Incumbency: Curse or blessing?

RUNWAY 2019 History suggests that incumbency tends to hurt ruling party politicians in the country more than it helps. To the extent that this trend holds in the next Lok Sabha elections, the BJP’s recent political dominance would appear to harm its prospects of retaining its majority. Winning power is one thing, but staying in power is an entirely different matter

As the Bharatiya Janata Party (BJP) gears up for next year’s general election, its political dominance would seem to grant it a great many advantages over the opposition. Most obviously, the party commands a majority of seats in the Lok Sabha—the first single-party majority obtained in three decades. Its leader, Prime Minister Narendra Modi, enjoys widespread popularity despite setbacks associated with demonetization and the patchy rollout of the Goods and Services Tax. At the state level, the BJP has doggedly worked to expand its footprint to parts of the country where it has not traditionally held sway. Today, the BJP and its allies govern 20 of twentynine states—a far cry from four years ago, when they called only eight states their own.

While the BJP is undoubtedly the front-runner in 2019, a closer look at the party’s political standing suggests its current dominance could, counter-intuitively, be a liability rather than an asset. For starters, the conditions that paved the way for the historic mandate in 2014 are unlikely to materialise in 2019. That year presented something akin to a perfect electoral storm—a historically unpopular Congress Party, a lagging economy, and a charismatic prime ministerial candidate in Modi. The end result was a once-in-a-generation electoral rout of Congress. The BJP’s triumph, in turn, rested on running the table in a relatively small number of states—a feat many party sympathizers doubt can be replicated.

Furthermore, incumbency is something of a double-edged sword in India. In many democracies, incumbency provides politicians tools with which they can perpetuate their power. In India, research suggests that incumbent politicians are not just vulnerable—in many instances they are affirmatively disadvantaged. The BJP’s growing presence in India’s states helps offset the geographic concentration of its 2014 performance, but it does not fully inoculate the party from voter hostility; history suggests that state incumbency can also adversely impact a party’s chances in national elections. These indirect costs of ruling complicate the party’s electoral math; for a ruling party linking its stature to its consistent numerical strength in parliament, the negative spillovers linked to state incumbency is a source of concern.

CAN LIGHTNING STRIKE TWICE?

The first challenge the BJP faces is to repeat its dominance in its traditional strongholds. In the 2014 election, the BJP captured 282 seats out of the 428 seats it contested nationally (or 66%). While the BJP made inroads in new territory, most of its victories were concentrated in north and west India. In 2014, just eight states—Bihar, Chhattisgarh, Gujarat, Jharkhand, Madhya Pradesh, Maharashtra, Rajasthan, and Uttar Pradesh— accounted for 76% of the BJP’s parliamentary tally. All together, these states account for 273 seats, of which the BJP won 216 (nearly 80%). The efficiency with which the BJP translated votes into seats evinced the concentrated character of the party’s victory: with just 31%, it earned 52% of parliamentary seats (figure 1).

In addition, the BJP’s sweep in these core states rested in large measure on the Modi wave—a factor that is expected to be less potent in 2019. While Modi remains popular, the recent LoknitiCSDS-ABP News ‘Mood of the Nation’ survey found that support for the prime minister is tapering somewhat. As many as 44% in May 2017 named Modi their preferred prime ministerial choice. One year later, that percentage had declined to 34 %—marginally lower than where it stood in 2014, at 36%. A further deepening of these trends would suggest that some of Modi’s shine has worn off with the electorate.

THE CONGRESS PARTY’S HISTORIC UNPOPULARITY IN 2014

The second hurdle the BJP faces is to replicate its dominance over the Congress. Consider constituencies where the BJP and the Congress were the top two finishers; these head-to-head contests are good indicators of the parties’ relative popularity. In 2014, there were 189 such contests; of these, the BJP won 166, or 87%. The country had not witnessed such a lopsided battle since 1984 (figure 2)—when the Congress swept to power in the wake of Indira Gandhi’s assassination, winning 98% of direct BJP-Congress contests. The BJP’s decisive advantage over the Congress in 2014 can also be seen in the average margin of victory of their head-to-head contests, which touched 18%.

Today, the BJP and its allies maintain a sizable popularity advantage over their Congress counterparts, according to the Lokniti survey, but the gap is closing (see figure 3). In May 2017, 45% of respondents intended to vote for the BJP-led National Democratic Alliance (NDA) while just 26% sided with the Congress-led United Progressive Alliance (UPA); 29% favoured other, unaffiliated parties. One year later, the NDA’s share dropped to 37% while the UPA’s share grew to 31% (and the share belonging to other parties crept up to 32%).

