Digital tech and employment

Preamble

The purpose of this exercise is to examine the changes in the structure and evolution of employment as affected by the revolution in digital technology.

Data

The dataset used (the World Indicators of Skills for Employment or WISE panel) is drawn from the OECD data repository:

https://stats.oecd.org

## ── Attaching core tidyverse packages ──────────────────────── tidyverse 2.0.0 ──
## ✔ dplyr     1.1.1     ✔ readr     2.1.4
## ✔ forcats   1.0.0     ✔ stringr   1.5.0
## ✔ ggplot2   3.4.1     ✔ tibble    3.2.1
## ✔ lubridate 1.9.2     ✔ tidyr     1.3.0
## ✔ purrr     1.0.1     
## ── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──
## ✖ dplyr::filter() masks stats::filter()
## ✖ dplyr::lag()    masks stats::lag()
## ℹ Use the ]8;;http://conflicted.r-lib.org/conflicted package]8;; to force all conflicts to become errors

wise <- read_csv("Downloads/WSDB_05042023191139959.csv")

## Rows: 112888 Columns: 17
## ── Column specification ────────────────────────────────────────────────────────
## Delimiter: ","
## chr (11): LOCATION, Country, VAR, Variable, SEX, Sex, Unit Code, Unit, Power...
## dbl  (4): TIME, Year, PowerCode Code, Value
## lgl  (2): Reference Period Code, Reference Period
## 
## ℹ Use `spec()` to retrieve the full column specification for this data.
## ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.

wise_T <- wise %>% select(1, 3, 5, 7, 15) %>% 
  arrange(VAR) %>% 
  filter(SEX=="T") %>% 
  spread(VAR, Value) %>% 
  select(-2)
head(wise_T)

## # A tibble: 6 × 34
##   LOCATION  TIME ADU_EMPL ADU_EMPL_HS ADU_EMPL_LS ADU_EMPL_MS ADU_UNEMP
##   <chr>    <dbl>    <dbl>       <dbl>       <dbl>       <dbl>     <dbl>
## 1 ABW       1990       NA          NA          NA          NA      NA  
## 2 ABW       1991       NA          NA          NA          NA       5.1
## 3 ABW       1992       NA          NA          NA          NA      NA  
## 4 ABW       1993       NA          NA          NA          NA      NA  
## 5 ABW       1994       NA          NA          NA          NA       6  
## 6 ABW       1995       NA          NA          NA          NA      NA  
## # ℹ 27 more variables: ADU_UNEMP_HS <dbl>, ADU_UNEMP_LS <dbl>,
## #   ADU_UNEMP_MS <dbl>, BG_GSR_NFSV_GD_ZS <dbl>, EMPL <dbl>, EMPL_HS <dbl>,
## #   EMPL_LS <dbl>, EMPL_MS <dbl>, GDP_EMPL <dbl>, GDP_HOUR <dbl>,
## #   HUM_DEV_IND <dbl>, IT_CEL_SETS_P2 <dbl>, IT_NET_USER_P2 <dbl>,
## #   NY_GDP_MKTP_CD <dbl>, SELF_EMPL <dbl>, SI_POV_2DAY <dbl>,
## #   SI_POV_GINI <dbl>, SKIL_GAP <dbl>, SP_POP_TOTL <dbl>,
## #   SP_POP_TOTL_FE_ZS <dbl>, SP_URB_TOTL_IN_ZS <dbl>, UNEMP <dbl>, …

The set includes the following variables:

WSDB_codebook_variable <- wise %>%
  distinct(VAR, Variable) %>%
  select(VAR, Variable) %>% 
  arrange(VAR)

The countries covered are:

WSDB_codebook_country <- wise %>%
  distinct(LOCATION, Country) %>%
  select(LOCATION, Country)

The set covers from 1990 to 2014 (25 years). Some of the variables contain data by sex (M: male, F: female, T: total), but only the total is used.

Frameworks

To start, a few panel models are estimated (with country and year fixed effects). The dependent variables of interest are:

Employment rate,
Adult employment rate,
Employment rate; high skilled, and
Employment rate; low skilled.

