Substack

Saturday, September 14, 2013

Principles of a universal health insurance model

Health care reforms have been fiercely debated across the world for more than three decades. Unfortunately there has been limited consensus on even fundamental issues. I find this a bit surprising since there are certain basic and self-evident features of health care market, which have been widely researched, that distinguishes it from other markets. Here are atleast two very important features that distinguish health care from other services.

1. The provision of health care is different from any other service or good traded in the market. Even though a very rich body of established research has shown that health care market suffers acutely from problems of information asymmetry and inherent limits to competition (patients cannot shop either because of the urgency of their medical condition or because of limited choice in their geographical or insurance area). Health care is not broccoli.

2. An unregulated health insurance systems suffers from multiple sources of incentive distortions and resultant market failures. The disconnect between service delivery and price signal at all levels - doctors dispensing care, patients demanding care, providers selling diagnostic and procedures, employers paying premiums, or insurers paying providers - creates a web of incentives that works against keeping costs under control.

In light of this, a universal health insurance model would have to adhere to atleast the following five principles.

1. The widest possible risk pool is necessary to keep costs down. Healthy individuals would be averse to paying the high cost of insurance, though they would then impose negative externalities by seeking subsidized care (either as part of a public health care system or even in a private insurance scheme) when they get old. Mandatory enrollment is therefore critical to the success of any universal health insurance model.

2. The cost of insurance cannot be based on pre-existing conditions. If this were the case, insurers would cherry-pick and premiums would be highest for those most in need of health services, thereby negating the very objective of health insurance. Furthermore, it would force insurers into spending large amounts in screening and other administrative services, thereby reducing the efficiency of aggregate health care spending. Community rating - whereby insurers provide the same policy at same price to all people of an age group within an area irrespective of their pre-existing medical condition - has to be an essential requirement for any universal health insurance model.

3. Health care services have to be rationed. As people live longer and technology advances, the cost of health care will be raised by both supply and demand side factors. Given the fiscal constraints of governments, it may not be possible to deliver the best available care as part of a universal health insurance model. There would have to be rationing in atleast two dimensions. One, any universal health insurance model can be sustainable only if it restricts coverage to a basic set of services, with the boundaries of coverage being determined by the degree of fiscal space available for governments.

Two, as the choice of diagnostic and treatment options expand, accompanied with exponentially rising costs, we would also need to value human life and determine thresholds for certain critical and terminal care options. A robust system of determining cost-effectiveness of diagnostic and treatment options is essential to keep a lid on rising costs. The British NHS has the National Institute for Health and Care Excellence (NICE), which has the authority to adjudicate on the cost-effectiveness of new diagnostic and treatment options.

4. The pricing of health care services - diagnostic and treatment services - by care providers cannot be purely market determined. The most effective price discovery mechanism would be for suppliers, health providers, and consumers, insurers, to engage in negotiations at an appropriately large geographical level, and arrive at a clearly defined fee schedule for all services. In its absence, a government intermediated process of price fixing may have to do the role of price discovery.

5. Finally, a universal health insurance system, however well deigned, will not be affordable to the poorest without some degree of premium support. The extent of premium support would depend on the country's poverty levels.

Update 1 (9/23/2013)

Five percent of patients account for 50% of US health care spending, with the top 1% accounting for 20.2% of the spending. Most of this is concentrated in terminal care and mental health type conditions.

National Institute for Health Care Management

Update 2 (10/6/2014)
Excellent primer on health care spending in the US from Vox.

No comments: