India cannot bank on IT

I was just thinking yesterday, how the economies of different countries evolve, and how each economy manages to find its niche, its [[wp>Unique_selling_proposition|USP]]. The [[wp>Japanese_economy|Japanese]] had efficient manufacturing ([[wp>Just_In_Time|JIT]] etc) and expertise in miniature electronics, robotics and automobiles; the Chinese have over-flown world markets with low cost goods and are investing tremendously in manufacturing infrastructure across the board; the Germans have their automobile industry; Finland has Nokia and so on. And I realized, that so far, the only niche Indian economy has been able to find (in a big way) is basically IT.

Now there’s nothing wrong with IT, but just that in the long run India can’t bank on it. I had written this before, but didn’t fully understand the implications — India needs core infrastructure and manufacturing to really move forward. See the problem with IT is that its primarily a Human Resource based industry, rather than a core skill/infrastructure based industry. I mean, its not like Infosys and HCL and TCS bank upon hordes of brilliant hackers to write their code. Thats what [[wp>Six_Sigma|Six Sigma]] and other such certifications are for — if you take a whole bunch of people, put them through excellent training programs, put in elaborate procedures for design, architecture, code reviews and testing (basically practice good software engineering), then you’re unlikely to mess up.

However, IT does not build much in terms of core infrastructure. I mean sure companies will be laying down fibre across the country, all homes will have broadband connections, Internet access over cell phones will become easier, cheaper, faster and so on, but these are not the same as say good public health infrastructure, roads and railways, power generation and distribution, iron and steel.

One of the key differences is that knowledge transfer in software is very easy. If you give me the functional spec of a software, I will most likely be able to build it. But if you give me the functional spec of an electro-mechanical device or a chemical compond, its highly unlikely that I’ll be able to figure it out by myself. The other reason is the zero cost of replicating software. Software is almost trivially easy to mass produce and distribute. More traditional (and tangible) products are less so. Setting up a “factory” to produce software (or software based services, for that matter) is significantly easier than setting up an iron and steel plant.

At its core, the IT industry to me seems a lot about processes and management — how do you manage your human resources, your computing resources; what processes do you put in place; what does your quality assurance look like. And this is primarily because software and software creation have been largely commoditized. Writing beautiful code is an art; writing code that works and doesn’t fail is not. Traditional manufacturing on the other hand is far from being commoditized — knowledge transfer is hard and expensive, the process of discover and reverse-engineering is also substantially harder.

Since I’m not an economist, my understanding of these issues is not very deep and perhaps superficial. I’m trying to dig up some more data on whats going in the major sectors in the Indian economy to gain better understanding of the big picture. I was glad to see that Wikipedia’s entry on [[wp>Indian_economy|Indian economy]] was quite detailed and seems like a good starting point. Nonetheless, I think its safe to say that India can and should not bank on IT to pull it through the next few decades.


  1. Aaron Farr

    “Now there’s nothing wrong with IT, but just that in the long run India can’t bank on it.”

    Neither can any of the other economies you mention bank on any of their sectors either. I think you over-estimate the value of physical capital and under-estimate the value of human capital. Manufacturing has largely left the US, for example. Banking on all those factories and heavy infrastructure did nothing to keep those industries there when cheaper, more efficient, sources arrived. India needs to consider what it’s competitive advantage is and right now that’s people. In particular, educated people. Don’t discount that.

  2. diwaker

    @aaron: I’m not doubting the potential of human capital. But even utilizing that human capital to its fullest needs the appropriate infrastructure. Right now India has a competitive advantage simply because we crank out a whole lot of people who are conversant in english. But thats not a big edge. The education infrastructure is creaking under pressure and the quality of education in a lot of places is just terrible.

    The BPO industry took off in India because a lot of young, english speaking “junta” was easy to find. But this kind of human capital does not add much value.

  3. Diganta sarkar

    One thing I did not understand – “One of the key differences is that knowledge transfer in software is very easy.” !!!! I thought in manufacturing, the KT is much easier, as you don’t need to understand the logic and you’ve to ‘learn’ from experience how it works. Let me remind you that Chinese are not still inventing, neither Indians, inventing can be a difficult task in both fields. Overall I agree to your point that India should ramp up its’ infractructure and manufacturing bases, at least in coastal areas.

  4. diwaker

    @diganta: I think we don’t agree on the meaning of knowledge transfer. All I meant was that re-creating a product given just the functional specification is much easier in the software industry than in other industries. You don’t agree with that? I mean, even looking at a complex software like Matlab, I’m confident that given enough time, I can code it up. I’m not sure my friends in mechanical engineering feel the same about the latest offerings from BMW or whatever (ok maybe not the best example, but you get the picture).

    I agree, invention is probably equally hard in all industries. My primary concern is manufacturing, and building core skills and infrastucture in traditional industries.

  5. Diganta

    Well, for reverse engineering is more difficult in BMW case. But given the spec, it’ll be easier to reproduce BMW one, rather than Matlab.

  6. Jaya

    Well… The emphasis on physical infrastructure is fine. And of course it would be good to push on manufacturing. There is only a class of population which are employable in industries like Software of BPOs. But physical infrastructure is not equivalent to manufacturing!! Infrastructure is very important for Service/Knowledge based industries too.

Leave a Reply