Programming and Corruption: an Anomaly
As someone who has always been interested in politics and computer science, I never really put much thought into putting them both together – until a dark and rainy night when I decided to put Google Trends to the test in its ability to find interesting data correlations.
The issue is that some stereotypes are true to a degree (yes, we all have guns). I decided the only way to prove or disprove the many predefined stereotypes of mine would be to test with some data. Lo and behold, my question: is there any correlation (causation?) between programming language popularity and government corruption?
Yes, I am aware that computer language choice and government corruption are two very different subjects, and don’t necessarily say anything about the people that use them. It could all be a coincidence… or not.
Three notecards, 45-50 unique countries and a flashback to 5th grade math later, a correlation appeared. An anomaly, right in the numbers themselves. I couldn’t believe it. Some languages (Perl and Ruby) topped the leaderboards with primarily English speaking, first-world countries, while other languages (PHP and Java) tailed with a hodgepodge of developing countries, out of primarily Africa and Asia.
When I took those same countries for each respective language and calculated the average corruption ranking (from the Corruption Perception Index), the data was definitive and clear. There was indeed correlation between specific programming languages and the corruption level of the countries they were most popular in!
Per capita , Java and PHP are substantially more popular in corrupt countries, when compared to alternatives like Python and Perl.
Maybe a graph will help you visualize the percepted corruption for each language. The highest score achieved in 2013 for the corruption perception index was 91, for New Zealand and Denmark – and over two thirds of the countries surveyed scored below a 50 in the ranking.
Yes, correlation does not imply causation – however, the anomaly here clearly lies in the numbers (and a nice graph). Programming language popularity per capita does indeed have unexpected associations with the corruption ranking for each respective country. Where do you stand on the scale?