Skip to content

Chart ten years of G7 aid

By the end of this tutorial you'll have a chart of official development assistance (ODA) from each G7 country from 2015 to 2024, in constant 2024 dollars, and a note that says where every number came from.

What you'll learn

  • How to ask oda-data for an indicator, for chosen years and providers
  • How to read the columns every result has
  • How to convert to constant prices
  • How to add a group's total beside its members
  • How to get a plain-text description of a result

What you'll need

  • Python 3.12 or later
  • About 15 minutes, and about 10 MB of disk for the OECD's DAC1 table
  • matplotlib for the chart (pip install matplotlib)

Install oda-data:

pip install oda-data

Step 1: get one number

Start with one provider and one year. oda-data names each indicator by who defines it, the table it reads, and what it counts, so oecd.dac1.total_oda is the OECD's total ODA from its DAC1 table.

You should see one row:

  provider provider_name  year            indicator             basis     value
0        4        France  2024  oecd.dac1.total_oda  grant_equivalent  15448.26

import oda_data as oda

df = oda.get("oecd.dac1.total_oda", years=2024, providers="France")
print(df[["provider", "provider_name", "year", "indicator", "basis", "value"]])

The first call downloads DAC1 from the OECD and keeps it on disk, so it takes a few seconds. Later calls read the copy on disk.

Every result has the same columns. The dimensions come first, here provider and year, each code with its name beside it. Then come the indicator, its version, the basis, the unit (currency, prices, base_year), the scale, is_aggregate, and the value. Money is in millions, so France gave USD 15.4 billion in 2024.

Step 2: ask for ten years of the G7

Now widen the call. providers= takes codes, names, three-letter codes, and groups, and oda.groups.g7 is the seven G7 countries. years= takes a range.

The output should start like this:

  provider_name  year             basis     value
0        France  2015  net_disbursement   9039.30
1        France  2016  net_disbursement   9621.67
2        France  2017  net_disbursement  11330.86
3        France  2018  grant_equivalent  12136.17
4        France  2019  grant_equivalent  12211.36
5        France  2020  grant_equivalent  14124.70

g7 = oda.groups.g7

df = oda.get("oecd.dac1.total_oda", years=range(2015, 2025), providers=g7)
print(df[["provider_name", "year", "basis", "value"]].head(6))

Notice that the basis changes in 2018. The OECD counts ODA as net flows up to 2017 and as grant equivalents from 2018, and oda-data's default basis, headline, follows the OECD's own series. The basis column tells you which one each row holds.

Step 3: convert to constant prices

Money given in 2015 bought more than the same sum in 2024. To compare years, convert every row to constant 2024 US dollars. oda.constant("USD", 2024) builds that unit, and oda-data uses each provider's own deflator for each year.

  provider_name  year currency    prices  base_year         value
0        France  2015      USD  constant       2024  10496.283891
1        France  2016      USD  constant       2024  11148.510457
2        France  2017      USD  constant       2024  12805.872280

df = oda.get(
    "oecd.dac1.total_oda",
    years=range(2015, 2025),
    providers=g7,
    unit=oda.constant("USD", 2024),
)
print(df[["provider_name", "year", "currency", "prices", "base_year", "value"]].head(3))

France's 2015 ODA was USD 9.0 billion at the time and is USD 10.5 billion in 2024 dollars.

Step 4: add the G7 total

Pass oda.total(g7) beside the group to get one more row per year with the G7's total. Its is_aggregate column is True, so you can always tell a total from a member.

   provider provider_name  year  is_aggregate     value
70       g7  G7 countries  2015          True  111840.0
71       g7  G7 countries  2016          True  127607.0
72       g7  G7 countries  2017          True  130990.0

df = oda.get(
    "oecd.dac1.total_oda",
    years=range(2015, 2025),
    providers=[g7, oda.total(g7)],
    unit=oda.constant("USD", 2024),
)
totals = df[df.is_aggregate]
print(totals[["provider", "provider_name", "year", "is_aggregate", "value"]].round(0).head(3))

Step 5: draw the chart

Turn the members' rows into one column per country, then plot them. The total stays out of the chart, since it would dwarf every member.

import matplotlib.pyplot as plt

members = df[~df.is_aggregate]
wide = members.pivot(index="year", columns="provider_name", values="value")

ax = wide.plot(title="ODA from G7 countries, constant 2024 USD millions")
ax.set_xlabel("")
plt.savefig("g7-oda.png", dpi=150, bbox_inches="tight")

Open g7-oda.png. You should see seven lines, with the United States far above the rest and Germany second.

Step 6: describe the result

A chart in a report needs a note on its source. Build the same call with oda.query in place of oda.get. A Query runs once and keeps its result, and explain() describes it in plain text.

Total ODA (DAC1), headline basis: net flows 2015-2017, grant equivalent 2018-2024.
7 G7 countries, plus their total, which covers all 7 in every year it returns.
Constant 2024 USD using DAC deflators from the same DAC1 release.
DAC1 downloaded 2026-10-28. 80 rows.

q = oda.query(
    "oecd.dac1.total_oda",
    years=range(2015, 2025),
    providers=[g7, oda.total(g7)],
    unit=oda.constant("USD", 2024),
)
print(q.explain())

The note names the basis for each span of years, the providers, the unit, and the day the DAC1 table was downloaded.

What you learned

  • You asked for an indicator by id, with years= and providers=
  • You read the fixed columns, including basis and is_aggregate
  • You converted to constant prices with unit=oda.constant("USD", 2024)
  • You added a group's total with oda.total(group)
  • You described a result with oda.query(...).explain()

What's next