# RECORD 323b30d53cb25830c8397eeb95e569a5dc5a3044
# TITLE: Sub-National Spatial Price Indexes for Housing: Methodological Issues and Computation for Italy
# YEAR: 2022.0  VENUE: Journal of Official Statistics  DOI: 10.2478/jos-2022-0004
# AUTHORS: Ilaria Benedetti; L. Biggeri; T. Laureti
# PAGES: 26  CANDIDATE TABLE PAGES (1-based): [10, 13, 14, 15, 16, 18]
# ABSTRACT: Abstract It is essential to measure within-country differences in housing costs in order to evaluate costs of living, assessing and comparing poverty levels, quantifying salaries and disposable income of families and finally for designing housing policies at local level. To the authors knowledge, no studies have yet been carried out on the computation of Space Price Indexes for Housing Rents (SPIHRs). In this article we computed preliminary estimates of sub-national SPIHRs by using hedonic regression model, which is an extension of the Country Product Dummy method, for all the Italian regions. The hedonic regression is generally used to obtain multilateral spatial indexes, thus allowing us to obtain multilateral SPIHRs for the Italian regions. The estimates have been done using 2017 data from the Real Estate Market Observatory which is a part of the Italian Agency of Revenue and Tax. This data source is the most comprehensive source of information on Italian houses price rents with a wide geographical coverage, including data for each Italian municipality. The obtained results show significant differences across the Italian regions, thus highlighting the importance of calculating SPIHR in Italy on a regular basis and the need to continue researches in this field.


===== PAGE 1 (score 5, numeric-paren 0, table-words 0, var-words 5) =====

Sub-National Spatial Price Indexes for Housing:
Methodological Issues and Computation for Italy
Ilaria Benedetti1, Luigi Biggeri2, and Tiziana Laureti1
It is essential to measure within-country differences in housing costs in order to evaluate costs
of living, assessing and comparing poverty levels, quantifying salaries and disposable income
of families and finally for designing housing policies at local level. To the authors knowledge,
no studies have yet been carried out on the computation of Space Price Indexes for Housing
Rents (SPIHRs). In this article we computed preliminary estimates of sub-national SPIHRs by
using hedonic regression model, which is an extension of the Country Product Dummy
method, for all the Italian regions. The hedonic regression is generally used to obtain
multilateral spatial indexes, thus allowing us to obtain multilateral SPIHRs for the Italian
regions. The estimates have been done using 2017 data from the Real Estate Market
Observatory which is a part of the Italian Agency of Revenue and Tax. This data source is the
most comprehensive source of information on Italian houses price rents with a wide
geographical coverage, including data for each Italian municipality. The obtained results
show significant differences across the Italian regions, thus highlighting the importance of
calculating SPIHR in Italy on a regular basis and the need to continue researches in this field.
Key words: Spatial rental price indexes; hedonic regression; country product dummy method;
symbolic data analysis.
1.
Introduction
It is essential to measure within-country differences in housing costs in order to evaluate
costs of living, assessing and comparing poverty levels, quantifying salaries and
disposable income of families and finally for designing housing policies at a local level
(Bishop et al. 2017).
It is well known that when making comparisons, the sub-national (regional) values of
the economic indicators should be adjusted for sub-national price differences in order to
avoid misleading regional analyses and the consequent policy implications and outcomes,
which is done by computing sub-national Spatial Price Indexes (SPIs) as illustrated by
Biggeri et al. (2017) and Laureti and Rao (2018).
In this context, the computation of sub-national SPIs for housing costs is particularly
important since housing costs can be a substantial financial burden to households,
especially for low-income families.
These sub-national SPIs have proven to be useful for assessing and comparing poverty
levels and for implementing housing policies for the poor at local level, however they can
also be used as proxies of general sub-national household consumption SPIs or in
q Statistics Sweden
1 University of Tuscia, Via del Paradiso 47, 01100, Viterbo (VT), Italy. Emails: i.benedetti@unitus.it and
laureti@unitus.it
2 The University of Florence, Viale Morgagni, 59, 50134 Firenze, 50134, Italy. Email: Luigi.biggeri@unifi.it
Journal of Official Statistics, Vol. 38, No. 1, 2022, pp. 57–82, http://dx.doi.org/10.2478/JOS-2022-0004


