As Cambridge Analytica Hits Headlines, Here’s How Political Parties Are Using Big Data

Ashwani Singla, founder of India’s first science-based reputation management advisory Astrum, said that thousands of volunteers conduct door-to-door surveys and interact with the voter, which is further broken down into data and analysed.

New Delhi: It was in 2014 that then Gujarat Chief Minister Narendra Modi came to power with the BJP and became the Prime Minister. His thumping victory was a subject of great speculation and assumption all over the country. One striking fact about his campaign and eventual win was his use of big data and analytics to assure him there were no ‘last mile’ problems.

Cut to 2018, data and analytics is still the hot topic of discussion. Over the years since Modi brought in the renewed face of election campaign in India, opposition parties have used data analytics in streamlining their decision-making better and help in respective elections. But how does big data exactly work? And how important is it to politics in India as of today?

Union Minister Ravi Shankar Prasad has accused the Congress party of using the services of Cambridge Analytica(CA), a controversial offshoot of British company SCL Group that deals in data mining and analysis for electoral procedures.

While the ruling party accused the opposition, Cambridge Analytica claimed in its website that it had worked with the BJP too. Its India partner Ovleno Business Intelligence (OBI) stated on its website that it had worked with the JD(U) in 2010 and secured its win. Interestingly, BJP was in alliance with the party back then.

Combine the Past, Present to Figure out the Future
Big data is a serious amount of data that is being processed, churned in a complex environment of various possibilities. Analytics of the same data is the application of it in the elections.

For Ashwani Singla, founder of India’s first science-based reputation management advisory Astrum, the intersection point of big data and the analytics of it is India’s large voter base. “We will be going into 2019 with a voter base of almost 900+ million,” he said.

Shedding some light on how the procedure of big data and analytics works in the Indian political scenario, he said all the data that he has ever used was only that available on the public data, which includes data from the Election Commission, Census and NSSO. It’s an exhaustive data, he admitted, but the data is dated and not so rich considering demographic change (for example, the last updated Census data goes back to 2011).

He has his boots on the ground, thousands of volunteers going from door-to-door to conduct polls and to have conversations with the voter, which is further broken down into data and analysed.

“Every voter is not a homogenous whole. And in India, it is more complex because there are various differences based on language, caste, socio-economic backgrounds. There is, of course, an SC list, an OBC list that’s available in the public domain. But say, if I’m working in a particular district, how many OBC groups do I have there? I have to find out for myself because that data is not in the public domain,” he explained.

Once these door-to-door polls come back to the analytical team, they then figure out which direction the conversation among the voter is going.

One can look at farmers and women as two separate groups, but then an SC farmer might be different from an OBC farmer. “We have to go down to that micro level. Otherwise there’s no point,” Singla added. The data allows him and his team to figure out the trend, starting from a panchayat election to a general election. While he refused to divulge which political party he was associated with, Astrum’s website states that he steered Modi’s 2014 general election campaign.

“Based on voter history, I can figure out what my battlefield looks like. Where do I fight? Why is the voter voting in a certain way? How do I engage and who should I pitch as my candidate? Our polls tell us all that and help the party further in engaging with the voter at a very early stage,” he said.

The layers of data, he added, eventually are only a solid supplement to support booth-level activities. “Political parties help us with their insights, of course, but I am the one who knows how to read the data and to see the science behind it. It definitely helps in electoral strategy, but that’s something where people like me don’t have the decision power. I can only suggest based on my scientific research,” he said.

The Politics of it All
While Singla might be a non-political entity in the big picture of electoral process and data analytics, parties like the BJP and the Congress, too, have their own data analytics team. Earlier this year, Congress President Rahul Gandhi announced a data analytics department led by political economist Praveen Chakravarthy. Speaking to News18, Chakravarthy said there was nothing new in political parties using data.

“Analysis of data to glean insights for decisions is not new. Every sphere of society embraces use of data – business, arts, law etc. So does politics. There is nothing new nor different about using data, be it in politics or business,” he said.

Asserting that big data does not mean private data, Chakravarthy said that the Congress was just using lots of public data, including electoral data, data from party workers and that his team was a group of scholars and experts who “had their own methods of data analysis”.

The ruling Bhartiya Janata Party (BJP) too has its own team that does big data analytics for the party.

Amit Malviya, the national head of information and technology, leads the data team at the BJP. “Numbers tell us what has happened historically. It helps us get an insight to what the numbers are saying in terms of electoral arithmetic,” he said.

Adding that his team tries grill down to the booth-level, Malviya said that elections wasn’t arithmetic but it was all about chemistry. “Unless we look at it dispassionately, we can’t do much with it,” he added.

Once the analysis is in, the party builds strategies around it, which are purely political in nature. “Do the numbers mean that we will lose an election? No. But will it help in winning? Certainly. It is one of the inputs to an electoral strategy but an important one,” he added.

Malviya’s statement could be seen in the light of the recent elections that the BJP lost-Gorakhpur, which is considered to be Uttar Pradesh CM Yogi Adityanath’s bastion. Yogi accepted that overconfidence cost him and the party in the bypolls.

While the BJP lost Gorakhpur, the party did look at numbers in a few other states, including Tripura and where it won. “The state units handle by-polls,” Malviya said. The focus is now on Karnataka.

Considering the level of interest and hard-core campaigns in state election by Congress President Rahul Gandhi and PM Narendra Modi, both Chakravarthy and Malviya were asked whether the leaders look at the data they churn out.

Chakravarthy said that Gandhi was a believer in data for its objectivity. “Our approach to data is much more scientific than scheming,” he added.

Malviya refused to divulge details on the PM’s involvement in looking at data.