Quick cost analysis: Assuming they get good programming jobs, you’d be getting at most, say, 10000 USD per year per kid or $100,000 USD per kid. A country low on this list_per_capita) has a GDP of under 5000 USD. Assuming you want decent facilities and educators, you’ll need, say, 3 times the GDP per student per year. If you’re giving them 10 years of education, that’s $150,000 in cost. This doesn’t work out even assuming a 100% success rate in getting them very high-end jobs. If you go for a very, very cheap place you might be able to get that to, say, $5000 a year in expenses per kid which works out if you get good success rates.
So this gives some obvious ways to get this to work:
You need to go for really as cheap a country as you can find and take full advantage of tech to reduce costs
More than 10% for 10 years might be necessary.
Alternative sources for funding—alumni donations are the current system most places use but would be weird to have on top of mandatory payments
Don’t educate them for 10 years or only do part-time education for some of it. (Can you give them the netbooks and have them study on their own for half the year while they live with their family?)
I hope you realized I meant that GDP per capita. Assuming you did and still think it’s high, here’s a more detailed estimate:
Let’s arbitrarily pick the 10th country from bottom of the GDP per capita list by CIA Factbook: Niger. Now we can make some more concrete statements. On average Niger women have 7 children and presumably raise them on 2*GDP per capita income. This gives 0.28 of the GDP on raising a child. (Infant and childhood death rates seem to be negligible for this calculation, perhaps increasing average spending by 10% when assuming they make it to age 15.) So assuming a high quality education takes three times as much as the typical education/child raising of the time we’d get about 1 GDP per child per year.
Niger happens to also have the highest birth rate in the world. Picking the country above it on the list, Afghanistan, gives a birth rate of 5 children/woman and would give us about 1.2 GDP per child per year. In any case, my 3 GDP estimate looks rather high here, but on the other hand so was my estimate that all* the children so taught would go on to work at Google-equivalent pay scales. I might then round my guess to 1.5 GDP per child per year when including extra costs like PR or administering tests to see who to admit that are done on more children than are educated.
If you think that’s still rather high, then you probably disagree with my estimate that such a program would want to spend three times as much money on education and child raising as is standard in the given country.
Aside: how do you make links here that include parentheses in them?
Quick cost analysis: Assuming they get good programming jobs, you’d be getting at most, say, 10000 USD per year per kid or $100,000 USD per kid. A country low on this list_per_capita) has a GDP of under 5000 USD. Assuming you want decent facilities and educators, you’ll need, say, 3 times the GDP per student per year. If you’re giving them 10 years of education, that’s $150,000 in cost. This doesn’t work out even assuming a 100% success rate in getting them very high-end jobs. If you go for a very, very cheap place you might be able to get that to, say, $5000 a year in expenses per kid which works out if you get good success rates.
So this gives some obvious ways to get this to work:
You need to go for really as cheap a country as you can find and take full advantage of tech to reduce costs
More than 10% for 10 years might be necessary.
Alternative sources for funding—alumni donations are the current system most places use but would be weird to have on top of mandatory payments
Don’t educate them for 10 years or only do part-time education for some of it. (Can you give them the netbooks and have them study on their own for half the year while they live with their family?)
3x GDP/student/year? That’s an absurdly high estimate.
I hope you realized I meant that GDP per capita. Assuming you did and still think it’s high, here’s a more detailed estimate:
Let’s arbitrarily pick the 10th country from bottom of the GDP per capita list by CIA Factbook: Niger. Now we can make some more concrete statements. On average Niger women have 7 children and presumably raise them on 2*GDP per capita income. This gives 0.28 of the GDP on raising a child. (Infant and childhood death rates seem to be negligible for this calculation, perhaps increasing average spending by 10% when assuming they make it to age 15.) So assuming a high quality education takes three times as much as the typical education/child raising of the time we’d get about 1 GDP per child per year.
Niger happens to also have the highest birth rate in the world. Picking the country above it on the list, Afghanistan, gives a birth rate of 5 children/woman and would give us about 1.2 GDP per child per year. In any case, my 3 GDP estimate looks rather high here, but on the other hand so was my estimate that all* the children so taught would go on to work at Google-equivalent pay scales. I might then round my guess to 1.5 GDP per child per year when including extra costs like PR or administering tests to see who to admit that are done on more children than are educated.
If you think that’s still rather high, then you probably disagree with my estimate that such a program would want to spend three times as much money on education and child raising as is standard in the given country.
Aside: how do you make links here that include parentheses in them?
LW wiki > Comment formatting > Common problems > Links
AFAIK, “per capita” means ‘divided by the whole population’, not just the adult one. (Am I missing something?)
[Like this](http://en.wikipedia.org/wiki/List_of_countries_by_GDP_%28PPP%29_per_capita)(28and29being the ASCII values for(and)respectively in hexadecimal).Good point and thanks for the tip.