obat aborsi jakarta wa 081336238223 jual obat aborsi cytotec asli di jakarta9...
Noise pollution suvarnabhumi airport
1. Economic Valuation of Noise Pollution from the Suvarnabhumi Airport
Using Home Value under Hedonic Pricing Method
Pisit Puapan and Pat Pattanarangsun
National Institute of Development Administration (NIDA)
May 2008
For Submission to Professor Adis Israngkura
ECON 951 Environmental Valuation
2. CONTENT
1. Introduction 1
1.1 Objectives of the Study 1
1.2 Benefits from the Study 1
2. Theoretical Framework and Literature Reviews 2
2.1 Hedonic Pricing Method (HPM) 2
2.2 Literature Reviews 2
3. Methodological Framework 3
3.1 Data 3
3.2 Methodological and Model 3
3.3 Variables Descriptions 5
4. Empirical Results 6
4.1 Descriptive Statistics 6
4.2 HPM Results 7
4.3 Marginal Prices 9
5. Concluding Remark 12
5.1 Summary 12
5.2 Suggestions 12
References 13
Appendix 14
Appendix A: Data 15
Appendix B: Full Model Estimation 16
Appendix c: Reduced Model Estimation 18
3. 1. INTRODUCTION
Since the opening of Suvarnabhumi Airport in September 2006, a major problem of this
Airport was created from the noise pollution affecting the surrounding communities from
hospitals, temples, schools, universities, and houses. It is possible to use the home and land
value to estimate the value of noise pollution under the presumption that it negatively affects
home and land value by using the Hedonic Pricing Method (HPM). We have conducted
survey of the home prices within the noise contour area as well as non-noise contour area
around the Airport.
1.1 Objectives of the study
To study the value of noise pollution around the Airport using the home values around
the Airport by comparing home prices in the noise contour and non-noise contour areas
under the Hedonic Pricing Method
To estimate the values of specific attributes and characteristics of homes and the
environments that could not be valued with other approaches.
1.2 Benefits from the study
To know the value of the noise pollution created by the Airport and to use this information as
policy guideline for compensations for affected families/persons.
-1-
4. 2. THEORETICAL FRAMEWORK AND LITERATURE REVIEWS
2.1 Hedonic Pricing Method (HPM)
The hedonic pricing method is used to estimate economic values for ecosystem or
environmental services that directly affect market prices. It is most commonly applied to
variations in housing prices that reflect the value of local environmental attributes.
It can be used to estimate economic benefits or costs associated with:
environmental quality, including air pollution, water pollution, or noise
environmental amenities, such as aesthetic views or proximity to recreational sites
The basic premise of the hedonic pricing method is that the price of a marketed good is
related to its characteristics, or the services it provides. The hedonic pricing method is most
often used to value environmental amenities that affect the price of residential properties.
2.2 Literature Reviews
Leggett, C.G. and Bockstael, N.E. (2000) had conducted a study on the effect on house
prices of changes in nearby faecal coliform concentrations in Chesapeake Bay, Maryland to
estimate the water quality using the Hedonic Price Method. The authors estimate a single
stage OLS hedonic model to demonstrate the effect of changes in water quality on property
prices rather than using the characteristics of houses (rooms, bathrooms, etc.), Leggett and
Bockstael use the appraised value of houses. The study found that a change of 100 faecal
coliform counts per 100 mL is estimated to produce about a 1.5% change in property prices.
The study shows some specification problems that were identified (eg heteroscedasticity and
autocorrelation. However, the correction of these problems did not have substantial effects on
value estimates.
Rahmatian M. and Cockerill L. (2004) had done a study titled “Airport Noise and
Residential Housing Valuation in Southern California” using the Hedonic Pricing Method by
comparing Model 3 Functional Forms namely Linear, Semi-Log and Log-Linear to search
for the best estimate of the marginal implicit price of airport influence. The results indicate
-2-
5. that individuals consider airport proximity and airport flight patterns in their housing
purchases. This shows that there exist two distinct measurable price gradients that distinguish
large airports from small airports. In addition, homes located under the flight path of a large
airport have a price gradient that is significantly larger than homes located under the flight
path of a small airport.
