Building a Better National Targeting System for Improving Social Safety Net Programs: Indonesian Experience in Shifting from Commodity Subsidies to Targeted Subsidies (EN)
Dr. Bambang Widianto
Executive Secretary to the National Team for the Acceleration of Poverty Reduction
Office of the Vice President
Republic of Indonesia
Semelhante a Building a Better National Targeting System for Improving Social Safety Net Programs: Indonesian Experience in Shifting from Commodity Subsidies to Targeted Subsidies (EN)
The relative position of brazil in regards to a possible sdg on social protec...UNDP Policy Centre
Semelhante a Building a Better National Targeting System for Improving Social Safety Net Programs: Indonesian Experience in Shifting from Commodity Subsidies to Targeted Subsidies (EN) (20)
Building a Better National Targeting System for Improving Social Safety Net Programs: Indonesian Experience in Shifting from Commodity Subsidies to Targeted Subsidies (EN)
1. BUILDING A BETTER NATIONAL TARGETING
SYSTEM FOR IMPROVING SOCIAL SAFETY NET
PROGRAMS:
INDONESIAN
EXPERIENCE
IN
SHIFTING
FROM
COMMODITY
SUBSIDIES
TO
TARGETED
SUBSIDIES
Dr.
Bambang
Widianto
Deputy
for
Social
Welfare
and
Poverty
Allevia;on/
Execu;ve
Secretary
of
THE
NATIONAL
TEAM
FOR
THE
ACCELERATION
OF
POVERTY
REDUCTION
(TNP2K)
September
9,
2014
OFFICE OF THE VICE PRESIDENT
THE REPUBLIC OF INDONESIA
3. Despite a declining trend in poverty rates, this has slowed in recent years
28.28 million people live below the poverty line (March 2014)
4. Poor and vulnerable
communities make up 40% of the population
14.15%
Below the
Monthly Consumption per Capita (IDR)
Population
Source: Susenas (2010)
Poverty
Line
200,000 400,000 600,000 800,000 1,000,000
5. GROWTH IN CONSUMPTION 2008-2012
Poor
29 million
Near-poor
70 million
Middle income
100 million
High income
50 million
Average
Annual increase (%)
± ± IDR 250,00 0/p/month IDR 370,000/p/month ± IDR 750,000/p/month
12%
± IDR 370,000/p/month
40%
± IDR 750,000/p/month
80%
6. FUEL
PRICES
IN
VARIOUS
ASIAN
COUNTRIES
(USD/LITER)
MAY
2013
RON
>
90
RON
<
90
Not
Specified
0.63
0.99
1.01
1.21
1.18
1.53
1.42
1.30
1.29
Indonesia
Indonesia
Myanmar
Thailand
Phillipines
Singapore
Vietnam
Laos
Cambodia
7. FUEL
SUBSIDY
DISTRIBUTION
20%
Highest
20%
Second
Highest
20%
Middle
20%
Second
Lowest
Source:
10.07
6.14
NaTonal
StaTsTc
Office
(BPS),
March
2014
7
2.74
22.03
59.03
5.7
12.9
21.1
46.3
124.0
-‐
20.00
40.00
60.00
80.00
100.00
120.00
140.00
20%
Lowest
Amount
(Triliun
ID
IDRR)
)
(Trillion
DistribuTon
(%)
8. PRESSURE
FROM
INTERNATIONAL
CRUDE
OIL
PRICE
INCREASES
Fuel
and
Electricity
Subsidies
take
Funding
Away
from
Pro-‐poor
Development
Sectors
[in
IDR
trillion]
23.3
186.2
76.3
Energy
Subsidy
[in
IDR
trillion]
36.0
38.8
49.9
50.6
201.6
259.2
300.5
326.1
344.7
86.0
114.2
145.4
184.3
2009
2010
2011
2012
2013
2014
26.5
Fuel
45.0
82.4
165.2
211.9
210.0
246.5
Electricity
49.5
57.6
90.4
94.6
100.0
103.8
206.6
50.1
57.8
44.9
50.6
63.6
82.1
94.5
140.0
255.6
306.5
310.0
350.3
2009
2010
2011
2012
2013
2014
Health
EducaTon
Infrastructure
Sosial
Assistance
Energy
Subsidy
10. BASIC IDEA
SHIFT FROM COMMODITY SUBSIDIES
TO HOUSEHOLD SUBSIDIES
Commodity
subsidies
are
simple
but
unfair.
