Lab2
Diana Piskareva
2023-06-13
library(readxl)
library(tidyverse)## ── 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 3.4.2 ✔ 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 errorslibrary(ineq)
library(reshape2)##
## Присоединяю пакет: 'reshape2'
##
## Следующий объект скрыт от 'package:tidyr':
##
## smithslibrary(dplyr)df <- read_excel("/Users/DP/Downloads/GCIPrawdata.xlsx", skip = 2)
head(df) ## # A tibble: 6 × 14
## Country Year `Decile 1 Income` `Decile 2 Income` `Decile 3 Income`
## <chr> <dbl> <dbl> <dbl> <dbl>
## 1 Afghanistan 1980 206 350 455
## 2 Afghanistan 1981 212 361 469
## 3 Afghanistan 1982 221 377 490
## 4 Afghanistan 1983 238 405 527
## 5 Afghanistan 1984 249 424 551
## 6 Afghanistan 1985 256 435 566
## # ℹ 9 more variables: `Decile 4 Income` <dbl>, `Decile 5 Income` <dbl>,
## # `Decile 6 Income` <dbl>, `Decile 7 Income` <dbl>, `Decile 8 Income` <dbl>,
## # `Decile 9 Income` <dbl>, `Decile 10 Income` <dbl>, `Mean Income` <dbl>,
## # Population <dbl>sel_Year <- c(1980, 2014)
sel_Country <- c("Russian Federation", "Ukraine")
temp <- df %>% filter(Country %in% sel_Country & Year %in% sel_Year)
temp## # A tibble: 4 × 14
## Country Year `Decile 1 Income` `Decile 2 Income` `Decile 3 Income`
## <chr> <dbl> <dbl> <dbl> <dbl>
## 1 Russian Federation 1980 1126 1555 1891
## 2 Russian Federation 2014 1781 2880 3740
## 3 Ukraine 1980 1572 2260 2702
## 4 Ukraine 2014 1026 1553 1976
## # ℹ 9 more variables: `Decile 4 Income` <dbl>, `Decile 5 Income` <dbl>,
## # `Decile 6 Income` <dbl>, `Decile 7 Income` <dbl>, `Decile 8 Income` <dbl>,
## # `Decile 9 Income` <dbl>, `Decile 10 Income` <dbl>, `Mean Income` <dbl>,
## # Population <dbl>total_income <- temp[, "Mean Income"] * temp[, "Population"]
options(scipen = 999)
total_income## Mean Income
## 1 391702000000
## 2 944909800000
## 3 232704752827
## 4 207127001400test <- c(2, 4, 10, 22)
cumsum(test) ## [1] 2 6 16 38# Pick the deciles (Columns 3 to 12) in Row 1 (RF, 1980)
decs_RF80 <- unlist(temp[1, 3:12]) * 2
# Give the total income, assuming a population of 20
total_inc <- 20 * unlist(temp[1, "Mean Income"])
cum_inc_share_RF80 <- cumsum(decs_RF80) / total_inc
cum_inc_share_RF80## Decile 1 Income Decile 2 Income Decile 3 Income Decile 4 Income
## 0.03995742 0.09513840 0.16224273 0.23977999
## Decile 5 Income Decile 6 Income Decile 7 Income Decile 8 Income
## 0.32732434 0.42523066 0.53477644 0.65890703
## Decile 9 Income Decile 10 Income
## 0.80479063 1.00000000# For RF, 2014
# Go to Row 2 (RF, 2014)
decs_RF14 <- unlist(temp[2, 3:12]) * 2
# Give the total income, assuming a population of 20
total_inc <- 20 * unlist(temp[2, "Mean Income"])
cum_inc_share_RF14 <- cumsum(decs_RF14) / total_inc
# For the Ukraine, 1980
# Select Row 3 (Ukraine, 1980)
decs_Ua80 <- unlist(temp[3, 3:12]) * 2
# Give the total income, assuming a population of 20
total_inc <- 20 * unlist(temp[3, "Mean Income"])
cum_inc_share_Ua80 <- cumsum(decs_Ua80) / total_inc
# For the Ukraine, 2014
# Select Row 4 (Ukraine, 2014)
decs_Ua14 <- unlist(temp[4, 3:12]) * 2
# Give the total income, assuming a population of 20
total_inc <- 20 * unlist(temp[4, "Mean Income"])
cum_inc_share_Ua14 <- cumsum(decs_Ua14) / total_inc plot(cum_inc_share_RF80, type = "l", col = "blue",
lwd = 2, ylab = "Cumulative income share",
# xlim = c(0,10), ylim = c(0,1)
)
# Add the perfect equality line
abline(a = 0, b = 0.1, col = "black", lwd = 2)
title("Lorenz curve, Russia Federation, 1980") 
Interpretation: Since the straight line shows the situation of absolute
equality, the bend in our case demonstrates the situation of inequality,
so we can see that 40% of the population (OX axis) owns 20% of all
income (OY axis)
plot(cum_inc_share_RF80, type = "l", col = "blue",
lty = 2, lwd = 2, xlab = "Deciles",
