PROPOSAL

META ANALYSIS OF INFLUENCES UPON PROPERTY VALUES

DRAFT

Kenneth Acks

Since the mid 1960's thousands of "hedonic" regression models have been estimated, and more than 150 have been published in major journals. These models have regressed property prices upon a wide range of variables. Three sets of independent variables have been utilized to explain differences in sale prices--1) property characteristics, 2) neighborhood attributes, and 3) economic factors. Standard property characteristic variables include: number of bedrooms, number of bathrooms, number of rooms, square footage of the land and building, age of the house, existence of quality plumbing facilities, presence of a pool, fireplace, or garage, water frontage, the type of heating system, and date of sale. Neighborhood attributes include crime, school test scores, commuting time to employment and shopping centers, percentage nonwhite, income levels, measures of air pollution, proximity to landfills, hazardous waste sites, and refineries and the percentage of dilapidated houses in the area. Economic variables include income, inflation, and employment. Data sources for these regressions include the U.S. Bureau of the Census and other agencies, assessment and sale records, appraisal companies, Multiple Listing Services, and specially collected information.

Despite the number of studies, these analyses are rarely utilized by decision makers outside of academia. Appraisers seldom justify adjustments with regression studies, and arbitrarily revise comparable data for differences in size, time of sale, location, etc. based on their intuition and experience. Often these highly subjective adjustments merely serve to justify questionable deals or are inconsistent from report to report.

Similarly, although regressions are being used more frequently by lenders, they do not now base decisions to advance loans on such analyses. Banks do not evaluate a $200,000 loan by looking at the median sale price in a neighborhood, and then reduce the estimated value by 3.5% because the house has 3 bedrooms rather than the median number of 4; by 4.2% because it has 2 bathrooms whereas the median is 2.5, by $5,242 because the lot size is 10% below the median, and by 11% because the school test scores near the subject are 15% below districtwide results.

One of the reasons that these studies are not often utilized is that results of individual studies are not accorded a high degree of certainty. Many people feel, with some justification, that studies can prove anything. Many variables are necessarily omitted from any particular analysis due to data collection difficulties. Results may also be applicable only in a particular area. In addition, analysts may make errors with respect to the data or the model utilized. For example, larger homes in an area containing older houses, may happen to lie near a toxic waste site that has created well-publicized panic. Unless a variable representing proximity to the site is included it might appear as if price diminishes with size.

Furthermore, fielding original studies suitable to the needs of a bank can be expensive, as a large number of variables must be coded.

Meta analysis was created in the 1970's by psychometricians. It is becoming increasingly popular among social and biological scientists. The results of combining studies are reported with increasing frequency in major journals and daily newspapers, particularly with respect to medical controversies.

Economists have recently begun to add this tool to their bag of tricks. Environmental economists have utilized the technique to evaluate the benefits of air pollution reductions measured in contingent valuation and hedonic regression studies. The many conflicting studies analyzing the effects of schooling and educational expenditures upon income and economic growth have also been subjected to this test.

In "Can Markets Value Air Quality? A Meta-Analysis of Hedonic Property Value Models" (1995) V. Kerry Smith of Duke University and Resources for the Future and Ju-Chin Huang of East Carolina University conducted a meta- analysis of 36 hedonic property value models developed between 1967 and 1988, which produced 86 estimates of the marginal willingness to pay (WTP) for reducing particulate matter. Results using both Ordinary Least Squares and Minimum Absolute Deviation estimators suggest that market conditions and the procedures used to implement the hedonic models significantly influenced the estimates. The interquartile range lies between zero and $98.52 in 1982-84 dollars for a one-unit reduction in total suspended particulates (in micrograms per cubic meter). The mean WTP is nearly five times the median ($109.90 vs. $22.40) suggesting that outliers significantly influence summary statistics.

We will update and expand upon Smith's database and consider other property and neighborhood characteristics outlined above.

