Since The Economist is publishing a special report on technology in India and China this week, I thought it was about time I wrote something on the subject. That way when I read the special report I can either congratulate myself on my deep insight or slag off the hacks for getting it so obviously wrong. Since I spent last year working in Bangalore and the best part of this year in Xi'an and Beijing, I think my credentials are as good as anyone's. There's nothing controversial here for people who are familiar with the Indian and Chinese markets, but anybody whose only source of news is the Northern press might find it interesting.
Outsourcing is by far the most common context in which the Indian and Chinese IT sectors appear in the press. This is only to be expected since outsourcing is perceived as a threat to the interests of the middle classes in Northern societies. In fact this is to some extent a myth. Clearly outsourcing has had some kind of moderating influence on the salaries of IT workers in the USA and Europe. However it's still easy to get a visa to come to Northern countries if you have IT experience, and you can still earn good money as an IT professional.
By and large, outsourcing has not put really talented people permanently out of work, although it has certainly got rid of some of the chaff. Why is this? There are two reasons. The first is that outsourcing is really much harder than people might think. It is always less effective to distribute IT work than to do it on-site due to the inefficiencies introduced through not being able to communicate face-to-face. Secondly, IT professionals are not fungible, and it is as hard to find good IT workers in places like India and China as it is to find them anywhere else. Although the populations of India and China are huge, the number of people in those countries with a good enough education and sufficiently fluent English to work effectively in outsourcing isn't actually that great.
The fact that demand is exceeding supply is reflected in the fact that salaries for IT professionals in Bangalore are rising at between 10% and 20% per year. The market is tight as a drum. What you're seeing is the effect of arbitrage -- eventually IT salaries in India will peak at a level where the difference in the salary of an Indian IT worker and that of a Northern IT worker will represent precisely the cost of the inefficiency introduced through effects such as the communication overhead. Indeed this arbitrage has already had the interesting effect of creating an outsourcing market within the US, with companies creating divisions in places like Montana and Virginia. The amount of IT work to be done in Northern countries certainly isn't decreasing, and there isn't a proportionate increase in the availability of skilled IT professionals in India. The Indian Institutes of Technology (IITs), India's premier engineering universities, admit only 4,000 people per year.
Long term, unless there is a significant change in the Indian educational system that creates a step change in the number of available skilled graduates, we won't see outsourcing increasing significantly. Indeed if demand for IT work continues to increase in Northern countries, we'll see even more demand for IT professionals. Things aren't really any different in China since although the education system is superior it is relatively hard to find people with sufficiently good fluency in European languages.
Nicholas Negroponte's One Laptop Per Child initiative is one way in which such a step change could happen. But society itself needs to change in order for projects like this to succeed -- in India, in particular, discrimination against the lower castes is still so firmly entrenched that they are simply not able to participate in India's economic growth by and large. None of the simple fixes have worked, and there is no agreement on what to do next, let alone the political will to implement a complex solution.
The domestic market
The above analysis entirely ignores the effect of the domestic market for IT services in India and China. In fact, both of these markets are relatively immature. It is hard to get local work in China, which many amateur Sinologists put down to the need to establish guanxi -- the network of contacts through which work and favours are procured. However the root cause of the difficulty in finding work is simply that the demand that exists is either for huge projects or for start-ups. Although the government is forging ahead with IT projects, private enterprise in China is only just beginning to become computerised. As the IT revolution begins to work its way through the Chinese economy, demand will increase exponentially, and guanxi will simply not be sufficient as a means of procuring IT services. It is already the case when negotiating with large companies that they typically want to know how many thousands of developers we can supply. This in itself is a sign of the immaturity of the market, but it demonstrates its scale.
India arguably has a lead on China in terms of penetration of IT in business. If it indeed it has, given the blistering pace of change in China, it won't last. We observed one interesting difference between the two countries in the way our offices were constructed. The Indian construction company had twice the number of workers as the Chinese one -- but the Chinese team were equipped with modern power tools which made them far more efficient. Whatever the reasons behind this, it shows how determined China is to be at the forefront of technology adoption, and in general Chinese infrastructure is years ahead of India's. As Indian and Chinese businesses get computerised in order to gain competitive advantage, the demand for skilled workers will also increase -- and the only supply will be people without the necessary language skills to participate in the outsourcing sector. The number of such people will be large, but still a long way from being able to satisfy demand.
Writing software is still far from being an engineering discipline. Although some common pieces of software are commoditised (operating systems, web servers, content management systems), every business will require some level of customisation in order to cater to its peculiar needs. The more revolutionary the business model, the more customisation. Custom application development is labour intensive, risky, and expensive. The company I work for, ThoughtWorks, makes money from reducing the risks involved with application development through applying lean methodologies, aggressively adopting newly maturing technologies such as Ruby on Rails and Erlang, and having people who are smart enough to know the right technology and approach to use in a given situation. We're also working on products to enable teams to deliver better software with less risk.
This isn't enough though. Reaping productivity benefits through the techniques described above requires a high level of skill and understanding -- and there just isn't enough to go round. Servicing the huge domestic markets of China and India as they mature is going to require something really extraordinary. It is certain that companies in these countries will have business models that differ disruptively from established Western ones. This is already clear both from the unprecedented trade surplus and from the large number of companies that have seen their China strategies fail. The innovation in IT required to power these businesses is just beginning.
The market for IT services is bright, and its future is in India and China. As domestic markets mature in these countries, they will need a huge injection of IT expertise. This will enable them to beat the rest of the world in everything from financial services to manufacturing. Northern businesses will be caught short, and will need to innovate furiously in order to remain competitive. At the same time, IT's wider adoption is changing the way societies function, and we can only expect these changes to accelerate. For many of us, the trick will be to find a way to use those changes to deliver social justice. I am looking forward to being a part of this paradigm shift along with my colleagues in our new Beijing office.
Some of these ideas come from discussions with Michael Robinson and Roy Singham (ThoughtWorks' fearless leader). However they shouldn't be held to blame for errors, omissions or stupidity.