Draft

Abstract. I study the effects of ratings designed to capture the financial risk associated with apartment ownership in Sweden. Using a regression discontinuity design, I find that ratings move sales prices and real estate agents' pricing decisions substantially, but only after ratings started being displayed in online listings -- a change in exposition that greatly increased their salience. Mechanisms are explored, and results suggest that the effect of ratings on prices is driven by improved information transmission. A back-of-the-envelope calculation shows that the ratings remedy a substantial part of the consumer inattention to financial information that previous studies find.

Abstract. Large-scale standardized assessments have been proposed as one of the most promising policy responses to remedy the learning crisis in developing countries. This study provides the first evidence on the effectiveness of a large-scale and long-running program of this kind, in the context of India. Since 2011, over 7 million primary school students in 110,000 public schools have been assessed annually and received letter-grades to reflect performance. We focus on two main questions: the reliability of the assessments and the accountability pressures exerted by the letter-grading system. Using direct audits, the study finds that reported achievement levels in the official tests are significantly overstated, indicating active manipulation rather than measurement discrepancies. The ordinal ranks of schools, however, remain stable over time despite the inflation in test scores. Further, we employ a regression discontinuity approach to investigate the impact of the letter-grade system on school functioning and student performance. The results suggest that the grades achieve very little, with no significant changes observed in school management practices, inputs, or monitoring. Likewise, there is no evidence of improvements in student outcomes based on the assigned grades. These findings raise concerns about the view that standardized assessment programs can drive system-wide improvements in education in settings where state capacity is weak. In particular, they emphasize the need for complementary measures to ensure data integrity in order for such reforms to deliver results at scale.

Work in progress.