WDI 2022 Report

Data Loading

In this section, I load the World Development Indicators dataset for 2022 (Bank 2022).

import pandas as pd

# Load the dataset 
df = pd.read_csv("wdi.csv")

# Show a small preview (as a figure-like output)
df.head()

	country	inflation_rate	exports_gdp_share	gdp_growth_rate	gdp_per_capita	adult_literacy_rate	primary_school_enrolment_rate	education_expenditure_gdp_share	measles_immunisation_rate	health_expenditure_gdp_share	income_inequality	unemployment_rate	life_expectancy	total_population
0	Afghanistan	NaN	18.380042	-6.240172	352.603733	NaN	NaN	NaN	68.0	NaN	NaN	14.100	62.879	41128771.0
1	Albania	6.725203	37.395422	4.856402	6810.114041	98.5	95.606712	2.74931	86.0	NaN	NaN	11.588	76.833	2777689.0
2	Algeria	9.265516	31.446856	3.600000	5023.252932	NaN	108.343933	NaN	79.0	NaN	NaN	12.437	77.129	44903225.0
3	American Samoa	NaN	46.957520	1.735016	19673.390102	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	44273.0
4	Andorra	NaN	NaN	9.563798	42350.697069	NaN	90.147346	2.66623	98.0	NaN	NaN	NaN	NaN	79824.0

Figure 1: Preview of the WDI sample (first 5 rows). Source: World Bank WDI.

Exploratory Data Analysis (EDA)

This section explores GDP per capita, Life expectancy, and Inflation (CPI, annual %) for 2022.

1) GDP per capita (levels and distribution)

I examine summary statistics, missingness, and the distribution in both raw and log scales.

# quick summary in text output
import numpy as np
print("Missing values (gdp_per_capita):", df["gdp_per_capita"].isna().sum())
print(df["gdp_per_capita"].describe())

Missing values (gdp_per_capita): 14
count       203.000000
mean      20345.707649
std       31308.942225
min         259.025031
25%        2570.563284
50%        7587.588173
75%       25982.630050
max      240862.182448
Name: gdp_per_capita, dtype: float64

import matplotlib.pyplot as plt
x = df["gdp_per_capita"].dropna()
plt.figure()
plt.hist(x, bins=30)
plt.title("GDP per Capita (USD, 2022)")
plt.xlabel("USD")
plt.ylabel("Count")
plt.show()

Figure 2: Distribution of GDP per capita (USD, 2022). Source: World Bank WDI.

import numpy as np
import matplotlib.pyplot as plt
x = df["gdp_per_capita"].dropna()
plt.figure()
plt.hist(np.log10(x[x>0]), bins=30)
plt.title("GDP per Capita (log10 scale, 2022)")
plt.xlabel("log10(USD)")
plt.ylabel("Count")
plt.show()

Figure 3: Distribution of GDP per capita on log₁₀ scale (2022). Source: World Bank WDI.

Summary — GDP per capita (2022)
- The distribution is highly right-skewed: most countries cluster below ~$20k, with a long tail of high-income economies.
- On a log scale, the distribution looks much closer to symmetric (roughly log-normal), which is typical for income variables.
- Interpretation: comparing countries on the log of GDP per capita (rather than raw USD) is more informative and reduces the influence of outliers.

---

2) Life expectancy (levels and relation to income)

I inspect the distribution of life expectancy and its relationship with income (diminishing returns expected), a pattern first described by Preston (Preston 1975).

# brief summary
print("Missing values (life_expectancy):", df["life_expectancy"].isna().sum())
print(df["life_expectancy"].describe())

Missing values (life_expectancy): 8
count    209.000000
mean      72.416519
std        7.713322
min       52.997000
25%       66.782000
50%       73.514634
75%       78.475000
max       85.377000
Name: life_expectancy, dtype: float64

import numpy as np
import matplotlib.pyplot as plt


gdp_le = df[["gdp_per_capita","life_expectancy"]].dropna().copy()
x = np.log10(gdp_le["gdp_per_capita"].values)
y = gdp_le["life_expectancy"].values

m, b = np.polyfit(x, y, 1)
plt.figure()
plt.scatter(x, y, alpha=0.7)
plt.plot(np.sort(x), m*np.sort(x)+b, linewidth=2)
plt.title("Life Expectancy vs log10(GDP per Capita), 2022")
plt.xlabel("log10(GDP per Capita, USD)")
plt.ylabel("Life Expectancy (years)")
plt.grid(True, linewidth=0.3)
plt.show()

print(f"Correlation (life expectancy vs log10 GDP per capita): {np.corrcoef(x,y)[0,1]:.3f}")

Figure 4: Life expectancy vs log₁₀(GDP per capita), 2022. Source: World Bank WDI.

Correlation (life expectancy vs log10 GDP per capita): 0.841

Summary — Life expectancy vs GDP per capita
- There is a strong positive association: the correlation with log₁₀(GDP per capita) is about 0.84 in the output.
- The slope flattens at higher incomes, consistent with diminishing returns (gains in life expectancy are large at low incomes and smaller at high incomes), this finding aligns with macroeconomic theories discussed by Mankiw (Mankiw 2019). .
- Takeaway: income is a powerful predictor of population health, especially among lower-income countries. —

3) Inflation (levels and relation to growth)

I examine inflation levels and its same-year association with GDP growth.

