Measurement of Poverty and Social Exclusion in Japan

1
Measurement of Poverty and Social Exclusion in
Japan

• Aya K. Abe
• National Institute of Population and Social
  Security Research
• Tokyo, Japan

Second Townsend Memorial Conference, Measuring
Poverty The State of Art, 22-23 January 2010
2
Poverty Rates of OECD countries(Mid 2000s)
Poverty Rates of OECVD Countries (Mid 2000s)
Income measure
Data OECD(2008) Growing Unequal?
3
Political Context

• Upto 2008 Denial stage

4

• Oct. 2009 Announcement of Relative Poverty Rate
• Jan. 18, 2011
• Social Inclusion Special Team established under
  Prime Ministers Office

5
From Research community

• Up to 2000s some qualitative research on
  special risk groups, such as homeless population,
  single-mothers, etc.
• From 2000s Some attempts to calculate extent of
  poverty among general population using large
  surveys (mostly income) (Abe 2005, Komamura 2005,
  etc.)

6
Brief Introduction of my work

• 2003 Necessities Survey (n1350)
• Asking general public what is necessary
• 2003 Social Living Survey (n1520)
• Using above items regarded necessary, asked who
  are deprived of those items some social network
  questions
• 2006 Living Conditions Survey (n584)
• Asked Deprivation and social exclusion more
  emphasis on social exclusion
• Sample limited to one geographical area near
  Tokyo
• 2008 Social Living Survey (n1021)
• 2003 and 2006 survey questions combined, covered
  all areas of Japan
• 2008 Necessities Survey for Children

7
MethodologyDefining Essentials and Identifying
Who is Deprived
Is it essential?
Do you have it?
Yes
Yes
No
No
Do not want it
Cannot afford it
8
Socially Perceived Necessities in Japan
9
DataThe 2003 Necessities Survey

• Sample of 2000 adults (20 years ), randomly
  chosen from residents register all over Japan
• 1350 responses (response rate 67.5)
• For 28 items, asked respondents whether they
  thought it is necessary to live normally in
  Japan

10
of support for items
11
those who think the item is essential
Item JAPAN UK
Toys (that most of other kids have, such as dolls, blocks, soccer ball, baseball etc.) 12.4 84
Bicycle (including second-hand) 20.9 55
At least one pair of shoes (not second-hand) 40.2 94
Clothes (not second-had) 33.7 70
Own books 51.2 89
To be able to go to dentists (including check-ups) 86.1 Australia 94.7
UK question Toys (e.g. dolls, teddies)
Australia Community Understanding of Poverty
and Social Exclusion Survey 2006 in Saunders et
al. (2007) Data Japan Child Necessity Survey
2008 in Abe (2008), UK Office for National
Statistics Omnibus Survey 1999, in Gordon et al.
(2000)
12
supporting items 1st vs. 5th quintile (income)
saying Definitely required 1st quintile vs.
5th quintile
5th quintile (richest)
1st quintile (poorest)
13
supporting items by education
saying Definitely required College grads vs.
Mandatory education only
telephone
College or above (12 )
Dentist
Suits
Mobile
Internet
Wedding
Neighborhood clubs
Junior high school (9 yrs of education)
14
supporting items Women vs. Men
saying Definitely required men vs. women
Microwave
Separate bedroom
Women
fruits
Mobile
Video player
Men
15
supporting items by location
saying Definitely required those living in
village vs. those living in large cities
Living in Village
wedding
Dentist
Multiple bedrooms
Neighborhood clubs
Familys own bathroom
Family trip
Internet
Living in Largest 13 cities
16
supporting items by Age
saying Definitely required Above 70 years
old vs. Below 30 years old
telephone
Doctor
Above 70 years old
Wedding
Neighborhood clubs
Mobile
Video pl.
Internet
Below 30 years old
17
Patterns of Deprivation in Japan
18
Data2003 Social Living Survey

• Sample of 2000 adults (20 years ), randomly
  chosen from residents register all over Japan
• 1520 responses (response rate )

19
(No Transcript)
20
(No Transcript)
21
(No Transcript)
22
(No Transcript)
23
(No Transcript)
24
(No Transcript)
25
Comparison of Japan-Australia Deprivation patterns

• Why Deprivation approach?
• Assumption that most households consist of
  nuclear families- Not the case in Japan -gt raises
  questions re equivalence adjustment
• Presence of multi-generational hh
• Deprivation approach direct measurement of
  living standard, not relying on assumptions on
  resource sharing within families

Slides 25-XX Saunders, Peter Abe, Aya. 2009.
Poverty and Deprivation in Young and Old A
Comparative Study of Australia and Japan.
Poverty and Public Policy, Vol.2, Iss.1, Article
5 (2010).
26
Comparison of Japan-Australia Deprivation patterns

• Australia
• Community Understanding of Poverty and Social
  Exclusion (CUPSE) survey, 2006, Social Policy
  Research Centre (Saunders, Naidoo and Griffiths,
  2007).
• N2,700

• Japan
• The Social Living Survey, 2003, NIPSSR (Abe
  2006)
• N1,520

27
Difficulties

• Do we use the same list of items?
• -gt how do we account for differences in what is
  considered necessary?
• Do we use the items selected using the same
  methodology?
• What UNIT of comparison do we use?

