Geoscience Reference
In-Depth Information
as the remaining unexplained temporal heterogeneity and is a measure for the
quality adjusted development of house prices over time.
￿
Categorical covariates/attributes: A high quality of the heating system ( heat )
as well as of the bathroom and toilets ( bath ) should have an increasing effect
on house prices. Furthermore, the existence of an attic ( attic_dum ), a terrace
( terr_dum ), and a garage ( garage , further separated into good and bad quality)
should raise house prices.
5.2.2
Spatial Resolutions and Neighborhood Attributes
House prices with structural attributes are nested within three spatial resolutions
and hence associated with the respective neighborhood attributes, which we use on
the most detailed level available. We use various socioeconomic and demographic
attributes as well as measures of proximity to work and metropolitan areas, obtained
from the sources described in Brunauer et al. ( 2013 ), to explain spatial variation in
house prices per sq. m.
Level-1 is the individual level, on which house prices and housing attributes
are measured (see Sect. 5.2.1 ). In total, 3,231 observations are available on the
individual level after validation.
Level-2 is the municipal level. Observations are available in 946 of the 2,379
Austrian municipalities. On level-2, we employ the following covariates:
￿
Socioeconomic/demographic characteristics of the neighborhood: On the one
hand, we use the purchase power index ( pp_ind ), the average level of education,
indicated by the share of academics ( educ ), which both reflect disposable income
and should therefore affect prices positively. On the other hand, we use an age
index ( age_ind ), constructed as a population-weighted mean of 20 age cohorts,
which measures the average age of inhabitants. A high population age index,
reflecting excess of age, serves as a proxy for structural weakness and should
have a negative effect on house prices.
￿
Measures of proximity to work and metropolitan areas: Urban economic theory
states that commuting to centers of economic activity gives rise to a location
rent, which is why a high commuter index ( comm ), that is, many employees
commuting from the municipality, should tend to affect prices negatively.
However, close proximity to these centers also provides certain disamenities,
as the local infrastructure tends to match the needs of residential use worse.
Therefore, the effect of a low commuter index is unclear. Furthermore, as a
measure of centrality, we employ population density ( dens ). In densely populated
areas, land becomes more valuable, which is why we expect a positive effect of
this covariate.
Level-3 is the district level. Individual observations are available on 109 of 121
districts, only the inner districts of Vienna are missing. As each of these districts
has neighboring units, spatial effects can be regularized using the neighborhood
Search WWH ::




Custom Search