An Asset-Based Approach to Poverty Dynamics and Safety Nets: Research and Policy Questions

1
An Asset-Based Approach to Poverty Dynamics and
Safety Nets Research and Policy Questions
Christopher B. Barrett Cornell University Michael R. Carter University of Wisconsin



November 7, 2005 Inter-American Development
Bank Washington, DC
2
Overview

• An Asset-based Perspective on Poverty
• Poverty Traps and the Dynamic Asset Poverty
  Threshold
• Empirical Evidence on Poverty TrapsWhat We Know
  So Far
• Bifurcated Asset Dynamics (South Africa)
• Long-term Effects of Short-term Shocks (Honduras)
• Asset Smoothing and Its Human Costs (Zimbabwe)
• Exclusion from Informal Safety Nets (East Africa)
• Cash Transfer Programs Poverty TrapsWhat We
  Dont Yet Know
• Future Directions for Cash Transfer Programs

3
Evolving Views of Poverty

• Successive generations of poverty analysis
• 1st static income/expenditure analysis
• (headcount, poverty gap, FGT measures)
• 2nd dynamic income/expenditure analysis
• (chronic/transitory poverty distinction)
• 3rd static asset poverty analysis
• (structural/stochastic poverty distinction)
• 4th dynamic asset poverty analysis
• (behaviorally-based poverty lines)

4
Asset-Based View of Poverty
Transitions from poverty 1) Stochastic churning
(B to u(A)) 2) Structural via
accumulation (A to A) 3) Structural via
higher returns (u(A) to C)
5
Poverty Traps and the Dynamic Asset Poverty
Threshold

• Will structurally poor move ahead over time?
  Depends on underlying dynamics of asset
  accumulation.
• Lessons from empirical macroeconomics is growth
  characterized by unconditional convergence,
  convergence clubs, or threshold-based multiple
  equilibria?
• Key question do returns to productive assets
  (land, labor, etc.) increase locally in wealth?
• What causes such dynamics and locally increasing
  returns?
• Increasing returns to scale in income generating
  process
• Minimum investment levels/indivisibilities
• Uninsured risk

6
Exclusion from opportunities is key

• Social exclusion ethnic/gender barriers
• Financial exclusion credit/insurance access
• Two can be reinforcing (Mogues and Carter 2005)

7
A 4th Generation View
Utility
L2
UH
Dynamic Asset Poverty Line
Income Poverty Line
L1
UL
Static Asset Poverty Line
Initial Assets
AS
A2
A1
A
A
Poverty Trap
Dynamic Asset Poverty Line (Micawber Threshold)
AtA0 (dynamic equilibrium)
Next Periods Assets
8
Empirical Evidence on Poverty TrapsWhat We Know
So Far

• Theory thus suggests circumstances in which
  poverty traps might exist
• But what do we know about their actual existence
  and importance
• Brief review now of various empirical studies
  that test for different implications of poverty
  traps

9
Bifurcated Asset Dynamics(South Africa)

• South African data, 1993-1998 (KIDS study)
• Define and estimate asset index for each
  household i in each period t, Lt(Ait), such that
  asset weights (prices) depend on asset mix
• Index scaled such that it is measured in poverty
  line units (PLUs)i.e., the index tells us what
  fraction of the poverty line a households bundle
  of assets would be expected to generate
• Non-parametric estimation of asset dynamics
• Key findings
• Divergent dynamics
• Repelling Micawber Threshold at 2 PLUs
• Poverty trap equilibrium at 0.9 PLUs
• Corroboration by later qualitative and
  quantitative data

10
Bifurcated Asset Dynamics
Source Adato, Carter and May (2006). Exploring
Poverty Traps and Persistent Poverty In South
Africa Using Qualitative and Quantitative Data
JDS
11
Estimated South African Asset Dynamics
12
Long-term Effects of Short-term Shocks
(Post-Mitch Honduras)
Source Carter et al. (2005). The Long-term
Impacts of Short-Term Shocks Poverty Traps and
Environmental Disasters in Ethiopia and Honduras
13
Asset Smoothing Its Human Costs(Zimbabwe)
14
Asset Smoothing Its Human Costs(Zimbabwe)

• Those owning gt2 oxen liquidated animals at 3.5-6x
  rate of those owning 1-2 in response to 1994-95
  drought
• Drought persistently lowers growth rates of
  children 12-24 months, temporarily lowers BMI of
  women, but no effect on men or older
  pre-schoolers.
• The nutritional impact is larger and more
  persistent in households with lower levels of
  livestock holdings. Asset portfolio choice
  protect human or livestock capital
• Temporary shocks, even mild ones, can have
  long-term consequences
• Source Hoddinott (2006). Shocks and Their
  Consequences within and between Households, JDS

