Optimal Expectations and Limited Medical Testing

Untested individuals report perceived probabilities [of contracting Huntington's disease] that are much lower than their objective probability.

Many people who may be carrying a life-threatening genetic disease evidently prefer not to know about it. In Optimal Expectations and Limited Medical Testing: Evidence from Huntington Disease (NBER Working Paper No. 17629), co-authors Emily Oster, Ira Shoulson, and Ray Dorsey find that even when testing is relatively inexpensive and highly accurate, large numbers of persons at risk behave as though they believe that the burden of living with the anticipation of ill-health should be avoided. This, the researchers say, has an important impact on life decisions of patients as well as significant implications for policymakers.

Oster, Shoulson, and Dorsey study 1,001 at-risk individuals who have chosen not to undergo genetic testing for Huntington's Disease (HD), a hereditary neurological disorder that reduces life expectancy to about age 60. DNA testing for the disease is 100 percent accurate and carries little financial or time cost, but fewer than 10 percent of at-risk individuals are tested prior to the onset of symptoms. The authors also analyze decisions made by some research participants who chose to undergo testing for HD (usually after symptoms of the disease began to appear). They link information on the probability that these tested individuals are likely to contract HD with data on their life decisions, including childbearing, marriage, retirement, education, and participation in clinical research. The authors find that untested individuals express optimistic beliefs about their likelihood of having HD and that they make their life decisions as if they do not have HD. Individuals with confirmed HD behave quite differently.

When asked about their chance of having a genetic predisposition to HD, untested individuals report perceived probabilities that are much lower than their objective probability. Having adopted such overly optimistic beliefs, these individuals take overly optimistic actions. Oster, Shoulson and Dorsey observe that persons avoiding testing are neither making mistakes nor lacking information: they simply prefer to "consume happiness" in the anticipation period. But as the objective risk increases and people continue to behave as if they are not likely to contract the disease, the utility loss from this optimistic behavior becomes larger and larger, increasing the incentive to test. Depending on the dataset analyzed, as many as half or more of individuals will eventually be tested to "prove" what they know already from symptoms. A significant "cost" of testing is the loss of the option to believe that one is healthy, even if one has HD. This cost may be large enough to make the value of testing negative for some individuals, even ignoring any more traditional costs of the test, such as the time involved in testing or the medical costs associated with the test.

Although their study focuses on HD, the researchers demonstrate that similar patterns of low testing exist in cancer screening and HIV testing. They suggest that for a disease like HIV, higher testing rates might lead to better treatment and to less spread of the disease. Even with a non-contagious disease, overly optimistic individuals may under-save for ill health, thereby placing the burden of care on the government if and when the disease strikes.

--Matt Nesvisky

The Digest is not copyrighted and may be reproduced freely with appropriate attribution of source.
NBER Videos

National Bureau of Economic Research, 1050 Massachusetts Ave., Cambridge, MA 02138; 617-868-3900; email:

Contact Us