4 ms·

Some people say giving money directly to poor people is best. Some people say paying poor families to keep their children in school is best. It seems to me th

by MarkMc 10y ago

Some people say giving money directly to poor people is best. Some people say paying poor families to keep their children in school is best. It seems to me that in this case a randomised controlled trial (RCT) would be an ideal way to determine who is right: A third of people get free money, a third of people get stay-in-school money, and a third of people get nothing. Then see which treatment, if any, is more effective at increasing the standard of living. But you seem quite certain that data from such an experiment would not be meaningful. What kind of problems do you think would prevent an RCT from producing an accurate answer? Could we avoid these problems by instead running the trail in, say, South East Asia, or are RCTs doomed to produce unreliable data?

PieterH 10y ago
No single local action can increase the standard of living in any measurable way. To even think in those terms is to utterly misunderstand the scale and depth of the problems that face Africans as they work to construct a functioning economy and its institutions. Do you even understand what creates a "standard of living"? It is things like, "if I start a business, and people don't pay their bills, can I take them to court?" If the answer is "no" then you are stuck with a cash economy, no credit, no trust. Everything becomes more expensive. How do you create functioning courts? You need honest politicians. How do you get honest politicians? You build coalitions of power that boot out the thieves. How do you build such coalitions? You spend decades slowly growing your economy so the commercial middle classes get wealthy enough to become politically relevant and interested. How do you grow your economy? You cut the cost of transport and communications. How do you do that? Etc. etc. I'm not giving a recipe, I'm giving examples of how complex this is. The scale is such that you'd need absolutely staggering amounts of money and political will to run studies that would show useful data. I'm talking entire regions, even countries, changing policies. Cases like Rwanda focussing on the Internet. Give it a decade, then compare to Burundi, which focusses on (e.g.) child health and political stability. Anyhow, there are no easy answers for this and forgive me if it sounds like I think there are.

by MarkMc 10y ago

PieterH 10y ago

No single local action can increase the standard of living in any measurable way. To even think in those terms is to utterly misunderstand the scale and depth of the problems that face Africans as they work to construct a functioning economy and its institutions. Do you even understand what creates a "standard of living"? It is things like, "if I start a business, and people don't pay their bills, can I take them to court?" If the answer is "no" then you are stuck with a cash economy, no credit, no trust. Everything becomes more expensive. How do you create functioning courts? You need honest politicians. How do you get honest politicians? You build coalitions of power that boot out the thieves. How do you build such coalitions? You spend decades slowly growing your economy so the commercial middle classes get wealthy enough to become politically relevant and interested. How do you grow your economy? You cut the cost of transport and communications. How do you do that? Etc. etc. I'm not giving a recipe, I'm giving examples of how complex this is. The scale is such that you'd need absolutely staggering amounts of money and political will to run studies that would show useful data. I'm talking entire regions, even countries, changing policies. Cases like Rwanda focussing on the Internet. Give it a decade, then compare to Burundi, which focusses on (e.g.) child health and political stability. Anyhow, there are no easy answers for this and forgive me if it sounds like I think there are.