Discrete Choice Modeling of Landings Locations

Interest in sustainable fishing communities suggests the need to understand fishermen’s decisions about where to land fish. In this paper, we apply techniques used extensively to analyze fishing location choices to study landings location choices. We analyze detailed microdata from Norwegian groundfish vessels that land fish in fourteen different ports in Norway’s northernmost region of Finmark. We find that insights from the fishing location choice literature are relevant but do not translate perfectly to the landing location. As in fishing location choice, vessels are responsive to travel distance. However, fish stocks in the water are far more dynamic than changes in market opportunities on land. As a result, we find little evidence that measures of expected revenues drive behavior. The discrete choice models appear to parallel the fishing location choice models when run naively with simple backward looking averages of port-level revenues. But when state dependence (the tendency to return to the same location) is accounted for, the significance on expected revenue evaporates. A high percentage of vessels in our sample always land in the same port, and many of these always deliver to the same processor. We are currently in the process of splitting the sample and treating port switchers as arbitrageurs. This analysis will allow us to see if port switchers are responsive to changing economic opportunities over space