INCUMBENCY DISADVANTAGE

The first two factors — the concentrated nature of the BJP’s victory and the historic unpopularity of the Congress—are specific to the 2014 contest. But there are also broader structural dynamics of anti-incumbency that shape the playing field. Anti-incumbency is a favoured bugbear of election analysts, who regularly toss it out as an explanation for nearly every underwhelming performance by a ruling party. But that loose talk conflates its multiple dimensions. Voters may punish individual incumbent candidates or ruling parties, and partisan incumbency in one tier of government may affect elections in another.

One approach by Nirmala Ravishankar disaggregates the negative effects of incumbency in national elections into direct and indirect costs, with a focus on how ruling parties fare. The direct cost of ruling is simply the impact of belonging to the national ruling party on an individual incumbent’s performance in parliamentary elections. According to Ravishankar’s calculations ,based on data from 1977 to 2004, an affiliation with the national ruling party reduces a member of parliament’s (MP’s) chances of reelection by 9 percentage points.

While the precise drivers of anti-incumbent sentiment are not well specified, there are many possibilities. The declining fortune of incumbents appears to be a post-1991 phenomenon, prior to which a Congress affiliation was the most reliable method of ensuring reelection. As the Congress’s dominance eroded and uncertainty soared, no single party label had a decisive advantage. In addition, given the centralized nature of Indian political parties, individual MPs have relatively little room to manoeuvre. Their ability to develop an individual brand might matter less than their party’s overall reputation, which can ebb and flow with time. Another possibility is voter frustration at incumbents for consistently failing to live up to their pre-electoral promises. Whatever the precise mechanism, the data suggest that an affiliation with the ruling party has become a net negative for incumbent parliamentarians.

INDIRECT COSTS OF RULING

But incumbency may also have indirect effects, in which party control at the state level affects how the state votes in national elections. This is relevant for evaluating how the BJP’s expanded presence could impact the national vote.

On first glance, a party’s performance in national elections is unaffected by its affiliation with a state-ruling party. Specifically, state-ruling parties appear to do just as well as opposition parties in the state when those same parties contest national parliamentary polls. But, as Ravishankar notes, this naive analysis ignores the role of timing. In many democracies around the world, governments enjoy a honeymoon period during their first two years in office when their popularity is at its apex. After this initial honeymoon period subsides, the party’s standing declines.

This is exactly what occurs in the Indian context. If a national election is held in the first year of a state government’s tenure, the state-ruling party does very well in national polls: stateruling parties are 18% more likely to win election at the national level than their colleagues out of government. Those effects remain positive through the state government’s second year in office, after which incumbency becomes a liability. After year three of a government’s tenure, the state-ruling party begins to suffer in national elections, and that penalty increases as the gap grows between the state and national election. For the BJP, this means that in states like Haryana, Jharkhand, and Maharashtra (where it is currently in power until late 2019/early 2020), the party will potentially face an uphill battle in the twilight of its tenure.

This interaction between state and national political performance will be a dynamic worth watching. Indeed, the honeymoon effect means that assembly elections later this year in Chhattisgarh, Madhya Pradesh, and Rajasthan will be highly informative with respect to how those states will vote in 2019. Parties that take power in each of those states in 2018 would be in the early stages of their honeymoon periods, allowing them to capitalize when those states vote in the Lok Sabha elections. Indeed, election data from these three states over the last three election cycles indicate, in every case,that the party that performed better in the state assembly polls had the upper hand in national elections—whether measured by vote share, seat share, or successfully forming the government (see figure 4). While the 2018 state assembly polls in these three states might not predict the overall verdict, they are a good barometer of how those states will vote in the general election.

CONCLUSION

Arguments relying on anti-incumbency are not ironclad but probabilistic statements: they suggest what is likely to happen, on average, within some margin of error. But there are certainly individual outliers. Furthermore, speculation about voters’ preferences at this stage is fully abstracted from campaign effects. As demonstrated in recent elections in both Gujarat and Karnataka, Modi’s intensive campaigning moved the needle in the BJP’s favour. In addition, the identity of the candidate on the ballot, their party’s alliance, and the choice set of prime ministerial candidates meaningfully shape voting decisions. Party leaders can try to minimise anti-incumbency by denying many of their elected members an opportunity to stand for re-election, a common strategy in Indian politics. However, related research suggests that ushering in fresh faces to contest elections only diminishes, but does not eliminate, the penalty of an affiliation with the national ruling party.

History suggests that incumbency tends to hurt ruling party politicians in India more than it helps. To the extent that this trend holds in 2019, the BJP’s recent political dominance would appear to harm its prospects of retaining its parliamentary majority. Winning power in India is one thing, but staying in power is an entirely different matter. Milan Vaishnav (@MilanV) is senior fellow and director of the South Asia program at the Carnegie Endowment for International Peace. Matthew Lillehaugen was until recently a junior fellow at Carnegie. This article is part of the ‘India Elects 2019’ series, a collaboration between Carnegie and HT