They
are
not
pro-‐poor.
Have
a
big
impact
on
government
budgets.
Aggregate
poverty
data
is
not
adequate.
Targeted
subsidies
as
the
basis
of
social
assistance:
UncondiTonal
Cash
Transfers
(UCT),
Health
Care
(Jamkesmas),
Student
Aid
(BSM),
Rice
for
the
Poor
(Raskin),
etc.
11. SHIFTING
TO
MORE
TARGETED
PROGRAMS
World
Crude
Oil
Price
Increased
Since
the
Last
15
Years
Fuel
Subsidy
ReducTon
CompensaTon
Program
UncondiTonal
Cash
Transfers
CondiTonal
Cash
Transfer
Rice
for
the
poor
EducaTon
Health
Rural
Infrastructure
Community-‐Based
Development
12. TARGETING OPTIONS
Means
tesTng:
this
requires
high-‐quality
data
that
is
not
available
in
many
countries
and
may
be
expensive
to
put
in
place.
Geographical
targeTng:
transfers
are
provided
to
those
living
in
areas
with
a
high
incidence
of
poverty.
Community-‐based
targeTng:
uses
community
structures
to
idenTfy
the
poorest
members
in
a
community
or
those
eligible,
according
to
agreed
criteria.
Providing
benefits
to
those
recognized
as
belonging
to
a
specific
vulnerable
category
of
the
populaTon.
Self-‐targeTng:
for
example,
in
work
programs
that
offer
a
below-‐market
wage,
based
on
the
logic
that
individuals
choose
to
opt
into
the
program.
13. EXAMPLES OF SPECIFIC
VULNERABLE GROUPS
1. Most
Poor
(Fakir
Miskin)
2. Orphans,
Street
Children
3. Homeless
without
Support
4. Isolated
Tribal
CommuniTes
5. Mentally
Ill
6. Displaced
PopulaTons
14. SELF TARGETING:
KEROSENE CONVERSIONS TO LPG
Government
provides
free
small
bokles
(3
Kg)
of
LPG
to
poor
households,
small
restaurants,
food
vendors
and
other
micro
businesses.
70
60
50
Billion
Litres
59.7
1.5
39.3 36.8
40
30
20
10
0
2005 2008 2009
Fuel
Consump;on
Conversion from
Kerosene to LPG
(Estimation)
16. 30% Only
of poor people
received
Household Consumption (Decile)
Receiving Assistance (%)
17. REVISED
DATA
COLLECTION
METHODOLOGY
Goal: To reduce inclusion and
exclusion errors
Construction of Initial Lists of Targeted Households
Individual
data
from
other
programs
Consulta;ons
with
poor
households
Popula;on
Cencus
2010
Poor
Not
Poor
Beneficiaries
Non-‐Beneficiaries
Ini;al
list
of
targeted
households
18. PROCESS
OF
DEVELOPING
THE
UNIFIED
DATABASE
Data
collec;on
(PPLS
2011)
BPS*
Data
analysis
&
development
of
TNP2K**
PMT
models
Unified
database
Improvements
to
the
Methodology:
-‐
More
households
surveyed
(43%
vs.
29%
in
2008)
-‐
Use
of
census
data
as
a
starTng
point
-‐
Community
involvement
-‐
More
variables
collected
for
beker
poverty
predicTon
-‐
Improvements
to
Proxy
Mean
TesTng
(PMT)
methods
Note:
*
BPS:
NaTonal
StaTsTcs
Office
**
TNP2K:
NaTonal
Team
for
the
AcceleraTon
of
Poverty
ReducTon
19. DATA COLLECTION
Involved 120,000 enumerators
Using initial lists, enumerators surveyed every
individual household and collected information
for variables on their social and economic
status.
Initial list contained “the bottom“ 50% of
households.
Survey results were sent to TNP2K, and then
processed to produce the Unified Database.
The Unified Database contains information
only on the bottom 40% of households.