3. METHODOLOGICAL FRAMEWORK
3.1 Data
This paper obtains information to conduct the study from the following sources:
1) Websites:
- http://www.thaihomeonline.com
- http://classified.sanook.com
- http://www.ban4u.com
- http://www.pantipmarket.com
2) Books: Talad Ban ( F ), Arkarn Lae Teedin (( )
3) Phone Interview
3.2 Methodology and Model
The methodology and model are as follows:
step 1: Data collection on home values with various attributes/characteristics such as prices,
areas, utilized areas, number of restrooms, number of bedrooms, garage, distance
from the Airport, number of stories, house types (town house, single home) and
location with respect to Noise Contour.
-3-
6. step 2: Estimate the relationship between the home prices and other related variables
Dependent Variables
- Prices of houses
Independent Variables
Categorized into 3 groups
- Attributes such as number of floors, number of bedrooms
- Environmental variables1 such as noise and air particle pollution
- Community variables2 such as Crime Rate and average income of communities
step 3: Choose Functional Form of the Hedonic Pricing Model. The study chooses Semi-log
(Log-Lin) with the following form ln(P) = α0 + ΣβiZi
step 4: Conduct 1st Stage Hedonic Pricing Model Estimation to estimate the coefficients of
the model to estimate the home values with home located within and outside the noise
contour. (Moreover, we can estimate the values of specific attributes that are in the
model such as the value of garages).
step 5: Conduct the Ordinary Least Square (OLS) with the Eviews program and testing the
following:
Testing the CLRM assumptions such as Normality Test , Heteroskedasticity Test,
Serial Correlation and Multicollinearity Problem
Testing the Model such as Individual Significance, Sign and R2
1
Due to data unavailability, the authors can not obtain certain environmental data such as Decibel and PM10,
therefore, the study uses dummy variables to represent home located within and outside the noise contour
2
The study cannot obtain data on community and neighborhood, so these factors are not included in the model.
-4-
7. 3.3 Variables Descriptions
The descriptions of all variables are illustrated in Table 1
Table 1 Variables Descriptions
Variables Definition Units Expected Sign
P Sale price Baht N/A
LOT Total land area Square Wa +
AREA Total living space Square Meters +
FLOOR Number of floors floors +
BATH Number of Bathrooms rooms +
BED Number of Bedrooms rooms +
CAR Garage space cars +
DIS Distance to Suvarnabhumi airport kilometers +/-3
NOISE 1 if located in noise contour, 0 if not 0/1 -
TOWNHOUSE 1 if townhouse, 0 if single house 0/1 -
3
In the case of uncertain +/-, this is due to indeterminable effect of distance whether is positive or negative
since it is positive for transportation but negative for noise.
-5-
8. 4. EMPIRICAL RESULTS
4.1 Descriptive Statistics
The empirical results are illustrated in Table 2
Table 2 Descriptive Statistics
Mean Median Max Min Std.Dev
P 4040909 3860000 12790000 820000 2784737.3
LOT 68.64 57 287 15 56.85
AREA 335.23 288 1148 50 240.05
FLOOR 1.977 2 3 1 0.46
BATH 2.114 2 4 1 0.75
BED 2.591 2 6 2 1.00
CAR 1.477 2 4 0 0.95
DIS 12.682 12 32 7 4.89
NOISE No.of “0” (outside contour) = 28 and No.of “1” (within contour) = 16
TOWNHOUSE No. of “0” (single house) = 30 and No. of “1” = 14 (townhouse)
-6-
9. 4.2 HPM Results
The results from regression in a Full Model (inclusive of all variables) i.e.
