What are the representational outcome of the 2021 Kerala State election? What is the sociological composition of the new assembly and what has been the representation trajectory of various groups and communities, as well as other variables of social diversity such as gender and religion, over time?

The state of Kerala boasts an image of inclusive politics. In this article, we look at various socio-demographic variables to examine this claim, and see how various parties and alliances have fared with regard to inclusion.

No caste or community dominates the assembly

The first observation is that no group holds more than 25% of the 140 seats of the Kerala assembly. Many Vidhan Sabhas across India are dominated by a single group of castes or sometimes by a particular caste.

The largest group represented are the Muslims, with 24% of the seats. Members of the Other Backward Classes hold the same share of seats: 19% by Ezhavas legislators, 6% of Thiyas and 2% of other OBCs.

Upper Castes make for 20% of the seats. Of these, 19% belong to the Nair community. There are 21% of Christian MLAs, themselves divided into many different groups and communities (Syrian Christians, Roman Catholic, Jacobites, and so forth).

Similarly, Muslims are also divided into various groups and castes. Finally, 10% of the MLAs are members of the Scheduled Castes and 1% are from the Scheduled Tribes, as mandated by quotas.

Most major groups then have a roughly equal representation within the assembly, but some castes and communities predominate largely within them.

Over time, we clearly see that the gaps that existed between these groups has been closed. In 2016, the four major groups (Christians, Muslims, upper castes and OBCs) all occupied between 21%-22% of the seats. The historical domination of the upper castes ended in the late 1980s-early 1990s. The rise of religious minorities – both Muslims and Christians – has been slow but constant.

If we divide these large groups into smaller entities (the main group represented within each category and the other clubbed together), we see that there has been more diversity of trajectory than the previous chart suggests.

The most impressive change has been the rise of Ezhava representation from the late 1980s, and the disappearance of a range of other backward castes, like the Nadars, the Dheevaras, the Kulalas and various other groups.

The representation of non-Ezhava OBC groups collapsed after the 1996 election, won by the Left, led by EN Nayanar. The only exception is the rise of the Thiya community, from 2006 (Thiyas are historically considered as Ezhavas but in recent years, members of this community have objected to being clubbed with Ezhavas). In 2016 and 2021, A handful of Thiya legislators were elected in a cluster of seats near Kannur and Kozhikode.

Other groups have been more stable. There is a growing representation of Christians over time, from a low 13.6% in 1980 to 23.6% today. Nair representation has always oscillated between 18% to 22% with greater variations in some years. Like the non-Ezhava OBCs, non-Nair upper caste representation has been a constant decline, since 1977.

Comparing the representation of major groups within both alliances does not throw up many variations. First of all, representation patterns among candidates within these two alliances are practically identical through time.

The differences between the two alliances are not stark either. The Left Democratic Front, which won the election, fields more Ezhava candidates (22%) than the United Democratic Front (14.2%). The Left Democratic Front fields more Muslim candidates than its rival (24.2% against 15%). Both of them field similar share of Nair candidates (15.7% for the Left Democratic Front and 16.7% for the United Democratic Front).

Considering that both alliances also partner with small parties and independent candidates outside their alliance, including small Muslim parties like the National Secular Conference, these differences fade even more in reality.

In terms of outcome, we see a similar kind of diversity. In 2021, the 62 MLAs of the Communist Party of India (Marxist) belong to 18 different castes and communities. The Congress’ 21 MLAs belong to 11 different groups and communities, half of them (10) Christians.

We see then that one of the reasons for the relative stability of representational patterns comes from the fact that major groups are well represented within alliances, albeit with small variations.

One explanation for this is the social and political geography of caste, which reveals that these communities tend to be clustered in specific sub-regions or districts, and therefore get included by both alliances, owing to their dominating local presence.

Geography of caste and community representation

Plotting the caste or caste groups of the MLAs elected in 2021 shows that there are distinct clusters across Kerala’s three sub-regions.

Ezhavas are elected in three clusters across the three sub-regions: in Travancore, around Thiruvananthapuram and Kollam, in Central Kerala between Thrissur and Ernakulam, and in the Malabar region, in the Kannur and Pallakkad districts. Most Ezhava MLAs are elected in interior districts, as opposed to coastal districts.

