Quantifying the influence of desert riparian areas on residential property values. (Features).

Abstract

The appraisal profession has long recognized the positive influence on residential property values of proximity to attractive amenities. Statistical analysis has documented this effect for proximity to beaches, lakes, riverfront, wetlands, and urban open space. This article examines the impact of desert riparian areas, an important but vanishing natural amenity throughout the arid western United States. It also analyzes the effects of riparian corridors (proposed for protection) on property values in the northeast Tucson metropolitan area in southern Arizona.

**********

Desert riparian areas form ribbons of green across dry landscapes, with cottonwood, mesquite, and willow trees that depend on a shallow depth of groundwater to survive. Riparian corridors support a wide variety of birds and other wildlife that could not live in the desert without access to riparian areas. Groundwater pumping affects riparian areas when it causes the water table to drop beyond reach of the riparian plants. This article examines the contribution to property value that results from location in a riparian area by statistically analyzing real estate sales near riparian corridors proposed for protection from new groundwater wells in northeast Tucson.

Landowner Benefits from Riparian Corridors

Along riparian corridors, large stands of cottonwoods, mesquite, and willows support diverse birds and wildlife and provide a variety of benefits to nearby landowners. Homeowners enjoy scenic views, open space, bird and wildlife viewing, and a buffer from urban noise. Riparian trees provide shade, giving respite from summer heat and lowering home cooling costs. (1) Policies that protect riparian areas from groundwater table declines also help protect the water table for existing well owners. Declines in the water table not only affect riparian areas, but they also cause higher costs for well owners who must pump from deeper levels and may need to invest in deeper wells. Proposed limits on new wells in riparian buffer zones help assure landowners that when they refrain from drilling a new well to protect riparian corridors, others also must refrain. (2) This can prevent landowners from inadvertently damaging riparian resources that add to the value of their properties' value, but which no individual landowner can protect alone.

Statistical analysis of real estate transactions help identify the contributions to property values in a riparian area. The hedonic valuation method has been used for many decades to examine the impact on property value that is made by proximity to both desirable features (open space, beaches) and undesirable sites (landfills, airports). We use it here to examine the property value premium that riparian corridors provide to landowners.

Myrick Freeman presents an excellent description of the underlying theory and the uses of the hedonic method. (3) Ronald Ridker was the first to apply this theory by investigating the effect of air quality on residential housing prices in St. Louis. (4) Since then, there have been hundreds of studies using the hedonic method. The hedonic property price method statistically relates real estate sale prices to a set of factors that influence property prices (age, size of parcel, square footage of home, etc.). Using this method, the value of proximity to a riparian corridor can be quantified separately from the other components of value.

Numerous other studies have documented property value premiums associated with riparian areas. For instance, Mahan, Polasky, and Adams estimated the value of wetlands in the Portland, Oregon, area. Using 14,233 real estate transactions, these authors statistically linked sale prices of residential homes to a set of variables that includes proximity to urban wetlands. This study found that a home's value increased by $436.17 for being located 1,000 feet closer to the nearest wetland. (5)

In the early 1980s, King, White, and Shaw examined the effects of proximity to riparian habitat and other wildlife habitat on the value of single-family residences throughout the Tucson metropolitan area. After accounting for other factors that affect home sale prices (square footage, age of home, location, etc.), this study identified a 3%-5% premium in the sale prices of single-family residences located within one-half mile of large, open space preserves, riparian areas, and other wildlife habitat. (6)

The Data

Our analysis uses real estate and geographical information system (GIS) data for private property parcels within 2.5 miles of a 15-mile stretch of the Tanque Verde Wash and nearby riparian corridors proposed for protection in the Tucson metropolitan area (Pima County GIS). (7) The statistical model analyzes sales of single-family residences over four years (1996-1999). The data set includes 7,658 single-family residential home sales.

Statistical Models and Results

The model used in our statistical analysis:

Sale Price is a function of: Year Sold, P S, L S, A, G, L(D)

Table 1 provides definitions of these variables and summary statistics. The model includes a nonlinear relationship between distance and sale price because the marginal effect on home price is not constant with respect to distance from the riparian corridor. Instead, the property value premium is proportionally higher for homes close to the riparian area than for those located farther away. Table 2 summarizes the statistical results.

The model has an R-squared of 0.9125, which indicates that it is an excellent predictor of home prices. (8) The t-statistics indicate each independent variable that has a significant influence on home price. All coefficient estimates are significant at the 5% level--a common benchmark for evaluating statistical significance.

