We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

econ.EM

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Economics > Econometrics

Title: Algorithm is Experiment: Machine Learning, Market Design, and Policy Eligibility Rules

Abstract: Algorithms produce a growing portion of decisions and recommendations both in policy and business. Such algorithmic decisions are natural experiments (conditionally quasi-randomly assigned instruments) since the algorithms make decisions based only on observable input variables. We use this observation to develop a treatment-effect estimator for a class of stochastic and deterministic decision-making algorithms. Our estimator is shown to be consistent and asymptotically normal for well-defined causal effects. A key special case of our estimator is a multidimensional regression discontinuity design. We apply our estimator to evaluate the effect of the Coronavirus Aid, Relief, and Economic Security (CARES) Act, where hundreds of billions of dollars worth of relief funding is allocated to hospitals via an algorithmic rule. Our estimates suggest that the relief funding has little effect on COVID-19-related hospital activity levels. Naive OLS and IV estimates exhibit substantial selection bias.
Subjects: Econometrics (econ.EM); Machine Learning (cs.LG); Methodology (stat.ME); Machine Learning (stat.ML)
Cite as: arXiv:2104.12909 [econ.EM]
  (or arXiv:2104.12909v5 [econ.EM] for this version)

Submission history

From: Yusuke Narita [view email]
[v1] Mon, 26 Apr 2021 23:18:34 GMT (35652kb,D)
[v2] Mon, 28 Jun 2021 12:37:09 GMT (37462kb,D)
[v3] Mon, 4 Oct 2021 04:06:04 GMT (2824kb,D)
[v4] Sun, 14 Nov 2021 07:36:00 GMT (2827kb,D)
[v5] Mon, 6 Dec 2021 08:44:42 GMT (2706kb,D)

Link back to: arXiv, form interface, contact.