This article is ridiculous. I think it would be better titled "Education is Inappropriate for India"

The author essentially makes the argument that India can't afford high quality education. Hiring tons more teachers, buying millions more blackboards, books, etc. would require zillions of dollars as well as a great number of years of capacity-building. Hiring 25% more teachers would take 5-10 to build up to. And good luck getting teachers from Bangalore to move to the poorer parts of India like Bihar or UP ;)

Improving education is very expensive __period__ whether you are in India or Western Europe.

I agree very much with this perspective. A quick search of the web and a few back-of-the-envelope calculations confirm your point: if India devoted its entire education budget to buying XOs for all K-12 aged youth, it still wouldn't have enough money to accomplish this task.

The idea of one-computer-per-child as a solution to the educational challenges of countries like India is simply a pipe dream at this time, and, if taken seriously, would hard other more urgent areas of priority, such as seeing that all children can receive at least a basic primary education.

The XO or other computers can still play a small role in Indian education, but not on a one-computer-per-child basis.

India is a big enough and complex enough country to need or consider my opinion on it's educational system and how better to start introducing changes. So, I will not "offer" any.

However, I was wondering if a conflict of interest disclosure might be in order given that the author is an economist of a mobile oriented company (SMS sender company). "Free" netbook (and associated services) spreading could be seen as a threat/competitor to this business.
A statement that XO deployment is in no way conflicting with netCORE Solution Inc short-term interests and long-term planing, would be nice to have here. The author should be directly involved or at least aware, of both and could provide a credible statement.

@atanu: "There are well-known reasons for why the Indian education system is a failure but they will not detain us here. What is relevant is that none of factors have anything to do with technology. It is not a technical problem that leads to the dysfunctional system. So technology cannot be part of the fix that the system needs."

Let's say that the "unmitigated failure" you cite elsewhere in your post means that all aspects of the system are lousy. If we then abstain from proposing a "technological solution" for these problems, but instead look at the affordances of technology (communication, access to information) as enabling elements in in more appropriate solutions, does technology potentially play a role. Examples: administrative solutions (better school and teacher accountability), teacher performance (increased access to resources and professional development), student performance (increased access to better teaching, information, productivity tools) would technology play a role? (And in this instance we would almost certainly not be talking about buying laptops for all the kids, more something on the order of increasing the information and comms available to all school participants through much smaller installations.)

(I haven't mentioned assessment, which is obviously a huge part of the problem, because, frankly, IMHO technology doesn't even have the potential to play a real-world facilitative role. And without change in assessment, one is essentially without change in education.)

if the seeds sprout many may thrive,many may not. those that flower will spread more seed.

Author wrote:

"One of my primary arguments against the OLPC program in India is that it is out of sequence in the sense that there are other more pressing important problems with the education system, which not only will not be helped by technology, but indeed that the diversion of resources to the OLPC will exacerbate the existing problems."

That's exactly the reason most countries will no adopt the program. It is not a solution to their problems - they must solve their many problems before even thinking of implementing OLPC's ideas.

OLPC would probably be more efficient in healthy education systems.

I have always wondered why there is not a single wealthy nation that believes in one-computer-per-child programs ;-)

This is a powerful argument that OLPC is not appropriate in India - at scale, at the current time. However, it ignores the fact that prices change. It is ridiculous to imagine that 100 million XOs would cost $200 each; economies of scale would reduce the cost significantly. And those reductions in netbook prices are coming anyway, one way or another. There will be a day, within well under 10 years, when individual netbooks, especially in lots of tens of millions, will be price-competitive with low-cost textbooks. India should be prepared for that day. With 100 million kids, a pilot program of tens to hundreds of thousands is, quite simply, an appropriate preparation.

I understand that the situation is always more complex on the inside. Corrupt interests can favor cosmetic solutions as a way of distraction, and that should be opposed. Yet that should not extend to a blanket opposition to a relatively small pilot program just because it isn't the most-urgent reform.

I was reading "the life you can save" by peter singer. He points out that : if you have 100 people facing death and you offer to save 80 through aid people feel quite good. If you save 10% of 1000 people, people are less inclined to give, yet more people are saved. My point is doing something is better than doing nothing!

