Notes: DLHS samples. Outcome variables are isgirl (Panel A) and isbirth (Panel B). DID estimates compare the change in each outcome between 2004–2007 and 2009–2012 in Fatehgarh Sahib with the corresponding changes in the control districts. Controls: child’s birth order, mother’s age and education, household’s caste, religion, location (rural or urban), and an indicator of dwelling construction type (wealth proxy). Alternative control groups comprise states that were surveyed in 2012/13 and, respectively, had no girl-child CCT operating during the period (“None”), had a girl-child CCT but the status of the CCT did not change (“No Changes”), and are culturally similar to Punjab (“Northwest”). The Punjab sample includes 6,701 mothers; the None sample, 42,114; the No Changes sample, 31,320; and the Northwest sample, 16,146. p-values in parentheses. Census sample. The outcome variables are share of girls in the zero-to-six–year-old population (Panel A) and the size of the zero-to-six population (Panel B). The DID estimates compare the change in each outcome between 2001 and 2011 in Sirhind with the corresponding changes in the other blocks in Punjab. The controls are male and female literacy rates, labor force participation by gender, rural population share, and scheduled caste or tribe share, and block fixed effects. The treated area is the Sirhind block and the treatment period is 2011. p-values in parentheses.