The PANEL Procedure
- Overview
- Getting Started
-
Syntax
-
DetailsSpecifying the Input DataSpecifying the Regression ModelUnbalanced DataMissing ValuesComputational ResourcesRestricted EstimatesNotationOne-Way Fixed-Effects ModelTwo-Way Fixed-Effects ModelBalanced PanelsUnbalanced PanelsFirst-Differenced Methods for One-Way and Two-Way ModelsBetween EstimatorsPooled EstimatorOne-Way Random-Effects ModelTwo-Way Random-Effects ModelHausman-Taylor EstimationAmemiya-MaCurdy EstimationParks Method (Autoregressive Model)Da Silva Method (Variance-Component Moving Average Model)Dynamic Panel EstimatorsLinear Hypothesis TestingHeteroscedasticity-Corrected Covariance MatricesHeteroscedasticity- and Autocorrelation-Consistent Covariance MatricesR-SquareSpecification TestsPanel Data Poolability TestPanel Data Cross-Sectional Dependence TestPanel Data Unit Root TestsLagrange Multiplier (LM) Tests for Cross-Sectional and Time EffectsTests for Serial Correlation and Cross-Sectional EffectsTroubleshootingCreating ODS GraphicsOUTPUT OUT= Data SetOUTEST= Data SetOUTTRANS= Data SetPrinted OutputODS Table Names
-
Examples
- References
LAG, CLAG, SLAG, XLAG, and ZLAG Statements
-
LAG var
(lag lag
…lag
) …var
(lag lag …lag) / OUT= SAS-data-set;
Generally, creating lags of variables in a panel setting is a tedious process requiring many DATA step statements. The PANEL procedure enables you to generate lags of any series without stepping through individual time series. The LAG statement is a data set generation tool. You can specify more than one LAG statement. Analyzing the generated lagged data requires a subsequent call to PROC PANEL.
The OUT= option is required. The output data set includes all variables in the input set, plus the genereated lags, named using the convention varname_lag. The LAG statement tends to generate many missing values in the data. This can be problematic because the number of usable observations diminishes with the lag length. Therefore, PROC PANEL offers the following alternatives to the LAG statement. The following statements can be used in place of LAG with otherwise identical syntax:
CLAG replaces missing values with the cross section mean for that variable.
SLAG replaces missing values with the time mean for that variable.
XLAG replaces missing values with the overall mean for that variable.
ZLAG replaces missing values with 0 for that variable.
For all of the above, missing values are replaced only if they are in the generated (lagged) series. Missing variables in the original variables remain unaltered.
Assume that data set A has been sorted by cross section and by time period within cross section and that the variables are
Y, X1, X2, and X3. The following PROC PANEL statements generate a series with lags 1 and 3 of the X1 variable; lags 3, 6, and 9 of the X2 variable; and lag 2 of the X3 variable.
proc panel data=A;
id i t;
lag X1(1 3) X2(3 6 9) X3(2) / out=A_lag;
run;
If you want zeroing instead of missing values, then use ZLAG in place of LAG.
proc panel data=A;
id i t;
zlag X1(1 3) X2(3 6 9) X3(2) / out=A_zlag;
run;
Similarly, you can specify XLAG to replace with overall means, SLAG to replace with time means, and CLAG to replace with cross section means.