Data

All the data behind Scamstats is free to download and reuse. These files were made on 2026-11-20 and are remade every time the site is updated.

Downloads

Each country page can also be downloaded as a PDF, from the button under the country's name.

For programs

The same figures and source lists as the pages, as JSON for programs to read. Each address has a schema that lists its fields.

AI connector

Scamstats has a connector that AI assistants can use to answer questions with Scamstats figures and sources. It is free, read only and needs no sign-in.

The connector address is:

https://scamstats.org/mcp

Add it to Claude

In Claude, open Settings, then Connectors, choose Add custom connector, and paste the connector address. Claude can then answer questions about scams with Scamstats figures and sources.

Add it to ChatGPT

In ChatGPT, turn on developer mode in the settings, add a connector, and paste the connector address.

For developers

The connector is a Model Context Protocol server over Streamable HTTP. It is stateless and read only. It supports protocol versions 2025-06-18 and 2025-03-26. Any assistant that can use a remote connector can use the same address.

ToolWhat it doesExample question
List countries
list_countries
Lists the countries Scamstats covers, with the people scammed in a year and the share of adults in each.Which countries does Scamstats cover?
Get a country
get_country
Gives one country's figures, scam types, money returned and sources, from its two-letter code such as PE.How many people in Peru are scammed each year?
Compare countries
compare_countries
Compares one measure across up to 20 countries, and names any differences in how the figures were counted, such as different periods.Compare money lost in Peru, Mexico and Colombia.
Get the world estimate
get_world
Gives the world estimate: the people scammed in a year, the share of adults and the money lost.How many people worldwide are scammed each year?
Get a scam type
get_scam_type
Gives one scam type's people and money lost in every country Scamstats covers, from its code such as investment.How much is lost to investment scams in each country?
Search sources
search_sources
Searches the catalogue of scam data sources by country, publisher, topic or what is measured. Give at least one of query, area or measure.What scam data does the Canadian Anti-Fraud Centre publish?
Get a source
get_source
Gives one catalogue entry in full, with its editions and how Scamstats uses it, from its id.Tell me about the Federal Trade Commission's Consumer Sentinel data.

Citing

Every answer ends with this line:

Cite as: Scamstats (Armidale), scamstats.org

The figures are the ones on the country pages. A guide for AI assistants is at /llms.txt.

Snapshots

Each week the data is checked, and whenever a figure has changed a copy of the full data set is kept on Zenodo, so a figure can be cited as it stood on that date.

Every scam data source we know of is listed on the sources page.

What changed in each release is on the change log.

Licence and citation

Our data is licensed CC BY 4.0: you may copy, share and adapt it for any purpose, as long as you credit Scamstats. Figures taken from other publishers are credited to them in the list of sources and remain theirs.

Scamstats (Armidale), scamstats.org, data of 2026-11-20.

The country table

Shares are fractions (0.104 is 10.4%). Money is not rounded.

ColumnWhat it holds
iso2Two-letter country code (ISO 3166)
nameCountry name
yearThe year the figures are for
baseWhere the shares come from: a non-government survey, a government survey, or estimated
populationTotal population (United Nations)
adultsPeople aged 18 and over (United Nations)
gdp_usdGross domestic product in US dollars (World Bank)
gdp_per_person_usdGDP divided by population, in US dollars
share_scammedShare of adults scammed in 12 months
people_scammedAdults scammed in 12 months
share_lost_moneyShare of victims who lost money
people_lost_moneyVictims who lost money
share_reported_anyShare of victims who reported to any institution
reported_any_basisWhether that share was measured or estimated
share_told_policeShare of victims who told the police
people_told_policeVictims who told the police
cases_recordedCases recorded by the authorities in the year
currencyThe local currency (ISO 4217)
money_lostMoney lost to scams in the year, in the local currency
money_lost_usdMoney lost to scams in the year, in US dollars
money_lost_share_of_gdpMoney lost as a share of GDP
average_lossMoney lost divided by victims who lost money, in the local currency
average_loss_usdThe same in US dollars
average_loss_share_of_gdp_per_personThe average loss as a share of GDP per person
money_recoveredMoney returned to victims in the year, in the local currency (refunded or recovered, not merely frozen)
money_recovered_usdThe same in US dollars
share_recoveredMoney returned to victims as a share of money lost
recovered_basisWhether the money returned was published or estimated
unrecorded_loss_priceThe loss counted for a victim missing from official records, in the local currency
unrecorded_loss_price_usdThe same in US dollars
unrecorded_loss_basisWhether that loss was measured in the country or fitted from other countries
money_methodHow money lost was worked out
most_common_1The most common scam type
most_common_2The second most common scam type
most_common_3The third most common scam type

The scam type table

types.csv, in the full data set, holds people and money lost by scam type. Money by type adds up to the money lost.

ColumnWhat it holds
iso2Two-letter country code (ISO 3166)
nameCountry name
yearThe year the figures are for
codeThe scam type code; methods is stolen logins and devices
labelThe scam type
peopleAdults hit by the scam type in 12 months
currencyThe local currency (ISO 4217)
moneyMoney lost to the scam type in the year, in the local currency
money_usdThe same in US dollars
basisMeasured when the country's own data gives the figure, borrowed when other countries' figures do
donorsThe countries a borrowed figure came from, as two-letter codes separated by semicolons

The full data set

FileWhat it holds
types.csvPeople and money lost by scam type for every country, beside the country table
countries.csvThe countries on the site and their status
sources.csvEvery source cited, with publisher, title, link and date
series.csvEach series of figures: what it measures, for which country, from which source
observations.csvThe figures themselves, one row per series and period
assumptions.csvValues that are assumed or carried over, and where each comes from
estimates.csvHow each country estimate is worked out
categories.csvThe scam types and methods, with definitions and examples
category_map.csvHow each source's own categories map to ours
descriptions.csvThe text on each country page, with its sources
population.csvPopulation and adults by country (United Nations)
world.csvWorld population and adults (United Nations)
gdp.csvGDP by country and for the world (World Bank)
income-groups.csvWorld Bank income group of every country, for the world estimate
fx.csvExchange rates to the US dollar (World Bank)

How the figures are worked out.

Put the map on your site

The world map can be embedded in another page. Use the embed button beside the search box on the front page to copy the code.