ylab = "Cumulative income share")
# Add the perfect equality line
abline(a = 0, b = 0.1, col = "black", lwd = 2)
lines(cum_inc_share_RF14, col = "green", lty = 1, lwd = 2)
lines(cum_inc_share_Ua80, col = "magenta", lty = 2, lwd = 2)
lines(cum_inc_share_Ua14, col = "orange", lty = 1, lwd = 2)
title("Lorenz curves, Russia Federation and the Ukraine (1980 and 2014)")
legend("topleft", lty = 2.5:1, lwd = 2, cex = 1.2, legend =
c("RF, 1980", "RF, 2014",
"Ukraine, 1980", "Ukraine, 2014"),
col = c("blue", "green", "magenta", "orange")) 
Here we can immediately note that the Russian Federation in 1980 was
closest to economic equality, since this curve is closest to a straight
line, in turn, the situation in Ukraine in 2014 reflects the greatest
inequality, since the values of the curve are the furthest from a
straight line. Ukraine 2014 - 50% of the population have 20% of income,
Ukraine in 1980 almost completely coincides with Russia in 2014, so 80%
of the population have 50% of income
This may be due to several historical facts, the deterioration of the economic situation in Russia is associated with the crises of 2008-2009 (the global financial crisis worsened the economic state of the state, respectively, and the state of incomes of the population) and 2014 in connection with the annexation of Crimea by Russia and subsequent comprehensive sanctions, which reduced the incomes of the population, and consequently affected the level of economic inequality. The situation in Ukraine may be dictated by the fact that in 1980 Ukraine was part of the USSR, which affected the fact that this region received external support and was better funded, the situation worsened in 2014 due to the collapse of the USSR, which in the long term affected the development of the economy of Ukraine, also the global financial crisis affected situations of economic inequality, and may also be related to the unstable political situation and the beginning of the armed invasion of the Russian Federation on the territory of Ukraine, as well as Euromaidan and revolutionary events in the country
g_RF80 <- Gini(decs_RF80)
g_RF14 <- Gini(decs_RF14)
g_Ua80 <- Gini(decs_Ua80)
g_Ua14 <- Gini(decs_Ua14)
paste("Gini coefficients")## [1] "Gini coefficients"paste("RF - 1980: ", round(g_RF80, 2), ", 2014: ", round(g_RF14, 2))## [1] "RF - 1980: 0.24 , 2014: 0.33"paste("Ukraine - 1980: ", round(g_Ua80, 2), ", 2014: ", round(g_Ua14, 2))## [1] "Ukraine - 1980: 0.32 , 2014: 0.43"plot(cum_inc_share_RF80, type = "l", col = "blue", lty = 2,
lwd = 2, xlab = "Deciles",
ylab = "Cumulative income share")
# Add the perfect equality line
abline(a = 0, b = 0.1, col = "black", lwd = 2)
lines(cum_inc_share_RF14, col = "green", lty = 1, lwd = 2)
lines(cum_inc_share_Ua80, col = "magenta", lty = 2, lwd = 2)
lines(cum_inc_share_Ua14, col = "orange", lty = 1, lwd = 2)
title("Lorenz curves, RF and Ukraine (1980 and 2014)")
legend("topleft", lty = 2.5:1, lwd = 2, cex = 1.2, legend =
c("RF, 1980", "RF, 2014",
"Ukraine, 1980", "Ukraine, 2014"),
col = c("blue", "green", "magenta", "orange"))
text(8.4, 0.78, round(g_RF80, digits = 3), col = 'blue')
text(9.4, 0.6, round(g_RF14, digits = 3), col = 'green')
text(5.2, 0.37, round(g_Ua80, digits = 3), col = 'magenta')
text(6.4, 0.31, round(g_Ua14, digits = 3), col = 'orange')
This graph with Gini coefficients confirms the earlier assumptions, and
also confirms the assumption that the indicators of Russia 2014 (0.325)
and Ukraine 1980 (0.32) are almost identical
df$gini <- 0# Give us the number of rows in decile_data
noc <- nrow(df)
for (i in seq(1, noc)){
# Go to Row I to get the decile data
decs_i <- unlist(df[i, 3:12])
df$gini[i] <- Gini(decs_i)
}
summary(df$gini)## Min. 1st Qu. Median Mean 3rd Qu. Max.
