Edward Cherlin
In the middle of an OLPC debate on the Bytes for All Yahoo Group, Edward Cherlin has estimated the cost of OLPC's for Nigeria. Edward starts off accurately with the number of Nigerian students:
UNICEF puts the school-age population (6-18) in Nigeria at 45 million. One-year age cohorts (children born in a year, less deaths before reaching school age) are currently 4 million.But then his numbers start to go awry. He uses the outdated $100 laptop cost. OLPC 2B1 Children's Machines are now $140 dollars per laptop, or $208 per laptop on the initial install. So if we correct his numbers, we find a stunning initial cost for Nigeria to equip all its students with OLPC 2B1's and buy more for each cohort:
If only Nigeria were so rich, Mr. Cherlin. According to the CIA Fact Book, Nigeria's 2005 budget had revenues of $12.86 billion and expenditures of $13.54 billion. An OLPC purchase for all its students would absorb 73% of its entire government income while supplying laptops for each cohort of students and replacing worn out computers would be 13% of the government income per year.$100$208 laptops for all starts at$4.5 billion$9.36 billion, with annual expense of computers for first-graders at$400 million$560 million. Presumably students will get new computers at some regular intervals, say in four years, so that they have new computers again when starting fifth grade and ninth grade. So that's$1.2 billion$1.68 billion a year. This is Nigeria, so that can come out of oil income.
Insert Nigerian 419 Scam here
And that is before we add in some of Mr. Cherlin's more esoteric suggestions:
For a start, Nigeria could have its own educational satellite (something like $300 million, including launch costs, and a few million a year to run it). So even the remotest schools wouldn't have to pay fees for Internet connections, although somebody would have to pay for the satellite receivers (VSAT currently, about $900 each).Finally, there is the real and immeasurable costs of the massive amounts of corruption and mismanagement that would occur. Nigeria may have improved its Transperency International ranking but its still one of the most corrupt countries worldwide.
Overall, I would have to disagree with Edward Cherlin's very optimistic conclusion:
But I think I have made the point that Nigeria could afford to do this.No, Mr. Cherlin, Nigeria cannot afford to spend 73% of its total government income in one year on buying 4 million laptops from Negroponte's One Laptop Per Child. Like Argentina's OLPC numbers, Nigeria does not have the budget for one laptop per child. Not by a long shot.
Wrong as usual Wayan, but it's funny anyway. I wish I have my employer to calculate my salary the way you calculate the olpc cost (using anything to inflate it ;)
How can you extrapolate the cost per laptop to $208 based on the Libyan MOU? Here is your logic:
..."the MOU apparently details an agreement of 1.2 million computers, one server per school, a team of technical advisers to help set up the system, satellite internet service and other infrastructure for $250 million dollars.
Or a total cost of $208 dollars per laptop."
How can you divide total project cost (include bandwidth for satellite connection, internet services and other infrastructure - which could include training teachers or even new teacher jobs, local school improvements etc) by the number of laptop and declare that this is the cost per laptop?
This is the budget that Libya is setting to undertake an important reform of their education system, not to buy hardware.
Thanks for fixing my numbers, but you have misstated the rest of my argument. I said that the $1.68 billion a year could come out of oil revenue, not the whole $9 billion (which is a one-time cost that could be amortized over a number of years, by selling bonds tied to future oil revenue). I don't see what is "esoteric" about buying a satellite and installing the needed ground stations.
In short, I stand by my conclusions.
btw The white farmers from Zimbabwe recently given land in Nigeria say that Nigeria could feed all of Africa. Apparently creating a good enough agricultural college is the main requirement. It is necessary to train farmers, research crops and methods appropriate for Nigeria, and also research markets, including the problem of providing adequate nutrition for everybody in Africa. And get decent communications out to the villages. So the Return on Investment (ROI) is huge.
ROI--that is a good point. I predicted in
http://www.olpcnews.com/countries/libya/2b1_cultural_integration.html that the little machines are going to wind up being used by everyone in the village, not just the students they were distributed to.
Presumably all those villagers could figure out economically productive uses, like getting crop price data, information on improved agricultural practices, and so on. So the laptops could wind up paying for themselves ten times over.
I know that is not what Negroponte and his gang were thinking of, but when you put incredibly flexible and powerful tools in the hands of a group of people, they invariably find innumerable uses you never anticipated.
