Collaborative Research: Foundations for Comparative Naivete and Sophistication
Duke University, Durham NC
Investigators
Abstract
Consumers frequently make decisions that are inconsistent across time. For instance, they may be tempted to consume more at a given date than they had previously budgeted, leaving less for future consumption. Awareness of time-inconsistency problems allows consumers to take advantage of commitment devices to constrain their future decisions: A consumer who anticipates being tempted to under-save may channel some assets into illiquid savings instruments such as tax-sheltered retirement accounts to commit to saving that money until a later date. However, a growing body of evidence suggests that many people hold inaccurate beliefs about their future self-control and naively anticipate making more virtuous choices. Naivete about time-inconsistency may undermine the effectiveness of certain policy instruments like retirement accounts, since naive consumers will underutilize these commitment devices. The goal of this research project is to study when consumers make choices that are inconsistent across time and the degree to which they are able to predict their future deviations from current plans This project will develop a new measure of the level of naivete or sophistication consumers have about their own future choices and will provide new behavioral definitions to test this. The project will also develop a recursive and fully dynamic model that allows for using dynamic programming techniques bringing naivete to closer contact with workhorse models in finance and macroeconomics. The importance of this foundational theory is to understand naivete in a language broad enough to develop and understand the implications of interventions in cases where citizens have heterogeneous degrees of naivete and are best approximated by a variety of different formal models. This project has broader impact for a wide variety of public policy decisions ranging from health management to retirement savings.
View original record on NSF Award Search →