Military spending analysis
Teeda Khan
2025-11-15
#**Research Question:**
#Is there evidence that the average military spending as a percentage of GDP in the United States vastly differs from China between 2020 and 2023?
##**Statistical Hypotheses:**
#Null hypothesis Ho:There is no significant difference between the US and China in average military spending as a percentage of GDP from 2020-2023
# Alternative hypothesis Ha: here is a significant difference between the US and China in average military spending as a percentage of GDP from 2020-2023 #Data Description
#My data comes from the Our World in Data (OWID) Military Expenditure dataset.It calculates annual military spending as a percentage of individual nations GDP worldwide, and their differences year to year. This data set is constructed by world-renowned global sources such as the World Bank, The (SPRI) Military Expenditure Database, and other accredited statistical agencies.Since first publishing in 1949, to date, there are 8,366 rows of data, and 5 total columns- which include: "Country", "Code", "Year", "Military.expenditure....of.GDP", and "World.regions.according.to.OWID".
#Since 1949, this data set continues to cover annual military spending as a percentage of a nation's military spending,(GDP), or the relative percentage spent to each annual total economic output. This data set has been cited in various academic research and public policy analysis, showcasing the power and reliability of this data. I trust it to help me determine just how Covid may have influenced military spending trends between China and US through 2020-2023. And whether **There is evidence the average military spending as a percentage of GDP in China vastly differs from the United States between 2020 and 2023?**#
#You can view and find the full data set below:
"https://ourworldindata.org/grapher/military-spending-as-a-share-of-gdp-\nsipri.csv?v=1&csvType=full&useColumnShortNames=true"## [1] "https://ourworldindata.org/grapher/military-spending-as-a-share-of-gdp-\nsipri.csv?v=1&csvType=full&useColumnShortNames=true"#Set working directory and read data-set
head(military_data, 10)## Entity Code Year Military.expenditure....of.GDP.
## 1 Afghanistan AFG 1970 1.629606
## 2 Afghanistan AFG 1973 1.868910
## 3 Afghanistan AFG 1974 1.610825
## 4 Afghanistan AFG 1975 1.722066
## 5 Afghanistan AFG 1976 2.046087
## 6 Afghanistan AFG 1977 2.011475
## 7 Afghanistan AFG 2004 2.431254
## 8 Afghanistan AFG 2005 1.992066
## 9 Afghanistan AFG 2006 1.896234
## 10 Afghanistan AFG 2007 2.566267
## World.regions.according.to.OWID
## 1
## 2
## 3
## 4
## 5
## 6
## 7
## 8
## 9
## 10#Rename columns for clairty
military_rename<- military_data%>%
rename( Country = Entity,
Mil_Spend_GDP = Military.expenditure....of.GDP.,
Region = World.regions.according.to.OWID
) #Filter and create table to compare GDP % 2020-2023 between US and China
us_china_covid <- military_rename%>%
filter(Country %in% c("United States", "China"),
Year >= 2020 & Year <= 2023) %>%
select(Country, Year, Mil_Spend_GDP)
print(us_china_covid)## Country Year Mil_Spend_GDP
## 1 China 2020 1.756699
## 2 China 2021 1.605260
## 3 China 2022 1.630534
## 4 China 2023 1.664849
## 5 United States 2020 3.645188
## 6 United States 2021 3.404520
## 7 United States 2022 3.309477
## 8 United States 2023 3.304442sum(is.na(us_china_covid$Mil_Spend_GDP))## [1] 0head(us_china_covid)## Country Year Mil_Spend_GDP
## 1 China 2020 1.756699
## 2 China 2021 1.605260
## 3 China 2022 1.630534
## 4 China 2023 1.664849
## 5 United States 2020 3.645188
## 6 United States 2021 3.404520us_china_summary <- us_china_covid %>%
group_by(Country) %>%
summarise( Mean_Spend = mean(Mil_Spend_GDP, na.rm = TRUE),
SD_Spend = sd(Mil_Spend_GDP, na.rm = TRUE),
Min_Spend = min(Mil_Spend_GDP, na.rm = TRUE),
Max_Spend = max(Mil_Spend_GDP, na.rm = TRUE),
n_years = n()
)
us_china_summary## # A tibble: 2 × 6
## Country Mean_Spend SD_Spend Min_Spend Max_Spend n_years
## <chr> <dbl> <dbl> <dbl> <dbl> <int>
## 1 China 1.66 0.0662 1.61 1.76 4
## 2 United States 3.42 0.160 3.30 3.65 4#Clean the dataset and conduct exploratory data analysis (EDA) to better understand differences in % and get the Summary Stats
us_china_summary <- us_china_covid %>%
group_by(Country) %>%
summarise(Mean_Spend = mean(Mil_Spend_GDP, na.rm = TRUE),
SD_Spend = sd(Mil_Spend_GDP, na.rm = TRUE),
Min_Spend = min(Mil_Spend_GDP, na.rm = TRUE),
Max_Spend = max(Mil_Spend_GDP, na.rm = TRUE),
n_years = n()
)
us_china_summary## # A tibble: 2 × 6
## Country Mean_Spend SD_Spend Min_Spend Max_Spend n_years
## <chr> <dbl> <dbl> <dbl> <dbl> <int>
## 1 China 1.66 0.0662 1.61 1.76 4
## 2 United States 3.42 0.160 3.30 3.65 4#I will perform a two-sample t-test comparing: China’s mean military spending (% of GDP) vs. U.S. mean military spending (% of GDP) for the years 2020–2023.