In addition, this study will test the responsiveness of hedonic analyses to "scope" of the variables. In 1992 the National Oceanic and Atmospheric Administration established a distinguished panel of social scientists, chaired by two Nobel laureates, Kenneth Arrow and Robert Solow, to critically evaluate the validity of contingent valuation (CV) measures of nonuse value. The Panel provided an extensive set of guidelines for survey construction, administration, and analysis, and distinguished a subset of items for special emphasis. It described them as "burden of proof" requirements. The Panel stated that if a CV survey suffered from inadequate responsiveness to the scope of the environmental insult (among other maladies) we would judge its findings 'unreliable': This paper will also attempt to use meta analysis to see if studies are responsive to scope.

Furthermore, we will create a database of studies from which analysts can draw their own conclusions (as they may wish to eliminate certain studies) which can be updated, and which can provide a good source of impact data.

Finally for some of these variables, such as the impacts of air pollution, crime, or schooling we can test whether results are consistent with other modes of analysis such as contingent valuation, wage differential, appraisal practices, and travel cost studies. Comparisons of contingent valuation and hedonic regression analyses have been conducted in a classic article by David S. Brookshire, Mark Thayer, William Schulze, and Ralph d'Arge (1982) and a more recent analysis by Shabman and Stephenson (1996).

We will focus upon the following independent variables (1) number of rooms, (2) lot size, (3) age of the house, (4) crime rates, (5) school test scores, (6) commuting time to employment, (7) measures of air pollution, and (8) the proximity to landfills, hazardous waste sites, and refineries.

OUTLINE

I. Description and History of Hedonic Analysis

Muth, Rosen, Lancaster, ...

Usage in the Real Estate Industry

II. Description and History of Meta Analysis

Origins, current uses, economic meta analyses, Kerry Smith study, ...

III. Description of Meta Analysis Model Utilized

IV. Brief Descriptions of Studies Utilized and Variables Considered. (The Data)

V. Results

VI. Application

VII. Comparison of Hedonic and Other Results

VIII. Conclusion

I. DESCRIPTION AND HISTORY OF HEDONIC ANALYSIS

Hedonic Price Models are regressions of differentiated good prices on quantities of characteristics or attributes associated with each good. The estimated coefficients are termed hedonic prices and are interpreted as the consumers implicit valuation of the characteristics or attributes. Elsewhere such values are referred to as shadow prices or Arrow-Debreu prices.

The Model is of the form

P(z) = ' · z

Where P(z) is the market price of a product which can be described by the vector of attributes or characteristics z. The vector is the vector of implicit or hedonic prices

Origins

Hedrick Houthakker assumed that the characteristics of commodities provide utility to individuals and introduced a new approach to the problem of quality variation and to the theory of consumer behavior. This new approach to the theory of individual choices helps to explain a number of phenomena that traditional economic theory cannot easily explain. In recent years, several economists have adopted the new approach to the theory of individual choices and have extended Houthakker's analysis to study consumer behavior. The theory was developed by Grilliches (1961) and Lancaster (1966) who derived a theory that demand for goods in general is a function of their characteristics. Gary Becker (1965) also contributed to the development of this literature.

Economists in the field of public economics were working along parallel lines. Tiebout (1956) suggested that households shop around the many jurisdictions in a metropolitan area for one that provides each household with its preferred mix of local services. The theory of residential location developed by Muth (1969) in Cities and Housing has become a staple in the collection of models employed by urban economists to analyze urban structure. Muth included the role of income, and in a time series study of the national housing market between 1915 and 1945 found evidence that households make adjustments over time to reduce the disequilibrium in their housing consumption.

One of the first regressions using land prices as the dependent variable was conducted by Brigham (1965). Shortly thereafter Ridker and Henning (1967) developed one of the first applications of hedonic valuation to environmental questions.

However, Freeman (1971) argued that it is inappropriate to use the marginal prices derived from hedonic regressions to estimate that willingness to pay for a nonmarginal change in the environment. This is because the regression coefficient does not measure marginal willingness to pay but rather reflects the market equilibrating price or opportunity locus in a particular supply-demand situation. If the benefits of environmental improvements are to be measured from changes in property values a general equilibrium model of land rents is required.

