Modi was set to lose 2024 like Vajpayee in 2004. Here’s what changed

Why did Narendra Modi come back to power? Why did the NDA manage to secure a majority? Why did the BJP not face a more comprehensive defeat?

It is a mark of the power of false narratives that we have not started to ask these questions. Much of the post-election analysis in the media is still stuck in a mistaken and self-serving question: What explains this unexpected electoral setback for the BJP? The question is mistaken, for the surprise in this instance lies in the eyes of the beholder.

The commentariat has not started reflecting on the possibility that there is nothing surprising about a bad, non-responsive and arrogant government losing an election, that the shock was entirely the creation of the media. It is self-serving as it draws the discussion exactly to the kind of minutiae blame games that helps to cover up the real issues. But once the dust settles, we can hope for more attention to the real issues and ask the counterfactual question: How did the BJP scrape through in an election that it was going to lose badly?

A precedence from the past

A comparison with the election of 2004 is very instructive here. Recall that it was the conclusion of a ‘successful’ five-year term of the NDA government led by the charismatic Atal Bihari Vajpayee. Riding on the slogan of ‘India Shining’, he was ‘widely expected’ to come back to power in 2009. All the pre-election polls and indeed the exit polls forecast a clear majority for the ruling coalition. Yet, the results belied all expectations and left everyone bewildered. Just as it did this month.

Fortunately, we have a post-poll survey conducted by Lokniti-CSDS in 2004 that asked the same questions that have been asked by the Lokniti survey after this election, whose findings have been made public.

Modi’s popularity ratings in 2024 are not very different from those of Vajpayee in 2004. When people were asked to name their choice for the PM, 38 per cent named Vajpayee, as compared to 41 per cent for Modi this year. In 2004, the nearest rival was Sonia Gandhi at 26 per cent. This year it was Rahul Gandhi at 27 per cent. Peoples’ satisfaction with the Vajpayee government was a shade higher than that with the Modi government: net satisfaction (satisfied minus dissatisfied) was 29 per cent in 2004 compared to 23 per cent in 2024. The critical question about whether the incumbent government should be given another chance elicited similar responses: Vajpayee was favored 48 to 30 percent, similar to Modi’s 46 to 39.

 

  The striking similarities between what happened in 2004 and 2024

	2004	2024
People’s choice for the next PM
Vajpayee/Modi	38	41
Sonia/ Rahul Gandhi	26	27
Satisfied with Central Govt?
Satisfied	57	59
Dissatisfied	28	36
Another chance for incumbent Govt?
Yes	48	46
No	30	39
Poll projections
Average for NDA in pre polls	287	373
Average for NDA in exit polls	257	339
Actual seats for NDA (BJP)	181	292

Source for 2004 and 2024 survey data: Lokniti-CSDS National Election Study 2004 and 2024; Average for NDA in pre polls and exit polls are authors’ calculations.

 Yet, Vajpayee lost the 2004 election badly. Exit polls had predicted anything between 230 and 275 seats for the NDA. The alliance ended up with 181. The BJP was unseated and the UPA government was formed.

Why did that not happen in 2024? One good answer could be that the NDA’s starting point in 2024 was much higher than it was in 2004. Back then, the NDA was a fledgling coalition of 23 parties held together by Vajpayee whose BJP had won only 182 seats in the 1999 elections. An electoral setback pushed the BJP down by 44 seats. This is not very different from what happened to Modi’s BJP, down from 303 to 240.

A counterfactual scenario

At the same time, the question remains: Could the outcome have been substantially different? Could the BJP have been ousted from power in 2024? Once we stop focusing on the wrong question and look at the election outcome with new spectacles, we can see that Modi saved his government by the skin of his teeth.

The scenarios presented in Table 2 and Table 3 present us with various possibilities. Let’s take the final outcome of this election as the starting point. If there was an additional 1 percentage point swing against the NDA (loss to NDA and proportionate gains to its principal opponent), it would have lost 18 seats. A national-wide swing of 1.5 pp would have brought the NDA tally down to 261, well below the majority mark. It would have also brought the ruling coalition slightly below its principal rival: 261 for NDA and 263 for INDIA. Another half a percent would have taken it down further to 246 and INDIA above the majority mark at 275.

 Just 1.5 percentage point Uniform National Swing would have unseated the BJP

 

Scenario/ Seats	BJP	Allies	NDA	Congress	Allies	INDIA	Others
Actual outcome	240	52	292	99	135	234	17
If 1% swing against NDA	224	50	274	105	146	251	18
If 1.5% swing against NDA	213	48	261	111	152	263	19
If 2% swing against NDA	201	45	246	118	157	275	22

  Table 3 shows a more realistic scenario. Instead of assuming a uniform swing across all the states, it presents the likely outcome if a national-wide swing was distributed unevenly across different states, concentrated more in Hindi heartland states where there was a momentum against the BJP that could be pushed further. If we assume an additional swing of 2 percentage point each in Uttar Pradesh, Maharashtra, West Bengal, Bihar, Rajasthan, and Haryana, the NDA would be down to just 260 seats, well below the majority mark. In that scenario, the BJP would be at 214, too far away from staking a claim to form the government. All these would add up to just one percentage point national swing against the BJP.

 Just 2 percentage point swing focused in select states, amounting to just over 1 pp (-1.1pp) national swing would have unseated BJP

 

SCENARIO: 2 pp SWING AWAY FROM NDA
State Name	Change in NDA seats	Change in BJP seats
UP	-15	-13
Maharashtra	-5	-2
West Bengal	-4	-4
Bihar	-3	-2
Rajasthan	-3	-3
Haryana	-2	-2
All India total	-32	-26

 Remember this: Modi’s BJP was about 1 percentage point away from a comprehensive defeat that would have forced it to sit in the opposition.

Now the question for us and future historians to ponder is: what could have made a difference of 1 percentage point national vote? What may have helped the BJP avert this disaster? Did the BJP leaders know something that all of us did not know, or were prevented from knowing?

One obvious answer is alliances. It is now evident why the BJP had struck a strange alliance with Nitish Kumar (JDU) and Jayant Chaudhary (RLD), why it did a U-turn to join hands with Chandrababu Naidu (TDP) and why it pursued every ally in Bihar and Maharashtra. The allies not only shored up the numbers for the NDA, they also helped the BJP pick at least 10 additional seats in Andhra Pradesh, Bihar, Maharashtra, and Uttar Pradesh. In retrospect, we can also see the cost of TMC-Congress rift in West Bengal (3 seats) and the damage done by the VBA of Prakash Ambedkar (4) and AIMIM of Asaduddin Owaisi (1).

Consider another answer. Just imagine that the mainstream media had covered this election a shade less unfairly than it did. Lokniti-CSDS post-poll survey tells us that 83 percent of the voters had a TV at home, 66 percent watched news channels every day or sometimes (compared to 47 percent who received information from social media), all but a fraction named one or the other Godi media channel as the source of their information on elections. Just imagine what would have happened if TV news had presented a balanced picture of the performance of the Modi government? Let alone speak truth to power, if only they had not run blatant propaganda for the ruling party and the cringe-worthy interviews of the Supreme Leader? Imagine if they had simply reported that the 2024 election was not a one-way race, that it might be a close contest?

Remember, if just 1 out of the 66 people watching the sarkari propaganda had changed their mind, the Supreme Leader would have been the Leader of Opposition today.