REVIEW 4 cited by
Optimal Decision Rules Under Partial Identification
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
read the original abstract
I consider a class of statistical decision problems in which the policymaker must decide between two policies to maximize social welfare (e.g., the population mean of an outcome) based on a finite sample. The framework introduced in this paper allows for various types of restrictions on the structural parameter (e.g., the smoothness of a conditional mean potential outcome function) and accommodates settings with partial identification of social welfare. As the main theoretical result, I derive a finite-sample optimal decision rule under the minimax regret criterion. This rule has a simple form, yet achieves optimality among all decision rules; no ad hoc restrictions are imposed on the class of decision rules. I apply my results to the problem of whether to change an eligibility cutoff in a regression discontinuity setup, and illustrate them in an empirical application to a school construction program in Burkina Faso.
Forward citations
Cited by 4 Pith papers
-
Dynamically Consistent Statistical Decisions
Frequentist minimax rules often lack interim credibility; two axiomatized dynamically consistent criteria restore it while nesting Manski as-if and Gamma*-minimax.
-
Optimal treatment assignment rules under capacity constraints
Under capacity constraints, the optimal treatment assignment problem can be rewritten as an optimal transport problem, where a Bayesian rule is locally asymptotically optimal and the plug-in rule is not when welfare i...
-
When Compression Scores Cannot Decide: Information Boundaries for Group-Robust LLM Pruning
A compression statistic's observation leaves a fiber of compatible endpoint-risk tables, so only orders fixed across that fiber are identified, and stable scores can still select worse group-robust models.
-
Who With Whom? Learning Optimal Matching Policies
An entropy-regularized optimal transport method learns welfare-optimal two-sided matching policies with estimated costs, supported by a non-asymptotic regret bound and calibrated simulations suggesting about one perce...
Discussion (0). Continue with ORCID to comment.