It occurred to me that I could do some birthday-related data analytics to estimate PM Najib Razak's popularity since his birthday was a few weeks ago. I didn't remember the date, just saw a few tweets wishing he well (I did have to memorize the birthdays of all PMs when I represented Perlis in Kuiz Perdana, but that was before Najib and Pak Lah). Anyway, I thought I should compare how many birthday wishes Najib received last year versus this year, you know after the 1MDB brouhaha and all. This could be a proxy metric to compare his popularity before and after the 1MDB scandal.
Granted, Twitter users may represent only a slice of Malaysian population, some wishes may or may not be sincere, and like other surveys, the results are subjected to sampling errors. However, this is a controlled, non-random, approach using the same methodology in identifying birthday wishes in English and Malay (original tweets and retweets) to @NajibRazak within the month of July in 2015 versus the same month in 2014. As such the relative percentage difference is more important than the absolute numbers (raw data here).
The verdict is not good. Najib's birthday wishes dropped by a whopping 61% this year as compared to that in 2014.
Although this is a proxy estimate, I would argue that it is much better (indirect, revealed preference) than asking a room full of UMNO Ketua Bahagian and Majlis Tertinggi (direct, stated preference) as people tend to lie when been asked directly in a survey to protect their interests.