===== PAGE 10 (score 20, numeric-paren 0, table-words 4, var-words 2) =====

quotations for all of the Italian municipalities and on the average quotations of the for the
20 Italian regions according to type of dwellings and conservation status. Quotations for
poorly conserved dwellings are very few to compute adequate descriptive statistics.
Table 1 shows the variability of the rental quotations observed for the 8,046
municipalities and across the Italian regions and how type of dwelling and conservation
status affect the distribution of the minimum and maximum rental prices. As illustrated in
Table 1, the highest rental value is observed for well-finished apartments, for which the
rental price ranges from 3.99 to 5.37 EUR/m2. In particular, the mean rental price ranges
from 4.90 to 6.37 EUR/m2 for excellently conserved apartments (the maximum rental
price shows as well the highest level of variability). Contrastingly, the lowest rental value
is observed for economic apartments, for which the rental prices range from 2.73 to 3.72
EUR/m2.
From the second part of the Table 1, we can see that also the variability of the average
rental values across regions is very high and this surely justify the importance to compute
the SPIHRs for the Italian Regions.
Table 1.
Descriptive statistics based on the quotations available for all the municipalities. Italy, 2017.
Conservation status
Total
Excellent
Normal
Type of dwelling
Min
Max
Min
Max
Min
Max
Detached
Mean
3.60
4.86
4.72
6.12
3.36
4.58
Standard deviation
2.04
2.78
2.42
3.39
1.86
2.55
Coefficent of variation
0.57
0.57
0.51
0.55
0.55
0.56
Skewness
2.41
2.65
2.50
2.85
2.36
2.45
Well-finished
Mean
3.99
5.37
4.90
6.37
3.83
5.19
Standard deviation
2.09
2.94
2.54
3.76
1.96
2.73
Coefficent of variation
0.52
0.55
0.52
0.59
0.51
0.53
Skewness
2.39
2.94
2.89
3.89
2.13
2.32
Economic
Mean
2.74
3.73
3.50
4.38
2.68
3.68
Standard deviation
1.59
2.13
1.97
2.53
1.54
2.08
Coefficent of variation
0.58
0.57
0.56
0.58
0.58
0.57
Skewness
2.41
2.25
3.43
3.12
2.19
2.09
Total
Mean
3.46
4.67
4.58
5.91
3.27
4.47
Standard deviation
2.00
2.73
2.44
3.46
1.85
2.53
Coefficent of variation
0.58
0.59
0.53
0.59
0.57
0.57
Skewness
2.36
2.68
2.66
3.26
2.20
2.31
Journal of Official Statistics
66


===== PAGE 11 (score 12, numeric-paren 0, table-words 2, var-words 1) =====

5.
The Specification of the Hedonic Models and the Estimation Procedure
As regards the estimation strategy, it is worth noting that, as already mentioned in
Section 4, the OMI database provides minimum–maximum range of rental prices
according to OMI zones, type of dwelling and conservation status. Therefore, for each
observation two rental prices are specified: the minimum and the maximum rental value.
As a consequence our dependent variable in model (1) is an interval-valued variable which
is defined for a given set of N entities S ¼ {s1, : : : , sN}, by an application Y: S ! B such
that si 7! Y si
ð Þ, Y si
ð Þ ¼ li; ui


, B is the set of intervals of an underlying set O # R, In
other words, the value of the interval-valued variable Y for each si [ S is usually defined
by the lower and upper bounds li and ui of Ii ¼ Y(si), In order to estimate the hedonic
regression model [1] with rij ¼ lij; uij


, we used the Symbolic Data Analysis approach
(SDA) which provides a framework for the representation and analysis of data that
comprehends inherent variability (Brito 2014).
Several approaches have been introduced for estimating regression models with
symbolic interval data, Billard and Diday (2000) proposed a first approach which
consisted in fitting a regression line to the centres of the intervals or midpoints defined as
cij ¼ (lij þuij)/2. Therefore, model (1) can be expressed as:
lncij ¼
X
M
j¼1
ajAjCþ
X
K
k¼1
bkCikC þ 1ijC
ð2Þ
Table 2.
Descriptive statistics based on averages computed for the 20 Italian regions. Italy, 2017.
Conservation status
Total
Excellent
Normal
Detached
Min
Max
Min
Max
Min
Max
Min
2.51
3.44
3.00
3.00
2.50
3.42
Max
6.04
8.14
10.09
13.00
6.60
8.97
Standard Deviation
2.04
2.78
2.42
3.39
1.86
2.55
Well-finished
Min
1.86
2.43
2.46
3.04
1.68
1.23
Max
5.59
7.46
9.41
12.58
5.24
7.00
Standard Deviation
2.09
2.94
2.54
3.76
1.96
2.73
Economic
Min
1.25
1.67
1.62
2.18
1.23
1.66
Max
7.50
11.65
6.30
7.97
6.30
7.97
Standard Deviation
1.59
2.13
1.97
2.53
1.54
2.08
Total
Min
1.64
2.16
2.51
3.10
2.51
3.10
Max
5.29
7.22
9.18
12.28
9.18
12.28
Standard Deviation
2.00
2.73
2.44
3.46
1.85
2.53
Benedetti et al.: Sub-National Spatial Price Indexes for Housing
67