It's "Argentina". Not "Argintina's" xD
LF:
"This is the budget that Libya is setting to undertake an important reform of their education system, not to buy hardware."
That's exactly my point in using the $208 per laptop figure. To implement the OLPC model, it will not just be $130 million for 1 million laptops. Millions more will be spent to integrate it into the educational system. Millions more that makes the true cost, the total cost of 2B1 ownership above $200 per laptop.
Ed:
While Nigeria could mortgage its future oil revenues to educate its children, is this untried, untested, and still unbuilt, future technology really the best option for $9 billion?
As Eduardo mentions, end users will come up with inventive income generating activities that we cannot even imagine - income generation that could exceed the value of OLPC's in the classroom setting.
But wouldn't Nigeria's education system benefit more from $9 billion spent on One Teacher Per Class, or better yet, One Well Trained, Highly Motivated, Very Respected Teacher Per Fully Supplied and Maintained Classroom?
As unfair as it is, it is possible that introducing a program like this will not happen at every grade level at the same time.
Are any of these countries considering applying this technology to say, three of the median grade levels? To start out with 4th, 5th and 6th graders perhaps, who will get a few years use of out them, are old enough to comprehend the concepts of technology and yet still young enough for this to affect their education by leaps and bounds?
Over the last few years I have worked with girls in the 4th, 5th and 6th grade levels and this seems an opportune time to introduce a technological and social network like the one these laptops will bring.
Sorry to say the older students will have less time with the technology, and the younger ones can be educated by the current means for a while longer.
As far as the math above... it would make sense to subtract the laptops for graduating 'seniors' at whatever grade level education there stops as they would be recycled back into the system.
Wayan wrote:
> [Cherlin] uses the outdated $100 laptop cost. OLPC 2B1 Children's Machines
> are now $140 dollars per laptop, or $208 per laptop on the initial
> install.
I'm sorry, Wayan, but I take back my thanks "for fixing my numbers". Thanks for trying, anyway.
The current price is not fixed for all time, or even past 2007. There are still plans to get the price down to $100.
The story on the Libyan MOU that you linked to to support this price
http://www.olpcnews.com/countries/libya/one_laptop_per_libyan_child.html
says
"Arising out of an August meeting between OLPC's Nicholas Negroponte and the Libyan leader Colonel Muammar Abu Minyar al-Qaddafi, the MOU apparently details an agreement of 1.2 million computers, one server per school, a team of technical advisers to help set up the system, satellite internet service and other infrastructure for $250 million dollars."
So $208 is not the cost of the laptop. It includes all of the other human costs and infrastructure, much of which is made up of one-time costs. You can't extrapolate that to the full system cost, particularly if you don't allow for economies of scale and the manufacturing learning curve. In addition, some of these costs are for outside experts. The target countries will train their own experts during the initial rollout, and have much lower costs going forward. After a while, schoolchildren will train themselves to be the experts, and costs will drop further.
Furthermore, my figures allowed for all the extra expenses. I took as a starting point someone else's claim that we would need to spend as much as the cost of the laptops again, and considered what that would buy. For example, more teachers with more trainincg; schools and roads; a national fiber optic network with more telephone switches and Internet routers, local DNS servers, resolvers, and so on; putting all government services on the Web; digitizing the national literature; licensing textbooks for electronic distribution; translations; and a lot more.
So yes, it costs more than the bare laptop, and yes, it is more than $100 per laptop currently, but no, my figures are not off the mark. They were labelled as estimates, since so many costs are variable or even unknown today.
Here, let's try another. India, with cohorts of 23 million annually, or about 250 million of school age.Let's suppose that the costs run $200 per child, including laptop plus a portion of infrastructure, teacher training, and whatnot for the first few years, but that it comes down to, say $125 after a time, certainly by the time the first cohort trained on these computers starts turning out teachers, techs, and the rest. Let us also suppose that children get three laptops during their schooling, in first, fifth, and ninth grades.
So the initial layout can be in the low nine figures, as in Libya (a million units plus the rest of the bundle for $200 million). A full rollout would come to $30 billion over a number of years, with annual costs thereafter of $9 billion.
