When the government is experimenting: Field study on households’ behavior

When households make financial decisions, one may expect them to be both informative and rational—just as any other financial agent. In this paper, we question whether this hypothesis is error-free and correct by using a unique large-scale event study. Our results do not support the axioms of financial theory. However, our results are harmonious with the experimental literature, which, to be sure, suggests that agents use information inefficiently; they ignore it or misuse to the effect of bringing about suboptimal or even bad financial results. We use a unique event—an entirely new valuation mechanism that was "enforced" on the pension system and on the households—in order to study the behavior of individual investors. We find that under actual conditions—more similar to "real life" than any deliberated and designed experiment can give rise to—and while information and time are abundant, investors do not behave as if they are informed and rational. They behave in this uninformed and irrational way regardless of the availability of information and the clear-cut possibility of utility-maximizing decision-making. Only when agents are provided with direct information and when the detailed consequence is analyzed for them, do households make a rational decision. We suggest that the gap between experimental conditions and actual situations may be the reason for the indecisive results appearing in the current experimental literature.