Christian Hansen’s Research Page

 

Code for IVQR

 

Below are links to MATLAB and Ox code for performing IVQR estimation and inference as developed in “Instrumental Quantile Regression Inference for Structural and Treatment Effect Models” (with Victor Chernozhukov) and “Instrumental Variable Quantile Regression” (with Victor Chernozhukov).  Along with the code, each file contains examples illustrating how the code may be implemented; the data for the examples may also be downloaded below.

 

1.  MATLAB Code             

2.  Ox Code                          

3.  Data for examples

 

Code for Weak Instrument Robust Inference

 

Below are links for the Stata code and data used in the empirical example in “A Simple Approach to Heteroskedasticity and Autocorrelation Robust Inference with Weak Instruments” (with Victor Chernozhukov).  The data are taken from Acemoglu, Johnson, and Robinson (2001) “The Colonial Origins of Comparative Development: An Empirical Investigation”.  The code illustrates the basic procedure and may easily be modified for other data sets and to provide inference that is robust to autocorrelation or clustering.

 

1.  Stata Code for weak instrument robust inference

2.  Data

 

Code for Sensitivity Analysis for IV (from “Plausibly Exogenous”)

 

Below are links for Stata code that produces some of the results from “Plausibly Exogenous” (with Tim Conley and Peter Rossi).  The code illustrates the basic procedure and may easily be modified for other data sets.  The file with the Stata code also includes sample data.

 

1.  Stata Code for IV sensitivity analysis

 

 

Working Papers

Only unpublished work appears here.  A complete list of research including publications may be found on my CV.

 

1.  Finite-Sample Inference In Econometric Models via Quantile Restrictions”, (with Victor Chernozhukov and Michael Jansson, portions of this paper previously appeared in a working version of the paper “An IV Model of Quantile Treatment Effects” (.pdf file)) forthcoming Journal of Econometrics

2.  Admissible Invariant Similar Tests for Instrumental Variables Regression” (with Victor Chernozhukov and Michael Jansson) forthcoming Econometric Theory

3.  Estimation with Many Instrumental Variables” (with Jerry Hausman and Whitney Newey, formerly “Weak Instruments, Many Instruments, and Microeconometric Practice”) forthcoming Journal of Business and Economic Statistics

4.  Plausibly Exogenous” (with Timothy Conley and Peter Rossi)

5.  Instrumental Variables Regression with Flexible Distributions” (with James McDonald and Whitney Newey) forthcoming Journal of Business and Economic Statistics

6.  A Penalty Function Approach to Bias Reduction in Nonlinear Panel Models with Fixed Effects” (with C. Alan Bester) forthcoming Journal of Business and Economic Statistics

7.  Bias Reduction for Bayesian and Frequentist Estimators” (with C. Alan Bester)

8.  Identification of Marginal Effects in a Nonparametric Correlated Random Effects Model” (with C. Alan Bester) forthcoming Journal of Business and Economic Statistics

9.  Inference with Dependent Data Using Cluster Covariance Estimators” (with C. Alan Bester and Timothy Conley)

10.  Flexible Correlated Random Effects Estimation in Panel Models with Unobserved Heterogeneity” (with C. Alan Bester)

 

Other Material

 

Technical Appendix for “Generalized Least Squares Inference in Panel and Multilevel Models with Serial Correlation and Fixed Effects” Journal of Econometrics (October 2007).

Technical Appendix for “Asymptotic Properties of a Robust Variance Matrix Estimator for Panel Data when T is Large” Journal of Econometrics (December 2007).

Working paper version of "The Reduced Form: A Simple Approach to Inference with Weak Instruments" (with Victor Chernozhukov, published as “The reduced form: A simple approach to inference with weak instruments” Economics Letters, July 2008) with additional tables and discussion excluded from published version.         

 

Abstract for Working Papers

 

 Finite-Sample Inference In Econometric Models via Quantile Restrictions”, (with Victor Chernozhukov and Michael Jansson, portions of this paper previously appeared in a working version of the paper “An IV Model of Quantile Treatment Effects” (.pdf file))

 

Under minimal assumptions finite-sample confidence bands for quantile regression models can be constructed. These confidence bands are based on the ``conditional pivotal property" of estimating equations that quantile regression methods aim to solve and will provide valid finite-sample inference for both linear and nonlinear quantile models regardless of whether the covariates are endogenous or exogenous. The confidence regions can be computed using MCMC, and confidence bounds for single parameters of interest can be computed through a simple combination of optimization and search algorithms. The methods may usefully complement existing inference methods for quantile regression when the standard assumptions fail or are suspect.

 

 

 Admissible Invariant Similar Tests for Instrumental Variables Regression” (with Victor Chernozhukov and Michael Jansson)

 

                This paper studies a model widely used in the weak instruments literature and establishes admissibility of the weighted average power likelihood ratio tests recently derived by Andrews, Moreira, and Stock (2004).  The class of tests covered by this admissibility result contains the Anderson and Rubin (1949) test.  In addition, it is shown that the tests proposed by Kleibergen (2002) and Moreira (2003), both of which are robust to weak instruments, belong to the closure of (i.e. can be interpreted as limiting cases of) the class of tests covered by our admissibility result.

