Species-specific time series analysis of the impact of Alaska policy decisions on salmon harvest yields

Abstract

Throughout Alaska's history, the volume of the yearly salmon harvest has been an issue of much debate. As early as the 1940s and 1950s, decreasing salmon harvests caused territorial leaders to push for Alaska statehood such that the resources could be governed by the people that lived there. Since then, various policy shifts in both 1959 and 1.974 have been credited with supporting higher and higher salmon yields. This thesis incorporates historical data from 1914-2013 for salmon harvests of Chinook, sockeye, coho, pink, and chum salmon along with environmental factors associated with the Pacific decadal oscillation (PDO) index to investigate the role that policy decisions in Alaska's history have played in the salmon industry. This cointegration relationship is studied using an autoregressive distributed lag (ARDL) bounds testing approach to create statistical models of the time series data for each of the five salmon species. Results for both short-run and long-run impacts are analyzed for each species as well as the salmon market as a whole. The conclusions show that while there is a long-run cointegration relationship between oceanic conditions and Alaska salmon harvest, the policy changes in 1959 had no statistically significant impact on long-run salmon yield. In addition, the changes in 1974 including the Limited Entry Act and bolstered support for salmon hatcheries did have a strongly significant affect on the volume of salmon harvests. The ARDL models presented here offer a new look at historical Alaska policy while taking into account the interconnectedness of the market with the environment.