Published by QJE Aug 2015
A large body of lab research on time inconsistency shows evidence for existence of present bias. However, several articles document that dynamic inconsistency in monetary payment choices disappears when transaction costs and payment reliability (Andreoni and Sprenger, 2012a) and appears again when payment risk is added (Andreoni and Sprenger, 2012b). These gives a risk explanation for present bias found in the monetary payment experiments.
The authors of this paper move out of monetary choices and into consumption behaviors. They design a seven week longitudinal experiment in which subjects allocate and reallocate effort (negative leisure consumption) over time. They find significant evidence for present bias in the domain of effort but no aggregate evidence of present bias in the domain of money. Furthermore, the results show that at price 0 roughly 60% of subjects prefer commitment to flexibility and such commitment demand correlates with present bias.
There are two highlights in this paper. First, real effort setting is proved to be a better way to illustrate existence of present-bias than monetary payment setting. Second, allocation of effort can directly interpreted as allocation of negative leisure consumption, leading the conclusion to a more general scope. However, it should be noted that investigation on other kinds of consumption is still in need to verify the external validity of this paper.
Reviews of lab research on dynamic inconsistency of time preference up to early 2000s: Frederick, Loewenstein and O'Donoghue, 2002
Critics on monetary experiments: Laibson and Schuldt, 2008