The 2018 assembly election in Meghalaya has for the moment led to a hung verdict. While we wait for a resolution of the conundrum, data can reveal information on the outcome of this election, as well as on longer term trends in Meghalaya electoral politics.

A stable turnout

Turnout was stable in Meghalaya, at 86.8% (1 percentage point down compared to 2013). Women have been outvoting men and while we do not yet have data on gender-based voting for 2018, it is safe to assume that 2018 was no exception.

Turnout in Meghalaya elections does not vary much from one constituency type (general or reserved) to another. Participation tends to be higher in the Jaintia Hills (in the east) and in the Garo Hills (in the west). The capital, Shillong, registers lower turnout than the rest of the state, at an average of 75.2% turnout.

The Garo Hills is traditionally a site of high political mobilisation, which translates into higher participation. This is the only sub-region in Meghalaya where participation increased, compared to 2013.

Stable number of political players

Contrary to most Indian states where there has been over the past 15 years an inflation of candidates, the total number of candidates in Meghalaya elections remain stable. Contrary to most Indian states also, the average number of candidates per seat is quite low, at six candidates per seat. Fifty eight percent of the candidates retained their deposit in 2018, which is quite high.

Again, contrary to most states, the number of parties contesting elections remains stable (14 against 15 in 2013). The number of parties represented in the assembly keeps to slowly increasing however, attesting to a gradual fragmentation of the political space in Meghalaya.

No clear strongholds in Meghalaya elections

The constituency-wise party winners’ map reveals that the two main parties – Congress and National People’s Party – do not really have particular strongholds. Both parties tend to do well in the Garo Hills and the Congress usually has an edge in the Khasi Hills area.

This time, the National People’s Party progressed in the Jaintia Hills sub-region, although this region is the smallest of the three, in terms of seats (seven). The National People’s Party won three seats, two of them thanks to two Congress sitting MLAs who shifted their allegiance to the NPP ahead of the election.

Long-term trends in terms of vote share reveal some strong stability. The Congress oscillates between 28% and 35% of vote share since 1978. With 28.5% of vote share in 2018, the Congress goes back to its 2003 performance. It drops more than six points of vote share compared to 2013. Most of those losses are located in the Khasi Hills sub-region.

Barring the Nationalist Congress Party, which has literally vanished from the political stage, most small regional parties have gained vote share in these elections. The total vote share of independent candidates, traditionally high in Meghalaya, has decreased by 17 points, from 27.7% to 10.8%. This means that voters were not ready this time to waste their vote on small parties or independent candidates and opted strategically for the party that seemed likely to win the seat locally.

The distribution of votes across small parties, combined with the good performance of the National People’s Party, explains why the Congress drops nearly 13% of the seats. Not only did the National People’s Party perform well, but the Congress candidates were also confronted to other strong smaller regional party candidates, whose vote share tends to be concentrated in specific seats rather than throughout the state. Thus, these elections were really a multi-corner fight for the National People’s Party and the Congress.

The relative good performance of small actors (strong in a few seats each) explains why none of the two main parties could convert their minority of votes into a majority of seats, like in states such as Uttar Pradesh, or Gujarat.

The National People’s Party makes the most of its vote share, converting 20.6% of vote share into 32% of the seats. The Congress gets 35.6% of the seats for 28.5% of the votes.

Vote share density maps reveal that main party performance remains fairly distributed across the territory, despite the sub-regional concentration of seats. This basically confirms again that elections in Meghalaya are fought seat by seat, and that local issues and local context prime over state-level (or national-level) narratives.

There isn’t much variation in the Congress performance between 2013 and 2018. Congress loses more vote in the Jaintia Hill sub-region and in the Eastern part of the Khasi Hill region.

The National People’s Party vote share performance map confirms sub-regional trends. In 2018, the National People’s Party scored much higher in clusters of seats in the extreme east of the state (the Jaintia Hills) and in the west (Garo Hills).

Its performance in and around the capital is extremely poor. In fact, the National People’s Party hardly contested any seat in the eastern half of Meghalaya in 2013. Its geographic expansion strategy did pay off. Not only the National People’s Party was able to conquer new territory, but it could also consolidate its hold in its earlier area of implantation, in the west.

The BJP’s relatively good performance – it increases its vote share from 1.3% to 9.6% – is fairly distributed across the territory. The two seats it could win are located in and near the capital (South Shillong and Pynthorumkhrah), where turnout was the lowest in the state.

The dispersion of the BJP vote explains why it did not get more than two seats. But thanks to the improved BJP performance however, the aggregate vote share of main parties has regained its pre-2013 level. Most voters do privilege main players.

