BEING AN HOMELESS IN MILAN:
A DESCRIPTIVE ANALYSIS
Michela Braga (University of Milan)
joint with Lucia Corno (Bocconi University)
Workshop "Social Minima"
fRDB - Milan, January 12th, 2009
MOTIVATION
Homelessness is a public policy issue in many
developed countries
…but the lack of reliable data on this population limits
effective strategies to prevent and reduce the
phenomenon and creates no incentive for academic
research in economics
OUTLINE






Paper contribution
The existing surveys
Methodology
 Count
 Interviews
Results
 Count: number/localization
 Interviews: refuse - answer rate
 Descriptive statistics (disaggregated by street/shelter/slums)
• Demographics
• Labor/income
• Social links
• Help
How to ameliorete data collection
Current research
PAPER CONTRIBUTION

Quantitative and qualitative data collection:



First Census of homeless in Milan
=> count and localization
Extensive data collection on different aspects
=> questionnaire
Are homeless people different from the general
population? If yes, in which dimensions?
WHY IS IT IMPORTANT?

Information on the number and characteristics of the homeless
is necessary for program planning

Quantitative and qualitative data are necessary to quantify economic
resources to reduce homelessness and to prevent it with policies

Baseline survey for further studies => program evaluation

Cross countries analysis: gap between Italian and international
research:
 In US, systematic data collection year by year starting from
the early 80’s
 In Europe some attempts have been made
…but in a non systematic way
 No data available in Italy
THE EXISTING SURVEYS

Only few countries
homelessness

provide
official
statistics
on
S – NIGHT APPROACH using PUBLIC PLACES METHOD:


U.S. Census Bureau large-scale effort in 1990 to count homeless people
at shelters and selected street sites
In Australia homeless census started in 1996 and takes place every 5
years

HOMELESS MANAGEMENT INFORMATION SYSTEM: U.S.
Department of Housing and Urban Development (HUD) requires
counts every two years on a national sample of 80 communities in
different geographical areas during a given period using a service
based enumeration

CAPTURE RECAPTURE APPROACH: for street homeless who
tend to not use shelters
METHODOLOGY
Procedure in two steps integrating different methodologies

COUNT
Point in time survey using the S - Night
approach (Shelter and Street Night) full
census of the whole city



INTERVIEW
Costs: monetary, human, time
Benefits: accuracy, limit under estimates
Extensive and representative survey in the
following days

Trade off between accuracy of the data
collection and loss of observations
THE COUNT
January 14th, 2008
POPULATION DEFINITION
All individuals that in the given night reside in

places not meant for human habitation, such as cars,
trucks, parks, doorways, sidewalks, stations, airports
(unsheltered homeless);

emergency shelters (sheltered homeless);

people living in disused areas/shacks/slums.
THE COUNT

City divided in 66 sufficiently small census blocks







Pre established itinerary to be followed with a complete list of all
streets in the census block
Localization of 5 headquarters to distribute materials to volunteers
(torches, hot tea, etc)
Informing the homeless for the next day interviews with a flyer
Collect information on the exact localization




Reduce risk of double count (3/4 hours for each block)
Simultaneous full census of the whole city
After 10 p.m.
Necessary for the survey
Detection of some observable characteristics (sex, average age, place )
Collect the lists of names in each shelter
Detection of disused areas and cross-check of their dimensions with
previous control
Partire da
Example: Area N. 9
girare a sinistra in
tornando controllare
girare a sinistra in
girare a sinistra in
tornando controllare
girare a sinistra in
girare a sinistra in
girare a sinistra in
girare a sinistra in
girare a sinistra in
girare a sinistra in
girare a sinistra in
girare a sinistra in
andare in
tornando girare in
girare a sinistra in
girare a destra in
girare a destra in
tornando girare a destra in
continuare in
tornando girare in
continuare in
tornare in P.zza Conciliazione
P.le Baracca
C.so Magenta
poercorrere il lato sinistro
via Aur. Saffi
P.zza Giovane Italia
C.so Magenta
via Ruffini
P.zza S.M delle Grazie
C.so Magenta
via Caradosso
via Sassi
P.zza Virgilio
via Metastasio
C.so Magenta
via Monti
via Carducci
via Leopardi
via Carducci
Pzza Cadorna
via Boccaccio
controllare P.zza Conciliazione
via Bazzoni
percorrerla in ambo i sensi
P.zza Tommaseo
controllare aiuole panchine
via Petrarca
via Mascheroni
percorrerla in ambo i sensi
via Rovani
via Sebeto
via Mascheroni
via Tamburini
via Tasso
percorrerla in ambo i sensi
via Tamburini
via Pontebba
via Tamburini
via XX Settembre
controllare tutte le corsie con le aiu
tornare in P.zza Conciliazione
percorrere il lato mancante di
girare a sinistra in
girare a sinistra in
continuare fino a
via Boccaccio
via Gioberti
via Boccaccio
Piazzale Cadorna
Parco Sempione
THE HOMELESS POPULATION
STREET

