From the abstract:
The contention that consumers systematically “undersave” for retirement is a frequent example provided by adherents to behavioral economics and behavioral law and economics to purportedly illustrate their theories. Although frequently asserted, the claim that people systematically undersave is rarely assessed empirically.
This article, written for the Georgetown Institute for the Study of Markets and Ethics Symposium on “The Ethics of Nudging,” examines available data on how many people fail to save and the reasons why they do not. According to available evidence, the overwhelming number of households saves enough or more than they need for retirement; only a small minority does not seem to save enough. Those who do not save for retirement lack the money to do so or allocate available resources to paying down consumer and student loan debt. Behavioral economics theories explain little of the observed patterns of saving or non-saving behavior. Moreover, behavioral economics itself suggests that many people probably oversave for retirement and makes no effort to reconcile these offsetting biases.
More fundamental, once it is recognized that there is an opportunity cost to saving more — one must consume less today, borrow more, or work more — the theoretical validity of the claim that people undersave because of behavioral biases is suspect. Given the inherently subjective nature of opportunity cost, a central planner cannot be confident that he can make people better off by influencing their consumption expenditures across time than he could by shifting consumption expenditures across different goods and services today. It is concluded that there is little reason to believe that people would be made better off by nudging them to save more for retirement.