The samples in Brazil, Chile, China, Colombia, India, Indonesia, Malaysia, Mexico, Peru, Saudi Arabia, Singapore, South Africa, Thailand, Turkey, and the UAE are more urban, more educated, and/or more affluent than the general population. The survey results for these markets should be viewed as reflecting the views of the more “connected” segment of their population.
It's usually an annotation because Internet/phone penetration among the rural, uneducated, and poor in those countries isn't great. They don't have means to survey these people. Surveying the people who do have access to Internet is representative of what "normal people" feel.
The US has ~91% Internet penetration, while China only has 73% and India only 43%.
That's a factor sure, but someone living in rural country and urban city will have different happiness index based on their living conditions and satisfaction drivers. So comparing X with Y = flawed results.
Yes, I check the data makeup of every survey I view. The sample data that is not using the same criteria for every participant skews the results. If I ask 10 politicians if they think they are doing a good job I get one set of results, but if I ask 10 McDonald employees if their elected politicians are doing great job, results will be different. If you can see the difference there, then the problem is you, not me.
The real question is why western countries consistently score so low on these reports. If western countries are so great and free, then you'd expect people to be reporting high level of happiness across the board.