How this was built

Where the data comes from

All data is sourced from City of Toronto records. But here's the twist: The City's Elections Financial Disclosure system currently only goes back to 2018, and the most recent Open Data contribution is also from 2018. For older elections, we're relying on archived copies of previous data files that were removed from the Open Data Portal. We kept copies, of course.

Alas, the previous Open Data records for campaign contributions did not attach any dates to the contributions. As such, only elections from 2018 forward include dates in the data.

What the City does and does not verify

Nothing here has been audited. Candidates submit their own list of contributors who donated more than $100 to campaigns. Errors are common. We noticed several places where postal codes and names included obvious typos.

Corporate and union money

Corporations and unions were banned from directly contributing to council and mayoral campaigns starting with the 2010 election. That did not cover school board trustee campaigns — they were still able to accept these kinds of donations until a province-wide ban took effect in 2018. Registered third party advertisers can still accept corporate and union contributions.

Postal codes and privacy

For privacy reasons, we only publish the first three characters — also known as the Forward Sortation Area, or FSA — of postal codes from individual contributors. For corporate, union and third party registrants, we publish the full address.

While the City publishes full postal codes in its data release, it only keeps the information available for a limited time. Since this data is intended to be a permanent record of campaign contributions, redacting the full postal code seemed like a reasonable concession for privacy.

Full postal codes are used to match likely duplicates, but full postal codes are redacted as part of the build process. They do not exist anywhere on this server or in the online version of this database.

How contributors are identified

Initially, contributors were matched when the full name and postal code were the same across multiple contributions. A secondary process then suggested other likely duplicates, based on factors like nicknames (Mike vs. Michael) and postal codes that closely matched. There were many cases where postal codes were the result of obvious data entry errors. For example, a number of 3s were entered as 8s. This is probably because handwriting can be messy.

Election results and money per vote

A person may seek only one municipal office at a time. Anyone here with two campaigns in the same year registered for the first, left before the nomination deadline, and registered for the second. Those are separate campaigns with separate money — contributions cannot be moved between them — and they are never combined.

A candidate whose name doesn't appear in the final results is our cue to conclude that the candidate ended their campaign before the final nomination deadline. These campaigns show no votes and no money-per-vote figure, because they couldn't have received any votes. Note that in some other cases, candidates may have informally ended their campaign after the nomination deadline — often by pledging their support for another candidate — but their name still appeared on the ballot.

Money raised per vote received is not campaign spending. This is just a simple calculation that divides the amount raised from contributions — including self-funding — by the number of votes received. This archive holds no record of how much money was actually spent, so we can't account for rare cases where candidates raised a lot of money but — for whatever reason — didn't actually spend it.

Where mayoral donors also gave

Election pages with a contested mayoral race show a grid of how each mayoral campaign’s donors spread their support across the other major mayoral campaigns in the archive: those with 500 or more individual donors to that campaign itself.

Only individual donors count. A candidate’s gifts to their own campaign are left out: they are flagged as the candidate’s own from 2010 on, and for 2006, where there is no such flag, a gift filed under the candidate’s own name is treated the same way. Donors are the same people the search shows, matched on name and postal code plus the records combined by hand. A donor who moved, or filed under another name, and has not been combined is counted as two people, so overlaps can be slightly understated.

Each cell compares the rate at which any two candidates received contributions from the same person. 2.0 means these two candidates shared common contributors at twice the normal rate, while 0.1 means they had fewer common contributors than a typical pair of candidates. Hover over or tap a cell for the share and the number of donors the two campaigns have in common. Someone who ran for mayor more than once is pooled across all those runs. Contributions they received when they ran for council or trustee (if applicable) are not included.

What this cannot tell you

• Spending. Only contributions are here. Expenses, surpluses and deficits are in the filed financial statements, which the City publishes as scanned PDFs.
• Rebates. Toronto refunds part of many contributions. None of that is reflected in these amounts.
• Contributions under the disclosure threshold, and anything a campaign did not file.
• Dates before 2018, and street addresses for individuals in any year — the City has never published either.
• Council by-elections before 2018. They predate the EFD system and are not in the Open Data releases either.
• Some French-language school board races. The 2006 and 2010 releases include no contributions to candidates for the Conseil scolaire catholique MonAvenir, and the City's online data has none for that board in any election since 2018. The few shown here were entered by hand from filed statements. In 2022, every French-board seat was either won by acclamation or had its election voided, so there are no vote counts for those races. The two voided seats were filled in a 2023 by-election.

What the totals include

• Only itemized contributions. Campaigns aren't required to report the names of contributors who give $100 or less. This database only includes contributors who gave a campaign more than $100 in total. As such, every total and every money-per-vote figure is lower than what the campaign took in.
• Less anything returned. Totals subtract contributions a campaign gave back. Campaigns often return donations if they're found to be against the rules. The City data only has records of returns from 2018 forward. The 2006-2014 releases don't indicate whether money was returned.
• Self-donations in 2006 are inferred. The 2006 data doesn't say which contributions came from candidates funding their own campaigns. When the contributor's name exactly matches the candidate's, the database counts that as a self-donation. In later elections, the City does flag candidate donations to themselves.

Corrections

Notice something wrong? We want to fix it. Get in touch with any errors or suggestions via email.

Special Thanks

The first version of Campaign Cash was developed in collaboration with Jeff McCartney. We thank him for all his work and for laying the foundation for this project.