Module 2 Poverty Dynamics

1. Measuring
poverty
2. Multidimensional
poverty
3. Poverty Dynamics
4. Panel Data
5. Inference with
Panel Data
6. International
Poverty
Comparisons
Module 8
Tackling Poverty
7. Vulnerability to
Poverty
JONATHAN HAUGHTON
8. Tackling Poverty
j h a u g h t o n @ S u ffo l k . e d u
JUNE 2017
Outline
June 2017
1.
Poverty Profile
2.
ADePT
3.
Challenges:
a.
Education
b.
Health
c.
Nutrition
d.
Microfinance
e.
Employment guarantee schemes
f.
Insurance
4.
Targeting
5.
Tax, spending incidence
JH: POVERTY MEASUREMENT AND ANALYSIS COURSE
2
What is a poverty profile?
Systematic effort to determine
◦ Who the poor are
◦ Where the poor are
◦ Correlates of poverty
Essential input into Poverty Reduction Strategy Papers
A prelude to policy
Data intensive
June 2017
JH: POVERTY MEASUREMENT AND ANALYSIS COURSE
3
ADePT
An application designed to simplify and speed up the production of tables and graphs on
poverty and inequality
World Bank: 15 poverty assessments annually
Now version 5.4; easy to download
Next two slides: Sample output, Vietnam Urban Poverty Survey, 2009.
Hanoi vs. Ho Chi Minh City
Assumes 10,000 dong/person/day poverty line
June 2017
JH: POVERTY MEASUREMENT AND ANALYSIS COURSE
4
Sample ADePT
output (1)
June 2017
JH: POVERTY MEASUREMENT AND ANALYSIS COURSE
5
Sample ADePT output (2)
June 2017
JH: POVERTY MEASUREMENT AND ANALYSIS COURSE
6
Poverty Strategy: MKUKUTA in Tanzania
National Strategy for Growth and Reduction of Poverty 20052010
◦ Growth and reduction in income poverty: Central
◦ Good macroeconomic management, modest inflation, adequate fiscal discipline
◦ Improvement in quality of life & social well-being
◦ Primary schooling now almost universal, but low quality
◦ Governance and accountability
◦ “Empowerment”. Surveys find “a lot” of corruption
Goals OK; monitoring essential
`
June 2017
Needs to be founded on data and analysis
JH: POVERTY MEASUREMENT AND ANALYSIS COURSE
7
Evidence-based policy
RCTs have become popular
e.g. Do flipcharts work in Kenya?
◦ Not always possible (e.g. macro policy)
◦ External validity often suspect
◦ But hope is that accumulation of evidence will guide
Two Recent Books
◦ Banerjee & Duflo: Poor Economics
◦ Karlan and Appel: Beyond Good Intentions
June 2017
JH: POVERTY MEASUREMENT AND ANALYSIS COURSE
8
1. Education
Supply encourages enrolment, raises wages
Africa: primary enrolment from 54% to 70% (1999-2006)
Duflo: Indonesia’s school building program
Demand sometimes strong
Benefits from nudges: Conditional cash transfers (Mexico); cash transfers (Malawi)
E.g. for private schools
Challenge: improving quality
Private vs. public? After-school tutoring? Materials? Absenteeism? Tracking? Information? Computer-assisted learning? All need
appraisal.
June 2017
JH: POVERTY MEASUREMENT AND ANALYSIS COURSE
9
2. Health
Low-hanging fruit often unpicked. Why?
◦ Bed nets reduce malaria.
Kenya: $6 cost; $0.75 price kills demand; free helps
◦ Oral rehydration
Cheap, easy; yet 1.5m children die of diarrhea p.a.
◦ Chlorination
Automatic with piped water; cheap to add
Benefits hard to see; externalities; time inconsistency (so we need nudges – e.g. Progresa doctor visits; Mexico piso firme).
June 2017
JH: POVERTY MEASUREMENT AND ANALYSIS COURSE
10
3. Nutrition
Food is desirable per se, for health, for work
◦ India: % saying they don’t have enough food
◦ 1983: 17%.
2004: 2%
◦ China: Jensen & Miller: subsidized staple, less was bought
Why so little attention to micronutrients?
