Three essays on applied economics

In this study, three areas of concern in applied economics are examined. In the first essay, growth indicator data from the Penn world data bank was analyzed using a relatively new technique, Synthetic Control Method, to assess public choice concerns. Malaysia's economic performance was measured under Dr. Mahathir bin Mohamad. Dr. Mahathir bin Mohamad is often credited with Malaysia’s economic success post-1980. However, it is well known that the Mahathir regime centralized power in the Office of the Prime Minister and extended state capacity, creating a system of government susceptible to corruption. This corruption eventually made global headlines in 2015 with the 1MDB $5 billion dollar scandal. Thus, while Malaysia did experience growth that coincided with the reign of Dr. Mahathir, it is likely that his regime limited this positive change. We evaluate the economic impact of the Mahathir regime using the Synthetic Control Method. We find that while Malaysia did experience growth throughout Dr. Mahathir’s leadership, GDP per-capita was far below (approximately $4,000 per-capita below 9-years following the treatment) its potential as measured by Synthetic Malaysia. Thus, this study provides evidence of a negative economic effect from power centralization and enhanced state capacity. Several robustness tests confirmed this conclusion. The second essay examined the profitability of a small scalable, multiproduct biorefinery that uses onsite biomass delivered with cotton bales at a gin. A Bayesian normal regression model was used to simulate 10,000 data points based on 15 years of biomass collected from a gin in Lubbock County and 15 years of precipitation data. We evaluated the use of cotton gin trash to produce electricity, much of which can be sold at peak prices to local farmers for irrigation and, by electrolysis, ammonia fertilizer. This is done by evaluating the profit maximization models for different possible scenarios using linear optimization model. This adaptability allows the biorefinery to be profitable in both severe drought and high yield years for multiple combinations of ammonia and electric power production. Simulating weather and price variability simultaneously, we present an EV Frontier that maps profitability to variance for different combinations of electricity and ammonia. We scaled the gin to allow for economies of scale and evaluated accordingly. The returns on investments were presented for models on the frontier. Additional tests were performed on biomass sensitivity and price sensitivity. Most of the models on the EV frontier were resilient to changes in prices and quantity of biomass. Third and final essay compared different methods of creating a housing submarket previously proposed. We defined submarkets as a function of both geographic and structural factors. To provide a precise basis for comparison, all 4 models were considered for the same city (Atlanta) and the same period (2015-2016). Housing data from the Tax Assessor's office, Census data, and publicly available statewide middle school average test scores were used. Based on the reduction in prediction error in housing prices, the best model was selected. All four models showed significant reductions in prediction errors compared to the market. This confirms the existence of submarkets in Atlanta. The simplest method outperformed the others. Using this method, the study area was grouped based on a priori demarcation (parcel districts), and then houses in the parcel districts were grouped based on structural types to create eight submarkets. The second-best model is the hierarchical linear model, which creates seven submarkets from adjacent school zones. The PCA and K means model with two submarkets had the worst performance.