In order to overcome the above deficiencies researchers have followed the suggestions of Freeman (1975) and Rosen (1974) and used the hedonically estimated prices as inputs into a second stage regression through the inclusion of household characteristics including income. Unlike previous work in the area, Rosen assumed that consumers are not producers and that all the commodities with their ultimate characteristics are available and traded in the market. Shortly thereafter, Oates (1976) provided the first empirical test of a corollary of Tiebout's hypothesis--that the price of a house reflects not only its structural characteristics and those of the neighborhood surrounding it, but also the quality and cost of the public services provided by the community in which it is located.

Epple (1984 and 1987) demonstrated that most of the work that uses the hedonic approach is unsatisfactory because estimation methods do not yield consistent estimates. Bartik (1987) and Palmquist (1984) are two exceptions.

Since the mid 1960's thousands of "hedonic" regression models have been estimated, and more than 150 have been published in major journals. These models have regressed property prices upon a wide range of variables. Three sets of independent variables have been utilized to explain differences in sale prices--1) property characteristics, 2) neighborhood attributes, and (3) economic factors. Standard property characteristic variables include: number of bedrooms, number of bathrooms, number of rooms, square footage of the land and building, age of the house, existence of quality plumbing facilities, presence of a pool, fireplace, or garage, water frontage, the type of heating system, and date of sale. Neighborhood attributes include crime, school test scores, commuting time to employment and shopping centers, percentage nonwhite, income levels, measures of air pollution, proximity to landfills, hazardous waste sites, and refineries and the percentage of dilapidated houses in the area. Economic variables include income, inflation, and employment. Data sources for these regressions include the U.S. Bureau of the Census and other agencies, assessment and sale records, appraisal companies, Multiple Listing Services, and specially collected data.

Furthermore, fielding original studies suitable to the needs of a bank can be expensive, as a large number of variables must be coded.

II. DESCRIPTION AND HISTORY OF META ANALYSIS

Origins, current uses, economic meta analyses,

In order to test whether hedonic studies can provide useful guidance for real world decision makers we will conduct a "meta analysis". Meta analysis uses the results of other studies as the data for statistical analysis to distill systematic conclusions regarding relationships between variables. Data are collected from a series of studies, which are then converted to a common scale and reanalyzed using various statistical techniques. The basic null hypothesis underlying the meta-analytic statistical model is that all estimated values of the dependent variables (made comparable by appropriate adjustments if necessary) are equal to a grand mean under specified design conditions. Statistical testing of the null hypothesis can use a variety of procedures such as cross tabulations, order statistics, and t-tests. Simple averages are used but MANOVA, Nonorthogonal Multivariate Analysis of Variance is perhaps the most popular test. Regression analysis is often used to explain differences in results.

It was created in the 1970's by psychometricians. This type of research is becoming increasingly popular among social and biological scientists. The results of combining studies are reported with increasing frequency in major journals and daily newspapers, particularly with respect to medical controversies.

A search of the Carl's Uncover database of scholarly journals turned up 1,546 references to meta analysis, and 27 books were found in the Library of Congress Catalogue. The vast majority of these references were in the field of medicine and related disciplines where life and death decisions are being made on the basis of meta analysis.

Economists have recently begun to add this tool to their bag of tricks. However, relatively few meta analyses have made their way into journals. A search of Carl's Uncover using the keywords meta analysis, and economic(s), value(s), labor, output, GNP, consumption, capital, finance, investment, costs(s), benefit(s), trade, real estate, property, government, and policy revealed only the following 14 references (excluding extraneous results)

1. Andrews, Rick "The Determinants of Cigarette Consumption: A Meta Analysis" Journal of Public Policy & Marketing Spring, 1991

2. Boyle, Kevin J. "What Do We Know About Groundwater Values? Preliminatry Analysis " American Journal of Agricultural Economics December, 1994

3. Doucouliagos, Chris "Worker Participation and Productivity in Labor-Managed Firms" Industrial & Labor Relations Review October, 1995