===== PAGE 13 (score 113, numeric-paren 92, table-words 1, var-words 0) =====

apartments. In this way useful information may be provided to economic and social policy
makers. The results of the estimates are reported in Tables 3A and 3B respectively. The
mid-point included in model estimation (2) is an arithmetic mean of the min and max
rental values. Estimation results of model (2) computed with the geometric mean are not
significantly different from the estimated results obtained using the arithmetic mean.
Table 3A.
Hedonic regional models for all dwelling types with different territorial classifications, first semester
(january–june), year 2017.
Hedonic model with districts
Hedonic model with
homogeneous zones
Model
[2]
Model
[3]
Model
[2]
Model
[3]
Midpoint
Min
Max
Midpoint
Min
Max
Regions (ref. Italy)
North
Piemonte
20.079
20.082
20.075
20.065
20.070
20.061
(0.005)
(0.005)
(0.005)
(0.005)
(0.005)
(0.005)
Valle d’Aosta
0.216
0.240
0.197
0.227
0.251
0.209
(0.022)
(0.022)
0.023
(0.022)
(0.022)
(0.022)
Lombardia
0.173
0.191
0.159
0.174
0.192
0.161
(0.005)
(0.004)
(0.005)
(0.004)
(0.004)
(0.004)
Liguria
0.426
0.436
0.419
0.423
0.433
0.416
(0.008)
(0.008)
(0.008)
0.007
0.007
(0.008)
Trentino-
0.412
0.422
0.403
0.415
0.425
0.406
Alto Adige
(0.008)
(0.008)
(0.008)
(0.008)
(0.008)
(0.008)
Veneto
0.102
20.176
20.174
20.176
20.176
20.174
(0.009)
(0.009)
0.010
(0.009)
(0.009)
(0.009)
Friuli-Venezia
0.020
20.022
0.048
0.028
20.014
0.056
Giulia
(0.009)
(0.009)
(0.009)
(0.009)
(0.009)
(0.009)
Emilia-
0.222
0.235
0.213
0.218
0.231
0.209
Romagna
(0.007)
(0.007)
(0.007)
(0.007)
(0.007)
(0.007)
Centre
Toscana
0.435
0.436
0.435
0.433
0.434
0.432
(0.007)
(0.007)
(0.007)
(0.007)
(0.007)
(0.007)
Umbria
20.071
20.074
20.068
20.080
20.083
20.078
(0.013)
(0.013)
(0.013)
(0.013)
(0.013)
(0.013)
Marche
0.024
0.021
0.026
0.024
0.022
0.027
(0.008)
(0.008)
(0.008)
(0.008)
(0.008)
(0.008)
Lazio
0.387
0.388
0.386
0.386
0.388
0.385
(0.008)
(0.008)
(0.008)
(0.008)
(0.008)
(0.008)
Abruzzo
20.179
20.210
20.157
20.182
20.212
20.159
(0.008)
(0.008)
(0.008)
(0.008)
(0.008)
(0.008)
South and Islands
Molise
20.308
20.351
20.276
20.284
20.328
20.251
(0.023)
(0.023)
(0.023)
(0.023)
(0.023)
(0.023)
Campania
20.174
20.168
20.179
20.195
20.188
20.201
(0.005)
(0.005)
(0.005)
(0.005)
(0.005)
(0.005)
Puglia
20.122
20.104
20.134
20.150
20.131
20.163
(0.008)
(0.008)
(0.008)
(0.008)
(0.008)
(0.008)
Benedetti et al.: Sub-National Spatial Price Indexes for Housing
69


