Week 5: Prediction and Causation

Instructions

This week we discuss: What is the role of machine learning and big data in a world of evidence-informed public policy? What are “prediction policy problems” and (how) do they differ from other policy problems? Are RCTs becoming obsolete?

Required readings

  • Kleinberg, J., J. Ludwig, S. Mullainathan, et al. “Prediction Policy Problems”. In: American Economic Review 105.5 (2015), pp. 491-95. DOI: 10.1257/aer.p20151023. Focus in lecture.
  • Bansak, K., J. Ferwerda, J. Hainmueller, et al. “Improving Refugee Integration Through Data-Driven Algorithmic Assignment”. In: Science 359.6373 (2018), pp. 325-329. DOI: 10.1126/science.aao4408. Focus in seminar.
  • Glaeser, E. L., S. D. Kominers, M. Luca, et al. “Big Data and Big Cities: The Promises and Limitations of Improved Measures of Urban Life”. In: Economic Inquiry 56.1 (2018), pp. 114-137. DOI: 10.1111/ecin.12364. Background reading.

Further reading

  • The Behavioural Insights Team. 2017. Using Data Science in Policy.

  • Kleinberg, J., H. Lakkaraju, J. Leskovec, et al. “Human Decisions and Machine Predictions”. In: The Quarterly Journal of Economics 133.1 (2018), pp. 237-293. DOI: 10.1093/qje/qjx032.

  • Athey, S. “Beyond Prediction: Using Big Data for Policy Problems”. In: Science 355.6324 (2017), pp. 483-485. DOI: 10.1126/science.aal4321.

  • Acemoglu, D. Harms of AI. Working Paper 29247. National Bureau of Economic Research, 2021. DOI: 10.3386/w29247.

Acknowledgments

Photo credit: https://unsplash.com/photos/ZiQkhI7417A.