pith. sign in

arxiv: 2604.13008 · v1 · submitted 2026-04-14 · 📊 stat.ME · stat.AP· stat.ML

Nonparametric efficient inference for network quantile causal effects under partial interference

Chao Cheng , Fan Li This is my paper

Pith reviewed 2026-05-10 14:18 UTC · model grok-4.3

classification 📊 stat.ME stat.APstat.ML
keywords quantile causal effectsnetwork interferencepartial interferencenonparametric efficiencycross-fittingasymptotic normalitycausal inference
0
0 comments X

The pith

Nonparametric efficiency theory delivers consistent estimators with parametric rates for network quantile causal effects under partial interference.

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

The paper develops a nonparametric efficiency theory to estimate causal effects on the quantiles of outcomes when one person's treatment affects others only inside the same cluster. The resulting estimator converges at parametric rates and stays consistent and normal even when nuisance functions are estimated with flexible, adaptive techniques. A three-way cross-fitting step makes it possible to skip estimating the full conditional distribution of the outcome. A sympathetic reader would care because many interventions create within-group spillovers and quantile effects capture impacts across the outcome spectrum rather than averages alone.

Core claim

We develop a general nonparametric efficiency theory for estimating network quantile causal effects under partial interference. This theory produces a nonparametrically efficient estimator that is consistent and asymptotically normal with parametric convergence rates. The estimator supports flexible, data-adaptive estimation of complex nuisance functions by using a three-way cross-fitting procedure that avoids direct estimation of the conditional outcome distribution. The methods show adequate finite-sample performance in simulations and are applied to a clustered observational study.

What carries the argument

Nonparametric efficiency theory combined with three-way cross-fitting for network quantile causal effects under partial interference.

If this is right

  • The estimator is consistent and asymptotically normal with parametric rates.
  • It allows flexible data-adaptive estimation of nuisance functions.
  • Simulations confirm good finite-sample behavior.
  • The approach applies directly to clustered observational studies.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same efficiency bounds and cross-fitting technique could support estimation of other causal functionals beyond quantiles.
  • Researchers studying interventions in schools or workplaces could apply this to measure effects on outcome tails within clusters.
  • Similar procedures might enhance efficiency in causal analyses with other forms of dependence among units.

Load-bearing premise

The partial interference assumption holds, meaning treatment effects do not spill across clusters, and the three-way cross-fitting procedure successfully avoids direct estimation of the conditional outcome distribution.

What would settle it

A simulation study or real data analysis where the estimator fails to achieve the parametric convergence rate under verified partial interference would falsify the claims of nonparametric efficiency.

Figures

Figures reproduced from arXiv: 2604.13008 by Chao Cheng, Fan Li.

Figure 1
Figure 1. Figure 1: Illustration of the three-way cross-fitting procedure with [PITH_FULL_IMAGE:figures/full_fig_p020_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: The quantile direct effect, overall effect, and spillover effects under the incremental [PITH_FULL_IMAGE:figures/full_fig_p030_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: The quantile direct effect, overall effect, and spillover effects under the incremental [PITH_FULL_IMAGE:figures/full_fig_p031_3.png] view at source ↗
read the original abstract

Interference arises when the treatment assigned to one individual affects the outcomes of other individuals. Commonly, individuals are naturally grouped into clusters, and interference occurs only among individuals within the same cluster, a setting referred to as partial interference. We study network causal effects on outcome quantiles in the presence of partial interference. We develop a general nonparametric efficiency theory for estimating these network quantile causal effects, which leads to a nonparametrically efficient estimator. The proposed estimator is consistent and asymptotically normal with parametric convergence rates, while allowing for flexible, data-adaptive estimation of complex nuisance functions. We leverage a three-way cross-fitting procedure that avoids direct estimation of the conditional outcome distribution. Simulations demonstrate adequate finite-sample performance of the proposed estimators, and we apply the methods to a clustered observational study.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

0 major / 0 minor

Summary. The paper develops a general nonparametric efficiency theory for estimating network quantile causal effects under partial interference. This theory leads to a nonparametrically efficient estimator that is consistent and asymptotically normal with parametric convergence rates. The estimator allows for flexible, data-adaptive estimation of complex nuisance functions using a three-way cross-fitting procedure that avoids direct estimation of the conditional outcome distribution. The methods are illustrated with simulations showing adequate finite-sample performance and applied to a clustered observational study.

