{"id":"arxiv/2608.13822","url":"https://arxiv.org/abs/2608.13822","tags":["math.ST","econ.EM","stat.ME"],"score":0,"title":"Sharp Minimax Theory for Randomized Experiments","domain":"math.ST","source":"arxiv","authors":["Timothy Sudijono","Edgar Dobriban","Eric Tchetgen Tchetgen"],"reviews":0,"subject":"math.ST","summary":"We study minimax-optimal designs and estimators for estimating the sample average treatment effect in finite population randomized experiments, where both design and estimator are unrestricted. For binary potential outcomes, we show this minimax risk is equivalent to the minimax risk $ρ_n^*$ of an estimation problem with $2$ unknown parameters. We leverage this reduction to establish a second-order risk expansion $ρ_n^* = n^{-1} - Cn^{-4/3} + o_n(n^{-4/3})$ for an explicit constant $C$ related to the Airy function. The minimax risk is attained by Bernoulli randomization with a nonlinear…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-08-13","reactions":{}}