===== PAGE 14 (score 81, numeric-paren 60, table-words 3, var-words 1) =====

Table 3A.
Continued
Hedonic model with districts
Hedonic model with
homogeneous zones
Model
[2]
Model
[3]
Model
[2]
Model
[3]
Basilicata
20.639
20.623
20.651
20.626
20.611
20.637
(0.012)
(0.012)
(0.012)
(0.012)
(0.012)
(0.012)
Calabria
20.362
20.359
20.364
20.372
20.369
20.374
(0.008)
(0.008)
(0.008)
(0.007)
(0.007)
(0.008)
Sicilia
20.333
20.346
20.323
20.336
20.349
20.326
(0.007)
(0.007)
(0.007)
(0.007)
(0.007)
(0.007)
Sardegna
20.149
20.147
20.151
20.135
20.132
20.136
(0.010)
(0.010)
(0.010)
(0.010)
(0.010)
(0.010)
Type of dwelling
(ref. well-finished apartment)
Economic
20.175
20.176
20.174
20.175
20.176
20.174
apartment
(0.004)
(0.004)
(0.004)
(0.004)
(0.004)
(0.004)
Detached
0.125
0.130
0.121
0.130
0.136
0.126
(0.004)
(0.004)
(0.004)
(0.004)
(0.004)
(0.004)
Status of the dwelling
(ref. Normal )
Excellent
0.194
0.226
0.170
0.196
0.228
0.172
(0.005)
(0.005)
(0.005)
(0.005)
(0.005)
(0.005)
District (ref. Central )
No
No
No
Semi-central
0.326
0.316
0.333
(0.007)
(0.007)
(0.007)
Outskirts
0.116
0.114
0.118
(0.005)
(0.005)
(0.005)
Suburban
0.031
0.033
0.030
(0.005)
(0.005)
(0.005)
Rural and
20.255
20.255
20.255
extra-urban
(0.005)
(0.005)
(0.005)
Homogeneous
zones
No
No
No
Yes
Yes
Yes
Intercept
1.058
1.058
1.374
1.189
1.020
1.332
(0.004)
(0.004)
(0.004)
(0.004)
(0.004)
(0.004)
Correlation
0.975
0.974
Breusch-Pagan test
51565
51467
Adjusted R2
0.463
0.478
0.478
0.485
0.471
0.471
RMSE
0.375
0.374
0.381
0.368
0.367
0.372
AIC
47.719
267.210
267.210
45.462
269.667
269.667
BIC
47.977
266.644
266.644
45.969
268.564
268.564
N
54,214
54,214
54,214
54,214
54,214
54,214
Source: our elaboration from OMI data 2017.
Note: standard error in brackets; all the coefficients are significant at p , 0.01.
Journal of Official Statistics
70


