Indian vs. American programmers: Whose code is best?

If there were ever an IT Olympics, in which software developers could compete to solve programming problems, the event would likely take place on Gild, an online career development community where IT professionals from around the world do just that.

On Gild, which launched September 2010, software developers participate in programming competitions, earn certifications and look for jobs. Through the competitions and certification exams, programmers can assess their math, logic, communication and software development skills, and compare their capabilities to other programmers across the globe. Nearly 500,000 developers have taken more than 1 million assessments, according to Gild.

CEO Sheeroy Desai says Gild's coding competitions and assessments take the form of a problem statement, followed by a snippet of code. Gild members then have to state whether there's a problem with the code (and what it is), if there's a better solution to the problem (and what it is), or the outcome of the code.

Based on the results of Gild's various assessments, the site recently compared Indian programmers' math, logic, software development and communication skills with those of American programmers. (Gild defines Indian programmers as those living in India and American programmers as those residing in the United States. The career development site asks users to identify where they're from when they register.)

According to Gild's data, Indian programmers appear to be better at math and logic than American programmers. The Indian developers who participated in math and logic assessments outscored their American counterparts by 11%.

But Americans lead at software development. They slightly outperformed Indian developers on mainstream programming languages, such as C, Java and SQL, where they scored eight percent higher on C and nine percent higher on Java and SQL.

Gild's data shows that American software developers are particularly good at Web programming. When tested on PHP and HTML, American programmers' scores were respectively 53 and 27% higher than their Indian counterparts.

Not surprisingly, American developers also have a better grasp of the English language, as noted in their scores on tests of their English communication skills, which were 33% higher than Indian developers.

Desai, who was born in Pakistan, says Gild undertook the comparison of Indian and American programmers for two reasons: One, because the company had enough statistically valid data to compare the two groups, and two, because Gild wanted to see how Indians' and American's technical skills matched up.

"There are lots of perceptions about how good different programmers in different parts of the world are, but there is no hard data," says Desai.

The CEO believes Gild's findings corroborate conventional wisdom regarding the strengths of Indian programmers and American programmers.

When asked if he was concerned about the way the study pits American software developers against Indian software developers, Desai says fanning the flames of competition and animosity was never Gild's intention.

"This is not a way to justify or not justify offshoring," he says. "Our motivation was to collect a remarkable amount of data that we don't think anyone else has and see what we find. This [data] by no means says that any individual programmer in the U.S. is better than any individual programmer in India, and vice versa. These are trends. If I'm a CIO looking to hire programmers in India who are going to be working on mainstream programming languages, I'll find good people there. If I'm looking for Web programmers, I'm better off looking for those people in the U.S."

Meridith Levinson covers Careers, Project Management and Outsourcing for CIO.com. Follow Meridith on Twitter @meridith. Follow everything from CIO.com on Twitter @CIOonline and on Facebook. Email Meridith at mlevinson@cio.com.

Read more about careers in CIO's Careers Drilldown.

This story, "Indian vs. American programmers: Whose code is best?" was originally published by CIO.

What’s wrong? The new clean desk test
Join the discussion
Be the first to comment on this article. Our Commenting Policies