Finally, parties’ strike rate provides another measure of performance. Not every party contests all seats and the conversion of seats won against the number of seats contested (a.k.a. strike rate) gives a better appreciation of party performance.

In the present case, party-level strike rates confirm the other party performance charts. Not only the National People’s Party contested more seats this election (51 against 32), but its conversion rate was also much higher (37.3% against 6.3%). By comparison, the Congress’s strike rate decreases while the party contested every seat in both elections.

More volatility than meets the eye

An examination of constituency-level trends reveals more volatility and churning than the aggregate trends would suggest.

In particular, there is a great deal of volatility between parties and candidates. Many seats change hands and these elections saw a large number of turncoats contesting on main parties’ tickets.

The first measure of competitiveness is the victory margin, or the distance in votes between the winner and the runner-up. The shorter the margin, the closer the election.

As in 2013, 27 seats were won with less than 5% of victory margin. As in previous elections, this means that small numbers of vote could have drastically changed the result. The National People’s Party won 10 seats with extremely low margin, against six won by Congress. The 11 remainder seats were won by small regional parties, who spoiled the bipolar contest between the two main parties.

A comparison of the distribution of victory margins between 2013 and 2018 reveals that the 2018 was a slightly more competitive election than the previous one, with a slightly flatter pattern.

The constituency-wise winners’ margin map reveals that victory margins tend to be higher in NPP strongholds, although there are variations across sub-regions.

Another measure of volatility is the number of seats that exchanged hands. This year, a staggering 77% of the seats (46 out of 60) have elected an MLA from a different party than in the previous election.

The previous chart reveals that this is a longstanding trend in Meghalaya elections. Individual incumbency runs very high, regardless of overall party performance.

Part of the explanation for this phenomenon is the number of turncoats that ran in these elections.

The defeat of the turncoats

This year, 63 candidates switched party affiliation before these elections. The BJP was the largest recipient of turncoats, with 14 candidates borrowed from other parties. The Congress lost 17 former candidates, including eight incumbent MLAs.

There is no evidence however that this strategy paid off. Only two turncoats out of the 14 BJP ran got re-election, both incumbent MLAs (Alexander Laloo Hek, from Congress, and Sanbor Shullai, from NCP). Only six of the 17 Congress turncoats were re-elected. And only one of the 13 UDP turncoats got re-elected. In total, only 17.5% of the turncoat candidates got re-elected (11 out of 63), on a variety of party affiliations.

Incumbent performance was not great either. 44 sitting MLAs out of 59 re-ran, most of them on Congress tickets (21) and on NPP tickets (10). Only 19 were re-elected (10 for Congress and 4 for NPP). Counting for the seven ex-MLAs who were re-elected, the new assembly counts 52.5% of newcomers.

One implication, which may at first seem paradoxical, is that while overall trends are stable, the Meghalaya assembly has to deal with very inexperienced legislators after almost every election.

It’s a no for NOTA

NOTA usually registers better scores in tribal areas, such as Chhattisgarh, Jharkhand, or the tribal belt of Gujarat. It is not the case in the tribal dominated state of Meghalaya. Less than 1% of voters have pushed the NOTA button in this election, that is less than fifteen thousand voters. Contrary to other states, NOTA did not make any difference in these elections (a higher NOTA vote share could alter the fate of the winner in a closely fought election).

Finally, there is another register in which Meghalaya, alongside its North-Eastern neighbors, distinguishes itself. Despite a higher than usual number of women candidates – 31 out of 361 (they were 25 in 2013) – only three women were elected in the assembly – two Congress and one National People’s Party. Agatha Sangma, daughter of PA Sangma and former Union Minister, barely wins her seat in South Tura, West Garo Hills, with a paltry 28.5% vote share. The dispersion of women candidacy between parties and the fact that a third of them run as independent candidates explain partly why so few women make it to the Meghalaya assembly.

In conclusion, a constituency-level analysis of election results in Meghalaya reveals far greater volatility and instability than overall results suggest. The good performance of small player ensures that the disproportionality effect of the electoral system does not lead to a skewed representation of parties, relative to their vote share performance. Local factors predominate and the presence of a large number of turncoats contributes to the blurring of the lines between parties. The result is a highly fragmented political landscape which, this time, has led to a hung assembly.

Coalition negotiations have started. It remains to be seen how these play out.

The Trivedi Centre for Political Data team is led by Gilles Verniers, Assistant Professor of Political Science at Ashoka University and co-Director, TCPD. Basim-U-Nissa, Mohit Kumar, Ashish Ranjan and Sudesh Kumar have contributed to the data. Raw data available at http://lokdhaba.ashoka.edu.in.