Pith. sign in

REVIEW 3 major objections 4 minor 46 references

Robust Instrumental Variables: Sharp Rates and Inference under Adversarial Contamination

T0 review · 3 major / 4 minor · reviewed 2026-08-03 · deepseek-v4-flash

Pith's one-line read A winsorized-mean modification of 2SLS is minimax-optimal under adversarial contamination, and keeps clean-data Gaussian inference whenever sqrt(n) eta^(1-1/m) vanishes.

desk verdict Promising robustification of 2SLS with sharp worst-case rates, but every positive result leans on unverified companion-paper lemmas; worth refereeing once those are in hand. read the letter →

arxiv 2607.29532 v1 pith:DFQ4JPDG submitted 2026-07-31 econ.EM math.STstat.TH

classification econ.EMmath.STstat.TH MSC 62F3562F1262J05
keywords instrumentalvariablestwo-stageleastsquareswinsorizedmeanadversarialcontaminationminimaxratesrobustinferenceweakidentificationfinite-sampleconcentration
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper asks how much data contamination 2SLS can survive before its estimates and inference break, under a worst-case model in which the identities and values of altered observations may depend on the clean data. It proposes W-2SLS, a drop-in replacement that substitutes quantile-winsorized means for every sample average in the 2SLS formula. The paper proves that W-2SLS's estimation error is of order eta_n^{1-1/m} + n^{-1/2} under finite m-th moments, and that no estimator can improve on the dependence on the contamination fraction. It also draws the exact boundary for first-order-free robustness: when sqrt(n) eta_n^{1-1/m} -> 0, W-2SLS has the same limiting Gaussian law as clean-sample 2SLS, so standard confidence intervals and tests remain valid. A winsorized Anderson-Rubin test keeps weak-instrument validity, and a finite-sample analysis shows W-2SLS concentrates much better than 2SLS even on clean data.

What carries the argument

The central object is the quantile-winsorized mean: replace observations below the empirical epsilon-quantile by that quantile and observations above the empirical (1-epsilon)-quantile by that quantile, then average. W-2SLS inserts this estimator in place of every sample average in the 2SLS formula, with a winsorization level set to epsilon_n = 1.01 eta_n + lambda log(n)/n, deliberately slightly larger than the contamination budget. The load-bearing estimates are the companion lemmas controlling the distance between the winsorized mean of contaminated data and the arithmetic mean of clean data, and the distance between coordinatewise winsorization and joint winsorization; these make the whol

What would settle it

Use the two distributions Q0 and Q1 from Lemma D.1 in the just-identified location model, contaminate through the map (11) with fixed eta_n and n, and evaluate any candidate estimator's worst-case error over the class Q_m. The theorem predicts the worst-case error is at least 0.25 eta_n^{1-1/m} with probability at least 0.5(1-(e/4)^{eta_n n/2}); an estimator with uniformly smaller worst-case error, or numerical evidence that the total-variation bound (e/4)^{eta_n n/2} is violated, would refute the sharpness claim.

Watch

Extended reading notes

Core claim

The central claim is that replacing the fragile sample averages in 2SLS by quantile-winsorized means yields an estimator that is simultaneously minimax sharp under adversarial contamination with finite m-th moments, achieving an estimation error of order eta_n^{1-1/m} + n^{-1/2}; asymptotically first-order equivalent to clean-data 2SLS, with the same N(0, Omega) limit, whenever sqrt(n) eta_n^{1-1/m} -> 0; and feasible for inference via a positive-semidefinite winsorized covariance estimator and a winsorized Anderson-Rubin test that is also valid under weak identification. Matching lower bounds show the conditions on eta_n are necessary: no estimator can be uniformly consistent if eta_n does

Load-bearing premise

The engine behind every positive result is a set of bounds on quantile-winsorized means under heavy tails and adversarial contamination that the paper imports from the authors' companion work; if those companion lemmas are not correct, the rate, inference, and concentration claims collapse.