Twenty six of the 32 Muslim candidates elected are from the Malabar region, mostly in Malappuram district, then Kozhikode, Palakkad, Kannur and Kasaragod districts. Four more Muslim legislators have been elected in Central Kerala and in Travancore (Aroor and Eravipuram constituencies).

Christians get elected across the three sub-regions but more won in Travancore, near Kottayam, Pathanamthitta and Alappuzha. Eight of the 14 MLAs of Ernakulam district are Christian. These urban seats appear smaller on the map.

The 2016 map looks eerily the same as the 2021 map, which confirm that the stable nomination patterns we saw earlier are grounded in geography.

That of course does not come as a surprise. Communities get elected where they are found in larger numbers. But there are many other factors at play than simply numerical strength. To what extent do these group command influence over local political economies, local institutions? Is there competition from other group or the fact that both alliances co-opt members of these communities election after election preclude the possibility of competition?

We see therefore that caste and community diversity is an aggregate concept that rests on the perpetuation of local patterns of dominance, sustained by political parties through candidate selection.


Another illustration of the limits of the Kerala inclusion model is women’s representation. Barring 1996, women’s representation in the Kerala assembly has been one of the lowest in the country.

Between 2001 and 2016, they barely occupied 5%-6% of the seats. In an assembly of 140 this time, it means only seven women legislators. Out of 1,939 MLAs elected in Kerala since 1965, only 84 have been women, or 51 individual women to be more precise, accounting for multiple victories by the same candidates.

We also see that women’s participation has been declining in recent years despite a significant overall rise of women candidates. In 2021, they made up 11% of all candidates, but got only 8% of the seats. This is a higher ratio than in the previous four elections, but still a very low number for a state that otherwise boosts the best women-welfare related statistics in the country.

To understand the gap between nomination rates and victories, we need to break down these numbers by alliances and parties. The following chart shows without ambiguity that the level of women representation depends on the performance of the Left Democratic Front, under whose tickets nearly all of the women elected over the past four elections have been contesting.

Since 2006, the United Democratic Front has nominated 33 women out of 525 candidates. Shockingly, only two of them have been elected over the period. PK Jayalakshmi was elected in 2011, in Mananthavady, on a Congress ticket, and KK Rema, the lone candidate of the Revolutionary Marxist Party of India, was elected in 2021 in Vadakara. Both were also first-time contestants.

Rema is the widow of a Communist Party of India (Marxist) dissident who was brutally killed in 2012.

The nomination chart for the two alliances show that the United Democratic does nominate women candidates, although at a much fewer than the Left Democratic Front (about half the number before 2021). It did slightly better in 2021 but without great results, since only one of its 12 women candidates got elected.

The Bharatiya Janata Party-led National Democratic Alliance nominated 14.5% of women candidates. But none got elected.

Party-wise nominations show the within the Left Democratic Front, the Indian Union Muslim League does not nominate women candidates, except for Advocate Noorbena Rasheed, who ran in Kozhikode South in 2021 and was the runner-up. She was Indian Union Muslim League’s first woman candidate since 1996, when Qamarunnisa Anvar contested in Calicut-II constituency. She also came second.

With 10.8% of women candidates, the Congress does nothing to compensate for the absence of women in its main alliance partner. There lies the first explanation for the poor performance of the United Democratic Front in terms of gender inclusion.

A second explanation probably lies with the voters of the United Democratic Front, who do not throw their support to women candidates as much as to male candidates. Among Congress candidates, women got an average of 36.7% of vote share, against 38% for the men. It could also be the case that women are nominated in more difficult seats.

The difference in terms of outcome is even starker. The 11 women elected in 2021 have all contested on the tickets of the CPM (8) or Communist Party of India (2), with the exception of KK Rema of the Revolutionary Marxist Party of India.

Seven of these 11 women were first-time contestants. The only woman MLA to have served more than two terms is the health minister, KK Shailaja, who won her seat with the largest margin recorded in the state in years.

Thus, not only does the map of women representation illustrate the state of under-representation of women, it also shows that the few women who do get elected fail to last in politics and do not get elected from the same constituency.

Muslim representation on the rise

Kerala is, with Assam, one of the rare states in India where Muslim representation is substantial and on the rise (the erstwhile state of Jammu and Kashmir technically no longer has a state assembly).

