While most evaluations of education programs in developing countries examine effects one or two years after a program has been introduced, this study does so over an extended duration of a program. Administered in Punjab, Pakistan, the program offers cash benefits to households conditional on girls' regular attendance in secondary grades in government schools. The study evaluates the evolution of the program's effects on girls' secondary school enrollment numbers over roughly a decade of its existence. The program was targeted to districts with low adult literacy rates, a targeting mechanism that provides an observed, numerical program assignment variable and results in a cutoff value. Recent advances in regression discontinuity designs allow the study to appropriately fit key features of the data. The study finds that the program had positive effects on girls’ secondary school enrollment numbers throughout the period and that these effects were stable. This pattern is observed despite a loss of more than 60 percent in the real value of the cash benefit over the period. The findings are consistent with potential behavioral explanations, such as the program making girls' education salient to households or catalyzing a shift in social norms around girls' education.