The Strange Case of One Laptop per Child

Charles Dickens invented the term “telescopic philanthropy” for people, like his Mrs. Jellyby, who are preoccupied with the lives of people in faraway places and ignore their own children and others at home. There is another type of telescopic philanthropy, where the telescope is aimed not far away but firmly at one’s own navel. Consider Nicholas Negroponte, a computer scientist at MIT, who has established a charitable organization to help poor children in the developing world. What would a computer scientist think that poor children need that they currently lack? Food, medicine, clean water? Or maybe a laptop computer?

Negroponte’s organization, One Laptop per Child, has developed a laptop that costs about $100 to $200 to manufacture. It has few features but can survive the rugged environment that OLPC imagines poor children live in. According to one press report:

The laptop is designed to withstand harsh conditions such as rain and dust. It has a screen that can be read under intense sunlight. Its battery lasts for 12 hours and can be recharged with the use of a solar panel or a pull cord.

More details here.  The OLPC website says that its mission is "to provide children around the world with new opportunities to explore" and that the laptop will enable "nations of the emerging world [to] leapfrog decades of development."

According to the website, children need to “learn learning,” and a “computer uniquely fosters learning learning by allowing children to ‘think about thinking,’ in ways that are otherwise impossible. Using the [laptop] as both a window on the world, as well as a highly programmable tool for exploring it, children in emerging nations will be opened to both illimitable knowledge and to their own creative and problem-solving potential.” This opportunity will be given to children in even the “most remote regions of the globe.”

So let’s pick a remote place—say, Mali, where almost three quarters of the adult population make less than $1 per day, the adult literacy rate is 46 percent, 60 percent of children attend primary school, and 15 percent of children attend secondary school. Will laptops be given to the illiterate children who don’t go to school? Computers might teach children to learn to learn, but one would think that a prerequisite for learning to learn is being able to read what’s on the screen—schools are still necessary for children to learn to learn to learn. Indeed, OLPC does not hand out laptops to children; it deals with education ministries. So the only children who are going to receive laptops are those lucky enough to be going to school in the first place.

Most of the children in places like Mali grow up to be subsistence farmers; many others will end up as unskilled laborers in factories. Even if laptops teach these people to “learn to learn,” it is unlikely that such a skill will provide much value for them where opportunities are so limited. Of much greater value, aside from medical care and the like, would be basic literacy, rudimentary training in some locally useful skill, or—for that matter—access to roads that are in reasonably good condition.

Still, if the education ministries of governments want to buy cheap laptops for children, that’s fine. But this raises another question, which is why commercial firms wouldn’t supply such laptops if, in fact, they have so much pedagogic value in poor countries, and government officials realize this.

The answer is that commercial firms, in fact, do supply such laptops. Intel, for example, has manufactured $400 laptops, and has sold a bunch of them to the Mexican government. Intel’s laptops cost twice as much as OLPC’s, but presumably the Mexican government thinks they are more than twice as valuable. OLPC laptops are cheap because they have few features—maybe too few features for Mexican students, while too many for the students in Mali. Other companies also offer inexpensive laptops. One might imagine that different laptops in different countries have different features, reflecting differences in local conditions. A Mexican child who studies in school and at home does not necessarily need a laptop that is waterproof, can be read in the glare of the sun, and has a solar powered battery.

OLPC is wilting under the competitive pressure exerted by Intel, and now complains that Intel is unfairly disparaging its product, and shouldn't compete with OLPC by offering governments a less cheap but perhaps superior alternative.  It's hard to credit this argument: the education ministries are perfectly capable of evaluating the different products, and if they prefer Intel's, they should not buy OLPC laptops.  (And if the ministries are incompetent, then it hardly matters what the market looks like.)  OLPC has also complained that Intel sells at below cost in order to drive OLPC out of business. But Intel has to compete not only with OLPC but with all the other for-profit companies that, like Intel, believe they can make profits by selling cheap laptops. Selling below cost is suicide in such a wide open market.

But isn’t it a bad thing that Intel is making profits off the sale of laptops to poor countries? OLPC, by contrast, is a nonprofit. But nonprofits that make products no one wants don’t deserve any praise, even if they don’t take a profit. There is a great deal of competition now in the cheap laptop market, so it seems unlikely that Intel is making enormous profits anyway.

