Number theory: How parties used social media in Gujarat elections
Digital campaigns are expected to play an increasingly important role in Indian elections as more Indians start using social media. An analysis of social media use in the recently concluded Gujarat elections, however, shows that all parties do not attach equal importance to this medium of campaign.
Digital campaigns are expected to play an increasingly important role in Indian elections as more Indians start using social media. An analysis of social media use in the recently concluded Gujarat elections, however, shows that all parties do not attach equal importance to this medium of campaign. To be sure, on platforms such as Facebook, social media’s use as a campaign tool may depend on parties’ ability to pay for advertisements. Here are five charts that show this.
How Twitter was used by contesting candidates
Twitter remains a minuscule part of the social media landscape in India. Only 23.6 million Indians use the platform against over 300 million active users of Facebook in January, according to Statista. However, an analysis of the platform’s use shows how parties and candidates use social media in their campaigns. Since Twitter does not allow political advertisements, our analysis focuses on how contesting candidates participate on the platform.
Among different parties, the Bharatiya Janata Party (BJP) candidates were the biggest users of the platform, with 81% of their candidates using it. This number was just 55% for the Congress and 43% for the Aam Aadmi Party (AAP). To be sure, presence on the platform need not translate into its active use. For example, the places where a high proportion of candidates were present on Twitter were not necessarily the places where candidates were most active on the platform during the campaign period.
Another interesting aspect of Twitter use by candidates was the language in which they used the platform. Of the 45,000 tweets made by the contesting candidates, more than 37,000 (80%) were in Gujarati, only about 3,000 in Hindi, and a little over 2,000 in English. Hindi tweets performed the best, receiving an average of 364 likes and 84 retweets, with English tweets coming in second at an average of 218 likes and 47 retweets. One explanation of these statistics is that they don’t necessarily reflect engagement by the people of Gujarat because people can engage with a tweet from any location. A tweet in English or Hindi is expected to have more engagement simply because the pool of people who understand the language is bigger. In case the statistics do reflect the engagement by people in Gujarat – this is also not an entirely unlikely possibility when the tweet is from a candidate in a state election – the strategy of using Twitter in a local language may be misguided.
How parties advertised on Meta (Facebook)
Campaigning on Facebook plays a much more important role since it allows for political advertisements and reaches a much larger population. In these elections, parties spent a total of ₹4.5 crore on 6,713 political ads, which received a total of 600 million impressions. On average, an ad ran for about five days. Similar to the strategy seen on Twitter, more than 85% of these ads were in Gujarati. The BJP spent three quarters of all the money spent on ads on Facebook ( ₹3.35 crore). It means that even the combined expenditure of the Congress ( ₹58.7 lakh) and the AAP ( ₹55.3 lakh) was less than half of the money spent by the BJP.
An important difference between Twitter and Facebook campaigns is that Facebook campaigns are run both by candidates and parties, who tend to centralize spending on political advertisement. Only 5% of AAP’s contesting candidates advertised on Facebook, against 15% of BJP’s candidates and 25% of Congress’ candidates.
Changes in strategy compared to Punjab elections
How did the Gujarat election compare with the previous ones in terms of Facebook advertisements? Analysis of the December elections in Gujarat and the Punjab elections in March shows that overall, parties ran fewer ads in Gujarat, but spent more money on them. The overall expenditure on Meta’s advertisements rose by approximately 32%, and the impressions received nearly doubled.
Some of these trends could just be the result of the importance of the Gujarat elections compared to the Punjab elections for various parties. For example, among the three main parties, the Congress ran the most ads in Punjab, while the BJP ran the most ads in Gujarat. However, the number of ads that the BJP ran in Gujarat was still smaller than the number Congress ran in Punjab.
What’s noteworthy here is how the parties have changed their strategies with respect to Instagram and Facebook. The share of advertisements that ran only on Facebook or Instagram increased, whereas the share of ads that ran on both platforms decreased. This was a result of a shift in the BJP’s strategy, and not the Congress or the AAP. It is possible that the BJP has caught on to the fact that the two platforms are used by different demographic groups, and as a result, may require different advertising strategies.
Despite the significant expenditure on such advertisements, it is also important to note here that such spending remains a fraction of total campaign expenditure. For example, the state units of only the three main parties discussed here spent ₹19 crore on the Punjab elections, according to expenditure data submitted by them to the Election Commission of India (ECI) and analysed by the Association for Democratic Reforms (ADR). In comparison, all parties combined spent just ₹3.4 crore on Facebook during the elections in Punjab. This is not surprising. A decreasing but significant share of Indians does not have access to the internet. It is a corollary of that fact that only a fraction of those who access the internet are likely to use social media. Digital campaigns are, therefore, add-ons rather than substitutes of traditional forms of campaigning.
(Shivam Gangwani is a Junior Research Fellow at Trivedi Centre for Political Data, Ashoka University.)