Versions Compared


  • This line was added.
  • This line was removed.
  • Formatting was changed.
Comment: Migrated to Confluence 4.0

Measuring Progress out of Poverty

Date last updated: Feb 20, 2007 – vinodhn

Update: UI Screens for this feature are now available `here <>`_.

Also see PPI Survey Conventions

To view the screens, download and extract the zip folder, click on login.htm, and if your browser gives you a warning select "Accept blocked content". Click "Login" button, click on "admin" tab, and follow the Functional Spec below for configuring and Viewing PPI tools and Completing a PPI Survey.

Business Need

Microfinance is a powerful tool in helping people move out of poverty. Measuring a client's poverty status and tracking this status over time is an important indicator that MFIs use to evaluate their social impact. Grameen Foundation's Social Performance Management Group has developed a Progress Out of Poverty Index. The feature described below will integrate the Progress Out of Poverty index into Mifos so that MFIs can store and report on their clients' movement out of poverty. For more background on this feature, see:

This includes the survey and indexes used to track poverty in 6 countries: India, Bangladesh, Haiti, Mexico, Pakistan, and the Philippines. This document also includes a good overview of GF's approach to poverty measurement.

Feature Overview

There are 5 terms that will be used when describing this functionality:

Poverty Index Survey: ---------------------

This is a country-specific survey that includes anywhere from 8 - 15 questions used to determine the Poverty Index Score (see below). The answers to each question carry a "score". The answers' scores are summed off-line and a Poverty Index Score is determined. The Poverty Index Score is the last field of each survey. These surveys are specific to a country. For example, the ownership of a pressure cooker may be an important factor to consider in India but may not be a relevant factor in Africa.

  • Example question from India survey: Does your household own a scooter/motorcycle. Answer: No, score 0 points / Answer Yes, score11 points.
  • Example question from the Pakistan survey: In the past 2 weeks, did anyone in your household eat any tomatoes. Answer: No, score 0 points / Answer Yes, score 5 points.
  • Example question from Haiti survey: In the past 12 months, did the household receive any money or gifts remitted from abroad. Answer: No, score 0 points / Answer Yes, score 7 points.

Poverty Index Score:

This is a score from 0 - 100 that is calculated from answers to the poverty survey. A total score of 0 is an indication of being extremely poor, a total score of 100 is an indication of being non-poor. As mentioned below, the value of this field is calculated by the application based on the answers to the survey questions.

Poverty Likelihood Chart:

A Poverty Index Score corresponds to a Poverty Likelihood that is detailed on a Poverty Likelihood Chart-- the probability that the client is non-poor, at-risk, poor, or very poor. For example, in India, if a client receives a poverty index score of 52 means that the client has a 2.5% chance of being very poor, a 10.8% chance of being poor, and a 86.7% chance of being non-poor. These probability %s will be used to calculate the overall poverty rate/likelihood of a larger portfolio and are not necessarily useful on at an individual client level. These Poverty Likelihood Charts are also country specific.

MFI defined poverty bands

While understanding the poverty probabilities for the overall portfolio is very important, MFIs also need to definitively identify their clients as belonging to a specific poverty buckets. Based on these buckets, they can offer particular products or track loan repayments/retention rates/recruitment rates of clients that fall within these buckets. The feature described below will allow an MFI to define what PI Score ranges map to which poverty buckets.

  • Example: An MFI may decide to use the following buckets: _ Poverty Index Scores from 0 - 24 => Very Poor _ Poverty Index Scores from 25 - 44 => Poor _ Poverty Index Scores from 45 - 69 => At Risk _ Poverty Index Scores from 70 - 100 => Non Poor

Client Poverty Status:

This is a data attribute stored at the client level. In Mifos, the look-up values for this field are configurable by the MFI. Currently this field is a user-edited field that can be changed at any time. The functionality below describes how the poverty defined bands would be used to determine this value.

