Model use data: EmpLuk and psi Method: Tool Variables, GMM and FGLS. Code copy from Econometric use R, author Nguyen Chi Dung. This paper has many errors.
library(tidyverse)## Loading tidyverse: ggplot2
## Loading tidyverse: tibble
## Loading tidyverse: tidyr
## Loading tidyverse: readr
## Loading tidyverse: purrr
## Loading tidyverse: dplyr## Conflicts with tidy packages ----------------------------------------------## filter(): dplyr, stats
## lag(): dplyr, statslibrary(plm)## Loading required package: Formula##
## Attaching package: 'plm'## The following objects are masked from 'package:dplyr':
##
## between, lag, leaddata("EmplUK")
head(EmplUK)## firm year sector emp wage capital output
## 1 1 1977 7 5.041 13.1516 0.5894 95.7072
## 2 1 1978 7 5.600 12.3018 0.6318 97.3569
## 3 1 1979 7 5.015 12.8395 0.6771 99.6083
## 4 1 1980 7 4.715 13.8039 0.6171 100.5501
## 5 1 1981 7 4.093 14.2897 0.5076 99.5581
## 6 1 1982 7 3.166 14.8681 0.4229 98.6151trang = table(EmplUK$firm)
View(trang)Can see unbalance data, every firms were studied with other numbers of year, maximum is 9, so set ylim = 10 in barplot of firms.
Visualization:
barplot(trang, ylim = c(0, 10)) 
mydata = data.frame(trang)
table(mydata$Freq) ##
## 7 8 9
## 103 23 14fyear = table(EmplUK$year)
library(fBasics)## Loading required package: timeDate## Loading required package: timeSeries## ## Rmetrics Package fBasics## Analysing Markets and calculating Basic Statistics## Copyright (C) 2005-2014 Rmetrics Association Zurich## Educational Software for Financial Engineering and Computational Science## Rmetrics is free software and comes with ABSOLUTELY NO WARRANTY.## https://www.rmetrics.org --- Mail to: info@rmetrics.orgbarplot(fyear, col = qualiPalette(12, "Set3"))
- Tool_var: lag(wage, 1) is tool variable of wage:
ivPanel = EmplUK %>% plm(emp ~ wage + capital|lag(wage, 1) + capital,
data = .,
inst.method = "bvk")
summary(ivPanel)## Oneway (individual) effect Within Model
## Instrumental variable estimation
## (Balestra-Varadharajan-Krishnakumar's transformation)
##
## Call:
## plm(formula = emp ~ wage + capital | lag(wage, 1) + capital,
## data = ., inst.method = "bvk")
##
## Unbalanced Panel: n=140, T=6-8, N=891
##
## Residuals :
## Min. 1st Qu. Median 3rd Qu. Max.
## -15.3000607 -0.3058983 0.0015485 0.3278123 19.6057783
##
## Coefficients :
## Estimate Std. Error t-value Pr(>|t|)
## wage -0.228807 0.078561 -2.9125 0.003692 **
## capital 1.026579 0.065893 15.5794 < 2.2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Total Sum of Squares: 3832
## Residual Sum of Squares: 2729.2
## R-Squared: 0.28796
## Adj. R-Squared: 0.15392
## F-statistic: 151.322 on 2 and 749 DF, p-value: < 2.22e-16ms is tool variable of wks:
setwd("D:/TKT/RImportData")
library(foreign)
trang = read.dta("psi.dta")
table(trang$t)##
## 1 2 3 4 5 6 7
## 595 595 595 595 595 595 595#Thiet lap du lieu mang
trang = plm.data(trang, index = c("id", "t"))
ivPanel = trang %>% plm(lwage ~ exp + exp2 + wks|exp + exp2 + ms,
data = .,
model = "within",
inst.method = "bvk")
summary(ivPanel)## Oneway (individual) effect Within Model
## Instrumental variable estimation
## (Balestra-Varadharajan-Krishnakumar's transformation)
##
## Call:
## plm(formula = lwage ~ exp + exp2 + wks | exp + exp2 + ms, data = .,
## model = "within", inst.method = "bvk")
##
## Balanced Panel: n=595, T=7, N=4165
##
## Residuals :
## Min. 1st Qu. Median 3rd Qu. Max.
## -4.114922 -0.119156 0.039161 0.198531 2.160009
##
## Coefficients :
## Estimate Std. Error t-value Pr(>|t|)
## exp 0.1408101 0.0547014 2.5742 0.01009 *
## exp2 -0.0011207 0.0014052 -0.7975 0.42520
## wks -0.1149742 0.2316926 -0.4962 0.61976
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Total Sum of Squares: 240.65
## Residual Sum of Squares: 946.63
## R-Squared: 0.094827
## Adj. R-Squared: -0.056669
## F-statistic: -886.733 on 3 and 3567 DF, p-value: 1- Hausman - Taylor Estimation Method
formul = lwage ~ occ + south + smsa + ind +
exp + I(exp^2) + wks + ms + union + blk + fem +
ed|exp + I(exp^2) + wks + ms + union + ed
hausmantaylor = trang %>% plm(formul, data = ., model = "ht")