All of the variables, except Indistance, are linear with respect to sale price. Their coefficients represent the marginal implicit price for the variable. That is, the coefficients are the amount that a home's sale price will change for a one-unit change in the independent variables. The model also demonstrates how square footage, garage size, and age affect the value of the home. The size of the land parcel has a positive effect, as does increased living space. Distance from the riparian corridor has a negative impact on a home's sale price. The variable for distance to the riparian corridor is included in the statistical model as a logarithmic function. Its regression coefficient is equal to the proportional change in home price per unit change in distance. (9)

Figure 1 illustrates the statistical results for a typical 2,000-square-foot home in the study area. (10) The statistical model indicates a sale price for this home of $192,107 when located 0.10 mile from the riparian corridor. However, if the identical home were located 1.5 miles from the riparian corridor, its price would fall to $181,466. The difference of $10,640 (6%) is the increased property value due to proximity to the riparian corridor. While this study examines only one portion of the metropolitan Tucson area, property value premiums also can be expected near riparian corridors elsewhere in Arizona's desert communities. Table 3 shows the contributions of each variable in the model to the value of a representative home.

Homes in close proximity (within 0.5 mile) to the wash are priced at a considerable premium compared to those farther away. Table 4 shows how a representative home increases in value as its location is moved closer to the riparian corridor. Table 4 indicates that a 15-year-old, 2,000-square-foot home with a one-car garage (which is located on a 0.25-acre parcel) would drop in market value by 6% if its location were changed from 0.10 of a mile from the riparian corridor to 1.5 miles away. The statistical analysis indicates that the zone of influence on property value extends about 1.5 miles from the center of the riparian corridor, with only a small influence on property value beyond 1.0 mile. Riparian corridors have a notable positive effect on single-family residential property values, which increases with a home's proximity to the riparian corridor.

Overall Single-Family Residential Property-Value Premiums

This part of the study considers the riparian corridor's effects on property value for all homes located within 1.5 miles of the corridor. The property value premium for each home is the difference between home value at the home's actual distance and the value as predicted by the statistical model at 1.5 miles from the riparian corridor. Homeowners capture this premium when they sell their homes. The statistical model indicates that the property-value premium for owners of the 25,560 single-family residences located within 1.5 miles of the center line of the Tanque Verde Wash is $103.1 million. Seventy-five percent of this riparian-property-value premium ($77.3 million) goes to homeowners located within a 0.5 mile of the riparian corridor. For homeowners located within the proposed riparian buffer zone, policies that limit new wells could provide some property value protection.

Conclusions

Riparian corridors contribute substantially to the property values of nearby homes, particularly those within die first 0.5 mile of the riparian area. This report documents a property value premium of $103.1 million for 25,560 single-family residences that lie within 1.5 miles of riparian corridors proposed for protection. Three-quarters of this premium ($77.3 million) goes to homes located within a 0.5 mile of the riparian area. While this study focused on northeast Tucson, property-value premiums can be expected for residences located near riparian corridors in other desert cities.

Table 1

Variables Used to Examine Property Value Impacts *

Variable Name  Description

Dependent
 variable
 sale price    Sale price of home
Independent
 variables
  1996         Variable representing a 1996 home sale (1,594 sales)
  1997         Variable representing a 1997 home sale (1,832 sales)
  1998         Variable representing a 1998 home sale (2,339 sales)
  1999         Variable representing a 1999 home sale (1,893 sales)
Parcel size    Size of land parcel, measured in square feet
Living space   Total living space of home, measured in square feet
Age            Age of home in years at time of sale
Garage         Number of garaged parking spaces
Distance       Distance in miles to center line of riparian corridor
Summary Statistics

Variable Name            Average    Minimum    Maximum

Sale price ($)          182,295.10    29,900  1,855,584
Parcel size (sq. ft.)    23,782.23  2,023.70  3,606,709
Living space (sq. ft.)    2,052.16       800      7,765
Age (years)                  15.37         0         98
Garage (# of spaces)          1.54         0          6
Distance (miles)              0.84      0.10       2.50

* The data was checked for inaccuracies by examining each variable in
turn and inspecting for extreme values. For example, we dropped five
observations from the data set that listed the area of the land parcel
to be under 200 square feet, 10 which listed zero bathroom fixtures, and
one observation which indicated that the home had 200 garaged parking
spaces.
Table 2

Statistical Results: Single Family Residences *

                                                          P>[absolute
Variable      Robust Coefficient  Standard Error    t     value of t]

1996              -58537.37          5663.41      -10.34     0.00
1997              -55185.27          5750.45       -9.60     0.00
1998              -49514.99          5703.83       -8.68     0.00
1999              -31320.59          6034.53       -5.19     0.00
Parcel SIZE            0.43             0.07        6.32     0.00
Living space         105.88             3.22       32.93     0.00
Age                 -560.38            81.94       -6.84     0.00
Garage              6290.79          1128.88        5.57     0.00
LnDistance         -3929.19           890.43       -4.41     0.00

* Ordinary least squares regression was used to estimate the hedonic
price model, using SPSS and Stata statistical software. The results were
then checked for heteroskedasticity using the Cook-Weisberg
(Breusch-Pegan) test. The tests indicated heteroskedasticity and White's
(Huber's) method was used in a second regression to recover consistent
estimates for the standard errors of the coefficients. Number of obs =
7685; F(9, 7649) = 11186.99; Prob > F = 0.0000.
Table 3