4
P = f(LOT,AREA,BATH,BED,FLOOR,CAR,DIS,NOISE,TOWNHOUSE) are as follows:
ln(P) = 13.6475 - 0.0017LOT + 0.0017AREA + 0.2378BATH - 0.0616BED
(48.86)* (-0.415) (2.350)* (2.071)* (-0.727)
+ 0.1055FLOOR + 0.1974CAR + 0.0103DIS - 0.190 NOISE - 0.1379 TOWNHOUSE
(0.833) (2.191)* (1.149) (-1.927)** (-1.420)
Adj.R2 = 0.8839
F-Stat = 37.377
Note: “*” and “**” denote 5% and 10% level of significance respectively
From the empirical results and testing of the Full Model, it has the problem of
multicollinearity as shown in the correlation matrix in Appendix B.4 which cause most
variables to be insignificant despite high R2. Therefore, the authors exclude some of the
variables to test the following 6 models:
1. Model excluding AREA, BED
2. Model excluding AREA, CAR
3. Model excluding LOT, CAR
4. Model excluding AREA, CAR, DIS
5. Model excluding AREA, BED, DIS
6. Model excluding LOT, CAR, DIS
4
See Appendix B
-7-
10. Table 3 comparison of the results from 6 Models
#1 #2 #3 #4 #5 #6
LOT * * * *
AREA * *
FLOOR * * * *
BATH * * * *
BED * * * **
CAR * *
DIS
NOISE * * * * * *
TOWNHOUSE ** ** ** ** ** **
# independent variables 7 7 7 6 6 6
- sig at α = 5% (10%) 4 (5) 5 (6) 4 (5) 5(6) 4(5) 3(5)
- sig & correct sign 4 (5) 4 (5) 3(4) 4(5) 4(5) 3(4)
Adjusted R-squared 0.8645 0.8632 0.8730 0.8647 0.8674 0.8721
Note: 1. “*” and “**” denote 5% and 10% level of significance respectively
2. the variable with wrong sign is “BED” for all cases in which variable “BED” is
significant
-8-
11. All Models have similar Adjusted R2. The authors choose Model number 4 which has the
most significant variables as follows:
ln(P) = 13.8262 + 0.0087LOT + 0.3071FLOOR + 0.2697BATH 0.1825BED
(53.568)* (5.157)* (3.135)* (2.376)* (-2.527)*5
– 0.2731NOISE – 0.1878TOWNHOUSE
(-2.891)* (-1.819)**
Adj. R2 = 0.8647
F-Stat = 46.805
Note: “*” and “**” denote 5% and 10% level of significance respectively
This is the Model that can explain home prices from the attributes/characteristics using the
Hedonic Pricing Model as follows:
ln(P) = 13.826 + 0.0087LOT + 0.307FLOOR + 0.2707BATH - 0.183BED - 0.190NOISE - 0.138TOWNHOUSE
4.3 Marginal Prices
4.3.1 Formula for semi-log model
ln(P) = α0 + ΣβiZi
(1/P)⋅(∂P/∂Zi) = βi
(∂P/P)/∂Zi = βi
Therefore marginal price of Zi = P(Zi) = ∂P/∂Zi = βiP
5
BED or number of bedrooms shows opposite sign from our expectation, but the authors retain the variable in
the Model since the study aims to value noise pollution and this attribute may be due to consumers’ behaviors
which may need further study.
-9-
12. 4.3.2 Marginal price calculation6
From the formula in 43.1, the marginal prices of attributes can be calculated as follows:
1. P(LOT) = 0.0087*4040909.09 = 35220.56
2. P(FLOOR) = 0.3071*4040909.09 = 1241003.59
3. P(BATH) = 0.2697*4040909.09 = 1089631.14
4. P(BED) = -0.1825*4040909.09 = -737328.52
5. P(NOISE) = -0.2731*4040909.09 = -1103370.23
6. P(TOWNHOUSE) = -0.1878*4040909.09 = -758878.69
4.3.3 Interpretation of marginal prices
The interpretation of the marginal prices can be done from the variables as follows:
Inside and outside noise contour (NOISE)
(= 1 if inside, = 0 if outside)
from P(NOISE) = -1103370.23
∴ the price of house which is outside noise contour is 1103370 baht higher than the
one which is inside noise contour
ln(Pout) - ln(Pin) = 0.273050 ln(Pout / Pin) = 0.273050
Pout / Pin = e0.27305 = 1.314
∴ the price of house which is outside noise contour is 1.314 (or 31.4% greater than)
times the one which is inside noise contour
Townhouse and Single house (TOWNHOUSE)
(= 1 if townhouse, = 0 if single house)
from P(TOWNHOUSE) = -758878.69
∴ the price of single house is 758879 baht higher than the price of a townhouse
ln(Psingle) - ln(Ptown) = 0.187799 ln(Psingle / Ptown) = 0.187799
