Health Insurance Public Policy

U.S. spends $700 billion on unnecessary medical tests

Peter Orszag, director of the Congressional Budget Office, estimates that 5 percent of the nation’s gross domestic product-—$700 billion per year –goes to tests and procedures that do not actually improve health outcomes…The unreasonably high cost of health care in the United States is a deeply entrenched problem that must be attacked at its root.

This quotation comes from a Progressive Policy Institute (PPI) report.  There is little doubt that much of health care is unnecessary or at least is not worthwhile in the cost-benefit sense.  However, how do we fix this problem?  PPI has some suggestions which the Healthcare Economist will scrutinize.

  1. Prospective Payment.  Currently, a majority of physicians are paid on a fee-for-service basis.  This encourages physicians to work harder (they get paid more for doing more services), but also encourages them to recommend unnecessary treatments to patients.  My own research finds that when specialists are paid on a fee-for-service basis, surgery rates increase 78% compared to when they are paid on a capitation or salaried basis.  Using a prospective payment system would give physicians an incentive to under-provide services.  Further, insurance companies could collect rents by enrolling only healthier patients so that the cost of care would be less for these individuals.  Prospective payment could work for specific diagnoses (as in the DRG system), but one must worry about DRG creep.  Also, if there is a high variance in the cost of treating a specific disease, than a fee-for-service compensation may be superior.   While the prospective payment system does have appeal in cutting costs, it could reduce patient access to medical care as well.
  2. Let individuals choose their own plan. This proposition has great appeal for those who favor consumer choice.  Everyone likes choice.  However, issues of adverse selection can negate any welfare gains from additional choice.  High risk individuals have a hard time getting insurance and if they do the price is often unaffordable.  PPI suggests setting a up a local area purchasing pool to counteract the issues of adverse selection. I am not exactly sure how the PPI proposal would be implemented (do insurance companies charge a fixed rate for all individuals enrolled? do they adjust premium by age or sex?)  Would enrollment in this pooling mechanism be mandatory?  If so, this could drive down costs and significantly reduces issues of adverse selection.  However, mandatory enrollment in the pool could also reduce consumer choice.
  3. Create a “Health Fed”.  “A Health Fed, as former Sen. Tom Daschle has proposed, would set national goals for health-care spending and patient outcomes based on the potential gains for integrated care.”  This I think is a horrible idea.  Having the federal government try to reduce costs and improve quality at such a high level is likely to be expensive and counter-productive.  Spending money on medical is not a bad thing; good health is one of the items individuals value most in life.   Thus, if the federal government set medical spending limits, this could lead to rationing.  What we want to happen is to reduce medical spending for unnecessary or wasteful medical procedures.  I don’t think a “Health Fed” would be very helpful in accomplishing this goal.


  1. The problem with all prospective payment systems, DRGs, capitation, risk-sharing, profit sharing, and solutions such as individual health accounts, Medicare Part D… is that they all involve mismanaged risk assumption by entities incapable of handling these risks.

    These risk transfers are insurance risk transfers. Big insurers write lots of policies, achieving substantial overall risk reduction. Big insurers can charge less for identical policies because their operating results are not as variable as small insurers. So the risk premium a large insurer charges is less than the risk premium a smaller insurer should charge to manage the risks it assumes based solely on portfolio size. The myth is that insurance risks can be transferred for a payment that is, on average, adequate. The truth is that smaller insurers are less efficient than larger insurers.

    What does this mean? When risks are transferred to health care providers – I call it “Professional Caregiver Insurance Risk” they become very, very inefficient insurers. Two, equally clinically efficient, providers, one a fee-for-service and one capitated, must provide different levels of service to achieve the same overall financial results – i.e. profitability. The capitated provider must provide less service per dollar of revenue because it must prepare for the possibility of higher than expected losses at the end of the contract. The illusion, perhaps delusion, is that capitated providers are more efficient.

    In fact, the risk borne by a smaller insurer (provider) is a function of the square root of the differential in size for the ceding insurer’s portfolio and the accepting provider’s portfolio: 1/10th the size – 3.162 times the variability in loss ratios for the provider’s portfolio. To compensate for this increased risk – capitated providers must reduce their average costs – because when they are already efficient, they can’t reduce variability. So, for every dollar paid through an efficient capitation system, less service is bought then with traditional indemnity insurance.

    The flaw underlying arguments favoring risk sharing is that providers willingly harm patients by over-prescribing, over-testing, performing unnecessary diagnostic and treatment procedures, to generate more revenue, BUT that capitated providers will not under-prescribe, defer or refuse necessary diagnostic and treatment procedures, to reduce costs and increase their net revenues. This, of course, ought to be proven rather than merely assumed.

    A better approach would be to assume, based on the inherent inefficiency of transferring insurance risks to health care providers, that less service will be furnished when insurance risks are transferred, and that providers who would harm patients with unnecessary treatments will deny care that harms their patients if their economic self-interest is best served by doing that. True, it will always be a small percent of providers and affect different patients. Healthy patients are often hurt by fee-for-service over-treatment whereas sick patients are often hurt when capitated providers deny necessary treatments.

    All assumptions that transferring insurance risks to health care providers are credible ways to fund health care, ought to be proven, rather than assumed. Basic probability, statistical, and actuarial theory suggests that transferring insurance risks to smaller entities is fundamentally unsound. Not recognizing or understanding insurance and risk doesn’t mean these risks do not exist and that accepting these risks does not adversely affect provider operations.

Leave a Reply

Your email address will not be published. Required fields are marked *