A few weeks back, we all witnessed the drama of the #HungerIndex. Made me wonder if something similar is done in the #HappinessIndex as well? Year on year we are deemed as one of the unhappiest nations in the world.
Here’s what I found:
The index is calculated by asking random nation’s citizens a few questions on the likes of “How would you rate your current life on a scale of 1-10?” Here’s where it gets interesting: in Finland that’s consistently among the happiest nations, a single person represents about 5500 individuals; In India, ONE person represents about 5 Lakh!!!! 1010 were interviewed in Finland against only 3000 in India. This basically means that had they called 25 Bangaloreans at 9:00 a.m. who are probably pissed for being stuck at Silk Board junction, the entire city would be considered unhappy.
Now, I decided to crunch some numbers further (Most of the data is available online). There is something called correlation in the field of Data Science. A correlation number of 0 means there’s no correlation between two datasets. A number near 1 means there’s heavy correlation (Negative means there’s an opposite relation). Example: If column one is KL Rahul’s averages against a bowler and column two is a bowler’s strike rate (representing quality), you’d get a negative number. If you change the second column to a bowler’s overall economy rate, the number would be close to 1, and similarly if you change it to IPL bowlers it’d be close to 0 (since there’s little deviation and he’s consistent there).
I compared the happiness score against Per Capita Income (Adjust for Purchasing Power Parity) and against Per Capita Income (Nominal). The correlation is 0.71 and 0.73 respectively. This basically means the more money you have the happier you are. Why even do they survey then??? This level of correlation is so high, that you can fire the entire team and have a person with basic excel skills publish the happiness index every year.
These numbers are so biased towards developed nations! Although they do adjust for error, it’s stupid #DataScience to extrapolate from such a small number. I don’t even want to get into the , : How do they adjust for language issues and other diversities??? Do they verify the backgrounds told by the interviewees, do they take into account timing as well (I’m assuming a farmer is as happy during harvest season as the Bangalore CS guy near his appraisal).
The ‘Consistent effort to taint India’s image’ remark by our Prime Minister seems to be true!