readoecd provides clean, tidy access to OECD economic data directly from R.

There is an existing R package called OECD, written by Eric Persson and first published on CRAN in 2016. It is a generic API wrapper - rather than providing named functions for specific datasets, it exposes the raw OECD SDMX API directly. The typical workflow requires the user to first browse available datasets with search_dataset(), then retrieve dimension codes with get_data_structure(), then construct a filter string manually before finally calling get_dataset(). This gives maximum flexibility but places the burden of knowing the OECD's dataset catalogue and dimension codes entirely on the user.

The CRAN version (v0.2.5, released December 2021) is currently broken. In July 2024, the OECD migrated their API from the old endpoint (stats.oecd.org) to a new SDMX 3.0 REST API (sdmx.oecd.org). The CRAN version still points at the old endpoint, which no longer returns data. Despite being broken, the package still attracts around 700–1,400 downloads per month - reflecting how many analysts depend on OECD data in R and have no working alternative.

Eric's GitHub version is partially updated (expersso/OECD). He updated the package in August 2025 to target the new API, but this version has not been submitted to CRAN and is not widely known. It remains a generic wrapper: it still requires users to know dataset IDs and construct their own filter strings.

readoecd takes a different approach. Rather than wrapping the raw API, it provides curated named functions for the most commonly needed OECD datasets. Users do not need to know anything about SDMX, dataset identifiers, or dimension codes - get_oecd_gdp(c("AUS", "GBR"), start_year = 2010) just works. Output columns are consistent across all functions, and data is cached locally so repeated calls are instant.

install.packages("readoecd")

# Or install the development version from GitHub
# install.packages("devtools")
devtools::install_github("charlescoverdale/readoecd")

Each function in readoecd returns a single, opinionated series - for example, get_oecd_cpi() returns the annual headline inflation rate, and get_oecd_health() returns total health expenditure as a share of GDP. If you need a different cut of the data, the best starting point is the OECD Data Explorer, which lets you browse and filter any OECD dataset interactively.

Examples of things you might want that go beyond what readoecd currently provides:

• CPI by expenditure category (food, housing, transport) rather than the headline total
• Health expenditure by financing source (government-only vs out-of-pocket) rather than the all-sources aggregate
• GDP in constant prices (volume, real growth) rather than current prices
• Tax revenue by tax type (income tax, VAT, corporate tax) rather than the total
• Unemployment by age group or sex rather than the all-persons total
• Trade in goods and services separately rather than the combined current account

library(readoecd)

Source: OECD National Accounts (Main Aggregates, Table 1). Annual GDP measured using the expenditure approach at current prices, in millions of US dollars at exchange rates. Coverage is broad - most OECD members have data from the 1990s, some earlier. GDP in USD PPP is used for cross-country comparisons where available; the package falls back to exchange-rate USD if PPP is not published for a given country-year.

# Largest OECD economies in the most recent year
gdp <- get_oecd_gdp("all", start_year = 2015)
latest <- gdp[gdp$year == max(gdp$year), ]
head(latest[order(-latest$value), c("country_name", "year", "value")], 5)
#>        country_name year    value
#>   United States     2023 27357916
#>         Germany     2023  4457178
#>           Japan     2023  4212945
#>  United Kingdom     2023  3081866
#>          France     2023  2923489

# G7 comparison since 2000
g7 <- get_oecd_gdp(c("CAN", "FRA", "DEU", "ITA", "JPN", "GBR", "USA"),
                   start_year = 2000)

Source: OECD Prices database (COICOP 1999 classification). Annual year-on-year percentage change in the Consumer Price Index for total household expenditure, not seasonally adjusted. This is the standard harmonised measure of headline inflation used for cross-country comparisons. Coverage varies by country but is generally available from the 1970s–1980s onwards.

# Inflation through the post-COVID price surge
cpi <- get_oecd_cpi(c("AUS", "GBR", "USA", "DEU"), start_year = 2018)
cpi[cpi$year == 2022, c("country_name", "year", "value")]
#>    country_name year value
#>       Australia 2022   6.6
#>         Germany 2022   8.7
#>  United Kingdom 2022   9.1
#>   United States 2022   8.0

Source: OECD Labour Force Statistics (Harmonised Unemployment Rates). Monthly seasonally-adjusted unemployment rate as a percentage of the labour force, for persons aged 15 and over. The harmonised series adjusts for differences in national survey methodologies, making it the standard series for cross-country comparisons. Coverage varies by country but generally starts from the 1980s–1990s.

# Track unemployment through the COVID shock
unemp <- get_oecd_unemployment(c("AUS", "GBR", "USA"), start_year = 2018)

# Peak unemployment by country during 2020
peak_2020 <- unemp[substr(unemp$period, 1, 4) == "2020", ]
aggregate(value ~ country_name, data = peak_2020, FUN = max)
#>    country_name value
#>       Australia  7.47
#>  United Kingdom  5.10
#>   United States 14.70

Source: OECD Revenue Statistics Comparator. Total tax revenue as a percentage of GDP, covering all levels of government (central, state, and local) and all tax types (income, payroll, property, goods and services, and other taxes). The standard measure of overall tax burden used in cross-country fiscal analysis. Annual data; most countries covered from the early 1990s.

