Winning back the rural heartland


back to issue

THE performance in the rural parts of India by the Bharatiya Janata Party (BJP) is perhaps the most impressive element of its sweeping victory in the 2019 election. Low food inflation, slow economic growth, and notebandi (India’s ‘demonetization’ exercise that severely restricted money stock) caused a number of analysts to suggest that rural India was facing economic distress. This seemed to be borne out by recent state election results in Chhattisgarh, Madhya Pradesh and Rajasthan – all once considered BJP stronghold states that the party lost to its rival Congress – less than six months before the national election. Further analysis of the electoral results from these states confirmed that the BJP had been particularly punished in rural areas in the state elections,1 and it seemed natural to assume these trends would continue to the national election. But the BJP actually increased its tally from 282 seats in 2014 to 303 seats in 2019, further consolidating its vote share in rural India.

What explains the BJP’s extra-ordinary electoral performance in rural India? And what lessons can we draw from this performance in thinking about the Indian voter? A brief discussion on the nature of the crisis which plagued the rural economy before the 2019 elections would be useful. A lower middle income country like India is always keen to increase its economic growth rates. A comparison of compound annual growth rate (CAGR) of Gross Value Added in agriculture during the first term of the Narendra Modi government (2014-15 to 2018-19) shows that it was marginally less than what its immediate predecessor, the second United Progressive Alliance (UPA) government (2009-10 to 2013-14) achieved.

However, it is not lower growth which was the biggest problem for farmers under the government. What really put a squeeze on farm incomes under the present government was a worsening of the terms of trade for agriculture. Simply speaking, terms of trade for agriculture can be defined as the ratio of prices received by the farmers and prices paid by them. As is obvious, this will be some kind of a ratio of food and non-food prices. When Narendra Modi assumed office as the prime minister for the first time in May 2014, food inflation was at double digit levels. The end of his term was marked by near-zero, even negative food inflation. It was largely the sharp fall in food inflation, which has allowed the government to keep headline inflation numbers in check. Food items have an almost 50% share in the commodity basket, which is used to calculate India’s benchmark inflation rate, namely the Consumer Price Index (CPI). That inflation was not an issue under this government was a claim which constituted an important part of the BJP’s electoral narrative.


A negative or near-zero food inflation, combined with significantly positive non-food inflation, means that a farmer will have to sell more and more physical quantities of his output to be able to buy the same amount of non-agricultural goods. One of the answers to the question of why the BJP overcame headwinds of rural distress in the 2019 elections is to be found in the different trajectories of food and non-food inflation and the inability of the political opposition to successfully take up this issue. Here is one explanation: Given the extent of poverty and dominance of food in the average consumption basket, higher food prices are not an unambiguous good in the Indian economy. While the farmers stand to make economic gains when food prices increase, a large section of the population in both urban and even rural (such as agricultural workers who do not own any land and are net buyers of food) are adversely affected by it. This means that low food inflation is a burden on the farmers but a boon for large parts of the economy.

This is exactly why no opposition party mounted any serious effort to blame the government for low food prices. This is not to say that the rest of the economy is not affected by low food inflation leading to low farm incomes leading to a squeeze in rural demand. However, the economic processes involved here are far too complicated to be obvious to the electorate.


What the opposition offered to the voters was a bunch of clichéd promises such as farm loan waivers and assured procurement prices, which were clearly not enough to take on the BJP’s popularity which comprised many things, from nationalism to caste equations, going for it – not to mention a powerful party organization to counter any narrative of rural distress.

The most rural parts of the country are in its ‘Hindi heartland’, a region consisting of about 40% of India’s population, which has for some time constituted a stronghold for the BJP. It is also a region largely characterized by two party contests between the BJP and the Congress. If the BJP were to falter in rural areas, the Congress was likely to be a disproportionate beneficiary – especially as the Congress is associated with a time in which farmers received better prices for their crops and the Mahatma Gandhi National Rural Employment Guarantee Scheme (MGNREGS).


But, contrary to state-level electoral trends, the BJP ceded no ground to the Congress in the national election. In 2019, the BJP and Congress had 190 head-to-head contests, meaning that they were the top two finishers in the constituency. In these 190 constituencies, the BJP won 175 contests for a whopping 92% strike rate against the Congress. In 2014, the BJP had a similarly large strike rate of 88% against the Congress.