Contributions of the Variables to Property Values

Variable      Increase in Home Value for a
              One-Unit Increase in the
              Variable (For the Average
              Home)

Parcel sale   $0.43 per sq. ft.
Living space  $105 per sq. ft.
Age           $560 per year (negative)
Garage        $6290 per garaged parking spac
Distance      $4650 per mile (negative)
Table 4

Single-Family Residential Values and Proximity to Riparian Corridor

Distance (miles)  Property Value  % Change from 1.5 Miles

      0.1            $192,106               6.0
      0.3            $187,790               3.5
      0.5            $185,783               2.4
      1.0            $183,059               0.9
      1.5            $181,466               0.0
Figure 1

Statistical Results of a 2,000-Square-Foot, One-Car Garage, 15-Year-Old
Home on One-Quarter Acre *

                       Percentage     Distance from
Home Value  Increase    Increase   Riparian Centerline

 $181,466                               1.50 miles
            + $1,593    (+ 0.9%)        1.0 mile
            + $4,317    (+ 2.4%)        0.50 mile
            + $6,321    (+ 3.5%)        0.30 mile
 $192,107   + $10,641   (+ 5.9%)        0.10 mile

* Based on 1999 northeast Tuscon home values. Colby and Wishart, 2002,
Department of Agricultural and Resource Economics, The University of
Arizona.

Note: Table made from line graph

Note: Support for this research was provided by the Arizona Agricultural Experiment Station. The authors appreciate assistance in obtaining data given by staff of the Arizona Department of Water Resources. Greg Fitzpatrick provided tireless work on the report production and design with Nancy Bannister. Dr. Harry Ayer and Dr. Satheesh Aradyula provided peer review of this report with useful comments and suggestions. The authors are solely responsible for the research findings.

(1.) A 1996 survey of homeowners in three different parts of Pima county found that 64% of respondents believed that loss of riparian vegetation near their home would have a negative impact on their property. K. Novak, "The Importance of Xeroriparian Habitat to Single Family Residents in Unincorporated Pima county," Master's Thesis, School of Renewable Natural Resources (University of Arizona, 1997).

(2.) Governor's Water Management commission, State of Arizona, Final Report (December 2001). Available from the Arizona Department of Water Resources.

(3.) Myrick A. Freeman Ill, The Measurement of Environmental and Resource Values (Washington, D.C.: Resources for the Future, 1993).

(4.) Ronald Ridker and John Henning, "The Determinants of Residential Property Values with Special Reference to Air Pollution," The Review of Economics and Statistics (49:2, 1967): 246-257.

(5.) Mahan, Brent, Stephen Ploasky, and Richard Adams, "Valuing Urban Wetlands: A Property Price Approach," Land Economics (76:1, 2000): 100-113.

(6.) D. King, J. White, and W. Shaw, "Influence of Urban Wildlife Habitats on the Value of Residential Properties," in Wildlife Conservation ion in Metropolitan Environments (Columbia, Maryland: National Institute for Urban Wildlife, 1991).

(7.) Pima County Geographic Information System, Land Information System for ArcView, Pima County Technical Services, Version 11.0.

(8.) R-squared (the co-efficient of multiple correlation) measures the proportion of the total variation in the independent variable that is explained by the model composed of the independent variables (Jack Johnston and John DiNardo, Econometric Methods, New York: McGraw-Hill, 1997). A review of six other published hedonic price studies revealed an average R-squared of 0.722, with a range of 0.4148-0.939.

(9.) The graph starts at 0.10 mile instead of at zero because homes are not located in the center of washes.

(10.) Jack Johnston and John DiNardo, Econometric Methods (New York: McGraw-Hill, 1997).

References

Freeman, Myrick A. III. The Measurement of Environmental and Resource Values (Washington, D.C.: Resources for the Future, 1993).

Johnston, Jack, and John DiNardo. Econometric Methods (New York: McGraw-Hill, 1997).

Mooney, Sian, and Ludwig M. Eisgruber. "The Influence of Riparian Protection Measures on Residential Property Values." Journal of Real Estate Finance and Economics (22:2, 2001): 273-286.

Bonnie G. Colby is professor of agricultural and resource economics at the University of Arizona, where she has been a faculty member since 1983. She has authored numerous publications on the economics of water issues, including the books Water Markets in Theory and Practice and Indian Water Rights: Negotiating the Future. She has provided invited testimony on these matters to state legislatures around the West, and to Congress. She served on the National Research Council's Committee on Western Water Management, the Committee on Managing Glen Canyon Dam, and the National Academy of Science committee investigating use of economic methodology by the Army Corps of Engineers for billion-dollar proposed projects on US waterways.

Steve Wishart, a native Minnesotan, is finishing his MS degree in agricultural and resource economics at the University of Arizona. He plans to complete a PhD in this field and is a champion cyclist in his free time.