The day in conversation for 2026-11-07
discussion
Transcript
The discussion: Iris: That moment condition paper really dives deep into how higher-order Neyman orthogonality fundamentally changes how we handle those nuisance parameter errors in estimation, specifically by requiring derivatives up to order q to vanish at the true values. Warren: It’s interesting because what I’m less familiar with is the specific mechanism for achieving that debiasing—how exactly are they using von Mises expansions and rooted trees to derive those closed-form expressions? Is it a purely algebraic simplification or does the tree structure introduce some non-linear complexity in the resulting function space? Ruth: The connection I see there is how that vanishing derivative requirement relates to our earlier work on model misspecification; we’ve always struggled with sensitivity when moving away from simple moment conditions, but this suggests a structured way to control that sensitivity based on the order q. Have you looked at how this specific q relates to the complexity of the underlying economic system we usually try to capture? Warren: I haven't focused on the connection yet because I’m more interested in how that concept interfaces with decision-making under uncertainty. For instance, if we adopt that orthogonality, does it simplify the approximation guarantees we might get from robust mechanism design papers, like the one about second-price auctions with reserves? Iris: That’s a good pivot point. If the estimation error is better controlled by this orthogonality structure, doesn't it provide a more stable foundation for applying those approximation bounds in mechanism design? I
More episodes
- Daily Summary for 2026-11-19
- 2608.21274-Recommendation Quality and the Concentration of Consumption: Experimental Evidence from Netflix
- 2408.02757-A nonparametric test for diurnal variation in spot correlation processes
- 2306.16393-High-Dimensional Canonical Correlation Analysis
- 2605.17299-Geometric Brownian motion with intermittent entries and exits
- 2607.04753-Fooling Yourself: how narratives shape beliefs
- 2408.14671-Inference on Treatment Effects with Repeatedly Mismeasured High-Dimensional Controls
- 2201.06898-Difference-in-Differences Estimators for Treatments Continuously Distributed at Every Period
- 2501.15307-Influence Functions: A Unified Differentiation-Based Framework
- The day in conversation for 2026-11-19