All of this raises the question why anyone would think that charitable organizations should be in the business of designing and manufacturing computers. Real charitable organizations can buy computers from companies like Intel and donate them to children in poor countries if they think that’s a good idea; there is no reason why they should buy worse computers from a nonprofit company. The profit incentive will encourage entrepreneurs to develop yet cheaper and better laptops. In the same way, nonprofits buy food, tractors, seeds, and other goods from for-profit companies, and then donate them to poor people in poor countries. Why should laptops be any different from these other products?

Of course, if people set up manufacturing businesses and want to forgo profits, that is their business. Prices will be marginally lower, and that is a good thing, unless quality suffers. However, it is another matter when they ask people for donations, as OLPC has. OLPC has offered the following deal: buy two laptops for $400, and you get one; a child in the developing country gets the other; and you get a tax deduction for the $200 that can be attributed to the donation. Anything wrong with this?

Yes. You don’t really want that $200 laptop.  (In fact, a quick google search reveals that if you want a cheap laptop, you can get quite a number of more feature-rich laptops for as little as $300. They won’t have pull cords for recharging the battery, but you didn’t really need that feature, did you?) So you are paying $400 for a laptop you don’t want, and for a laptop for a developing-world child who might want it, or think he wants it, but would really be much better off with a decent education, clean water, clothes, food, and so forth. For $400, you can do much better, for yourself and for poor children. Set aside $200 toward a laptop that you could use. Then give the other $200 to (say) Oxfam, which says that it could do the following with it:

$20  Buys enough maize to feed a family of four in the Rift Valley of Ethiopia for six months.

$30  Buys books to help 10 girls in Afghanistan learn to read and write.

$50  Grows substantially more crops in a region of scarce resources, harsh terrain, and prevalent malnutrition in Peru.

$100 Provides a young student living in poverty in Mali with the vocational training and financial support necessary to start her own weaving business.

$200  Provides disaster preparedness training and technical support to two families in El Salvador.

It takes little insight to see that laptops would be low on the list of priorities of the developing-country poor.  One Laptop per Child makes as much sense as One iPod per Child or One Snowmobile per Child.

The last word should go to Nigeria’s education minister, Igwe Aja-Nawachuku. Nigeria initially ordered $100 million of OLPC laptops, but later canceled the order.

What is the sense of introducing one laptop per child when they don't have seats to sit down and learn, when they don't have uniforms to go to school in, when they don't have facilities?

Comments

[minderbender]

I think that at the time OLPC was conceived, there weren't cheap laptops suitable for use in the developing world. In fact, one version of the story has it that Intel is marketing its laptops in the developing world to preempt open-source software. I don't know if that story makes sense or not - ask Professor Picker.

I'm skeptical that $100 is better spent on a laptop than on other resources . See here:

http://www.povertyactionlab.com/projects/project.php?pid=6

and here:

http://www.povertyactionlab.com/projects/project.php?pid=21

and here:

http://www.povertyactionlab.com/projects/project.php?pid=2

However, I think there is considerable value in experimentation, perhaps on a smaller scale than Negroponte envisioned. Some people believe that computers, and particularly the internet, have a huge impact on the dispersal of information and democratic participation. These people could be wrong, but it might be worth a few million dollars to find out if they're right. If there's a big externality here - if there's option value in waiting to see how it turns out in some other developing country - you can see why developing countries might experiment suboptimally.

See also:

http://fakesteve.blogspot.com/2007/09/finally-customer.html
http://online.wsj.com/article/SB119586754115002717.html?mod=home_we_banner_left

[DWAnderson]

This post reminds me of a friend who worked for a nonprofit doing micro credit loans in the Dominican Republic. When a for profit institution started making such loans on more favorable terms, instead being happy that it could move on to another market, the nonprofit was really irked that it was being subjected to competition.

[Kimball Corson]

What real good is a laptop without an ISP or other form of internet access? A bit like having a phone that is not connected. Overstated, to be sure, but I make my point.

[jason Nolan]

I guess it is best if we 'provide' food and not resources to foster independence? The OECD in a report I read almost a decade back on what was then called the digital divide, clearly stated that in their opinion that the access to safe drinking water and basic education for women were the two priorities for the OECD in terms of balancing the digital divide. There is no question what is important. But why stop there? If there are organizations trying to meet these needs, we should and do support them. There are also other needs, such as learning. Many of these countries do not have the ability to develop first language learning materials, let alone print and distribute them. Many do not have an encyclopedia or knowledge base in their first language (which wikipedia is working towards helping with). Most do not have access to the tools to share their stories with themselves, and perhaps the world around them. Computer are now what basic printing and publishing would have been when Caxton and Gutenberg popped into the scene.

