Psephologists failed to predict the massive win of BJP in the 2017 Assembly elections in Uttar Pradesh, even as they did forecast an edge for the saffron party. Times Now VMR and News 24 Chanakya had predicted a clean sweep for BJP and gave 200 and 285 seats respectively and 120 and 88 seats to SP-Congress alliance respectively. The other exit polls showed BJP leading but falling short of the majority.
After the elections were conducted successfully in all 403 constituencies in seven phases, political pundits started analysing and speculating the results, with many of them predicting a hung assembly. Manisha Priyam, a political analyst had predicted that BJP would be getting around 170-180 seats, SP-Congress settling at 100-110, and BSP securing only 90-100 seats. Ajit Jha, editor of India Today also predicted a hung assembly with BJP getting around 170 seats, SP-Congress at 130, and BSP at 90. Similarly, Sunita Aaron and Sharat Pradhan, both senior journalists had predicted more or less similar results. But to their dismay, BJP clean swept the election with securing a massive 312 out of 403 seats.
One of the major challenges before the analysts was to measure the impact of the demonetization drive on the mindset of people at the grassroots level. The elitist section was busy finding loopholes in the drive with almost forgetting to calculate the impression it had left on the common masses on the ground.
While we may criticise the psephologists for being unable to predict the results accurately, there are certain limitations which one comes across during the process of speculating the results. UP’s demography is made up of diverse overlapping identities of voters based on caste, language and religion which makes it difficult to grasp their voting patterns. It becomes a challenge to predict the impact of party hoppers such as Swami Prasad Maurya who left BSP to join BJP and Rita Bahuguna who joined BJP after leaving the Congress. Moreover, the formation of political alliances before the elections further dilutes the accuracy.