## 0.1791 0.3470 0.4814 0.4617 0.5700 0.7386gini_selected <- df%>% filter(Country %in% c("Russian Federation", "Ukraine", "Turkmenistan", "Kazakhstan"))
ggplot(gini_selected,
aes(x = Year, y = gini, color = Country)) +
geom_line(linewidth = 1) +
theme_bw() +
ylab("Gini") +
ggtitle("Gini coefficients for the post-Soviet countries")
This graph almost duplicates the graph with Gini coefficients, since
there is a noticeably similar dynamics, for example, the peak of
inequality in Kazakhstan (1996-1997) is reflected both here and in the
previous graph, as well as a sharp decrease in inequality in Ukraine -
the absolute minimum in the graph (2004-2005). This graph also shows a
fairly uniform and stable situation before 1990 and a sharp increase in
inequality after - as a consequence of the collapse of the USSR. The
difference between these two charts is that there is a significant
difference between the peak of Russia and Kazakhstan (1996-1997), in
Russia 10% of the richest earned 30 times more than 10% of the poorest,
in Kazakhstan it was more than 50 times. On the graph with Gini
coefficients, this indicator is approximately equal to 0.5 for both
Russia and Kazakhstan during this time period. This may be due to the
fact that in the Gini coefficient we take into account all the values of
various population groups, and here we take two conditionally “extreme
points”
df$ratio90_10 <- df$`Decile 10 Income`/df$`Decile 1 Income`
df$ratio90_50 <- df$`Decile 10 Income`/df$`Decile 5 Income`
df$ratio50_10 <- df$`Decile 5 Income`/df$`Decile 1 Income`ratios_selected <- df%>% filter(Country %in% c("Russian Federation", "Ukraine", "Turkmenistan", "Kazakhstan"))
ggplot(ratios_selected,
aes(x = Year, y = ratio90_10, color = Country)) +
geom_line(linewidth = 1) +
theme_bw() +
ylab("Interdecile P90/P10") +
ggtitle("Interdecile P90/P10 for the post-Soviet countries")
This graph almost duplicates the graph with Gini coefficients, since
there is a noticeably similar dynamics here, for example, the peak of
inequality in Kazakhstan is reflected both here and in the previous
graph, as well as a sharp decrease in inequality in Ukraine - the
absolute minimum in the graph. This graph also shows a fairly uniform
and stable situation before 1990 and a sharp increase in inequality
after - as a consequence of the collapse of the USSR
ggplot(ratios_selected,
aes(x = Year, y = ratio90_50, color = Country)) +
geom_line(linewidth = 1) +
theme_bw() +
ylab("Interdecile P90/P50") +
ggtitle("Interdecile P90/P50 for the post-Soviet countries")
ggplot(ratios_selected,
aes(x = Year, y = ratio50_10, color = Country)) +
geom_line(linewidth = 1) +
theme_bw() +
ylab("Interdecile P50/P10") +
ggtitle("Interdecile P50/P10 for the post-Soviet countries")
In general, we can trace a trend similar to the rest of the graphs, but
if we talk about the differences between these two graphs, we can see
that the peaks are distributed differently, so in the second graph the
inequality that arose in 1996-1997 is more clearly visible, this may be
due to the fact that it most affected the poorest strata However, on the
same graph after 2000, it is noticeable that inequality seems to have a
more “smooth” character without sharp jumps, perhaps this is due to the
fact that further crises affected the poorest and the “middle” equally.
In the first chart after 2000, there are more angles and sharp drops,
that is, it is noticeable how the ratio between the “middle” and the
richest changed - the situation improved during periods, it got worse
during periods, but basically it all developed in leaps and bounds