28
(No Transcript)
29
(No Transcript)
30
(No Transcript)
31
(No Transcript)
32
Comparison by Household Type
Similarities

• The number of items differs between JP and AUS,
  thus comparison of the absolute values of MDS or
  of deprivation do not mean much. Instead, we
  need to look at patterns and ordering of family
  types within each country.
• Similarities Sole parents are the most deprived,
  followed by WA singles, WA with children. Least
  deprived are older couples, WA couples w/o
  children, older singles.

33
Comparison of hh types remarkably similar

• ? Elderly single vs. WA single
• (AUS) ltlt
• (JP) ltlt
• ? Elderly single vs. Elderly couple
• (AUS) gtgt
• (JP) gtgt
• ? WA single vs. WA couple w/o children
• (AUS) gtgt
• (JP) gtgt
• ? WA couple with children vs. WA couple w/o
  children
• (AUS) gt
• (JP) gtgt
• ? WA couple with children vs. Lone parents
• (AUS) ltlt
• (JP) ltlt

34
(No Transcript)
35
Overlap analysis

• Le us define Low income Dlt2 to be consistent
  poverty
• Aus Consistent poverty is spread evenly at
  around 8 across all households (except sole
  parent hh).
• Jp the differential is very large across
  different household types strikingly high in
  elderly singles sold parents

36
Conclusion Aus-Jap comparison

• The ranking of poverty as measured by income
  differs between AUS-JP, but it is very similar if
  poverty is measured by deprivation.
• Perhaps deprivation captures the real
  occurrence of poverty which is shared among
  countries?
• From the overlap analysis, consistent poverty is
  more concentrated in Japan.

37
Measuring Social Exclusion in Japan
38
DataThe 2006 Living Conditions Survey (LCS)

• Sample of 1600 adults (20 years ), randomly
  chosen from the residents register in the
  southern Kawasaki City
• Kawasaki is located between Tokyo and Yokohama, a
  part of industrial belt.
• The southern part host many factories, and the
  city received influx of migrant laborers from
  rural sections of Japan.
• 584 responses (response rate 36.5)

39
Survey Concept

• It should capture economic impoverishment not
  only by income, but also by material deprivation
• It should capture how an individual is excluded
  (forced out) from various public constructs
  within a society, e.g. public schemes such as
  public pension and public health insurance,
  public services such as transportation and
  utilities, and public spaces such as libraries
  and sports facilities
• It should capture exclusion from private spheres,
  e.g. lack of social relations (communication with
  others, meeting family obligations, doing
  activities with others) and social networks
  (support in need)
• it should measure degree of individuals
  involvement with society, e.g. social
  participation such as participation in local
  communities (neighborhood organizations, womens
  clubs, PTA, etc.), civic activities (voting,
  political involvement, etc.), and personal
  communities (alumni clubs, sports and hobby
  circles, etc.)

40

• These should enforced lack, rather than
  preference.
• It should not only capture the enforced lack due
  to economic constraints, but also due to other
  constraints (health, family, work, social, etc.),
  and should be able to distinguish them

41
8 dimensions chosen

1. (lack of ) basic human needs,
2. material deprivation,
3. exclusion from systems and services,
4. (lack of) leisure and social participation,
5. inadequate housing,
6. (lack of ) social relation,
7. subjective poverty, and
8. income poverty

42

• Almost all items were asked whether they are
  wanted but cannot be obtained, not wanted (or
  not interested) or are obtained.
• For most of items, the survey also asks the
  reason why that item cannot be obtained
  Economic, family, work, health, other reasons.

43
Example Social Participation

• Very few indicate economic reasons for not being
  able to participate in social activities
• Family/Work reasons is most often cited.

44
Social Exclusion Indexes

• Threshold for determining those who are
  "excluded" are decided by the author so that the
  exclusion rate will be 10 to 20 of the
  respondents.

45
Example Exclusion from Systems

• Almost none stated economic reasons
• Access is most often cited reason.