Sure it would be nice if all children had access to a decent education system, and food etc. etc. but we should do what we can and not hold back because we cannot help everyone

The base cost of XOs for 100 million children works out to be approximately $20 billion. That does not include recurring cost of use and ownership, such as replacement, repair, and support. That could add at least 20 percent more, or $4 billion recurring per year. That's more than the entire public budget for education in India.

That's underestimating. The Flicking mind mentions quotations in several parts that upkeep like support, teacher training, etc. should be closer to equal the cost of the purchase of the technology.

There's infrastructure costs: Installing and maintaining internet access. Servers, power (some places XOs have been distributed can't power them all at once), utility bills, and support in the form of helping with problems (hardware/software), finding new applicable applications, managing OS updates, help incorporating into curriculum, etc.

Someone mentioned high volumes driving down prices... XO-1 has been out for a while and the price keeps going UP.

Atanu is right, and he is wrong. He is correct that India cannot handle olpc as it stands today.

However, the X0-2 is planned to sell for $75, and we can expect prices to go down from there. Eventually a large portion of the Indian population, which is so poorly served by its educational system, is just going to buy XO's on their own for their own children to use. If the government is hopeless, then the solution is to simply go around the government.

I have made this point before a number of times for other cases. No one has agreed, and no one has presented reasons it is wrong. Is there some sort of rule here against thinking outside the box?

Oops, I see that kiko likes the idea

http://www.olpcnews.com/countries/peru/olpc_peru_far_from_goals.html#comments

But I still would like to get some feedback from other people.

India and China have got to insert into the digital age all of their children or we are all doomed. It is not complicated really:

Laptop + Internet = Better Society

If India, China, Obama and others start ordering millions, it'll accelerate the development of XO-2 which thus will lower the cost per laptop below $75, increase the battery life x10 to then really start making the OLPC goal, to every child in the world, much closer to reality.

Obviously, Atanu Dey is right. But he is wrong too. It is a matter of perspective.

Indeed, India's spending on education went from "3.23 percent of GDP in 2000-2001 to 2.88 percent in the recent times"
http://en.wikipedia.org/wiki/Education_in_India

In contrast, the developed world spends around 5% of GDP on education (OECD average). And in India this

That would also hold for money spend on XOs. To get an XO in every child's hands could end up costing $1000s per laptop in "overhead" and "extra costs". With a very large risk of most laptops never reaching the children anyway.

So far, Atanu Dey is right.

Where he is wrong in my opinion is about the future. At some moment, India will be forced to chose between improving education or falling back into poverty. Economic growth requires an educated workforce.

And the most efficient way of improving the efficiency (productivity) of the system will be by using technology, more specific, ICT. At that point, the OLPC, or some other program like it, will be in focus again as a fast way of getting information, communication, and modern teaching methods to teachers and children.

But I do not think Atanu will mind be proven wrong here.

Winter

So, bypassing government education ministries seems appropriate. OLPC should open more channels for non-governmental distribution of XO laptops, including retail sales of quantity one.

I like the anecdote about learning from a hundred slim textbooks. I myself got thru highschool before I ever touched a computer. There are many aspects to education, only a few of which can benefit from expensive technology. Laptops require a lot of infrastructure and tech support. Maybe we are working on the wrong project if we want to improve education for the world's poorest children. Maybe we need to start a new project: One Textbook Per Child (OTPC), to be followed up eventually with Ten Textbooks Per Child (TTPC), then One Hundred Textbooks Per Child (OHTPC). Eventually ebooks on a laptop will become practical.

Your article is an excellent reality check, Atanu. Thank you.

