The programme we were evaluating worked with women's collectives across Bihar and Odisha — self-help groups federated into producer collectives, supported by a combination of livelihood inputs and social empowerment activities. The existing monitoring framework, which we inherited at the start of the engagement, was organised around outputs: number of groups formed, number of members trained, number of loans disbursed, value of goods sold through collective marketing channels.
These are defensible indicators. They are also insufficient for the programme's actual theory of change, which was not primarily about income but about agency.
The mismatch became clear in the first round of data collection. Programme data showed strong performance on outputs. Community-level qualitative work showed something more textured: women who had been in the collectives for three or more years described changes in their lives that the monitoring framework did not capture — and, more importantly, did not weight correctly relative to the income gains the programme prioritised in its reporting.
We ran a series of co-design sessions with women in both states to understand what outcomes they themselves used to assess whether the programme had made a difference in their lives. The methodology drew on Most Significant Change — asking women to narrate specific instances of change they had experienced, then working through what made those instances significant. We did not start with a proposed indicator list. We started with stories.
Three outcome areas emerged consistently that the original framework had missed or underweighted.
The first was voice in public-facing interactions. Women across sites described a change in how they handled encounters with government officials: at the panchayat office, at the ration shop, at the health centre. The change was not about having more information, though that was part of it. It was about not being dismissed, and about knowing what to do when they were. Several women described specific incidents — a ration entitlement restored, a delayed MGNREGA payment followed up — as markers of something the income figures could not represent.
The second was within-household negotiation. Women in longer-standing collectives described changes in how household financial decisions were made. Husbands consulting them about expenditures above a certain threshold. Being told, rather than not told, when a significant purchase was planned. These are not outcomes that a loan-disbursement indicator captures, and they are also not straightforwardly measurable by an external enumerator asking a standard question about financial decision-making.
The third was their daughters. In multiple groups across both states, the change women named as most significant was something they observed in their daughters: confidence in school, willingness to argue, a sense that their own futures were negotiable. Women in their thirties and forties were describing an intergenerational change they attributed, at least partly, to what their daughters had seen them do in the collective.
The co-designed framework incorporated all three areas, with indicator sets built from the women's own language rather than external frameworks. This required upfront time — roughly four weeks of iteration — and it required managing the tension between indicators that are meaningful to participants and indicators that are legible to funders. That tension is real; it does not resolve neatly.
The methodological lesson is not simply that participatory MEL design produces better indicators, though it often does. It is that the process of co-designing the measurement framework is itself a form of programme data. How women describe change — which changes they weight, which they dismiss as trivial, which they attribute to the programme versus other factors — tells you something about the programme's theory of change that no external evaluation framework can produce.
One important caveat: co-design at the collective level risks amplifying the voices of more confident, better-connected women within those collectives. We built in separate discussion sessions with newer members and with women who had dropped out of groups, which produced somewhat different indicator priorities and served as a useful check.
Bihar and Odisha NFHS-5 district data on women's autonomy indicators are available through the DHS Programme. NRLM's data portal carries SHG membership and loan disbursement data by state.