4. Greenwald, Rob, Larry V. Hedges, and Richard D. Laine "When Reinventing the Wheel is Not Necessary: A Case Studyy in the Use of Meta-Analysis in Education Finance" Journal of Education Fincance Summer, 1994, volume 20, number 1

5. Jarrell, Stephen "A Meta-Analysis of the Union-Nonunion Wage Gap" Industrial & Labor Relations Review October, 1990

6. Loomis, J.B. "Economic Benefits of Rare and Endangered Species" Ecological Economics September, 1996

7. Marra, Michele C. "Kansas Wheat Yield Risk Measures and Aggregation" Journal of Agricultural and Resource July, 1994

8. Phillips, Joseph M. "The Effect of State and Local Taxes on Economic Development" Southern Economic Journal October, 1995

9. Phillips, Joseph M. "Farmer Education and Farmer Efficiency: A Meta Analysis" Economic Development and Cultural Change October, 1994

10. Smith, V. Kerry and Ju-Chin Huang "Can Markets Value Air Quality? A Meta-Analysis of Hedonic Property Value Models" Journal of Political Economy 1995 Volume 103 #1

11. Smith, V. Kerry and Laura L. Osborne "Do Contingent Valuation Estimates Pass a 'Scope' Test? A Meta Analysis" Journal of Environmental Economics and Management Volume 31, Number 3 November, 1996 pages 287-301

12. Smith, V. Kerry "What have we learned since Hotelling's Letter?" Economics Letters March, 1990

13. Van den Bergh Jeroe "Meta Analysis of Environmental Issues" Regional Urban Studies May, 1997

14. Vanhonacker, Wilfrie Using Meta Analysis Results in Bayesian Updating Journal of Business & Economic Statistics October, 1992

Environmental economists have utilized the technique to evaluate the benefits of air pollution reductions measured in contingent valuation and hedonic regression studies. The many conflicting studies analyzing the effects of schooling and educational expenditures upon income and economic growth have also been subjected to this test.

Some recent meta analyses are described below.

Hedonic Property Value Meta Analysis

Two Damage Calculation Methods Found to Produce Relatively Close Results

In "Contingent Valuation and Revealed Preference Methodologies: Comparing Estimates for Quasi-Public Goods", Richard T. Carson, Nicholas E. Flores, Kerry M. Martin and Jennifer L. Wright (CFM&W) conduct a "meta analysis" of 83 studies which compares "contingent valuation" (CV) and "revealed preference" (RP) valuation methods. Contingent valuation estimates are derived from surveys, while revealed preference methods utilize statistical analyses of home sales, travel costs, and/or "averting behavior". Revealed preference estimates are affected by the definition of the good, the relationship specified by the researchers, and other assumptions, such as the value of time and the number of sites analyzed. The 83 studies generated 616 comparisons.

The works examined by CFM&W provide a wide range of valuations, ranging from a recreational fishing day on the Blue Mesa Reservoir in Colorado to a statistical life. There is a significant amount of variation in goods considered, in implementation of valuation techniques, and in other factors. The contingent valuation studies were obtained from a bibliography of over 1,600 contingent valuation papers listed in Carson et. al. (1994) conducted between 1966 and 1994. Due to potential biases arising from reliance upon only the published literature, the authors included unpublished dissertations, conference papers, and government reports.

Carson, Flores, Martin and Wright report three sets of results. One set utilizes the full sample. The second trims off the smallest 5% and the largest 5% of the CV/RP ratios. The weighted sample uses the mean CV/RP ratio for each study as the study's observation

The table below presents the results

Study Set	Mean	Median	95% Confidence Interval	Correlation Coefficient Pearson/Spearman
Complete	0.89	0.747	0.813-0.960	0.83/0.78
Trimming smallest & largest	0.774	0.747	0.736-0.811	0.91/0.88
Weighted	0.922	0.936	0.811-1.034	0.98/0.92

The authors also regress the CV/RP ratios from the trimmed dataset on a set of dummy variables representing the technique used. The estimated coefficients suggest the CV estimates run about 20% to 40% lower than their RP counterparts, and that estimates for health goods are closer than others.

