World's cheapest Tablet - The Aakash

As far as memory serves, the tagline mentioned above was (and is) used when ‘The Aakash Tablet was first talked about. Not being a gadget-goer at all, this post coming from me is rather surprising. But what really compelled me to write about this issue was a recent article I read some place – ‘what really went wrong with the Aakash?’

For those who would like a mention of its features - The Aakash is a low-cost computer with a 7-inch touch screen and 256 MB RAM running under the Android 2.2 operating system. It was in 2011, that India's IT minister Sh. Kapil Sibal announced the launch of the Aakash tablet for Rs 2,500 with much fanfare. It is part of the ‘One Laptop per Child’ idea. That moment, I recall my brother telling me with much excitement, ‘And I shall book this digital leveler ASAP’. (Well, he is a computer engineer after all!). And so, he became one of those 1.4 million users who registered for the gadget. (:O)

But I guess, that was the ONLY happy part. A series of blunders followed. First, there was an IIT-Rajasthan versus DataWind quarrel over specification issues, making headlines. According to a statement by Datawind's CEO, Suneet Singh Tuli, the company supplied 10,000 tablets to IIT-Rajasthan which were part of an initial order of 1,00,000 tablets that Datawind was supposed to supply to the government at a price of $49.98 per unit. But after only a few hundred tablets reached the students, IIT-Rajasthan started rejecting the Aakash tablet, allegedly after reports that the device was failing during pilot testing and did not meet the quality criteria set by the institute. IIT-Rajasthan was then removed from the Aakash project, and IIT-Bombay was then chosen.

The process became murkier when DataWind entered into a war-of-words with its ex-assembly vendor, Quad Electronics. While the ‘registered’ customers awaited the delivery of Aakash, the Hyderabad-based assembler of the tablet, Quad Electronics, sued DataWind alleging that the British Indian firm failed to procure its contracted inventory or pay for the tablets. DataWind retorted saying it won't procure any more tablets from Quad and would settle the issue legally.

And to top it all, the prototype turned out to be a disaster. Some phones in the market worked faster than this contraption. The battery wouldn’t last two hours if a user tried to play video files on it. The touch screen, apparently, wasn’t ‘touchy’ enough. Similar computing devices with superior capabilities have been brought out of Chinese factories but India seems to have lost the plot; what could have been an incredibly compelling story is now nothing but a stillborn.

What I’m trying to put across through this post is that almost three years ago, when we first heard of an ultra low-cost tablet to be launched by the Indian government, then known as ‘the $35 tablet’, India made it to the headlines across the world. ‘Steve Jobs innovates for the rich; this is for us’, one of my friends had remarked. So what went wrong – Is it all the quarrels? Is it the MHRD, which claimed to come up with an effective program in too short a period? Or is it the failure of the initial prototype? Well, it is ALL of these.

And now the news shows the launch of an upgraded version of the tablet. The Aakash II is reported to have improved hardware with Google’s Android 4.0 OS, 256MB of RAM and 2GB flash storage. And it shall hit the market around December this year. So, lets just wait-n-watch on this one before commenting any further.

 

Cut-offs : Fallacies of the critics

Just watched the following discussion on the Delhi University cut-offs (click here for the video). Apart from Chetan Bhagat's revealingly thoughtless accusations and assertions, there are several other clarifications that are in order. I am in no way defending status quo. However, what I am saying is that we could see the education system as a set of 'markets' or domains - the high school domain, the graduation domain, the post graduation domain and so on. There are several 'market failures' in each of the domain. Trying to set right one of the domains without correcting the failures in the other domains would be inefficient. The familiar theory of second best is inapplicable here because the failures in the other domains are not intrinsically immune to being corrected. With this in mind, I begin my discussion of the DU cut off system.

Firstly, the appalling suggestion by Chetan Bhagat that SRCC and other 'good' DU colleges should set up more campuses. At one level, it is hypocritical on his part ('excellence is exclusive', according to Bhagat. Click here for the video) to oppose dilution of quality in IITs and to suggest something similar in DU. However, I do not intrinsically subscribe to the fact that excellence is exclusive, and hence leave such a shallow argument at that. More importantly, to set up new campuses for SRCC, Hindu etc requires a whole lot of infrastructure. While we can talk about diminishing returns to a factor, we must consider what exactly is the limited factor we are talking about. In view of my previous discussion of market or (in the case of education, when 'market' doesn't sound all that right) domain failures, the domain failure that I see here is one of post graduation. 

