Explaining framing effects is one of the main challenges faced by decision theories. This research experimentally examines how different contextual frames influence competitive behaviour in a Bertrand duopoly
game (repeatedly played under a stranger matching), unexplored so far. The design comprises four frames:
one abstract (a beauty-contest framing), two meaningful (the standard Bertrand framing and an access-to-river
framing) and one evocative (a take-from-fund framing). Our findings show that, at first, the evocative frame
differentiates from the rest mostly in market prices. While the evocative frame induces subjects to behave
closer to the theoretical predictions initially, the others need some repetitions until convergence is achieved.
Differences across frames eventually vanish at the end. During the transition, in the Bertrand frame, a quicker
decay in prices is observed due to the behavioural reactions to historical market prices. Lastly, irrespectively
of frames, behavioural reactions to immediate past information allow to explain strategic interaction in the
long-run: a force-balance situation which is consistent with the related literature on price floors in Bertrand
games.
