In: Statistics and Probability
Why do the two-stage least squares estimators give biased but consistent estimators?
When the explanatory variables in the regression model are correlated with stochastic disturbance terms,the resulting estimates are biased as well as inconsistent.
In such a situation ,two stage least squares is an estimation procedure.
The basic idea behind 2SLS is to replace the (stochastic)
endogenous explanatory variable by a
linear combination of the predetermined variables in the model and
use this combination as the explanatory variable in lieu of the
original endogenous variable.
In this technique the estimators are biased due to the violation of assumption of Linear model...but as the sample size tends to infinity estimators will be consistent.
The estimates obained are consistent, that is, as the sample size increases indefinitely, estimates converge to their true population values. The estimates may not
satisfy small-sample properties, such as unbiasedness and
minimum variance. Therefore, the results obtained by applying these
methods to small
samples and the inferences drawn from them should be interpreted
with due caution.
The statistical justification of the 2SLS is of the large sample
type. When there are no lagged endogenous variables, the 2SLS
coefficient estimators are consis-
tent if the exogenous variables are constant in repeated samples
and if the disturbance [appearing in the various behavioral or
structural equations] . . . are independently and identically
distributed with zero means and finite variances. . . .
If these two conditions are satisfied, the sampling distribution of
2SLS coefficient estimators becomes approximately normal for large
samples. When the equation system contains lagged endogenous
variables, the consistency and large-sample normality of the 2SLS
coefficient estimators require additional condition, . that as the
sample increases the mean square of the values taken by each lagged
endogenous variable converges in probability to a positive limit.
If the [disturbances appearing in the various structural equations
are] not independently distributed, lagged endogenous variables are
not independent of the current operation of the equation system ,
which means these variables are not really predetermined. If these
variables are nevertheless treated as predetermined in the 2SLS
procedure, the resulting estimators are not consistent.