The authors conclude that Contingent Valuation estimates are "smaller but not grossly smaller", than their Revealed Preference counterparts. For the complete dataset, 1.0 is just outside the upper end of the 95% confidence interval for the mean CV/RP ratio. For the trimmed dataset one can clearly reject the hypothesis that the mean CV/RP ratio is 1.0. For the weighted dataset the mean ratio is not significantly different from 1.0. In every case the correlation coefficient estimates are significant at < 0.001, thus providing support for the convergent validity of the two basic approaches to nonmarket valuation of quasi-public goods. Based on the available CV/RP comparisons, arbitrarily discounting CV estimates by a factor of two or more, as some, including the influential National Oceanic and Atmospheric Administration, have proposed, appears to be unwarranted, and, in fact, make these estimates diverge from observable behavior. CV/RP ratios are greater than 2.0 in only 5 percent of the complete sample and 3 percent of the weighted sample.

Contingent Valuation Studies of Visibility Pass Meta Analysis Scope Test

Using meta analysis V. Kerry Smith and Laura Osborn found that Contingent Valuation (CV) studies on the value of changes in visibility range are responsive to the amount or scope of visibility offered. Willingness-to-pay (WTP) rises with the percentage increase in visibility across five contingent valuation studies at U.S. national parks. Because these findings indicate that summary models also fit the set of data reasonably well, they suggest a consistent economic relationship between WTP and the proportionate change in range.

The authors originally considered 13 studies, but five were chosen due to their use of comparable methods and their focus on air quality as a key element in visibility. These studies also presented the change in air quality in a way that permits computation of the change in visible range. Each study used photos to depict visibility. The table below presents a summary of the results

WILLINGNESS TO PAY IN 5 VISIBILITY STUDIES

Authors/

Year Published/

Location
#

obs.
WTP per month

1990$
Inter-quartile range of mean WTP Mean change in visibility

Mean Median

Rowe et. al

1980

Navaho Recreation Area
6 $9.27 $8.64 $6.83-$10.82 0.50

McFarland et al

1983

Grand Canyon & Mesa Verde National Parks
8 $2.75 $2.68 $1.69-$3.73 1.18

Schulze et al

1983

Grand Canyon, Mesa Verde & Zion Parks
20 $8.50 $7.00 $4.42-$11.67 0.79

Chestnut & Rowe

1990

Grand Canyon, Yosemite & Shendoah Parks
72 $4.35 $4.20 $3.15-$5.48 0.62

Balson et. al.

1990

Grand Canyon Park
10 $0.46 $0.12 $0.007-$0.97 0.955

The dependent variable for this meta analysis is the estimated mean willingness to pay for visibility change. The first commodity-related explanatory variable distinguishes eastern and western parks. The second variable related to whether or not the respondents were questioned about declines in visibility rather than increases. The last commodity variable concerns the extent of the change--whether visibility changes are to occur at a single park or for a region as a whole. Several explanatory variables described the survey design, including (1) the elicitation format (iterative bidding, direct question, or payment card/checklist), (2) whether the interview took place at the park, and (3) whether respondents live in the state where the park is located.

The authors conclude that regardless of the sample composition or model specification there is a statistically significant positive relationship between willingness to pay and proportionate improvement in the visible range.

Assessing Environmental Externality Costs for Electricity Generation

In "Assessing Environmental Externality Costs for Electricity Generation", prepared for Northern States Power Company William H. Desvouges, F. Reed Johnson, and H. Spencer Banzhaf (DJ&B) of Triangle Economic Research examined the effects of six pollutants upon human health (morbidity and mortality), agriculture (reduced crop yields), and materials (stone and metal corrosion and surface soiling) arising from six types of pollutants: particulate matter, sulfur dioxide, carbon monoxide, nitrogen oxide, lead and ozone. The authors reviewed over 400 individual health studies as well as EPA criteria documents and agricultural studies.

