Page 2

July 2001

Cognition and Wealth: The Importance of Probablistic Thinking

by Lee A. Lillard and Robert J. Willis

Executive Summary

In recent years there has been an increase in private and company pension plans that allow workers to make contributions to their own retirement accounts and to make their own decisions about how much they contribute, into what type of account they contribute, and about when and how much is paid out to them. It has been argued that at least some portion of Social Security funds should be managed as voluntary private accounts of this sort. Whether this policy move would be harmful or beneficial depends in part on how well individuals and households would manage these accounts. Critics are concerned that many people, especially those with lower income to begin with, will not make investment decisions that will give them the greatest return on their money. This, they argue, will widen the already large inequalities in wealth among older Americans. Research that addresses this question is needed for informed retirement policy.

In this Issue in Brief, we summarize work that investigates one aspect of financial decision-making: the ability to make precise judgments, or guesses, about the likely occurrence of future events, or as we refer to it, probabilistic thinking. We use data from the Health and Retirement Study (HRS) to create a measure of this kind of thinking and then relate it to household choices about the riskiness of investments, and hence the rate of growth in the value of those assets. In short, we find that a large proportion of older Americans tend to make imprecise probability judgments and that this is related to holding a financial portfolio containing less risky assets with lower rates of return. This is important information for policymakers to consider when weighing the costs and benefits of individual retirement accounts in the Social Security system.

The Data

The Health and Retirement Study (HRS) is a longitudinal, nationally representative study of older Americans. The survey began in 1992 with an initial cohort of 12,652 individuals from 7,607 households, with at least one household member born from 1931 to 1941. In this study, we use questions asked in the 1998 wave that surveyed over 22,000 Americans over the age of 50. We construct an index of probabilistic thinking using questions that ask the respondent about topics ranging from personal life expectancy and date of retirement to beliefs about the rate of inflation and future Social Security policy. For each question, the respondent is asked how likely they think various events might be. The respondent is told to give a number between 0 and 100 where “0” means “no chance at all” and “100” means the event is absolutely sure to happen. The HRS also contains extensive information about respondents’ financial assets, which we use in our analyses.

The Logic of our Argument

In general, the probability questions give good information. For example, answers to questions about life expectancy match life table probabilities surprisingly well. However, there are also a large proportion of respondents who typically answer ‘0’ or ‘50’ or ‘100.’ We call these focal answers and assume that they reflect a large degree of uncertainty about the true probability of the occurrence of an event. By contrast

[continued on page 4]

Page 1 2 3 4 5 6