It is the most obvious thing to any Economics student at Delhi University, for example, that even India's best economics degree suffers from a severe and crippling shortage of quality teachers. One of my co-interviewees at the Rhodes Scholarship interview confessed that even fresh postgraduates with no teaching experience and unverified aptitude would be invited to teach at one of the colleges. If we go about increasing the intake without considering such factors, we will just create a situation where Economics, for example, will be the new engineering - lots of third grade institutions that add nothing to the students. New colleges cannot be set up unless we increase the quality of output from our postgraduate institutions. Period.

Second is the system of cut-offs. I agree with critics that treating all state and central boards as equivalent is criminally objectionable, to state it mildly. However, we must realise that it is a court order that forbids colleges from taking out percentiles in each of the different boards (this is what I have read somewhere, but am unable to find a link right now). At best, Delhi University has committed a crime of inaction in not challenging such an order.

Secondly, to argue that we must de-link admissions from the Class XII boards is pointless. If admissions to even the liberal arts or pure sciences are de-linked from the boards, what is the purpose of having the boards in the first place? There is a domain failure in that the boards do not reflect the aptitude of students*. Do we correct it by making the boards more relevant, or do we correct it by having this pointless appendage called the boards, and then also have an entrance examination for various courses? Again, we must realise that there will be multiplicity of entrance exams, since an Economics (H) entrance exam will have to be different from a Philosophy (H) entrance exam and so on. The rot begins at the boards level, also because there is an incentive for state boards to inflate marks as as to better the lot of their students. Hence, basic economics demands that it is at this level that the problem must be addresses. Delhi University in this case is the victim, not the perpetrator.

Finally, we come to the suggestion of having interviews. Given my experience at St. Stephen's College, I believe that interviews are the best way of gauging suitability of a student for a particular course. I will not feign knowledge of why other colleges do not have interviews, for I am sure there must be some reason. In case the reason is some rule of Delhi University, then more autonomy must be granted to the 'institutes of excellence'. If there is some other reason, then we must address it.

A little footnote about media coverage of the issue. The way the media handles this issue year after year has something to say about the intelligence level of those running such shows. Remember the ruckus over the 100% cut-off at SRCC last year? Firstly, they failed to qualify the 100% cut-off. It was for science students who now wanted to pursue B.Com (H) now; hence, it was the result of a 4% gap between commerce students and science students that SRCC had committed itself to (this year's lower cut off, hailed by the media, is just a sleight of hand, with the 4% margin having been reduced). Even now, they say that the cutoff for science is 99.25% at Hindu, without qualifying that statement as to which course, and for students from which stream. Is such misleading coverage not worth condemning? Secondly, one student cleared the 100% cut-off this year. Even in my year when the science cut-off for the same course was 98.75%, only one science student cleared the cut-off (incidentally, my dear friend Manchit Mahajan). If one student cleared the cut-off anyway, how does it make a difference whether it is 98.75 or 100? The media certainly needs to be more intelligent.

-------------------

* - I have no hesitation in saying that my marks in physics and chemistry did not reflect my aptitude for the subject. The only justification I see is either that I worked hard, or that I severely over-estimate others' aptitude. However, hard work cannot always be a substitute for aptitude per se, as my CAT experience taught me.

Objectivism and Altruism

I just read this article that attacks Ayn Rand's objectivist theory at several levels. As someone who has barely read a quarter of The Fountainhead, I base my understanding of Rand on this (rather biased) article. Click here for reading the article.

Let me not even attempt a critique of Wikipedia claims was Rand's espousal of full, pure, uncontrolled, unregulated, laissez-faire capitalism. For even those with the most cursory knowledge of economics, it is not difficult to disprove such an assertion. There are so many market failures that unregulated capitalism cannot be optimal even in theory. Marx, of course, takes it further - he says that capitalism is a doomed regime that will one day fall under the weight of its own contradictions. In practice, the numerous crises have taught us that capitalism by itself cannot stand. However, trying to critique Rand at this point would end up becoming a lesson in basic economics, which I do not wish to detail here.

It is the second assertion, which I have accidentally dealt with earlier, that I will elucidate a bit here. As I have discussed in my previous article on norms and compliance (click here to read), many kinds of socially optimal behaviour can be made individually optimal using a set of incentives. In fact, according to the Folk Theorem, if an individual is patient enough, then any outcome in which the payoff to the individual is at least as much as his worst-case scenario, can be sustained. Given that a young person would not discount the future heavily, we can abstract away from the details and state that almost any social outcome can be sustained as a Nash equilibrium.

