Prime Minister Narendra Modi conducted a roadshow while inaugurating an Expressway and made an indirect campaign speech at a rally in a neighbouring district just a day before the Lok Sabha by-election in Kairana, Uttar Pradesh, on May 28. The Bharatiya Janata Party lost Kairana.
In the Modi era, a BJP victory in any Assembly election is attributed to the “Modi magic”. This is due to the unprecedented time Mr. Modi devotes to campaigning in State elections. For example, Mr. Modi participated in 34 rallies in Gujarat, 31 in Bihar, 24 in U.P., and 21 in the Karnataka Assembly elections. After the BJP secured 104 seats in Karnataka, commentators again lauded the “Modi magic” and “wave” for taking the BJP past 100 seats.
Problematically, these are anecdotal conjectures, unfounded on systematic post-poll surveys probing the effect of a Modi rally on last-minute voter decisions. In Gujarat, for example, the Lokniti-Centre for the Study of Developing Societies (CSDS) conducted a post-poll survey which found that 43% of respondents decided their candidate in the last two weeks, and more of them voted for the BJP than the Congress. This tilted an evenly balanced election (as seen in the last CSDS prepoll survey) towards the BJP. The CSDS researchers believe that it is “quite obviously” Mr. Modi’s campaigning which saved the BJP from defeat.
In Karnataka, most opinion polls conducted before Mr. Modi began his rallies got the vote share correct, and the seats wrong. Siddaramaiah was the most popular Chief Minister choice, yet the Congress lost its majority in the Assembly. The exit polls had given contradictory results. Given the unreliability of opinion/exit polls, how do we treat them as conclusive evidence? This is the problem of attributing BJP’s performance, especially in Karnataka, to the Modi magic before cogent studies.
Complex elections in a complex and diverse country like India are reduced here to a national charismatic leadership, at the expense of local, caste, class/economic, factional factors (surveys find development and a candidate’s work as the most important voting issues, not leadership). The fact that the Congress increased its vote share to 38% (almost 2% over the BJP, and translating into nearly 7 lakh votes) during an ostensible Modi wave is discounted here. Mr. Modi himself claimed that the performance in Karnataka was “unprecedented and unparalleled.” But the BJP had won more seats in 2008 (110) before the Modi magic.
This is not to argue that Mr. Modi's popularity does not matter; after all, he is the most popular national leader at present, and by a huge distance. But the crucial question is, how does this influence voter decisions? The arguments of Modi magic after each election are also cyclical because they begin with what they are supposed to explain. By this logic, is the Kairana loss due to the failure of the Modi magic?
Moreover, they ignore the material basis: immense resources spent in creating the “magic”: full-page advertisements in newspapers, the amount spent on electioneering (supposedly the highest in Karnataka), etc. This itself is made possible by 90% of corporate donations going to the BJP recently. If the Modi magic is so overwhelming, why is an unprecedented ₹4,343 crore budget needed for publicity by the Modi government? And why does it need “the divisive innuendoes, falsehoods and fear mongering” (according to CSDS researchers) of the Modi campaign in Gujarat which supposedly swung the election?
Instead of anecdotal evidence, here are the results of the constituencies (including cities which might have two-three constituencies but excluding the metro city of Bengaluru which has many constituencies) that Mr. Modi conducted rallies in. We can safely surmise that BJP wins with good/big majorities would have been victories even without the effect of a Modi rally. Contrastingly, close wins (even if we keep a higher margin of around 15,000 votes) might have been made possible by a Modi rally.
Of the 35 such constituencies (including in Congress/Janata Dal (S) areas of influence), the BJP got 16: it won 11 from other parties/Independents (Belgaum South, Belgaum North, Bellary City, Gulbarga Rural, Mangalore City South, Mangalore City North, Raichur, Shimoga, Shimoga Rural, Tumkur City and Udupi) and retained five (Chikmagalur, Chitradurga, Gulbarga South, Hubli-Dharwad Central and Hubli-Dharwad West). Only six constituencies (Gulbarga South, Gulbarga Rural, Raichur, Shimoga Rural, Tumkur City and Udupi) were with a majority of approximately 15,000 or less showing the possible influence of a Modi rally in the win.
The Congress won 16 (Belgaum Rural, Bellary, Chamarajanagar, Gulbarga North, Hubli-Dharwad East, Kolar Gold Field, Bidar, Mangalore, Raichur Rural, Yeshwanthpur, Bangarapet, Koppal, Gadag, Jamakhandi, Babaleshwar and Chikkodi), and the JD(S) three (Bidar South, Kolar and Tumkur Rural). Of these, eight (Belgaum Rural, Bellary, Bidar, Bidar South, Kolar, Kolar Gold Field, Raichur Rural, Tumkur Rural) were captured from the BJP/splinter-of-BJP/Independents despite the Modi rally.
These figures suggest the gains/losses for the BJP after the Modi rally were roughly evenly matched. And that we need in-depth studies about voter decisions conducted before and after elections to conclusively establish the effect of a Modi rally. Otherwise, the breathless recital of the mantra of a Modi wave might turn out to be a vacuous exercise after every election.
Nissim Mannathukkaren is Chair, International Development Studies, Dalhousie University, Canada. He tweets @nmannathukkaren