Reading ability conflates SES creativity gaps.

Who is more creative: high- or low-socioeconomic status (SES) individuals? This question is the focus of intense debates within both the scientific community and society at large, since creativity has been linked to innovation, productivity, and wealth generation. This question is, however, hard to answer; in particular, because creativity is typically assessed through standardized tests that build on reading and writing proficiency, which might conflate the relationship between SES and creativity. To overcome this challenge, we combine high-quality data on reading ability and experimental variation in reading requirements embedded in creativity assessments in a series of studies with 6-12th graders in Brazilian schools. We first document that established creativity measures exhibit sizable SES gaps, but that these gaps become much smaller and no longer significant once we parse out students' reading ability (Study 1). Next, in two randomized control trials, we have students complete divergent thinking tasks while experimentally varying reading requirements: in the control group, students had to read question prompts, just like in standard assessments (e.g., PISA); in the treatment group, enumerators read the prompts on their behalf. In both conditions, enumerators wrote down students' answers. We find that high-SES students outperform low-SES ones only when they read the prompt, but not when performance does not depend on reading ability (Studies 2 and 3).