Can the BJP Do It Again?

Just a couple of years ago, the Bharatiya Janata Party (BJP) seemed certain to return to power in India’s 2019 national election. The BJP commanded a majority of seats in parliament, and its leader, Prime Minister Narendra Modi, seemed untouchable.

But nothing is certain in Indian politics.

In the last 18 months, India has seen increasing frustration among its farmers, high levels of joblessness, and an otherwise listless economy. In 2017, the BJP almost lost the state election in Gujarat, Modi’s home state and one which the BJP has ruled for more than 20 years. The frustrations finally boiled over. In December 2018, the BJP lost three key states in Chhattisgarh, Madhya Pradesh, and Rajasthan (two of which the party had ruled for 15 years).

Modi remains highly popular, but frustration with the direction of the economy has exposed a chink in his armor. Capitalizing on this opportunity, a bevy of regional parties have sought to create a “unified opposition” to oust the BJP from power.

The election has suddenly become competitive. Will the BJP and the prime minister be able to hang on to power?

Core BJP Support

While one often thinks about the core support base of Indian parties in terms of caste, occupation, or economic wealth, the BJP’s core support is best thought of in geographic terms. The previous national election was instructive in this regard.

In 2014, the BJP won 282 seats, a majority of the 543 seats available in the lower house of India’s bicameral parliament, known as the Lok Sabha, on just a 31 percent vote share. This extraordinary mismatch between seats won and vote share is due to the geographic concentration of BJP’s support. At the outset, we should be clear that in any first-past-the-post electoral system like India’s, there will be a discrepancy between vote share and seat share. This is because parties can still lose a constituency after procuring a significant vote share, and they can win a constituency by getting support from a fraction of the constituency (but more than the second place party).

The discrepancy between votes and seats for the BJP in 2014 tells us that whenever the BJP fielded a competitive candidate in 2014, it almost always won the seat. This only happens when support is geographically concentrated. Six states alone – Bihar, Gujarat, Madhya Pradesh, Maharashtra, Rajasthan, and Uttar Pradesh – contributed 194 seats to the BJP, 69 percent of the total number of seats won by the BJP. In these six states, the BJP won an astounding 91 percent of the seats it contested; we can call this effectively a “strike rate” of 91 percent.

This geographic BJP support base also structures the party system and party competition in India in important ways. In the vast majority of the core BJP territory, it is in head-to-head battles with the Congress Party. Indeed, of the 189 constituencies in which the BJP and Congress were the top two vote getters in 2014, the BJP had a commanding strike rate of 88 percent. By contrast, even in 2014, the BJP’s strike rate was about even (49 percent) in the rest of the constituencies. In short, the electoral fates of the BJP and the Congress are likely to be mirror images of each other because of direct competition. Furthermore, the spectacular decline in support for the Congress over the past few decades is not just due to the creation of new regional and caste-based parties; it is also a result of the BJP consolidating its electoral support base in its core regions.

The BJP is well aware that it will need to hold onto these core regions in large measure if it is to come back to power in 2019. Indeed, increasing disenchantment with the BJP in these regions — particularly among farmers and upper castes — has led the BJP to announce a flurry of policies in the run up to the election. For instance, this is the underlying reason for the BJP to announce income support that will largely benefit farmers and a scheme of job quotas based on economic status that will benefit the upper castes. It remains to be seen whether these last minute declarations are enough to bring disenchanted groups back into the fold.

The Turnout Effect

What changed in 2014? The answer is quite simple: The Modi Wave. But will it happen again?

Armed with a charismatic leader like Narendra Modi, who energized the youth, the BJP swept to power. The 2014 national election registered India’s highest-ever voter turnout rate, at 66 percent, a sharp increase from the 58 percent turnout in 2009.

There is a strong empirical relationship between the increase in turnout and support for the BJP in 2014. In the 116 constituencies in which the BJP contested the election and where the increase in the turnout rate was less than 5 percent, the BJP only had a strike rate of 24 percent. On the other hand, in the 312 constituencies in which the BJP contested and the turnout increase was more than 5 percent, the BJP had a strike rate of 81 percent.