I could go on and on... but if your goal is to keep people in a ghetto, then just send them food. If you want to give people access to the resources that might over time lead them to develop and learn, give them access to basic resources and add the ability to learn and construct their own knowledge and share it with others.

IMHO of course.

BTW, the laptop doesn't require an ISP... it uses mesh networking.

[jason Nolan]

"The last word should go to Nigeria’s education minister, Igwe Aja-Nawachuku. Nigeria initially ordered $100 million of OLPC laptops, but later canceled the order."

If you do a bit of research you can see the international pressure going on there. Oh, and they realized that if children have computers and can communicate with each other, people will realize how Nigera's government is screwing people on the oil issue. This quote supports the value of an OLPC, because in the end it will be come an uncontrollable voice for whatever people think is important to them. :)

[guava]

The statement "OLPC has few features" is ill-founded. The review that you linked to, specifically, evaluates the OLPC's merits on criteria like web-surfing and online video rendering. The XOs are designed to be educational tools more than an internet browser, and the reviewer would have done well to explore some of the laptop's other features (of which there are many). The reviewer was comparing the OLPC to a modern laptop, that he wanted for watching videos on youtube, which is not an intended use of the XO.

For example, and this ties to the "laptops without ISPs" comment, the XOs are connected together with a mesh network such that two (or more) laptops in range will have communications between themselves even in the absence of a connection to the world wide web. For kids wishing to interact with each other, this is arguably enough. Morover, when one of those computers is connected to the internet, it is as if both of them are.

[Ted Lemon]

So let's get this straight: your position is that it's better to buy people fish than teach them how to fish? That it's better to have a school and uniforms than to have learning? Have you even looked at the OLPC's feature set? It's really cool - I'd much rather have an OLPC than one of those Intel laptops. The main advantage of the Intel laptop is that it runs Windows, which presumably is how OLPC has been FUDded out of those contracts you mentioned.

Sure, the free market is a great thing, but sometimes someone has to think outside the box, and Intel isn't doing it in this case. IMHO Negroponte's big mistake is that he isn't trying to fund the OLPC project out of commercial sales.

[james119]

I think you're missing the point, Eric.

It is a mistake to compare the OLPC to the computer you have sitting there on your desk. The aim was to create a computing device that addressed a specific set of needs, not a traditional "personal computer."

Engineers meet challenges like this all of the time; take for instance the Tivo, which is nothing more than an underpowered PC running a customized version of Linux. According to you, I shouldn't have purchased my Tivo because it is a "worse" computer than a traditional desktop computer. But darn it, try as I might, as much as I point my remote control at my desktop computer, it never does record Law and Order for me. Hmm...

[P3T3RK3Y5]

There are some of us that believe an idea is more important than a physical object. This notion even spawned a movement... they call it the "Information Age".

You’re not seriously suggesting we give a man a fish?

The OLPC is completely sufficient for its purposes (in my mind, systematically combating poverty and injustice), particularly when you see the google’s webapps overtaking microsofts apps; and wikipedia the equal of britannica in accuracy, not to mention blogs and other media hosting sites. and it all makes sense because the third world is quick to build wireless networks - which is why everyone has a cell phone.

what Intel did is immoral.

[Haja Sheriff]

While I have my doubts on OLPC and their model, I do believe that OLPC has done what few other companies have done. At a time when the PC's were selling for $1000 or more, OLPC announced (though without doing due diligence) a $100 Notebook. It caused a paradigm shift which has brought us to a level where you have Intel Class Mate PC's & Asus EeePC's which are all in the sub $300 range.

OLPC also has developed technologies, which will be of terrific use in the coming years.

So we should do give credit where it is due.

This not about selling a $100 notebook (incidentally it is up at $200 now)to the poor children, but to move a HUGE segment of the population from dire poverty a few steps up - and giving them a vision of what they could be doing. To do that, one cannot operate in a silo mode, but work together with the Governments, build content, build an ecosystem of local partners who will deliver the content, and with other agencies .These as a team need to help oversee the next steps to get these students experience technology and move up the value chain.

Technology does play a key part in helping peoples lives.