Editorial extensions

If this is right

  • W-2SLS has estimation error of order eta_n^{1-1/m} + n^{-1/2}; matching lower bounds show no estimator can uniformly improve the dependence on the contamination fraction.
  • When sqrt(n) eta_n^{1-1/m} -> 0, W-2SLS has the same N(0, Omega) limit as clean-sample 2SLS, so robustness is first-order free for t-tests and confidence intervals.
  • The winsorized Anderson-Rubin test is asymptotically chi-squared and remains valid under weak identification, heteroskedasticity, and adversarial contamination.
  • A positive-semidefinite winsorized covariance estimator makes feasible inference possible without sample-average constructions.
  • Even with no contamination, W-2SLS's uniform finite-sample deviation radius grows only like sqrt(log(1/delta)), while 2SLS's grows like sqrt(1/delta).

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Editorial extension: the explicit thresholds give practitioners a contamination-budget check: if a bound on the fraction of possibly manipulated observations satisfies the stated condition, standard 2SLS output remains trustworthy without redesigning the specification.
  • Editorial extension: the coordinatewise-versus-joint winsorization device used for the covariance estimator is not specific to IV; the same idea should carry over to other moment-based estimators, such as GMM, where averages also enter nonlinearly.
  • Editorial extension: the finite-sample concentration gap is empirically testable by rerunning just-delete-two-observations sensitivity analyses with W-2SLS; the theory predicts far fewer flipped conclusions than reported for ordinary 2SLS.
  • Editorial extension: the non-existence of honest adaptive confidence sets suggests that any practitioner claiming robustness to an unknown contamination level must be relying on assumptions beyond finite moments, not on the estimator alone.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 4 minor

Summary. The paper proposes W-2SLS, a robustified 2SLS estimator that replaces the sample averages in the 2SLS moment conditions by quantile-winsorized means. Under an adversarial contamination model in which an adversary may alter up to an η_n fraction of observations, with full knowledge of the clean sample, the paper claims that W-2SLS attains the estimation error O_P(η_n^{1-1/m} + n^{-1/2}) under m-th moment assumptions, and that this rate is minimax sharp via matching lower bounds. It further claims that centered Gaussian inference with the same asymptotic covariance as clean-data 2SLS is possible when √n η_n^{1-1/m} → 0, and that this condition is necessary. The paper also constructs a PSD heteroskedasticity-robust covariance estimator, a winsorized Anderson-Rubin test valid under weak identification and adversarial contamination, and finite-sample uniform deviation bounds showing W-2SLS has sub-Gaussian concentration while ordinary 2SLS does not. The lower-bound arguments are largely self-contained, but the upper-bound and inference results rely on lemmas whose key inequalities are deferred to a companion paper by the same authors that is cited as 'to appear'.

Significance. If the companion-paper lemmas are correct, the paper makes a substantial contribution. It gives a simple closed-form estimator with a sharp minimax rate under a very flexible contamination model, identifies the exact threshold for root-n consistency and for clean-data Gaussian inference, and provides feasible inference. The lower-bound constructions are elegant and self-contained, and the paper is honest about the dependence of the upper bounds on external results. The finite-sample concentration comparison between 2SLS and W-2SLS is also a valuable contribution. However, the central positive claims are not currently verifiable from the manuscript alone, because the key step converting contamination into the η^{1-1/m} rate is delegated to a not-yet-available companion paper. This is a transparency and verifiability problem, not circularity.