20. PERCENT OF THE POPULATION WITH
SIMILAR SOCIO-ECONOMIC
CHARACTERISTICS
Exclusion
Error
Includes
24.7
million
households,
or
around
96.4
million
individuals
Includes
15.5
million
households
or
65.6
million
individuals
Inclusion
Error
Includes
5.7
million
households
or
28.6
million
individuals
Near
Poor/
Vulner-‐
able
Poor
60
%
40
%
25
%
11,66%
21. WHICH OF THESE HOUSEHOLDS WILL RECEIVE
SOCIAL ASSISTANCE?
…
due
to
the
number
of
household
members,
the
number
of
dependents
and
the
wife’s
employment
status,
the
household
on
the
right
is
the
real
beneficiary
of
social
assistance.
At
first
glance,
this
household
would
be
the
beneficiary,
BUT
…
22. MANAGING UNIFIED DATABASE
Program
Services
(Opera;on)
Research
Informa;on
System
• Ensure
that
programs
use
the
Unified
Database.
• Provide
technical
support
to
the
programs.
• Ensure
the
validity
of
various
studies
to
improve
targeTng.
• Monitor
&
evaluate
the
use
of
the
Unified
Database.
• PMT
modeling
and
analysis
of
cost-‐effecTveness
for
future
data
collecTon
(presumably
next
in
2014).
TNP2K TARGETING UNIT TASKS:
Œ
Ž
• IT-‐based
management
• Provide
informaTon
extracted
from
the
Unified
Database
through
IT,
media.
23.
24. NATIONAL TARGETING SYSTEMS
USING THE UNIFIED DATABASE
Eligibility criteria
social assistance program
Unified
Database
for
social
assistance
Beneficiary
list
for
Beneficiary
List
of
Beneficiary
List
of
social
assistance
programs
Beneficiary
List
of
Social
ProtecTon
Program
Social
ProtecTon
Program
Social
ProtecTon
Program
Set
by
each
program.
For
example,
for
PKH,
the
criteria
was
set
by
the
Minister
of
Social
Affairs:
extremely
poor
households
with
elementary
school-‐aged
children
or
pregnant
mothers.
Data
by
name
and
address.
Contains
informa;on
on
the
bogom
40%
of
the
popula;on.
Names
and
addresses
of
eligible
beneficiaries
for
social
assistance
programs.
25. Started 2013
25% of households with the lowest socio-economic status
or 15.5 million poor and near-poor households.
For accessing:
BLSM, BSM, Raskin and the JKN card
26. Data Update by Combining
Top Down and Bottom Up
PT. Pos
Households Village Level
Deliberation
Recapitalisation
TNP2K’s Unified Database
27. Online Complaints Service (LAPOR!) with
UKP4
Number of Complaints as of July 2014
Complaints
received
Followed
up
Finished/
complete
29. UNCONDITIONAL CASH TRANSFER
Program
descripTon
and
size:
Each
beneficiary
family
received
IDR
100,000
per
month,
paid
quarterly,
from
October
2005
to
December
2006.
2005-‐2006
program
budget
was
IDR
23
trillion.
2008
program
budget
was
IDR
13
trillion.
In
2013,
the
Government
of
Indonesia
implemented
the
uncondiTonal
cash
transfers
(UCT)
program
for
15.5
million
poor
and
near-‐poor
families,
as
compensaTon
for
inflaTonary
effects
linked
to
fuel
price
increases.
Each
family
received
IDR
150,000
per
month
for
four
months
2013
program
budget
was
IDR
12
trillion
30. Reasons
for
Providing
Cash
Transfer
as
Compensa;on
for
Rising
Fuel
Prices
Recipients
of
cash
transfers
can
benefit
immediately.
Cash
is
easier
for
beneficiaries
when
making
adjustments
in
their
consumpTon
needs.
In
terms
of
programme
implementaTon,
giving
cash
is
more
efficient
and
the
distribuTon
costs
are
cheaper.
31. Fuel
Price
Increases
and
Necessary
Compensa;on
for
the
Poor
Premium
Fuel
Price
Increases
• If
(IDR)
Fuel
Price
Increases
(%)
Baseline
+
Addi;onal
Infla;on
Linked
to
the
Consumer
Price
Index
(pp)1
Baseline
+
Addi;onal
Infla;on
Incurred
by
the
Poor
(pp)
Compensa;on
for
Poverty
Line
Increases
(IDR)
Compensa;on
Amount
per
month
(IDR)
2,000
30.77
1.8
3.861
695,077
115,846
3,000
46.15
3.2
6.864
1,235,692
205,949
4,000
61.54
4.6
9.868
1,776,308
296,051
fuel
prices
rise
by
IDR
3,000
to
total
IDR
9,500,
it
would
be
necessary
to
compensate
+/-‐
IDR
200,000/household/month
for
6
months.
