Hi, I guess it is a convenient way to study corruption in a "consumer-choice" setting. Bounded rationality could play a role, where people will take bribes maybe not because they really want, but because not taking it would disrupt the going system and maybe cause a punishment by peers. So asking and taking bribes might become the non-optimal, least disruptive shortcut solution. Still, inserting the loss-aversion paradigm, one could argue that punishments could bring this system out of balance towards a more Pareto efficient solution.
These are just some quick thoughts. Hope it helps.