Frequently Asked Questions

The Poverty Probability Index (PPI^®) is a poverty measurement tool for organizations and businesses with a mission to serve the poor.  The PPI is statistically-sound, yet simple to use: the answers to 10 questions about a household’s characteristics and asset ownership are scored to compute the likelihood that the household is living below the poverty line – or above by only a narrow margin.  With the PPI, organizations can identify the clients, customers, or employees who are most likely to be poor, integrating objective poverty data into their assessments and strategic decision-making.

A business or organization must use the PPI scorecard that has been developed for the country in which they operate;  PPIs have been developed for dozens of countries. There are two steps to using the PPI to determine a person’s level of poverty.

• The survey and score: The PPI survey contains 10 verifiable questions that a field agent can ask their clients in 5 to 10 minutes. The questions are simple – “What material is your roof made out of? How many of your children are in school?” The survey respondent chooses an answer from multiple choices. It is important that the PPI administrator ask and interpret the survey questions consistently across all clients and as directed by the PPI guidelines in order to maintain accuracy. For many PPIs, there is a corresponding document that provides guidance on how to interpret a question in complicated situations.  Each answer is given a value, and the total value of all the answers is the survey respondent’s PPI score.
• Poverty likelihood look-up table: The PPI administrator uses the PPI look-up table to convert the PPI score to a likelihood that the respondent’s household is living below a poverty line. The look-up table allows the PPI administrator to determine the household’s likelihood of living below multiple national and international poverty lines.

In November 2017, the PPI technical team updated the methodology behind the development of the PPI. To learn more, read our blog post.

An important note is that existing PPIs remain valid and will be continue to be supported by the PPI Help Desk. However, the new methodology will be used to create new PPIs, and update existing PPIs from this point on. You can see the PPIs that can be updated with new available data here.

Click here for information on the PPI development pipeline. 

Social performance is defined as “the effective translation of an institution’s mission into practice in line with accepted social values.” Social performance describes how well an organization is achieving its mission, or social goals. Organizations that use the PPI have a mission to positively impact the lives of people living in poverty through information, health services, employment opportunities, financing, or a combination of interventions. These organizations use the PPI to better understand their performance against these social goals.

The development and update of every PPI is coordinated by the PPI team and includes the input of a variety of stakeholders. The PPI team tests every PPI in the field and creates supporting documentation in the interest of transparency.

The questions, responses, and weights on the PPI scorecard and look-up table are derived from each country’s most recent national household expenditure or income survey. These surveys typically contain 200 to 1000 questions. Of these, ten questions are derived for the PPI scorecard, based on a balance of the following criteria:

1. The question has a strong correlation with poverty, i.e., there is statistical significance that households who answered the question a certain way are below the poverty line. Example: “What is the level of education attained by the head of the household?” 
2. The question is inexpensive to collect, easy to answer quickly, and simple to verify. Example: “Of what material is the roof of the residence made?”
3. The question is liable to change over time as poverty level changes. Example: “Does the household own a motorbike or car?”

After the scorecard questions are selected, the scoring system is developed so that the lowest possible score is 0 (most likely poor) and the highest is 100 (least likely poor). Each PPI scorecard is published with a Design Documentation Memo. Read this document for your country’s PPI for a detailed account of how the PPI was made and why it is statistically sound. 

Organizations and businesses that currently use the PPI differ in their business models but share a mission to help those living in poverty. PPI users either work directly with the world’s poor, or they are intermediaries who influence or invest in a network of other organizations that work directly with the world’s poor. PPI users include but are not limited to microfinance organizations, impact investors, public health organizations, social enterprises, and nonprofit networks.

Your organization may operate in a highly-impoverished area, but without objective poverty data on the households you reach, social performance management will rely on assumptions. Organizations that collect poverty data from all or a statistically significant sample of households they reach are able to:

• measure poverty outreach (i.e. the portion of customers, clients, or employees who live below the poverty line),
• assess the performance of the intervention among the poor and poorest, and
• track poverty levels over time.

With these data, management staff can make informed strategic decisions and can provide stakeholders with objective evidence that the organization is reaching the poor.

The PPI scorecard, look-up table, and supporting documents are free for download on this site. However, the process of implementing the PPI will cost the organization staff time. The amount varies, depending on how the organization decides to implement the PPI. The least expensive method is to add the PPI survey to existing information fields that are collected on all incoming clients, customers, or employees.

This method may only require an additional 5-10 minutes of a representative’s time per client.  It is more expensive to administer the PPI a time when the representative is not already scheduled to visit with the client, customer, or employee, which is often the case when an organization administers the PPI to a sample as an independent research exercise.

The organization must also budget time for PPI data analysis. The PPI has an especially high return on investment for organizations with a staff member who is dedicated to monitoring social performance because PPI data allows that staff member to conduct more efficient and accurate analyses. 

An organization must have a working management information system (MIS) and Microsoft Excel to make effective use of PPI data.

