Counting trans lives, miscounting realities

Social stigma and data gaps limit transgender enumeration in India, complicating the State’s policy design

In January 2026, the Karnataka government announced findings from a statewide survey on transgender persons. This is undeniably significant because the National Family Health Survey (2019-21) and the Periodic Labour Force Survey do not even enumerate transgender persons as a distinct demographic category, despite their recognition as a separate gender category by the Supreme Court of India.

The recent Karnataka survey identified 10,365 transgender persons in the state, which is, however, only about half the population recorded in the 2011 National Census. It also significantly diverges from estimates produced by other datasets. Electoral data from the Karnataka assembly elections (2023) listed 4,927 registered transgender voters, while Election Commission figures from the 2024 Lok Sabha polls recorded 5,012 transgender electors. In contrast, community activists’ macro-estimates for 2024-25 place their population at over 200,000. A recent ethnographic study conducted as part of an ongoing project at the Institute of Social and Economic Change (ISEC), Bengaluru, finds an approximate estimate of more than 300,000 transgender people in Karnataka.

Inconsistencies are visible in other state datasets as well. The statewide transgender survey conducted in Kerala (2014-15) estimated their population at over 25,000, which is an exponential divergence from the mere 3,902 persons recorded in the 2011 Census. These inconsistencies across surveys signal layered processes of undercounting, which deepens the statistical opacity surrounding transgender population in India.

Thus, what is at stake is not merely demographic precision but persistent data inconsistency that shadows transgender realities in India. The questions that arise are: why do transgender populations appear differently across datasets? And why do the State and its enumeration regimes continue to find transgender populations difficult to count?

Why do the numbers differ? The first explanation lies in methodology. Typically, population surveys rely on door-to-door enumeration, simultaneously mapping spatial coverage and demographic triangulation across documentation systems. In contrast, relying on community outreach and voluntary disclosure makes surveys contingent on network reach and willingness to self-identify. This introduces data distortions: those outside welfare databases, facing bureaucratic barriers in updating certificates, or geographically dispersed beyond outreach clusters risk exclusion. The dataset thus becomes vulnerable to network reach and assumptions of homogeneous identity assertion, rather than ground realities.

The second explanation lies in definitional rigidity. Due to the Euro-American origin of the term ‘transgender’, many local non-binary identities do not neatly fit into this category. Therefore, who is counted as transgender varies across categories, welfare eligibility criteria, medical certification regimes, and community self-identification practices. Moreover, there are significant variations in terms of individuals who currently identify as transgender; those who transition into binary male or female identities after medical or social transition; and those who occupy a liminal space within ritual and cultural frameworks. Thus, rigid definitional categories struggle to resolve who counts. In turn, it becomes a regulatory act to compress the very population into homogeneity, which it claims to represent.

Third, stigma shapes participation. Uneven response rates, due to a wariness of being included in state databases, are widely acknowledged as a structural challenge in surveys that involve gender minorities. Disclosure can carry perceived risks, ranging from employment precarity and housing discrimination to heightened institutional scrutiny. For transgender community, where historical interactions with the State have been marked by surveillance, exclusion, or bureaucratic friction, willingness to self-identify within official data systems remains understandably uneven.

Fourth, in states like Karnataka that experience a high rate of migration and mobility, demographic capture is bound to be complex, especially for transgender communities who often have limited access to structured labour markets. Consequently, they often inhabit circulatory geographies, moving between districts, informal economies, and seasonal labour networks. Thus, typical survey methods struggle to capture populations that are spatially fluid and temporally mobile.

Finally, household rejection produces statistical erasure. A significant proportion of the transgender population lacks family acceptance. Expulsion from natal homes pushes them into alternative kinship structures unrecognised by typical survey methods, rendering them demographically invisible.

These discrepancies are not technical errors. They emerge from structural conditions that make transgender lives difficult to enumerate within conventional demographic frameworks. It is also precisely what makes enumeration consequential to frameworks of recognition, resource allocation, and policy design.

Beyond the statistics

Counting matters because numbers produce policy legitimacy. Welfare boards require demographic baselines, reservations depend on quantification, and budget allocation follows population scale. Therefore, any survey of the socially and historically marginalised transgender communities must move beyond extraction towards collaboration. Such an enumeration requires partnership with community networks, involving experts, and most importantly, recognising the dynamic forms of transgender identity and a method that is informed by their realities. Without such shifts, surveys risk becoming instruments of symbolic inclusion rather than structural transformation.

While the Karnataka survey signals administrative willingness to map the welfare needs of the marginalised and produce targeted interventions, which is welcome, it also reveals the epistemic limits of demographic knowledge itself in other Indian states. Transgender lives exceed the statistical frames designed to contain them. They move across everyday experiences of precarities, kinship reconfigurations, and identity transitions that resist bureaucratic fixity. Enumeration, in this context, becomes necessary, but often insufficient. There is already a lot of work in trans scholarship which argues that AI’s facial recognition software is a regulatory regime for trans people and how it perpetuates biases.

The noted inadequacies in transgender enumeration undermine the sustainable formulation of economic, housing, and social policies for them, which are missing in most Indian states. But Karnataka and Kerala provide examples for other Indian states to follow, not only to make these populations inclusive but also to contain inconsistencies in their counting, as most states move to present their budgets.

(The writers are, respectively, a post-doctoral scholar and a professor at the Institute for Social and Economic Change)

Disclaimer: The views expressed above are the author's own. They do not necessarily reflect the views of DH.