no image

Retrospective PPI?

Lindsay Dakan
• Akola Project
• Uganda
• 02/26/19
• 1 Comment

Hi PPI users- my organization is currently struggling with a lack of baseline data on beneficiaries' poverty levels at the start of the project. We want to measure both the organization's poverty reach over time, as well as changes in poverty rates among beneficiaries over time. Would it be valid to conduct a retrospective PPI survey, asking for example "What type of toilet facility did the household mainly use in [year before the start of the project]?" (given the usual caveats about recall bias). What factors should we take into account if we were to go down this road?

Thanks in advance for any insights!


1 Comment
Hi Lindsay, Thank you for your question. We would generally advise against re-framing questions in the PPI to ask about the household's status at a point in the past, due to accuracy issues with retrospective, recall data. This has been a finding in several published papers. See this paper for example, which "finds that the magnitude of the recall error increases over time, in part because respondents resort to inference rather than memory" -- This recall error would have obvious implications for the accuracy of poverty estimates obtained by using the PPI. Best regards, Sharada