Since 1967, there has been a constant gradual rise of Muslim representation both among the candidates and the MLAs. There were fewer Muslim candidates in 2021 compared to 2016 (19.3% against 22.6%), but there is one more Muslim legislator (33) than five years ago.

According to the 2011 Census, 26.6% of the population of Kerala are Muslims. They have thus acquired a near proportional representation over the past three elections, which is not the case among the candidates. This reflects the demographic concentration of Muslim communities that we mentioned earlier.

Those figures, however, are a bit misleading as they could suggest that the nominations of Muslims are on the rise across the board. If we isolate the four main parties (Communist Party of India (Marxist), Communist Party of India, Congress and Indian Union Muslim League), we see that their nomination of Muslim candidates is in fact stable, even though more and more Muslim get elected on their tickets.

The next two charts give the detail by party of nomination and representation of Muslims in Kerala. It confirms that the Indian Union Muslim League is the main provider of representation to Muslims, followed by the Communist Party of India (Marxist), which has contributed more to Muslims’ representation since 2006.

We also see that in many elections, a handful of independent Muslim candidates also get elected (three of the six independents elected in 2021).

Thus, the Indian Union Muslim League no longer has a monopoly on Muslims’ participation in election as candidates and legislators. The greater inclusion of Muslim in other parties is what accounts for the recent rise of overall representation.

The map reveals that most Muslims are elected in a large cluster in the south of the Malabar region, in Malappuram district and around.

Kerala’s ‘new’ political class

Until recently, it was a given that most sitting MLAs would be given a chance to retain their seat in the following election. With little inter-caste competition for tickets at the ground level, parties do not have many incentives to throw away their incumbent MLAs ahead of an election.

This election was different, as the Communist Party of India (Marxist) implemented a “maximum two-term policy” that was officially intended to infuse new blood in the party and reward long-term party workers who had for long been working in the shadow of local bosses. This leads to the drop of re-running incumbent ratio, at 56%.

The re-election ratio, or strike rate, of these sitting MLAs is however quite high, around 80% average since 2011. This is rare in India as in most states, voters are more inclined to reject their incumbent candidates than here. Once again, it may have to do with the stability of representational patterns and party affiliations. It may also mean that many MLAs in Kerala are in fact doing a good job.

The number of first-time MLAs is the highest it has been since 2006.

Seventy eight MLAs are serving a first term, most of them (49), without surprise on a Left Democratic Front affiliation.The veteran of the assembly is Oomen Chandy, former two-time Congress chief minister, elected for the 12th time in Puthupally, where he contested first in 1970.

The following chart shows that the “maximum two-term” policy was applied with a few exceptions, as three candidates who had served more than two terms did contest on Communist Party of India (Marxist) tickets in 2021.

They were Pinarayi Viajayan, the chief minister and six-time MLA from Dharmadam; K Radhakrishnan, former speaker of the Kerala Vidhan Sabha and former minister, five-time MLA from Chelakkara; and KK Shailaja, minister for health and four-time MLA from Mattanur.

But by and large, the rule was applied.

Kerala’s shrinking professional political class

As a result of this Communist Party of India (Marxist)-induced turnover of long-standing sitting MLAs, the size of Kerala’s “stable” political class (MLAs having won more than two terms in a given assembly) has become very small. Only 34 of the 140 MLAs meet that criteria: ten from the Congress, six each from the Communist Party of India (Marxist) and Communist Party of India, five from the Indian Union Muslim League, three independents and four on other party tickets.

One notable turncoat this year was PJ Joseph, a ten-time MLA from Thodupuzha who returned to the Kerala Congress after having run twice in the dissident faction of Kerala Congress (M). The advocate Mons Joseph, a five-time MLA from Kaduthuruthy, performed the same movement.

The chart also shows that few of the longer-standing candidates lost their race, only six out of 40. This goes on to show that Kerala’s professional politicians tend to be solidly anchored in their constituencies, which makes the two-term policy of the Communist Party of India (Marxist) unusual.

Kerala’s strong party affiliations

Turncoats are simply not a phenomenon in Kerala. Year after year, the number of individuals contesting on a different party ticket is very small. The only two turncoats who contested on a strong party ticket lost. Advocate Gafoor Illis, the Communist Party of India (Marxist) candidate in Tirur, lost. He had previously contested the same seat on a National Secular Conference ticket. So did Stephen George, the Kerala Congress (M) candidate in Kaduthuruthy, formerly with the Kerala Congress (AMG).

