51社区黑料

Michael Gilraine

It Takes Time: The Asymmetric Impacts of Imposing and Removing a Teacher Incentive Scheme

Teachers play a vital role in the education system, but there is considerable debate about how to enhance teaching quality. One natural approach grounded in economic principles is pay-for-performance -- offering financial rewards to teachers based on the improvement in their students' test scores. Many school districts have implemented performance incentives, but evaluations of such programs have yielded mixed conclusions -- some programs help, but others do not. One possible reason for the conflicting results is that only certain teachers are motivated by extra pay.

This project explores the possibility that only some teachers respond to monetary incentives by investigating both the introduction and termination of a pay-for-performance program called Mission Possible that ran in North Carolina from 2006 to 2017. If teachers who choose to work in schools with incentive programs are exactly those teachers who respond to monetary incentives, then we expect there to be a dramatic asymmetry between the program鈥檚 introduction and termination. Specifically, we would expect little effect when the program is introduced as the majority of teachers are not responsive to incentives. Over time, however, incentive-responsive teachers sort into schools offering the incentives, causing a large drop in test scores upon the program's termination. This research will thus uncover the impact of pay-for-performance on student test scores and shed light on whether improvements are driven by existing teachers responding to incentives or by the gradual movement of incentive-responsive teachers into participating schools.

Chenyu Hou

Measuring Narratives about Macroeconomy using Large Language Models

When we make economic decisions, we rely on more than just statistics; we use stories, or "narratives," to make sense of what鈥檚 happening. For example, the idea that supply-chain disruptions caused recent inflation is a powerful narrative that shaped public opinion and expectations. This project seeks to understand where these economic stories come from, how they spread, and how they influence the beliefs of both consumers and professionals.

To achieve this, our project uses advanced Large Language Models (LLMs) to "read between the lines" of a vast collection of texts, including Federal Reserve reports (Beigebooks) and major news articles from since 1970. Unlike typical text analysis that just identifies topics, our method is designed to extract the underlying logic鈥攖he causal links that form a narrative (e.g., "an increase in input costs leads to an increase in prices"). By comparing the narratives from policymakers to those presented in the media, we can track how economic stories are told and how they might change in transmission.

The findings from this project will be useful for a wide range of audiences. For academics and researchers, it provides a novel methodology for studying economic narratives. More importantly, it helps policymakers understand how the public forms its economic beliefs. By knowing which stories are resonating, they can communicate policy more effectively, anticipate public reactions, and avoid unintended consequences, ultimately fostering a better-informed public discourse on the status of the economy.