They began by showing graphs of the exponential rise in software revenues in both China and India since the 1990s, but when broken down into exports and imports, it becomes clear that Indian software is predominently written for exportation while Chinese software is for the domestic market. And although India does not have a similarly developed hardware industry, when that sector is analyzed, Chinese hardware is overwhelmingly exported while what hardware India does make is for domestic consumption.

A number of explanations are typically given for the difference, notably India’s English language proficiency, its higher education system that created a large labor pool of software engineers, and the overbearing regulation that was not extended to Indian software firms. The authors of this book believe that while these are necessary explanations, they are not sufficient. Using a variety of data, including firm-level interviews with 300 Chinese and Indian companies, they think they have flushed out the answer.

Their research suggests that Indian management, not labor, and their pool of larger, better educated professionals were largely responsible. The management can be applauded for seeking quality certifications for Indian software firms and utilizing the diaspora ties. Further, they strategically partnered with far more American software companies than the Chinese did – 60% of surveyed Indian firms had Western partners, compared to only 12% in China. (There was a lot of data thrown into the presentation that focused on the software industry, but I didn’t copy most of it down.) A final reason offered by the authors, more tentatively, was a cultural explanation – Indians tend to be more outspoken and tolerant of ambiguity. Because software creation is a creative enterprise, perhaps they have an inherent comparative advantage.

During the Q&A, Professor Mike Nelson offered some helpful insights from his time with the American IT industry:

• In hardware, you can thrive with 2-3 clients whereas in software, you need many more. Therefore, overcoming the “foreignness” of China is more of a factor than in India where multiple Western clients can be easily courted due to the relative institutional familiarity.
• Timezones shouldn’t be discounted – India is apparently much easier to schedule with than China.
• Given India’s relative governance instability, software (with lower fixed costs) is a more flexible industry – Wipro or Infosys can leave localities more easily than OEMs.

Overall a very interesting talk that adds great data to the debate while debunking commonly held beliefs like the importance of Y2K.

A useful comparison is found in the map below which represents national gross domestic products (though the coloring is the opposite of Firefox’s – red is poor). As you might expect, the poorer countries align with those who downloaded fewer versions of Firefox.