===== PAGE 15 (score 139, numeric-paren 114, table-words 1, var-words 0) =====

Table 3B.
Hedonic regional models for economic apartments with different territorial classifications. First
semester (january–june), year 2017.
Hedonic model with districts
Hedonic model with
homogeneous zones
Model
[2]
Model
[3]
Model
[3]
Model
[2]
Model
[3]
Model
[3]
Midpoint
Min
Max
Midpoint
Min
Max
Regions (ref. Italy)
North
Piemonte
20.044
20.064
20.028
20.033
20.054
20.016
(0.013)
(0.013)
(0.013)
(0.013)
(0.013)
(0.013)
Valle d’Aosta
0.146
0.174
0.125
0.161
0.189
0.141
(0.038)
(0.038)
(0.038)
(0.038)
(0.038)
(0.038)
Lombardia
0.196
0.213
0.183
0.190
0.207
0.178
(0.013)
(0.013)
(0.013)
(0.013)
(0.013)
(0.013)
Liguria
0.452
0.445
0.457
0.450
0.443
0.455
(0.016)
(0.016)
(0.016)
(0.016)
(0.016)
(0.016)
Trentino
0.673
0.691
0.659
0.641
0.666
0.623
Alto Adige
(0.177)
(0.179)
(0.178)
(0.175)
(0.175)
(0.175)
Veneto
0.238
0.246
0.153
0.237
0.250
0.158
(0.026)
(0.026)
(0.026)
(0.026)
(0.026)
(0.026)
Friuli Venezia
0.020
20.027
0.052
0.031
20.017
0.064
Giulia
(0.019)
(0.019)
(0.019)
(0.019)
(0.019)
(0.019)
Centre
Emilia-
0.213
0.223
0.205
0.211
0.221
0.203
Romagna
(0.016)
(0.016)
(0.016)
(0.016)
(0.016)
(0.016)
Toscana
0.390
0.399
0.384
0.389
0.397
0.382
(0.017)
(0.017)
(0.017)
(0.017)
(0.017)
(0.017)
Umbria
20.070
20.065
20.074
20.076
20.071
20.079
(0.027)
(0.027)
(0.027)
(0.027)
(0.027)
(0.027)
Marche
0.014
0.012
0.016
0.016
0.013
0.018
(0.016)
(0.016)
(0.016)
(0.016)
(0.016)
(0.016)
Lazio
0.433
0.427
0.438
0.436
0.430
0.441
(0.018)
(0.018)
(0.018)
(0.018)
(0.018)
(0.018)
Abruzzo
20.217
20.249
20.193
20.217
20.249
20.193
(0.016)
(0.016)
(0.016)
(0.016)
(0.016)
(0.016)
South and Islands
Molise
20.360
20.402
20.329
20.339
20.382
20.307
(0.035)
(0.035)
(0.035)
(0.035)
(0.035)
(0.035)
Campania
20.212
20.214
20.211
20.230
20.231
20.230
(0.013)
(0.013)
(0.013)
(0.013)
(0.013)
(0.013)
Puglia
20.140
20.117
20.156
20.158
20.135
20.175
(0.016)
(0.016)
(0.016)
(0.016)
(0.016)
(0.016)
Basilicata
20.734
20.720
20.743
20.724
20.712
20.733
(0.020)
(0.020)
(0.020)
(0.020)
(0.020)
(0.020)
Calabria
20.368
20.363
20.371
20.373
20.368
20.375
(0.015)
(0.015)
(0.015)
(0.015)
(0.015)
(0.015)
Sicilia
20.415
20.427
20.405
20.419
20.431
20.409
(0.014)
(0.014)
(0.014)
(0.014)
(0.014)
(0.014)
Sardegna
20.215
20.212
20.218
20.193
20.190
20.197
Benedetti et al.: Sub-National Spatial Price Indexes for Housing
71


===== PAGE 16 (score 48, numeric-paren 30, table-words 5, var-words 0) =====

No evidence of functional form misspecification was observed from the Ramsey test
(1969). The estimation results of models (2) and (3) at regional level for all the dwelling types
are reported in Table 3A by considering two model specifications. The first three columns
refer to districts, while the last three columns refer to homogeneous territorial zones.
As regards the estimation results, the best fitting-models are obtained when homogeneous
zones are included. Moreover, model (3) is preferable to model (2) as it allows for the
correlation between 1ijL and 1ijU and the estimation of the full variance–covariance matrix
of the coefficients. As shown in the last rows of Table 3, the correlation of the residuals in the
minimum and maximum rent equations is equal to 0.975. We can reject the hypothesis that
Table 3B.
Continued
Hedonic model with districts
Hedonic model with
homogeneous zones
Model
[2]
Model
[3]
Model
[3]
Model
[2]
Model
[3]
Model
[3]
(0.020)
(0.020)
(0.020)
(0.020)
(0.020)
(0.020)
Status of the dwelling
(ref. Normal)
Excellent
0.193
0.246
0.153
0.198
0.250
0.158
(0.012)
(0.012)
(0.012)
(0.012)
(0.012)
(0.012)
District (ref. Central)
Semi-central
0.312
0.305
0.317
(0.012)
(0.012)
(0.012)
Outskirts
0.117
0.116
0.117
(0.008)
(0.008)
(0.008)
Suburban
0.009
0.010
0.008
(0.009)
(0.009)
(0.009)
Rural and
20.294
20.295
20.293
extra-urban
(0.008)
(0.008)
(0.008)
Homogeneous
zones
No
No
No
yes
yes
yes
Intercept
1.078
0.906
1.222
1.046
0.876
1.189
(0.011)
(0.011)
(0.011)
(0.011)
(0.011)
(0.011)
Correlation
0.978
0.978
0.977
0.977
Breusch-Pagan
test
15,345
15,345
15,322
15,322
Adjusted R2
0.457
0.463
0.449
0.477
0.482
0.470
RMSE
0.373
0.376
0.374
0.365
0.369
0.366
AIC
13.873
222.242
222.242
13.275
222.888
222.888
BIC
14.080
221.781
221.781
13.698
221.958
221.958
N
16,040
16,040
16,040
16,040
16,040
16,040
Source: our elaboration from OMI data 2017.
Note: standard error in brackets; all the coefficients are significant at p , 0.01.
Journal of Official Statistics
72


