One of the factors ascribed to the rise of the Bharatiya Janata Party (BJP) under Narendra Modi ahead of the 2014 general elections was the party’s appeal to youth. With about 40% of registered voters in the 2014 elections being under the age of 35, and the BJP relying heavily on data and social media messaging, analysts concluded that the party’s popularity among younger voters was crucial in its victory in those elections.
And there is data to support this hypothesis.
One of the benefits of opinion and exit polls is that they make it possible to analyse voting patterns using data that is not available in official elections. Most opinion and exit polls collect demographic data on the respondent, such as age, religion, caste, income, household size, etc., and this is an invaluable resource for political parties, analysts and anyone else whose livelihood might depend on the outcome of elections.
In most cases, pollsters don’t make this data public, possibly because there is greater value to be gained by selling it to interested parties. When such data is occasionally released, however, it can provide significant insight. And it is one such release that supports our hypothesis —CVoter’s exit poll following the recent Karnataka assembly election.
In a series of tweets, CVoter founder Yashwant Deshmukh showed the voting pattern by different demographic cuts, which made it clear that the BJP has higher appeal among the younger voters, with the Congress’s vote share going up as the voters got older. (See Chart 1)
So, why does the BJP have a demographic worry? Because India’s population is not getting younger.
Using data from the 2011 census (the most recent such data available) and actuarial tables (from the Insurance Regulatory and Development Authority of India (Irdai), we can project the expected population of different age groups in 2019. Comparing it to the population distribution in the 2014 elections (using similar data), it appears that India is getting marginally older.
Based on census data, the median age of the voter in the 2014 general elections was 37.2, and this goes up to 38.0 in the 2019 general elections (we assume that every voter who was aged 15 and above in the 2011 census was eligible to vote in 2014. And that every voter aged 10 and above in 2011 will vote next year, subject to survival).
In other words, the population of first-time voters in 2019 is not sufficient to lower the average age given that everyone is now 5 years older (one caveat, though, is that we use actuarial data to measure deaths, and since these are based on people who are insured, there might be a slight bias).
And while the BJP still seems to be doing well in terms of attracting the first-time voter (looking at the Karnataka data above), the question of whether the party will continue to attract these voters as they get older remains. While we don’t have longitudinal data (on whether people actually switch away from the BJP as they get older—though some polling agencies might have some of this data), the fact that the average voter is not getting younger can be a source of worry for the BJP.
However, not all states are becoming older, or by an equal extent. Assuming zero migration across states, the median voter in 2019 in Uttar Pradesh, India’s largest state, will actually be younger than the median voter in the state in 2014. The median age of the Bihar voter in 2019 is only marginally higher than the median age in the state in 2014. They are both states that the BJP did rather well in 2014. On the other hand, the median voter in Kerala, which is India’s oldest state and where the BJP’s presence is negligible, has got older by over a year. (See Chart 2)
What must also worry the BJP further is that India’s religious composition is changing, and in a way that might affect the party adversely in the long run. While the change in overall share of different religions between the 2014 and 2019 elections is negligible (proportion of Hindus comes down by 10 basis points from 80.13% to 80.03%, while that of Muslims goes up by 15 basis points from 13.76% to 13.91%), census 2011 data shows that nearly one-sixth of the first-time voters in 2019 (people in the 10-14 age group in 2011) will be people classified as “Muslim” by the census (See Chart 3)
The BJP definitely has something to worry about!