** Chapter 9.3 ** for W.H. Greene, Econometric Analysis 6th ed. *****************
* (c) Noel Roy 2008
*
*				               THE POOLED REGRESSION MODEL
*
*===============================================================================
* Example 9.1 Wage Equation
*
READ (TableF9-1.txt) /NAMES
* Add Exp Squared to the list of regressors.
GENR exp2=Exp**2
* Estimate the model by OLS with OLS standard errors.
* The dependent variable is in LOG form. The LOGLIN option corrects the reported
* elasticities and the value of the log-likelihood to account for this.
*
OLS lwage exp exp2 wks occ ind south smsa ms fem union ed blk /LOGLIN
*
* Resestimate using the HETCOV option to obtain the White heteroskedaticity
* consistent standard errors.
*
OLS lwage exp exp2 wks occ ind south smsa ms fem union ed blk/ HETCOV LOGLIN
*
* The POOL command provides features for estimating pooled cross-section
* time-series models. The NCROSS= option can  be used to specify the number
* of cross-sectional units in the data. Alternatively, with balanced panels
* such as this one, the NTIME= option can be used to specify the number of 
* time units in the data. The Beck-Katz robust standard errors (9-3) can be 
* obtained by using the HETCOV option.
*
PAR 4500
* In some implementations, the default memory allocation is insufficient to
* estimate a pooled regression model of this size. The memory can be increased with 
* the PAR command, which specifies the memory in kilobytes.
*
POOL lwage exp exp2 wks occ ind south smsa ms fem union ed blk/ NTIME=7 HETCOV LOGLIN
*
* For reasons that are not clear, the panel robust standard errors do not agree with 
* those reported in Table 9.1.
STOP
*===============================================================================
* Updated September 22, 2008