Broad brush strokes


I am a huge fan of Troy Hunt I read everything he posts and have gone through a fair few of his plural sight courses. I did however take issue with a recent post that attempted to document his relative experiences while outsourcing work in India, China and the Philippines. The post "Offshoring roulette" appears to be the sum of a series Troy's professional observations as a Software Architect overseeing work conducted in the Asia Pacific.

Before I dive into what I think the problems are from my perspective, I will say that Troy attempts to limit the discussion to his own personal observations, and takes the time to share that he spent several years as a minority, specifically pointing out that three of those were spent in Singapore as teenager.

My problem, however, is that I simply do not think he has provided enough information on his source data to make accurate comparisons between India, China and the Philippines. Some of what Troy discusses appears to be perfectly and logically correct (cost, language barriers, etc.) but I believe he has casually incorporated certain myths into his analytical framework which is dangerous, and if we just accept them, no matter how hard we try, the bias becomes intrinsic and implacable.

The following is a list of reasons why making broad sweeping observations about diverse populations is problematic:

  • Distilling a country into a few paragraphs will always by incomplete and inaccurate.
  • When ever I have see comments of work conducted with bad developers of European descent, for some reason that does not seem to reflect back to their nation of origin. You simply assume you are dealing with a bad developer.
  • Having a deep understanding of cultural and social anthropology is an important prerequisite to comparing and contrasting cultures.
  • Trained anthropologist have been known to struggle with their own bias.
  • Having a proximity to other cultural groups does not make you qualified to speak on their behalf.
  • My experiences as a minority in two countries does not make me qualified to speak on the experiences of other minority groups.
  • Considering the appropriate level of sensitivity surrounding cultural comparisons neglecting to share information on test and control groups is lazy and possibly negligent.

Describing and comparing cultures is difficult but possible, however, if you do not have a background in anthropology and have not accounted for statistical blind spots in your data sources your analysis should be considered flawed on many, many levels. For example, if the groups you have access to are not representative of the whole country it may be more appropriate to compare one corporation to another, or maybe one group of graduates to another, either way we are a really long way off from comparing countries.

Let’s put this another way, could you in a couple of sentences describe the primary attributes of a UK developer? Or a US developer? Do you find them relatively genial? Arrogant? Or how about the differences in west coast vs. east coast developers? Delhi vs. Bangalore?

When you take time to really consider this, it just sounds ridiculous.


Photo from Women of Color in Tech

Comment Section

Comments are closed.