Three Essays on Pricing in Socially-Optimal Markets for Differentiated Goods
Open Access
- Author:
- Ghosh, Gaurav Somenath
- Graduate Program:
- Agricultural Economics
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- July 15, 2009
- Committee Members:
- James Samuel Shortle, Dissertation Advisor/Co-Advisor
James Samuel Shortle, Committee Chair/Co-Chair
Anthony Mark Kwasnica, Committee Member
Edward C Jaenicke, Committee Member
Richard C Ready, Committee Member - Keywords:
- Spatial Econometrics
Emissions Trading
Experimental Economics
Pollution Control
Hedonics - Abstract:
- In my first essay I report results from an analysis where representative methods of estimation from the Classical and Bayesian approaches to statistical inference are empirically compared. The chosen Classical methods are based on Least Squares and Maximum Likelihood and the chosen Bayesian method is the Hierarchical model. Each method is applied to a spatial hedonic property value model. The resulting estimates are then compared using nonparametric tests. The comparisons are then used to make inferences on the relative accuracy, precision and quality of the different methods of estimation. The Hierarchical Bayesian and Classical Maximum Likelihood methods are found to supply the best estimators of the spatial regression model and predictors of house price. Significant differences are also found in the relative accuracy and precision of the methods. I infer that the Hierarchical Bayesian and Classical Maximum Likelihood methods are best suited to prediction and estimation of spatial hedonic property value models. In my second essay I compare two tradable permit markets in their ability to meet a stated environmental target at least cost when some polluters have stochastic and non-measurable emissions. The environmental target is of the safety-first type, which requires probabilistic control of emissions. One market is built around the trading ratio, which defines the substitution rate between stochastic and deterministic pollution, and is modeled on existing markets for water quality trading. The other market is built around a new definition of the traded commodity as a multi-attribute good, where the attributes supply information to the market on the environmental risks associated with stochastic pollution. The latter market is found to out-perform the trading ratio market in its ability to satisfy the safety-first environmental target at least cost. This result comes about because polluters are able to directly price risk in the latter market. In the former market risk is not a factor in the trading decision and can only be controlled under highly restrictive conditions. In my third essay I report results from an economic experiment where the two markets developed in the previous essay are compared in a testbed that captures important features of existing markets for water quality trading. In the interests of tractability these features were abstracted from in the previous theoretical analysis. One feature is that of oligopsony and the second feature is that of a discrete trading environment where polluters generate credits by implementing one of a small set of emission-reducing technologies. The experimental results indicate that the market with multi-attribute goods generates a superior environmental outcome to the trading ratio market. Furthermore, the average cost of pollution control is lower in the former market. Market power is independent of the type of market institution, but I do find that large buyers have more market power than small buyers. Finally, I find that sellers of credits learn to resist market power as they gain experience, but at the cost of market efficiency since their resistance causes a fall in the number of trades. Overall, the results support the thesis of the second essay, that the market with multi-attribute goods generates better environmental outcomes than the trading ratio market.