Type of Document Dissertation Author Soto, Patricia Author's Email Address email@example.com URN etd-04132004-144402 Title Spatial Econometric Analysis of Louisiana Rural Real Estate Values Degree Doctor of Philosophy (Ph.D.) Department Agricultural Economics & Agribusiness Advisory Committee
Advisor Name Title Lonnie R. Vandeveer Committee Chair Alvin R. Schupp Committee Member Faik A. Koray Committee Member Michael E. Salassi Committee Member Richrad F. Kazmierczak, Jr. Committee Member Sun Joseph Chang Dean's Representative Keywords
- hedonic analysis
- rural real estate market
- spatial econometrics
Date of Defense 2004-03-19 Availability unrestricted AbstractThe general purpose of this study was to conduct a spatial analysis of the dynamics of rural land values in Louisiana. Specifically, spatial econometric procedures and hedonic price analysis were used to evaluate the impact of land characteristics on land prices across rural land markets in Louisiana.
Initially, hedonic models were estimated by ordinary least squares (OLS) procedures to test for the presence of spatial autocorrelation using Lagrange Multiplier tests. Results suggested that there was spatial autocorrelation in the error terms. Hedonic models were then estimated using maximum likelihood (ML) spatial error techniques. Log likelihood numbers and likelihood ratio tests where used to compare OLS and ML model estimation, and the ML was better for these data.
Information for this study includes sales that were collected for the time period January 1, 1993 through June 30, 1998, and data collected as a part of this study for the period July 1, 1998 through June 30, 2002. Data on 3,542 Louisiana rural land sales were collected during the two periods using mail survey techniques. Geo-reference of these sales indicated that sales were evenly dispersed throughout the state. Results from the data indicate that there is a substantial variation in rural real estate prices across the state.
Results from hedonic model estimation showed that cropland, pastureland, government program cotton base acreage, month of sale, value of improvements, paved road access, reasons for purchase residential, commercial and investment; residential, commercial, and highway influences; statistical metropolitan areas, and inverse of travel time had statistically significant positive influences on per acre land values. Meanwhile, size of tract, distance to nearest town, travel time to nearest town, flood influence, and reasons for purchase farm expansion and recreational had a statistically significant inverse relationship with per acre rural land values.
Marginal implicit prices were estimated using the results from models estimated by OLS and ML spatial procedures. Results indicated that, in several instances, marginal implicit prices were overestimated or underestimated when using results from OLS estimation. In general, spatial econometric techniques can be used to improve the accuracy of rural land value estimates.
Filename Size Approximate Download Time (Hours:Minutes:Seconds)
28.8 Modem 56K Modem ISDN (64 Kb) ISDN (128 Kb) Higher-speed Access Soto_dis.pdf 1.19 Mb 00:05:29 00:02:49 00:02:28 00:01:14 00:00:06