major comments (3)
  1. [Appendix G, Lemma G.1, Eq. (G.9)] The proof of Lemma G.1 is not self-contained. The key inequality (G.9) — that the contaminated winsorization points α̂ and β̂ lie between clean quantiles Q_{c1ε}(S1) and Q_{1−c1ε}(S1) — is asserted by reference to Lemma B.5 of Kock and Preinerstorfer (2026, to appear). This containment is exactly what converts the contamination budget η into the ε^{1-1/m} bound through (G.10). The lower bounds on c1 and c2 are also delegated to Lemma B.3 of the same companion paper. Since Proposition 2.1, and hence Theorems 2.2, 2.3, 2.4, 4.1, and 5.2, all funnel through Lemma G.1, these results are currently unsupported unless the companion-paper lemmas are reproduced or supplied in full. The manuscript should either include complete proofs of these lemmas or state them as assumptions/conditions with proofs available in the appendix.
  2. [Appendix C, Theorem 2.4; Appendix E, Theorem 4.1] The consistency proofs for the PSD covariance estimator Ω̂ and the winsorized Anderson-Rubin covariance matrix rely on Theorem 2.1 of Kock and Preinerstorfer (2026, to appear) and on Lemma G.2, which itself invokes Lemma B.5, Lemma B.3, and Lemma C.1 of the same companion paper. These lemmas are not reproduced. In particular, Lemma G.2's bound (G.12) depends on the same quantile-containment facts and on bounds for c1 that are not derived here. Given that Theorem 2.4 is the basis for feasible inference and Theorem 4.1 is the basis for weak-identification-robust inference, the same external dependency applies. The authors should provide the missing proofs or otherwise make the companion results available for verification.
  3. [Section 5, Theorem 5.2 and Eq. (F.7)] The finite-sample sub-Gaussian deviation guarantee for W-2SLS is obtained by applying Theorem 2.1 of the companion paper with m=2. The constant B in Eq. (F.7) is defined in terms of companion-paper quantities A(1.01,1), B(1.01,1), l(1.01,1), and u(1.01,1), whose derivation is not included. Thus Theorem 5.2 is also conditional on the companion paper. Since this theorem drives the striking Corollary 5.3 comparison between 2SLS and W-2SLS, the proof should be completed or the companion results should be stated in full.
minor comments (4)
  1. [Appendix G, Lemma G.1] The probability bound in Lemma G.1 reads 'with probability at least 1 − 4/6 δ − 1/M', which is awkwardly written and could be misread as 1 − (4/6)δ − 1/M. Please clarify, e.g., '1 − (2/3)δ − 1/M', and ensure the intended meaning is transparent.
  2. [Section 2.2, Eq. (7)] The choice ε_n = 1.01 η_n + λ log(n)/n uses a fixed 1.01 factor. The paper motivates this as 'slightly more than η_n', but it may help to state explicitly that the 1.01 factor is arbitrary and can be replaced by any constant strictly greater than 1; the proofs appear to only need λ_1 > 1.
  3. [Section 3, Theorem 3.1] The lower-bound construction in Lemma D.1 is a useful concrete example, but the notation Q_{1,n} in Lemma D.1 and Q_1 in the proof is introduced with a slight inconsistency. Please harmonize the notation.
  4. [Appendix G, Eq. (G.14)] The bound in (G.14) is stated after a chain of inequalities; for readability, please indicate which lines use Cauchy-Schwarz, Hölder, and Lemma C.1 respectively, since these steps are important for verification.

Circularity Check

0 steps flagged · score 2.0 of 10

No significant circularity: W-2SLS rates follow from a general winsorized-mean lemma; the main caveat is an unverified companion-paper dependency, not a definitional or fitted-input circularity.

full rationale

After tracing the derivation chain, I find no step in which a claimed output is equivalent by construction to an input. The upper-bound story is: (i) Theorem 2.2 follows from Proposition 2.1 plus clean-data 2SLS; (ii) Proposition 2.1's proof says 'we do by verifying the conditions of Lemma G.1' (Appendix A), so the entire upper bound is delegated to the general winsorized-mean lemma; (iii) Lemma G.1 itself is a statement about i.i.d. S_i with E|S_1|^m<∞ and |{i: tilde S_i != S_i}|≤ηn, not about β or W-2SLS. The same holds for Lemma G.2/Theorem 2.4 and Theorem 4.1. These lemmas are in turn proved from Lemmas B.5, B.3, C.1 of Kock & Preinerstorfer (2026); those are also general quantile-containment/moment bounds, not statements about IV, and they are not fitted to data. This is a heavy, unverified (to-appear) self-citation dependency, so the manuscript is not fully self-contained, but it is not circular: the cited statements have assumptions that do not include the target result. The lower-bound side (Theorem 3.1) is self-contained; its proof uses only Lemmas D.1-D.2 and a TV bound, and it never invokes W-2SLS. Section 5's comparison is likewise a fresh lower-bound construction versus the companion theorem with η=0. No fitted parameter is renamed as a prediction; the only user input η_n is a pre-specified budget, and Remark 3.1/Theorem 3.2 show it cannot be estimated. Hence no Eq. X = Eq. Y by construction.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