Substantial, but not impossible. Now, what is the return on this investment? Trillions of dollars per year in new economic growth. Parity with Europe and the US on a per capita basis would be a GDP of something over $40 trillion dollars/year, which will take several decades to achieve. The first country to do all this would set a world record for national economic growth. I'll take a deal like that any day. If you were a government that cared about people, wouldn't you?
Let's see, if every country did that, it would come to at least a quarter quadrillion $/yr Gross Global Product. Wow!
On October 20, 2006 Wayan wrote:
> Ed:
> While Nigeria could mortgage its future oil revenues
> to educate its children, is this untried, untested,
> and still unbuilt, future technology really the best
> option for $9 billion?
Come, now, Wayan, you know perfectly well that nobody does it that way. Nigeria would start with the same $200 million package that Libya signed up for, and wouldn't do the full rollout until it had a handle on the training and infrastructure issues, and had an adequate curriculum and a set of textbooks in at least Yoruba, Igbo, and Hausa. The project would then have to ramp up over a number of years, with manageable increases in expenditures from year to year.
> As Eduardo mentions, end users will come up with
> inventive income generating activities that we
> cannot even imagine - income generation that could
> exceed the value of OLPC's in the classroom setting.
Well, I'm glad that you and I agree on something.
This leads to another important point--rising tax revenues to pay for more of the program.
> But wouldn't Nigeria's education system benefit more
> from $9 billion spent on One Teacher Per Class, or
> better yet, One Well Trained, Highly Motivated, Very
> Respected Teacher Per Fully Supplied and Maintained
>Classroom?
No. And not only because that wouldn't provide the immediate increases in income that you yourself pointed to above. Rather, the Laptop program would be the fastest way to increase teacher competence. The Laptop is better and much cheaper than a well-stocked library and a set of old-fashioned science labs. I would personally give up chairs and desks before a computer, and I expect that the schoolchildren will agree with me once they see what the Laptop can do.
Posted by: wayan on October 20, 2006
Edward Cherlin said:
>No. And not only because that wouldn't provide >the immediate increases in income that you >yourself pointed to above.
"Immediate" is not the word that springs to my mind when trying to demonstrate the merits of an OLPC roll out over other types of education spending - such as building/repairing schools, equipping classrooms and training teachers. Both endeavours would provide results, neither would provide immediate results. Moreover, when in life does "faster" = "better" - especially when politicians and large sums of money are involved.
The laptop project is _not_ going to provide an immediate increase in income for subscriber countries and Wayan didn't say it would. If the project works, OLPC target countries should benefit financially from a better educated population, but repairing schools, improving student/teacher ratios and properly funding education systems would be a different means to the same end. That's not to say that either method is "better" or "cheaper", but one of these methods is proven and the other unproven.
The laptop scheme - in its current "cards close to the chest iteration" - is essentially a high stakes gamble using education budgets as collateral. Is OLPC's scheme worth risking a generation's education for?
Edward Cherlin (also) said:
>Rather, the Laptop program would be the >fastest way to increase teacher competence.
How did you come to this conclusion? I have seen very little evidence that OLPC has anything to do with teachers - I thought the plan was for the kids to automagically teach themselves with their little green godesses.
If anyone could point me in the direction of evidence to the contrary, I'd be grateful - I'd like to write about OLPC and teacher training as it raises interesting pedagogical issues - particularly the use of real-time collaboration in the classroom.
"But wouldn't Nigeria's education system benefit more from $9 billion spent on One Teacher Per Class, or better yet, One Well Trained, Highly Motivated, Very Respected Teacher Per Fully Supplied and Maintained Classroom?"
Substitute Nigeria above for any other country, and you have the main reason why this project is doomed from the start.
Interesting article over at IT & Telecom Digest about Intel vs. OLPC in Nigeria:
http://www.ittelecomdigest.com/cover.htm
the nigeria numbers are crazy. billions for laptops. nothing for teachers. brazil? india? they would need to mortgage the country for these computers. chinese computers. where is local manufacturing? local solutions by local people?
Hello Wayan,
Thank you for calculating these numbers. However, there's one huge glaring error, based on what I'll assume is your and Edward Cherlin's (and perhaps Unicef's?) western mindset. You agree with Edward when he assumes that everybody between the ages of 6 and 18 is in school:
You wrote:
> Edward starts off accurately with the number of
> Nigerian students:
>
>> UNICEF puts the school-age population (6-18) in
>> Nigeria at 45 million.
