The Hausman and Taylor (1981) model is a hybrid that combines the consistency of a fixed-effects model with the efficiency and applicability of a random-effects model. One-way random-effects models assume exogeneity of the regressors, namely that they be independent of both the cross-sectional and observation-level errors. In cases where some regressors are correlated with the cross-sectional errors, the random effects model can be adjusted to deal with the endogeneity.
Consider the one-way model:
The regressors are subdivided so that the variables vary within cross sections whereas the
variables do not and would otherwise be dropped from a fixed-effects model. The subscript 1 denotes variables that are independent
of both error terms (exogenous variables), and the subscript 2 denotes variables that are independent of the observation-level
errors
but correlated with cross-sectional errors
(endogenous variables). The intercept term (if your model has one) is included as part of
in what follows.
The Hausman-Taylor estimator is an instrumental variables regression on data that are weighted similarly to data for random-effects estimation. In both cases, the weights are functions of the estimated variance components.
Begin with and
. The mean transformation vector is
and the deviations from the mean transform is
, where
is a square matrix of ones of dimension
.
The observation-level variance is estimated from a standard fixed-effects model fit. For ,
, and
, let
To estimate the cross-sectional error variance, form the mean residuals . You can use the mean residuals to obtain intermediate estimates of the coefficients for
and
via two-stage least squares (2SLS) estimation. At the first stage, use
and
as instrumental variables to predict
. At the second stage, regress
on both
and the predicted
to obtain
and
.
To estimate the cross-sectional variance, form , with
and
After variance-component estimation, transform the dependent variable into partial deviations: . Likewise, transform the regressors to form
,
,
, and
. The partial weights
are determined by
, with
.
Finally, you obtain the Hausman-Taylor estimates by performing 2SLS regression of on
,
,
, and
. For the first-stage regression, use the following instruments:
, the deviations from cross-sectional means for all time-varying variables
, for the ith cross section during time period t
, where
are the means of the time-varying exogenous variables for the ith cross section
Multiplication by the factor is redundant in balanced data, but necessary in the unbalanced case to produce accurate instrumentation; see Gardner (1998).
Let equal the number of regressors in
, and
equal the number of regressors in
. Then the Hausman-Taylor model is identified only if
; otherwise, no estimation will take place.
Hausman and Taylor (1981) describe a specification test that compares their model to fixed effects. For a null hypothesis of fixed effects, Hausman’s
m statistic is calculated by comparing the parameter estimates and variance matrices for both models, identically to how it
is calculated for one-way random effects models; for more information, see the section Specification Tests. The degrees of freedom of the test, however, are not based on matrix rank but instead are equal to .