Indian voters do not vote on the basis of development or policies, according to a recent study, confirming the results of the 2019 general elections.

Despite a raft of economic woes afflicting the country, ranging from joblessness to farm distress, the Bharatiya Janata Party has swept to power in India, winning 56% of seats.

The study, Do Citizens Enforce Accountability For Public Goods Provision?, by Oxford University scholar Tanushree Goyal took into account election results across 14 states that hold 90% of India’s population. It linked the performance of the Pradhan Mantri Gram Sadak Yojana (prime minister’s rural roads project) to the results of state and central elections between 1998 and 2017.

A flagship programme of the earlier BJP-led government headed by Atal Bihari Vajpayee, Pradhan Mantri Gram Sadak Yojana was launched in 2000. By 2018, the programme had resulted in more than 5,50,000 km of roads being laid across rural India at a cost of Rs 28 million crore.

Even when the programme was successfully implemented in villages, bringing connectivity, progress and better access to public facilities, voters did not vote back the same government, the study showed. For example, between 1998 to 2003, the Congress government in Rajasthan constructed 13,634.43 km of road across the state. Yet, in the 2003 elections, the Congress lost an average vote share of 9.6% and the BJP came to power in Rajasthan. Between 1999-2004, the Telugu Desam Party governing the then united Andhra Pradesh, constructed 8167.56 km of roads. However, in the 2004 election, the Telugu Desam Party lost its vote share by 7.3% and the Congress came to power.

“My empirical analysis suggests that while the world’s largest rural roads provision programme provided all-weather roads to over 200,000 Indian villages, most of which lacked paved roads and desperately needed all-year market access, the electoral effects of roads provision were close to zero over time, space and electoral levels,” Goyal said in her study. “The very few times the results are significantly positive or negative, they are far too feeble and inconsistent, to have an electoral impact for the incumbent or incentivise future policy provision.”

The Pradhan Mantri Gram Sadak Yojana was picked to ascertain the presence or absence of a policy vote in India because roads are a highly visible public good, easy for the voter to observe and evaluate, Goyal explained. “Moreover, there is causal evidence that PMGSY [Pradhan Mantri Gram Sadak Yojana] brings the poor better access public services in health, education, and agriculture,” she said.

Shift away from vikas

In the campaigning months that preceded the 2019 general elections, the mention of infrastructure projects has steadily declined in the speeches of Prime Minister Narendra Modi. The election campaign leading saw an abundance of everything except debates on policies, former finance minister P Chidambaram noted in a comment in The Indian Express. Prime Minister Modi has been accused by the Opposition of focussing on the Congress party, its president and his family instead of the vikas (development) that he had promised on coming to power.

In a series of articles titled Modi’s Report Card, had assessed the BJP government’s flagship projects such as the rural electrification programme, rural jobs scheme and the Swachh Bharat (clean India) mission.

The Oxford study has dire implications for democracy, according to Goyal. “...if voters remain unresponsive to policy provision, electoral incentives for public goods provision or improving service delivery, may diminish over time, leading to abandonment or poorer implementation of these crucial programmes”, she said in a blog post.

Parties pay for ‘cost of ruling’

Goyal used detailed, lowest-level roads data from the Pradhan Mantri Gram Sadak Yojana and aggregated it to national and state-level constituencies to cover approximately 11,000 electoral races – parliamentary and assembly – for almost the entire duration of the roads programme.

The researcher has argued that if accountability is not observed for roads, “it is unlikely that we observe it for other policies”. Given their links to improved general social and economic welfare, roads are widely regarded as a “signal of competence” in the Indian context, as per Goyal.

Harish Ramaswamy, professor of political science, Karnatak University, Dharwad, traced the disconnect between votes and development to illiteracy. “One of the arguments about the Indian democracy is that it thrives because of illiteracy,” he said. “The electoral manifesto is read only by 1%-2% of the voters.”

However, in her study Goyal has also explored the possibility of voters enforcing more accountability with higher education and better availability of information. She found that awareness about the scheme and even visible proof of road construction did not influence voters in both national and state elections. Information is not a constraint for this programme but its electoral effects remained null.

Since citizens do not vote on policy performance, incumbency does not bring political parties an advantage in the Indian context. On the contrary, it results in the “incumbency disadvantage” or the “cost of ruling”, according to Goyal.

“Incumbents have a comparative advantage over oppositions because they can use their policies to signal competence more effectively, while opposition cannot do so,” she said. “However if citizens’ do not take performance into account, incumbency does not bring this comparative advantage.”

What this meant for 2019 elections

“This election was not fought on the basis of any policies,” Goyal said. “It doesn’t matter how the government has performed in the last five years and that is why Modi speaks about things like nationalism and air strikes. Look at how BJP’s discourse has changed from 2014 to 2019. Then it was a mix of development and Hindutva but now it has changed only to identity and caste.”

So if development is not electorally rewarding then why should any government invest time and funds in constructing roads?

“Constructing roads may not help but not providing them may do great harm,” said Anupam Manur, economic research fellow at Takshashila, a think tank. If the economy is not doing well it works against political parties.”

There could also be other reasons why people do not vote for roads, said Sandeep Shastri, national coordinator of Lokniti Network, a think tank, and pro-vice chancellor, Jain University. “In many places, people lose their lands and farms for the construction of the roads and they may also receive inadequate compensation. Moreover, PMGSY is a centrally funded scheme that is implemented by the state government and people may not know who to give credit for the project.”

Shastri also questioned the representativeness of roads as an important parameter to vote for. “[W]e also need to see what constitutes building blocks of development from the people’s prism. Roads is only one factor and people consider multiple factors, like unemployment and price rise while going to vote,” he said.

In her paper, Goyal ruled out the possibility that the presence of both members of parliament and members of legislative assembly might lead to attribution errors. She also ruled out the likelihood of citizens not favouring the incumbent government for roads provision due to corruption concerns. “I find that citizens do not respond to variation in quality of roads: they neither reward good-quality roads nor punish for poor quality,” she wrote in her blog.

On the contrary, Shastri said parties’ development records matter in another way. “Democratic battles are battles of perception. In 2014 Modi’s campaign was strong on the non-performance of Congress. They were voted to power as they could build that perception. For the opposition development becomes a tool to attack the incumbent,” he said.

Votes are cast on ‘emotional plank’

While the research does not dwell on what Indians vote for if not for development, Goyal is of the opinion that it is caste. “But this is not to say that their choices are unwise or uninformed,” she said. “Caste plays a central role in Indian social and political life. While this could have bad outcomes for development and policy, it simply means that the electorate cares about other things.”

There is a need to reimagine caste today, Shastri added: “Caste should not be looked as a primordial identity. It is a modern political identity which is used to seek advantages. People also vote on the basis of caste as their life experiences are the same.”

Votes are cast more on an emotional plank than on caste, Ramaswamy of Karnatak University said. “They do vote on the basis of caste but that is a small percentage. The remaining vote a certain way because they are emotionally moved and that is why the parties are talking about nationalism or Ram Mandir. Every party tries to milk the emotional card to their benefit,” he said.

Voter preference could be viewed as a hierarchy of needs with individual benefits right on top, followed by community benefits and lastly benefits from public goods, said Manur from Takshashila. If the benefits are dispersed, the voter does not see it as a personal gain. “For most voters at the time of voting they are not thinking of the newly-constructed roads but of individual gains. That is why in an election cycle, political parties work towards providing public goods in the initial years and towards the end entice voters with freebies. Free saris, laptops, money go a long way in winning votes,” he said.

This article first appeared on IndiaSpend, a data-driven and public-interest journalism non-profit.