As I said before I will not “offer’ any suggestions.
It seems to me that the whole argument is that the money for few pilots, in the case of India this may translates to ~1 mill laptops, will be better spend in books and other necessities and that the educational system has so many fundamental “malfunctions” that no positive outcome can result from these pilots.
What would be nice to know is how are we SURE that the money, if not spend on laptops will be spend on books or blackboards, as opposed to not spend at all or spend in an even less productive/more elitist activity. Given the questions about India’s educational policies and practices, this would be hard to exclude. Some info here would be handy (even the link to the respective blog entry :-)
For the second point, it may be so. I wouldn’t know, though I could assume a positive influence on the recipients’ education at least. But placing computers and communication devices in rural areas even with barebones infrastructure may result in unexpected activities beyond education. I remember reading stories of businesses set-up around a single cell-phone. Imagine a connected laptop. Even if 1% of these machines go to rural areas and the rest goes to middle- high- class schools, the “fringe benefits” can be major.
Obviously if all the machines go in areas to fill an educational gap, instead of “replacing” or “improving” a functioning setting in the expense of others, then the benefits are obvious in the education front too.
Is there any indication for the proposed initial deployment scheme? This may at least tell we are eating someone else’s cake or we are feeding more peopel with the same one (even if they might get a more generous piece).

Obviously my point is not if the XO is a good investment or value for India right now, but that no hardware or software company wants it spreading in this emerging market, exactly because of the praises that you offered for the machine and the program (in a different setting of course).
Don't you agree with that?

Some of the benefits that OLPC gives is probably overzealously pushed by Negroponte. No One knows Moore's law better than himself. The great promise he is jumping about is just a 2 year leap in moore's law by not buying the $400 laptop from a dell. 2 years down the line the same config from dell could cost $200. Of course dell would not probably like to make a 50$ device to protect its laptop market. But cellphone companies the android phones etc would probably do that job. So mr negroponte is just helping u skip a about 3 years ahead if everything was really smooth.

The bigger problem is that he is demanding a $20 billion experiment. You cannot experiment with 1/6th the population of the planet just to see if some idea works. Even a company like Google first tests out ideas on its own employees, then does a preview then a beta and then a final product.

Stuff like Nepal, Peru etc. are like these large scale betas. It takes TIME to see if such tests delivered on their promise. While I am aware of how bad the indian system is, it probably is better than that of rwanda/nepal. So if it can work wonders in those places, it may have promise in india too. But to see if works, you have to wait at least 5 to 8 years.

The pilot projects have passionate teachers. The average teacher still would probably care little about education. So overzealous claims by the pilot projects would also need to be taken with a pinch of salt.

By the time we all know whether such an experiment on such a large scale is really necessary, by moore's law, the cost of the olpc or competitors would have come down to say $10.(Just 2+2+2+2 8 years away). Moore's law facor of 16... It would no longer be a big deal and any parent would buy it for his kid, coz a computer that costs Rs 400 or Rs 300 by 2016 would be worth every pisa even if was just an ebook reader. Buying a something that also serves as a mobile phone would be more than enough of a deal for even the poorest.

Politicans in our country pull off gimmicks like distributing tvs around to voters. No I am not joking. This was an election promise of one of our state governments which was implemented. I do not think a laptop for $10 in 2015 would be a big deal. So even if olpc vanishes away, I'd suspect every citizen of India would have a laptop by 2016.

India's education system is not a nimble web 2.0 like company. It is a huge monstrous crawly dinosaur like creature. Dreaming that someone can barge in and turn it upside down in no time is just day dreaming. Taking 5 to 8 years to deploy olpc in india is as good as allowing market forces to sell the laptops to consumers including rural consumers directly. And if that is the case I would prefer my tax money not blown away in a multi billion $ experiment.

But I do hope a lot of open educational content would be generated through various initiatives including olpc which would be accessible to students across the globe when having a laptop would be a no brainer.

Right now I hope India uses a wait n watch policy with some pilot projects to check out some pedagogical claims.

I would suggest to visit the schools of the Parikrma Foundation in Bangalore once http://www.parikrmafoundation.org/home.htm and to check why Shukla Bose understand IT education is so important http://www.youtube.com/watch?v=5wl5eXQyf4w

So: Atanu, can you clarify by stating what (if anything) would be the minimum things which would have to happen before a pilot program would be appropriate? (No magic pony requirements, please.) And when that happens, what would be an appropriate pilot program size for India? (Keep in mind, there are economies of scale even in pilot programs.) I suspect our points of view are closer than you think.

