Thanks so much Nick, this is really helpful and interesting. We do base eligibility on objective criteria, such as asset ownership, housing situation, consumption, etc. We hesitate to exclude folks based on very broad categories such as location (although we do restrict eligibility by district for research and logistics reasons). So far this has been working reasonably well, as you can see from the data above on income amongst the selected students. Very interesting point about the government schools, we’ll discuss that.
On the fraud, yes I’d be grateful for any advice you have!
That’s interesting Johannes—I would be pretty nervous about income, and consumption questions yielding accurate results. I wouldn’t call family income or consumption measures objective criteria really as it relies on what people tell you—perhaps parental qualification level might be better if you do want to get an objective idea of likely income?
I might be confused but isn’t the data about income just from your own surveys, what is it proving exactly?
I think the example of 7 out of 30 students falsifying or attempting to results, is decent evidence of the lengths people will go to to get their kids or themselves onto a program like this.
If you are going to rely on subjective interviewing, I think the IPA validated poverty scores are pretty good, you might be using them already. Some of the questions are pretty regionally dependant so not super reliable (there’s a silly one about whether you eat rice or not, which of course doesn’t work if you already grow rice as a staple), but at least it is a researched and validated survey https://www.povertyindex.org/country/uganda
On the fraud front sorry I missed the explanation, I would have thought you could just have candidates sit a standard English proficiency exam in the German Embassy? It seems like an easy fix thankfully.
One of your biggest risk as you grow I would say is in student selection. Your people employed in Uganda will have so much incentive to favour family members, take bribes for places etc. This is why in this case I would use as many truly objective poverty measures as possible even if it is not technically the “best” way of assessing poverty. The first thing I would do would be get photos/photocopies of the highest level of education of their parents and their siblings as this will give you a pretty good idea.
I would also seriously consider separating people who identify potential candidates and gather the poverty data, with the people doing the assessing (if you don’t already). I would have people submitting the objective information to a central office, which would then select people based on criteria. I would be interested to see your current criteria as well.
If it was me I would be excluding based on location to make selecting more reliable. I would definitely exclude Kampala, I struggle to see strong arguments against that and perhaps even other major towns as well.
As a side note I would also be interested to see a cost-effectiveness analysis (you may have published one already).
Anyway I’m super interested and think this is super cool.
Thanks so much Nick, this is really helpful and interesting. We do base eligibility on objective criteria, such as asset ownership, housing situation, consumption, etc. We hesitate to exclude folks based on very broad categories such as location (although we do restrict eligibility by district for research and logistics reasons). So far this has been working reasonably well, as you can see from the data above on income amongst the selected students. Very interesting point about the government schools, we’ll discuss that.
On the fraud, yes I’d be grateful for any advice you have!
Johannes
That’s interesting Johannes—I would be pretty nervous about income, and consumption questions yielding accurate results. I wouldn’t call family income or consumption measures objective criteria really as it relies on what people tell you—perhaps parental qualification level might be better if you do want to get an objective idea of likely income?
I might be confused but isn’t the data about income just from your own surveys, what is it proving exactly?
I think the example of 7 out of 30 students falsifying or attempting to results, is decent evidence of the lengths people will go to to get their kids or themselves onto a program like this.
If you are going to rely on subjective interviewing, I think the IPA validated poverty scores are pretty good, you might be using them already. Some of the questions are pretty regionally dependant so not super reliable (there’s a silly one about whether you eat rice or not, which of course doesn’t work if you already grow rice as a staple), but at least it is a researched and validated survey
https://www.povertyindex.org/country/uganda
On the fraud front sorry I missed the explanation, I would have thought you could just have candidates sit a standard English proficiency exam in the German Embassy? It seems like an easy fix thankfully.
One of your biggest risk as you grow I would say is in student selection. Your people employed in Uganda will have so much incentive to favour family members, take bribes for places etc. This is why in this case I would use as many truly objective poverty measures as possible even if it is not technically the “best” way of assessing poverty. The first thing I would do would be get photos/photocopies of the highest level of education of their parents and their siblings as this will give you a pretty good idea.
I would also seriously consider separating people who identify potential candidates and gather the poverty data, with the people doing the assessing (if you don’t already). I would have people submitting the objective information to a central office, which would then select people based on criteria. I would be interested to see your current criteria as well.
If it was me I would be excluding based on location to make selecting more reliable. I would definitely exclude Kampala, I struggle to see strong arguments against that and perhaps even other major towns as well.
As a side note I would also be interested to see a cost-effectiveness analysis (you may have published one already).
Anyway I’m super interested and think this is super cool.
Nick.