Here’s another conundrum we stumbled across today, possibly the killer. SDDNCF has a database of all its clients. We filter this database for basic criteria such as age, neighborhood/cluster, and occupation to create the survey masterlist for each city. For Lahore we have about 400 potential respondents 18 years or older in 12 professions across 8 or so neighborhoods. The problem is that basic functional literacy (sufficient to compose an SMS) is a criterion for this project but literacy level is not explicitly listed in the database, only the highest level of schooling completed by head of household which is often not the client herself. So we’re left guessing until we arrive in the field and actually meet the clients and family.
Incertainty
Yesterday about 90% of respondents were sufficiently literate. Today it’s the reverse. This has to do with the specific demographics in each neighborhood and family and caste and the nature of trainings conducted in each cluster and luck. As such none of my colleagues can predict or explain the demographic differences causing the discrepancy between yesterdays batch of 18 to 22 year old tailoring trainees and today’s. They and their surroundings looked exactly the same to me.
Now we’re back to square one. We can’t teach ABCs and we can’t realistically expect people to learn on their own even if the incentives for participating in this network were as magnificent as I say they are. Thus the only criterion that really matters and the one we can’t control for internally is literacy. ask anyone for the key to unlock the poverty problem and they’ll tell you it’s education. So here is the simultaneity problem in real life: how/why do you get people to use a tool that requires a certain degree of education when if they had that degree in the first place you wouldn’t be targeting them as needy. The answer is there is no easy answer. We could commit in years and teach a few of the 50 million illiterate people in Pakistan but that’s the job of the entire society and certainly not a feasible objective for a six month pilot. So what’s the middle ground – the most likely to be feasible and produce the greatest value for the greatest number in the little time we have left – I can’t really say anymore. I never actually knew in the first place.
So, for the sake of an effective pilot, can you target the areas where you discover higher literacy rates and recruit more people there – or does that screw up the validity of the findings?