A1162
Title: Regression discontinuity designs under interference
Authors: Laura Forastiere - Yale University (United States) [presenting]
Abstract: Interference takes place whenever a treatment on one unit affects the outcome of another unit, and such a phenomenon can also occur in regression discontinuity designs (RDDs). For instance, in conditional cash transfer programs for education, the eligibility to the program and the potential receipt of cash transfers may affect eligible children's schooling choices, which in turn may influence schooling choices of their peers. In this setting, assignment to the individual treatment and to the spillover exposure, which incorporates through a mapping function the exposure to the treatment of interfering units (e.g., friends, classmates), is determined by a unit's score and the scores of other interfering units. Unlike the standard RDD, the presence of spillover exposure to other units may give rise to complex, multidimensional boundaries on a multidimensional score space. A method is provided to characterize such boundaries for a broad class of exposure mapping functions and derive generalized continuity assumptions to identify the proposed causal estimands. Next, an estimation method that can handle multidimensional and potentially heterogeneous multi-scores, including complex dependencies, is developed. Finally, the proposed methodology is applied to the PROGRESA/Oportunidades data to estimate the direct and indirect effects of receiving cash transfers on children's schooling attendance.