Figure 1 displays the predicted victory margins for the BJP in 2014 and 2019 in the seats it competed head-to-head against the Congress, using a statistical technique known as LOESS. What is noticeable is a secular increase in the margins of victory between 2014 and 2019, with the greatest gains and highest margins of victory coming in the most rural regions.

In order to characterize the extent to which a parliamentary constituency is rural in nature, we made use of 2015 satellite data from the European Space Agency. Analysing these data provide clean estimates of the percentage of the land area that can be classified as ‘rural’ – a far more robust measure than the out-of-date 2011 Indian Census that is also susceptible to statutory categorization errors. We are grateful to Shamindra Nath Roy at the Centre for Policy Research for providing us with this data.

A number of macroeconomic indicators suggested that the Indian economy was, and continues to be, quite sluggish. A simple theory of economic voting suggests that the BJP should have been hurt at the polls this time, especially in the rural belt. But politics is far more complicated. Armed with a well financed and well organized party machinery, the BJP was far more effective at ‘reaching’ the voter, rural or urban. How did the BJP accomplish this feat?


First, one cannot deny the sheer popularity of Prime Minister Narendra Modi. Even after the state election defeats in December 2018, polling data found the levels of support for Narendra Modi surging. A nationwide survey conducted by Axis-My India in March 2019, as India was about to go to the polls, found that 52% of respondents preferred Narendra Modi as the next prime minister (as compared to only 33% for Rahul Gandhi).2 Notably, this was an eight percentage point increase in Modi’s popularity from six months earlier.

Many political analysts have tried to analyse these election results through ‘issues’ like joblessness and rural distress. These results suggest a principle that is well understood by scholars of voting behaviour. Voters were drawn to Narendra Modi first, deciding reasons to support him thereafter – whether it be quality of leadership, standing strong against Pakistan, or because of the centrally sponsored schemes. Indeed, a significant amount of research points to the idea that voters select a preferred candidate or party and allow them to frame how issues should be perceived.3

Second, the extraordinary penetration of Narendra Modi into all means of communication – whether it was television, the print media, social media, or WhatsApp – is noteworthy. In order for the popularity of Narendra Modi to translate into electoral support, the BJP needed to control the means of communication with the voter. While data cannot fully capture Prime Minister Modi’s dominance in communication, a suggestive data point is the prevalence of Google searches about a politician. Although users of Google are likely to be skewed towards a younger, more educated, and urban population, the penetration of various political leaders on Google is a reasonable indicator of the extent to which politicians are able to reach citizens – even in rural India which is rapidly becoming more connected to social media and WhatsApp. When one looks at Google searches about politicians over the election period, a whopping 75% of the searches were about Narendra Modi, compared to just 12% about Rahul Gandhi.4


Narendra Modi is India’s first ‘media prime minister.’ Since the day Modi was elected, actually much before, his image had been carefully choreographed to appeal to the voter and maximize votes. It is a phenomenon that has been seen in the United States with presidents like Ronald Reagan and Bill Clinton, but never has India had such a media and image savvy prime minister. This control over the means of communication with the voter has made Narendra Modi’s character nearly unassailable. This is perhaps why the Congress ‘chowkidar chor hai’ (the watchman is a crook) jibe at Prime Minister Modi failed to capture the imagination of the electorate.


Finally, the BJP’s most noticeable advantage is in its party organization. With Amit Shah as party president, the BJP built a well oiled, highly efficient machine which effectively melds quantitative data with a dedicated party cadre.5 In the rural countryside, the strength of the BJP’s party organization confered an extraordinary advantage – especially against the Congress. In the Hindi belt, it is common to visit villages where only BJP workers (and no Congress workers) are visible. This ‘last mile’ efficiency of the BJP in reaching the voter guaranteed control over communication between Narendra Modi and the citizen.