The main regressor of interest is a proxy for the extent of digital technology use, either:

Access to mobile phones or
Access to Internet

The control variables are:

Labour productivity (GDP/worker)
Poverty incidence
Population
Public expenditure on education
Share of urban population
Share of female population

## 
## Attaching package: 'plm'

## The following objects are masked from 'package:dplyr':
## 
##     between, lag, lead

wise_T_plm <- pdata.frame(wise_T, index = c("LOCATION", "TIME"))
wise_d <- as.data.frame( table(index(wise_T_plm)) ) %>% 
  arrange(TIME)

Models

Scatterplots

wise_T %>% 
  filter( TIME>=1995 ) %>% 
  ggplot( aes(x = IT_CEL_SETS_P2, y = EMPL) ) +
  geom_point(color="darkgray") +
  facet_wrap(~ TIME) +
  labs(title = "Employment rate vs access to mobile phones", x = "(%)", y = "(%)") +
  geom_smooth(method = "lm", se = FALSE, color="black", lwd=.5)

## `geom_smooth()` using formula = 'y ~ x'

## Warning: Removed 2611 rows containing non-finite values (`stat_smooth()`).

## Warning: Removed 2611 rows containing missing values (`geom_point()`).

wise_T %>% 
  filter( TIME>=1995) %>% 
  ggplot( aes(x = IT_CEL_SETS_P2, y = ADU_EMPL ) ) +
  geom_point(color="darkgray") +
  facet_wrap(~ TIME) +
  labs(title = "Adult employment rate vs access to mobile phones", x = "(%)", y = "(%)") +
  geom_smooth(method = "lm", se = FALSE, color="black", lwd=.5)

## `geom_smooth()` using formula = 'y ~ x'

## Warning: Removed 3364 rows containing non-finite values (`stat_smooth()`).

## Warning: Removed 3364 rows containing missing values (`geom_point()`).

wise_T %>% 
  filter( TIME>=1995 ) %>% 
  ggplot( aes(x = IT_CEL_SETS_P2, y = EMPL_HS ) ) +
  geom_point(color="darkgray") +
  facet_wrap(~ TIME) +
  labs(title = "Employment rate (high skilled) vs access to mobile phones", x = "(%)", y = "(%)") +
  geom_smooth(method = "lm", se = FALSE, color="black", lwd=.5)

## `geom_smooth()` using formula = 'y ~ x'

## Warning: Removed 3785 rows containing non-finite values (`stat_smooth()`).

## Warning: Removed 3785 rows containing missing values (`geom_point()`).

wise_T %>% 
  filter( TIME>=1995 ) %>% 
  ggplot( aes(x = IT_NET_USER_P2, y = EMPL) ) +
  geom_point(color="darkgray") +
  facet_wrap(~ TIME) +
  labs(title = "Employment rate vs access to Internet", x = "(%)", y = "(%)") +
  geom_smooth(method = "lm", se = FALSE, color="black", lwd=.5)

## `geom_smooth()` using formula = 'y ~ x'

## Warning: Removed 2622 rows containing non-finite values (`stat_smooth()`).

## Warning: Removed 2622 rows containing missing values (`geom_point()`).

wise_T %>% 
  filter( TIME>=1995) %>% 
  ggplot( aes(x = IT_NET_USER_P2, y = ADU_EMPL ) ) +
  geom_point(color="darkgray") +
  facet_wrap(~ TIME) +
  labs(title = "Adult employment rate vs access to Internet", x = "(%)", y = "(%)") +
  geom_smooth(method = "lm", se = FALSE, color="black", lwd=.5)

## `geom_smooth()` using formula = 'y ~ x'

## Warning: Removed 3364 rows containing non-finite values (`stat_smooth()`).

## Warning: Removed 3364 rows containing missing values (`geom_point()`).

wise_T %>% 
  filter( TIME>=1995 ) %>% 
  ggplot( aes(x = IT_NET_USER_P2, y = EMPL_HS ) ) +
  geom_point(color="darkgray") +
  facet_wrap(~ TIME) +
  labs(title = "Employment rate (high skilled) vs access to Internet", x = "(%)", y = "(%)") +
  geom_smooth(method = "lm", se = FALSE, color="black", lwd=.5)

