Identifier

etd-04242012-101116

Degree

Doctor of Philosophy (PhD)

Department

Agricultural Economics

Document Type

Dissertation

Abstract

Regional economists and policy makers are interested in forecasting economic changes that are likely to take place at local and state levels after exogenous shocks to an economy; that is, create disequilibrium conditions in terms of supply and demand. Impacts of such shocks could be observed at the level of employment, unemployment, commuting patterns, assessed property values, property and sales taxes and local level of expenditures in several categories. The objective of the first essay was to model the employment change decompositions of different effects in two major industries using a shift share analysis technique in context of Louisiana parishes before and after hurricanes Katrina and Rita. A correlation analysis test was performed to identify whether a distinct regional industry effect can be identified separately from a sub-region local effect in shift-share analysis. Results from the test indicated that the distinctiveness of spatial neighboring region effect and the localized effect was evaluated and they were two separate effects. The objective of my second essay was to model the Louisiana labor market for purposes of improving forecasting accuracy in regional economic modeling. Specifically, this was performed through the use of alternative regional econometric estimators in Community Policy Analysis System (COMPAS) models for Louisiana. Results suggested that panel data models increased forecasting performance compared to other models in the study, if measured in terms of traditional error measures. However, the mean comparison test suggested that panel models do not always display statistical improvement in forecasting. The third and final objective of my dissertation was to evaluate if a fiscal module under the COMPAS framework (an equilibrium model) fits better under a disequilibrium economic environment. I found that both a simple naïve model with one year lagged expenditure as well as a lagged expenditure model with revenue capacity variables significantly increased forecasting performance relative to the traditional supply/demand equilibrium model of the public sector. I also found weak evidence suggesting that in cases where the equilibrium model is used in a cross-sectional setting, quantile regression may improve forecasting performance given the attribute of lumpy public goods.

Date

2012

Document Availability at the Time of Submission

Student has submitted appropriate documentation to restrict access to LSU for 365 days after which the document will be released for worldwide access.

Committee Chair

Fannin, James Matthew

Share

COinS