India’s Janani Suraksha Yojana, a Conditional Cash Transfer Programme to Increase Births in Health Facilities: An Impact Evaluation

Publication Details

The Lancet, June 2010, vol.375, iss.9730, pp.2009-2023. Available From:

Link to Source
Stephen S. Lim, Lalit Dandona, Joseph A. Hoisington, Spencer L. James, Margaret C. Hogan, Emmanuela Gakidou
South Asia
Health Nutrition and Population
Health Services, Hospitals- Secondary & Tertiary, Mortality, Preventive Health and Health Behavior, Primary Health- including reproductive health
Equity Focus
Evaluation design
Difference-in Difference (DID), Others
Journal Article


This article evaluates the impact of a conditional cash transfer programme aimed at reducing maternal, perinatal and neonatal mortality in India. The analysis focuses on Janani Suraksha Yojana (JSY), a large-scale conditional cash transfer programme launched by the Indian government in 2005. The programme provides a financial incentive to pregnant women delivering in a government or accredited private health facility. In a majority of states, only women from low-income households are eligible for the programme, but ten states (“high-focus states”) offer cash benefits to all women, regardless of their socioeconomic status. Women delivering in a participating health facility receive 600 rupees (1,000 rupees in high-focus states) in urban areas and 700 rupees (1,400 rupees in high-focus states) in rural areas.
The authors use three different approaches to assess the programme’s effectiveness. First, they conduct an exact-matching analysis, matching JSY births to non-JSY births on the basis of state of residence, urban or rural location, poverty, wealth quintile, caste, education, parity and maternal age. The regression specification controls for covariates such as the number of live births, birth interval, maternal education, caste or tribe, and religion. Second, they carry out a with-versus-without analysis, comparing pregnant women who received JSY to pregnant women who did not. The specification includes the same control variables as for the exact-matching analysis. Third, they use a difference-in-differences (DID) approach, using a model that controls for district-level variables such as maternal age, maternal education, rural or urban location, and distance to the nearest health facility. The data originate from the India District-Level Households Survey (DLHS), which gathered information on maternal and child health, family planning, reproductive health and use of maternal and child health-care services. Over 620,000 observations were collected between 2002 and 2004, and about 720,000 observations were collected between 2007 and 2009.

Main findings

The study finds significant variation in programme uptake across districts. Uptake rates were lower than 5 percent in 141 districts, ranged from 5 to 30 percent in 342 districts, and were higher than 30 percent in 128 districts. National statistics show that uptake rates were higher among women having 1–11 years of education (approximately 12 percent) than among women with no education (8 percent) or among women with more than 11 years of education (9 percent).
The three methodological approaches show significant improvements in the use of antenatal care and the number of in-facility births. The exact-matching analysis indicates that the programme increased the use of antenatal care by 10.7 percent and in-facility deliveries by 43.5 percent. The with-versus-without analysis presents similar results, with an 11.1-percent increase in the use of antenatal care and a 43.9-percent increase in the number of in-facility births. Finally, DID estimates show a 10.9-percent increase in the use of antenatal care and a 49.2-percent increase in the number of in-facility deliveries.
The programme’s impact on health outcomes varies depending on the methodology used. Matching estimates show that programme participation reduced perinatal mortality by 3.7 deaths per 1000 pregnancies and neonatal mortality by 2.3 deaths per 1000 live births. Using the with-versus-without comparison, the authors find that the programme led to a reduction of 4.1 perinatal deaths per 1000 pregnancies and to a reduction of 2.4 neonatal deaths per 1000 live births. However, none of the three methodological approaches showed any impact on maternal mortality, and no statistically significant effects were found on any of the three health outcomes using the DID approach.

Scroll to Top