Policy impact estimation

IFS is involved in assessing the effectiveness of a number of labour market programmes, tax and transfer programmes and social programmes in a variety of fields, from education and training, to labour supply, childcare, health and welfare. Determining whether such policy interventions work and whether their cost is justified is of crucial importance in the presence of limited public resources, and allows policy decisions to be guided by evidence on programme effectiveness.

Estimating the causal impact of a programme is difficult because one can never observe the outcome that programme participants would have experienced had they not participated. Constructing this unobserved counterfactual for programme participants is the central issue that evaluation methods need to overcome. In addition to the evaluation of specific government interventions, our research contributes to the development of econometric and statistical methods to address the evaluation problem.

Fixed effect estimation of large T panel data models

| Working Paper

This article reviews recent advances in fixed effect estimation of panel data models for long panels, where the number of time periods is relatively large. We focus on semiparametric models with unobserved individual and time effects, where the distribution of the outcome variable conditional on covariates and unobserved effects is specified parametrically, while the distribution of the unobserved effects is left unrestricted.

Find out more

Cross-fitting and fast remainder rates for semiparametric estimation

| Working Paper

There are many interesting and widely used estimators of a functional with finite semi-parametric variance bound that depend on nonparametric estimators of nuisance func-tions. We use cross-fitting to construct such estimators with fast remainder rates. We give cross-fit doubly robust estimators that use separate subsamples to estimate different nuisance functions.

Find out more

Posterior distribution of nondifferentiable functions

| Working Paper

This paper examines the asymptotic behavior of the posterior distribution of a possibly nondifferentiable function g(θ), where θ is a finite-dimensional parameter of either a parametric or semiparametric model. The main assumption is that the distribution of a suitable estimator θn, its bootstrap approximation, and the Bayesian posterior for θ all agree asymptotically.

Find out more