Page 131 - Contributed Paper Session (CPS) - Volume 1

CPS1198 Firano Z. et al.
characteristics. Thus, one of the important factors is the value of neighbouring
properties. In fact, the higher the value of a good, the greater the probability
that a neighbouring good will have such a high price (phenomenon of real
estate mimicry). In this design, it is necessary to take into account the spatial
autocorrelation that measures neighbourhood effects. Spatial autocorrelation
is based on the observation that spatialized observations in cross sections are
not independent. Spatial autocorrelation is defined as the correlation, positive
or negative, of a variable with itself arising from the geographic location of the
3
data. .

4. Results
To test this spatial autocorrelation we opted for the Geary index (1954)
which measures the local spatial dependence between real estate. The value
obtained from the Geary index is less than one, making it possible to reject
the hypothesis of non-existence of spatial autocorrelation. In this respect, we
can say that real estate prices in Rabat-Hay Riad and Témara are auto
correlated in space. In other words, the value of goods is influenced by
neighbouring prices. The existence of the spatial correlation up to the fifth
neighbourhood in the region corroborates the hypothesis of the influence of
neighbouring prices on the real estate sales value. Therefore, we proposed a
model that takes into account this spatial dependence effect estimated using
the instrumental approach to correct the low exogeneity bias. Models were
estimated according to the following specification:
,
, = −, + ∑ + + , 4
. ,
=1;=1

Where i is the individual dimension describing real estate and j is the
nature of the property. Indeed, we have distinguished in the estimation
between the different categories of real estate (apartment, villa, land,... etc.).
Therefore, we have 8 categories of property and more than 11,000 properties
in the region of Rabat-center, 6,500 properties on Hay Riad and 20,400 on the
city of Témara.

3 Julie Le Gallo. Econométrie spatiale (1, Autocorrélation spatiale). [Rapport de recherche]
Laboratoire d’analyse et de techniques économiques(LATEC). 2000, 45 p., Table, ref. bib. : 5 p.
<hal-01527290>.
4 , is a combination of supposedly random residues and effects specific to types of real estate.

120 | I S I W S C 2 0 1 9