No new entities are postulated; W-2SLS is a construction from existing winsorized means. The main unverified inputs are external lemmas, not invented objects.

free parameters (3)
  • Winsorization tuning epsilon_n = 1.01 eta_n + lambda log(n)/n = 1.01 eta_n + lambda log(n)/n (lambda > 0 fixed, e.g., 1)
    The 1.01 factor and lambda are chosen by hand; the theorems require epsilon_n > eta_n and epsilon_n -> 0, and hold for any fixed lambda > 0, so this is a design choice rather than a fitted constant.
  • Robustness budget eta_n = specified by researcher; not estimated
    All positive results and the implementation of W-2SLS require the researcher to commit to a maximum contamination fraction; Remark 3.1 shows the true fraction cannot be estimated at precision o(eta_n).
  • Failure probability delta in Section 5 = user-chosen, e.g., 12 n^{-lambda}
    The finite-sample deviation radius of W-2SLS uses epsilon = log(12/delta)/n and the bound depends on log(1/delta); delta is a user-specified confidence level, not fitted to data.
assumptions (4)
  • domain assumption Clean data are i.i.d. with E|z_l x_k|^m, E|z_l z_j|^m, E|u z_l|^m finite, rank[E(zz')]=L, rank[E(zx')]=K (Assumption 2.1).
    This is the moment and identification framework for all positive results.
  • domain assumption The adversary may alter at most eta_n n observations, with identities and replacement values depending on the clean sample (Section 2, equation (2)).
    This defines the contamination model; the guarantees are uniform over all such contamination rules.
  • domain assumption Instrument exogeneity E(z_1 u_1)=0 and rank[E(u_1^2 z_1 z_1')]=L for inference results (Theorems 2.2-2.4).
    Needed for clean-sample 2SLS asymptotic normality and for consistent covariance estimation.
  • standard math Correctness of the companion-paper quantile-winsorized mean bounds: Lemmas B.5, B.3, C.1 and Theorem 2.1 of Kock and Preinerstorfer (2026).
    Invoked in Lemma G.1, Lemma G.2, and the proofs of Theorems 2.4, 4.1, and 5.2; not proved in this manuscript.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Robust Instrumental Variables: Sharp Rates and Inference under Adversarial Contamination." pith.science (2026). https://pith.science/paper/DFQ4JPDG

@misc{pith2026260729532,
  author       = {Pith},
  title        = {Pith review of: Robust Instrumental Variables: Sharp Rates and Inference under Adversarial Contamination},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/DFQ4JPDG}},
  note         = {Machine review of arXiv:2607.29532}
}
abstract

Because 2SLS is built from sample averages, a small number of observations can have a disproportionate effect on estimates and inference. We introduce W-2SLS, a simple drop-in robustification that replaces these averages by quantile-winsorized means. We analyze W-2SLS under adversarial contamination, which permits both the identities and the reported values of the contaminated observations to depend on the realized clean sample and therefore accommodates targeted or strategic manipulation. Under finite $m$-th moments, W-2SLS attains the minimax-sharp rate $\eta_{n}^{1-\frac1m}+n^{-1/2}$, where $\eta_n$ is the fraction of observations that may be altered. Matching lower bounds identify the exact contamination thresholds for uniform consistency, root-$n$ estimation, and centered Gaussian inference with the same first-order law as clean-sample 2SLS. When $\sqrt{n}\eta_{n}^{1-\frac1m}\to 0$ robustness is first-order free. We also construct feasible heteroskedasticity-robust inference and a winsorized Anderson--Rubin test valid under weak identification and adversarial contamination. Finally, even without contamination, ordinary 2SLS can have poor uniform finite-sample concentration, whereas W-2SLS admits confidence-calibrated sub-Gaussian deviation guarantees.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

