How analytics make a difference in elections
Businesses have been successfully using Big Data and Analytics to understand consumer behaviour and increase their own competitiveness. While data has revolutionised Business Analytics, it is being explored in politics too.
The use of data and analytics in politics is not new. In 2012, Obama had made good use of analytics for election campaigns and raised significant public awareness. Today, leading parties in India have begun to utilise the power of electoral data, and analytics towards micro-targeting.
The upcoming Karnataka elections are going to be no different. Data analytics teams have been set up by parties to help them strengthen their campaigns, craft strategies to promote voter engagement, increase memberships, raise donations, and nudge the voter behaviour in their favour with accurate messaging, on a near real-time basis.
Today, the electoral analytics is a reality because of all the right ingredients that are readily available - abundant electoral and election data, superior computing power, and advances in scientific modeling techniques as well as the talent to put it into action.
All these have a significant role in driving the outcome predictions of the elections. IndiaVotes, one of the largest public elections resource of the country, offers data analysis of more than 200 past national and state elections in the form of dashboards, and visually stimulating data models, which are accessible publicly. While this generates aggregated insights, several experiments are being further conducted to get the foresight right with data and analytics.
Let us understand how data and analytics are being leveraged in the 2018 Karnataka Elections. First, the fundamental question is to mine the percentage of voters and non-voters. This year, Karnataka has about 4.9 crore electors in 224 assembly constituencies. The next question is how to increase the probability of those who vote. If they vote at all, which party are they going to vote for? Next, analytics plays a critical role in analysing specific segments to influence the voting preferences of the undecided voters to swing favourably in predicting the seats.
Moving ahead, opinion polls have played a critical role in understanding which party is leading and which is not. This is where predictive analytics comes in handy. And, over time this becomes even more interesting because the time factor can swing the fortunes completely. For example, in the 2016 US Elections, while the odds were in favour of Hillary Clinton, a couple of months later the time factor turned the table for Donald Trump.
Decoding the hidden data
In the recent times, in addition to the structured data, the unstructured data too has played the role of a catalyst in enhancing the predictability of the seats. This is where text analytics comes into the picture. Text Analytics involves an analysis of the conversations, emails, print and online text, messages, and social media coverage about the party candidates, and about the issues faced by the electors in the region.
The analysis is contextual because the issues are confined to a given region and for a given time during the elections. For example, the issues could range from unemployment, safety, corruption, economic development or law and order. Here, voter sentiment analysis provides a reality check to the parties, and an advanced topic-modeling analytics highlights the core issues that really matter to the people. The reality check and key topics help the parties to stay focused on connecting with the electorate, and to understand how the voters perceive the candidate and the party’s standpoint to the key issues.
In the coming future, technological advancements coupled with analytics will emerge as an indispensable part of the elections. A new approach using Blockchain as an enabling technology for elections is emerging, which can streamline many aspects of elections.
One of the blockchain implementations is followmyvote.com. It is online and transparent, but doesn’t compromise on privacy and ensures that votes that are being cast by individuals reach the ballot box in a secure way. Blockchain adds accountability, and most importantly, it prevents frauds. It makes every vote immutable, unlike the penetrable electronic voting system. But, until a blockchain technology enabled election system is in place, analytics will continue to pave the way and bridge the gap between electors and parties.