To obtain externality costs for each scenario the model 1) assessed ambient air quality, 2) predicted emission quantities and compositions for specific generation scenarios, 3) modeled the transport and dispersion of these emissions, 4) calculated the exposures to people, crops and materials, 5) assessed potential injuries resulting from these exposures and 6) estimated the willingness to pay to avoid these injuries.

The authors also used meta analysis techniques to synthesize information. The meta analyses also provides a means to estimate a range of valuations so that uncertainty can be quantified.

DJ&B also utilize "health state indexes" to derive value estimates for the full range of health effects. These indexes--developed by health scientists to prioritize treatment of physical conditions--combine many attributes of health (such as comfort and mobility) into a single scale. DJ&B also explore the relationship between health state and willingness to pay for health effects when willingness to pay estimates are available.

The study also uses a comprehensive database and modeling system for agricultural policy analysis developed by the Food and Agricultural Policy Research Institute (FAPRI) The model utilizes county level information to predict yield trends, acreage planted, acreage harvested, state price linkages, value of production and deficiency payments. The often significant effects of deficiency payments, the 1985 Farm Bills O-92 voluntary acreage reduction program, and the conservation reserve program are incorporated. The model provides production values and deficiency payment estimates for given ambient air pollution levels.

Finally the study uses a Monte Carlo simulation to estimate uncertainty. A Monte Carlo simulation takes estimated ranges for all parameters, randomly selects a value from each of these ranges, and then combines the estimates. The result produces one possible damage estimate. This sequence is repeated 400 times. The resulting distribution of outcomes yields an expected value, and an estimate of the 90 percent confidence interval around the most likely value. A total of 18 million calculations are made for each scenario.

One difficulty in utilizing meta analysis is the the tendency of journals to accept only strong effects or statistically significant findings, which may lead to an upward bias in the magnitude oof reported effects. As evidence accumulates results which depart from those off past studies are looked at more suspiciously by reviewers, who are often authors of previous studies. Authors may supress results that show small or insignificant, or which are at variance with accepted wisdom.

III. Description of Meta Analysis Model Utilized

IV. Brief Descriptions of Studies Utilized and Variables Considered. (The Data)

V. Results

VI. Application

VII. Comparison of Hedonic and Other Results

VIII. Conclusion

REFERENCES

Kenneth Arrow, Robert Solow, Paul R. Portney, Edward E. Leamer, Roy Radner and H. Schuman (January 15, 1993) "Report of the NOAA Panel on Contingent Valuation" Federal Register Volume 58 Number 10 pages 4601-4614

T.J. Bartik The Estimation of Demand Parameters in Hedonic Price Models Journal of Political Economy 95 pages 81-95 (1987)

Gary S. Becker A Theory of the Allocation of Time Economic Journal (!965)

Brigham, Eugue F. 1965 The Determinants of Residential Land Values" Land Economics 41 (August) pages 325-334

David S. Brookshire, Mark Thayer, William Schulze, and Ralph d'Arge "Valuing Public Goods: A Comparison of Survey and Land Hedonic Approaches" American Economic Review Volume 72 March, 1982 pages 165-177

Richard T. Carson, Nicholas E. Flores, Kerry M. Martin and Jennifer L. Wright "Contingent Valuation and Revealed Preference Methodologies: Comparing Estimates for Quasi-Public Goods" Land Economics February, 1996 Volume 72, Number 1 pages 80-99

D. Epple "Closed Form Solution to a Class of Hedonic Equilibrium Models," Carnegie Mellon University 1984

D. Epple Hedonic prices and implicit markets: Estimating demand and Supply Functions" Journal of Political Economy 95 pages 59-80 (1987)

European Commission (1995) Externalities of Energy: ExternE Project. For the Directorate General XII, Prepared by Metroeconomica, CEPN, IER, Eyre Energy-Environment, ETSU, Ecole des Mines

A. Myrnick Freeman III (1971) "Air Pollution and Property Values: A Methodological Comment" Review of Economics and Statistics 53 (November) pages 415-416

A. Myrnick Freeman III (1974) On Estimating Air Pollution Benefits from Land Value Studies," Journal of Environmental Economics and Management 1 (May) pages 74-83