46 extracted references · 4 linked inside Pith

  1. [1]

    Estimation of the parameters of a single equation in a complete system of stochastic equations,

    Anderson, T. W. and H. Rubin(1949): “Estimation of the parameters of a single equation in a complete system of stochastic equations,”The Annals of mathematical statistics, 20, 46–63. 51

  2. [2]

    In a small moment: Class size and moral hazard in the Italian Mezzogiorno,

    Angrist, J. D., E. Battistin, and D. Vuri(2017): “In a small moment: Class size and moral hazard in the Italian Mezzogiorno,”American Economic Journal: Applied Economics, 9, 216–249

  3. [3]

    Maimonides rule redux,

    Angrist, J. D., V. Lavy, J. Leder-Luis, and A. Shany(2019): “Maimonides rule redux,”American Economic Review: Insights, 1, 309–324

  4. [4]

    Minimax m-estimation under adversarial contamination,

    Bhatt, S., G. F ang, P. Li, and G. Samorodnitsky(2022): “Minimax m-estimation under adversarial contamination,” inInternational Conference on Machine Learning, PMLR, 1906–1924

  5. [5]

    Star wars: The empirics strike back,

    Brodeur, A., M. L ´e, M. Sangnier, and Y. Zylberberg(2016): “Star wars: The empirics strike back,”American Economic Journal: Applied Economics, 8, 1–32

  6. [6]

    Manipulation of social program eligibility,

    Camacho, A. and E. Conover(2011): “Manipulation of social program eligibility,” American Economic Journal: Economic Policy, 3, 41–65

  7. [7]

    Challenging the empirical mean and empirical variance: a deviation study,

    Catoni, O.(2012): “Challenging the empirical mean and empirical variance: a deviation study,”Annales de l’IHP – Probabilit´ es et Statistiques, 48, 1148–1185

  8. [8]

    A general decision theory for Huber’sϵ- contamination model,

    Chen, M., C. Gao, and Z. Ren(2016): “A general decision theory for Huber’sϵ- contamination model,”Electronic Journal of Statistics, 1935–7524

Show all 46 references
  1. [9]

    High-dimensional robust mean es- timation in nearly-linear time,

    Cheng, Y., I. Diakonikolas, and R. Ge(2019): “High-dimensional robust mean es- timation in nearly-linear time,” inProceedings of the thirtieth annual ACM-SIAM sym- posium on discrete algorithms, SIAM, 2755–2771. ˇC´ıˇzek, P.(2016): “Generalized method of trimmed moments,”Journ...

  2. [10]

    On the LambertWfunction,

    Corless, R. M., G. H. Gonnet, D. E. G. Hare, D. J. Jeffrey, and D. E. Knuth (1996): “On the LambertWfunction,”Advances in Computational Mathematics, 5, 329–359

  3. [11]

    How do hospitals respond to price changes?

    Dafny, L. S.(2005): “How do hospitals respond to price changes?”American Economic Review, 95, 1525–1547. 52

  4. [12]

    All-in-one robust estimator of the Gaussian mean,

    Dalalyan, A. S. and A. Minasyan(2022): “All-in-one robust estimator of the Gaussian mean,”Annals of Statistics, 50, 1193–1219

  5. [13]

    The causes and conse- quences of test score manipulation: Evidence from the New York regents examinations,

    Dee, T. S., W. Dobbie, B. A. Jacob, and J. Rockoff(2019): “The causes and conse- quences of test score manipulation: Evidence from the New York regents examinations,” American Economic Journal: Applied Economics, 11, 382–423

  6. [14]

    Robust sub-Gaussian estimation of a mean vector in nearly linear time,

    Depersin, J. and G. Lecu ´e(2022): “Robust sub-Gaussian estimation of a mean vector in nearly linear time,”Annals of Statistics, 50, 511–536

  7. [15]

    Robust estimators in high-dimensions without the computational intractability,

    Diakonikolas, I., G. Kamath, D. Kane, J. Li, A. Moitra, and A. Stewart(2019): “Robust estimators in high-dimensions without the computational intractability,”SIAM Journal on Computing, 48, 742–864

