if we allowed the 60-64 age group into Medicare, the cost per beneficiary would be less than the cost being paid to insurance companies for those 60-64 year olds? What is the average premium paid for that age group, counting the people on employer insurance whose premiums are less?

            And you are further saying that if we allowed 60-64 year olds into Medicare, that would be cheaper than paying the subsidies for the 60-64 year olds who are buying individual insurance and are getting subsidies? Wouldn’t that plan shift millions of 60-64 year olds onto Medicare, which we pay for, instead of private insurance, which they or their employers pay for?

            • No, I’m saying it could be made an option, the profit from which could be used to reduce the deficit for the senior population in the future.

      • Jeffrey, I think there’s little question that covering a broader age group in Medicare would cost less, in total, than covering the group with private insurance. The real issue is “cost less to whom?” and “who pays?”