Printer Friendly

Determinants of vacant land values and implications for appraisers.

When real estate professionals, especially appraisers, are called on to appraise land, they must determine the effects and importance of area factors such as distance to the central business district (CBD) and population, as well as traditional property factors such as parcel size. Land sales in a "growth" location within a metropolitan area are evaluated and compared in an attempt to better understand the variables that influence land prices. The area examined in this study is easily accessible to major limited access highways, thus enhancing its desirability for development and growth.

The metropolitan area's growth coincides with the pattern of development for many modern cities characterized by Arthur Sullivan as a system of urban villages, or suburban subcenters where increases in population, retail stores, and medical service providers follow a pattern consistent with a multicentric suburbanized city.(1) It is important to understand factors affecting vacant land values, especially in rapidly growing areas. The use of geographic information systems (GIS) technology in the land valuation models is also demonstrated.(2)

The sales data are based on recorded transfers of vacant land (except for single-family residential lots) as reported by the Clerk of Courts' offices in Caddo and Bossier Parishes, Louisiana, during January 1984 through February 1995. The area under study has been identified by developers, appraisers, and the authors as a growth location for business activities. The location of land sales is shown in figure 1 with a geographic information systems thematic map based on sales price per square foot.

The study area is located in the northeast sector of the metropolitan area near 1-20 and 1220. This area has shown rapid development in recent years, with higher concentrations of sales in the $90,000 range [ILLUSTRATION FOR FIGURE 2 OMITTED]. Growth in this area has been attributed to both the expansion of Barksdale Air Force Base, the addition of riverboat casinos, intraurban migration (from Shreveport and Bossier City), and immigration from people outside the metropolitan area. (The two riverboats in Bossier City, along with one in Shreveport, are permanently docked and each has gross revenues of approximately $13 million per month.)

The ordinary least squares procedures are employed to estimate a land valuation model where the logarithm of the price per acre is the dependent variable. The price is adjusted for inflation, using the Consumer Price Index. During the time frame examined, 1984-1995, the local economy experienced a moderate decline in real estate prices, followed by a leveling off with price increases near the end of the time period. In general, vacant land in this area did not experience large declines, probably because most landowners adopted a wait-and-see approach. The independent variables are:

1. The zoning classification (ZB)

2. The logarithm of parcel size in number of acres (LACRES)

3. The logarithm of the distance to the central business district in number of miles (LCBD)

4. The logarithm of the population in the region (LPOP)

The metropolitan area has not used zoning to preserve farmland. (Zoning changes from agricultural to business are routinely granted unless the new use will be detrimental to neighboring residential uses; agricultural to residential zoning use is also routinely granted.) Rather, land is often maintained in an agricultural classification by the owner to take advantage of the low property tax burden. Typically zoning changes have been relatively easy to secure except for occasional cases involving controversial uses such as video poker. In the model the zoning variable (ZB) takes on a value of 1 for business zones and 0 otherwise.

Distance to the CBD, calculated using GIS procedures, is included as a measure of accessibility. It is expected that locations closer to the CBD will have higher values.

Figure 1 shows the individual land sales along with the major roads, railroads, and hydrography features. The individual property sales data are imported into the GIS as datapoint files and then assigned specific geographic locations with latitude and longitude fields through an address-matching procedure. Attribute files containing census data are imported and merged with the geographic files according to census tracts. To develop additional location or neighborhood characteristics for each of the land sales in this study, a separate layer consisting of regions (circles with a one-mile radius around each land sale) are established using GIS software. The GIS program assigns attribute values of the underlying census tracts to the regions based on a geographic area-weighted procedure that assigns values according to the proportion of each tract contained in the region. For example, if one half of a census tract's area is included within a region, then one half of the underlying variable's value would be included for that region.
TABLE 1

Statistics for Sample Land Sales in the East Area: 150 Observations
in 1984-1995

Variables Mean Standard Deviation Minimum Maximum

LPA 10.63 1.48 6.59 13.30
LACRES 0.43 1.69 -2.79 6.40
LPOP 8.44 0.41 7.10 9.16
LCBD 1.32 0.38 -0.12 2.13
ZB 0.59 0.49 0 1.00

LPA = Log of price per acre
LACRES = Log of acres
LPOP = Log of population
LCBD = Log of distance to CBD
ZB = Business zoning


After the regions are established, the variable, the size of the population, is calculated using GIS procedures. It is also expected that greater population leads to higher demand for retail and business land use and thus higher land value. The population variable is based on the underlying census block data and includes 1990 census data as well as projections for the years 1984-1990 and 1991-1995.(3) Summary statistics for all the variables are reported in table 1.

RESULTS

Table 2 shows the regression results. In the model, the dependent variable is the logarithm of the price per acre adjusted for inflation. The number of acres is significant with a negative regression coefficient, indicating that, as might be expected in vacant land sales, the smaller parcels tend to sell for more per acre than do the larger parcels. Often the larger parcels involve more development work and sometimes result in unused segments sold to other businesses later.
TABLE 2 Regressions of Log of Price per Acre on Property Variables:
150 Observations in 1984-1995

Variables Parameter t-value

INTERCEPT -2.584 -0.654
LACRES -0.3217 -6.327(a)
LPOP 1.674 3.033(a)
LCBD -1.432 -2.333(b)
ZB 1.8807 10.845(a)

LACRES = Log of acres
LPOP = Log of population
LCBD = Log of distance to CBD
ZB = Business zoning

a. Significant at 99% level,
b. Significant at 95% level,
Adjusted [R.sup.2] = 0.6457.