## `geom_smooth()` using formula = 'y ~ x'

## Warning: Removed 3785 rows containing non-finite values (`stat_smooth()`).

## Warning: Removed 3785 rows containing missing values (`geom_point()`).

Estimates: two regressors

fe_model1_c0 <- plm(EMPL ~ IT_CEL_SETS_P2 + GDP_EMPL, data = wise_T_plm, effect = "individual", model = "within")
summary(fe_model1_c0)

## Oneway (individual) effect Within Model
## 
## Call:
## plm(formula = EMPL ~ IT_CEL_SETS_P2 + GDP_EMPL, data = wise_T_plm, 
##     effect = "individual", model = "within")
## 
## Unbalanced Panel: n = 116, T = 1-23, N = 1598
## 
## Residuals:
##       Min.    1st Qu.     Median    3rd Qu.       Max. 
## -25.575464  -1.593224   0.032944   1.662955  14.946188 
## 
## Coefficients:
##                   Estimate  Std. Error t-value  Pr(>|t|)    
## IT_CEL_SETS_P2  3.8829e-02  3.1953e-03 12.1520 < 2.2e-16 ***
## GDP_EMPL       -1.0233e-04  3.8811e-05 -2.6367  0.008459 ** 
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Total Sum of Squares:    21688
## Residual Sum of Squares: 18553
## R-Squared:      0.14454
## Adj. R-Squared: 0.07691
## F-statistic: 125.029 on 2 and 1480 DF, p-value: < 2.22e-16

fe_model2_c0 <- plm(ADU_EMPL ~ IT_CEL_SETS_P2 + GDP_EMPL, data = wise_T_plm, effect = "individual", model = "within")
summary(fe_model2_c0)

## Oneway (individual) effect Within Model
## 
## Call:
## plm(formula = ADU_EMPL ~ IT_CEL_SETS_P2 + GDP_EMPL, data = wise_T_plm, 
##     effect = "individual", model = "within")
## 
## Unbalanced Panel: n = 88, T = 1-23, N = 974
## 
## Residuals:
##     Min.  1st Qu.   Median  3rd Qu.     Max. 
## -10.5509  -1.1637   0.0000   1.1512  11.1793 
## 
## Coefficients:
##                   Estimate  Std. Error t-value Pr(>|t|)    
## IT_CEL_SETS_P2  3.9064e-02  3.2282e-03 12.1008   <2e-16 ***
## GDP_EMPL       -4.7048e-05  3.7582e-05 -1.2519   0.2109    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Total Sum of Squares:    7773.4
## Residual Sum of Squares: 5613.2
## R-Squared:      0.27789
## Adj. R-Squared: 0.20519
## F-statistic: 170.095 on 2 and 884 DF, p-value: < 2.22e-16

fe_model3_c0 <- plm(EMPL_HS ~ IT_CEL_SETS_P2 + GDP_EMPL, data = wise_T_plm, effect = "individual", model = "within")
summary(fe_model3_c0)

## Oneway (individual) effect Within Model
## 
## Call:
## plm(formula = EMPL_HS ~ IT_CEL_SETS_P2 + GDP_EMPL, data = wise_T_plm, 
##     effect = "individual", model = "within")
## 
## Unbalanced Panel: n = 43, T = 1-21, N = 576
## 
## Residuals:
##    Min. 1st Qu.  Median 3rd Qu.    Max. 
## -9.4200 -1.2393  0.1358  1.3401  6.9426 
## 
## Coefficients:
##                   Estimate  Std. Error t-value Pr(>|t|)   
## IT_CEL_SETS_P2  1.1126e-02  3.9443e-03  2.8209 0.004969 **
## GDP_EMPL       -6.0291e-05  4.8858e-05 -1.2340 0.217746   
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Total Sum of Squares:    2940.7
## Residual Sum of Squares: 2879.9
## R-Squared:      0.020671
## Adj. R-Squared: -0.060479
## F-statistic: 5.60401 on 2 and 531 DF, p-value: 0.0039043

fe_model1_n0 <- plm(EMPL ~ IT_NET_USER_P2 + GDP_EMPL, data = wise_T_plm, effect = "individual", model = "within")
summary(fe_model1_n0)