  8. [16]

    Diakonikolas, I. and D. Kane(2023):Algorithmic high-dimensional robust statistics, Cambridge University Press

  9. [17]

    Truth-telling by third- party auditors and the response of polluting firms: Experimental evidence from India,

    Duflo, E., M. Greenstone, R. Pande, and N. Ryan(2013): “Truth-telling by third- party auditors and the response of polluting firms: Experimental evidence from India,” The Quarterly Journal of Economics, 128, 1499–1545

  10. [18]

    Tax rates and tax evasion: Evidence from “missing imports

    Fisman, R. and S.-J. Wei(2004): “Tax rates and tax evasion: Evidence from “missing imports” in China,”Journal of Political Economy, 112, 471–496

  11. [19]

    Occasionally misspecified,

    Forneron, J.-J.(2023): “Occasionally misspecified,”arXiv preprint arXiv:2312.05342

  12. [20]

    Upcoding: evidence from Medicare on squishy risk adjustment,

    Geruso, M. and T. Layton(2020): “Upcoding: evidence from Medicare on squishy risk adjustment,”Journal of Political Economy, 128, 984–1026

  13. [21]

    A guided tour of Chernoff bounds,

    Hagerup, T. and C. R ¨ub(1990): “A guided tour of Chernoff bounds,”Information Processing Letters, 33, 305–308

  14. [22]

    Robust and heavy-tailed mean estima- tion made simple, via regret minimization,

    Hopkins, S., J. Li, and F. Zhang(2020): “Robust and heavy-tailed mean estima- tion made simple, via regret minimization,”Advances in Neural Information Processing Systems, 33, 11902–11912

  15. [23]

    Rotten apples: An investigation of the preva- lence and predictors of teacher cheating,

    Jacob, B. A. and S. D. Levitt(2003): “Rotten apples: An investigation of the preva- lence and predictors of teacher cheating,”The Quarterly Journal of Economics, 118, 843–877. 53

  16. [24]

    A simple robust procedure in instrumental variables regression,

    Jiao, X.(2024): “A simple robust procedure in instrumental variables regression,” Tech. rep

  17. [25]

    Two-stage Huber estimation,

    Kim, T.-H. and C. Muller(2007): “Two-stage Huber estimation,”Journal of statistical planning and inference, 137, 405–418

  18. [26]

    Robustness, infinitesimal neigh- borhoods, and moment restrictions,

    Kitamura, Y., T. Otsu, and K. Evdokimov(2013): “Robustness, infinitesimal neigh- borhoods, and moment restrictions,”Econometrica, 81, 1185–1201

  19. [27]

    Outlier robust inference in the instrumental variable model with applications to causal effects,

    Klooster, J. and M. Zhelonkin(2024a): “Outlier robust inference in the instrumental variable model with applications to causal effects,”Journal of Applied Econometrics, 39, 86–106. ——— (2024b): “Resistant Inference in Instrumental Variable Models,”arXiv preprint arXiv:2403.16844

  20. [28]

    High-dimensional Gaussian approxima- tions for robust means,

    Kock, A. B. and D. Preinerstorfer(2025): “High-dimensional Gaussian approxima- tions for robust means,” . ——— (2026): “Winsorized mean estimation with heavy tails and adversarial contamina- tion,”Electronic Journal of Statistics (to appear)

  21. [29]

    Two-stage bounded-influence estimators for simultaneous- equations models,

    Krasker, W.(1986): “Two-stage bounded-influence estimators for simultaneous- equations models,”Journal of Business & Economic Statistics, 4, 437–444

  22. [30]

    Resistant estimation for simultaneous-equations models using weighted instrumental variables,

    Krasker, W. and R. Welsch(1985): “Resistant estimation for simultaneous-equations models using weighted instrumental variables,”Econometrica, 1475–1488

  23. [31]

    Robust estimators for simultaneous equations models,

    Krishnakumar, J. and E. Ronchetti(1997): “Robust estimators for simultaneous equations models,”Journal of Econometrics, 78, 295–314