I suppose if the opportunity cost of the pilot program were zero, it can be anywhere and in any size.

Let me give some examples of when the opportunity cost is not zero. Suppose the Indian government spends $10 million in a pilot program. That money could have alternative uses in education -- such as fully educating all the way to high school 10,000 students. (That is 10 years times 100 dollars times 10,000.)

That argument holds regardless of whether the Indian government paid for it (from my tax money) or whether a very high net worth individual or some foundation donated the money -- the money can be used more efficiently in providing education to a lot of people.

To do a real test, I suggest that two pilots -- one with the OLPC and one which uses text books and teachers but spends the same per student. That would tell us something.

In any case, I am rehashing things that I have written about at length on my blog. If you are interested, please do take a look at Formula for Milking the Digital Divide and OLPC Rest in Peace.

As Bryan says, this post is based on a ridiculous premise - that because a system is broken, applying an otherwise valuable change is 'inappropriate'.

Opportunity cost is a fine way to assess new projects. One of OLPC's strengths is the way it draws attention to education, and reminds people and policymakers alike of the importance of focusing on young children to ensure they stay in school and get something out of it.

You claim that funds spent on one initiative related to education are necessarily drawn from a fixed education budget, which is regularly not the case. OLPC (here for once I mean the capital project, not the lowercase global movement) has the iconic capacity to inspire a change of vision for what a solid education entails, including how it is supported across a country.

You criticize the cost of a defined and tangible project -- whose very public details allow you to do so even without doing too much homework -- by comparing it with a vague and unquantified class of alternatives.

Your claims that notebooks and blackboards are more effective uses of resources, all else being equal, is interesting but unproven. Can you point to groups doing such work and rough estimates of their costs? Certainly there are corner cases where this is true, and corner cases where laptops hit a sweet spot. Is it a 40/60 ratio? 99/1? Direct school improvements suffer from higher overhead and corruption costs, are restricted to changing student experience in the classroom, and do not provide significant new incentive to attend school.

You could make the argument that India is bubbling over with students who need nothing but access to a dozen more books, hot meals, and a larger blackboard to realize their potential and graduate from highschool. If you have evidence to support this, that's a real incentive to find out how inexpensively you can print out a billion books and install a million room-sized slates and erasers.

But in many cases (including I expect some parts of India) this is not all that is needed. Children need a way to effectively learn on their own outside of school. Or they have books but feel they must pay tuition to actually 'get an education'. In which case you might argue that a massively distributed tutoring network is the lower-opportunity-cost solution, and one that will strengthen social bonds at the same time. Another fascinating and possibly compelling argument needing real assessment and definition of a solution.

You are good at bold comparisons; it is hard not to be impressed by your writing. But I am put off by the hand-waving. I do not think you will improve education without spending more time researching solutions.

You overstate the cost of repair and maintenance of laptops (or at least of XOs in particular), and do not estimate the cost of books and blackboards. You compare the cost of a 40,000 laptop deployment with that of education 10,000 students from K through 12, and assume that the latter will lead to more high school graduations.

> I suggest that two pilots -- one with the
> OLPC and one which uses text books and
> teachers but spends the same per student.

Let's get specific. You said that the average cost per student per year is $100. Assume that you can get your choice of laptops whose total cost of ownership is $300 over the course of 3 years (an XO would be significantly less and last longer, but let's make this impersonal and assume you may not like green.) One school receives a laptop for every child, that the children own and take home, and you can spend another $100 per student per year at the other school however you like. What would you spend it on?

SJ write, "As Bryan says, this post is based on a ridiculous premise - that because a system is broken, applying an otherwise valuable change is 'inappropriate'."

No, that's a complete misreading of the argument. The argument is that the system is broken and that it needs help. The significant point is that any proposed solution has to meet the budget constraint. The XO fails to meet the simple arithmetic test -- there is really nothing much more to be said if it fails that basic test.

