Tuning-Free Tests for Treatment Effect Heterogeneity via Expectile Processes

Haokun Lu; Xiaojun Song

Home/Archive/arXiv

arXiv · 2610.25986

Tuning-Free Tests for Treatment Effect Heterogeneity via Expectile Processes

Haokun Lu·Xiaojun Song

View original source ↗Read source document ↗Export metadata ↓

Abstract

This paper develops tuning-free tests for treatment-effect heterogeneity. By characterizing the null hypothesis as constancy of the expectile treatment-effect curve, we construct test statistics as continuous functionals of the mean-centered sample expectile process. We establish a uniform Bahadur representation, derive the asymptotic properties of the test statistics under the null and fixed alternatives, and show nontrivial power against local alternatives approaching the null at the $n^{-1/2}$ rate. For feasible inference, we develop a fast multiplier bootstrap that consistently approximates the nonpivotal null distribution while capturing the first-order effect of estimating the unknown common shift. The procedure is tuning-free, requiring no density or score-function estimation, bandwidth selection, or repeated refitting. The framework further incorporates a pointwise-studentized refinement and a covariate adjustment that reduces asymptotic variation without altering the population discrepancy. Monte Carlo comparisons with existing methods show generally reliable size control, competitive power, and substantial computational savings. An application to a large-scale water-conservation experiment illustrates the proposed tests.

Explore related subjects

econ.EM

Keep this discovery

Explore connections, maps & timelines

BibTeX RIS

Haokun Lu, Xiaojun Song. 2026-10-19. Tuning-Free Tests for Treatment Effect Heterogeneity via Expectile Processes. https://arxiv.org/abs/2610.25986

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

What Impulse Response Do Instrumental Variables Identify?

We study what local projection instrumental variable (LP-IV) regressions identify when an instrument is correlated with several structural shocks, such as defense and non-defense spending shocks. LP-IV identifies an instrument-specific affine combination of the component impulse responses, possibly with negative weights. Over-identified 2SLS or GMM yields another such combination, and the over-identification test can reject even with valid instruments. With as many instruments as shocks, the component responses are identified when each shock has a policy variable (e.g., sectoral spending) and its impact on the other policy variables is restricted. We illustrate with government spending and tax multipliers.

econ.EM↗

arXiv2026

Identifying Elasticities in Autocorrelated Time Series Using Causal Mechanism Graphs

Instrumental variable (IV) methods are widely used to estimate the price elasticity of demand from observational data. However, naive IV estimators may be inconsistent in settings with autocorrelated time series. We argue that causal graphs taking into account time structure can simplify IV identification and help select consistent estimators. More precisely, we propose to model equilibrium conditions using causal mechanism time graphs. These (acyclic) graphs allow us to maintain the assumption of a simultaneous determination of prices and quantities and come with an interpretable notion of intervention. They further allow us to exploit recent advances in graphical inference to derive valid IV estimators that are consistent under weak assumptions. Finally, we argue that observing significant differences between the estimates of presumably valid estimators can help to reject false model assumptions, thereby improving our understanding of underlying economic dynamics. We apply our approach to the German electricity market, estimating the price elasticity of demand from simulated and real-world data. The findings underscore the importance of accounting for structural autocorrelation in IV-based analysis.

econ.EM↗

arXiv2026

Granular Instrumental Variables in Large Panels: Identification and Inference Across Strong, Nearly Weak, and Weak GIV

Granular Instrumental Variables (GIV), introduced by Gabaix and Koijen (2024), use idiosyncratic shocks to large units as instruments for aggregate endogenous variables. I develop asymptotic theory for GIV in large panels where both the cross section and time series grow ($N,T\to\infty$). I make two main contributions. First, I show that estimating the individual shocks used to construct the granular instrument has a first-order effect on the asymptotic variance of the GIV estimator. Existing methods do not account for this effect, leading to incorrect standard errors. I derive corrected inference. Second, I characterize how instrument strength changes as $N$ grows, yielding strong, nearly-weak, and weak regimes. In the newly characterized nearly-weak regime, the instrument vanishes asymptotically but the estimator remains consistent and asymptotically normal, at a slower rate, when $T$ grows sufficiently fast relative to $N$. I apply the theory to estimate short-run supply and demand elasticities for crude oil, copper, and natural gas.

econ.EM↗