I want to take a slightly different topic for this week's post to set up a discussion on a topic a bit closer to home (software engineering) for next week's post. Please bear with me.
The Schumpeter column of the April 23rd 2011 edition of The Economist, "The case against globaloney," gave me a different perspective on globalization and got me thinking about the impact of globalization on the software industry so far and where we are going next. To date, my philosophy has been that software as a commodity lends itself to instant, almost costless transportation around the globe both as part of the manufacturing process and for delivery to end users. Taken together with the uniform distribution of intelligence around the world, it is not surprising that those of us working in the software industry find ourselves at the leading (bleeding) edge of globalization.
Indeed, if we can assume that other great intangible commodity, money (which is really just software after all), is available anywhere there is more money to be made then the future path of software industry globalization will follow the recent past. That is, software development will be done wherever there is intelligent life on the planet. Software development density will be proportional to the available access to the education, training, digital infrastructure and tools necessary to support cost-effective participants.
However, the Schumpeter article reviews a new book by Pankaj Ghemawat of IESE Business School in Spain called, "World 3.0." Mr Ghemawat has assembled some impressive data points to debunk some of my assumptions about the forces of globalization:
- Only 2% of students are at universities outside their home countries (hard to believe based on my visits to US Computer Science and engineering facilities)
- Only 3% of people live outside their country of birth (again, this runs contrary to my personal experience but is perhaps easier to believe)
- From a few years ago, <1% of US companies have foreign operations (we do)
- Only 7% of directors of S&P 500 companies are foreigners ( I wonder what % of customers are?)
- <20% of Internet traffic crosses national borders
Perhaps most telling of all for the future of software development globalization, nearly a quarter of North American and European companies shortened their supply chains in 2008! Are Mr Ghemawat's statistics an indication that globalization is proceeding slower than we thought? An indication that a stable equilibrium has been reached? Or an indication that globalization never really existed in the first place outside of a few specialized examples such as those noted in Friedman's "The World is Flat?"
I have to say that as a software engineer, "The World is Flat" resonated much more with my life experience than "World 3.0." However, also as a software engineer, data is data and I must consider the implications for our industry of the Schumpeter article and another article I read in the Philadelphia Inquirer over the weekend by Donald L Bartlett and James B Steel, "Offshoring stole many U.S. Programming Jobs." Hopefully, I'll have some more thoughts and conclusions on this topic in next week's post.