For our purpose of our argument, it is sufficient to say that Rand's argument that only selfish acts can be sustained, is false and myopic. Human society is a society of repeated social interactions. If interaction between any two individuals occurred only once, then Rand would be correct. However, since we need to interact at several times, we need to also consider how our action today would affect the state that will materialise tomorrow. While a more complete discussion can be found in my article I referenced earlier, let me give a simple example. When someone wants to photocopy my notes, then it would be in my selfish interests to not give those notes because the notes provide me with a comparative advantage that will help me score more. Sharing would reduce my comparative advantage. However, I might need notes in the future, and refusal to share notes today might jeopardize my chances of getting them in the future. Hence, the 'altruistic' task of sharing notes today becomes the Nash equilibrium.

Why I put 'altruistic' in quotes is because even this behaviour is selfish in that I still take the cost-and-benefit (to me) approach while deciding whether to share notes or not. However, the key thing to note here is that this selfishness differs from that of Rand in that it has a social context of incentives. In case the societal norms are removed and exchange of notes is made depersonalized (every person uploads his notes voluntarily on a network and others can download it, such that the uploader does not know who is downloading), then cooperation would break down and nobody would share notes. Instead, if anyone who refused to share notes was flogged, then everyone would share notes. The basic point is that seemingly selfish behaviour can also be modified and contained.

Going by my reading of Ms.Bekiempis' article, I can strongly assert that Rand was wrong - selfish behaviour isn't the only sustainable ones. Better outcomes can be sustained, and very often without the need for the Government.

Bollywood, Superstars and Signalling

Most of us with even the slightest affinity for sense and sensibility would cringe after having a look at the 100-crore club, arguably the most elite club in Bollywood right now. For every 3 Idiots and Ghajini, there are Ready, Rowdy Rathore and Bodyguard. The common thread running through these films is obvious - the presence of a 'superstar' - the three Khans, Akshay Kumar or Hritik Roshan. What is it that a superstar brings to the film? Why is it so elusive for films without a superstar to make it to this club, and why is it so easy for otherwise pathetic films with a superstar to make it?

Is it good acting? We would all agree that there are films with better acting than the ones listed above. Is it a good script? Except for Aamir Khan's movies, it can hardly be said about the others. Moreover, there are films with far better scripts that do not even aspire being in the club. I believe that the reason might be signalling. A superstar signals to the audience several things, that I elucidate upon for the rest of this article.

The first problem is the problem of asymmetric information. To see the power of signalling, take a simple example. Would you rather buy a television from a rickety, old, desolate shop or a swanky, new, branded shop in a crowded mall? I believe most would choose the latter. The need for some kind of guarantee that the shop will be around tomorrow to fulfill its warrantee is pressing. There are two similarities between this example and our movies example. Firstly, a television is a durable good, which we use very often. Similarly, we watch movies very often and would prefer to watch good movies than bad movies. Secondly, even though we have no prior information about the quality of the television, we would place our bets depending on the decor of the shop that sells it. Similarly, in situations of asymmetric information, we would tend to choose our movies based on the star. There is a certain standardisation that superstars bring - lots of song and dance sequences, good cinematography and so on. As risk-averse consumers, we would prefer to compromise on the expected quality as long as the variability in quality is reduced.

Secondly, there is the question of correlates. Producers who can afford to have superstars in their films typically can also spend a lot more on marketing than smaller films can. They can bombard our senses with custom-made promos. They can afford the most prominent billboards. They can hire the best musicians, get the most popular actresses to do item numbers and so on. Five years down the line, would people remember Agneepath more or Chikni Chameli more? My guess is the latter. In this sense, I look at films with superstars as behemoths. Yes, they do fail very often too. However, as a percentage, I am sure the chances of success are greater. At the very least, the music stays. Sheila ki Jawani, anyone?

The third and last theme that I explore (or rather, leave as a puzzle) is why the allure of the superstar doesn't wade after delivering several crass films one after the other. On one hand, I can assert that even Shah Rukh Khan needs a Chak De India to make sure that his stock doesn't sink too fast. However, how does one then explain Salman Khan's unfettered run? Why people would subject themselves repeatedly to such films escapes me. Maybe for a large majority of the people, the theater is an escape from reality, and a laugh, no matter how silly, is still a laugh. That requires us to dig deeper, something that I do not do here.

