|
Activity Number:
|
614
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 3, 2016 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Transportation Statistics Interest Group
|
|
Abstract #318788
|
View Presentation
|
|
Title:
|
Quantifying the Causal Effect of Speed Cameras on Road Traffic Accidents via an Approximate Bayesian Doubly Robust Estimator
|
|
Author(s):
|
Daniel Graham* and Haojie Li
|
|
Companies:
|
Imperial College London and Southeast University
|
|
Keywords:
|
Bayesian inference ;
Doubly robust ;
Propensity Score ;
treatment effect
|
|
Abstract:
|
Doubly robust (DR) estimation combines outcome regression (OR) with inverse propensity score (PS) weighting to derive an average treatment effects (ATEs) estimator which is consistent and asymptotically normal under correct specification of either the OR or the PS models. DR estimators are typically constructed as solutions to estimating equations based on a set of moment restrictions. Standard Bayesian methods are difficult to apply because restricted moment models do not imply fully specified likelihood functions. This paper applies a Bayesian bootstrap approach to derive approximate posterior predictive distributions that are DR for average treatment effect (ATE) estimation. We apply the estimator to study of the effects of speed cameras on road traffic accidents in the UK. Our empirical results show significant reductions in the number of accidents of all severities at speed cameras sites.
|
Authors who are presenting talks have a * after their name.
