The Exit Polls for the 2019 Lok Sabha Elections were released by various agencies after voting in the final phase was completed. While a majority of the pollsters predicted the NDA to secure a majority in the Lower House of the Parliament, conspicuously, there were two which predicted the NDA to come up short.
One of them was the ABP-Nielsen Exit Poll that gave the NDA 267 seats, 5 short of a majority. It was later revealed that they possibly manipulated the data to reach their favoured conclusions.
The other poll that predicts the NDA to fall short of 272 is the NewsX Neta Exit Poll. NewsX Neta predicts the NDA to secure 242 seats, 30 short of the 272 required for a majority. The UPA is predicted to win 164 while Others are expected to win 136.
“The BJP, which had won 71 of the 80 seats in Uttar Pradesh in 2014, is set to suffer huge loss at the hands of the Mahagathbandhan. The saffron party is expected to win only 33 seats this time, a loss of more than 50 per cent over 2014. The Mahagathbandhan is expected to win 41 seats (BSP 22, SP 20 and RLD 1) and the Congress 4 seats,” the exit poll said.
“The Congress is likely to witness a resurgence in all the states where it is fighting the BJP directly, which include Assam, Delhi, Gujarat, Madhya Pradesh, Rajasthan, Karnataka and Maharashtra. However, one interesting thing to note is that states like Madhya Pradesh and Rajasthan, which recently went the Congress way in the Assembly elections, seem to be voting strongly for the BJP in the Lok Sabha elections, highlighting the popularity of Prime Minister Narendra Modi in these regions,” it stated.
The funny bit is the predictions appear to base a lot on their app. Pratham Mittal, Founder, Neta App said, “Over 47 lakh voters participated in the Neta exit poll while over 2.5 crore voters participated in its opinion poll across all the 542 constituencies. These results are a statistical derivation based on the two. To ensure a pertinent sample size, the Neta app has used multiple mediums like the app, IVR calls and SMS to gather data on evolving political inclinations”.
The app currently has only 1 million downloads. And yet, since yesterday, the NDA numbers have fallen by 2 seats already and is now predicted to win 240. More of a problem is the fact that there’s no guarantee that the data received through the app is authentic, it’s not too difficult to imagine how it can be hoodwinked. For instance, a single person can very easily own multiple phones and register the app on their phones which in turn can affect the authenticity of the data.
Also, if the data can change within the space of 24 hours, one ought to question their methodology and the over-dependence on easily manipulated app results. Again, the data itself appears to be tainted with too many falsities and too much is left to the internal biases of the interpreters of it.
For Exit Polls to be accurate, not just the size of the sample, but also the quality of the sample would matter greatly. The sample would need to be a representative sample of all demographics, economic strata, a mix of educated and the uneducated, age variety etc. One would also need to ensure that the sample size has an equal representation of all stratas of the society from all constituencies.
The app has 10 lakh downloads while the Neta Exit Polls claims to have surveyed 47 lakh voters. Thus, more than 20% of the data surveyed could well be inauthentic. If the percentage of inauthentic is even half that or one-fourth of that or even one-tenth, it’s more than enough to completely scar the results.
Thus, the 242 given to the NDA by the NewsX Neta Exit Poll ought not be taken too seriously. Like the ABP Nielsen Exit Poll where the Director admitted to manipulating the data, this exit poll, too, suffers from serious flaws.