IITs, IIMs and the Signalling Race

The first time I read an interview with this year's JEE topper, I felt a certain degree of deja vu. At the end of the day, any 'success story', be it the boards, JEE, Civil Services or CAT, bears a strong resemblance to each and every other 'success story'. Of course, much of it is due to the media trying to highlight the same set of attributes, and hence it would be unfair to blame the toppers for being conformists. However, when I talked to my friends about it, something struck me as especially alarming. While we all know that much of the intake of our top business schools comes from the top engineering colleges, what was striking was that this year's IIT topper said he wanted to go to an IIM even before he entered an IIT.

The tragedy, unfortunately, is that it is unfair to single him out for criticism. Perhaps lots of students have the same aim of going through the IIT-IIM route. There is certainly a tremendous 'brand value' associated with this route. Going to the best institutions in the country in their respective fields is certainly a very difficult task. It is perhaps the most effective signal that a person can offer to potential employers in the labour market. I can also think of how this is the best way to humour one's ego. In an education system that is geared towards mass production (and here I am not being critical of India in specific), the fraction of students who can qualify for both these institutions is indeed very, very small. Truth be told, I would have considered it an honour to be in such elite company.

That said, what I am unable to grasp is why somebody who wants to do an MBA eventually would want to spend four years studying engineering. My experience at the IIM interviews this year has convinced me that it is easier for a non-engineer with the same aptitude (read CAT score) to clear the interviews than it is for an engineer. What sense would it then make for someone who eventually wants to go to an IIM to spend four years at an IIT?

There are two reasons that I can possible think of. Firstly, the person might intrinsically enjoy engineering as a discipline and even though he does not want to pursue a career in it, he might enjoy studying it. I am not sure how far this is the case for engineers, and hence cannot make a statement on it. Secondly, being in an IIT might expose a student to a plethora of activities that are unheard of in other institutions. For starters, the scale of cultural events at IITs is far beyond what one can even imagine at Delhi University. Being able to access such opportunities can help an individual hone one's skills and hence not only improve his career prospects, but also help him grow as an individual. Hence, even though I doubt how many students actually rationalise it in this way, it does make sense individually to follow the IIT-IIM path.

Does it make for a social optimal? Most signalling devices are pareto inefficient - they are a drain on resources that are now used to create signals, but could have instead been used for productive purposes. An IIT is an effective, but expensive signal. The kind of money spent on a single IIT student is exceptional for a country at India's level of development. Certainly, this is a case of individually sensible decisions to become socially sub-optimal. There is thus a pressing need to devise better signals that drain fewer productive resources on those whose eventual aim is to get management education.

Imperfections in Education

Here is a thought-provoking article by Manan Vyas about how complete education is a luxury:

http://www.mananvyas.com/thoughts/why-complete-education-is-a-luxury/

There is certainly no disagreement about the fact that for most of our population, complete education is indeed a luxury. The purpose of this article is to further expand on what Manan says, and also introduce certain disagreements with the referenced article.

While I have no disagreement with the advocacy of a dual education system, it must be noted that there are two more fundamental market imperfections at play here, one of which supports the broad-based education system that we currently have.

Before I deal with those those issues, a subtle clarification must be sought: I am a firm believer in the fact that education has an intrinsic value - not only does it get me a good job, but it makes me feel good about life. Of course, the article does not seem to disagree with that at any stage. It accepts this by calling complete education a luxury, in the sense that this intrinsic value of education is important but out of reach for a majority of our population. However, it must be noted that a good (in this case, education) is not a luxury forever. At some point, with rising incomes, it becomes a necessity. Hence, are we arguing that the dual education system will be progressively dismantled over time as complete education ceases to be a luxury? Why is it that Germany, a fairly rich country, would still find complete education a luxury? It is the most basic of economic phenomena - the Engel's law - which will ensure that education cannot be a luxury forever. Why then do we see this paradox?

Now returning to the two market imperfections that I referenced to earlier.

Firstly, there is the issue of imperfect capital markets. Why can't Sushil borrow to finance his children's educational expenses and then use the higher income potential to pay off this loan either himself or through his children? Of course, such loans in themselves are a luxury for most Indians. However, it is an aspect of education that I saw was not discussed in the article. Rather than provide the agent with an X (broad-based education) or Y (vocational education) choice, financial inclusion will provide him with a whole gamut of choices. This change is, of course, not going to happen any time soon, maybe not even in the next quarter of a century. Moreover, given the long gestation period of such loans, it might even be difficult or impossible to enforce such contracts. However, my point is that it is theoretically possible for Sushil to want to send his kids to school, given his low current income.

