Using the new PPI in Ethiopia

Hello,

I am from an organization called Enveritas that collects data on coffee and cocoa smallholder farmers across the world. We use your 10-question PPI within our survey to calculate PPI among our respondents. This year we used the new questions for Ethiopia, but we noticed a drastic change in the % of respondents who fall above or below the poverty and extreme poverty measures ($3.20 and $1.90 per person per day). There are a couple reasons why this is so, but the biggest factor is likely the new question that asks whether the household has consumed beef within the last 7 days. Our survey is conducted during coffee harvest season only (October - January) and there are strong cultural, religious, and seasonality factors related to meat consumption. The second longest Christian Orthodox fasting (no animal products consumed) occurs during harvest which means that certain regions with higher populations of Christian Orthodox farmers will answer no to this question, reducing those scores. Then there is the seasonality effect. Coffee harvest is a time of economic wealth for all coffee farmers, many of which eat meat daily or even for every meal during this time period. However, this is not sustained throughout the year, causing significantly higher scores from the previous PPI (which we used before this year) and from the rest of the year among the same farmers/regions.

I was wondering if anyone else using the new PPI index in Ethiopia has experienced something similar. Or if you have any advice for if we can substitute this question with another similar one that does not have to do with meat or dairy consumption.

Thanks,

Morgan