Uncategorized

Single payer = single data point

Bernie Sanders and others are advocating a Medicare for All system.  At first glance this is a sensible approach.  Why not allow the government to provide equitable health insurance coverage for all?  Don’t we want to let all Americans have high quality care?  Why not use the power of the federal government to help negotiate lower rates for physicians, hospitals and providers and drive down drug prices?

The key issue is one of information.  Consider the case of a State Medicaid Agency (SMA).  Let’s say there is a new treatment and they need to decide whether or not to cover it.  Oftentimes the SMA may look to other states or commercial insurers within their state to see how they are doing this.  By having a decentralized system, one can learn from others without putting everyone’s eggs in one basket.  In a centralized system, there is only one US-based data point for policymakers to go off of.

Second, how are prices to be set?  For instance, Medicare currently has a Medicare Wage Index for hospitals and geographic practice cost index (GPCI) for physicians which adjust payments based on the wage rate in local areas.  The government is able to do this because wage rates for health care practitioners depends on reimbursement from both public and private payers. If the federal government paid the vast majority of provider wages, it is not clear how one would determine whether the government should pay more or less in certain areas, despite clear differences in cost of living.  If the government overpays, suppliers get extra profits at taxpayer expense; if the government underpays, then providers may exit the market and individuals may have limited access to care. 

Similarly, if a private health insurers has no providers in their network in a given area, it will be very difficult for them to sign up members.  Thus, whereas a local insurer with detailed knowledge of market conditions will know about these provider exits, federal government workers may have less of a good sense of whether provider exits are leading to access limitations. 

What about new classes of treatments such as digital therapeutics.  If/how to cover treatments that are “administered” digitally is uncharted territory.  Having many small payers being able to experiment with different coverage approaches is helpful for understanding what did and did not work and understanding what consumers value.  With a single payer system

In summary, in a static, unchanging world, a single payer system works great. In the real world with a lot of innovation, new treatments, changing prices, and new technologies, a decentralized structure is needed.

For those who want a more egalitarian system, a “Medicare Part C” for all where the government subsidizes private insurance is useful.  To prevent adverse selection, the government could make risk adjustment payments as well.  Such as approach would balance the coverage goals with a more dynamic, decentralized marketplace while also insuring that health plans are incentivized to provide high-quality care to even the sickest patients.

2 Comments

  1. This is an excellent discussion point but I think the value of the private market is idealized and not consistent with what really goes on. Varied rates in different markets have more to do with different degrees of leverage, depending on the dominance of providers (including, it seems, venture capitalists), insurers or hospitals, than it does on experimentation. Prices of new therapies have more to do with what is charged and how many patients and providers buy in than on actual value. This is a herding phenomenon, well known to cognitive scientists. Whether rates are set by one group (CMS) or by many (private insurers) makes little difference. What’s important is that the process should be based on evidence, not leverage. Germany’s model for drug pricing seems an excellent example of a rational approach that could be applied more broadly to other realms (negotiated prices depending on worth relative to existing treatments with increases based on new data only). This could be tweaked with the addition of updates based on the CPI, geographically adjusted. Single payer, single data may be a more rational approach than multiple payer, biased data.

Leave a Reply

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