• A
compensaTon
period
of
6
months
is
considered
adequate
because
inflaTon
tends
to
return
to
normal
levels
by
that
point.
32.
33. PT. POS INDONESIA
NO.
DESCRIPTION
NUMBER
1.
Post
Office
Branches
3,892
2.
Mobile
Services
3,062
3.
Cars
and
Motorcycles
10,523
4.
Employees
28,900
5.
Online
Post
Offices
3,500
6.
Delivery
People
9,867
37. POTENTIAL FOR INJURY:
Long queuing times, particularly for the elderly
Queuing
Time
Wai;ng
Times
for
UCT
Beneficiaries,
60+
38. POTENTIAL FOR INJURY:
Long distances to the nearest post office
Distance
from
collecTon
point
(PT.
Pos)
Time to Collection Point (Phase 1)
79.72%
2.21% 1.20%
16.87%
Less than 1 hr 1 - 2 hrs 3 - 5 hrs more than 5 hrs
39. 0
20
40
60
80
100
Rice
Kerosene
Repay
debt
Health
EducaTon
Others
Capital
Gasoline
2nd
payment
UCT
WAS
USED
FOR
BASIC
NECESSITIES
40. UCT
DID
NOT
REDUCE
TOTAL
WORKING
HOURS
Near-Poor and Below
2005 2007 Diff
Household (HH) Head
UCT 39.2 37.7 1.5**
Non-UCT 41.0 39.8 1.2**
Difference -1.8 -2.1 0.3
Spouse
UCT 30.1 31.6 -1.5**
Non-UCT 33.2 33.4 -0.3
Difference -3.0 -1.8 -1.2
Other HH Member
UCT 37.8 35.6 2.2**
Non-UCT 39.1 37.5 1.6**
Difference -1.3 -1.9 0.6
** Sign. at 5%
41. Cash Assistance for Poor
Students in Elementary,
Middle and High School
(BSM)
42. Years in education
The dropout rate both among the poor between
grades and stages of education is very high. “ “
Percent
43. Less
than
10% receive BSM
of poor people
Household Expenditure (Consumption) per Decile
Percent of 6-18-year-olds that receive BSM
45. IMPROVING
BSM
TARGETING
ACCURACY
USING
KPS
UNIFIED DATABASE
Children/parents bring
their KPS +
Family Card +
additional proof
to their school/
madrasah
MINISTRY OF EDUCATION &
CULTURE / MINISTRY OF
PROVINCIAL
DISTRICTS / CITIES
Schools/madrasah collect card summaries and
information on students for sending to the district/city
levels
RELIGION
47. Using KPS to Improve
Targeting Accuracy for BSM
School-based
Households-based (March 2014)
Source: Susenas 2009, SPS TW IV 2013 and TW I 2014
School-based
Households-based (March 2014)
Coverage of Beneficiaries (%)
Coverage of Beneficiaries (%)
Elementary School Middle School
49. The number of
poor decreased
4.25 million
in 5 years
32.53
28.28
14.15%
11.25%
2009 2014
Number of poor (million)
Poverty rate (%)
0.37
0.41
2009 2012
From 2009 to 2012
inequality continues to rise
50. Growth in Consumption and the Poverty Line
2010-2014
Percent (%) Decile 1 Decile 2 Decile 10
Average Growth in Consumption 2010-2014
Changes in the Poverty Line
51. Growth in Consumption and the Poverty Line
2013-2014
Percent (%) Decile 1 Decile 2 Decile 10
Average Growth in Consumption 2013-2014
Changes in the Poverty Line
52. ANNUAL
INFLATION:
FOOD
AND
NON-‐FOOD
Annual Inflation – Food (%)
Annual Inflation – Non-Food (%)
Food inflation is
always higher
compared with non-food
inflation. As
such, the burden on
the poor is heavier.
Annual Inflation