Both alliances cultivate strong party affiliations and favour nominations of party workers. This is similarly to what we saw in Tamil Nadu. As a result, candidates have no incentives to leave their parties as their support base – and party workers – would not follow them.

An assembly of elders

If the “maximum two-term” policy was meant to infuse new blood in the party and in the assembly, it did not exactly infused young blood. The median age of legislators in 2021 is 53 against 51 in 2016. Thirty six MLAs are above 65, 44 are between 55 and 65. The youngest member of the assembly is 27-year-old Advocate KM Sachindev, the Scheduled Caste Communist Part of India (Marxist) MLA from Balusseri. At 73, his party colleague Thottathil Ravindran from Kozhikode is the oldest member of the assembly.

Congress MLAs overall were more youthful. Seven of them are younger than 35. Otherwise, the age distribution of Communist Party of India (Marxist) candidates is more skewed towards the seniors. One reason is that the party tends to nominate long-standing party associates, as a reward for long careers spent on consolidating local party structures. This is a model that rewards local political experience.


One would expect the state with the highest literacy ratio to have the most educated assembly in the country. That does not seem to be the case as even though most MLAS hold a graduate degree or above (80), there are still 36 MLAs who haven’t studied past the 10th standard and 18 who stopped after the 12th standard.

These less-educated MLAs are found mostly among the Communist Party of India (Marxist) (24) and the Communist Party of India (5), perhaps underlying an element of class inclusion among their candidates. Otherwise, the educational profile of legislators across parties does not vary particularly.

UDF with the richest candidates loses

Another peculiar aspect of Kerala politics is the differentiated role of money between parties. Not that money does not matter to communists, but there is a stark difference of average personal wealth between the main Left Democratic Front candidates and the main United Democratic Front candidates.

In 2021, 70% of all Indian Union Muslim candidates were crorepatis, against 43.7% for Congress. The Communist Party of India (Marxist) and the Communist Party of India are lagging behind with 26.2% and 20% of crorepati candidates.

However, the difference diminishes when one compares the MLAs from the candidates. Interestingly, the richer Communist Party of India (Marxist) MLAs seemed to have performed much better than their poorer counterparts. This goes to show that even if money isn’t everything in Kerala politics compared to, say. Karnataka or Tamil Nadu, it is still something and does make a difference.

That does not seem to be the case though for Communist Party of India (Marxist) MLAs who are marginally richer than the Communist Party of India candidates. The phenomenon observed with Communist Party of India (Marxist) MLAs was already there in 2016, signalling another change in the political culture of the party.


What do we learn from this exercise? First that the Kerala model of inclusion is not incompatible with the perpetuation of local phenomenon of social and political exclusion. Most seats tend to be trusted by members of the same caste over time and across party affiliation. They just happen to be different castes and groups across the territory so when one aggregates the data, it creates an illusion of inclusion.

A well distributed balance of power between major groups at the state level is then compatible with local patterns of dominance. In fact, one could say that the latter reinforces the former.

The Kerala model of inclusive politics thus has this important limitation. The other important limitation is the little space carved by parties for women candidates, in a state where women’s participation in elections is high, and where women participation in every aspect of public life other than state politics is also high. The United Democratic Front, and its Indian Union Muslim League partner, are to blame, but so are the communists, who have been slow at making efforts in that regard.

The 2021 election looks a lot like the 2016 election, signalling that the changes that took place with the rise of Pinarayi Vijayan at the helm of state politics in Kerala are on the way of becoming trends.

Gilles Verniers is Assistant Professor of Political Science at Ashoka University, Co-Director of TCPD and Visiting Senior Fellow at the Centre for Policy Research.

Basim-U-Nissa is Affiliate Researcher at TCPD.

Neelesh Agrawal are Research Engineers at TCPD.

Ayaan Sagar, Jenish Raj Bajracharya, Mayank Sharma, Niharika Mehrotra, Paravi Sapra, Prashasti Agarwal, Priavi Joshi, Pulari M. Baskar, Rashmi Guha Ray, Shreya, and Shreyashree Nayak have contributed to the data, with the TCPD team.

We particularly thank Hasna Shahitha for caste and community data collection and Dr Nissim Mannathukkaren, Associate Professor in the Department of International Development Studies, Dalhousie University, for vetting and consolidating the data.