## Oneway (individual) effect Within Model
## 
## Call:
## plm(formula = EMPL ~ IT_NET_USER_P2 + GDP_EMPL, data = wise_T_plm, 
##     effect = "individual", model = "within")
## 
## Unbalanced Panel: n = 115, T = 1-23, N = 1532
## 
## Residuals:
##      Min.   1st Qu.    Median   3rd Qu.      Max. 
## -28.31412  -1.53849   0.12649   1.69112  15.16707 
## 
## Coefficients:
##                   Estimate  Std. Error t-value Pr(>|t|)    
## IT_NET_USER_P2  6.0789e-02  6.6547e-03  9.1347  < 2e-16 ***
## GDP_EMPL       -7.5165e-05  4.1903e-05 -1.7938  0.07306 .  
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Total Sum of Squares:    19302
## Residual Sum of Squares: 17297
## R-Squared:      0.10384
## Adj. R-Squared: 0.030375
## F-statistic: 81.9807 on 2 and 1415 DF, p-value: < 2.22e-16

fe_model2_n0 <- plm(ADU_EMPL ~ IT_NET_USER_P2 + GDP_EMPL, data = wise_T_plm, effect = "individual", model = "within")
summary(fe_model2_n0)

## Oneway (individual) effect Within Model
## 
## Call:
## plm(formula = ADU_EMPL ~ IT_NET_USER_P2 + GDP_EMPL, data = wise_T_plm, 
##     effect = "individual", model = "within")
## 
## Unbalanced Panel: n = 87, T = 1-23, N = 967
## 
## Residuals:
##       Min.    1st Qu.     Median    3rd Qu.       Max. 
## -9.7756145 -1.2079777  0.0014338  1.1738169 10.7901611 
## 
## Coefficients:
##                  Estimate Std. Error t-value Pr(>|t|)    
## IT_NET_USER_P2 5.2875e-02 6.2300e-03  8.4871   <2e-16 ***
## GDP_EMPL       4.5112e-06 4.3363e-05  0.1040   0.9172    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Total Sum of Squares:    7705.3
## Residual Sum of Squares: 6005
## R-Squared:      0.22067
## Adj. R-Squared: 0.14256
## F-statistic: 124.305 on 2 and 878 DF, p-value: < 2.22e-16

fe_model3_n0 <- plm(EMPL_HS ~ IT_NET_USER_P2 + GDP_EMPL, data = wise_T_plm, effect = "individual", model = "within")
summary(fe_model3_n0)

## Oneway (individual) effect Within Model
## 
## Call:
## plm(formula = EMPL_HS ~ IT_NET_USER_P2 + GDP_EMPL, data = wise_T_plm, 
##     effect = "individual", model = "within")
## 
## Unbalanced Panel: n = 43, T = 1-21, N = 576
## 
## Residuals:
##      Min.   1st Qu.    Median   3rd Qu.      Max. 
## -8.991621 -1.141267  0.082597  1.390876  7.026626 
## 
## Coefficients:
##                  Estimate Std. Error t-value Pr(>|t|)
## IT_NET_USER_P2 0.00129874 0.00682115  0.1904   0.8491
## GDP_EMPL       0.00004371 0.00005007  0.8730   0.3831
## 
## Total Sum of Squares:    2940.7
## Residual Sum of Squares: 2922.9
## R-Squared:      0.0060634
## Adj. R-Squared: -0.076297
## F-statistic: 1.61964 on 2 and 531 DF, p-value: 0.19895

Estimates: many regressors

fe_model1_c <- plm(EMPL ~ IT_CEL_SETS_P2 + GDP_EMPL + SI_POV_2DAY + SP_POP_TOTL + XGDP_FSGOV + SP_URB_TOTL_IN_ZS + SP_POP_TOTL_FE_ZS, data = wise_T_plm, effect = "individual", model = "within")
summary(fe_model1_c)

