Debatable Assumptions

July 25, 2007 • Commentary
This article appeared on TCS​dai​ly​.com on July 25, 2007.

“there will exist other better, yet more utopian, proposals…when we reject those ‘better proposals,’ we end up arguing that they are excessively utopian. Why are these reforms too utopian while ours are not?”
–Tyler Cowen

In public policy debates, it is common to claim that the other side has different goals. “They just want to screw the poor.” “My opponent is a Communist.” And so on.

However, it is possible to agree on goals but disagree on assumptions about what will achieve those goals. In fact, I think that it is reasonable to assume that political opponents disagree primarily in their assumptions rather than in their goals. In Tyler Cowen’s terminology, our disagreements are about what is feasible rather than about what is desirable.

If my assumption is correct, then political debate would be more enlightening if we asked proponents to focus on their assumptions. When you say that your policy proposal would be an improvement, what are the key assumptions behind that statement?

Scoring the Assumptions

For example, suppose that someone proposes that health insurance should be made mandatory. Below are a set of assumptions that might be consistent with such a policy. In parentheses, I give a score that indicates my degree of belief in the assumption. The scale goes from 1 to 5, where 5 means that I strongly accept the assumption, and 1 means that I am highly skeptical of the assumption.

  • People who do not buy health insurance will nonetheless obtain medical treatment. (5)
  • When the uninsured are treated, the cost of their care necessarily falls on someone else. (2)
  • People who are uninsured tend to fail to obtain preventive care. (5)
  • The uninsured fail to obtain preventive care because they are uninsured. (2)
  • This failure to obtain preventive care substantially raises costs in the long run. (2)
  • The government can design a mandatory health insurance package that balances the needs of consumers, health care providers, and the providers of the insurance (which could be either private insurance companies or taxpayers or some combination). (2)

I do not believe that the cost of treating the uninsured necessarily falls on someone else. As one of my correspondents frequently points out, the taxpayers could provide funding for the treatment in the form of a loan rather than a gift, and the government could be quite aggressive about collecting any such outstanding loans.

One can argue that in practice we do fund the treatment of the uninsured through cost‐​shifting. It is only fair to compare mandatory health insurance with this reality, not to a hypothetical alternative. Point taken.

Many people are more convinced than I am that lack of health insurance causes major health problems, due to failure to obtain preventive care. There is an urban legend that 20,000 people die each year for lack of health insurance. The Institute of Medicine may be the original source for this claim, in which case it is based on superficial correlations, not on controlled studies that would be required in order to make a claim about, say, a new pharmaceutical. In careful comparisons of similar groups of people, it is rare to find significant differences in longevity based on levels of health care spending, which makes it unlikely that one can demonstrate a causal relationship between health insurance and longevity.

The famous RAND experiment found that decreasing insurance coverage reduced the use of preventive care, but without a significant overall adverse effect on health outcomes. Another analysis, by Amy Finkelstein, showed that Medicare increased health care utilization substantially, again with little noticeable impact on health outcomes.

I can believe that there is a significant group among the uninsured who take poorer care of their health than the rest of us. But I am not convinced that forcing those people to buy health insurance will make them take better care of themselves. It might even do the opposite.

My main area of skepticism about mandatory health insurance is the ability of government to design a health insurance package. In Massachusetts, for example, provider lobbies have a long history of dictating coverages that most consumers do not want. This is adversely affecting the Massachusetts health plan, where the premiums are likely to be double what was promised, if not higher.

The Assumptions Behind My Proposals

We can also examine the assumptions behind my own proposals on health care. For example, I have frequently recommended raising the age of eligibility for Medicare for people currently aged 50 and younger. The age might be 72, with gradual increases for younger workers as longevity continues to rise. I would state the assumptions behind this approach as follows.

  • Within the next two decades, we will have to make major cuts in Medicare benefits in order to make the system sustainable. (5)
  • Raising the age of eligibility now gives individuals the opportunity to make plans to provide saving and insurance to cover the reduction in Medicare coverage. (5)
  • Many individuals actually would increase their use savings and purchases of private insurance rather than face devastating financial setbacks due to medical expenses between age 65 and 72. (3)

This exercise points out that the weakest assumption in my approach is that individuals would adapt to a higher age of eligibility and make appropriate long‐​term plans. Of course, any alternative approach to Medicare is also going to rest on some weak assumptions, such as an assumption that “cost control” will be effective.

The Corporate Setting

Inside corporations, government agencies, and other bureaucracies, acrimonious debates also break out. In these settings, I also would recommend assuming that everyone shares the same goals and framing the argument in terms of assumptions. When someone argues for a new marketing initiative or a particular systems development project, everyone should try to articulate the chain of assumptions that justify the proposal. Then have people score the assumptions. Scoring the assumptions is a way of taking the emotion out of disagreements. Divergence in scores will help to highlight the risks and uncertainties that surround the decision, so that perhaps people can find ways to test the critical assumptions before proceeding further.

The next time you hear someone make a proposal that you find outrageous, try getting at the assumptions behind it. See if you can identify which assumptions are 4’s and 5’s in terms of the advocate’s confidence but 1’s and 2’s in terms of your own confidence. My guess is that this approach will lead to more constructive discussions and less personal antagonism.

Sometimes, assumptions simply are not compatible. If one person assumes that life begins at conception and someone else does not, there is not much room to resolve the issue through compromise or further study. Also, there may be situations where people really do disagree on goals, so that going over assumptions is futile. However, my bet would be that much more often differences in assumptions are at the heart of disagreement.

About the Author
Arnold Kling

Senior Affiliated Scholar, Mercatus Center at George Mason University