The XO is not anymore a solution to the broken Indian education system than proposing low-cost BMWs is to India's transportation system. I defy anyone to prove that a very low-cost BMW is not a better solution for the needs of an individual than say a bicycle. But I also defy anyone to do the arithmetic and show that collectively the low-cost BMW is even remotely affordable compared to a mix of transportation alternatives such as bicycles, autorickshaws, buses, light rail, private cars, etc. If subsidies have to be provided to a large number of people who cannot afford any transportation at all, the first goal should be to see that at least everyone has access to a bicycle, and then if money is left over, then to help others who do need more costly transportation modes, even up to subsidizing low-cost BMWs for some.

My point is that subsidizing low-cost BMWs for a few is silly if that is at the cost of not providing a very large number of people bicycles.

Of course to makers of low-cost BMWs, that a huge number of people will not ever get bicycles is of little interest. They just want to sell as many low-cost BMWs as possible -- and damn those who are left on their own.

SJ wrote, "Your claims that notebooks and blackboards are more effective uses of resources, all else being equal, is interesting but unproven."

And your claim that giving a few a costly device at the expense of denying a very large number something that has been proven for a very long time (the whole of the 20th century) is proven?

SJ wrote, "You could make the argument that India is bubbling over with students who need nothing but access to a dozen more books, hot meals, and a larger blackboard to realize their potential and graduate from highschool."

I could make that argument but I don't. Nobody needs "nothing but" -- everyone would like to have everything. Given a choice between -- (A) a few good textbooks, basic hot meal, a blackboard, and a competent teacher, and (B) a laptop, internet access, excellent meals with huge servings, and all the rest of it -- nobody would choose (A).

Arguing that (B) is not better than (A) is asinine, and reading that argument into my piece is equally so.

The claim is that (1)there is a set of alternatives all of which have the potential of serving the need, and (2)that one should choose the best alternative which meets the budget constraint.

To restate my point for the n-th time, I claim that the XO solution lies outside India's budget constraint.

To makers of XO, I would ask them to consider the low-cost BMW scenario -- since they are likely to be more objective about BMWs than about the OLPC project.

SJ wrote, "You are good at bold comparisons; it is hard not to be impressed by your writing. But I am put off by the hand-waving."

Why, thank you, SJ, for your kind words. If you knew me, you would not dismiss hand-waving. It's what pays my bills. But before I add all the hand-waving, I do the arithmetic so that I don't sprout patent nonsense. The hand-waving bit is just to keep my customers entertained because the result of all the arithmetic is often oh so boring.

I hope this is my last comment on the matter in this thread.

I have been nothing if not sincere in my piece that the XO is a great laptop and will be very useful in education. I don't think that point is ever in doubt. As someone deeply interested in education, I am delighted that the OLPC project exists and I wish it all its deserved success.

It is a great pity that every student in India cannot be provided with a laptop. The undeniable sorry fact is that India is desperately poor.

Once in an airline magazine, I had seen an ad which said, "If the solution to Africa's AIDS epidemic was just a glass of clean water, Africa would still be out of luck" (or something to that effect.) Expensive drugs to combat AIDS in poor populations is out of the running from the get go.

Somewhat analogously in India, even if the solution to India's education challenge were just some basic schools with low-tech traditional teaching tools, even then India would be out of luck. High technology gizmos for the 100 million or so who need education is out of the question.

Funny how a country with so many billionaires that also spends ~$15 billion/year educating (200-300k) kids abroad can not afford textbooks and blackboards because of the "badged constrains" of its $10 Billion education budget (100mil x$100).
Are the two related by any chance?

Isn't it time to think outside "budget constrains"? With $100/student-year, NOTHING can be done. Splitting the breadcrumbs better or adding few more is not going to feed anyone, specially when there is apparently plenty of food to feed everyone. India is a country with a huge number of poor people but not necessarily a poor country. Arguing that this or that can not be done because all we have is $10 Billion for 100 million people, what it effectively does is KEEPING THE BUDGET at $10 billion.
Maybe is conceived as the prudent path. But can a modest increase to 10.5 or 11 billion change much? Adding another 2 books or a teacher every 1000 kids or a school for every million kids, is it going to revitalize India's education?