## Oneway (individual) effect Within Model
## 
## Call:
## plm(formula = EMPL ~ IT_CEL_SETS_P2 + GDP_EMPL + SI_POV_2DAY + 
##     SP_POP_TOTL + XGDP_FSGOV + SP_URB_TOTL_IN_ZS + SP_POP_TOTL_FE_ZS, 
##     data = wise_T_plm, effect = "individual", model = "within")
## 
## Unbalanced Panel: n = 60, T = 1-15, N = 326
## 
## Residuals:
##       Min.    1st Qu.     Median    3rd Qu.       Max. 
## -10.212485  -1.505233   0.033734   1.485356  10.212485 
## 
## Coefficients:
##                      Estimate  Std. Error t-value  Pr(>|t|)    
## IT_CEL_SETS_P2     5.4382e-02  7.7465e-03  7.0202 1.941e-11 ***
## GDP_EMPL          -3.0738e-04  9.6590e-05 -3.1823 0.0016396 ** 
## SI_POV_2DAY        1.9607e-02  3.6393e-02  0.5388 0.5905135    
## SP_POP_TOTL        3.7868e-08  4.1913e-08  0.9035 0.3671033    
## XGDP_FSGOV        -1.1978e+00  3.1914e-01 -3.7533 0.0002154 ***
## SP_URB_TOTL_IN_ZS  3.2333e-01  1.0507e-01  3.0773 0.0023129 ** 
## SP_POP_TOTL_FE_ZS -3.3224e+00  6.0623e-01 -5.4805 1.005e-07 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Total Sum of Squares:    3039.1
## Residual Sum of Squares: 2023
## R-Squared:      0.33436
## Adj. R-Squared: 0.16474
## F-statistic: 18.5858 on 7 and 259 DF, p-value: < 2.22e-16

fe_model2_c <- plm(ADU_EMPL ~ IT_CEL_SETS_P2 + GDP_EMPL + SI_POV_2DAY + SP_POP_TOTL + XGDP_FSGOV + SP_URB_TOTL_IN_ZS + SP_POP_TOTL_FE_ZS, data = wise_T_plm, effect = "individual", model = "within")
summary(fe_model2_c)

## Oneway (individual) effect Within Model
## 
## Call:
## plm(formula = ADU_EMPL ~ IT_CEL_SETS_P2 + GDP_EMPL + SI_POV_2DAY + 
##     SP_POP_TOTL + XGDP_FSGOV + SP_URB_TOTL_IN_ZS + SP_POP_TOTL_FE_ZS, 
##     data = wise_T_plm, effect = "individual", model = "within")
## 
## Unbalanced Panel: n = 29, T = 1-13, N = 158
## 
## Residuals:
##       Min.    1st Qu.     Median    3rd Qu.       Max. 
## -6.3049573 -1.1863342 -0.0082171  0.7762636  6.8234607 
## 
## Coefficients:
##                      Estimate  Std. Error t-value  Pr(>|t|)    
## IT_CEL_SETS_P2     3.6488e-02  1.1762e-02  3.1023 0.0023864 ** 
## GDP_EMPL          -3.9673e-05  1.4223e-04 -0.2789 0.7807651    
## SI_POV_2DAY        2.0881e-01  5.8021e-02  3.5989 0.0004628 ***
## SP_POP_TOTL        1.7979e-07  9.4691e-08  1.8987 0.0599659 .  
## XGDP_FSGOV        -1.4725e+00  5.0810e-01 -2.8981 0.0044521 ** 
## SP_URB_TOTL_IN_ZS  1.9995e-01  3.3734e-01  0.5927 0.5544714    
## SP_POP_TOTL_FE_ZS -1.8012e+00  1.6421e+00 -1.0968 0.2748688    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Total Sum of Squares:    1156.8
## Residual Sum of Squares: 672.36
## R-Squared:      0.41878
## Adj. R-Squared: 0.25204
## F-statistic: 12.5578 on 7 and 122 DF, p-value: 4.747e-12

fe_model3_c <- plm(EMPL_HS ~ IT_CEL_SETS_P2 + GDP_EMPL + SI_POV_2DAY + SP_POP_TOTL + XGDP_FSGOV + SP_URB_TOTL_IN_ZS + SP_POP_TOTL_FE_ZS, data = wise_T_plm, effect = "individual", model = "within")
summary(fe_model3_c)