# Tax burden: Nordics vs Anglo-Saxon economies
tax <- get_oecd_tax(c("DNK", "SWE", "NOR", "AUS", "GBR", "USA"),
                    start_year = 2010)

latest_tax <- tax[tax$year == max(tax$year), c("country_name", "year", "value")]
latest_tax[order(-latest_tax$value), ]
#>    country_name year value
#>         Denmark 2022  46.0
#>          Sweden 2022  42.6
#>          Norway 2022  42.2
#>  United Kingdom 2022  35.3
#>       Australia 2022  29.5
#>   United States 2022  27.7

Source: OECD National Accounts (General Government Accounts, NAAG). General government net lending/borrowing as a percentage of GDP, consistent with the System of National Accounts (SNA) definition. A positive value indicates a surplus; a negative value indicates a deficit. Annual data; most OECD members covered from the 1990s.

# Fiscal positions during and after the pandemic
deficit <- get_oecd_deficit(c("AUS", "GBR", "USA", "DEU"), start_year = 2018)
deficit[deficit$year %in% 2019:2022, c("country_name", "year", "value")]
#>    country_name year  value
#>       Australia 2019  0.02
#>       Australia 2020 -4.45
#>       Australia 2021 -4.97
#>       Australia 2022 -1.70

Source: System of Health Accounts (SHA). Total current health expenditure as a percentage of GDP, covering all financing sources - government schemes, compulsory health insurance, voluntary health insurance, and out-of-pocket payments. Based on the internationally standardised SHA 2011 framework developed jointly by the OECD, WHO, and Eurostat. Annual data; coverage typically from 2000 onwards.

# Which OECD country spends the most on health?
health <- get_oecd_health("all", start_year = 2015)
latest_hlth <- health[health$year == max(health$year), ]
head(latest_hlth[order(-latest_hlth$value), c("country_name", "year", "value")], 5)
#>    country_name year value
#>   United States 2022  16.6
#>         Germany 2022  12.7
#>          France 2022  11.9
#>     Switzerland 2022  11.3
#>       Australia 2022   9.7

Source: Education at a Glance (EAG) - UOE Finance indicators. Total expenditure on educational institutions as a percentage of GDP, aggregated across all levels of education (ISCED 0–8, from early childhood through tertiary). Covers both public and private expenditure. Published annually; data typically available with a two-year lag, covering most OECD members from the mid-2000s.

# Education spending trends across selected countries
edu <- get_oecd_education(c("AUS", "GBR", "USA", "KOR", "FIN"),
                          start_year = 2005)

latest_edu <- edu[edu$year == max(edu$year), c("country_name", "year", "value")]
latest_edu[order(-latest_edu$value), ]

Source: OECD Income Distribution Database (IDD). Gini coefficient of disposable household income (after taxes and transfers), on a scale from 0 (perfect equality) to 1 (maximum inequality). Follows the OECD METH2012 methodology for comparability across countries and over time. Annual data; coverage varies by country, typically from the mid-1990s.

# Income inequality: how do OECD countries compare?
gini <- get_oecd_inequality("all", start_year = 2010)
latest_gini <- gini[gini$year == max(gini$year), ]

# Most and least equal OECD economies
head(latest_gini[order(latest_gini$value), c("country_name", "year", "value")], 5)
head(latest_gini[order(-latest_gini$value), c("country_name", "year", "value")], 5)

Source: OECD Productivity Database (PDB). GDP per hour worked in USD PPP where available, falling back to GDP per capita. GDP per hour worked is the standard measure of labour productivity used for cross-country comparisons, as it adjusts for differences in average working hours across countries. Annual data; coverage varies by country.

# Productivity rankings across all OECD members
prod <- get_oecd_productivity("all", start_year = 2010)
latest_prod <- prod[prod$year == max(prod$year), ]
head(latest_prod[order(-latest_prod$value), c("country_name", "year", "value")], 5)

Source: OECD Balance of Payments Statistics. Annual current account balance with the rest of the world, in millions of US dollars at current exchange rates. A positive value indicates a surplus (exports exceeding imports of goods, services, income, and transfers); a negative value indicates a deficit. Coverage is annual; most countries available from the 1990s.

# Identify persistent surplus and deficit countries
ca <- get_oecd_current_account("all", start_year = 2010)
latest_ca <- ca[ca$year == max(ca$year), ]

# Largest surpluses
head(latest_ca[order(-latest_ca$value), c("country_name", "year", "value")], 5)

# Largest deficits
head(latest_ca[order(latest_ca$value), c("country_name", "year", "value")], 5)

# See all available country codes
list_oecd_countries()

# Clear the local cache to force fresh downloads
clear_oecd_cache()

All get_oecd_*() functions return a tidy long-format data frame with consistent columns:

Downloaded data is stored in your user cache directory (tools::R_user_dir("readoecd", "cache")). Subsequent calls return the cached copy instantly. Use refresh = TRUE to force a re-download, or clear_oecd_cache() to wipe all cached files.

readoecd uses the OECD Data Explorer REST API (sdmx.oecd.org). The OECD periodically migrates dataset identifiers - if a function stops working, check for a package update or file an issue.

OECD, SDMX, international economic data, GDP, unemployment, trade, education, health, environment, OECD Data Explorer, open data, cross-country comparison, R package

MIT