Three Essays in Health Economics

Abstract

<p><b>This dissertation contains three essays on economic analysis of the New Zealand healthcare market particularly relating to publicly funded elective surgeries and the pharmaceutical industry.</b></p> <p>In Chapter 1, using administrative data on all patients booked for publicly funded elective surgery in 2014, we investigate whether ethnic disparities exist in waiting time for elective surgeries in New Zealand. Unlike existing studies on the topic, we extensively control for sample selection bias, clinical severity, and regional factors affecting resource supply and healthcare utilization. We find evidence of Maori, Pacific, and Asian patients waiting longer for elective surgery when com-pared to New Zealand Europeans. Our results indicate that Maori wait on average 3 percent longer for elective surgery, however, they are not consistent across all model specifications. We find that Pacific and Asian patients wait 8 - 9 percent longer for elective surgeries compared to NZE. Our results are considerably stronger for Pacific and Asian compared to Maori across all model specifications. Furthermore, these differences are pervasive across several surgical categories and are as high as 22 percent across angiography procedures. We also find evidence that waiting time in New Zealand is geographically inequitable. Although an increasing number of empirical studies find evidence of inequities in waiting time socioeconomic status, our analysis reveals no such effects once regional differences affecting waiting time are controlled for.</p> <p>In Chapter 2, using administrative data on patients who have been booked and treated for publicly funded elective surgeries over 2011-2015, we investigate how supply of and demand for elective surgeries respond to waiting times in New Zealand. Unlike existing studies, we endogenize waiting times in a system of three equations, which nests the conventional demand-and-supply model. Each structural equation is over-identified by excluded instruments, and the exclusion restriction is justified on a priori grounds while over-identifying restrictions are tested. The analysis finds that, in the case of New Zealand, the demand for elective surgery is inelastic, whereas supply is elastic. From a policy perspective, the results suggest that a long-term increase in supply will lead to a permanent reduction in waiting times and thus improving timely access to care.</p> <p>In Chapter 3, I develop an intuitive model based on the traditional newsvendor framework that enables pharmaceutical wholesalers to efficiently forecast demand and reduce inventory while adjusting for required customer service level (CSL) targets. First, using historical demand data for a major pharmaceutical wholesaler in New Zealand, I present several demand forecasting models and assess their performance in predicting demand, based on several forecast error metrics. I find that artificial neural network (ANN) models are generally stable for demand forecasting and sometimes outperform traditional demand forecasting methods. Second, using the demand fore-cast as an input, I derive a periodic review inventory model based on the newsvendor inventory framework. Numerical analysis of the proposed model shows that, compared to current practice, the model enables wholesalers to significantly reduce both inventory levels and ordering, while maintaining a very high non-stock-out probability constraint and CSL target close to 100 percent. I conduct separate analyses for both government subsidized low value-high demand drug and non-subsidized low demand-high value drug and find the results to be consistent across both types of drugs.</p>