Fishery managers face many challenges when setting effective policies. This includes working with fisherman to set the total allowable catch (TAC), preventing overfishing, and monitoring the status of a fishing industry based on imperfect data. This dissertation focuses on the last two issues. In particular we focus on how measurement error and estimation issues can impact fishery policy using two common economic models, the stochastic frontier model and the Schaefer production model. Both of these models use production data on inputs and outputs to estimate a production function. Using panel data from the Mid-Atlantic surfclam fishery from 2001-2009, Chapters 2 and 3 examine measurement error correction methods that policy makers can use when instrumental variables are not available. We examine this particular fishery because both logbook data from vessels, and scientific estimates of the biomass are available. In Chapters 2 and 3 we make use of estimates of the measurement error variance, and apply a simulation based correction method known as simulation extrapolation (SIMEX). SIMEX is a simulation based method for reducing bias in parameter estimates caused by measurement error. After estimating both models under a naive analysis, SIMEX is used to obtain less biased estimates of technical efficiency, production and biological parameters. In the last chapter we revisit the stochastic frontier model, estimating both a maximum likelihood and Bayesian model. We conclude the dissertation by discussing the results from each chapter, and their implications for fishery policy.