summary(hausmantaylor)## Oneway (individual) effect Hausman-Taylor Model
## Call:
## pht(formula = formul, data = .)
##
## T.V. exo : exp, I(exp^2), wks, ms, union
## T.V. endo : occ, south, smsa, ind
## T.I. exo : ed
## T.I. endo : blk, fem
##
## Balanced Panel: n=595, T=7, N=4165
##
## Effects:
## var std.dev share
## idiosyncratic 0.02304 0.15180 0
## individual 58.38642 7.64110 1
## theta: 0.9925
##
## Residuals :
## Min. 1st Qu. Median 3rd Qu. Max.
## -1.8254038 -0.0643299 -0.0021546 0.0669370 1.9446714
##
## Coefficients :
## Estimate Std. Error t-value Pr(>|t|)
## (Intercept) 8.0007e+00 3.4251e+00 2.3359 0.01950 *
## occ -2.1452e-02 1.3643e-02 -1.5724 0.11587
## south -2.1349e-03 3.3949e-02 -0.0629 0.94986
## smsa -4.2543e-02 1.9230e-02 -2.2123 0.02694 *
## ind 1.9173e-02 1.5289e-02 1.2541 0.20982
## exp 1.1315e-01 2.4454e-03 46.2703 < 2.2e-16 ***
## I(exp^2) -4.1901e-04 5.4028e-05 -7.7553 8.812e-15 ***
## wks 8.3480e-04 5.9354e-04 1.4065 0.15958
## ms -2.9846e-02 1.8790e-02 -1.5884 0.11219
## union 3.3223e-02 1.4769e-02 2.2495 0.02448 *
## blk -2.7984e+01 1.6539e+01 -1.6920 0.09064 .
## fem 6.7401e+00 4.2331e+00 1.5922 0.11133
## ed -1.6246e-01 2.1432e-01 -0.7580 0.44843
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Total Sum of Squares: 886.9
## Residual Sum of Squares: 93.974
## F-statistic: 2919.47 on 12 and 4152 DF, p-value: < 2.22e-16- GMM method
gmmPanel = EmplUK %>% pgmm(log(emp) ~ lag(log(emp), 1) + lag(log(emp), 2)
+ log(wage)+ lag(log(wage), 1)
+ log(capital)
+ log(output)
+ lag(log(output), 1)|lag(log(emp), c(2:99)),
data = .,
effect = "twoways",
model = "twosteps")
summary(gmmPanel)## Twoways effects Two steps model
##
## Call:
## pgmm(formula = log(emp) ~ lag(log(emp), 1) + lag(log(emp), 2) +
## log(wage) + lag(log(wage), 1) + log(capital) + log(output) +
## lag(log(output), 1) | lag(log(emp), c(2:99)), data = ., effect = "twoways",
## model = "twosteps")
##
## Unbalanced Panel: n=140, T=7-9, N=1031
##
## Number of Observations Used: 611
##
## Residuals
## Min. 1st Qu. Median Mean 3rd Qu. Max.
## -0.6190677 -0.0255683 0.0000000 -0.0001339 0.0332013 0.6410272
##
## Coefficients
## Estimate Std. Error z-value Pr(>|z|)
## lag(log(emp), 1) 0.474151 0.185398 2.5575 0.0105437 *
## lag(log(emp), 2) -0.052967 0.051749 -1.0235 0.3060506
## log(wage) -0.513205 0.145565 -3.5256 0.0004225 ***
## lag(log(wage), 1) 0.224640 0.141950 1.5825 0.1135279
## log(capital) 0.292723 0.062627 4.6741 2.953e-06 ***
## log(output) 0.609775 0.156263 3.9022 9.530e-05 ***
## lag(log(output), 1) -0.446373 0.217302 -2.0542 0.0399605 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Sargan Test: chisq(25) = 30.11247 (p.value=0.22011)
## Autocorrelation test (1): normal = -1.53845 (p.value=0.12394)
## Autocorrelation test (2): normal = -0.2796829 (p.value=0.77972)
## Wald test for coefficients: chisq(7) = 142.0353 (p.value=< 2.22e-16)
## Wald test for time dummies: chisq(6) = 16.97046 (p.value=0.0093924)- GFLS Estimation Method
fgls = EmplUK %>% pggls(log(emp) ~ log(wage) + log(capital), data = .,
model = "pooling")
summary(fgls)## NA
##
## Call:
## pggls(formula = log(emp) ~ log(wage) + log(capital), data = .,
## model = "pooling")
##
## Unbalanced Panel: n=140, T=7-9, N=1031
##
## Residuals
## Min. 1st Qu. Median Mean 3rd Qu. Max.
## -1.80696 -0.36552 0.06181 0.03230 0.44279 1.58719
##
## Coefficients
## Estimate Std. Error z-value Pr(>|z|)
## (Intercept) 2.023480 0.158468 12.7690 < 2.2e-16 ***
## log(wage) -0.232329 0.048001 -4.8401 1.298e-06 ***
## log(capital) 0.610484 0.017434 35.0174 < 2.2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## Total Sum of Squares: 1853.6
## Residual Sum of Squares: 402.55
## Multiple R-squared: 0.78283Thank you!!