Limitations of Hedonic Pricing – Project topics materials

Get Complete Project Material File(s) Now! »

Focus variables

The focus variables of this paper are the distance-based amenity variables, including dis-tance to city centers, permanent open spaces and other natural amenities. We predicted that as distance to these variables increases the price of the house will decrease, suggest-ing a negative relationship. Based on our regression, our hypothesis was proven to be true for all except for one amenity variable. The distance to Vättern, the nearest lake and the distance to nearest park all hold this negative relationship with housing prices which we predicted in the section 4, the same holds for the distances to city centers. This is in line wit the economic theory of Buchanan (1965), Tiebout (1956), Bruckner et al. (1999) Von Thünen (1826) all of which, with different approaches, state that housing prices will be higher when the house is located near one of these amenities.
Being the second biggest lake in Sweden and the sixth in Europe, Vättern holds an impact of decreasing house prices of 0.117% according to model 1 and 0.104% according to model two, when increasing the distance to Vättern by 1%. Due to the location of Jönkö-ping city and the U-curved shape of the southern part of lake Vättern, this distance vari-able might be captured in other variables such as the distance to the Jönköping variable. Moreover, due to the extensive length of the Vättern’s southern coast line, the different house price levels represented along this region are exposed to significant volatility.
We also wanted to examine how much of an impact the distance to the closest lake and park has on house prices. The result from these variables are as expected, but with mixed significance. Proximity to nearest lake has a decreasing impact of more than 0.02% of the price if the distance to the lake increases by 1%, with a significance level of 10% for model 1 and 5% for model 2. The park-proximity variable holds a similar relationship with house prices and we can observe that we get a negative and statistically significant value of 0.020 for model 1 and a negative value without any significance of 0.014 for model 2. This implies, according to model 1, that the house price will drop slightly with 0.02% as we increase the distance to the nearest park by 1%.
Distance to permanent open spaces (POS) is however positively related with distance to the observed houses, but not to a significant level. This is the contrary to our expectations and thus deviates from the result of the work by (Geoghegan, 2002 & Geoghegan et. al, 1997) where she states that a neighbourhood consisting of a large fraction of permanent open spaces will face an increase in property values. However, a positive relationship may be explained by the definition of what permanent open space amenities are. Since we have distance to parks and lakes as separate variables, the distance to nearest POS-variable may include less attractive open spaces such as swamp marks, farm lands and other government protected environmental areas that may not be very desirable. The dis-tance to the nearest POS-variable therefore absorbs more of the open spaces that does not increase the value of the property. This is shown by the positive but not significant coef-ficient values of 0.016 for model 1 and 0.017 for model 2.
As additional focus amenities, we chose to include distances to city centers. For the pur-pose of this paper, we have chosen to include Jönköping, Huskvarna, Bankeryd and Gränna as city centers as they capture complete or parts of the definition of a CBD (Thü-nen, 1966). The cities are all negatively correlated with property prices and this is in line with our prediction regarding distances to the nearest urban centers. This is explained theoretically in the monocentric city model where the bid-rent curve suggest that the rent is higher near the central business district. In this case, it might be up for debate to define some of our city centers of choice as being central business districts. However, in order to pove a point, this definition is assumed to hold for all city centers and thus is part of the explanation for the negative relation with house prices and distances to CBD’s. Furthermore, close distance to the cities of Jönköping, Huskvarna, Bankeryd and Gränna do all absorb other city amenities than being the main employer. Such city amenities are restaurants, bars, shopping, local buzz and other leisure amenities.
The results from our regression models all held true to our expectations of the signs and all four of the distances between property and city centers show a negative relation with house prices at a 5% significance level. The most prominent city-distance variable is surprisingly the distance to Bankeryd. This variable shows a coefficient value of -0.238 where both regression models support the negative relationship argued above, and may be interpreted as a de-crease in the house price of 0.238% alongside an increase in the distance between Bank-eryd and the property by 1%. As Jönköping is the biggest city center amongst these vari-ables, we expected this variable to have more impact on prices as it plays a crucial role as main employer of the region. Contrarily, the distance to Jönköping city was the second most influential city-distance variable and presented that with a 1% increase in distance, we observed a decrease of the house price equivalent to more than 0.15%. The other two city-distance variables of Gränna and Huskvarna are both also statistically significant at a 5% level but have a lower impact on house prices. The reaction of increasing the dis-tances to Gränna and Huskvarna by 1% are equivalent to a price reduction of 0.1% and 0.05%, respectively, according to model 1.
The reasoning behind Bankeryd being the most prominent city-distance variable instead of Jönköping is unclear to us since we expected Jönköping to have the biggest impact due to the comparably higher supply of positive externalities being created in the bigger city of Jönköping. However, according to Yang and Fujita (1983) high income families may tend to locate outside the city centers regardless of the amenities and externalities sup-plied by the city center, and this might provide one reasoning of why the Bankeryd vari-able contributes the highest effect on house prices.

READ The functions and objectives of government

Control variables

regression model 1 includes dummy variables for each year beginning in 2001 and ending in year 2011. These dummies provide a clear indication of the development of the prices on the housing market in Jönköping municipality. All values are significant at a 5% significance level except for the Dummy 2001 variable, meaning that there was not enough changes in housing prices between the year 2000 and 2001. The property prices tend to increase each year according to the estimates, except for the years 2001 and 2005. In 2007 and 2008 house prices experienced minor stagnation due to the recession during these years. If we observe the development of the dummy coefficients over time, there is a very clear increasing trend that corresponds to the overall trend observed in real house prices in Sweden over these specific years. The last year of our observation shows that the prices had risen about 100% in nominal terms since 2000, which is a fair estimate in accordance to the report by Statistics Sweden which states that the overall developmen of house prices in Sweden increased by 196% in nominal terms since 1990 (Statistics Sweden).

1 Introduction
2 Theoretical Framework
2.1 Consumer Theory
2.3 Amenity valuation
3 Models.
3.1 Hedonic pricing
3.2 Limitations of Hedonic Pricing .
4 Data
4.1 Functional form
5 Data results and analysis
5.1 Focus variables
5.2 Control variables.
5.2.1 Land value and focus variables
6 Conclusion
References
Appendix