408 individuals
SHELTER

1152 individuals
DISUSED
AREAS

2300 adults
Total adult population: 3863
Street homeless
Sheltered
homeless
THE SURVEY
January 15th, 16th, 19th 2008
SAMPLING

Sampling procedure:



Street: all population
Shelter: Random sample proportional to the shelter
dimension. Over – sampling for the small ones and under –
sampling for the big ones
Disused areas: Stratified random sample according
 City administrative division (9 areas)
 Official area classification (authorized, non authorized,
shacks, abandoned buildings, disused areas, ride men);
 Dimension: small (n≤30), medium (30<n<100) and big
(n≥ 100)
THE SURVEY

Few interviewers (75) to minimize answer bias and to exploit the
learning by doing effect;

Interviewers trained



to produce accurate and complete questionnaire
to approach the homeless
to avoid risky situations

2 volunteers/assistants for each interviewer;

Voucher to avoid time consuming interviews “Ticket Service”;

Questionnaires in different languages (IT, EN, RUM);

Average length of each questionnaire: 30’
THE HOMELESS POPULATION

STREET
408 individuals: census - 34.5% interviewed




SHELTER

1152 individuals, sample 500 - 84% of the
sampled interviewed



DISUSED
AREAS

12% refusal rate
15% not found
16.4% sleeping
21% not found
2% refusal rate
6.7% not found
7.3% no time
2300 adults, sample 525 - 66.5% of the
sample interviewed

33.5% no time
Total adult population: 3860
Final Sample: 910 homeless
Socio – demographic characteristics
% Females
% Italians
Street
10
56
Shelters
16
40
Disused areas
49
11
The countries of origin are in line with those found in the general population:
European (especially Romania), African (Tunisia, Morocco, Egypt), Asian
…as expected, especially new immigrants have no house
Immigration year
120
100
80
60
40
20
2007
2006
2005
2004
2003
2002
2001
2000
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
1987
1986
1970
-198
0
1980
-198
5
0
<1 97
0

Different from the general population, the homeless
are especially men (72% vs. 48%) and immigrants
(68% vs 5.8%)
… but the distribution over sex and nationality varies
significantly among the three sub samples
<1 96
0

Socio – demographic characteristics



Current civil status is significantly different from the one found in the general
population
 HL: 32.4% married, 35.4% single, 4.1% widow, 18.9% divorced, 8% other
 GP.:50.4% married, 40,5% single, 7.7% widow, 1.5% divorced
High incidence of mortality in their kids and parents (especially for those in the street)
Family as insurance against adverse shocks
Street
Shelters
Slums
Panel A: Marital Status
Widow/er
9.52
3.54
2.29
Married
8.93
21.46
57.02
Separated/Divorced
29.76
28.3
2.87
Single
44.64
39.39
25.79
Other
4.17
6.84
11.46
Don't answer
2.98
0.47
0.57
Children
47.02
50
68.77
At least 1 child dead
10.13
4.25
5.06
35.06
46.75
28.14
44.41
17.7
24.22
Panel B: Children
Panel C: Parents
Mother dead1
Father dead1
Age

Homelessness affects adults in the central part of their life (average age 39.9)
=> failures in individual life projects (lack/loss job, family relationships,
divorces..)
…but the total population is spread across all age groups

The homeless population is a little bit younger than general population (42.6)
for the high incidence of immigrants. All categories are older than in the
general population
 HL: Italian M=51.1 Foreign M=35 Italian F = 45.6 Foreign F=35.2
 GP: Italian M=41.6 Foreign M=30.4 Italian F = 44.5 Foreign F=31.3

Average age is higher among street homeless (49) than among sheltered
homeless (43). Population young in disused areas (30.7) as in general
population (30.9 years)

Differently from the general population males are 4 year older than females
Education
All sample
None
Elementary school
Middle school
High school
University




14.45
21.68
33.16
25.19
5.53
Italian
8.88
29.28
39.47
19.41
2.96
Foreign
17.11
18.05
30.14
27.94
6.75
Street
10.71
18.45
34.52
30.36
5.95
Shelter
6.84
17.45
34.43
32.78
8.49
Disused
areas
25.5
28.37
30.95
13.47
1.72
General
population
6.8
26.4
31.7
27.2
7.9
Education distribution is in line with the one found in the general
population
Higher proportion of people with no education
More educated people tend to stay in shelter
As in the general population, on average, immigrants are more educated
than native born
Native have 8.2 years of education
 Immigrants have 9.7 years of education
…but the higher education level reflects their age structure