===== PAGE 17 (score 18, numeric-paren 0, table-words 7, var-words 4) =====

this correlation is zero on the basis of the Breush-Pagan test of independence results. The
correlation is strong, so the efficiency gains with the SUR estimation is great (Cameron and
Trivedi 2009).
Table 3A shows different distributional patterns as rental prices vary greatly across the
Italian regions, thus confirming territorial heterogeneity. The lowest rental prices were
observed for Basilicata and Molise while the highest rental prices were observed in
Trentino Alto-Adige and Liguria (both for midpoint and minimum- maximum models).
As expected, the localization factor plays an important role in determining rental prices.
However, it is interesting to note that the rents of houses located in central areas of the
municipalities cost 11.4% – 11.8% less than the rents of dwellings on the outskirts as shown in
Table 3A while houseslocatedinrural areascost 25.5% less thanhouseslocatedin centralareas
of the municipalities. This effect may be due to the heterogeneity of houses included in the
district classification which does not take into account the socio-economic characteristics of the
territorial area. More specifically, the district variable captures an average effect of the
locational characteristics since it is obtained by considering rental prices related to all the
municipalities included in each region. Yet, Italy is characterized by the presence of numerous
small municipalities in which the rent values are smaller than those observed in the big
municipality and the more expensive houses (well-finished dwelling and detached houses) are
located in semi-central and outskirts districts. Contrastingly, wehave found that Italian regional
capitals, such as Florence (Tuscany), Rome (Lazio) or Milan (Lombardy), show a different
behaviour of housing rental prices with more expensive houses located in the central districts.
When considering a detailed specification for the localization factors (OMI zones), the
effects of type-of-dwelling and property status do not vary: detached houses cost 13.6%
and 12.6% more than rented well-finished apartments, while excellently conserved houses
cost 22.8% and 17.2% more than those with normal conservation status.
Table 3B reports hedonic regional model estimation results for economic apartments by
using models (2) and (3) specifications. As regards the estimation results, the best fitting-
models are obtained when homogeneous zones are included, and model (3) is preferable to
model (2). The number of observations is equal to 16,040 and estimated results confirm
high territorial heterogeneity across the Italian regions, highlighting disparities between
the Northern and Southern regions. As regards the localization factors, the coefficients
associated to the districts dummy continue to assume the same signs but are lower than the
coefficients estimated in Table 3A for all the dwelling type thus confirming the influences
of houses rental prices in small municipalities.
Table 4 reports the SPHIRs obtained when homogeneous territorial zones are included
in the models (2) and (3) if we consider the 20 Italian regions and compare every region to
the overall mean (Italy ¼ 100). The SPHIRs are calculated both for all dwelling type and
the economic apartments.
On observing regional models reported in Table 4, our results highlight high territorial
heterogeneity and confirming significant rental price differences across the Italian regions
thus demonstrating that rents are higher in the Northern-Central regions than in the South in
respect to the whole of Italy. Basilicata proved to be the less expensive region for rental
properties while Tuscany, Liguria while the highest rents were found in Trentino-Alto
Adige. Significant differences were observed in the estimated SPHIRs, when all of the
dwelling type were included with the exception of the estimation for economic apartments.
Benedetti et al.: Sub-National Spatial Price Indexes for Housing
73


