Replication package for: Made in Heaven, Matched by Parents — Can Arranged Marriage Restrict Labour Market Autonomy and Participation of Women?

Published: 8 October 2026| Version 1 | DOI: 10.17632/w8ygss8n3z.1
Contributors:
,

Description

This dataset supports the article "Made in Heaven, Matched by Parents: Can Arranged Marriage Restrict Labour Market Autonomy and Participation of Women?" by Sugata Bag and Anirban Kar (World Development). Research question. The study tests whether the type of marriage arrangement — parent-arranged, jointly-arranged, or self-matched — causally affects two distinct dimensions of a woman's labour market engagement in India: her autonomy over work decisions (whether she decides whether to work) and her actual labour force participation. The central hypothesis is that parent-arranged marriage, by distorting matching in favour of more patriarchal husbands, reduces post-marriage labour market autonomy, and that autonomy and participation are separate outcomes governed by different constraints. Data and construction. The data are derived from the India Human Development Survey-II (IHDS-II), 2011–12, a nationally representative survey of Indian households. The deposited analytical dataset is a subset of IHDS-II restricted to the variables used in the analysis, together with constructed outcomes (work autonomy and two labour force participation measures), controls, and the instrumental variables. The instruments are built from the lagged prevalence of each marriage type within state–caste–birth-cohort cells, capturing the persistence of community marriage norms across generations. The raw IHDS-II microdata are publicly available from ICPSR (36151) and are not redistributed here. What the data show. Using an instrumental-variables strategy estimated via a Conditional Mixed Process framework, the analysis finds that self-matched marriages raise women's work autonomy by approximately 30 percentage points and participation by 14–16 percentage points relative to parent-arranged marriages, with jointly-arranged marriages falling in between. Autonomy gains consistently exceed participation gains — a systematic "wedge" — which the heterogeneity patterns attribute to external structural barriers (limited job availability, care burdens, commuting and safety constraints) rather than to differences in preferences. Use. The accompanying Stata do-files allow users to reproduce all reported estimates, tables, and figures, and to examine the sensitivity of the results to alternative instrument constructions and specifications.

Files

Steps to reproduce

1. Download and unzip the replication package. It contains three folders: do_files/ (Stata code), data/ (the analytical dataset autonomy_analysis.dta), and output/ (empty; results are written here), plus a README. 2. Obtain the raw IHDS-II data if you wish to rebuild variables from scratch: it is publicly available from ICPSR (https://doi.org/10.3886/ICPSR36151.v6). This is optional — the provided analytical dataset already contains all constructed variables and instruments. 3. Install Stata 17 or later and the required user-written packages: cmp, ivreg2, ranktest, and estout (run: ssc install cmp, ivreg2, ranktest, estout). 4. Open do_files/00_master.do and set the global macro "root" to the folder where the package is located (one line only). 5. Run 00_master.do. It executes all steps in sequence: variable construction (01), instrument construction (02), and the analysis files producing Table 1 (03), Tables 2–3 (04), and the 2SLS, placebo, and instrument-validity tests (05). All tables and figures are written to the output/ folder. Total runtime is approximately 30–60 minutes, with the CMP models being the most time-intensive. The README documents the mapping of each do-file to the corresponding table and figure in the paper.

Institutions

Categories

Economics of Gender, Development of Economics

Licence