The population variable is significant and positive for this area, indicating that as the population increases, land values also increase. Distance to the CBD is significant and negative, indicating that a greater distance from the CBD decreases land values. In recent years, more new activities and land sales have actually taken place farther from rather than closer to the CBD.

The existence of a B zoning classification (business zoning) is proof positive for the area, an indication that this type of zoning contributed, as expected, to higher land prices. Business zoning is the prevalent type of zoning (often with retail, hotel, and medical facilities uses) in this study with some residential, agricultural, and industrial zoning classes identified to a lesser extent.

This model may be used to estimate vacant land values. For example, using the average values for the independent variables as found in table 1 results in a price per acre of $41,357. This is found by substituting 0.43 for LACRES (which translates into 1.54 acres), 8.44 for LPOP (equivalent to 4,628 persons), 1.32 for LCBD (equivalent to 3.74 miles) and 0.59 for zoning. For example, the equation:

Log price = -2.584 + -0.3217 (log of acres) + 1.674 (log of population) - 1.432 (log of distance from CBD) + 1.88 (zoning class value)

becomes:

Log price = -2.584 + (-0. 3217)(0.43) + (1.674)(8.44) - (1.432)(1.32) + (1.88)(0.59)

The result is a value estimate of 10.63 for the log of price per acre, which is $41,357 per acre. In a separate test, the year 1984 is used as a base year, and a dummy variable for each year from 1985 to 1995 was added to the model. The purpose is to see whether time is an important explanatory variable in addition to inflation. The results show that land value was significantly different from 1984 in only three years: 1985, 1987, and 1995. Thus, it appears that the inflation-adjusted model accounts for the effects of time.

This study provides an approach to studying land sales in an area identified as a growth area in an urban subcenter. The distance to the CBD is generally significant and negative. This finding is not surprising since more of the development has been taking place on the fringes of the urban area, and as expected, population was positive, showing that increasing population results in higher values. The existence of business zoning always has a positive influence on the price per acre, and the size of the parcel is inversely related to sales price.

1. A discussion of these suburbanization concepts is found in Arthur M. Sullivan, Urban Economic (Irwin Publishing: Homewood, Illinois, 1990), chapter 10.

2. For further discussion of GIS applications, see Douglas Bible and Robert J. Aalberts, "Geographic Information Systems," The Appraisal Journal (October 1992): 483-492; Douglas Bible, "Using GIS to Enhance MLS," The Appraisal Journal (October 1995): 440444; Devlin Fung, Hsiang-te Kung, and Melvin C. Barber, "The Application of GIS to Mapping Real Estate Values," The Appraisal Journal (October 1995): 445-452.

3. PCensus provided updates to the 1990 census for population and occupation for 1995. Estimates for years not included were made using linear interpolation. PCensus data is available through Tetrad Computer Applications, Inc., on the Internet at www.tetrad.com. The forecasts are based on estimates provided by the WEFA Group, econwpa.wustl.edu/EconFAQ/StatSoft/AREWEFGro.html.

Douglas S. Bible, PhD, is the Oscar Cloyd Professor of Real Estate at Louisiana State University, Shreveport. He received his PhD from Ohio State University, MBA from Indiana University, Bloomington, and BS from Bowling Green State University, Ohio. Contact: College of Business Admin.; LSU in Shreveport; Shreveport, LA 71115. dbible@pilot.lsus.edu.

Chengho Hsieh, PhD, is associate professor of finance at Louisiana State University, Shreveport. He received his PhD from Louisiana State University, Baton Rouge, MBA from the University of Missouri, Columbia, and BS from National Chengchl University, Taipei, Taiwan. Contact: chsieh@pilot.lsus.edu.
COPYRIGHT 1999 The Appraisal Institute
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1999 Gale, Cengage Learning. All rights reserved.

 Reader Opinion

Title:

Comment:



 

Article Details
Printer friendly Cite/link Email Feedback
Author:Bible, Douglas S.; Hsieh, Chengho
Publication:Appraisal Journal
Date:Jul 1, 1999
Words:1769
Previous Article:The effect of pipeline ruptures on noncontaminated residential easement-holding property in Fairfax County.
Next Article:How North American appraisers value contaminated property and associated stigma.
Topics:


Related Articles
Appraisal considerations in distressed markets.
Case study: an alternative technique to the land residual method.
Analysis of land value under differing land use restrictions.
Logical steps in property valuation.
URAR site comments often poorly developed.
Estimating the effect of a view on undeveloped property values.
Appraisal practice and Japan's bubble economy.
Real estate and business value: a new perspective.
How to analyze land values in the luxury market.
How Australian Appraisers Assess Contaminated Land.

Terms of use | Copyright © 2014 Farlex, Inc. | Feedback | For webmasters