Significance. If the theoretical results are correct, this work would be significant as it extends nonparametric efficiency theory to quantile causal effects in the presence of partial interference, a common setting in network and clustered data. The achievement of parametric rates with data-adaptive nuisances and the innovative use of three-way cross-fitting to sidestep direct conditional distribution estimation are strengths that enhance the practicality of the method. This could facilitate more robust inference on distributional treatment effects in interference settings.

Simulated Author's Rebuttal

0 responses · 0 unresolved

We thank the referee for their positive summary of our work on nonparametric efficient inference for network quantile causal effects under partial interference. The referee accurately captures the main contributions, including the efficiency theory, the three-way cross-fitting approach that avoids direct conditional distribution estimation, and the achievement of parametric convergence rates with data-adaptive nuisances. We are pleased that the potential significance for robust inference on distributional effects in interference settings is recognized. Since the report lists no specific major comments, we have no points to address point-by-point at this stage and will incorporate any minor revisions as recommended.

Circularity Check

0 steps flagged

No significant circularity

full rationale

The paper derives a nonparametric efficiency theory for quantile causal effects under partial interference, yielding a consistent asymptotically normal estimator at parametric rates via data-adaptive nuisance estimation and three-way cross-fitting. This construction follows standard semiparametric efficiency arguments applied to the clustered interference setting and does not reduce any claimed result to a fitted parameter or self-referential definition by construction. The partial interference assumption and cross-fitting procedure are externally motivated and do not presuppose the target estimator's properties. No load-bearing self-citation chains or ansatz smuggling appear in the derivation chain.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The approach rests on standard causal inference assumptions plus the partial interference structure; no free parameters or invented entities are evident from the abstract.

axioms (1)
  • domain assumption Partial interference: treatment of one unit affects only units within the same cluster.
    Stated as the setting for network effects in clustered data.

pith-pipeline@v0.9.0 · 5423 in / 1140 out tokens · 22003 ms · 2026-05-10T14:18:48.400232+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

41 extracted references · 41 canonical work pages

  1. [1]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION word.in bbl.in " " * FUNCTION format....

    work page

  2. [2]

    (2002), Instrumental variables estimates of the effect of subsidized training on the quantiles of trainee earnings, Econometrica\/ , 70, 91--117

    Abadie, A., Angrist, J., and Imbens, G. (2002), Instrumental variables estimates of the effect of subsidized training on the quantiles of trainee earnings, Econometrica\/ , 70, 91--117

    work page 2002

  3. [3]

    Aronow, P. M. and Samii, C. (2017), Estimating average causal effects under general interference, with application to a social network experiment, The Annals of Applied Statistics\/ , 11, 1912--1947

    work page 2017

  4. [4]

    G., Hudgens, M

    Barkley, B. G., Hudgens, M. G., Clemens, J. D., Ali, M., and Emch, M. E. (2020), Causal inference from observational studies with clustered interference, with application to a cholera vaccine study, Annals of Applied Statistics\/

    work page 2020

  5. [5]

    and Feller, A

    Basse, G. and Feller, A. (2018), Analyzing two-stage experiments in the presence of interference, Journal of the American Statistical Association\/ , 113, 41--55

    work page 2018

  6. [6]

    (2018), Uniformly valid post-regularization confidence regions for many functional parameters in z-estimation framework, Annals of statistics\/ , 46, 3643

    Belloni, A., Chernozhukov, V., Chetverikov, D., and Wei, Y. (2018), Uniformly valid post-regularization confidence regions for many functional parameters in z-estimation framework, Annals of statistics\/ , 46, 3643

    work page 2018

  7. [7]

    J., Klaassen, C

    Bickel, P. J., Klaassen, C. A., Bickel, P. J., Ritov, Y., Klaassen, J., Wellner, J. A., and Ritov, Y. (1993), Efficient and adaptive estimation for semiparametric models\/ , volume 4, Springer

    work page 1993

  8. [8]

    Cattaneo, M. D. (2010), Efficient semiparametric estimation of multi-valued treatment effects under ignorability, Journal of econometrics\/ , 155, 138--154

    work page 2010

  9. [9]

    (2024), Doubly robust estimation and sensitivity analysis for marginal structural quantile models, Biometrics\/ , 80, ujae045