◦ Iodine
◦ Iron: In fish sauce in Indonesia; $1 p.a., +$46 earn
◦ Deworming in Kenya: improves schooling too
Effects of fortification hard to see; other priorities
June 2017
JH: POVERTY MEASUREMENT AND ANALYSIS COURSE
11
4. Microfinance
Not just microcredit
Inherently expensive to lend small sums without collateral
◦ Group borrowing can help, but limited
◦ Adverse selection, moral hazard are problems
Don’t oversell: not a panacea
◦ Seems to help (Khandker & Pitt on Grameen; Duflo et al. on Spandana in Hyderabad)
June 2017
JH: POVERTY MEASUREMENT AND ANALYSIS COURSE
12
5. Employment guarantees
Maharashtra.
◦ Started 1965. Low wage. Guaranteed manual work.
◦ Reduced poverty severity, but not headcount
◦ Originally: Relief operation to ameliorate distress. Can it be a program to fight poverty?
India: NREGS.
◦ Since 2005. 41m households. How effective?
Argentina: Trabajar II (1997) workfare
◦ Raised incomes (by half of gross wages paid); 4/5 of beneficiaries from poorest quintile
June 2017
JH: POVERTY MEASUREMENT AND ANALYSIS COURSE
13
6. Insurance
Against old age; crop failure; illness
Why used so little?
◦ E.g. Fewer than 20% in Gujarat, Andhra Pradesh, bought even minimal crop insurance
Standard problems:
◦ Adverse selection: only bad risks sign up
◦ Moral hazard: given insurance, careless
Other issues:
◦ Government spoiled market? Poor understanding by clients? Trust in provider? Time inconsistency?
June 2017
JH: POVERTY MEASUREMENT AND ANALYSIS COURSE
14
Constructing Social Protection: Indonesia
Context:
◦ 12% poor (29 million out of 240 million)
◦ Next 70 million still “vulnerable”
◦ 2008-2012: Rising inequality
Post AFC strategy:
◦ Food; jobs; access to health, education; credit for small enterprises. [Basri & Papanek]
Political support for targeting poverty
June 2017
JH: POVERTY MEASUREMENT AND ANALYSIS COURSE
15
Social Protection in Indonesia
Several programs, poorly coordinated.
◦
◦
◦
◦
◦
Rice for poor (RASKIN). Rp13 trn. 25% of population.
Scholarships (BSM). Rp19 trn. To rise to 25% of population.
Social health assistance for the poor (JAMKESMAS): Rp5 trn. 35% of population.
Unconditional cash transfers (BLT): Rp14 trn. [Low support.]
Conditional cash transfers (PKH): Rp1 trn in 2009. [Moderate support.]
◦ Due to expand to 3m hh by 2014, 6.5m by 2015; now 1.2m. Benefit c. 10%.
◦ Does it work? Well targeted? Efficient?
◦ PNPM . Cash for work.
◦ Self targeting; funds go directly to local body. Cost effective.
National Team for the Acceleration of Poverty Reduction (TNP2K), VP. Think
tank
June 2017
JH: POVERTY MEASUREMENT AND ANALYSIS COURSE
16
Targeting: Poverty Alleviation in Indonesia
Modest programs
Targeting not great
June 2017
JH: POVERTY MEASUREMENT AND ANALYSIS COURSE
17
How To Target the Poor
Means testing
◦ Requires high-quality data
Geographical
Community-based
◦ Local people identify peers who are poor
Population groups
◦ E.g. Infirm, old
Self-targeting
◦ E.g. food for work; free 3kg LPG bottles
Hybrid: Conditional cash transfers:
◦ Mexico: Progresa/Oportunidades [Paul Shultz; rigorous impact evaluation]. Brazil.
Bangladesh [Faria Huq]
June 2017
JH: POVERTY MEASUREMENT AND ANALYSIS COURSE
18
Identifying the Poor
Poverty mapping: identify poor areas
Construct proxy means testing methods
◦ By district (n=500); based on detailed survey data
◦ Index = f(housing info, assets, hhsize, education, work status, …)
Collect data on proxies
◦ E.g. 43% of households, 2011; + census data; + local community data.
Identify poor, unify database, issue cards.
Allow system for appeals; revisions; updates
June 2017
JH: POVERTY MEASUREMENT AND ANALYSIS COURSE
19
Determinants of Poverty
June 2017
JH: POVERTY MEASUREMENT AND ANALYSIS COURSE
20
At Least Get Tax, Spending, Right
Fuel subsidies
Taxes: which, and at what rates?
Spending: how target?