Zvi Grilliches (1961) Hedonic Price Indices for Automobiles: An econometric analysis of quality change in "Models of Income Determination, Studies in Income and Wealth" Volume 28 Columbia University Press for NBER New York

Zvi Grilliches, ed. (1971) Price Indices and Quality Change: Studies in New Methods of Measurement Cambridge, Mass.: Harvard University Press

Hagler Bailly Consulting, Inc. (1995) The New York State Externalities Cost Study (Dobbs Ferry, New York) Oceana Publications

Alan J. Krupnik and Dallas Burtraw "The Social Costs of Electricity: Do the Numbers Add Up" Resources for the Future Discussion Paper 96-20 May, 1996 1616 P Street, NW Washington, DC 20036 (202) 328-5000 http://www.rff.org

Kelvin J. Lancaster A New Approach to Consumer Theory Journal of Political Economy" 74 pages 132-156 (April, 1966)

R. Lee (1995) Externalities Study: Why Are the Numbers Different? Draft Paper prepared for the Third International Workshop on Externality Costs, Ladenburg, Germany, May 27-30

R. Lee, A.J. Krupnick, D. Burtraw, et. a l. (1995) Estimating Externalities of Electric Fuel Cycles: Analytical Methods and Issues, Estimating Externalities of Coal Fuel Cycles and additional volumes for other fuel cycles (Washington, D.C.), McGraw-Hill/Utility Data Institute

R.F. Muth "Household Production and Consumer Demand Functions" Econometrica 34 pages 699-708 (!984)

Wallace E. Oates "The Effects of Property Taxes and Local Public Spending on Property Values: An empirical study of tax capitalization and the Tiebout hypothesis Journal of Political Economy 6 pages 231-242 (1976)

R. B. Palmquist "Estimating demand for the characteristics of housing" Review of Economics and Statistics 66 pages 394-404 (1984_

R. Ridker and J. Henning "The Determinants of Residential Property Values with Special Reference to Air Pollution Review of Economics and Statistics 49 (May) pages 246-257

S. Rosen Hedonic Prices and Implicit Markets: Product Differentiation in Pure Competition, Journal of Political Economy 82 pages 34-55 (!974)

Leonard Shabman and Kurt Stephenson "Searching for the Correct Benefit Estimate" Land Economics Volume 72, Number 4 November, 1996 pages 433-449

V. Kerry Smith and Ju-Chin Huang "Can Markets Value Air Quality? A Meta-Analysis of Hedonic Property Value Models" Journal of Political Economy 1995 Volume 103 #1

V. Kerry Smith and Laura L. Osborne "Do Contingent Valuation Estimates Pass a 'Scope' Test? A Meta Analysis" Journal of Environmental Economics and Management Volume 31, Number 3 November, 1996 pages 287-301

C. Tiebout " A Pure Theory of Local Expenditures Journal of Political Economy 645 pages 416-424 1956

Triangle Economic Research (1995) Assessing Environmental Externality Costs for Electricity Generation, prepared for Northern States Power Company, Minnesota Durham, N.C.

Authors/ Year Published/ Location	# obs.	WTP per month 1990$		Inter-quartile range of mean WTP	Mean change in visibility
Authors/ Year Published/ Location	# obs.	Mean	Median	Inter-quartile range of mean WTP	Mean change in visibility
Rowe et. al 1980 Navaho Recreation Area	6	$9.27	$8.64	$6.83-$10.82	0.50
McFarland et al 1983 Grand Canyon & Mesa Verde National Parks	8	$2.75	$2.68	$1.69-$3.73	1.18
Schulze et al 1983 Grand Canyon, Mesa Verde & Zion Parks	20	$8.50	$7.00	$4.42-$11.67	0.79
Chestnut & Rowe 1990 Grand Canyon, Yosemite & Shendoah Parks	72	$4.35	$4.20	$3.15-$5.48	0.62
Balson et. al. 1990 Grand Canyon Park	10	$0.46	$0.12	$0.007-$0.97	0.955