National datasets differ on women's land rights (WLR) because they use different criteria in their calculations. Some estimates include only agricultural land, others include homestead land, while still others include land leased for cultivation, as we detail later.

Women's access to and control over productive resources, especially ownership of agricultural land, is key to their economic empowerment, social status, physical security, and their own and their families' overall well-being. The United Nations' Sustainable Development Goals and the Food And Agriculture Organization's voluntary guidelines endorse gender equality, with the former adding two SDG indicators (1.4.2 and 5 a 1) for annual reporting on the status of women's land rights. India currently uses the Agriculture Census (2015-16) to report on women's land rights to track SDG progress and for the FAO's Gender Land Rights Database.

To measure the real progress in women's empowerment, accurate figures are needed: How much land do women hold, is it family or individual ownership, and what is the size and quality of the land owned? Is the type of land owned agricultural or other, what is the basis of ownership--documentary or perceived? What type of tenure is involved? Without sex-specific data for these, estimates drawn up using different methodologies can be misleading, as women's land rights scholar Bina Agarwal said in a recent paper.

Across the board, various data-sets show women's land ownership in India to have risen slowly but steadily over the years, though varying across states and with rural areas showing bigger improvements. With countries now measuring efforts to narrow the gender gap by acting on inheritance laws, it has become imperative for India to monitor 'gendered' land rights data. India has a rich land data ecosystem, said Shipra Deo, director for women's land rights at the non-profit Landesa, which can be used to gauge the true extent of women's land rights in the country and help close the gender gap by aligning with sustainable goal indicators.

Datasets, diversity and patriarchy

Estimates of women's land ownership such as the Population Census, the Socio-Economic Caste Census (SECC) and the more recent and frequent India Human Development Survey (IHDS) are reliable, robust and open-access, mostly collected and reported by government agencies, yet none offers a clear picture. The National Family Health Survey (NFHS), in 2015-16, started providing data on the percentage of women who own land alone or jointly.

The NFHS data, as explained in this paper on its fourth round of surveys, have several weaknesses related to gendered land rights. For example, it seeks land rights information from only 15% of sample households; from women in the 15-49 years age group and men in the 15-54 years age group; and clubs together all lands, agricultural and non-agricultural. This paper, and another presented at the World Bank Land Conference 2017, showed up the use of diverse methodologies in terms of granularity, sampling techniques, periodicity, dissemination methods as well engendering.

Further, survey data may not match those in land records. For instance, NFHS-4 figures are almost four times those of IHDS 2011-12, and present a geographic pattern that is contrary to earlier findings of intensive local research, according to this April 2020 paper by Bina Agrawal, professor, development economics and environment at the University of Manchester's Global Development Institute. While these and other national analyses show more land ownership among women in the southern states, NFHS-4 figures show 50% of women in Bihar and 23% in Kerala own land.

That women continue to be marginalised in terms of land ownership and also in land-data ecosystems was also highlighted in this IndiaSpend story on the the WLR index constructed by the Center for Land Governance by pulling together diverse databases in 2018. Published in the State of Land Report 2018, the index put India's average WLR at 12.9%. It used multiple datasets to arrive at this estimate--women's operational land holdings from the Agriculture Census of 2011, the share of adult women owning farm land from IHDS 2011-12, the share of women-headed households owning land from the Socio-Economic Caste Census 2011, and the share of women owning house and/or land (solely or jointly) from NFHS-4.

In all these datasets, there are wide variations in state- and district-level numbers, our analysis showed. These differences can be traced to regional, historical and socio-cultural diversities and the fact that land is a state subject and inheritance of agrarian lands is often decided by state succession laws. The family is taken as a unit and land recorded in the name of the head of the household, almost always a male. While some datasets refer to homestead land and property, others collect information only about agricultural land, and still others refer to both.

How data on land ownership, disaggregated by gender, is recorded differs by state, though most states do not have a gender column in their land records. Women's identity and land ownership is subsumed in the identity of the household, a major reason why women are still not widely recognised as farmers despite positive policy intent.

Slow progress

All datasets do, however, highlight a steady though slow increase in WLR. This could be due to regional contexts, especially the historical, socio-cultural and agrarian relations that define women's role in society and farming, but recent legal and institutional reforms too may have helped--the Hindu Succession Act amendment of 2005 empowers the daughter to assume legal rights in ancestral property by birth; policy changes to recognise joint titling by women farmers; institutional incentives such as stamp duty reduction/waiver for registering property in a woman's name. The cause-effect relationships, however, continue to be academically contested and realigned.

In 2015-16, women's ownership of land, independent or otherwise, was estimated to be 28.3%, according to NFHS-4. Nearly 50% women in Arunachal Pradesh and Bihar owned land alone or jointly, the highest number, while 9% did in Himachal Pradesh, NFHS-4 data show. Rural India appears to be more gender equitable on WLR at 31.4% than urban India (22.9%).

Women's land ownership data are yet to be released in NFHS-5 for 2019-20. Instead, the fact-sheets provide data for women's ownership--independent or otherwise--of a house and/or land. This is bound to be higher than land ownership.

On average, about two in five women (41.6%) own a house and/or land, alone or jointly, according to the recently released data from the first phase of NFHS-5 for 2019-20. Among 17 states and five Union territories for which results have been made available so far, Karnataka leads at 67.6%; the Andaman and Nicobar Islands had the least at 15.8%.

While NFHS-5 has only released limited data so far, in terms of averages, it shows a 2.6-percentage-point increase in women who own a house and/or land (39% to 41.6%) when compared with NFHS-4 data for the 32 states and six UTs for which data are now available. Other datasets show similar increases--for example, the Agriculture Census data showed a 1.1-percentage-point increase between 2010-11 and 2015-16.

How to fix this

To monitor its progress towards SDG targets for 2030, India started reporting SDG indicators in 2018. On SDG indicators relating to women's land rights, India's reporting falls short of the standard set by UNSTAT's metadata, the UN's information repository on data related to SDG indicators.

For example, Indicator 1.4.2 on 'Ending Poverty' is reported differently in NITI Aayog's SDG Dashboard. Instead of sex-disaggregated data on the proportion of adults with secure land rights, it reports 'Percentage of households living in kutcha houses'. For 'Gender Equality' Indicator 5 a 1, it reports 'Percentage of female operated holdings' using the Agriculture Census data. It should instead have reported, by sex, the proportion of total agricultural population that owns agricultural land using other datasets that provide more periodic and reliable gender disaggregated data.

NFHS and other datasets can improve SDG reporting of WLR in India if their methodology is aligned with the SDGs' and their design and collection made gender-sensitive, Shipra Deo of Landesa told IndiaSpend.

The Agriculture Census excludes agriculture labourers, cites the gender of the head of the household only, considers land holding to be different from land ownership, and does not include forest land.

There is a clear case for clearer definitions that must be consistently used across relevant national datasets, for improving WLR reporting, said Rita Sinha, former secretary at the Department of Land Resources, Government of India. Such consistency, she underlined, is imperative for well-informed policy to ensure gender parity in land ownership and also important for appropriate reporting of SDG Indicators.