Assessment of Enforcement Mechanism for Pork Traceability System in Korea Using Mixed Strategy Nash Equilibrium

Abstract

The National Agricultural Products Quality Management Service (NAQS) of Korea initiated enforcing the food traceability system since 1994. However, the effectiveness of NAQS’ enforcement has been in dispute. To provide a framework for valid assessment of NAQS’ enforcement mechanism, the Mixed Strategy Nash Equilibrium, is used. Also, to cover the long range of seller’s and NAQS’ strategies, a distribution-based model is developed, along with a mean-based model. The analysis is done for pork labeling because pork is the sensitive food item for traceability and labeling. Also, the analysis is limited to year 2009 due to data availability. The results from the mean-based model suggest that NAQS needed to increase the raid probability for pork significantly from the current 0.1% to 16% in 2009. The cheating probability for pork labeling is calculated as 1.5% in 2009. However, it was found that the data on fines and revenues by cheating are highly skewed. This made implications from the mean-based model unreliable. Through the sensitivity analysis, it was found that the court needs to increase the amount of fine and the most frequently observed Mixed Strategy Nash Equilibrium would be for the seller to cheat with probability of 1.3% and for NAQS to raid with probability of 0.1%, which was quite close to the actual raid probability of NAQS in 2009.

The National Agricultural Products Quality Management Service (NAQS) of Korea initiated enforcing the food traceability system since 1994. However, the effectiveness of NAQS’ enforcement has been in dispute. To provide a framework for valid assessment of NAQS’ enforcement mechanism, the Mixed Strategy Nash Equilibrium, is used. Also, to cover the long range of seller’s and NAQS’ strategies, a distribution-based model is developed, along with a mean-based model. The analysis is done for pork labeling because pork is the sensitive food item for traceability and labeling. Also, the analysis is limited to year 2009 due to data availability. The results from the mean-based model suggest that NAQS needed to increase the raid probability for pork significantly from the current 0.1% to 16% in 2009. The cheating probability for pork labeling is calculated as 1.5% in 2009. However, it was found that the data on fines and revenues by cheating are highly skewed. This made implications from the mean-based model unreliable. Through the sensitivity analysis, it was found that the court needs to increase the amount of fine and the most frequently observed Mixed Strategy Nash Equilibrium would be for the seller to cheat with probability of 1.3% and for NAQS to raid with probability of 0.1%, which was quite close to the actual raid probability of NAQS in 2009.