用Frontier4.1做一个SFA的分析,最终输出结果待估参数和t值是一样的,标准差都等于1。除此之外,每个截面每一年的效率值都是一样的。请高手指教是可能是什么原因引起的。不甚感激!
程序的命令:
1 1=ERROR COMPONENTS MODEL, 2=TE EFFECTS MODEL
CHN3-dta.txt DATA FILE NAME
CHN3-out.txt OUTPUT FILE NAME
1 1=PRODUCTION FUNCTION, 2=COST FUNCTION
y LOGGED DEPENDENT VARIABLE (Y/N)
28 NUMBER OF CROSS-SECTIONS
34 NUMBER OF TIME PERIODS
952 NUMBER OF OBSERVATIONS IN TOTAL
9 NUMBER OF REGRESSOR VARIABLES (Xs)
n MU (Y/N) [OR DELTA0 (Y/N) IF USING TE EFFECTS MODEL]
y ETA (Y/N) [OR NUMBER OF TE EFFECTS REGRESSORS (Zs)]
n STARTING VALUES (Y/N)
IF YES THEN BETA0
BETA1 TO
BETAK
SIGMA SQUARED
GAMMA
MU [OR DELTA0
ETA DELTA1 TO
DELTAP]
NOTE: IF YOU ARE SUPPLYING STARTING VALUES
AND YOU HAVE
得出的结果:
Output from the program FRONTIER (Version 4.1c)
instruction file = CHN3-INS.TXT
data file = CHN3-dta.txt
Error Components Frontier (see B&C 1992)
The model is a production function
The dependent variable is logged
the ols estimates are :
coefficient standard-error t-ratio
beta 0 -0.44334839E+01 0.54351746E+00 -0.81570220E+01
beta 1 0.20789932E+01 0.17172396E+00 0.12106599E+02
beta 2 0.33481947E+00 0.21848595E+00 0.15324531E+01
beta 3 -0.17703306E+00 0.18704136E-01 -0.94649153E+01
beta 4 -0.14215624E-01 0.55654005E-01 -0.25542859E+00
beta 5 0.91153448E-01 0.41344420E-01 0.22047340E+01
beta 6 -0.15915487E+00 0.31863504E-01 -0.49948955E+01
beta 7 0.44863861E-02 0.95386696E-03 0.47033667E+01
beta 8 -0.11908538E-01 0.69761053E-02 -0.17070468E+01
beta 9 0.23845368E-01 0.40546349E-02 0.58810148E+01
sigma-squared 0.81916965E-01
log likelihood function = -0.15482761E+03
the estimates after the grid search were :
beta 0 -0.41243141E+01
beta 1 0.20789932E+01
beta 2 0.33481947E+00
beta 3 -0.17703306E+00
beta 4 -0.14215624E-01
beta 5 0.91153448E-01
beta 6 -0.15915487E+00
beta 7 0.44863861E-02
beta 8 -0.11908538E-01
beta 9 0.23845368E-01
sigma-squared 0.17664248E+00
gamma 0.85000000E+00
mu is restricted to be zero
eta 0.00000000E+00
iteration = 0 func evals = 20 llf = 0.26294767E+03
-0.41243141E+01 0.20789932E+01 0.33481947E+00-0.17703306E+00-0.14215624E-01
0.91153448E-01-0.15915487E+00 0.44863861E-02-0.11908538E-01 0.23845368E-01
0.17664248E+00 0.85000000E+00 0.00000000E+00
gradient step
pt better than entering pt cannot be found
iteration = 1 func evals = 28 llf = 0.26294767E+03
-0.41243141E+01 0.20789932E+01 0.33481947E+00-0.17703306E+00-0.14215624E-01
0.91153448E-01-0.15915487E+00 0.44863861E-02-0.11908538E-01 0.23845368E-01
0.17664248E+00 0.85000000E+00 0.00000000E+00
the final mle estimates are :
coefficient standard-error t-ratio
beta 0 -0.41243141E+01 0.10000000E+01 -0.41243141E+01
beta 1 0.20789932E+01 0.10000000E+01 0.20789932E+01
beta 2 0.33481947E+00 0.10000000E+01 0.33481947E+00
beta 3 -0.17703306E+00 0.10000000E+01 -0.17703306E+00
beta 4 -0.14215624E-01 0.10000000E+01 -0.14215624E-01
beta 5 0.91153448E-01 0.10000000E+01 0.91153448E-01
beta 6 -0.15915487E+00 0.10000000E+01 -0.15915487E+00
beta 7 0.44863861E-02 0.10000000E+01 0.44863861E-02
beta 8 -0.11908538E-01 0.10000000E+01 -0.11908538E-01
beta 9 0.23845368E-01 0.10000000E+01 0.23845368E-01
sigma-squared 0.17664248E+00 0.10000000E+01 0.17664248E+00
gamma 0.85000000E+00 0.10000000E+01 0.85000000E+00
mu is restricted to be zero
eta 0.00000000E+00 0.10000000E+01 0.00000000E+00
log likelihood function = 0.26294767E+03
LR test of the one-sided error = 0.83555056E+03
with number of restrictions = 2
[note that this statistic has a mixed chi-square distribution]
number of iterations = 1
(maximum number of iterations set at : 100)
number of cross-sections = 28
number of time periods = 34
total number of observations = 952
thus there are: 0 obsns not in the panel
效率值,我把第一年和最后一年的贴出来吧。
第一年的:
efficiency estimates for year 1 :
firm eff.-est.
1 0.65250098E+00
2 0.67490256E+00
3 0.74122366E+00
4 0.63577046E+00
5 0.66461441E+00
6 0.99554976E+00
7 0.83006209E+00
8 0.10000000E+01
9 0.87943593E+00
10 0.69933649E+00
11 0.61799536E+00
12 0.73787988E+00
13 0.73335899E+00
14 0.54311324E+00
15 0.74745797E+00
16 0.62938970E+00
17 0.75355811E+00
18 0.97454013E+00
19 0.73481618E+00
20 0.76184131E+00
21 0.80715852E+00
22 0.53357264E+00
23 0.60501712E+00
24 0.58416434E+00
25 0.53944350E+00
26 0.86675913E+00
27 0.65811517E+00
28 0.65977248E+00
mean eff. in year 1 = 0.72361965E+00
最后一年的:
fficiency estimates for year 34 :
firm eff.-est.
1 0.65250098E+00
2 0.67490256E+00
3 0.74122366E+00
4 0.63577046E+00
5 0.66461441E+00
6 0.99554976E+00
7 0.83006209E+00
8 0.10000000E+01
9 0.87943593E+00
10 0.69933649E+00
11 0.61799536E+00
12 0.73787988E+00
13 0.73335899E+00
14 0.54311324E+00
15 0.74745797E+00
16 0.62938970E+00
17 0.75355811E+00
18 0.97454013E+00
19 0.73481618E+00
20 0.76184131E+00
21 0.80715852E+00
22 0.53357264E+00
23 0.60501712E+00
24 0.58416434E+00
25 0.53944350E+00
26 0.86675913E+00
27 0.65811517E+00
28 0.65977248E+00
mean eff. in year 34 = 0.72361965E+00