Perhaps Unicef puts the school-AGE population at 45 million, but the population IN SCHOOL is far less than that. The majority of Nigerians today get an elementary school education (ages 6-9) and no more.
In the 70's when I was living in Nigeria, the government launched the "Universal Primary Education" program to attempt to get every child access to an elementary school education. That program failed miserably.
More recently, Nigeria launched another similar program "Universal Basic Education", to attempt to get everyone access to at least an elementary school education.
The last actual figures that I could find for how many children are ACTUALLY in school is from 1999, when there were 16.8 million students in public elementary schools and 1.9 million students in private elementary schools. That's about 18.7 million students in elementary school. In contrast, in 1999, there were 4.4 million students in post-elementary schools.
That brings us to a total in school of around 23.1 million.
(Cites: The Guardian Online: May 6, 1999 and Victor Dike: http://www.africaeconomicanalysis.org/articles/gen/education10204234737htm.html )
That is far far less than the 45 million that you are accepting as fact.
Please correct this false assumption--your article will be far better when using accurate numbers.
Trevor Z.
Trevor,
OLPC is not targeting children in school. They are targeting ALL children in a country, precisely because so many are not in school. They ask governments to buy enough XO's for One Laptop Per Child, not one laptop per child in school.
The post's assumption is correct and the numbers are accurate.
But you do bring up a major problem with OLPC's idea: How can we expect the Nigerian government to equitably distribute laptops when it has not been able to equitably distribute other resources?
Amazing. I checked the OLPC website: http://laptopfoundation.org/participate/ and you're right. How could an organization based in the Western world expect to distribute laptops to kids not in school? That's worse than ridiculous, it's comical in a very sad way. The naivete' boggles my mind. Is some American planning on trooping out into the bush of Nigeria on foot with a huge backpack filled with laptops?
In that case, I withdraw my critique of your article, and move it to OLPC. They need to make their goal something more reasonable, like one laptop per SCHOOLchild.
Trevor
Forget your Western myopia, Trevor. We expect *Nigerians* to trek into the bush with laptops, the way they deal with distributing River Blindness medicine. (See Rx for Survival, a PBS series)
And the way the Sinhalese and Tamil members of the Sarvodaya Movement handle village development in Sri Lanka, with a government equally uninvolved with and indeed hostile to the welfare of the poor. Sarvodaya has built a school in about half of the villages in Sri Lanka, or rather, has taught the villagers how to build and staff their own schools. And clinics, and roads, and village banks, and...
http://www.sarvodaya.org/
More OLPC Nigeria budget calculations for XO laptops, this time from Allen on another OLPC News post: http://www.olpcnews.com/implementation/plan/technology_implementation.html
I've edited it for clarity & present it below:
To really see these numbers in perspective, lets compare expenditure on education as part of the GDP.
The gross domestic product of Nigeria on Purchasing Power Parity
2006 $188.5 billion
needed for XO: $2.3B ~ 1.2% GDP
http://www.airninja.com/worldfacts/countries/Nigeria/gdp.htm
Total expenditure as a % of GDP for OECD countries (2003)
http://www.oecd.org/dataoecd/7/32/37344685.xls
(watch out, an Excel file)
Examples (all education combined, 2003):
Germany 4.7%
USA 5.7
Poland 5.8%
Finland 6.5%
All OECD 5.5%
Brazil 4.7%
Chile 3.7%
Israel 7.0%
Russia 3.7%
In this light, tripling the Nigerian expenditure on education might be a no-brainer. If they have something useful to spend it on. And the politics behind it might also be a problem. Note that the last president intended to spend more oil money on education and research. I have no idea about the current administration.
But as the OLPC cannot deliver 13 million XOs in a single year, this investment would be spread over several years.
Winter
Some of the comments above sound interesting, but all seems to forget that nigeria is a federation of states, each with its own budget and priorities.
Best plan will off course target individual states, while some states may be interested, others may not be.
Emma, thanks for pointing that "little" detail. A perfect, good example of how most of this discussion lacks some grounding in the realities of the countries that are to become participants.
when all is quiet like this, I hope the Siemens are not wilbrosing some officials to overlook the technical implication of the olpc