1a

1
Implementing USAID Poverty Assessment
ToolsMaterials Developed by The IRIS Center
at the University of Maryland
2
What is Poverty?How have you measured it?
3
Methodology Identifying the Very Poor

• Very poor households have non-standard and highly
  variable sources of income ? poverty must be
  measured using expenditure data.
• Expenditure surveys are too costly and
  time-consuming to conduct on all beneficiaries ?
  short-cut tools should be developed and tested.
• Can measure at household level, divide by number
  of household members for approximation
• IRIS testing methodology evaluates both the
  accuracy and the practicality of shortcut poverty
  assessment tools.

4
What is a Poverty Assessment Tool?

• Includes
• Sets of poverty indicators and coefficients in
  poverty level calcualtion
• Integration into program implementation who
  implements the tool on whom and when
• Data entry and analysis MIS or other data
  collection system/template
• Training materials for users

5
Background Legislation

• US Congress half of all USAID microenterprise
  funding must reach the very poor
• Microenterprise Mincrofinance, Business
  Development Services ( Enabling Environment)
• Very poor
• Bottom 50 below a national poverty line OR
• Under PPP1/day international poverty line
• USAID must develop and certify at least two tools
  to measure this level of outreach

6
Background (cont.)

• USAID-certified tools are objective, absolute,
  accurate, and low-cost
• Tools were tested for practicality before
  certification

7
Accuracy Tests
8
Balancing Errors Accuracy in the Aggregate

• Attention should be paid to the errors of
  classification classifying the very poor as not
  very poor (undercoverage) or the not very poor as
  very poor (leakage).
• Measuring accuracy at the aggregate level can
  allow a cancelling out of errors which does not
  occur at the individual household level.
• See handout PAT Errors Example and the Note on
  Assessment and Improvement of Tool Accuracy for
  more details.

9
Tests of Accuracy

• Testing indicators for their ability to act as
  proxies for poverty
• Tests of accuracy completed in Bangladesh,
  Uganda, Kazakhstan, and Peru
• Sampling nationally representative sample of 800
  randomly selected households
• Full reports available at http//www.povertytools.
  org

10
Design of Tests in Four Countries

• Two-step process obtains data on
• - poverty indicators from a Composite Survey
  Module (compiled from existing indicators and
  literature),
• and
• - Benchmark per-capita-expenditures from an
  adapted LSMS Consumption Expenditure Module.

11
Analysis of Eight LSMS Data Sets

• Objective
• Assess robustness of results from main study over
  larger number of countries, using methodology and
  set of indicators as similar as possible to 4
  field countries.
• 8 LSMS data sets
• Africa Ghana, Madagascar
• Asia India, Vietnam
• Eastern Europe and FSU Albania, Tajikistan
• Latin America and Caribbean Guatemala, Jamaica

12
From Accuracy to Practicality Developing Tools
13
Developing Poverty Assessment Tools

• Preliminary accuracy testing across 12 countries
  yielded 110 indicators that were best at
  predicting poverty.
• These 110 indicators were divided into 6
  prototype questionnaires.
• Each questionnaire tested in at least 3 countries
  and 2 regions of the world.

14
Objectives of the Practicality Tests

• Test-drive data collection methodology
• Test indicators for applicability, difficulty
• Use this information to create final tools that
  balance accuracy and practicality

15
Practicality Criteria

• Low risk of misreporting or manipulation
• Cost for implementation and for client
• Ease of data collection and analysis

16
Participants in the Practicality Testing

• Prototype tools tested by 17 microenterprise
  practitioners in 14 countries.
• Three data collection methodologies were tested
  household interview, intake, and ongoing
  monitoring.
• Feedback provided via reporting from implementing
  organizations and in-country debriefs.

17
Discerning Practicality Lessons

• IRIS gathered all information from the
  participating practitioners about the six
  questionnaires and three methods of
  implementation.
• Analysis of this data yielded lessons on which
  indicators and data collection practices were
  impractical.
• Lessons applied to create new tools for 12
  countries.

18
Certified USAID Tools
19
Characteristics of Certified Tools

• Tools are country specific.
• Poverty indicators may vary substantially between
  tools, as they were selected for accuracy in each
  country individually.
• Indicators were selected according to data from
  the national level, not regional. Regional data
  is more accurate but very costly to collect. Some
  indicators may be impractical in some regions of
  a country, but on the overarching national level,
  the indicator proved to be a good predictor of
  poverty.
• Tools were created via a process of trade-offs
  between accuracy and practicality. Similar to the
  difference of using an atomic clock or a sundial
  to measure time.

20
Characteristics of Certified Tools (cont.)

• Tools are accurate at predicting the aggregate
  levels of very poor households (by balancing
  errors).
• Tools are also practical to implement by taking
  field experience into account.
• Tools were developed in an iterative and
  collaborative process between USAID, tool
  designers, and the broader microenterprise
  community.
• Tools are for public use.

21
What the final certified tools will DO

• Measure aggregate number of clients above or
  below the legislative poverty line

22
What the tools are NOT designed to do

• Measure complex nature of poverty
• Measure multiple types of poverty
• Measure relative poverty
• Target clients for inclusion in the program
  (results are only known in the aggregate)
• Measure impact or movement out of poverty for
  individual clients