It would appear that is a matter of political priorities (OK, corruption). If it is how do you change it/
"Revolution" is a way but is a pretty messy and unproven solution.
Here is an "appealing to the system" alternative. Instead of spending more money on teachers (many, for a looong time and no-one profits), books (many authors, low tech, labor intensive) buildings (many constructors and labor intensive) etc, have local manufacturers build the computers, with a guaranteed marker of 100 million units and profit margins of choice. More money for education may "suddenly appear" under these conditions.

So an OLPC-like project may actually be more likely to increase "education spending" than the other labor-intensive/low(no)-profit options. If we assume that we have a corrupt system not really interested in education, the only way to improve education spending is as a by-product. If "trickle-down" economics is implemented in India, an OLPC-like project could be its implementation in education. Build (primarily) computer manufacturing companies and (as a side project) improve kids education.
The only problem is if you do not want to build any computer manufacturing or in no way want to better educate your people. Is this the case? I hope not.
Of course there is always "revolution" or wait for another 25-50 years for things to improve and the $100/student, become $200 or $300. Till then India we'll live with the "budget constrains" that God-forbid, can not be changed.
Maybe an OLPC pilot is not such a bad idea after all.

So the basic premises against OLPC in India are as follows:
OLPC/XO is an excellent idea/implementation but not for India (even for a pilot) because:
India is too poor
Implementation of even a pilot will cut money from more effective venues
Increase of the Education budget is out of the question under any circumstances
India government and particular the education department is way too corrupt to be trusted with anything

Although a lot of Q&A when on after the post, 3 of my question went unanswered
"no hardware or software company wants it (XO/Linux) spreading in this emerging market. Right?" Asked twice. The first time a misleading answer was given and the second was just ducked.
"how are we SURE that the money, if not spend on laptops will be spend on books or blackboards, as opposed to not spend at all or spend in an even less productive/more elitist activity?" ducked
And the one above (looking for a way out of this corruption hydra) "Which social groups will enforce the required changes in education and beyond (obviously against the will of this corrupt government) and how?" No response 3 days now.

Obviously he can not answer everything but did answer most, including:
Al Quaeda is a non-profit organizations (in response to value of non-profits),
He is happy with FOR-profit products (in response to XO being non-profit),
The middle and upper class of India is spending $15billion/year to educate their kids abroad because the corrupt government does not allow corporations to open private Universities (in response to the dismal education budget).
A government with a 200 billion total budget (in a 3+ trillion economy) "reportedly" has 1.5+ trillion in private personal accounts abroad (in response to distribution of wealth in India),
Government and corruption is THE problem in India (India stand above the middle in the corruption list and only the second time in India's recent history, the prime minister just got re-elected).
As long as the government is involved, nothing can be done in India (in response to locally produced laptops as an "incentive" for educational budget increase).

I did search his blog for answers since my requests for a relevant post were also ignored. No relevant post there too. Found posts like this though:
"Dismantling the bureaucracy would be the first step to fixing the problem of corruption in India, followed by reduction of the public sector"( http://www.deeshaa.org/2003/10/21/corruption-in-india/ ) or
"Liberalization of the economy has given us some gains but certainly not enough liberalization has been done. What the government has to do is to reduce the interference of the government in the economy so that the economy can be truly free to grow." ( http://www.deeshaa.org/2007/06/13/fake-pms-speech-part-teen/ )
Which does provide a clear view of his preferences and priorities.

I also search Google Scholar to see what makes Atanu Dey expert in Economics, Education or even plain Politics (besides the 4000+ blog posts of course). Nothing. Only his PhD. No peered review publication and no citation by the people in the respective fields (2004-2009). And there are plenty of peered reviewed publications in the field (7000+ only in his PhD subject).

So despite the abundant and smooth writing, the fundamentals, the howto and the prospective are still missing, other than "the market will fix everything by itself" mentioned above (and by now, we all know how this ends up).
There is no doubt that India has major problems but it would appear to me that in its core the opposition to OLPC pilots is mostly based on "Karma Raegamonics" and good old political opposition. There is not necessarily something wrong with either, but lets be clear about it.

"In Indian politics, you should not believe before you see."
So why not see an OLPC pilot then?...

