Roland Benjamin: The Problems with Medicare-for-All

Hyperbole aside, Robert Whitcomb’s Projo op-ed from October 19 can be summarized by his one statement:

In short, extend Medicare to everyone.

He proceeds to use exaggerated estimations of private insurance overhead costs and completes his argument by saying:

Then we would not have to hang our heads in shame that Americans are the most unhealthy people of any developed nation.

The administrative assumptions of Whitcomb’s argument are suspect at best, and fatally flawed at worst. A white paper from the Manhattan Institute, published a few weeks ago, analyzes the holes in the Medicare-for-All debate. To support his claim of efficiency, Whitcomb estimates that non-benefit expenses in the private insurance market are near 25%, while the same expenses are only 2% for Medicare administration.

The Manhattan paper cites all the usual sources from both sides of the debate like Krugman, Council on Affordable Health Insurance, Rand, Commonwealth, Kaiser Family Foundation, etc. and points out a few interesting analyses. Notably, administrative costs of Medicare are nowhere near the 2% Whitcomb describes in his piece (nor are private costs anywhere near the 25% number). The paper uses conservative estimates of 14% private (while citing estimates as low as 11%) and 6% Medicare to make the claim that the theorized savings by transition are impossible to realize. Private carrier rates are confirmed locally by a study from the Rhode Island Office of the Health Insurance Commissioner citing total benefit payouts by carriers at 85.2% of premiums (or 14.8% non-benefit spending, including some percentage for taxes). The Medicare rate shifts dramatically, for example, when fraud expenses are classified as non-benefit expense instead of actual benefit spending. An apples to apples comparison using consistent accounting methods shows the true cost of Medicare between 6% and 8%.

The Manhattan study then goes on to describe the evolving nature of spending by those who would be newly insured under the policy and underscores the deficit that will be realized based on static assumptions. Finally, the white paper describes the response to tax policy that will inevitably result. Migrating the funding system to a Medicare style payroll tax will increase explicit labor cost in the US.

Currently, payroll taxes cover Medicare, private insurance premiums cover the health care of active employees and dependents and some income taxes cover another portion of health care spending. In the three buckets (or more), the costs tied to labor are less explicit and more dispersed. Take all of those costs and apply them in one bucket that is explicitly tied to labor and there will be a problem. Like it or not, corporate behavior responds to tax policy with amazing agility and immediacy. Increase the cost to employ directly, and work will move to regions where labor costs less. Or in the best case, capital investments facing a lower ROI bar will be exploited where labor can be replaced with automation. Reduce the workforce, or the value of a skilled workforce, and the tax base drops, perpetuating the spending deficit in an otherwise well intentioned policy.

The economic argument of Medicare for All has the underpinning that the cost of the uninsured is at the root of the health care inflationary problem. Yet I have not seen convincing data that this is true. No studies that I have seen refute either of these two facts:

• Uncompensated care to the uninsured is not a top tier cost driver. Kaiser shows uncompensated care around 2.05% of all health care spending, while others show the impact at less than one percent. Defensive medicine accounts for anywhere from 3 to 10 times that amount depending on the definition. Rx sales of the top 15 drugs alone exceed the total amount of uncompensated care. And so on and so on…
• Uncompensated care to the uninsured has not inflated at the same pace as overall spending. In fact, the inflation rate of spending by the uninsured has kept pace or lagged behind normal inflation since 2000 (this comes right from the Medical Expenditure Panel Survey MEPS). Meanwhile, the spending by the insured has increased at alarming rates. The economic problem lies with those currently insured. Bringing more individuals into that system, and providing the same or more insulation from real prices will only aggravate the situation.

Since most of the uncompensated care is funded by government sources, the real uncompensated costs are relatively small. So the economic case for Medicare for All makes some pretty big leaps. There is little evidence that supports the economic urgency argued by Medicare for All advocates.

The only argument to be made for insuring all via Medicare type arrangement then is a moral one. While well intentioned, the demographics of the uninsured suggest that the moral argument may be shaky as well. With 93% either opting out of available insurance for non-affordability reasons, without insurance for less than 4 months, or without citizenship status, there may be as few as 4 million people who would truly benefit from the coverage. These are the people that are at or beyond their ability to afford or access coverage today. This would also be the population most likely impacted when the inevitable labor shifts occur in response to changing tax policy. To use the same hyperbole that Whitcomb employs, would it be moral to give someone insurance, only to have their job become prematurely obsolete or permanently displaced?