How Is Econometrics Changing? (Josh Angrist, Guido Imbens, Isaiah Andrews)

TL;DR

Econometrics is evolving with new methods, tech influences, and debates on causality.

Transcript

♪ [music] ♪ - [Narrator] Welcome to Nobel Conversations. In this episode, Josh Angrist and Guido Imbens sit down with Isaiah Andrews to discuss how the field of econometrics is evolving. - [Isaiah] So, Guido and Josh, you're both pioneers in developing tools for empirical research in economics. And so I'd like to explore where you feel like the... Read More

Key Insights

• The field of econometrics is shifting towards more credible identification strategies and causal questions, moving away from traditional models.
• Regression discontinuity designs and instrumental variables are increasingly complementary, offering robust methods for causal inference.
• There is a growing trend of using randomized trials creatively, especially in educational settings, to understand causal effects.
• Empirical problems should drive econometric research, ensuring that methods are applicable and relevant to real-world issues.
• The use of local average treatment effects (LATE) has expanded, sometimes beyond formal results, raising concerns about its intuition being overextended.
• There is a call for shorter, more focused econometric papers to enhance clarity and understanding.
• The tech sector's demand for econometrics skills is rising, influencing the development of empirical methods and potentially altering graduate education funding models.
• The debate between reduced form and structural approaches continues, with concerns about misinterpretations and the need for clear communication.

Questions & Answers

Q: How is the field of econometrics evolving?

Econometrics is evolving towards more credible identification strategies and a focus on causal questions rather than traditional models. This shift is characterized by the increased use of regression discontinuity designs and instrumental variables, which offer robust methods for causal inference. The field is also embracing creative uses of randomized trials to better understand causal effects, particularly in educational settings.

Q: What role do empirical problems play in econometric research?

Empirical problems are crucial in driving econometric research, ensuring that the methods developed are applicable and relevant to real-world issues. The speakers emphasize the importance of having a clear connection between empirical practice and methodological development to produce useful and impactful research. Empirical problems push researchers to refine methods and address practical challenges.

Q: What concerns exist about the use of local average treatment effects (LATE)?

There are concerns that the intuition behind local average treatment effects (LATE) is sometimes overextended beyond formal results, leading to potential misinterpretations. While LATE provides valuable insights into causal effects, its application in settings without formal results can be problematic. Researchers caution against relying solely on intuition and emphasize the need for rigorous application to ensure accurate conclusions.

Q: Why is there a call for shorter econometric papers?

There is a call for shorter econometric papers to enhance clarity and understanding. Longer papers can make it difficult for readers to grasp the core insights and contributions, potentially hindering the dissemination of knowledge. By focusing on concise, clear results, researchers can improve communication and ensure that their work is accessible and impactful.

Q: How is the tech sector influencing the field of economics?

The tech sector's growing demand for econometric skills is influencing the development of empirical methods and potentially altering graduate education funding models. Tech companies offer exciting opportunities and data for research, but this trend raises questions about the role of public funding in graduate education. As more students enter the tech sector, there may be implications for how economics programs are structured and funded.

Q: What is the debate between reduced form and structural approaches?

The debate between reduced form and structural approaches involves differing perspectives on the assumptions and methods used in econometric analysis. Reduced form approaches focus on empirical relationships without specifying a detailed model, while structural approaches rely on economic theory to inform model assumptions. The speakers express concerns about misinterpretations and emphasize the need for clear communication to avoid confusion.

Q: What impact does the demand for econometric skills have on graduate education?

The rising demand for econometric skills in the tech sector may impact graduate education by challenging traditional funding models. As more students pursue careers in tech, questions arise about the justification for public funding of graduate education, which has traditionally focused on training educators and scholars. This trend could lead to changes in how programs are structured and funded.

Q: How are randomized trials being used creatively in econometrics?

Randomized trials are being used creatively in econometrics to explore causal effects in settings where traditional experiments are not feasible. For example, researchers are using financial aid to encourage schooling as a way to study educational outcomes. This approach allows for the exploration of causal questions in innovative ways, expanding the scope of empirical research and providing valuable insights into complex issues.

Summary & Key Takeaways

• Econometrics is evolving with a focus on causal inference and credible identification strategies, moving beyond traditional models. Regression discontinuity and instrumental variables are key methods in this shift. The field is also embracing creative uses of randomized trials, particularly in educational research.

• Empirical problems should guide econometric research, ensuring methods are relevant and applicable. The local average treatment effects framework has been widely adopted, but its use sometimes extends beyond formal results, raising concerns about over-reliance on intuition.

• The tech sector's growing demand for econometric skills is influencing empirical method development and could impact graduate education funding. The debate between reduced form and structural approaches persists, with calls for shorter, clearer papers to improve understanding and communication.