[D Wu]

Not all third world nations are alike and I think there needs to be a distinction made here when talking about OLPC

I think the target for OLPC is not poverty-stricken nations like Mali but rather developing nations which has enough social infrastructure to support laptops in the classroom, hence the target being education ministries. The goal is not to combat poverty but rather give the next generation a leg up in crossing the so-called digital divide.

OLPC has been key in spurring Intel to develop their Classmate competitor, but as is the case with almost every non-profit I've dealt with, as soon as competition enters, they throw a hissy fit and look for the UN or the government to referee. People in non-profits have this terrible mentality that they are doing good (they are - mostly) but only they can do it and only they are doing it for "goodness sake". Frankly, it's insufferable.

Word to the wise, if a resume crosses your desk and the most recent posting is a non-profit - Pass.

[alexandre van de sande]

There are some wrong assumptions there:

1-That the OLPC product is inferior to a normal laptop. It's subjective, but it has some features that puts gives it competitive advantage when comparing with any "for-profit" laptop

2-That a laptop bought from a "real computer manufacturer" would suffice. There was not a laptop in such a price point before OLPC announced it's own, and maybe there never would be. Also the focus on low power consumption instead of higher processing power helped benefit the whole industry. The MIT is a research institute, not simple charitable organization. The greatest donation of OLPC was MIT brainpower.

3-That a laptop should be low in a country priority. Of course other infrastructure investment, such as roads are equivalent important, but education, access to information, ability to collaborate and communicate is not to be underestimated.

[David Rand]

The importance of a profit margin that establishes a value chain cannot be understated. Local entrepreneurs--who have a vested interest in making sure that systems are up and running, maintained, and repaired when broken--create local wealth and opportunity as well as providing a platform that helps educate the kids. Look at NComputing's virtual desktop products and business model. It is working everywhere from the FYR of Macedonia to North Carolina, and about 70 countries in between.

[Steve Fullerton]

You are obviously an educated person, but I don't think you understand the philosophy of the OLPC Foundation. It is not about laptop machines it is about education.

Education is a big part of the solution, and in some cases the entire solution to most of the developing world's problems, be it poverty, peace, pollution, political unrest, global warming, etc.

The OLPC is not designed for Darfur --- Only God and Jimmy Carter can help them. They need food and shelter before they can think of eucation.

The OLPC is for emerging countries exactly like Peru where there is some food, educational infrastructure, and stability but a lack of educational resources.

Brazil pays $20US/year for crappy textbooks. The OLPC can hold 500-1000 textbooks and an OLPC school server 100x that.

In the Peru pilot, providing OLPCs brought children back into the school system.

50% of the world's children live in rural areas that lack electricity in the home. the current Intel machines and alternatives won't work for them due to power consumption requirements --- although the OLPC hopes they will fix this.

As for your examples:

$20 Buys enough maize to feed a family of four in the Rift Valley of Ethiopia for six months. (This is the job of the UN and World Food Fund. Italy is funding OLPCs for Ethiopia)

$30 Buys books to help 10 girls in Afghanistan learn to read and write. ($3 per book; OLC 500 books/$200=$2.50 per book, but they can be updated over the internet to always be current and tailored to the curriculum, even if weaving or road construction is the curriculum)

$50 Grows substantially more crops in a region of scarce resources, harsh terrain, and prevalent malnutrition in Peru.(education will allow them to apply and customize food production techniques to dramatically increase the economic efficiency of food production)

$100 Provides a young student living in poverty in Mali with the vocational training and financial support necessary to start her own weaving business.(ibid)

$200 Provides disaster preparedness training and technical support to two families in El Salvador.(ibid)

Applications are being developed for TB sputum testing via plug-in microscope to an OLPC and electronic transmission via internet to pathologist anywhere in the world. Like wise for Malaria blood testing. Think about that.

Again --- Education is a big part of the solution, and in some cases the entire solution to most of the developing world's problems. If you give a person a fish, they can eat for a day; if you teach a person to fish ...

[JP Chilumula]

Dear Prof Posner,

I believe there are some underlying assumptions in your case against OLPC. I might ramble and digress a bit, but let me take a stab. Most of this would be India specific, but may apply to other countries as well.

1. Are services Mutually Exclusive? Are these others cost efficient?
"What would a computer scientist think that poor children need that they currently lack? Food, medicine, clean water? Or maybe a laptop computer?"
Are these mutually exclusive? Does delivering clean water exclude food? Delivering education doesn't exclude any of those. There are other people working on those areas. Why should everyone work in delivering food, medicine, clean water and nothing else?