===== PAGE 18 (score 17, numeric-paren 0, table-words 2, var-words 2) =====

As regards economic apartments, the SPHIRs estimated for Trentino-Alto Adige are not
significant due to the reduced number of observations. In this case, the SPHIRs estimated
for Veneto and Lazio are higher than the SPHIRs estimated for all types of dwelling, while
the SPHIRs estimated for Toscana, Abruzzo, Molise, Sicilia and Sardegna are lower than
the SPHIRs estimated for all dwelling types. These findings may be influenced by two
contrasting effects which emerge when estimating SPHIRs at regional level: the inclusion
of tourist destinations characterized by higher rents and small municipalities with lower
rents. Taking into account the various housing characteristics provided by the OMI data
set, it is clear that the role played by tourism is stronger than the effect of the prevalence of
low-cost houses with economic apartments.
In order to explore the territorial heterogeneity of rents within each Italian region,
Figures 1–4 show sub-national SPHIRs estimate with model (3) if we consider
homogeneous territorial zones for the Italian regions (Italy ¼ 100).
Figures 3 and 4 show sub-national SPHIRs with the corresponding intervals at 95%.
Although most regions have small confidence intervals, few regions have overlapping
confidence intervals. Overlapping confidence intervals were observed between Lombardia
and Liguria SPIHRs and Calabria and Sicilia SPIHRs for the model estimated for all
dwelling types and for the economic apartments. With the aim of making multiple
Table 4.
Regional SPIHRs (Italy ¼ 100), year 2017.
All dwelling types
Economic apartments
Model [2]
Model [3]
Model [2]
Model [3]
Region
Midpoint
Min
Max
Midpoint
Min
Max
North
Piemonte
93.72
93.28
94.09
96.76
94.70
98.37
Valle d’Aosta
125.48
128.58
123.26
117.52
120.78
115.18
Lombardia
119.00
121.12
117.45
120.98
122.99
119.48
Liguria
152.73
154.18
151.58
156.87
155.81
157.66
Trentino-Alto Adige
151.39
152.95
150.06
Veneto
110.09
83.82
84.02
126.69
128.45
117.14
Friuli-Venezia Giulia
102.83
98.61
105.81
103.15
98.28
106.59
Emilia-Romagna
124.37
126.00
123.23
123.43
124.76
122.52
Centre
Toscana
154.13
154.31
154.08
147.48
148.78
146.58
Umbria
92.29
92.01
92.53
92.66
93.14
92.37
Marche
102.45
102.19
102.70
101.59
101.35
101.82
Lazio
147.11
147.34
147.01
154.65
153.66
155.47
Abruzzo
83.36
80.89
85.26
80.49
77.94
82.45
South
Molise
75.31
72.05
77.80
71.27
68.28
73.55
Campania
82.29
82.90
81.80
79.43
79.36
79.46
Puglia
86.09
87.72
84.94
85.38
87.33
83.98
Basilicata
53.47
54.30
52.89
48.48
49.09
48.06
Calabria
68.96
69.16
68.83
68.90
69.20
68.70
Sicilia
71.44
70.53
72.15
65.79
64.96
66.45
Sardegna
87.41
87.62
87.26
82.41
82.73
82.15
Source: our elaboration from OMI data 2017.
Journal of Official Statistics
74


===== PAGE 19 (score 0, numeric-paren 0, table-words 0, var-words 0) =====

Source: our elaboration from OMI data 2017 
SPIHRs on minimum values
Piemonte
ta/Vallée d'Aoste
Lombardia
Trentino−Alto Adige/Südtirol
Veneto
Friuli−Venezia Giulia
LiguriaEmilia−Romagna
Toscana
Umbria
Marche
Lazio
Abruzzo
Molise
Campania Puglia
Basilicata
Calabria
Sicilia
Sardegna
40
60
80
100
120
140
160
180
200 40
60
80
100
120
140
160
180
200
SPIHRs on maximum values
Piemonte
ta/Vallée d'Aoste
Lombardia
Trentino−Alto Adige/Südtirol
Veneto
Friuli−Venezia Giulia
LiguriaEmilia−Romagna
Toscana
Umbria
Marche
Lazio
Abruzzo
Molise
Campania Puglia
Basilicata
Calabria
Sicilia
Sardegna
Fg. 1.
SPIHRs for Italian regions for all dwellings types, year 2017.
Source: our elaboration from OMI data 2017 
SPIHRs on minimum values
40
60
80
100
120
140
160
180
200 40
60
80
100
120
140
160
180
200
SPIHRs on maximum values
Piemonte
ta/Vallée d'Aoste
Lombardia
Trentino−Alto Adige/Südtirol
Veneto
Friuli−Venezia Giulia
LiguriaEmilia−Romagna
Toscana
Umbria
Marche
Lazio
Abruzzo
Molise
Campania Puglia
Basilicata
Calabria
Sicilia
Sardegna
Piemonte
ta/Vallée d'Aoste
Lombardia
Trentino−Alto Adige/Südtirol
Veneto
Friuli−Venezia Giulia
LiguriaEmilia−Romagna
Toscana
Umbria
Marche
Lazio
Abruzzo
Molise
Campania Puglia
Basilicata
Calabria
Sicilia
Sardegna
Fig. 2.
SPIHRs for Italian regions for economic apartments, year 2017.
Benedetti et al.: Sub-National Spatial Price Indexes for Housing
75