  24. [32]

    Agnostic estimation of mean and covariance,

    Lai, K. A., A. B. Rao, and S. Vempala(2016): “Agnostic estimation of mean and covariance,” in2016 IEEE 57th Annual Symposium on Foundations of Computer Science (FOCS), IEEE, 665–674

  25. [33]

    Robust multivariate mean estimation: The optimality of trimmed mean,

    Lugosi, G. and S. Mendelson(2021): “Robust multivariate mean estimation: The optimality of trimmed mean,”Annals of Statistics, 49, 393–410

  26. [34]

    Adaptive robust confidence intervals,

    Luo, Y. and C. Gao(2026): “Adaptive robust confidence intervals,”The Annals of Statistics, 54, 1128–1152. 54

  27. [35]

    How much should we trust the dictator’s GDP growth esti- mates?

    Martinez, L. R.(2022): “How much should we trust the dictator’s GDP growth esti- mates?”Journal of Political Economy, 130, 2731–2769

  28. [36]

    Risk Protection, Service Use, and Health Outcomes under Colombia’s Health Insurance Program for the Poor,

    Miller, G., D. Pinto, and M. Vera-Hern ´andez(2013): “Risk Protection, Service Use, and Health Outcomes under Colombia’s Health Insurance Program for the Poor,” American Economic Journal. Applied Economics, 5, 61–91

  29. [37]

    Statistically optimal robust mean and covariance estimation for anisotropic Gaussians,

    Minasyan, A. and N. Zhivotovskiy(2023): “Statistically optimal robust mean and covariance estimation for anisotropic Gaussians,”arXiv preprint arXiv:2301.09024

  30. [38]

    Efficient median of means estimator,

    Minsker, S.(2023): “Efficient median of means estimator,” inThe Thirty Sixth Annual Conference on Learning Theory, PMLR, 5925–5933

  31. [39]

    Robust and efficient mean estimation: an ap- proach based on the properties of self-normalized sums,

    Minsker, S. and M. Ndaoud(2021): “Robust and efficient mean estimation: an ap- proach based on the properties of self-normalized sums,”Electronic Journal of Statistics, 15, 6036–6070

  32. [40]

    Finite-sample properties of the trimmed mean,

    Oliveira, R., P. Orenstein, and Z. Rico(2025): “Finite-sample properties of the trimmed mean,”arXiv preprint arXiv:2501.03694

  33. [41]

    Robust generalized method of moments: a finite sample viewpoint,

    Rohatgi, D. and V. Syrgkanis(2022): “Robust generalized method of moments: a finite sample viewpoint,”Advances in Neural Information Processing Systems, 35, 15970– 15981

  34. [42]

    Romano, J. and A. Siegel(1985):Counterexamples in probability and statistics, Wadsworth & Brooks/Cole

  35. [43]

    Robust inference with GMM estimators,

    Ronchetti, E. and F. Trojani(2001): “Robust inference with GMM estimators,” Journal of econometrics, 101, 37–69. Sølvsten, M.(2020): “Robust estimation with many instruments,”Journal of Econo- metrics, 214, 495–512. W agenvoort, R. and R. W aldmann(2002): “On B-robust instrume...

  36. [44]

    Wooldridge, J.(2010):Econometric analysis of cross section and panel data, MIT Press. 55

  37. [45]

    Channeling fisher: Randomization tests and the statistical insignif- icance of seemingly significant experimental results,

    Young, A.(2019): “Channeling fisher: Randomization tests and the statistical insignif- icance of seemingly significant experimental results,”The Quarterly Journal of Eco- nomics, 134, 557–598. ——— (2022): “Consistency without inference: Instrumental variables in practical appl...

  38. [46]

    On the robustness of two-stage estimators,

    Zhelonkin, M., M. G. Genton, and E. Ronchetti(2012): “On the robustness of two-stage estimators,”Statistics & Probability Letters, 82, 726–732. 56

Pith tools

Reviewed August 3, 2026 · model on record in the stance chip above.