Study Detail



Medium : Urban Amenity / Greenspace / Bushland

Tyrvainen, L. and Miettinen, A. (2000)



Country Finland  Location Salo 
Measured
Reduction in terrace house price from a 1km increase in distance from a forest park Increase in terrace house price associated with a view onto forest 
Units
% decrease in price % increase in price 
Method
Hedonic Price Method 

Key Results
Values
  Currency Year Value Australian$ 2002 Other Currency
Click (choose) in a row above to display a Key Value in another currency
Dose Response Relationships
 
Hedonic Price Relationships
 
Transfer ?
Click here if you are you considering transferring these estimates to another site

Annotated Bibliography
STUDY (FULL REFERENCE)
Tyrvainen, L. and Miettinen, A. (2000) Property prices and urban forest amenities, Journal of Environmental Economics and
Management, Vol. 39, pp. 205-223.

TECHNIQUE
Hedonic Price Method

FOCUS AND LOCATION
Affect on terrace house prices in Salo, Finland of proximity to and views of forested areas.

SITE & SOCIOECONOMIC CHARACTERISTICS
The district of Salo is located 110kms to the northwest of Helsinki. The district contains two municipalities, Salo and Halikko. In 1996,
Salo had a population of 23,000 while Halikko had a population of 9,000. In most housing areas detached houses or terraced houses
predominated. The town of Salo had approximately 1,100 ha of forest, the majority of which (74%) was found in the urban fringe.
Housing expanded into these hill areas only in recent decades. The urban forests in the district of Salo were quite fragmented and
different in respect of both tree species and age-class distribution. Most areas were pine-dominated, but also species such as oak,
linden, maple and hazelnut are frequent.

KEY RESULTS
Reduction in terrace house price from a 1km increase in distance from a forest park: 5.9%
Increase in terrace house price associated with a view onto forest: 4.9%

COMMENTS/SUMMARY
There were four different variables included initially to measure urban forest amenities. Due to issues of multicollinearity this was
reduced to two variables: the presence or absence of a forest view and distance to the nearest forested area. Although these two
variables are also related themselves the following justification was offered for treating them as distinct variables in the price equation.
The average distance to a forested area from a dwelling with a view was only 30 metres, whereas for dwellings without a view, the
average distance to the nearest forested area was 365 metres.

A model was also estimated where distance to a forested park was classified using dummy variables representing 0.005-0.099km,
0.100-0.299km, 0.300-0.599km and 0.600-0.999km. In this model, distance from the forest was statistically significant for only values
less than 600 metres. The paper also presents a method, with a hypothetical example, to use the results of the hedonic analysis in
estimating the aggregate value of a forest area.


 
Evaluation Criteria
 
Benefit transfer
Evaluation
HEDONIC PRICE METHOD (HPM)


WAS THE ENVIRONMENTAL GOOD CAREFULLY MEASURED?
Yes. Distance in kms to the nearest forest park and whether or not the property had a view to forest.

WAS PRIMARY DATA USED TO MEASURE ECONOMIC IMPACT?
Yes. Dwelling sales sample was collected from owner-occupied terraced house sales in Salo and Halikko. Selling prices and basic
structural data re characteristics of dwellings was collected from stamp duty record files. Distances to forest areas were measured off
maps.

WERE RESULTS AFFECTED BY HOUSEHOLD INCOME?
Owners socioeconomic characteristics such as income were not considered in the analysis.

WERE RESULTS CORRELATED WITH OTHER FACTORS?
A range of explanatory variables were tested including housing attributes (such as floor area, number of bedrooms, roof type,
construction material, presence of sauna, type of kitchen, age of building, yard size) locational attributes (such as distance to the centre
of Salo, shopping centre, school, swimming pool) and four different variables measuring forest amenities (distance to nearest wooded
recreation area, direct distance to the nearest forested area, relative amount of forested areas in the housing district and the view from
the dwelling window).

WERE SOCIOECONOMIC DIFFERENCES ACCOUNTED FOR?
No.

OTHER ECONOMIC/ECONOMETRIC PROBLEMS
Various models showed signs of multicollinearity and heterscedasticity. Testing led to a change in the preferred functional form and
inclusion of variables based on several estimation trials and information on the muticollinearity of variables.

SURVEY SIZE
All apartment sales in terraced housing over 3 years, 1984 to 1986. Sample size of 590.

OTHER
Data was restricted to terraced houses because the characteristics of such housing vary less than those of detached houses or blocks of
flats.

The period of data collection was a reasonably stable price period. Tests were undertaken for temporal stability.

Tests were also undertaken for spatial autocorrelation. None was found.


16 July, 2001


Copyright | DECCW home | Privacy | Feedback