Of course, none of this would be possible without significant financial resources. The BJP receives nearly all of the electoral financing in the system. After the controversial ‘electoral bond’ system was implemented to allow for anonymous political donations and remove certain limits on corporate donations, one can only calculate to whom donations are given, not the source. The data from fiscal year 2017-2018 provided by the Association for Democratic Reforms (ADR) shows that the BJP received 210 crore out of the total of 222 crore in electoral bond financing, a whopping 95% of all electoral financing through the electoral bond method.6

Assessing the impact of the BJP’s control over communication and party machinery on electoral outcomes is challenging. Perhaps one of the best indicators of the quality of party machinery is the extent to which a party can bring its own supporters to the polls. Historically, an increase in voter turnout at the constituency level has been associated with anti-incumbency in the constituency.7 There is a political accountability logic to this empirical regularity. When voters are aggrieved, their frustrations compel them to come to the polls and vote out the incumbent. But BJP insiders insist that their superior party machinery has flipped this empirical relationship. According to their political mobilization logic, an increase in voter turnout occurs precisely where the BJP machinery is more active in getting their own supporters to the polls – that is, a turnout increase which follows a pro-incumbent logic for the BJP.


In 2014, an increase in turnout at the constituency was strongly associated with the BJP victory. In constituencies where turnout was less than five percentage points, the BJP had only a 24% strike rate, but with a turnout increase of more than five percentage points, the BJP had an 81% strike rate.8 The 2014 national election saw, at the time, the highest ever turnout in Indian history at 66%. The strong turnout for the BJP in 2014 is consistent with the narratives of strong anti-incumbency for the then ruling Congress as well as the result of the strength of BJP’s voter mobilization. The 2019 national election, with the BJP as an incumbent, affords us an opportunity to disentangle the anti-incumbency and political mobilization narratives.


The 2019 national election saw very high voter turnout once again. At 67%, it was the highest turnout ever recorded in Indian history, besting the record turnout just five years earlier. Figure 2 plots the predicted turnout change against the percentage of rural land area in a constituency, once again using the LOESS technique.

As can be surmised, some of the highest increases in turnout were associated with the most rural of areas. This empirical relationship can also help in adjudicating between the two narratives of turnout. Were rural citizens more likely to vote because they were frustrated with economic distress? Or, were they more likely to vote because rural areas are precisely where the BJP’s strong party organization is effective? As we now know, the BJP increased its vote share most significantly in rural areas – suggesting that turnout in rural areas was more a function of political mobilization than anti-incumbency. Indeed, a simple predictive plot of the probability of BJP victory against turnout (using binary logistic regression) shows that turnout is positively related to BJP’s chances for victory – providing definitive evidence of the political mobilization narrative (see Figure 3).


Rural areas, which have much more stable population flows, strong social networks and established patterns of caste and religious interaction, make it much easier for parties to mobilize voters. Thus, as the BJP continues to build and strengthen its party organization, it is solidifying its rural base of voters. Despite significant evidence of economic distress in rural areas, and recent state-level setbacks, the BJP demonstrated in this election that it can win back the rural vote on the strength of its organization. This is another demonstration of the power of the BJP’s ascendant party machine.


* Views expressed are personal.


1. Neelanjan Sircar, ‘BJP Strike Rate Drops in Both Rural, Urban Areas’, The Hindustan Times, 12 December 2018. https://www. areas/story-4nxyl3Y22knj4b8Z0H zEbK.html. Accessed 7 July 2019.

2. These data are from the ‘Political Stock Exchange’ presented on India Today, using data from Axis-My India. https://www. Accessed 7 July 2019.

3. See, for instance, Gabriel Lenz, Follow the Leader? How Voters Respond to Politicians’ Policies and Performance. University of Chicago Press, 2012.

4. ‘On Google, Narendra Modi’s Stock Continues to Rise’, Live Mint, 20 May 2019. Accessed 7 July 2019.

5. For a clear exposition on how the BJP’s party organization has evolved, see the recent work by Prashant Jha, How the BJP Wins: Inside India’s Greatest Election Machine. Juggernaut Books, Delhi, 2007.

6. Niranjan Sahoo and Niraj Tiwari, ‘How Electoral Bonds Distort India’s Democracy.’ Accessed on 7 July 2019.

7. Neelanjan Sircar, ‘Lok Sabha Elections 2019: How Does Voter Turnout Impact Election Results?’ The Hindustan Times, 3 May 2019. Accessed 7 July 2019.

8. Neelanjan Sircar, ‘Voter Turnout and the Modi Wave’, Transitions Blog, 8 July 2014. Accessed 7 July 2019.