Of course, the empirical relationship between turnout and BJP support need not imply that new or young voters supported the BJP, but there is survey evidence to suggest that the youngest voters were disproportionately supporters of the BJP. According to a post-poll survey conducted by Lokniti/CSDS in 2014, newly enfranchised young voters, aged 18 to 22, supported the BJP over Congress 39 percent to 19 percent. On the other hand, voters over the age of 56, who presumably had voted in multiple elections, supported the BJP over Congress 31 percent to 23 percent. Recent reports suggest that 45 million new, young voters will be added to the electoral rolls for 2019, with the pool of eligible voters swelling by about 5 percent. It remains to be seen whether the BJP will retain its large advantage among this population.

There is a certain synergy between the BJP’s geographic base and its ability to bring new voters to the polls. The variation in turnout is broadly driven by state-level factors, and a quick analysis of the data finds that increases in turnout were primarily driven by states that have historically had a low turnout, where a significant proportion of the voters did not feel included in the system. In fact, if one looks at the 15 largest states in India, the six states with the lowest turnout are precisely the same six states named above that accounted for 69 percent of the BJP’s seats in 2014. Each of these states saw at least a 9 percent increase in turnout between 2009 and 2014.

While it is clear that there was a groundswell of support for Modi and the BJP in 2014, there is much we simply do not know. Who were these new voters turning out at higher rates? Will they come back to the polls this time? For its part, the BJP claims that its superior party machinery will be able to manufacture turnout rates on par with 2014. But last time, the BJP and now Prime Minister Modi played the role of oppositional politics to perfection in the face of widespread disenchantment with the ruling Congress. This time the BJP is the ruling party. Can it stir the same passion in 2019?

Expanding the BJP Geography

For as well as the BJP did in 2014, it did not do a good job of breaking into states with strong regional parties and identities (particularly of the non-Hindi-speaking variety). Given the BJP’s strong performance in the states of Uttar Pradesh (UP) and Bihar this may seem like an odd statement — but the BJP was not “breaking into” these states. In Bihar, the BJP was a part of the ruling coalition until June 2013 and it was a formidable party in UP all throughout the 1990s owing to the Ram Janmabhoomi movement.

But things may be changing. Consider the state of West Bengal, a state where the BJP received its highest ever vote share at 17 percent in 2014. The BJP contested all 42 constituencies and won just two seats (one of which it had also won in 2009). It finished in the top two in just three other constituencies. Today, the Left Front and Congress (the main opposition parties) have all but fallen apart in West Bengal, and the BJP is likely to be the biggest beneficiary. In the recently conducted village/panchayat elections, the BJP secured more than 30 percent of the contested gram panchayat (village council) seats in the districts of Alipurduar, Purulia and Jhargram — which have significant proportions of India’s indigenous or scheduled tribe (ST) community.  Nonetheless, the ruling Trinamool Congress remains very strong through much of West Bengal (with 39 percent vote share in 2014), and the recent panchayat results also suggest that the Left Front and the Congress may still have enough votes to split the support for the BJP. For these reasons, as of now, even if the BJP is to see an increase in its tally in West Bengal, it is likely to be modest.

In 2014, the BJP won just three of the 12 seats it contested in Andhra Pradesh, one of the eight seats it contested in Tamil Nadu, one of the 21 seats it contested in Odisha (where it looks to do better this time), and none of the 18 seats it contested in Kerala. In the five major states of Andhra Pradesh (now divided into Telangana and Andhra Pradesh), Kerala, Odisha, Tamil Nadu, and West Bengal — the BJP had a measly strike rate of just 7 percent in 2014. As it turns out, if one looks at the 15 largest states in India again, these five states are among the six highest turnout states in India.