    Cheng, C., Hu, L., and Li, F. (2024), Doubly robust estimation and sensitivity analysis for marginal structural quantile models, Biometrics\/ , 80, ujae045

    work page 2024

  10. [10]

    and Li, F

    Cheng, C. and Li, F. (2025), Inverting estimating equations for causal inference on quantiles, Biometrika\/ , 112, asae058

    work page 2025

  11. [11]

    --- (2026), Semiparametric causal mediation analysis of cluster-randomized trials for indirect and spillover effects, Biometrics\/ , 82, ujag017

    work page 2026

  12. [12]

    (2018), Double/debiased machine learning for treatment and structural parameters, The Econometrics Journal\/ , C1--C68

    Chernozhukov, V., Chetverikov, D., Demirer, M., Duflo, E., Hansen, C., Newey, W., and Robins, J. (2018), Double/debiased machine learning for treatment and structural parameters, The Econometrics Journal\/ , C1--C68

    work page 2018

  13. [13]

    and Hansen, C

    Chernozhukov, V. and Hansen, C. (2005), An IV model of quantile treatment effects, Econometrica\/ , 73, 245--261

    work page 2005

  14. [14]

    and van der Laan, M

    D \' az, I. and van der Laan, M. J. (2012), Population Intervention Causal Effects Based on Stochastic Interventions, Biometrics\/ , 68, 541--549

    work page 2012

  15. [15]

    (2007), Efficient semiparametric estimation of quantile treatment effects, Econometrica\/ , 75, 259--276

    Firpo, S. (2007), Efficient semiparametric estimation of quantile treatment effects, Econometrica\/ , 75, 259--276

    work page 2007

  16. [16]

    M., and Mealli, F

    Forastiere, L., Airoldi, E. M., and Mealli, F. (2021), Identification and estimation of treatment and interference effects in observational studies on networks, Journal of the American Statistical Association\/ , 116, 901--918

    work page 2021

  17. [17]

    R., Fr \"o lich, M., and Melly, B

    Frandsen, B. R., Fr \"o lich, M., and Melly, B. (2012), Quantile treatment effects in the regression discontinuity design, Journal of Econometrics\/ , 168, 382--395

    work page 2012

  18. [18]

    and Melly, B

    Fr \"o lich, M. and Melly, B. (2013), Unconditional quantile treatment effects under endogeneity, Journal of business & economic statistics\/ , 31, 346--357

    work page 2013

  19. [19]

    M., Halpern, C

    Harris, K. M., Halpern, C. T., Whitsel, E. A., Hussey, J. M., Killeya-Jones, L. A., Tabor, J., and Dean, S. C. (2019), Cohort profile: The national longitudinal study of adolescent to adult health (add health), International journal of epidemiology\/ , 48, 1415--1415k

    work page 2019

  20. [20]

    M., and Zhou, W.-X

    He, X., Pan, X., Tan, K. M., and Zhou, W.-X. (2023), Smoothed quantile regression with large-scale inference, Journal of Econometrics\/ , 232, 367--388

    work page 2023

  21. [21]

    (2007), Smoothed rank regression with censored data, Journal of the American Statistical Association\/ , 102, 552--559

    Heller, G. (2007), Smoothed rank regression with censored data, Journal of the American Statistical Association\/ , 102, 552--559

    work page 2007

  22. [22]

    (2023), Doubly robust estimation of direct and indirect quantile treatment effects with machine learning, arXiv preprint arXiv:2307.01049\/

    Hsu, Y.-C., Huber, M., and Yen, Y.-M. (2023), Doubly robust estimation of direct and indirect quantile treatment effects with machine learning, arXiv preprint arXiv:2307.01049\/

    work page arXiv 2023

  23. [23]

    (2022), Average direct and indirect causal effects under interference, Biometrika\/ , 109, 1165--1172

    Hu, Y., Li, S., and Wager, S. (2022), Average direct and indirect causal effects under interference, Biometrika\/ , 109, 1165--1172

    work page 2022

  24. [24]

    Hudgens, M. G. and Halloran, M. E. (2008), Toward causal inference with interference, Journal of the american statistical association\/ , 103, 832--842

    work page 2008

  25. [25]

    (2021), Causal inference with interference and noncompliance in two-stage randomized experiments, Journal of the American Statistical Association\/ , 116, 632--644

