First-difference estimator

In statistics and econometrics, the first-difference (FD) estimator is an estimator used to address the problem of omitted variables with panel data.

It is consistent under the assumptions of the fixed effects model.

In certain situations it can be more efficient than the standard fixed effects (or "within") estimator, for example when the error terms follows a random walk.

[1] The estimator requires data on a dependent variable,

, for a set of individual units

The estimator is obtained by running a pooled ordinary least squares (OLS) estimation for a regression of

The FD estimator avoids bias due to some unobserved, time-invariant variable

, using the repeated observations over time: Differencing the equations, gives: which removes the unobserved

and eliminates the first time period.

[2][3] The FD estimator

is then obtained by using the differenced terms for

, are notation for matrices of relevant variables.

Note that the rank condition must be met for

Let and, analogously, If the error term is strictly exogenous, i.e.

, by the central limit theorem, the law of large numbers, and the Slutsky's theorem, the estimator is distributed normally with asymptotic variance of Under the assumption of homoskedasticity and no serial correlation,

, the asymptotic variance can be estimated as where

, is given by and To be unbiased, the fixed effects estimator (FE) requires strict exogeneity, defined as The first difference estimator (FD) is also unbiased under this assumption.

If strict exogeneity is violated, but the weaker assumption holds, then the FD estimator is consistent.

Note that this assumption is less restrictive than the assumption of strict exogeneity which is required for consistency using the FE estimator when

, then both FE and FD are consistent under the weaker assumption of contemporaneous exogeneity.

The Hausman test can be used to test the assumptions underlying the consistency of the FE and FD estimators.

, the FD and fixed effects estimators are numerically equivalent.

[6] Under the assumption of homoscedasticity and no serial correlation in

, the FE estimator is more efficient than the FD estimator.

This is because the FD estimator induces no serial correlation when differencing the errors.

follows a random walk, however, the FD estimator is more efficient as