G²LM|LIC - Long–Run Social Norms and Female Labour Force Participation in India

1. Overview
The project studies the long-run relationship between social norms, structural transformation, urbanization, and female labour force participation (FLFP) in India. It is motivated by India's low female labour force participation and by the possibility that historically persistent patriarchal norms mediate women's responses to economic development.
The research links historical information on social norms to modern census and socioeconomic records. The project description envisages an integrated gender-data resource covering approximately 5,000 social groups over a roughly 100-year period, with historical ethnographic sources linked to contemporary gender-disaggregated outcomes.

2. Data Sources and Temporal Coverage
According to the project description, the broader research design combines historical and modern sources. These include ethnographic data from the early 1900s and from 1988–1990, the 1931 Census of India, and the 2012 Socio-Economic and Caste Census (SECC). The intended integrated resource therefore spans approximately a century.
The currently supplied Culture data packet documents group-level social-norm data and supporting crosswalks, descriptions, classifications, and mappings. The supplied README does not document the detailed provenance, construction procedure, observation counts, or variable definitions inside each file; those points should not be inferred without the corresponding codebooks or data.

3. Units of Analysis and Geographic Coverage
The project focuses on social groups at the jati/community level and on local labour markets at the village and town level. Depending on the source and analytical file, observations may therefore represent social groups, localities, social-group-by-locality combinations, or social-group-by-time observations.
Geographic coverage is India. The available metadata also indicate mappings from specific communities to broader, state-specific social groups.

4. Dataset Structure
The available culture-data metadata indicate that the core norm files are group-level datasets that record whether individual social norms are present or absent for each group. Supporting files provide community-code/name correspondence, norm descriptions and classifications, alternate or cleaned community names, and mappings to broader state-specific social groups.

The metadata supplied do not establish which of the two group-level norm files should be treated as the canonical analytical file, how consolidation differs from cleaning, or which keys should be used for every merge. Those details should be added from the processing scripts or a technical codebook before publication.

5. Methodological Design
The first component of the project studies whether the relationship between economic transformation and FLFP differs across social groups and whether historical patriarchal norms are associated with different income and substitution responses of female work. The analysis is intended to distinguish within-group patterns from cross-group patterns.
The second component studies long-run changes in gender norms in relation to urbanization. The project description states that social-group time-series data from the early twentieth century through the late 1980s will be used and that instruments for urbanization are intended to support causal identification. The exact instruments and estimating equations are not specified in the supplied metadata.

6. Relationship to the Female Urban Mobility Study
A separate supplied replication README concerns The Female Urban Mobility Advantage in South Asia by Sam Asher, Aditi Bhowmick, Maurizio Bussolo, Paul Novosad, Jonah Rexer, and Sara Sethia. Its documented workflow is to edit config_edu_mobility.do to set local paths and then run make_edu_mob_results.do; it is intended to be used with the associated replication-edu-mobility-data.zip data packet. This replication package should be documented separately from the Culture data packet unless the repository deposit explicitly establishes that it is part of the same publication package. The supplied metadata alone do not document a file-level dependency between the culture files and that replication code.