46
Share of Respondents Excluded, by key social
variables
47
Some Key findings

• (Gender)
• Men show higher rate of exclusion than women
  (lack of social participation and subjective
  poverty)
• Same as PSE (social participation), but even more
  so.
• (Age Group)
• The income poverty shows an U shape curve, but it
  does not seem to translate directly to Lack of
  Basic Needs and Material Deprivation.
• Those at 50s seem to be at higher risk of many
  dimensions of S.E., e.g. Housing, Subjective
  poverty, Lack of Social Participation, Lack of
  Social Relations
• Concurs with the fact that the suicide rate for
  men peaks at age 50-59. 95 of homeless persons
  are men, a half of which are in their 50s.

48

• (Household type)
• Working age single-person households are by far
  the most at risk of social exclusion Basic
  Needs, Material deprivation, housing deprivation,
  and lack of social participation
• (Working Status)
• Not being in the labor force, by itself, does not
  seem to indicate higher risk of social exclusion.
• In fact, housewives and retired persons are at
  lower risk of social exclusion in some dimensions
  (subjective poverty, housing, social partipation)
• However, not-being in the labor force for other
  reasons does indicate higher risk of social
  exclusion in 6 dimensions, including
  non-financial dimensions such as Exclusion from
  Systems and Lack of Social Relations.
  Involuntary detachment from the labor force is
  associated with social exclusion.

49

• (Education Level)
• Low education attainment (up to Junior high
  school age 15, the compulsory education in
  Japan) is a strong link to social exclusion, not
  only for financial dimensions (income poverty,
  lack of basic needs, material deprivation,
  housing deprivation), but also for non-financial
  dimension s (exclusion from systems, lack of
  social relations).
• High education attainment (college) is
  associated with lower risk of social exclusion
  (systems, and basic needs)

50
Social Exclusion and Earlier Disadvantages

• Outside Japan, there are many studies linking
  childhood poverty to adult outcomes.
• However, in Japan, there are very few studies
  connecting earlier life disadvantages and poverty
  and/or social exclusion, since there has not been
  much accumulation of panel data sets.
• The studies using the Japanese Panel Survey of
  Consumers (JPSC) , the only long enough panel
  data set, has shown those who divert from
  standard life course, such as those who divorce
  and who do not marry, are more prone to becoming
  poor (Iwata Nishizawa 2005). However, JPSC
  only covers women in a certain cohort.

51
This Study

• The survey was designed to capture major
  disadvantageous events (Independent variables)
  childhood poverty, divorce, prolonged illness or
  injury, involuntary lay-off.
• Outcome (Dependent) variables include both
  financial poverty as well as social exclusion.
• Control variables are current income, sex, age
  class, single-elderly, current household type
  (has children, single-person household), current
  working status.

52
Results of Regression for Social Exclusion Indexes
53
Key Findings

• (Lay-off)
• Having experience of a lay-off has positive and
  significant effect on current material
  deprivation, housing deprivation, lack of social
  participation, lack of social relations,
  exclusion from systems and subjective poverty,
  even after controlling for current income, age,
  sex, working status, and household type.
• (Divorce)
• Having experience of divorce has positive and
  significant effect on basic needs and housing
  deprivation, even after controlling for current
  marital status (single-person hh).

54
Key Findings

• (Prolonged illness and injuries)
• Having experience of prolonged illness and
  injuries (which caused one to be out of work or
  school for more than one month) has positive and
  significant effect on exclusion from systems.

55

• (Childhood poverty)
• Having experienced childhood poverty (living
  standard at age 15 was very low (1) out of
  scale of 5) has positive and significant effect
  on current lack of basic needs, even after
  controlling for current income, age, household
  type and other disadvantages such as divorce and
  lay-offs.
• The causal relationship is indicated.
• (Control variables)
• Income Negative and significant in all but one
  (systems exclusion), including social relations
  and participation.
• Gender Men are (social relations,
  subjective poverty)
• Age does not seem to have that strong of a
  effect
• Work sattus (Exclusion from systems) ??

56
Conclusions

• Sections of population most vulnerable to income
  poverty is not most vulnerable to social
  exclusion.
• Possible new vulnerable group men in their
  50s.
• Disadantages in earlier stages of life seem to
  exhort influences on some aspects of current
  social exclusion, even after controlling for
  current income, work status, household type, etc.
  
• The catch-phrase of the former PM Abe a society
  in which one can re-challenge DOES NOT seem to
  hold.
• Childhood poverty seems to have irrevocable
  continuing effect on adult well-being not only
  via education and occupation (and thus income),
  but by another path.