Dynamic Structural Econometrics

Short notes and worked examples on estimating structural models — with an emphasis on dynamic discrete choice (DDC) methods. Files render as standalone HTML pages.

Quick start for students: Click a topic card to open the HTML page. The current focus is on Bus Engine CCP, Unobserved Heterogeneity, and Dynamic Games. Additional notes extend the toolkit to continuous choice models and auctions.

Core notes for this workshop

Bus Engine Replacement — CCP Estimation core

Rust (1987) bus engine replacement model as the workhorse DDC example.
Conditional Choice Probability (CCP) estimator per Hotz & Miller (1993).
Arcidiacono & Ellickson (2011) formulation and implementation details.
Full-information MLE (à la Rust) included for comparison.

Unobserved Heterogeneity — EM Algorithm core

Step-by-step EM algorithm with a simple integer-data illustration.
Application to DDC models following Arcidiacono & Miller (2011).

Dynamic Games — Entry/Exit core

From single-agent DDC to multi-agent, simultaneous-move settings.
Entry/exit game as the empirical IO benchmark example.
Adapting single-agent estimators to fixed-point equilibria in games (Pesendorfer and Schmidt-Dengler (2008)).

Additional notes

Continuous Choice — Consumption/Saving supplement

Hansen & Singleton (1982): Euler-equation estimation framework.
Estimating utility parameters using aggregate consumption and asset returns.
Simple introduction to moment-based estimation of continuous-choice structural models.

Auctions — Static & Dynamic supplement

Estimation for first-price and second-price sealed-bid auctions.
Data-limited settings (e.g., only the winning bid) and identification strategies.
Extensions to dynamic ascending (English) auctions under frictionless bidding.

Note: If this link 404s, the file may still be named DynamicStructuralEconometrics_DynamicAuctions.html. Update the filename here after the rename.