Administer the PPI to the same group of households or two equally representative samples of households at a regular interval. If an individual household’s poverty likelihood changes, you can infer that the household’s economic standing has changed. For example, if a household had a 65% likelihood of living below the poverty line last year and a 54% likelihood of living below the poverty line this year, you can infer that the household is moving towards being non-poor.

If the poverty rate of a group changes over time, you can infer that the number of people in the group who live below the poverty line has changed. For example, suppose last year 58 percent of your clients lived below the poverty line. Then suppose that this year only 39 percent of the same group of clients lived below the poverty line. This means that every 19 out of 100 clients moved out of poverty.
 

Many businesses and organizations serve a mix of people above and below the poverty line. The proportion of clients that live below the poverty line represents the organization’s poverty outreach. This is also called the poverty rate of the group. To find your poverty outreach, you must average the poverty likelihoods (not the scores) of your PPI respondents. For example, if you administered the PPI to 100 people and their average poverty likelihood was 58%, you would conclude that 58% of those 100 people, or 58 people, live below the poverty line.

Yes, you can estimate person-level poverty rates by taking a household-size-weighted average of the household-level poverty likelihoods. To do so, follow these steps:

1. Multiply each household poverty likelihood by the number of household members.
2. Add all products from step 1.
3. Divide by the total number of individuals in question. Each country's PPI design memo provides more detail on this process and an example, typically found in section 2.2.

On July 15, 2016, the PPI moved from Grameen Foundation as its organizational ‘home’ to Innovations for Poverty Action (IPA) and will now be governed by the Steering Members that make up the new PPI Alliance. This announcement addresses questions that PPI Users may have regarding this change and how it might impact them.

If there were a one-to-one relationship between scores and likelihoods, there would be nothing wrong with averaging scores to calculate the poverty rate of a group. However, this is not the case and averaging scores will consistently lead to an incorrect poverty rate. For a full explanation, read our blog post on why averaging PPI scores won't work.

When using the PPI and reporting its results, it’s helpful to understand the difference between the terms poverty likelihood and poverty rate—sometimes referred to as estimated poverty. These terms are not interchangeable and express different concepts, so it is important to use them correctly. For a full explanation, read our blog on Poverty Rates vs. Poverty Likelihoods.

The PPI supports assertions of impact but cannot prove it alone. The PPI is a measurement tool, similar to a ruler. A ruler can be used to measure someone’s height at two separate times and show that someone has grown, but the measurements don't indicate how. The ruler cannot attribute the increase in height to any particular cause. Likewise, a program aimed at alleviating poverty can measure poverty levels over time, but it cannot use the PPI alone to measure its success. However, the PPI can be used as an integral component of an impact evaluation framework that assesses projects serving the poor. With the PPI and other data obtained through such a framework, managers can make assumptions about causality that inform their decisions.

Yes. When doing so, the international poverty lines ($1.25/day, $2.50/day, etc.) are recommended since these lines were developed by the World Bank precisely to be compared across multiple countries.

However, PPI-calculated poverty rates from different countries are not perfectly comparable. Please consider the following when reviewing data from multiple countries:

• Different countries measure and/or value expenditure  in slightly different ways, meaning that Country A might consider a person poor compared to the $1.25/day poverty line, while Country B looking at the same individual might conclude the opposite.
• Some countries adjust the absolute value of poverty lines for regions within the same country. A country could define just one poverty line or it could vary the value by region, as is the case in Zambia, where there are nine values for each poverty line.
• PPIs are made when data and funding become available, so they are not created at the same point in time (or even within the same year sometimes). Changes in a country’s economy may, over time, affect the results of a country’s PPI.

As noted in each PPI Design Documentation, it is not possible to know the extent to which a given scorecard will become inaccurate over time. The change in accuracy depends on the overall change in the economic situation of each country; this change could happen very quickly or could progress more slowly. The goal is to update each PPI every five years, which is generally how often countries update their national household survey, and meeting that goal is dependent on data availability, funding and interest. 

The PPI is a public good. As such, you may use it and reference it on your website freely. If you plan to use the PPI logo, we kindly request that you let us know.

The PPI scorecard is based on data from a nationally representative group of households. When you apply the survey to a group of households that are not nationally representative – such as a group of farmers or a group of women – there will be 'out of group bias'. Organizations that use the PPI will usually encounter out of group bias because they have designed their services to meet the needs of households and individuals with specific characteristics, or because they do not operate across the entire country.

Most organizations cannot avoid this bias, and since the bias will be different in every case, we cannot predict the bias ahead of time. Those responsible for analyzing PPI results and making strategic decisions based on those analyses should consider how their unique context may affect PPI results.

You cannot alter the PPI's questions, answer options, answer values, or look-up table without rendering the PPI inaccurate. Even a small change can have a big impact on the PPI's accuracy. The questions, responses, and scores on the PPI scorecard and look-up table are derived from recent national household expenditure or income survey s and should not be changed. The PPI was calibrated using indicators that were found through statistical means to be the most powerful indicators of poverty within a country. While the relevance of a particular indicator may vary from region to region within a country, the overall predictive power of the PPI remains strong. Changing any component of the PPI will undermine its predictive power and result in poor data. Learn more about how the PPI is designed.