Background In India, the poor are increasingly forced to live in highly populated urban dwellings with inadequate living conditions including deficient service provision. This in turn affects their food security and negatively influences their health. In December 2012, Dilli Annashree Yojna (DAS), an unconditional cash transfer scheme was launched for the urban poor in Delhi to address food security and, ultimately, nutrition and health. Our study attempts to explore DAS for new elements like unconditionality and transfer to female head of family, use of benefits, operational issues and perceptions on its influence on health.

Methods We conducted a qualitative assessment, based on document analysis followed by focus group discussions (FGDs) and in-depth interviews (IDIs) with DAS beneficiaries. Data were collected in June 2014, three months after the scheme had been withdrawn. Potential respondents were informed about date and venue of FGDs and the first 15 participants who arrived at the venue were recruited. The rest were offered a chance to participate in the IDIs. Four FGDs and 7 IDIs were conducted. Data from FGDs and IDIs were analysed after sorting out the responses according to study-related objectives. Responses were analysed for similarities and contradictions. Relevant quotes were used to justify and strengthen the findings.

Findings The DAS policy documents clearly outline the unconditional character of cash transfer for food subsidy and the transfer of cash to female heads of family as novelties of the scheme, and specify criteria for inclusion in the scheme. In practice, dissemination of DAS implementation details and identification of beneficiaries depended on the proximity of active Gender Resource Centres. The scheme's cash transfers were generally erratic. Only two families reported a regular monthly deposit of INR 600. Others reported transfers between INR 600 and 18,000, in irregular deposits. Access to the transferred money was easier for families who received updates on deposits, i.e. with bank accounts tagged to their mobile phones, and more difficult for older heads of family not adept at this technology, who then had to rely on younger family members.

All respondents felt that the inbuilt unconditionality allowed them to appropriately prioritise their needs. Female respondents particularly appreciated transfer to female heads of families. One male participant opposed this arrangement. Respondents also expressed a preference for the Targeted Public Distribution System that was seen as offering more regular supplies.

According to the respondents, the received transfers were first and foremost spent on children's education and related expenses. This was followed by purchase of food supplies and immediate health needs. Not having to borrow money for purchasing food, as long as DAS was functioning, was mentioned. One lady attributed her good sleep to the scheme. Few families were able to meet emergency health needs. Still fewer families, those who had received larger transfers, were able to purchase assets.

Discussion Communication gaps and lack of community involvement in the enrolment process let to localized implementation of DAS and loss of uniformity. Where DAS was implemented, the unconditionality enabled beneficiary families to prioritize their needs. Spending was made on choices specific to household level context, within the constraints defined by the actual amounts transferred. Transfer to the female head of the family enabled gender empowerment; female respondents reported additional decision-making space in household spending. In families not under immediate risk of food insecurity, no change in food mix was observed. In families with this risk, however, the transfers ensured not only basic staples for an assured period but also an improvement in food mix. Spending made to acquire health services could directly translate to health benefits. Priority spending on education, as reported, could possibly engender long-term benefits, e.g. prevention of passing on poverty to the next generation.

Recommendations Identification of beneficiaries and communication with them need substantial improvement. Particularly, information on what is expected from the beneficiaries and what they can expect from the scheme is lacking. Regular monthly transfers of INR 600 are more likely to be used for food then larger cumulated sums, and should thus be preferred. The positive effect on gender empowerment justifies maintaining the transfer arrangement with female heads of family as recipient.

No competing interest.