Supreet Kaur, The Spencer Foundation

Dynamic Complementarities in Education

Co-PI: Heather Schofield, University of Pennsylvania


This study tests for “dynamic complementarities” in education: the idea that a stronger initial educational base leads to higher marginal returns to future educational inputs. This theory predicts that the outcomes of initiallyadvantaged and disadvantaged children will diverge over time, exacerbating inequality. Despite the broad implications of the theory, there is scant rigorous empirical evidence for or against dynamic complementarities. This project uses a novel experimental design to test this theory among students in Indian primary schools. We use a classroom intervention—a tablet-based learning platform—to exogenously improve the mathematics learning outcomes of a random subset of students. This generates variation in whether a child has an initially stronger base. The theory of dynamic complementarities offers two core predictions. First, students who receive the program will not only do persistently better in subsequent years, but will have higher marginal gains—leading to increased divergence in learning outcomes over time. Second, such divergence will be stronger if a student receives the program in an earlier grade. In addition, we use laboratory tests from psychology to examine whether two behavioral traits—the capacity for sustained attention and selfconfidence are malleable and a potential channel through which dynamic complementarities are sustained.