 

 

Estimation with Many Instrumental Variables” (with Jerry Hausman and Whitney Newey, formerly “Weak Instruments, Many Instruments, and Microeconometric Practice”)

 

                Using many valid instrumental variables has the potential to improve efficiency but makes the usual inference procedures inaccurate.  We show that using the Bekker (1994) standard errors improves accuracy.  We obtain this finding in empirical work, simulations, and in the asymptotic theory.  Use of the Bekker (1994) standard errors in t-ratios leads to an asymptotic approximation order that is the same when the number of instrumental variables grow as when the number of instruments is fixed.  We also give a version of the Kleibergen (2002) weak instrument statistic that is robust to many or many weak instruments.

 

 

 Plausibly Exogenous” (with Timothy Conley and Peter Rossi)

 

Instrumental variables (IVs) are widely used to identify effects in models with potentially endogenous explanatory variables. In many cases, the instrument exclusion restriction that underlies the validity of the usual IV inference holds only approximately; that is, the instruments are ‘plausibly exogenous.’ We introduce a method of relaxing the exclusion restriction and performing sensitivity analysis with respect to the degree of violation. This provides practical tools for applied researchers who want to proceed with less-than-perfect instruments. We illustrate our approach with empirical examples that examine the effect of 401(k) participation upon asset accumulation, demand for margarine, and returns-to-schooling.

 

 

Instrumental Variables Regression with Flexible Distributions” (with James McDonald and Whitney Newey)

 

                Instrumental variables are often associated with low estimator precision.  This paper explores efficiency gains which might be achievable using moment conditions which are nonlinear in the disturbances using flexible parametric families for the error distributions.  We show that these estimators can achieve the semiparametric efficiency bound when the true error distribution is a member of the parametric family.  Monte Carlo simulations demonstrate low efficiency loss in the case of normal error distributions and potentially significant efficiency improvements in the case of thick-tailed and/or skewed error distributions.

 

 

A Penalty Function Approach to Bias Reduction in Nonlinear Panel Models with Fixed Effects” (with C. Alan Bester)

 

In this paper, we consider estimation of nonlinear panel data models that include individual specific fixed effects.  Estimation of these models is complicated by the incidental parameters problem: that is, noise in the estimation of the fixed effects when the time dimension is short generally results in inconsistent estimates of the common parameters due to the nonlinearity of the problem.  We present a penalty for the objective function that reduces the bias in the resulting point estimates.  The penalty function involves only cross-products of scores and the hessian matrix and so is simple to construct in practice.  The form of the penalty also provides interesting intuition into how the bias reduction is working.  We present simulation results that suggest that the penalized optimization approach may substantially reduce the bias in nonlinear fixed effects models.

 

 

Bias Reduction for Bayesian and Frequentist Estimators” (with C. Alan Bester)

 

                We show that in parametric likelihood models the first order bias in the posterior mode and the posterior mean can be removed using objective Bayesian priors.  These bias-reducing priors are defined as the solution to a set of differential equations which may not be available in closed form.  We provide a simple and tractable data dependent prior that solves the differential equations asymptotically and removes the first order bias.  When we consider the posterior mode, this approach can be interpreted as penalized maximum likelihood in a frequentist setting.  We illustrate the construction and use of the bias-reducing priors in simple examples and a simulation study.

 

 

Identification of Marginal Effects in a Nonparametric Correlated Random Effects Model” (with C. Alan Bester)

 

                In this paper, we consider identification and estimation of average marginal effects in a correlated random coefficients model without imposing strong functional form assumptions on the structural likelihood or the mixing distribution.  Identification is achieved through imposing that the mixing distribution depends on observed covariates only through an index function and the assumption that at least three time periods are available for each cross sectional unit.  We leave the functional form of the index function unrestricted subject to smoothness conditions.  We present identification results for this model and consider estimation of the marginal effects of interest.  We illustrate the use of the approach through a brief empirical example which considers the relationship between insider trading activity and trading volume.

 

 

Inference with Dependent Data Using Cluster Covariance Estimators” (with C. Alan Bester and Timothy Conley)

 

This paper presents a novel way to conduct inference using dependent data in time series, spatial, and panel data applications. Our method involves constructing t and Wald statistics utilizing a cluster covariance matrix estimator (CCE). We then use an approximation that takes the number of clusters/groups as fixed and the number of observations per group to be large and calculate limiting distributions of the t and Wald statistics. This approximation is analogous to `fixed-b' asymptotics of Kiefer and Vogelsang (2002, 2005) (KV) for heteroskedasticity and autocorrelation consistent inference, but in our case yields standard t and F distributions where the number of groups essentially plays the role of sample size. We provide simulation evidence that demonstrates our procedure outperforms conventional inference procedures and performs comparably to KV.

 

 

Flexible Correlated Random Effects Estimation in Panel Models with Unobserved Heterogeneity” (with C. Alan Bester)

 

                In this paper, we consider identification in a correlated random effects model for panel data. We assume that the likelihood for each individual in the panel is known up to a finite dimensional common parameter and an individual specific parameter. We allow the distribution of unobserved individual specific effects to depend on observed explanatory variables and make no assumptions about the particular functional form of this dependence.  This leads to a semiparametric problem where the parameters include a finite dimensional common parameter, θ and an infinite dimensional conditional density, q, that describes the distribution of unobserved individual specific effects. For a given likelihood, we establish restrictions on the space of functions H for the distribution of unobserved heterogeneity under which {θ,q} are identified. We show the model parameters may be consistently estimated by sieve maximum likelihood for a fixed panel length, T. The conditions on H, which include assumptions about the support of explanatory variables and smoothness of q in its arguments, are relatively mild and are similar to those required for nonparametric density estimation.