Do Short-Term Laboratory Experiments Validly Describe Long-Term Economic Interactions?

A key problem regarding the external validity of laboratory experiments is their short duration: while economic interactions in the field are often long lasting, lab experiments typically last only for an hour or two. To address this problem, we compare standard laboratory experiments (two Cournot duopoly variants) lasting one hour to treatments which last for nearly a month. In the long time span treatments, subjects play the same Cournot games and make the same number of decisions as in the short time span treatments, but they decide once a day over the internet. For both variants, we find no noticeable differences between treatments.