This data points to a major challenge for the BJP in extending its geographic base. Many of the non-Hindi-speaking states in India are highly mobilized (as the turnout shows) because regional parties are often the political manifestation of linguistic or caste social movements. The challenge for the BJP in these contexts is two-fold. First, the BJP must build an effective party cadre on the ground in places it has historically been absent (due to the traditionally limited linguistic and caste bases of the party). Second, unlike what it has achieved in its core supporting states by bringing out new voters, the BJP will have to convert voters who have previously supported other parties.

While there are good reasons to believe the BJP will expand its level of support outside its core geographies, especially in the northeast of India, the aggregate gains from this extension are unlikely to offset significant losses to the party in its core geographies. For this reason, the BJP’s recent policy declarations have sought to shore up its core geographic support rather than extending its geographic base.

The Case of Uttar Pradesh

The state of Uttar Pradesh (UP) merits special analysis. As India’s largest state, it sends 80 members of parliament to the legislature. It was also in no small part responsible for the BJP’s commanding performance in 2014. Out of 80 seats, 71 were won by the BJP (73 when one considers its coalition partners). This means that more than one out of every four seats the BJP won was in UP. Another sweeping victory for the BJP in the 2017 UP state election forced once bitter enemies, the Samajwadi Party (SP) and the Bahujan Samaj Party (BSP), to join forces to take on an ascendant BJP. Given the share of the BJP’s seats that came from UP in the last election, holding on to the state is crucial to win enough seats to form government in 2019.

To understand the plausible outcomes in UP, one has to make sense of how the new BSP-SP coalition is likely to perform. Until recently, these two parties were the top two parties in UP and bitter enemies. Their party organizations have been competing with each other and their respective caste bases, Yadavs and Jatavs (who are among the lowest or scheduled castes), do not like each other very much. Can anyone really expect them to work together well?

Actually, this has happened once before. In 1991 the BJP stormed to power in state and national elections (both of which were held in 1991) on promises of building a temple on the purported birth site of Ram, where a mosque (called the Babri Masjid) stood in the town of Ayodhya. In 1992, a group of Hindu nationalist activists destroyed the Babri Masjid. Sensing things were getting out of hand, the BSP and SP (originally under the Janata Party banner) formed an alliance and defeated the BJP in the 1993 election — although their coalition would fall apart in 1995 due to an incident in which BSP leader Mayawati accused then-SP leader Mulayam Singh Yadav of trying to kill her.

While it was a long time ago, and much has changed in UP, the 1993 election offers some lessons about what might happen this time. The BSP and the SP had similar results in 1991, with the BSP winning 10 percent and 11 percent vote share and the SP winning 13 percent and 11 percent vote share in that year’s state and national elections, respectively. But when the SP and BSP formed a coalition, they received 29 percent vote share, more than 6 to 7 percentage points higher than their combined vote shares in the previous election.

This is a trend that has also been observed in recent bypolls in UP (special elections for members of parliament when a seat has been vacated for any reason). In the bypolls of Kairana, Gorakhpur, and Phulpur, candidates contesting under the SP-BSP alliance have received more than the sum of the individual vote shares of the separated SP and BSP in the 2014 national election.

Why does this happen?

Coalitions are often “more than the sum of their parts” because they create a stronger opposition that draws more strategic voters. In 1991, an anti-BJP voter could have selected any number of parties, but he or she knew to vote for the SP-BSP coalition if the BJP was to be defeated in 1993. This is empirically borne out by the data. The SP-BSP coalition did not have much of an impact on the BJP’s vote share — it stayed between 32-33 percent in both 1991 and 1993. Rather, the coalition hurt the “other” opposition party, Congress. The Congress dropped from 18-19 percent in 1991 to 15 percent in 1993. This suggests that although the Congress may have introduced the charismatic Priyanka Gandhi to campaign for it in UP, the party’s vote share may suffer as voters strategically opt for a more formidable SP-BSP coalition.

