BJP gained in irregular colonies, lost SC areas
Arvind Kejriwal and his Aam Aadmi Party (AAP) must be ecstatic. A year ago, in the 2019 national election, the party did not win even a single assembly constituency (AC) segment of the 70 ACs in Delhi, while rival Bharatiya Janata Party (BJP) won 65 AC segments with a 57% vote share. Just a year later, the AAP has won 62 of 70 ACs on a roughly 54% average constituency-wise vote share in the state election, nearly replicating the 67 it won five years ago in the 2015 state election (also on 54% average vote share).
We’ve all seen this before. In 2015, political observers were shocked by the AAP’s landslide victory on the heels of a dominant BJP performance in the national election.
Today, an AAP landslide doesn’t cause the same kind of shock and awe.
In many ways, Delhi is an exceptional place. It is a city-state (or Union territory) with more than half of its population having been born elsewhere, and which eschews the homespun identity-based arithmetic that characterizes its neighbour, Uttar Pradesh. For years, its chief ministers, such as Sheila Dixit and Arvind Kejriwal, have sought to distinguish themselves in infrastructure and service delivery.
For the BJP, which has been out of power in Delhi for more than two decades (and counting), there is no governance narrative to speak of in the state.
The BJP staked its campaign on two prongs: bald-faced attempts at polarization that saw the demonisation of the Muslim community and the arrest of Dalit leader Chandrashekhar Azad; and a promise to regularise (i.e., provide land titles) to Delhi’s myriad unauthorised colonies.
Judging by the results, and a drop of nearly 20 percentage points from the 2019 national election, on the whole the BJP’s campaign fell flat across the board. But the data can tell us a little about how BJP’s twin strategies fared with the electorate.
In the protests in the aftermath of the passage of the Citizenship (Amendment) Act (CAA), the BJP attempted to reap political dividends from Hindu-Muslim polarization. In the midst of the protests, it jailed popular Dalit leader Chandrashekhar Azad.
Unfortunately, it is hard to generate good data on the Muslim composition of various assembly constituencies. However, we were able to generate meaningful estimates of the percentage of the population in the scheduled caste (SC) community. The religious polarisation may have benefited the BJP vis-a-vis the SC community, while the jailing of Azad could have hurt it. So, how did the BJP fare in areas with larger SC populations?
Chart 1 shows the predicted vote share of the BJP in this election (after controlling for its electoral performance in 2019 in a statistical model) as a function of the percentage of SCs in an AC. The implication is that the BJP disproportionately lost vote share in areas with higher SCs populations. An increase of 20 percentage points in the SC population predicted a further drop of 3 percentage points for the BJP. Further investigation, however, is required to truly measure the electoral preferences of the SC community in Delhi.
A second important element of BJP’s campaign was to promise land rights to dwellers in Delhi’s unauthorised colonies — a promise that has been made by many parties over Delhi’s history. But the BJP’s promise carried extra heft as it was back by the Prime Minister and passed in Parliament. How much of an impact did this have on BJP’s fortunes this time?
In order to get some sense of the answer, we scraped the polling station lists and matched polling stations to localities. The matching of localities required complicated judgment calls about how to classify them (as planned, unauthorised, etc.) — as many localities include many types of settlement. We chose the prevailing form of settlement, but this necessarily masked small slums and colonies inside the localities. Nonetheless, using this procedure, we were able to determine an approximate percentage of an AC that is unauthorised.
Chart 2 shows the predicted vote share of the BJP in this election (after controlling for its electoral performance in 2019 in a statistical model) as a function of the percentage of polling stations that are in unauthorised localities in an AC.
While there is a lot of variance, the data suggest that the BJP did stem the tide somewhat in areas with more unauthorised colonies. An AC with 30 percentage points more in unauthorised population predicted about 2 percentage points more for the BJP. Nonetheless, more disaggregated data is required to provide a true colony-level analysis of voting preferences in Delhi —which will be available when Election Commission of India publishes the election results at the micro-level.
This election further showed that there is a strong divergence between national and state elections. But this is not just about what national and state leaders can deliver.
An analysis of the BJP’s campaign also provides evidence that national and nationalistic issues, and polarisation that worked so well for the BJP in national politics may have limited appeal in state elections — of course, no change in strategy would likely have changed the electoral outcome in Delhi. Sooner or later, the BJP will have to burnish its governance credentials if it is to be electorally competitive across the country.
This is a lesson it should remember as it heads into elections in Bihar, Assam, and West Bengal.