The cost of keeping track

Johannes Haushofer

SECTOR

Inclusive Finance

PROJECT TYPE

Empirical tests

DOI

Location

Kenya

BEHAVIORAL THEME

Intertemporal choice | Temporal discounting | Time preference | Memory | Limited attention | Technology adoption
OVERVIEW

I show that a lump-sum cost of keeping track of future transactions predicts several known departures from the standard discounting model: decreasing impatience with dynamic inconsistency; a magnitude effect; a reversal of this magnitude effect in the loss domain; a sign effect; and an Andreoni-Sprenger (2012) type reduction of discounting and decreasing impatience when money is added to existing payoffs. Agents of this type “pre-crastinate” on losses and are willing to pay for reminders. These results speak to failures of technology adoption in developing countries, and empirical tests conducted in Nairobi, Kenya confirm the model’s predictions.

THEMATIC AREAS

This paper has argues that a number of features of empirically observed discounting behavior can be explained with a lump-sum cost of keeping track of future transactions. Such a cost will cause agents to discount gains more and losses less than they otherwise would; as a result, they will exhibit a sign effect in discounting, a gain-loss asymmetry in valuing future outcomes similar to that observed in loss aversion, and pre-crastination in the loss domain, i.e. a preference for incurring losses sooner rather than later. If the cost of keeping track is lump-sum, it also creates a magnitude effect in the gains domain, i.e. discounting large gains less than small gains; and a reversed magnitude in the loss domain, i.e. discounting large losses more than small losses.