The DIU / Transform Rural India study asked what looked like a simple question: what do women in rural and peri-urban India aspire to economically? We were commissioned to answer it through primary research at sites in different agro-ecological zones, with different caste compositions, and at different distances from labour markets.
The first round of fieldwork showed us that the question, as framed, was the problem.
Research instruments that ask about aspirations directly ("what kind of work would you like to do?", "what are your economic goals for the next five years?") draw a kind of answer that women in programme-heavy areas know well. Decades of intake forms for government schemes, NGO enrolment and baseline surveys have taught respondents which answers sound plausible and get them enrolled. The answers come back quickly and coherently, and they are much the same across sites: tailoring, teaching, leading a self-help group. Many women do want these things. The sameness comes from the framing of the question, which steers women toward one vocabulary and away from others.
Our first change was from preference questions to scenario questions. Instead of asking what a woman wanted to do, we asked what she would do if a specific constraint were removed: if her eldest daughter could walk to school, if the family had an extra ₹3,000 a month, if her husband worked different hours. Social desirability still affects scenario questions, though less, because the answer to "what would you do if X" is harder to pre-script than the answer to "what do you want".
The second change made mobility mapping a separate, drawn instrument. We asked women to draw or trace the limits of their regular movement: where they went each week, where they went seasonally, where they had been once or twice, and where they had never been but knew about. The maps gave us a spatial record of constrained aspiration. Women who named distant cities as places they might want to work had often never travelled more than 12 kilometres from home, and the distance between the aspiration and the places they had actually been was itself data.
The third change was to use life-history prompts with the older women in the sample. Women in their forties and fifties who had changed livelihoods, from agricultural labour to vegetable trading, or from home-based piece-rate work to a factory floor and back, knew from experience which transitions were possible, what they cost, and what happened at home once a woman's income became visible. Their accounts were more useful to the study than any number of preference rankings.
Across sites we found a pattern we called constrained aspiration. Women's aspirations were often very specific. A woman at one peri-urban site could name the exact salary that would make a factory commute worthwhile and the exact bus timing that made it impossible. The constraints were infrastructural and relational: husbands uncomfortable with visible earnings, mothers-in-law counting the labour the household lost when a woman left every day, and no childcare option anyone trusted. Women could describe in considerable detail the specific aspirations these constraints ruled out.
This matters for method because livelihoods programmes designed around stated preferences will consistently overestimate revealed demand for what they offer. Women asked whether they want income-generating opportunities nearly always say yes. That answer tells a programme almost nothing about uptake, about persistence, or about whether it has identified the binding constraint.
IHDS-II microdata and PLFS data on female labour force participation at the district level gave us reference points for describing the sample sites. Oxfam India's 2023 inequality report provided benchmarks on gender and economic participation.