Psingle / Ptown = e0.187799 = 1.207
6
Average P = 4040909.09 has been used in the estimation of marginal prices
- 10 -
13. ∴ the price of a single house is 1.207 times (or 20.7% greater than) the price of a
townhouse
Lot size (LOT)
from P(LOT) = 35220.56
∴ the value of lot size is about 35220.56 baht per 1 square Wa
(∂P/P)/∂LOT = βi = 0.008716
∴ when lot size increases 1 square Wa, the price of a house will increase by 0.8716%
No.of floors (FLOOR)
from P(FLOOR) = 1241003.59
∴ the value of one additional floor is about 1241003.59 baht
(∂P/P)/∂FLOOR = βi = 0.307110
∴ when there is one additional floor, the price of a house will increase by 30.71%
No.of bathrooms (BATH)
from P(BATH) = 1089631.14
∴ the value of a bathroom is about 1089631.14 baht
(∂P/P)/∂BATH = βi = 0.269650
∴ when there is an additional bathroom, the price of a house will increase by 26.97%
No.of bedrooms (BED)
from P(BED) = -737328.52
∴ the value of a bedroom is about -737328.52 baht
(∂P/P)/∂BED = βi = -0.187799
∴ when there is an additional bathroom, the price of a house will decrease by 18.78%
- 11 -
14. 5. CONCLUDING REMARK
5.1 Summary
Noise problem from Suvarnabhumi airport can be reflected from a difference between prices
of houses which are in and out of noise contour ≈ 1,103,370 Baht. Furthermore, other
attributes which may not be valued directly or easily can be determined by the calculation of
marginal prices from hedonic price function in the 1st stage.
5.2 Suggestions
5.2.1 Recommendations for further study
Collecting more observations to make the results more reliability.
Add more independent variables to capture other effects to home value and be
able to focus on the values of specific attributes
- Pollution level e.g. dB(A), PM10
- Community and neighborhood variables e.g. crime, income
Adjusting the Model such as
- Changing the Functional Form e.g. linear Boxcox
- Including the Interaction Term or Slope Dummy into the Model
Estimate demand (2nd Stage Hedonic) to measure welfare (in case of non-
marginal change)
5.2.2 Policy Recommendation
The concerned authorities should consider compensating the people affected by
the noise pollution within and outside the noise contour by 1.1 million Baht.
Consideration should be made for the timeline of residence such as the new
residents (after the Airport’s opening) have already benefited from lower home
prices, and should not be compensated for noise pollution. On the other hand,
old residents (living prior to the Airport’s opening) should be compensated
accordingly.
- 12 -
15. REFERENCES
Adair, A.S., Berry, J. N& McGreal, W. S. (1996). Hedonic modeling, housing submarkets
and residential valuation, Journal of Property Research, vol. 13, pp. 67-83.
Clark, D. E. & Herrin, W. E. (2000). The Impact of public school attributes on home sale
price in California, Growth and Change, vol. 31, pp. 385-407.
Espey, M. & Lopez, H. (2000). The impact of airport noise and proximity on residential
property values, Growth and Change, vol. 31, pp. 408-419.
Garrod G. and Keneth G. Willis. (2003). “A Primer on Non market Valuation”. 1st ed.
Edward Elgar Publishing Limited.
Leggett, C. G. & Bockstael, N. E. (2000). Evidence of the effects of water quality on
residential land prices, Journal of Economics and Management, vol. 39, pp.121-144.
Patricia A. Champ, Kevin J. Boyle and Thomas C. Brown. (1999). “Economic Valuation of
the Environment: Methods and Case Studies”. 1st ed. Kluwer Academic Publisher.
Rahmatian M. and Cockerill L. (2004) “Airport Noise and Residential Housing Valuation in
Southern California”. International Journal of Environmental Science & Technology.
vol.1, No.1, pp.17-25.
- 13 -