45
40
35
30
25
20
15
10
5
0
O
th
e
r
um
Le
en
ga
ts
lp
ro
bl
em
s
ily
Fa
m
D
oc
of
Im
m
ss
/la
ck
Lo
ig
ra
re
tio
la
n
tio
D
is
ns
ab
hi
ili
ps
t ie
s/
ill
D
ru
ne
g/
Fr
ss
A
e
e
lc
P
C
ho
ol
ho
iti
ol
ic
c
D
U
al
e
na
ep
r
ea
en
bl
e
so
de
to
ns
cy
pa
/C
y
iv
re
il
nt
w
ar
/m
or
tg
ag
C
es
on
vi
nc
t io
ns
G
am
bl
in
g
Immigrants_Milan data
Italians_Milan data
SF data
jo
b
%
FIRST REASON FOR
HOMELESSNESS


Unemployment is among the most cited causes of homelessness (consistent with
SF data) together with familiar problems
 Italians: family relationships (35.1%), loss of job (21%), drug/alcohol
dependency (9.2%), previous convictions (7.5%), eviction (5%), free choice
(8%)
 Immigrants: immigration/language problems/documents(27%), loss of job
(17.3%), family relationships (8.8%)
Failure in life project => crucial to design adequate policies for social inclusion
HOMELESSNESS AND PRISON
30%
70%
Prison After Homelessness
Prison Before Homelessness

High rate of criminality with respect to the general population


About 30% have been in prison at least once (39% of Italians and 23% of
immigrants)
70% of whom spent a period in prison after becoming homeless and 30%
of whom have been in prison before it
LABOR MARKET and
HOMELESSNESS

45% of the population were working before loosing the house
Job condition and loss of house
Employed
Not employed
Don't Answer

All sample
Total
Male
Female
44.64
51.21
27.38
54.6
47.89
72.22
0.76
0.9
0.4
Street
Total
Male
Female
80.14
82.68
57.14
17.02
14.96
35.71
2.84
2.36
7.14
Shelters
Total
Male
Female
47.41
48.18
43.28
52.59
51.82
56.72
Disused areas
Total
Male
Female
26.93
34.83
18.71
72.21
63.48
81.29
0.86
1.59
Possible to find a job being on street
Job found after loosing the house
All the sample
Sub samples
Total
Men
Women Italian
Immigrants
Street
Shelters
Disused
Areas
No
77.11
73.1
84.07
75.68
77.71
83.33
80.72
73.31
Yes
22.89
26.9
15.93
24.32
22.29
16.67
19.28
26.69
LABOR MARKET

Labor force participation is higher compared with the general population
All sample
Street
Shelters
Disused areas

All
74.39
57.14
78.3
77.94
Male
Female
76.8
68.08
59.59
40.91
79.83
70.15
84.83
70.76
Italian
Foreign
59.54
81.48
51.58
64.38
62.57
88.93
65.79
79.42
The 29.3% was employed at the time of the survey. Among unemployed
people the 17% worked during the previous month
 More than half of people are employed in the black market compared with
the 12.1% in the general population
 Only 13% have permanent contract and a significant percentage (20%) has
temporary contract while in the general population the percentages are
65% for permanent and 10% for temporary
All
sample
Permanent contract
Non permanent contract
Italian
13.12
9.3
Foreign
14.8
Street
Shelters
Disused
Areas
9.52
8.94
18.8
22.7
29.07
19.9
7.14
30.89
19.66
58.16
55.81
59.18
64.29
56.91
57.26
Don't know
1.06
2.33
0.51
2.38
0.81
0.85
Don't answer
4.96
3.49
5.61
16.67
2.44
3.42
Don't have a contract/ paid under table
LABOR MARKET

People are employed as low skilled workers, especially as factory workers (33%),
domestic workers, nannies, cleaners (15.3%), bricklayers, carpenters, electricians,
plumbers (9.4%), unskilled service workers (12.9%), cooks/waiters (5.9%)

Unemployed people look for a job through informal channels
All
sample

Street
Shelter
Disused
Areas
Friends/relatives
40.57
41.1
34.73
48.21
Work placement office (municipality)
15.28
8.22
16.41
16.41
Temporary work agency
19.06
10.96
24.81
14.36
Voluntary associations
5.66
2.74
5.73
6.67
Asking directly to firms
3.58
4.11
4.2
2.56
Asking to acquaintances
2.64
8.22
2.67
0.51
Newspapers
2.64
5.48
1.53
3.08
Social assistant/Public services
1.13
Internet
0.94
2.74
1.15
Cooperatives
2.45
2.74
3.05
1.54
Don't know
2.83
6.85
1.53
3.07
Don't answer
3.21
6.85
1.91
3.59
Individual reservation wage is 827 euro/month
2.29
INCOME