Then again, times are different and the BJP is much more popular today than it was in the 1990s, procuring 43 percent of the vote share with its allies in 2014 and 41 percent vote share in 2017. And unlike 1993, the SP and BSP have a much longer history of enmity now, and there are serious questions about their ability to work together. On the margins, the alliance is almost certain to shave off some seats from the BJP, but how many? If one adds the vote shares of the BSP and SP (along with minor partner Rashtriya Lok Dal) in 2014 and treats it as a single party, the BJP and allies still win a respectable 36 seats (although a big drop from 73). This a reasonable benchmark from which one can adjust based on how well or poorly the SP and BSP coordinate with each other.

Recent State Results and Farmer Distress

A year and a half ago, it was a commonly held view that the BJP had the 2019 election locked up. But recent state elections results in which the BJP performed well below expectations, particularly in Chhattisgarh, Gujarat, Madhya Pradesh, and Rajasthan, have many questioning whether the BJP is losing popularity. Of course, national elections are different from state elections, but how much do these results tell us about the popularity of the BJP?

The states that recently went to the polls — Chhattisgarh, Gujarat, Madhya Pradesh, and Rajasthan — have traditionally been predictive of the national outcome, as they go to the polls soon before the national election. A simple analysis of the 2012-2013 state contests and the 2014 national contests finds that state elections are strongly predictive of the national result. Data from the Election Commission of India (ECI) allow one to analyze national election results down to the state assembly constituency (AC) level. Using this data, one finds that a 10 percent increase in AC vote share in the 2012-2013 state contests predicted a 5.7 percent increase in vote share in the 2014 national election.

These states are also easier to analyze because they do not have complicated coalition calculations as in Karnataka and UP. Instead, they are basically head-to-head contests between the BJP and Congress. Even without data, there is some prima facie evidence that the state results track national results closely. The BJP won state contests in Chhattisgarh, Gujarat, and Madhya in 2007 and 2008, and also won these states in the 2009 national election, even though Congress won the national election handily. The BJP won all four states in 2012 and 2013, and swept them in the 2014 national election. Of the 91 parliamentary constituencies in these four states, the BJP won 88 seats, a strike rate of 97 percent. More than 30 percent of all seats the BJP won in 2014 came from just these four states.

So what ails the BJP in these states today? There is a growing perception that the ruling BJP has an “urban bias” in policy that hurts the economic well-being of farmers. The state of Gujarat provides a good example. When I visited the mandi (agricultural wholesaler) during the 2017 Gujarat state election, I was told that the market price for cotton had dropped from about 1,400 Indian rupees per 20 kilogram bundle, during Congress rule at the center, to about 950 rupees per bundle. Sharp drops in market prices for the two most prevalent crops in Gujarat — cotton and groundnut — coupled with financial distress caused by notebandi (India’s 2016 demonetization exercise, where much of the country’s cash supply was announced to be no longer legal tender) had serious economic impacts on farmers. While it is debatable how much the market price should be the responsibility of the government, the average farmer associates the drop in agricultural prices with attempting to keep down inflation and allowing imports to flood the market.

The BJP came to power in Gujarat for the first time in 1995 under Keshubhai Patel, and has yet to relinquish control of the state. Until the most recent election, the BJP’s performance had been remarkably stable, with the BJP never winning less than 116 (2012) or more than 127 (2002) seats out of the 182 seats in Gujarat. But the results in the 2017 Gujarat state election bucked that trend. While the BJP held on to power, it did so weakly — winning just 99 seats. Analysis of the state election showed significant erosion of BJP’s support in rural Gujarat. Using 2011 Census data, one can classify the proportion of a constituency that is rural, as well as the percentage of farmers. In Gujarat’s large cities like Ahmedabad and Surat, the BJP’s strike rate remained at 90 percent in both 2012 and 2017. But outside the big cities, the BJP’s strike rate dropped from 55 percent to 44 percent. Further analysis shows that this drop was due to farmer anger. When restricting the analysis to rural constituencies, one sees that when less than 50 percent of the workforce is engaged in agriculture, the strike rate dropped slightly, from 80 percent in 2012 to 73 percent in 2017. But when looking at those constituencies where more than 65 percent of the workforce was engaged in agriculture, the strike rate dropped from 49 percent in 2012 to 30 percent in 2017.

