World Bank API in Python

World Bank API in Python#

by Avery Fernandez and Michael T. Moen

The World Bank offers a suite of APIs providing access to a vast array of global development data, including economic indicators, project information, and more. These APIs enable users to programmatically retrieve data for analysis, application development, and research purposes.

Please see the following resources for more information on API usage:

NOTE: Please see access details and rate limit requests for this API in the official documentation.

These recipe examples were tested on March 10, 2026.

Setup#

Import Libraries#

The following external libraries need to be installed into your environment to run the code examples in this tutorial:

We import the libraries used in this tutorial below:

import matplotlib.pyplot as plt
import pandas as pd
import plotly.express as px
import pycountry
import requests
from time import sleep

DataBank Bulk Download#

The World Bank offers bulk downloads of specific datasets from DataBank Bulk Download or DataBank Bulk Microdata Download. If you require extensive or complete data, consider using these bulk downloads. For smaller requests or interactive analysis, the standard API calls work well.

1. Comparing GDP per Capita by Country#

In this example, we look at the GDP per capita of each country using the NY.GDP.PCAP.CD indicator. Note we examine both countries and dependent territories defined as having ISO 3166-1 alpha-3 (also known here as ISO 3) codes.

BASE_URL = "http://api.worldbank.org/v2/country/all/indicator"
indicator = "NY.GDP.PCAP.CD"    # Indicator code for GDP per capita (current USD)

params = {
    "date": 2024,
    "format": "json",
    "per_page": 500,
}
response = requests.get(f"{BASE_URL}/{indicator}", params=params)

# Status code 200 indicates success
response

<Response [200]>

# Extract JSON data from response
data = response.json()[1]

# Print first result
data[0]

{'indicator': {'id': 'NY.GDP.PCAP.CD',
  'value': 'GDP per capita (current US$)'},
 'country': {'id': 'ZH', 'value': 'Africa Eastern and Southern'},
 'countryiso3code': 'AFE',
 'date': '2024',
 'value': 1615.39635562849,
 'unit': '',
 'obs_status': '',
 'decimal': 1}

# Convert result to a pandas DataFrame and drop unnecessary columns
df = pd.DataFrame(data)
df = df.drop(columns=["country", "indicator", "unit", "obs_status", "decimal"])

df.head()
countryiso3code date value
0 AFE 2024 1615.396356
1 AFW 2024 1411.337029
2 ARB 2024 7583.811701
3 CSS 2024 19903.811231
4 CEB 2024 24604.806860

Below, we use the pycountry package to retrieve the names of te countries based on their ISO 3 codes. Since the World Bank also maintains listings for different aggregations, we also drop those from the data below:

def iso3_to_country(code):
    country = pycountry.countries.get(alpha_3=code)
    return country.name if country else None

df["country_name"] = df["countryiso3code"].apply(iso3_to_country)

# World Bank contains some aggregate regions (e.g., "World", "High income") that aren't
# officially defined countries, so we drop those rows
df = df.dropna(subset=["country_name"])

# Extract top 20 countries by GDP per capita
top20 = df.sort_values("value", ascending=False).head(20)

# Print top 20 countries
top20
countryiso3code date value country_name
179 MCO 2024 288001.433369 Monaco
70 BMU 2024 142855.374714 Bermuda
165 LUX 2024 137781.681659 Luxembourg
142 IRL 2024 112894.953241 Ireland
237 CHE 2024 103998.186686 Switzerland
220 SGP 2024 90674.066633 Singapore
196 NOR 2024 86785.433449 Norway
137 ISL 2024 86040.530985 Iceland
255 USA 2024 84534.040784 United States
208 QAT 2024 76688.694174 Qatar
114 FRO 2024 74119.660755 Faroe Islands
166 MAC 2024 72004.743702 Macao
102 DNK 2024 71026.483227 Denmark
188 NLD 2024 67520.421896 Netherlands
59 AUS 2024 64603.985631 Australia
60 AUT 2024 58268.878765 Austria
236 SWE 2024 57117.487744 Sweden
67 BEL 2024 56614.567950 Belgium
122 DEU 2024 56103.732318 Germany
84 CAN 2024 54340.347722 Canada 
fig, ax = plt.subplots(figsize=(10,6))

# Bars
ax.bar(top20["country_name"], top20["value"], color="#4C72B0")

# Gridlines
ax.grid(axis="y", linestyle="--", linewidth=0.6, alpha=0.7)
ax.set_axisbelow(True)

# Remove top/right spines
ax.spines["top"].set_visible(False)
ax.spines["right"].set_visible(False)

# Lighten remaining spines
ax.spines["left"].set_color("#CCCCCC")
ax.spines["bottom"].set_color("#CCCCCC")

# Labels and title
ax.set_title("Top 20 Countries by GDP per Capita (2024)")
ax.set_xlabel("Country")
ax.set_ylabel("GDP per Capita (USD)")

# Rotate country labels
plt.xticks(rotation=30, ha="right")

plt.tight_layout()
plt.show()
../_images/c8c8fd1d4faa041e9869dcaa8ba045897e03e6f31d390538654fefec119215f0.png

2. Comparing Scientific and Technical Journal Articles by Country#

In this example, we look at the “scientific and technical journal articles” of each country using the IP.JRN.ARTC.SC indicator and the total population of each country using the SP.POP.TOTL. Note we examine both countries and dependent territories defined as having ISO 3166-1 alpha-3 (also known here as ISO 3) codes.

BASE_URL = "https://api.worldbank.org/v2/country/all/indicator"
indicators = {
    "IP.JRN.ARTC.SC": "Article Count",
    "SP.POP.TOTL": "Population"
}

params = {
    "date": 2021,
    "format": "json",
    "per_page": 500
}

dfs = []
for indicator, colname in indicators.items():
    response = requests.get(f"{BASE_URL}/{indicator}", params=params)
    
    df_temp = pd.DataFrame(response.json()[1])
    
    df_temp = df_temp[["countryiso3code", "value"]]
    df_temp = df_temp.rename(columns={"value": colname})
    
    dfs.append(df_temp)

# Merge both datasets
df = dfs[0].merge(dfs[1], on="countryiso3code")

# Add country names
df["Country Name"] = df["countryiso3code"].apply(iso3_to_country)

# Remove aggregates like "World", "High income", etc.
df = df.dropna(subset=["Country Name"])

df.head()
countryiso3code Article Count Population Country Name
69 AFG 177.39 40000412.0 Afghanistan
70 ALB 241.09 2489762.0 Albania
71 DZA 6839.00 44761099.0 Algeria
72 ASM NaN 49225.0 American Samoa
73 AND 9.44 78364.0 Andorra 
# Calculate article count per million people
df["Article Count per Million"] = df["Article Count"] / df["Population"] * 1000000

df.head()
countryiso3code Article Count Population Country Name Article Count per Million
69 AFG 177.39 40000412.0 Afghanistan 4.434704
70 ALB 241.09 2489762.0 Albania 96.832549
71 DZA 6839.00 44761099.0 Algeria 152.788921
72 ASM NaN 49225.0 American Samoa NaN
73 AND 9.44 78364.0 Andorra 120.463478 
# Use Plotly to create an interactive choropleth map
fig = px.choropleth(
    df,
    locations="countryiso3code",
    color="Article Count per Million",
    hover_name="Country Name",
    hover_data={
        "Article Count per Million": True,
        "Article Count": True,
        "Population": True,
        "countryiso3code": False
    },
    color_continuous_scale="viridis",
    projection="natural earth",
    title="Technical Journal Articles by Country (2021)"
)

fig.show()