Low rate of participation into government program => few individuals receive
social assistance. How to reach the excluded?
First source of income
All
sample
Females
Street
Shelter
Disused
areas
No income
8.09
9.41
4.62
7.14
14.86
0.29
Welfare check
5.85
6.47
4.23
3.57
11.08
0.57
Unemployment benefit
0.64
0.74
0.38
0.6
0.47
0.86
Disabilities Insurance
2.23
2.79
0.77
2.98
3.77
Permanent work
10.85
11.47
9.23
6.55
6.84
17.82
Occasional work
22.02
23.82
17.31
17.86
26.18
18.97
Family/Relatives
13.3
7.65
28.08
4.76
5.42
27.01
Friends
4.57
5.29
2.69
8.33
4.48
2.87
Pension
4.04
4.85
1.92
6.55
6.13
0.29
Savings previous job/rent
1.17
1.47
0.38
0.6
1.89
0.57
Shelter subsidy
0.64
0.44
1.15
0.6
1.18
Church/voluntary association
0.74
0.88
0.38
1.19
1.18
Illegal activities
0.96
1.18
0.38
0.6
1.18
0.86
Secret activity
6.38
5.29
9.23
10.12
3.07
8.62
Don't know
Don't answer

Males
4.89
4.41
6.15
15.48
2.59
2.59
13.62
13.82
13.08
13.1
9.67
18.68
On average weekly income is 151 €. It increases in disused areas (164€) with
respect to street and shelter (140 and 145)
…not below the relative poverty threshold 246.5€ for a two persons hh but
insufficient to afford house expenditures
IN KIND HELP
80
70
60
50
40
30
20
10
0
Total


Italian
Foreign
In kind help Food
Street
Shelter
Clothes Medicines
Dised areas
People receive help especially from catholic associations, non profit organizations,
advocacy groups
Some categories appear disadvantaged
 Is there any distortion in the existing distribution mechanism?
 Are some groups self selected?
 How to reach all needy people?
HOMELESS AND HELP

In-kind help is the main
form of help
Generic help
last year
Yes
Financial help In k ind help
ever
50.77
No, I haven't asked/received anyone
41.65
for help
No, I don't need help

Family as the main
source, followed by
voluntary associations
21.87
63.41
73.74
35.83
5.6
Don't know
0.44
0.39
Don't answer
1.54
4
First source of help
Family
Voluntary associations
Friends
Church/parish
Social Services/Public administration
Employer/Ex-employer
Hospital/Doctor/Naga
Don't answer
Other
Obs.
0.76
35.27
24.03
20.35
8.87
7.36
1.08
1.08
1.08
0.87
462
SOCIAL NETWORKS

Can you tell me name and surname of your first 5 homeless friends?
Distribution of friends
0 links
1 links
2 links
3 links
4 links
5 links
Don't know/Don't answer
Mean
Observations


All Sample
%
49.34
16.15
11.54
4.73
5.38
5.16
7.69
1.09
910
Street
%
28.37
19.86
16.31
6.38
7.09
5.67
16.31
1.53
141
49% do not have any friends
Higher percentage of links for those on the street
Shelter
%
38.57
21.19
12.38
5
5.95
5.71
11.19
1.28
420
Slums
%
70.77
8.6
8.6
3.72
4.01
4.3
0
0.74
349
CRONICALITY vs. IN AND OUT

The 75% of the population never slept in a house after the first night on street



The 52% of the street homeless, the 67% of the sheltered homeless and the
93% of the population of disused areas
On average, the street homeless lost the house 4.3 years ago, the sheltered
homeless 3.3 years ago and disused areas inhabitants 11 years ago
People doing in and out rent a house/room, stay with relatives and parents for
short periods
CURRENT RESEARCH

Relationship between crime and social network by
exploiting a dyadic data structure
=> evidence of fascinating peer effect in the realm of
criminality

Determinants of the labor markets
=> variables affecting labor market behavior are in line
with the underlying theoretical framework of utility
maximization and labor-leisure choice
HOW TO AMELIORATE DATA
COLLECTION

Collect data on a regular basis
 Capture seasonality
 Monitor trends
 Verify efficacy and efficiency
costs/benefits analysis
of
applied

Capture re – capture approach

Exploit available administrative data from all servicers

Multi disciplinary approach
 Economy
 Sociology
 Psychology
policies
=>
POLICY INTERVENTIONS
To design adequate social inclusive
policies it is important to go over the
traditional iconography
Descargar

BEING AN HOMELESS IN MILAN: A DESCRIPTIVE ANALYSIS