LSDVC Procedure

@LSDVC is a procedure for computing bias-corrected estimators for dynamic panel models with individual effects. The main reference on the technique is Kiviet(1995). The model estimated is

\({y_{i,t}} = \gamma {y_{i,t - 1}} + {\alpha _i} + {X_{i,t}}\beta + {u_{i,t}}\)

Note that this allows for precisely one lag of the dependent variable, which you do not include in the regressor list. Also, note that you cannot apply this to one equation out of a panel VAR. The exogenous variables really need to be exogenous, not just predetermined, or they'll be subject to exactly the same type of bias as the lagged dependent variable.

@LSDVC( options ) y start end

# list of exogenous regressors only

Parameters

y	dependent variable
start, end	range to estimate, defaults to maximum range permitted by all variables involved in the regression.

Options

METHOD=[K1]/K2/K3/SIMPLE/FIXED/AH

K1, K2 and K3 are the three levels of correction proposed by Kiviet. SIMPLE is a related alternative which treats the "X'X" matrix from LSDV as fixed. FIXED is fixed effects, which is known to be inconsistent for small T. AH is Anderson-Hsiao, which is consistent but generally quite inefficient.

ITERS=# of recalculations of corrections for K1, K2, K3 or SIMPLE [1]

[PRINT]/NOPRINT

ROBUSTERRORS/[NOROBUST]

The most important option is the METHOD. The key choices there are METHOD=K1, METHOD=K2 and METHOD=K3 which are the three levels of correction given by Kiviet. Subsequent papers showed that there isn't that much to be gained in using the K2 and K3, but there's no harm in using them once they're available.

Example

compute n=94

cal(panelobs=6)

all 94//6

open data penngrow.txt

data(org=columns) / indiv year y x

* Least squares; doesn't allow for individual effects.

linreg y

# constant y{1} x{1}

* Fixed effects. Severe bias with so few data points

preg(method=fixed) y

# y{1} x{1}

* LSDV with Kiviet 3-term correction

@lsdvc(method=k3) y

# x

Sample Output

@LSDVC uses LINREG with the CREATE option to display the output. Some of the summary statistics (the Durbin-Watson in particular) should be ignored.

Linear Regression - Estimation by LSDV Corrected

Dependent Variable Y

Panel(6) of Undated Data From 1//2 To 94//6

Usable Observations 470

Degrees of Freedom 468

Skipped/Missing (from 563) 93

Mean of Dependent Variable 7.7077923723

Std Error of Dependent Variable 1.0295167233

Standard Error of Estimate 1.5323635075

Sum of Squared Residuals 1098.9285462

Durbin-Watson Statistic 0.0132

Variable Coeff Std Error T-Stat Signif

************************************************************************************

1. Y{1} 0.8008099027 0.0237986180 33.64943 0.00000000

2. X 0.1562586391 0.0194158345 8.04800 0.00000000