DEA와 토빗모형을 이용한 홍합 양식업의 생산성 분석

Abstract

This paper examines the efficiency and the influence factors on the efficiency of 31 mussel aquaculture households in Korea using DEA , Super-DEA, OLS regression models and Tobit analysis. Input variables of DEA and Super-DEA are labor costs, production management coast, a number of mussel seedlings, output variables are net profit and total sales. Independent variables of Tobit analysis and OLS regression models are age, facility area, cost-ratio, ship management cost. The estimation result of the DEA models is summarized as follows. Firstly, the average of mussel aquaculture household's technical efficiency score is 0.706 by applying CCR-I model. Secondly, the average of mussel aquaculture household's pure technical efficiency score is 0.762 and scale efficiency score is 0.922 by applying BCC-I model. It is proved that DMU 4, DMU 7, DMU 14, DMU 21, DMU 23, DMU 24, DMU 25 and DMU 27 produce mussel efficiently in CCR model. The result of Super-DEA shows that DMU 21 is the most efficient households in CCR model, DMU 14 is the most efficient households in BCC model. At the second step, the results of influence factors analysis show that age, cost-ratio, facility area and ship-management cost variables appear statistically significant at least twice in those tree efficiency measures. Therefore, It turns out that all independent variables under the study are the influence factors having an effect on each efficiency of the mussel aquaculture households.