How data can inform policy? Some examples… 1. Data from public budgets • Public expenditures and revenues are telling a lot about policy.

Download Report

Transcript How data can inform policy? Some examples… 1. Data from public budgets • Public expenditures and revenues are telling a lot about policy.

How data can inform policy?
Some examples…
1. Data from public budgets
• Public expenditures and revenues are telling a lot about
policy (and government efficiency)
– Accountability
– Budget transparency
– Allocative efficiency
Line-item vs. programme budget
Ministry of Health
Line-item Budget
Salaries & wages
Overtime
Building expenses
Programme Budget
6 000 General
150 Administration
Transport
800 Primary health care
750 & health promotion
Equipment
400
Shipping
125 Hospital services
Water & lights
15
Telephone
Printing
25 Training & medical
20 research
Consumables
12
TOTAL
8 297 TOTAL
462
4 326
2 817
692
8 297
Public Financial Management
Fiscal Discipline
(gov’t budget balance)
Allocative Efficiency
(public expenditure
planning)
Tax & aid policy
(revenue planning)
Operational efficiency
(implementation)
2. Data from Household Surveys
• Descriptive statistics – together they can be powerful
– Focus on the big picture of “issues and policy responsiveness”
– Can be used for highlighting vicious and virtuous policy cycles
(multidimensional model of child poverty)
• Exploring causality with multivariate statistics
– What is the role of certain factors (e.g. parental education) in
child outcomes
– Why certain policies work or do not work
Percentage of children experiencing severe deprivations in East Asia
Cambodia
100
Lao
Viet Nam
Thailand
91
90
80
75
74
70
70
66
60
55
51
50
46
39
40
34
30
20
22
21
19
16
16
26
26
24
16
14
12
11
8
10
16
14
2
7
2
1
7 7
1
3
2 1
0
Absolute
Severe
Poverty (2+ Deprivation
deps)
(1+ deps)
Shelter
Water
Sanitation Information Education
Food
Health
Source: MICS/direct communication with Bristol University
Rates of absolute poverty among children by household type, India
0
50
100
62
1 adult 1 child
60
1adult 2 children
79
1adult 3+ children
45
2 adults 1 child
41
2 adults 2 children
67
2 adults 3 children
77
2 adults 4 children
81
80
2 adults 5 children
2 adults 6+ children
3 adults 1 child
40
51
3 adults 2 children
62
3 adults 3 children
71
3 adults 4 children
75
3 adults 5 children
79
3 adults 6+ children
42
4 adults 1 child
47
4 adults 2 children
59
4 adults 3 children
65
4 adults 4 children
71
4 adults 5 children
74
4 adults 6+ children
5 adults 1 child
5 adults 2 children
38
39
49
5 adults 3 children
55
5 adults 4 children
68
5 adults 5 children
76
5 adults 6+ children
46
6+ adults and 1+ children
Adult Only Household
Children Only (Under 18)
43
72
Income Poverty Dynamics
Income poverty dynamics in the Maldives, 1997, 2004 and 2005
Percentage of households with less than Rf. 15 per person per day, Atolls
'poor'
'non-poor'
60%
23%
9%
11%
4%
1997
40%
5%
3%
37%
14%
29%
7%
32%
27%
2004
2005
Source: Dr. Fuwad Thowfeek, Statistics Maldives
Intergenerational income mobility:
your father earns 100 per cent more than mine
- what per cent impact will that alone have on our earning differences?
Canada
Finland
Sweden
Germany
France
Nepal
Pakistan
United States
United Kingdom
Malaysia
Peru
Brazil
Ecuador
0
20
40
60
80
100
Source: Dr Miles Corak Statistics Canada
Multidimensional child poverty concepts broaden policy focus
Albania: % of children 3-5 yrs old materially poor with poor nutritional
outcomes and not attending pre-school – Venn diagrams
MATERIALLY
POOR
POOR HEALTH
OUTCOME
1.3%
2.9%
7.3%
11.8%
20.2%
17.9%
NOT ATTEDING PRESCHOOL
15.8% are not poor,
have access to
preschool, clean water
and are in good health
22.7%
Source: 2002 LSMS. Note: Total number of children 450.
Angela Baschieri and Jane Falkingham (University of Southampton)
Anthropometric failure
and breastfeeding practices in Tajikistan
Nutritional status by breastfeeding pattern for children less than 18 months
35
30
25
20
Exclusively or partially
breastfed
15
Fully weaned
10
5
0
Wasted
Underweight
Source: MICS 2005 and
Baschieri and Falkingham, 2007
Stunted
Breastfeeding practices
• Most women in Tajikistan stop exclusively breastfeeding and switch
to a mix feeding pattern relatively early
– Amongst children aged 6-23 months under 5 percent are either
‘exclusively’ or ‘almost exclusively’ breastfed.
• As a result many children are exposed to the risk of poor nutrition
and associated adverse developmental consequences.
Is family land ownership an effective policy
against child malnutrition?
(results of multivariate analysis)
• We control for children age (months), region, mother education, wealth
quintile, ethnicity, sanitation, household access to land, ownership of
livestock
• We found that children living in a households with access to land have
higher probability of being underweight that those without access to land
3. International comparisons
•
•
•
•
Can be helpful for “big policy ideas”
Highlighting policy coherence and/or policy efficiency
Can stimulate policy transfer
Advocacy value
Income inequality and economic output: paths of
development
70
Brasil
So ut h- A merican
eco no mies ( early 90s)
60
Panama
Colombia
50
Gini coefficient
Chile
Venezuela
M o st develo ped F arSo ut hern East
eco no mies ( early 90s)
Costa Rica
West ern eco no mies
( early 90s)
M OL94
40
RUS94
EST94
BUL94
LIT94
CEE 1994
United Kingdom
Singapore
USA
Korea, Rep
Italy
POL94
30
ROM 94
M OL89
France
Canada
FRG
EST89
Netherlands
Sweden
LIT89
BUL89 RUS89
ROM 89 POL89
HUN94
CZE92
SLK94
HUN89
SLK89CZE89
Japan
CEE 1989
20
10
0
10
20
30
40
50
60
70
80
90
PPP estimates of GDP per capita (USA = 100)
Source: Children at Risk in Central and Eastern Europe: Perils and Promises, Regional
Monitoring Report - No. 4, UNICEF 1997.
100
Challenges in using statistics to inform
policy
• Existing concepts, data and availability
• Sensitivity analysis, robustness
– child focus
– thresholds
– economy of scale/equivalence of scale (income data)
• Design causal analysis: Need hypotheses plus data
to test them
• Overlaps of income and non-income dimension:
limitation