This theme of farmer distress continued through 2018, culminating with the BJP losing in Chhattisgarh, Madhya Pradesh, and Rajasthan in December 2018 after significant erosion in rural support. Two of those states (Chhhattisgarh and Madhya Pradesh) had been ruled by the BJP for 15 years. In the interim, other economic issues, like high rates of joblessness, have also cropped up. Responding to the frustrations of farmers, the ruling BJP cleared the way for an income transfer to farmers just before the elections. It remains to be seen if this is enough to allay the frustrations of farmers in India, or whether Modi’s charisma can offset the wrath of farmers that could not be avoided in recent state elections.

Nationalism and Military Action Against Pakistan

On February 14, a suicide bomber in Jammu and Kashmir’s Pulwama district killed 40 Central Reserve Police Force (CRPF) troops in the deadliest attack on security forces in Kashmir since 1989. The Pakistan-based terrorist group Jaish-e-Mohammed (JeM) claimed responsibility for the attack. In response, India launched airstrikes deep into Pakistan, and Modi received widespread support in India for a decisive response.

This helped the BJP change the public mood. The public had taken to complaining about the BJP’s performance on a host of economic issues, just before the elections were announced. It also allowed the BJP to shift its electoral narrative to one of nationalism and protecting the country with a decisive leader like Modi. How much of an impact will it have on the voter?

To get some sense of how military action against Pakistan might impact voters, it is useful to look at the 1999 national election. In 1998, the BJP and then-Prime Minister Atal Bihari Vajpayee came to power with a fairly unstable coalition. By July 1999, Indian forces had scored a decisive victory over Pakistan in the Kargil War. Due to a number of factors, including the Kargil War, the BJP called for early elections in 1999. The 1999 national election returned the BJP to power in a more stable coalition that would last five years (ending a period of instability in which India saw three national elections in three years).

The BJP’s average vote share (in seats it contested) rose from 36 percent in 1998 to 40 percent in 1999, pointing to a positive impact from the military action on BJP’s vote share. But the change in support was far from uniform. While the BJP gained 7.4 percentage points in Rajasthan, it actually lost vote share in three major states it contested — Karnataka, Maharashtra, and Uttar Pradesh (losing 7.3 percentage points in UP).

So what explains where the BJP gained vote share? The BJP primarily gained against the Congress. In head-to-head contests against the Congress, the BJP’s vote share increased from 44 percent in 1998 to 48 percent in 1999. On the other hand, the BJP actually lost vote share in direct competition against other (regional) parties — from 40 percent in 1998 to 36 percent in 1999. This underscores the benefit to the BJP when it is able to compete on “national narratives” against the Congress. When it must compete on regional terms, the nationalistic strategy is less successful, something we saw in the analysis of the 2014 national election as well.

Of course, the aggregate impact of the post-Pulwama airstrike may be entirely different. But the data we have suggests that, to the extent that the airstrike affects voters, the impact will be largely on head-to-head competition between the BJP and the Congress. These are exactly the areas that make up the BJP's core geographies, so it may serve the BJP’s electoral aims quite well. On the other hand, the Congress is significantly weaker today than it was in 1999, and the BJP will still have to perform well when it competing against regional parties.

Toward the 2019 General Elections

In 2014, everyone knew the Congress was going to lose power months before the election; the only question was how badly the Congress would lose. This time there is suspense. Most analysts believe that the BJP will be the single largest party after the election, but, given the commitment of many smaller parties to oust Modi from power, this may not be sufficient to return the BJP to power, even in coalition.

The broad contours of the campaigns are clear. The BJP will seek to structure its campaign around the charisma of Modi and national issues. The opposition parties, on the other hand, will try to keep the focus on the middling performance of the economy and constituency-level issues.

Election season has begun, and we are seeing what can euphemistically be called the pageantry of Indian democracy. Expect it to be crazy. Expect it to be chaotic. But, above all, expect the unexpected.