15
Exclusion from Informal Insurance

• There might be holes in informal safety nets
• Santos and Barrett (2005) on Ethiopian
  pastoralists social invisibility within the
  poverty trap
• Logit estimates suggest that transfers flow in
  response to shocks, but only to those who have
  not collapsed into the poverty trap.
• Those in the trap are significantly less
  frequently known smaller networks. Estimated
  39 have no effective social insurance network.
• Implication transfers to persistently poor have
  negligible crowding out effects.
• Lybbert et al. (2004), Lentz and Barrett (2005)
  and McPeak (2006) meager interhousehold
  transfers among east African pastoralists, no
  crowding out effects

16
Cash Transfer Programs Poverty TrapsWhat We
Dont Yet Know

• While the empirical is still thin and imperfect,
  hopefully it is sufficient to encourage a deeper
  look at poverty traps and what they might mean
  for programs like Progressa
• To introduce these ideas and implications, I
  would like to criticize my own study of a South
  African (unconditional) cash transfer scheme, the
  Child Support Grant (CSG)

17
Cash Transfers Poverty Traps
Source Agüero, Carter and Woolard (2005). From
Flows to Stocks The Impact of Unconditional Cash
Transfers on Human Capital
18
Cash Transfers Poverty Traps

• So what is long-term value of this human capital
  asset?
• Assume that
• Maintain z-score gain ? 2.1 cm gain in adult
  height
• Adopt Thomas-Strauss wage-height elasticity
  estimate of 2.4-3.3
• Implies adult monthly wage gain of R190-R262
• Wage gain accrues from 25-65 years old with 50\
  unemployment
• Results
• Present value at birth of expected wage gain
  R6500-7500
• Program cost R3400 (plus administration costs)
• Benefit-Cost 1.6-2.3
• But two critical questions to ask of this simple
  analysis
• Sufficient to surmount threshold?
• Sustainability of human capital gains given
  probability of shocks?

19
Future Directions for Cash Transfer Programs

• Progressa/Opportunidades compelling because
  targets well-being of current generation
  inter-generational transmission of poverty
• Yet we would seem to know relatively little about
  whether the flows and stocks of Progressa create
  basis for sustained accumulation for some or all
  beneficiaries
• Researchable question, but also one worthy of
  further experimentation
• Levels of support
• Basic asset grant
• Remedy exclusion (leverage transfer flows)
• Protection against shocks (perhaps only common
  shocks for incentive compatibility purposes)

20
Implications for Policy Policy Experiments

• In summary, shocks in the presence of poverty
  traps imply
• Long run micro (macro?) growth effects
• Costly chronic poverty
• Costly avoidance of persistent poverty (asset
  smoothing)
• Social protection policy built around this
  behavioral poverty line would appear to be
• Cost-effective
• Imply unpleasant triage?
• Would also seem to imply that ex ante insurance/
  credible safety nets have behavioral/growth
  implications

21
References

• Theory and Concepts
• Carter and Barrett (2006). The Economics of
  Poverty Traps and Persistent Poverty An
  Asset-based Approach, JDS
• Bifurcated Asset Dynamics
• Adato, Carter and May (2006). Exploring Poverty
  Traps and Persistent Poverty In South Africa
  Using Qualitative and Quantitative Data JDS
• Lybbert, Barrett, Desta and Coppock (2004),
  Stochastic Wealth Dynamics and Risk Management
  Among A Poor Population, EJ
• Barrett, Marenya, McPeak, Minten, Murithi,
  Oluoch-Kosura, Place, Randrianarisoa,
  Rasambainarivo and Wangila (2006), Welfare
  Dynamics in Rural Kenya and Madagascar, JDS
• Long-term Effects of Shocks
• Carter, Little, Mogues and Negatu (2006). The
  Long-term Impacts of Short-Term Shocks Poverty
  Traps and Environmental Disasters in Ethiopia and
  Honduras, WD
• Lybbert et al. (2004), EJ
• Asset smoothing/Consumption destabilization
• Hoddinott (2006) Shocks and Their Consequences
  within and between Households, JDS.
• Zimmerman and Carter (2003), Asset smoothing,
  consumption smoothing and the reproduction of
  inequality under risk and subsistence constraints
  JDE
• Barrett et al. (2006), JDS
• Exclusion from Informal Insurance
• Santos and Barrett (2005), Poverty traps and
  informal insurance Evidence from southern
  Ethiopia Cornell working paper.
• Lentz and Barrett (2005), Food Aid Targeting,
  Shocks and Private Transfers Among East African
  Pastoralists, Cornell working paper.
• Lybbert et al. (2004) EJ
• McPeak (2006), "Confronting the Risk of Asset
  Loss  What role do livestock transfers in
  northern Kenya play?" JDE

22
Thank you!