# brief summary
print("Missing values (inflation_rate):", df["inflation_rate"].isna().sum())
print(df["inflation_rate"].describe())

#| label: fig-growth-vs-infl
#| fig-cap: "GDP growth vs inflation, 2022. Source: [World Bank WDI](https://databank.worldbank.org/source/world-development-indicators)."
import numpy as np
import matplotlib.pyplot as plt

if "gdp_growth_rate" in df.columns:
    infl_growth = df[["inflation_rate","gdp_growth_rate"]].dropna()
    plt.figure()
    plt.scatter(infl_growth["inflation_rate"], infl_growth["gdp_growth_rate"], alpha=0.7)
    plt.title("GDP Growth vs Inflation (2022)")
    plt.xlabel("Inflation rate (annual %)")
    plt.ylabel("GDP growth (annual %)")
    plt.grid(True, linewidth=0.3)
    plt.show()

    r = np.corrcoef(infl_growth["inflation_rate"], infl_growth["gdp_growth_rate"])[0,1]
    print(f"Correlation (GDP growth vs inflation): {r:.3f}")
else:
    print("Column 'gdp_growth_rate' not found in the dataset.")

Missing values (inflation_rate): 48
count    169.000000
mean      12.493936
std       19.682433
min       -6.687321
25%        5.518129
50%        7.967574
75%       11.665567
max      171.205491
Name: inflation_rate, dtype: float64

Correlation (GDP growth vs inflation): -0.140

Summary — GDP growth vs inflation (2022)
- The cross-section shows a weak negative correlation (≈ −0.14), and the scatter is very noisy.
- Several extreme-inflation outliers pull the pattern around; excluding them would likely make the relationship even weaker.
- Same-year cross-sectional comparisons don’t identify causality—growth and inflation dynamics are time-dependent; panel/time-series analysis would be more appropriate.

---

Visualisations

Top GDP per Capita (Bar Chart)

As shown in (Figure 5), high-income economies dominate the upper tail of GDP per capita in 2022.

import matplotlib.pyplot as plt

gdp_top = (
    df[["country", "gdp_per_capita"]]
    .dropna()
    .sort_values("gdp_per_capita", ascending=False)
    .head(12)
    .iloc[::-1]
)

plt.figure(figsize=(9,6))
plt.barh(gdp_top["country"], gdp_top["gdp_per_capita"])
plt.title("Top 12 GDP per Capita (USD), 2022")
plt.xlabel("GDP per Capita (USD)")
plt.ylabel("Country")
plt.tight_layout()
plt.show()

Figure 5: Top 12 countries by GDP per capita (2022). Source: World Bank WDI.

Life Expectancy vs Income (Scatter)

In (Figure 6), life expectancy rises with income, with diminishing returns at high income.

Figure 6: Life expectancy vs. log₁₀(GDP per capita), 2022. Source: World Development Indicators — World Bank.

Key Statistics Table

Table Table 1 summarizes the indicators used in the analysis and supports the distributional findings in Figure 2 and the relationship in Figure 4.

import pandas as pd
import numpy as np

indicators = {
    "gdp_per_capita": "GDP per capita (USD)",
    "life_expectancy": "Life expectancy (years)",
    "inflation_rate": "Inflation rate (%)",
}

use_cols = [c for c in indicators if c in df.columns]
desc = (
    df[use_cols]
    .describe(percentiles=[], include="all")
    .loc[["count","mean","50%","std","min","max"]]
    .rename(index={"50%":"median"})
    .T
)

def fmt_row(name, row):
    row = row.copy()
    if name == "gdp_per_capita":
        for c in ["mean","median","std","min","max"]:
            row[c] = f"{row[c]:,.0f}"
    elif name in ["life_expectancy"]:
        for c in ["mean","median","std","min","max"]:
            row[c] = f"{row[c]:.1f}"
    else:  # inflation_rate or other %
        for c in ["mean","median","std","min","max"]:
            row[c] = f"{row[c]:.2f}"
    row["count"] = int(row["count"])
    return row

display_df = pd.DataFrame(
    [fmt_row(name, desc.loc[name]) for name in desc.index],
    index=[indicators[name] for name in desc.index]
)[["count","mean","median","std","min","max"]]

display_df

Table 1: Key statistics for selected indicators (2022). Source: World Bank WDI.

	count	mean	median	std	min	max
GDP per capita (USD)	203	20,346	7,588	31,309	259	240,862
Life expectancy (years)	209	72.4	73.5	7.7	53.0	85.4
Inflation rate (%)	169	12.49	7.97	19.68	-6.69	171.21

Interactive Dashboard

Use the controls to explore the 2022 indicators. The scatter updates live; the table below shows the filtered rows.

Interactive dashboard: filter and explore 2022 indicators.

References

Bank, World. 2022. “World Development Indicators 2022.” https://databank.worldbank.org/source/world-development-indicators.

Mankiw, N. Gregory. 2019. Principles of Economics. 9th ed. Cengage Learning.

Preston, Samuel H. 1975. “The Changing Relation Between Mortality and Level of Economic Development.” Population Studies 29 (2): 231–48. https://doi.org/10.1080/00324728.1975.10410201.