Net effects: average incidence
Marginal incidence
June 2017
JH: POVERTY MEASUREMENT AND ANALYSIS COURSE
21
Subsidies in Indonesia: Fuel and Electricity
Very long history; typical of oil producers
◦ But net oil importer since 2004
Some other countries too
◦ 2005: >2% GDP in Azerbaijan, Bolivia, Ecuador, Egypt, Indonesia, Jordan, Yemen
Large fraction of budget – [23% gov rev, 3.7% GDP, 2012]
◦ Crowds out other spending, including infrastructure
◦ Enough to spook ratings firms: S&Ps cut BB+ from positive to stable on May 3, 2013; stalled reform and
“weaker external profile”. Inefficient
Use too much energy
◦ But embedded in prices, and investment decisions
Weakly targeted
◦ Most benefits flow to non-poor
June 2017
JH: POVERTY MEASUREMENT AND ANALYSIS COURSE
22
Distribution
of Fuel
Subsidies
20 countries, 2005-2009
Direct effect: 2.6% of household consumption; indirect effect, 3.3%.
Source: Granado, Coady, and Gillingham. The Unequal Benefits of Fuel Subsidies. IMF, 2010
June 2017
JH: POVERTY MEASUREMENT AND ANALYSIS COURSE
23
Why Fuel Subsidies Persist
◦ Problem: High proportion of spending by poor, even if, absolutely, most benefits go to rich.
◦ Spending pattern similar to Thailand, for instance – see incidence analysis
◦ Difficult to undo
◦ May 14, 2013: Proposed raising
◦ Pgas 33% to Rp6,000/l,
◦ Pdiesel 22% to Rp5,500/l.
◦ Will add to inflation, which hits in short run (but not long-run).
◦ Politically, only replaceable if one can find better ways to target the poor
June 2017
JH: POVERTY MEASUREMENT AND ANALYSIS COURSE
24
Peru: Incidence of All Taxes
Progressive or not?
June 2017
JH: POVERTY MEASUREMENT AND ANALYSIS COURSE
Page 25
Peru: State Spending on Education 2000
June 2017
JH: POVERTY MEASUREMENT AND ANALYSIS COURSE
26
Education Spending Elsewhere
June 2017
JH: POVERTY MEASUREMENT AND ANALYSIS COURSE
27
Social Subsidies in Peru
Allocated based on reported usage
June 2017
JH: POVERTY MEASUREMENT AND ANALYSIS COURSE
Page 28
Peru: Distribution of Social Subsidies
June 2017
JH: POVERTY MEASUREMENT AND ANALYSIS COURSE
Page 29
Peru: Taxes Net of Social Spending
System seems progressive overall
◦ But
◦ Some taxes not included (e.g. CIT)
◦ Not all spending can be included
◦ Rely on our incidence assumptions
June 2017
JH: POVERTY MEASUREMENT AND ANALYSIS COURSE
Page 30
Peru: Marginal Incidence: Raise VAT by 1%
◦ 2000: assume VAT from 18% to 19%
◦ Of extra revenue:
◦ Education takes 7.6%
◦ Health takes 7.3%
◦ Social subsidies take 28.4%
◦ Remaining spending not allocable
June 2017
JH: POVERTY MEASUREMENT AND ANALYSIS COURSE
Page 31
June 2017
JH: POVERTY MEASUREMENT AND ANALYSIS COURSE
Page 32
Notes on the Reading
Jonathan Haughton & Shahidur Khandker (2013), Notes on Tackling Poverty.
Haughton & Khandker (2009), Handbook, chapter 15. This covers tax and benefit
incidence.
Abhijit Banerjee & Esther Duflo. 2011. Poor Economics: A Radical Rethinking of the Way
to Fight Global Poverty. PublicAffairs. An excellent treatment of microeconomic and
practical approaches, rooted in the use of RCTs.
Dean Karlan & Jacob Appel. 2011. More Than Good Intentions: How a New Economics is
Helping to Solve Global Poverty. Dutton. Often paired with the Banerjee/Duflo book;
covers many similar topics, but more breezily written.
World Bank. ADePT Software Platform.
http://econ.worldbank.org/WBSITE/EXTERNAL/EXTDEC/EXTRESEARCH/EXTPROGRAMS/EXTADEPT/0,,conten
tMDK:22595675~menuPK:7108374~pagePK:64168176~piPK:64168140~theSitePK:7108360,00.html
June 2017
JH: POVERTY MEASUREMENT AND ANALYSIS COURSE
33