We provide a framework for evaluating potential effects of introducing tradable quotas to a sector. The effects depend on the economies of scale and scope of the production technology, and on firms' ability and willingness to learn best practice methods (catching up) and to change their input and output composition (mix). To illustrate our approach, data from the Danish fishery are used to calculate the potential gains from introducing individually transferable fishing quotas. Data envelopment analysis is used to model the production technology. We find that pure reallocation is as important as pure learning, i.e. quota reallocation without catching-up is as valuable as learning best practice with fixed allocations.