Secondly, there will be imperfections even if financial market imperfections are removed. There are benefits from a well-educated workforce that an individual agent is unable to internalise. Of the top of my head, a person who is able to explain much of how the world around him works will be happier and more motivated - hence more productive. Even as a financial analyst, it would make someone happier when he knows himself how electricity works. However, since these positive externalities are not internalised by the agent, there will always be under-investment in education. Hence, leaving the agents to make choices on their own will produce a level of education that isn't socially optimal.

Another problem that I realised while writing this response, is one of agency. If you propose that a dual education system be introduced, do note that in the presence of imperfect information regarding the aptitude of the child and uncertainty regarding the job prospects after a particular stream (since all industries go through phases of growth and stagnation), some subjective choices need to be made. Who makes these choices? In a society such as India's, these decisions will often be made by the parent. It is not certain, and not even likely, that the agent's (parents') interests coincide with those of the principal (the child). Once introduced to a particular stream at a young age, the child faces high costs of making a switch. In such a case, having a monolithic educational system till the age of 16 might not be a bad idea. I am not, in any way, saying that this WILL happen. What I am saying is that this is another aspect to be considered. 

Norms and Compliance

Very often, we find ourselves in a conflict situation - a conflict between our personal (and often, primeval) desires and social norms. By the very fact that 'norms' need to be enforced through a variety of legal rules, religious beliefs and psychological conditioning, it is obvious that breaking these norms would be the more obvious thing to do in the absence of such norms. I sometimes feel amazed at the superstructure of social norms that we have built for ourselves. To me, it is quite surely man's greatest creation. Everything else - including finance and technology (arguably two of the greatest inventions of the last century) - could not have been possible unless we got together as a society and built these norms that could sustain cooperation.

Some of the basic ideas of this short article are from a chapter in our course in Development Economics (which most of us decided to skip for the exams, though).  The idea is simple - if we as humans did not coordinate (and by coordinate I encompass things such as trust and cooperation), then even basic activities would be impossible. Take even a simple barter economy. You make handicrafts and sell it for food. What forces you to ensure that you are exchanging good quality handicrafts? Once you get the food, even if the handicrafts turn out to be of poor quality, you have already made whatever gains you needed to. However, we observe adherence to a certain quality in our society. This is caused due to the repeated nature of the interaction. If I do not supply good quality handicrafts, then very few farmers would be willing to exchange their food in the future.

Take something more private. Why should we be kind to others? Here I am not considering the 'feel good' associated with being kind. Often, the 'feel good' is a part of our psychological conditioning, and is in no way 'good' by itself. However, we would, generally speaking, still find it in our interest to be kind, because of the hope of reciprocity. If I am kind to people today, they would be kind to me tomorrow, and that would be beneficial to me. Similarly, faithfulness. If there was no stigma attached to not being faithful in relationships, then our primeval urges would take over. In this case, I do not quite see why human society would find faithfulness as a useful virtue. Probably a bit more thinking would make it obvious.

Now when you come to think of it, every type of human society, and every activity in human society, is based on such social norms. It is exactly like a repeated form game. The Nash equilibrium in any one game might be for all players to do the socially imperfect action. However, because human society is based on repeated interactions, we are able to sustain socially optimal outcomes even when there is an incentive for individual players to renege.

Take a simple example - driving. Consider that a driver has to actions - follow traffic rules and don't follow traffic rules.  If all other players are following traffic rules, then your not following traffic rules doesn't make a difference and hence it is in your interest to not follow traffic rules. In case all other players are not following traffic rules, you would be stupid to be following traffic rules. Abstracting away from the intermediate outcomes where some people follow rules and some don't, we find that it is in every player's interest to not follow traffic rules (in the language of game theory, a dominant strategy). However, the outcome where nobody follows traffic rules is socially imperfect. Hence, we have a system of rules and fines that make it optimal for every individual player to follow traffic rules. This is the only way that the social optimal can be sustained.

Think about it - where would we be if socially optimal outcomes could not be sustained? The need to make coordination optimal has made humans create such an elaborate system of norms. In one word, it is simply mind-boggling.

A natural corollary of what I have argued, and one that I do not always agree with, is religion. Religion is probably the greatest of these social norms that has survived centuries. What is religion but a means to ensure that all of us behave a certain way? Is it not religion that is the last-ditch effort to ensure that we place an intrinsic value on traits such as truthfulness, honesty etc that are otherwise only functionally important? At this stage, I would leave this argument as such and leave it to the readers to think about it.