PS: I do check Google Scholar before I go to an MD (The near by hospital is a University Hospital!) and those who don't look for a second opinion for anything remotely serious

Please, let's have fewer ad hominem attacks. Here are the facts that I think we all agree on.

1. At current prices and budgets, giving XOs to a significant fraction of India's students is unfeasible. I think we'd all agree that the gap is somewhere between a factor of 3 and a factor of 20. This takes as given that the majority of any budget must go to pay teachers, build schools, etc., but leaves plenty of room for disagreement on the exact numbers.

1a. Corruption makes that worse. I'd say it multiplies the budget gap by a factor of less than 2, and reduces the effectiveness of almost any alternative including laptop programs, more books, or more teachers.

2. Laptop prices are going down. Moore's law's 18-month doubling time does not quite apply unless you keep a constant price point, but even if power:price at a given time follows a power law, it is still exponential growth with a doubling time of 3 years or under.

3. India's education budget per student can be expected to increase in absolute numbers, even if it decreases as percent of GDP.

4. We do not have good data as to the relative effectiveness of laptop programs relative to other comparable spending options.

Given these undisputed (AFAIK) facts, the numbers say: laptop programs will be economically feasible in India at scale in something between 3 and 12 years. It is in India's interest to have good data on relative effectiveness at such a time.

I argue that facts 2 and 3 support a pilot program. Atanu Dey argues that facts 1 and 1a weigh against one. I don't see that arguments about credentials help resolve this argument for any reader.

"Give people a voucher for the amount to be spent on education."

Wow. And I had thought that this argument had reached the point where nothing new was going to be said.

This suggestion is based on the (extremely ideological) idea that markets are the best solution to any economic problem. There are several obvious ways this fails in the case of education. First, education is, by definition, a case of asymmetric information, which is a known cause of market failure. Second, and more germane to the current question, markets systematically under-invest in public goods - such as educational research or open source software. Third, markets do not respond to externalities; for instance, high inequality and low social mobility are negative externalities which could easily be incidentally increased by a voucher program (especially given assymetric information).

In fact, when someone is completely oblivious to all of these points; points which are not only obvious, but which are also already explicit and central to the discussion; I begin to doubt whether they're open to non-ideological discussion.

In other words: sure, under such a system, no poor Indian would buy a laptop today. And then, when laptops become price-competitive with schoolbooks, a whole generation of students would have to blindly choose which they think is better, and blindly decide how to best use the laptop if they choose it. This is obvious and I've said it three times now, avoiding it is in fact the whole meaning of the words "pilot program", and if you can't address it then there's no point talking.

As far as opportunity cost is concerned, India it seems is still able to tolerate the forgone lifetime productive potential of an educated child in exchange for the dehumanising exploitation of child labour. Official estimates range from 20 million to 50 million child labourers. that's about 5 to 12 times more Indians than those who have access to broadband internet! so these choices are political. By even asking the question: why don't we do more for our children? we are asking policy makers to justify the status quo, a task that once undertaken begins the work of shifting the status quo.

In Oceania, where I work for OLPC, we are framing OLPC as a catalyst to encourage policy makers, governments and bureaucracies (both state and non-state actors) to FOCUS on delivering better quality education, while at the same time stimulating community level demand for that better quality.

If politicians in India were educated (and most weren't morons) this statement "Public spending in (on?) education is in the low single-digit percentages." wouldn't be true.

I do agree with the author though, money could be used better on essentials as opposed to the OLPC which is, relatively, quite expensive when you crunch the numbers. Consider this - most state education boards disallow the use of scientific calculators (and just use log tables and data sheets) because a lot of students cannot afford one.

Going through the comments on this post, I'd be curious to know if opinions have changed in the past year. With the price of devices falling, is there a way of improving education with them? e.g. E-book readers are already dropping in price. So far that just means more textbooks in a small package, at a price. What if the few really motivated had access to devices that could help them learn on their own (I know it comes down to content, but that is not an unsurmountable problem) or to complement what little they learn at school, including answering tests, comparing scores and so on. That would be one way of playing out the right to education dream into reality.