Though Twitter had around 2.5 to 3 crore users in India between 2017 and 2020, the period of my study, the real-time data it provided was far more useful than the data from the larger user bases of the other platforms. Twitter has become a source of breaking news and is perhaps the best indicator of the global pulse at any time. The privacy restrictions of platforms like Facebook make it almost impossible for the objective study of data on what is happening around the world at any given point of time. But Twitter provides highly relevant and region-focused trends on a real-time basis. This explains why many recent social movements such as #MeToo and #BlackLivesMatter became global trends on Twitter rather than any other social media platform. News and information travel the fastest on Twitter partly because the platform doesn’t force non-users to sign up to see its content. The majority of users keep their tweets public as well, which means they can be read by anyone, whether the reader has a Twitter account or not.

A study on political outreach during the 2019 Lok Sabha elections analysed the use of language, highlighting how Twitter extended its function well into being a means to mediate conversations between political actors and the mainstream media. The study found that politicians from Hindi-speaking states were much less likely to use English on Twitter compared to those in the rest of India. Government handles, meanwhile, are most likely to tweet in English compared to political parties, showing that the formal language of business continues to be English.

Another study showed that the platform had evolved from being the medium of choice for a select few in 2014 to becoming a channel accessible to the electorate outside the social media elite. Some earlier studies, such as Seizing the Moment: The Presidential Campaigns’ Use of Twitter During the 2012 Electoral Cycle had pointed out that targeted campaign messages on Twitter have the potential to create spillover effects in other media. This allowed me to take my study further and establish that there is increased Twitter engagement during the election season and that the tweets of politicians and political parties appear to have transmedia effects or implications. So I decided to concentrate my study on the effect of Twitter engagement on vote share.

The Twitter effect has been measured in many dimensions, but never in terms of actual vote share. The biggest electoral planks that affect vote share are jobs, development, corruption and farmers. Given the size of India’s population, jobs and development are obviously the top priority for most Indians, while corruption is an issue that can be used by opposition parties to cash in on the anti-incumbency factor. Farmers and agriculture are the drivers of India’s rural economy and most states have a sizeable population employed in agriculture-related industries.

While planning my study, I focused on three questions:

  1. Does higher engagement on Twitter result in positive election outcomes for politicians and political parties? If so, does engagement vary for each political party in different states, or do they get similar levels of engagement in all the states under study?

  2. Do politicians drive the agenda with their projection of issues? If so, what is the response they get from end users? Do all politicians talk about the same issues or is there a subtle difference in the kind of issues a certain political party or politician chooses to focus on?

  3. What is the impact made by the Twitter engagement level of political parties and their political leaders on the final vote share secured by political parties during elections?

After bringing myself up to speed with background research into these questions, I came up with three hypotheses. First, that the Twitter engagement of political parties and their political leaders would be different in all the states I would study. Second, that the level of engagement on Twitter on the issues of jobs, farmers, corruption and development would not be uniform. And finally, that the Twitter engagement of political parties and their political leaders is correlated to the vote share secured by the political parties in the elections

I knew I would face some challenges in my research. For one thing, while Twitter provides a huge amount of data, not all that data is significantly researchable. This is because there are a significant number of messages without any socio-political value, such as condolence messages, messages regarding social functions, updates on official visits by politicians, etc. And since tweets on political issues, particularly in the election season, have a transmedia effect, the tweets don’t remain limited to Twitter but influence the electorate via news channels, radio and print, bringing more people into the equation than just Twitter users.

For example, when the AAP convenor and chief ministerial candidate Arvind Kejriwal raised the issue of giving Delhi full statehood via a tweet, it made headlines on TV, radio and print, and reached more people than just Kejriwal’s followers on Twitter. Later, the tweets of both Kejriwal and the BJP’s chief ministerial candidate for Delhi, Kiran Bedi, made headlines in traditional media just minutes before counting began for the election.

In another case, during the heated election campaign for the 2017 Gujarat assembly elections, the inaccuracies in a graphic tweeted by Congress (INC) leader Rahul Gandhi about the rise in the prices of essential commodities flashed all over traditional media. The way the BJP attacked Rahul for putting out erroneous figures was also the subject of traditional media headlines and the way Rahul retaliated by saying that, unlike Prime Minister Modi, he is human and can make mistakes spread via traditional media as well.

In yet another case, this time in 2018, when the Karnataka assembly elections resulted in a hung assembly and Rahul Gandhi attacked the BJP on Twitter after the governor invited the BJP to form the government, traditional media jumped into the fray with headlines, editorials, talk shows and debates. In December 2019, when the BJP swept the bye-polls in Karnataka, winning 12 out of 15 seats and acquiring a simple majority in the process, Narendra Modi took to Twitter to thank the people of the state for their support. This too made national headlines on all television and print media platforms.

While the transmedia effect of certain tweets would make my study a little more difficult because it would require the study of the effect of tweets on other forms of media (television, radio and print) and how they were used by different media organisations, I found it useful as well, particularly for identifying the issues on which tweets were made more frequently. The prominent issues, I learned, about were jobs, development, corruption and farmers, which allowed me to identify workable research questions from those tweets.

My experience in journalism and, later, psephology helped me identify the Twitter handles that would be useful for my research, including those of political parties and those of the leading politicians of the states I would study. Then, having made my list of important handles, I tracked and analysed their tweets and retweets. I also reviewed all the print articles that used the tweets of these parties and leaders to assess the transmedia effect of the content. I had other sources of information too, including formal and informal interactions with journalists, politicians and people active on social media platforms.

Excerpted with permission from The Online Effect: Decoding X to Predict Election Outcomes, Sanjeev Singh, Bloomsbury India.