Feature Summary

Configuring PPI and MFI defined poverty Bands

  • From Admin section, under "Organizational Preferences", user clicks "Configure PPI Settings".
  • User is taken to the Configure Poverty Measurement tool for PPI. (Note, other poverty measurement tools might be added in the future).
  • Note: Currently we have configuration only for the PPI Tool. Going forward we will add configuration settings for other tools also.
  • From "Configure page", user selects the country, status of the tool and has option to define MFI poverty bands _ Country is selected from a drop down (NOTE: we could automatically determine this based on the HO's country, but this probably isn't required? Would potentially require us to add logic to the HO address page that flags a warning if a user tries to change the HO's country and this feels like too much complexity) _ The status of the tool is also selected from a drop down and has two options; Active and Inactive. * System displays a table to define poverty band: rows for each poverty status defined in the system and columns for the user to enter the PI ranges.
    • This is optional. ie, An MFI can choose to use the PPI survey, but not choose to have it's score impact the client's poverty status.
    • User clicks on Preview and is taken to the preview page. On this page, the user has the option to edit the configuration.
    • User enters "submit" and is taken back to the admin page.
  • Once a poverty tool is selected and the country configured, the "Attach a poverty Index" link that appears on the client detail page (see below) will appear and display the appropriate survey.
  • Poverty Index Surveys and their corresponding Poverty Liklihood Charts change every 4 - 5 years, so there doesn't need to be a UI interface for entering in this information. Changes to the survey and the chart will impact only future surveys and scores and do not impact previous scores.
  • Open Issues: * Should we create a separate PI specific "poverty status field"-- and potentially have separate fields for each measurement tool. Seems like a bad idea. Since we will be dealing only with the PPI tool initially, this should not be a problem as of now.
  • OOS (from initial version): * Option to only display the PI score in the survey and not the other questions. (However, we can get around this by making entering the answers to the questions optional).

User Flow: Viewing the PPI Settings

  • From Admin section, under "Organizational Preferences", user clicks "View PPI Settings"
  • The ViewPPIDetails page is shown which contains the details of the configuration for the PPI tool.
  • The user also has the option to Edit the configuration from here.

User Flow: Completing a PPI Survey

  • If a poverty measurement tool has been selected, then the Progress out of Poverty Index (PPI) Survey will appear in the drop-down list of surveys on the "attach a survey" page off the client detail page(ClientDetailsMeeting).
  • User clicks on "Attach a survey" link on the client detail page.
  • User selects the PPI Suvey from the drop down list and clicks on Continue.
  • The PPI survey is displayed and the user can enter the survey data for the client.
  • After completing the survey, the user then clicks on Preview.
  • The user also has the option of clicking on Clear All, which will clear all the selected answer options for the survey.
  • On clicking Preview, the user is taken to the attachPPISurveyPreview_client page where the entered answer options can be previewed along with the points for each answer.
  • The Status field is automatically calculated by Mifos based on the total score and the defined poverty bands.
  • Clicking on Submit, takes the user back to the Client page.
  • Note: If the poverty bands have been defned, then a poverty status field is attached to the client page (ClientDetailsMeeting), This will come under a new section called Poverty measurement information and will have the following fields:
    1. Poverty measurement tool:
    2. Poverty Band:

Additional Information

  • Note: PPI surveys can only apply to clients, not to accounts, groups or centers.
  • There will be a field for when the survey data was collected. Surveys can be "back-dated" without any restriction.
  • Mifos will use the last field of the survey, the "total score", to look-up poverty likelihood scores. The allowable range for the last field should be between 0 - 100.
  • If MFI has defined MFI specific poverty buckets, then upon submission of the survey-- the client's Poverty Field will be updated as appropriate. The Poverty Field will be displayed on the client detail page. The actual % of poverty likelihood will not be displayed, only stored in the back-end.

User Flow: Editing a PPI Survey

  • Users should be able to edit a PPI survey at any time. Any changes will update the poverty status.

User Flow: Deleting a PPI Survey

  • Users should be able to delete a PPI Survey. This will remove the associated poverty status and scores. If the most recent survey is deleted, then poverty field will revert back to the previous score.

User Flow: Generating Reports on PPI Scores

Examples of reports that we'll likely want to generate using PPI scores. More detail and explanation still needed:

  • Poverty MFI, branch or regional office, loan officer, etc.
  • Age of client, ethnicity, other social characteristic
  • Repayment and default rates
  • Type of income generating activity
  • loan product: size and type
  • time in program
  • reason for leaving MFI (ref client exit survey)
  • Change in poverty status using cross-sectional and longitudinal data
  • savings balance
  • use of other financial product (microinsurance, etc.)
  • Use of non-financial product (participation in literacy, training, health programs. etc.)
  • Education level of clients and children

PPI Scorecards

India Scorecard Mexico Scorecard Peru Scorecard Philippines Scorecard Bolivia Scorecard Pakistan Scorecard Morocco Scorecard Haiti Scorecard

– Main.vinodhn - 20 Feb 2007 Date last updated: Feb 20, 2007 – vinodhn