## Oneway (individual) effect Within Model
## 
## Call:
## plm(formula = EMPL_HS ~ IT_CEL_SETS_P2 + GDP_EMPL + SI_POV_2DAY + 
##     SP_POP_TOTL + XGDP_FSGOV + SP_URB_TOTL_IN_ZS + SP_POP_TOTL_FE_ZS, 
##     data = wise_T_plm, effect = "individual", model = "within")
## 
## Unbalanced Panel: n = 16, T = 1-13, N = 92
## 
## Residuals:
##      Min.   1st Qu.    Median   3rd Qu.      Max. 
## -5.094731 -1.282627  0.062453  1.086769  4.819162 
## 
## Coefficients:
##                      Estimate  Std. Error t-value Pr(>|t|)  
## IT_CEL_SETS_P2     2.1226e-02  2.0222e-02  1.0497  0.29754  
## GDP_EMPL          -2.1398e-04  1.9923e-04 -1.0740  0.28656  
## SI_POV_2DAY       -3.2447e-01  1.6530e-01 -1.9630  0.05368 .
## SP_POP_TOTL        5.7180e-07  5.9637e-07  0.9588  0.34100  
## XGDP_FSGOV        -8.7029e-01  7.4881e-01 -1.1622  0.24915  
## SP_URB_TOTL_IN_ZS -3.1984e-01  4.0461e-01 -0.7905  0.43196  
## SP_POP_TOTL_FE_ZS  3.3116e+00  2.8539e+00  1.1604  0.24989  
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Total Sum of Squares:    430.56
## Residual Sum of Squares: 358.89
## R-Squared:      0.16647
## Adj. R-Squared: -0.099292
## F-statistic: 1.96865 on 7 and 69 DF, p-value: 0.071926

fe_model1_n <- plm(EMPL ~ IT_NET_USER_P2 + GDP_EMPL + SI_POV_2DAY + SP_POP_TOTL + XGDP_FSGOV + SP_URB_TOTL_IN_ZS + SP_POP_TOTL_FE_ZS, data = wise_T_plm, effect = "individual", model = "within")
summary(fe_model1_n)

## Oneway (individual) effect Within Model
## 
## Call:
## plm(formula = EMPL ~ IT_NET_USER_P2 + GDP_EMPL + SI_POV_2DAY + 
##     SP_POP_TOTL + XGDP_FSGOV + SP_URB_TOTL_IN_ZS + SP_POP_TOTL_FE_ZS, 
##     data = wise_T_plm, effect = "individual", model = "within")
## 
## Unbalanced Panel: n = 59, T = 1-15, N = 320
## 
## Residuals:
##     Min.  1st Qu.   Median  3rd Qu.     Max. 
## -10.6298  -1.4613   0.0000   1.5433  10.6298 
## 
## Coefficients:
##                      Estimate  Std. Error t-value  Pr(>|t|)    
## IT_NET_USER_P2     8.8455e-02  2.2145e-02  3.9943 8.501e-05 ***
## GDP_EMPL          -1.7710e-04  1.1266e-04 -1.5721 0.1171790    
## SI_POV_2DAY       -4.0982e-02  3.7918e-02 -1.0808 0.2808160    
## SP_POP_TOTL        4.3522e-08  4.5286e-08  0.9611 0.3374387    
## XGDP_FSGOV        -1.0964e+00  3.4261e-01 -3.2003 0.0015476 ** 
## SP_URB_TOTL_IN_ZS  3.8497e-01  1.1418e-01  3.3717 0.0008634 ***
## SP_POP_TOTL_FE_ZS -3.2197e+00  6.5510e-01 -4.9148 1.595e-06 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Total Sum of Squares:    2997.4
## Residual Sum of Squares: 2237.2
## R-Squared:      0.25362
## Adj. R-Squared: 0.062614
## F-statistic: 12.3297 on 7 and 254 DF, p-value: 1.4269e-13

fe_model2_n <- plm(ADU_EMPL ~ IT_NET_USER_P2 + GDP_EMPL + SI_POV_2DAY + SP_POP_TOTL + XGDP_FSGOV + SP_URB_TOTL_IN_ZS + SP_POP_TOTL_FE_ZS, data = wise_T_plm, effect = "individual", model = "within")
summary(fe_model2_n)