t_res <- t.test(Mil_Spend_GDP ~ Country,
data = us_china_covid,
alternative = "two.sided",
var.equal = FALSE)
t_res##
## Welch Two Sample t-test
##
## data: Mil_Spend_GDP by Country
## t = -20.269, df = 4.0034, p-value = 3.475e-05
## alternative hypothesis: true difference in means between group China and group United States is not equal to 0
## 95 percent confidence interval:
## -1.991424 -1.511719
## sample estimates:
## mean in group China mean in group United States
## 1.664335 3.415907#Create table to compare GDP % 2020-2023
us_china_covid <- military_rename %>%
filter(Country %in% c("United States", "China"),
Year >= 2020 & Year <= 2023) %>%
select(Country, Year, Mil_Spend_GDP)
print(us_china_covid)## Country Year Mil_Spend_GDP
## 1 China 2020 1.756699
## 2 China 2021 1.605260
## 3 China 2022 1.630534
## 4 China 2023 1.664849
## 5 United States 2020 3.645188
## 6 United States 2021 3.404520
## 7 United States 2022 3.309477
## 8 United States 2023 3.304442sum(is.na(us_china_covid$Mil_Spend_GDP)) #checking the amount of missing values in this column ## [1] 0Create table to summarise the average GDP % spending across 2020-2023 between the US and China.
us_china_covid %>%
group_by(Country) %>%
summarise(Average_Mil_Spend = mean(Mil_Spend_GDP, na.rm = TRUE))## # A tibble: 2 × 2
## Country Average_Mil_Spend
## <chr> <dbl>
## 1 China 1.66
## 2 United States 3.42head(us_china_covid)## Country Year Mil_Spend_GDP
## 1 China 2020 1.756699
## 2 China 2021 1.605260
## 3 China 2022 1.630534
## 4 China 2023 1.664849
## 5 United States 2020 3.645188
## 6 United States 2021 3.404520sum(is.na(us_china_covid$Mil_Spend_GDP)) #checking the amount of missing values in this column ## [1] 0#Line graph to visualize the difference in spending trends
library(ggplot2)
ggplot(us_china_covid, aes(x = Year, y = Mil_Spend_GDP, color = Country)) +
geom_line(size = 1.2) +
geom_point(size = 2) +
labs(title = "Military Spending as % of GDP (2020–2023)",
y = "% of GDP",
x = "Year") +
theme_minimal()## Warning: Using `size` aesthetic for lines was deprecated in ggplot2 3.4.0.
## ℹ Please use `linewidth` instead.
## This warning is displayed once every 8 hours.
## Call `lifecycle::last_lifecycle_warnings()` to see where this warning was
## generated.
#**Key findings of the Line graph:**
#China stays around 1.60%–1.76% of GDP.
#The U.S. stays around 3.40%–3.64% of GDP
## The United States spends much more of its GDP on the military than China
#The US uses almost double what China does of their GDP in military expenditures.
## Neither Country shows dramatic increae or decrease during this period#Boxplot visulization
ggplot(us_china_covid, aes(x = Country, y = Mil_Spend_GDP, fill = Country )) +
geom_boxplot() +
labs(title = "Distribution of Military Spending (% of GDP)",
x = "Country",
y = "Military Spending (% of GDP)")+
theme_minimal()
#Key findings of box plot
#The box plot illistruates the comparision of the spending distribution differences per GDP between China and the US from 2020-2023
#The entire US boxplot is 3% above Chinas box plot
#The length of the United States IQR or Interquartile range is a little wider than china, which suggests the US's spending fluctuates more year to year than China.
#Chinas box-plot is very narrow, suggesting tight and consistent spending habits.
#China median ≈ 1.64%
#U.S. median ≈ 3.52%
#Finally, there are no outliers on either plots, suggesting stables values across the yearsTwo-sample test findings
#Welch Two Sample t-test
#data: Mil_Spend_GDP by Country
#t = -20.269, df = 4.0034, p-value = 3.475e-05
#alternative hypothesis: true difference in means between group China and group United States is not equal to 0
#95 percent confidence interval:
-1.991424 -1.511719## [1] -3.503143#sample estimates:
#mean in group China 1.664335
#mean in group United States 3.415907 Conclusion
# Welch T-test variables
#t = −20.269
#df ≈ 4.00
#p-value = 3.475e-05 ≈ 0.00003475
#Means:
#China: 1.6643
#US: 3.4159
#Assume your significance level is α = 0.05
#p-value=0.00003475<0.05
#*We reject the Null Hypothesis as there is strong statistical evidence that the average military spending as a percentage of GDP from 2020–2023 is significantly different between the US and China.* Thoughts on findings and future direction
#Firstly, in reference to the line graph; while both lines fall relatively flat, Chinas stagnent trend suggests that the Covid-19 pandemic did not influence large fluctuations in % of GDP military spending in China during China averaged a military spending of 1.6% while the US averaged 3.4% of its % GDP spending.
#Both spending trends dropping in similar fashion in 2021 shines a cast down that the US had to cut spending economically affected both countries within economic trade
#While China conservatively increased their military expenditure % from 2021-2022, the US continued to cutback on spending. Suggesting the pandemic caused the economies to react in different ways after both taking a similar drop in 2021.
#Future direction the data could illistruate
#Topic: Comparing absolute military spending rather than percentages
#Context: GDP percentage does not show the real spending power.
#Future work could compare: Total military budget (in billions) Growth rates of military spending