In the last couple of days, my thoughts have turned to the question of how much Medicare for All legislation could save compared to the present non-system of health insurance. One estimate of the savings we often see in articles on Medicare for All is $350-400 billion annually. This a substantial amount and over 10 years might amount to over $4 trillion. However, I think it may reflect a serious underestimate. In this post, I’ll discuss five categories of savings and talk about the difficulties involved on pinning down good estimates. I’ll also come up with some wild-assed guesses (WAGs). I don’t think these will are very reliable, but they may be better than ignoring a category of savings entirely and arriving at a total savings estimate that’s a gross under-estimate. Also, I’m asking Firedogs who read this post to critique my estimates and contribute whatever they can to refining them. I’d like a real community effort here, to get as close as we can to the truth about this issue. In particular, I ask Scarecrow, hipparchia, ralphbon, Hugh, wesgpc, libbyliberal, masslib, selise, kip sullivan, Valley Girl, NathanAschbacher, HealthSustainomics, blub, lambertsrether, Ian Welsh, jon walker, and montanamaven, along with anyone else who’s been reading and writing about the issue of savings that might be expected from Medicare for All, or who sees a flaw in the estimates we’re coming up with, for help with this.
I see 5 categories of immediate savings from Medicare for All: 1) savings on Administration and profit elimination, 2) savings from reduction in volume of emergency room care, 3) savings in provider reimbursement, 4) savings in pharmaceutical costs and 5) savings from elimination of Medicare Advantage programs. Of course, a big source of potential error in this discussion is my overlooking a category of savings. So anyone who can suggest another category will immediately have a big impact on this analysis. Let’s now examine each of the categories of potential savings.
The estimate of $350 – $450 billion in annual savings from lower overhead charges and elimination of profits comes from a study by Steffie Woolhandler and David Himmelstein, published in 2003, but using 1999 data. They compared this category of costs in the United States and Canada and found that the United States’s Administrative costs exceeded Canada’s by 15.6% of total national health expenditures. When that percentage is multiplied by a 2009 forecast of National Health Expenditures (NHE) of “$2.5-2.6 trillion; 15.6% of that would be $390-405 billion.”
There other ways to estimate this category of savings, however. The combination of Administrative Costs and profit margins in private health insurance is about 30% of its premium revenue, while Medicare’s Administrative costs are evidently about 2.5% its expenditures. So, theoretically, there should be a 27.5% savings there if we assume that Medicare provides all insurance for essential care and spends. Recently, the Kaiser Family Foundation released the results of a study which found that the average premium cost health insurance provided by one’s employer is $13,375 per family, and $4824 for individuals. Not having easy access to the number of family and individual policies issued, I assume that the average premium per covered individual across both types of policies is roughly, $4483. From recent census data for 2008, I estimated that roughly 168,000,000 are getting their insurance from private companies alone. The product of these last two numbers estimates private insurance company revenue in 2008 at about $752,000,000,000. This is an under-estimate because it doesn’t include people who have Government insurance, but also buy private insurance supplements. From the census data, I estimated that there are roughly, 33,257,000 with joint coverage, but I’m not at all clear on how to estimate the amount of revenue generated this way, or the overhead/profit margins involved. So, perhaps someone else can help with a better estimate of this. In any event, using the $752,000,000,000 and multiplying by 0. 275, we get $207 billion in Medicare for All Savings rather than $400 billion. Even adding in revenue from the 33 million or so who have supplemental private insurance, it seems doubtful that this estimate would get anywhere near $400 billion. This makes me uncomfortable because it’s nowhere near the Woolhandler – Himmelstein estimate, so I’d appreciate learning about the flaws in this alternative estimate, if someone has some comments on it.
Moving to the second category, many have made the point that health care providers pass through the cost of treating the uninsured in emergency rooms to insurers, both private and Government, and also that the amount passed through is $1100 per insured person annually. The total of insured persons, both in the private and Government-insured categories, was roughly, 255,000,000 in 2008. So, the total expenditures passed through by providers is about $281 billion annually. If Medicare for All were covering everyone, these emergency room expenditures for routine care of the uninsured would be saved.
Savings in provider reimbursement arise because Medicare has lower reimbursement rates than private insurance companies. For our purposes here, I’ll use a multiplier of 80% of private sector rates after overhead and profit are subtracted from premiums. This translates to $109 billion in savings we can expect from Medicare for All.
Americans currently spend roughly $250 billion per year on prescription drugs. These drugs cost about 40% less in Canada than they do here. This suggests that if a Medicare for All program negotiated with the pharmaceutical companies it might save $100 billion per year.
Lastly, the Medicare Advantage program is said to cost Medicare $150 billion in subsidies for private insurers annually. Passing Medicare for All would save this expense.
Admittedly, the above savings estimates are rough and open to criticism. The question I’m raising here, however, is whether they’re in the ballpark? If so, savings to be expected from passing Medicare for All total roughly $846 billion annually, and not $400 billion. Also, if my reasoning about the first category of savings is wrong and the Woolhandler-Himmelstein-derived estimate still applies, then the savings that might be expected from Medicare for All would be more than $1.040 trillion annually. Even if the first of these estimates for total savings is correct, Medicare for all could be expected to reduce health care expenditures as a percent of GDP from 17.8% to 11.7% as soon as it was implemented.
Finally, what I’m trying to do with this analysis is to show that the impact on savings if we implement Medicare for All is much greater than many of us previously thought, and that, provided the analysis is correct in outline, if not in every detail, the expected impact of implementing it is so great that there really is no other reasonable alternative to it that Congress is now considering. I don’t know if I’ve really made that case successfully, but I leave it to you to deconstruct or reconstruct my thinking, and then we”ll know better.
(Also posted at firedoglake.com where there may be more comments)