--------------------------------------------------------------------------------
log: /Users/baum/Desktop/To do/EC327_S2008/327fdfe.smcl
log type: smcl
opened on: 22 Jan 2008, 17:10:55
. use http://fmwww.bc.edu/cfb/boyan/udcl_irrc_3, clear
. keep if yeara>2004
(21755 observations deleted)
. xtset
panel variable: firmid (unbalanced)
time variable: yeara, 2005 to 2006
. * make into a balanced panel for these two years with complete data
. qui reg data18 data128 data6 data12
. keep if e(sample)
(231 observations deleted)
. bysort firmid: g ok = _N==2
. keep if ok
(479 observations deleted)
. tab yeara
fiscal year | Freq. Percent Cum.
------------+-----------------------------------
2005 | 1,320 50.00 50.00
2006 | 1,320 50.00 100.00
------------+-----------------------------------
Total | 2,640 100.00
. * first difference estimation (ssc install xtivreg2)
. xtivreg2 data18 data128 data6 data12 if ok, fd noconst
FIRST DIFFERENCES ESTIMATION
----------------------------
Number of groups = 1320 Obs per group: min = 1
avg = 1.0
max = 1
OLS estimation
--------------
Estimates efficient for homoskedasticity only
Statistics consistent for homoskedasticity only
Number of obs = 1320
F( 3, 1317) = 186.12
Prob > F = 0.0000
Total (centered) SS = 525542940.1 Centered R2 = 0.2903
Total (uncentered) SS = 531141338.7 Uncentered R2 = 0.2977
Residual SS = 373001157 Root MSE = 531.6
------------------------------------------------------------------------------
D.data18 | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
data12 |
D1. | .1330804 .0073978 17.99 0.000 .118581 .1475799
data6 |
D1. | -.0081666 .0009866 -8.28 0.000 -.0101004 -.0062329
data128 |
D1. | .2658948 .0556797 4.78 0.000 .1567645 .375025
------------------------------------------------------------------------------
Included instruments: D.data12 D.data6 D.data128
------------------------------------------------------------------------------
. * fixed effect estimation
. xtreg data18 data128 data6 data12 if ok, fe
Fixed-effects (within) regression Number of obs = 2640
Group variable (i): firmid Number of groups = 1320
R-sq: within = 0.2977 Obs per group: min = 2
between = 0.4813 avg = 2.0
overall = 0.4735 max = 2
F(3,1317) = 186.12
corr(u_i, Xb) = -0.7398 Prob > F = 0.0000
------------------------------------------------------------------------------
data18 | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
data128 | .2658948 .0557431 4.77 0.000 .1565398 .3752498
data6 | -.0081666 .0009877 -8.27 0.000 -.0101043 -.0062289
data12 | .1330804 .0074062 17.97 0.000 .1185511 .1476097
_cons | -381.0436 40.63884 -9.38 0.000 -460.7675 -301.3196
-------------+----------------------------------------------------------------
sigma_u | 1967.8384
sigma_e | 376.31125
rho | .96472092 (fraction of variance due to u_i)
------------------------------------------------------------------------------
F test that all u_i=0: F(1319, 1317) = 18.75 Prob > F = 0.0000
.
. * repeat for three years
. use http://fmwww.bc.edu/cfb/boyan/udcl_irrc_3, clear
. keep if yeara>2003
(19915 observations deleted)
. xtset
panel variable: firmid (unbalanced)
time variable: yeara, 2004 to 2006
. * make into a balanced panel for these two years with complete data
. qui reg data18 data128 data6 data12
. keep if e(sample)
(344 observations deleted)
. bysort firmid: g ok = _N==3
. keep if ok
(889 observations deleted)
. tab yeara
fiscal year | Freq. Percent Cum.
------------+-----------------------------------
2004 | 1,319 33.33 33.33
2005 | 1,319 33.33 66.67
2006 | 1,319 33.33 100.00
------------+-----------------------------------
Total | 3,957 100.00
. * first difference estimation (ssc install xtivreg2)
. xtivreg2 data18 data128 data6 data12 if ok, fd noconst
FIRST DIFFERENCES ESTIMATION
----------------------------
Number of groups = 1319 Obs per group: min = 2
avg = 2.0
max = 2
OLS estimation
--------------
Estimates efficient for homoskedasticity only
Statistics consistent for homoskedasticity only
Number of obs = 2638
F( 3, 2635) = 350.91
Prob > F = 0.0000
Total (centered) SS = 1033457052 Centered R2 = 0.2772
Total (uncentered) SS = 1045356767 Uncentered R2 = 0.2855
Residual SS = 746942589.9 Root MSE = 532.1
------------------------------------------------------------------------------
D.data18 | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
data12 |
D1. | .1224888 .00452 27.10 0.000 .1136297 .1313478
data6 |
D1. | -.0072491 .000863 -8.40 0.000 -.0089405 -.0055577
data128 |
D1. | .0732101 .0457331 1.60 0.109 -.0164251 .1628452
------------------------------------------------------------------------------
Included instruments: D.data12 D.data6 D.data128
------------------------------------------------------------------------------
. * fixed effect estimation
. xtreg data18 data128 data6 data12 if ok, fe
Fixed-effects (within) regression Number of obs = 3957
Group variable (i): firmid Number of groups = 1319
R-sq: within = 0.3223 Obs per group: min = 3
between = 0.5826 avg = 3.0
overall = 0.5650 max = 3
F(3,2635) = 417.72
corr(u_i, Xb) = -0.5965 Prob > F = 0.0000
------------------------------------------------------------------------------
data18 | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
data128 | .0983172 .0415973 2.36 0.018 .0167506 .1798838
data6 | -.0006023 .0007978 -0.76 0.450 -.0021666 .000962
data12 | .1050727 .0042303 24.84 0.000 .0967776 .1133678
_cons | -238.4987 21.01925 -11.35 0.000 -279.7146 -197.2828
-------------+----------------------------------------------------------------
sigma_u | 1355.899
sigma_e | 415.58627
rho | .91412382 (fraction of variance due to u_i)
------------------------------------------------------------------------------
F test that all u_i=0: F(1318, 2635) = 17.90 Prob > F = 0.0000
.
. log close
log: /Users/baum/Desktop/To do/EC327_S2008/327fdfe.smcl
log type: smcl
closed on: 22 Jan 2008, 17:10:57
--------------------------------------------------------------------------------