## Oneway (individual) effect Within Model
## 
## Call:
## plm(formula = ADU_EMPL ~ IT_NET_USER_P2 + GDP_EMPL + SI_POV_2DAY + 
##     SP_POP_TOTL + XGDP_FSGOV + SP_URB_TOTL_IN_ZS + SP_POP_TOTL_FE_ZS, 
##     data = wise_T_plm, effect = "individual", model = "within")
## 
## Unbalanced Panel: n = 29, T = 1-13, N = 158
## 
## Residuals:
##     Min.  1st Qu.   Median  3rd Qu.     Max. 
## -6.82459 -0.95466  0.00000  1.11562  7.00842 
## 
## Coefficients:
##                      Estimate  Std. Error t-value  Pr(>|t|)    
## IT_NET_USER_P2    -2.7109e-02  3.2328e-02 -0.8386 0.4033531    
## GDP_EMPL           4.6626e-04  1.7741e-04  2.6282 0.0096878 ** 
## SI_POV_2DAY        2.0958e-01  6.0105e-02  3.4868 0.0006802 ***
## SP_POP_TOTL        1.3811e-07  9.7701e-08  1.4135 0.1600441    
## XGDP_FSGOV        -1.2036e+00  5.2321e-01 -2.3005 0.0231205 *  
## SP_URB_TOTL_IN_ZS  9.6097e-01  3.7275e-01  2.5781 0.0111240 *  
## SP_POP_TOTL_FE_ZS -1.7880e+00  1.7225e+00 -1.0380 0.3012999    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Total Sum of Squares:    1156.8
## Residual Sum of Squares: 721.24
## R-Squared:      0.37653
## Adj. R-Squared: 0.19766
## F-statistic: 10.5254 on 7 and 122 DF, p-value: 2.6603e-10

fe_model3_n <- plm(EMPL_HS ~ IT_NET_USER_P2 + GDP_EMPL + SI_POV_2DAY + SP_POP_TOTL + XGDP_FSGOV + SP_URB_TOTL_IN_ZS + SP_POP_TOTL_FE_ZS, data = wise_T_plm, effect = "individual", model = "within")
summary(fe_model3_n)

## Oneway (individual) effect Within Model
## 
## Call:
## plm(formula = EMPL_HS ~ IT_NET_USER_P2 + GDP_EMPL + SI_POV_2DAY + 
##     SP_POP_TOTL + XGDP_FSGOV + SP_URB_TOTL_IN_ZS + SP_POP_TOTL_FE_ZS, 
##     data = wise_T_plm, effect = "individual", model = "within")
## 
## Unbalanced Panel: n = 16, T = 1-13, N = 92
## 
## Residuals:
##    Min. 1st Qu.  Median 3rd Qu.    Max. 
## -4.5639 -1.2001  0.0000  1.2614  4.7459 
## 
## Coefficients:
##                      Estimate  Std. Error t-value  Pr(>|t|)    
## IT_NET_USER_P2    -1.3084e-01  3.7758e-02 -3.4653 0.0009151 ***
## GDP_EMPL           5.9583e-04  1.9362e-04  3.0772 0.0029955 ** 
## SI_POV_2DAY       -4.2324e-01  1.4353e-01 -2.9487 0.0043520 ** 
## SP_POP_TOTL        3.1663e-07  5.5719e-07  0.5683 0.5717008    
## XGDP_FSGOV        -3.6470e-01  6.7892e-01 -0.5372 0.5928722    
## SP_URB_TOTL_IN_ZS  8.0631e-01  4.0536e-01  1.9891 0.0506535 .  
## SP_POP_TOTL_FE_ZS  7.0514e+00  2.2246e+00  3.1697 0.0022758 ** 
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Total Sum of Squares:    430.56
## Residual Sum of Squares: 310.57
## R-Squared:      0.27869
## Adj. R-Squared: 0.048711
## F-statistic: 3.80853 on 7 and 69 DF, p-value: 0.0014945