    Imai, K., Jiang, Z., and Malani, A. (2021), Causal inference with interference and noncompliance in two-stage randomized experiments, Journal of the American Statistical Association\/ , 116, 632--644

    work page 2021

  26. [26]

    (2024), Localized debiased machine learning: Efficient inference on quantile treatment effects and beyond, Journal of Machine Learning Research\/ , 25, 1--59

    Kallus, N., Mao, X., and Uehara, M. (2024), Localized debiased machine learning: Efficient inference on quantile treatment effects and beyond, Journal of Machine Learning Research\/ , 25, 1--59

    work page 2024

  27. [27]

    Kennedy, E. H. (2019), Nonparametric causal effects based on incremental propensity score interventions, Journal of the American Statistical Association\/ , 114, 645--656

    work page 2019

  28. [28]

    --- (2024), Semiparametric doubly robust targeted double machine learning: a review, Handbook of statistical methods for precision medicine\/ , 207--236

    work page 2024

  29. [29]

    D., and Hudgens, M

    Lee, C., Zeng, D., Emch, M., Clemens, J. D., and Hudgens, M. G. (2024), Nonparametric causal survival analysis with clustered interference, arXiv preprint arXiv:2409.13190\/

    work page arXiv 2024

  30. [30]

    Lee, C., Zeng, D., and Hudgens, M. G. (2025), Efficient nonparametric estimation of stochastic policy effects with clustered interference, Journal of the American Statistical Association\/ , 120, 382--394

    work page 2025

  31. [31]

    and Hudgens, M

    Liu, L. and Hudgens, M. G. (2014), Large sample randomization inference of causal effects in the presence of interference, Journal of the american statistical association\/ , 109, 288--301

    work page 2014

  32. [32]

    G., Saul, B., Clemens, J

    Liu, L., Hudgens, M. G., Saul, B., Clemens, J. D., Ali, M., and Emch, M. E. (2019), Doubly robust estimation in observational studies with partial interference, Stat\/ , 8, e214

    work page 2019

  33. [33]

    L., Sofrygin, O., Diaz, I., and Van der Laan, M

    Ogburn, E. L., Sofrygin, O., Diaz, I., and Van der Laan, M. J. (2024), Causal inference for social network data, Journal of the American Statistical Association\/ , 119, 597--611

    work page 2024

  34. [34]

    Papadogeorgou, G., Mealli, F., and Zigler, C. M. (2019), Causal inference with interfering units for cluster and population level treatment allocation programs, Biometrics\/ , 75, 778--787

    work page 2019

  35. [35]

    (2024), Minimum resource threshold policy under partial interference, Journal of the American Statistical Association\/ , 119, 2881--2894

    Park, C., Chen, G., Yu, M., and Kang, H. (2024), Minimum resource threshold policy under partial interference, Journal of the American Statistical Association\/ , 119, 2881--2894

    work page 2024

  36. [36]

    and Kang, H

    Park, C. and Kang, H. (2022), Efficient semiparametric estimation of network treatment effects under partial interference, Biometrika\/ , 109, 1015--1031

    work page 2022

  37. [37]

    V., Van Der Laan, M

    Phillips, R. V., Van Der Laan, M. J., Lee, H., and Gruber, S. (2023), Practical considerations for specifying a super learner, International Journal of Epidemiology\/ , 52, 1276--1285

    work page 2023

  38. [38]

    arXiv preprint arXiv:2107.12420 , year=

    Qu, Z., Xiong, R., Liu, J., and Imbens, G. (2021), Semiparametric Estimation of Treatment Effects in Observational Studies with Heterogeneous Partial Interference, arXiv preprint arXiv:2107.12420\/

    work page arXiv 2021

  39. [39]

    (2024), Estimating counterfactual distributions under interference, Proceedings of Machine Learning in Healthcare

    Shankar, S., Sinha, R., and Fiterau, M. (2024), Estimating counterfactual distributions under interference, Proceedings of Machine Learning in Healthcare. PMLR\/

    work page 2024

  40. [40]

    Sobel, M. E. (2006), What do randomized studies of housing mobility demonstrate? Causal inference in the face of interference, Journal of the American Statistical Association\/ , 101, 1398--1407

    work page 2006

  41. [41]

    Tchetgen, E. J. T. and VanderWeele, T. J. (2012), On causal inference in the presence of interference, Statistical methods in medical research\/ , 21, 55--75

    work page 2012