Essays on socially responsible operations with a focus on agricultural and health care industries

Abstract

This dissertation studies the socially responsible operations in agricultural and health care industries. In the first essay,we examine whether the government or non-governmental organizations (NGOs) can improve farmers' welfare by offering agricultural advice and market information. Towards this end, we consider a sit-uation where farmers decide whether to use market information to improve their production plans or adopt agricultural advice to improve their operations when they engage in Cournot competition under both uncertain market demand and un-certain process yield.We show that both farmers will use the market information to improve their profits in equilibrium. Hence,relative to the base case in which mar-ket information is not available, the provision of market information can improve the farmers' total welfare (i.e.,total profit for both farmers). Moreover, when the underlying process yield is highly uncertain or when the products are highly het-erogeneous, the provision of market information is welfare-maximizing in the sense that the maximum total welfare of farmers is attained when both farmers utilize market information in equilibrium. Furthermore,in equilibrium,whether a farmer adopts the agricultural advice depends on the size of the requisite upfront invest-ment.More importantly, we show that agricultural advice is not always welfare improving unless the upfront investment is sufficiently low. This result implies that to improve farmers' welfare, governments should consider offering farmer subsidies. In the second essay, we study the performance measurements of health care sys-tems.Many governments use waiting time per admission (i.e.,the waiting time that a patient spends per admission) as a measurement to evaluate the congestion level and the performance of the health care system. Adopting this criterion may force doctors to speed up their service and spend less time on each patient,resulting in a decline in service quality and an increase in readmission rate. To characterize both system congestion and service quality,we propose a new performance measurement,total waiting time,which is the total amount of time a patient spends in the sys-tem before being cured. We then consider the optimal design and control issues of the health care system based on this new measurement. We model the health care system as an M/M/1 queue with Bernoulli feedback,where the service rate is a de-cision variable of the health care provider (HCP) and the readmission rate (i.e.,the feedback) is increasing in the service rate.We study the decision problems for the three parties in the system: patients, the HCP and the social planner (government). We demonstrate that a naive adoption of the waiting time reduction target could even worsen the system performance,leading to a higher congestion level,lower accessibility for new patients and a busier HCP. We find that the social optimality cannot be achieved via the single pricing mechanism.Instead,the social planner needs to adopt a regulation mechanism with multi-dimensional control variables.