Identification and Inference i n Nonlinear Panel Models: An Adversarial Approach
Coauth ors: Irene Botosaru and Isaac Loh
Abstract:
In nonlinear panel models with fixed effects\, structural parameters and counterfactual object s such as the average structural function are typically partially identifie d\, and restrictions imposed conditional on the fixed effects induce a cont inuum of moment conditions. Existing approaches to identification and infer ence are largely model-specific. We propose a unifying approach that formul ates identification as a set-membership question in the space of probabilit y measures on observables. The central object is an adversarial discrepancy function whose zero set sharply characterizes the identified set. Point id entification need not be established a priori.
The discrepancy functi on is the value of a semi-infinite linear program\, which we solve by row-a nd-column generation with optimality certificates. For discrete outcomes\, the sample analogue admits the same representation. We derive its limiting distribution and a uniformly valid penalized bootstrap whose contact-set pe nalty mitigates conservativeness. Applied to binary choice panels\, the fra mework delivers the first sharp identification under sequential exogeneity with unrestricted errors and the first sharp identified set for probit with interval-censored covariates\, while nesting classical point-identificatio n results\, including conditional logit\, as special cases.
Organized by: Vadim Marmer
LOCATION:IONA 533 GEO:49.260872;-123.113952 URL;VALUE=URI:https://economics.ubc.ca/events/event/chris-muris-mcmaster-un iversity-econometrics-seminar/ END:VEVENT BEGIN:VTIMEZONE TZID:America/Vancouver BEGIN:DAYLIGHT TZOFFSETFROM:-0800 TZOFFSETTO:-0700 DTSTART:20260308T100000 TZNAME:PDT END:DAYLIGHT END:VTIMEZONE END:VCALENDAR