Besides, these services are resource intensive and not cost efficient. Hundreds of billions of dollars have been spent by Govt over a period of decades and nothing was accomplished on this front. The cost benefits of delivering these services to villages doesn't look good. New cities need to be built to bring in efficiencies of scale.

2. Prerequisite: Ability to read
.. "but one would think that a prerequisite for learning to learn is being able to read what’s on the screen"

Most kids enroll in school, but they drop-out and they would meet the ability to read prerequisite that you mention.

"Of the estimated population of 205 million in the age group 6-14 years on March 1, 2002, nearly 82.5 per cent was enrolled in schools. However, drop out in 2002-03 at the primary level was 34.9 per cent and at the upper primary level, it was 52.8 per cent." - http://en.wikipedia.org/wiki/Literacy_in_India

Kids drop-off because the curriculum is most uninteresting and parents find no value in that kind of education. A lot of parents are more than willing to pay for quality education, but Govt's have a vice-like grip on education and do not allow private players. I'm sure you are aware that all information channels to people are controlled by Govts. The primary information delivery mechanism is education. Private players and freedom of information (access to thousands of books, instead of 8 - 10 Govt propaganda!) doesn't auger well for the Govt.

Next, there are other technological efforts in making people literate. Take a look at Same Language Subtitling.

3. Naive understanding of Govts
"the education ministries are perfectly capable of evaluating the different products, and if they prefer Intel's, they should not buy OLPC laptops"
I believe your understanding of Govts is naive (so is OLPCs!). Govts are least bothered about education, literacy and such. Their sacred duty is steal money as efficiently as possible. Why do you think almost all money is spent in defense, irrigation projects and such? They have established efficient workflows to transfer money from those transactions. Intel's laptops and others would be favored, because there would be kickbacks to the officials. Say, Intel goes to an education minister and says, here is a laptop, lets charge it $400, you get back $100 (cash, no direct-deposit or mail-in rebate!), education minister would love that. Say, Negroponte goes to the same minister, talks passionately about education and all that, it won't work. There is no incentive for the minister. Surely, you know the power of incentives. Negroponte is too naive. It would have been a remarkable success if he offered the OLPC at $1000 and gave a margin of $600 to the minister. We would have seen it deployed everywhere.

The ministry of education in India said OLPC is pedagogically suspect. Of course, Govts are always suspicious of projects that give them little room to steal.

I feel that Negroponte should have taken a different route, like the one taken by Tesla Motors, build something cool and froody for the richest and then figure out how to make it cost efficient down the line. First, build a laptop for US children. It is going to sell easily in US, considering that US Govt spends about $1200/month/kid. Then, it would have been an easy sell to third world countries. "Look! Every kid in US has this laptop! We should get one too to become like US!"

4. Assumption about what children would do in future
"Most of the children in places like Mali grow up to be subsistence farmers"
In abt 2 - 3 decades, there won't be any water in majority of India to do farming. -- http://technology.newscientist.com/article/mg18925401.500

5. Assumptions about education

"What is the sense of introducing one laptop per child when they don't have seats to sit down and learn, when they don't have uniforms to go to school in, when they don't have facilities?"
Precisely the reason why countries should invest in OLPC. There will always be a scarcity of resources to invest on these not so useful things. Do we stop investing in textbooks when there is no furniture available?

Also, how do uniform and chairs/tables relate to education? Is there a correlation between learning and uniform? Is there a causation?

That one needs to be taught stuff in schools by a teacher (who knows stuff) is deeply rooted in Christian theology, delivering God's knowledge to the masses. Kids learn much faster on their own, by interacting with the environment and other people. Unfortunately, this rapid learning process is killing by the school-teacher teaching methodology. Kids lose all capacity to learn, think and be creative.

[Grubbs]

People are retarded. It takes TIME for the Third World impoverished to learn how to farm, for the crops to grow, for the businesses to get set up and for the kids to get their education. In the meantime, they need food and medicine, ne? Also remember that the human brain CANNOT learn without proper nutrition!!! Just giving kids breakfast before school increases their learning! Men not only need to learn how to fish, but they need fish until they can catch them themselves.

So the point is that $200 can go a long way towards both FEEDING families and EDUCATING them to become more self-sufficient.

Just because the OLPC isn't a good idea doesn't mean its detractors are heartless monsters.