The Government of Canada disclosed 59,637 contracts in 2025, worth $34.5 billion. It awarded $4.40 billion of that without competition, or about one dollar in eight.
That figure on its own says very little. Governments have legitimate reasons not to run a competition, and the disclosure records those reasons. The useful questions are which spending went uncompeted, by whom, and on what stated grounds.
Read that way, the data tells two stories that point in opposite directions.
Every figure here counts each contract once. The disclosure publishes a new row each time a contract is amended, and each row repeats the contract's whole value, so adding the rows counts the same money several times over. Section 6 explains what that costs.
How federal contracting is awarded
| Procedure | Contracts | Value | Share |
|---|---|---|---|
| Competitive, traditional | 28,614 | $23.63B | 68.6% |
| Competitive, open bidding | 8,075 | $5.13B | 14.9% |
| Non-competitive | 19,247 | $4.40B | 12.8% |
| Competitive, selective tendering | 3,277 | $1.16B | 3.4% |
| Advance contract award notice | 424 | $0.15B | 0.4% |
Non-competitive contracts are 32% of all contracts but only 12.8% of the money. They are mostly small.
Where the sole-sourcing is
Two departments hold three-quarters of it, and they hold it for opposite reasons.
| Department | Sole-sourced | Contracts | Share of its own spending | Share of all sole-sourcing |
|---|---|---|---|---|
| National Defence | $2.09B | 1,755 | 12% | 47.6% |
| Shared Services Canada | $1.14B | 269 | 55% | 25.9% |
| Fisheries and Oceans Canada | $0.15B | 3,240 | 2% | 3.3% |
| Public Services and Procurement Canada | $0.14B | 1,470 | 7% | 3.2% |
| Royal Canadian Mounted Police | $0.14B | 1,253 | 16% | 3.2% |
| Global Affairs Canada | $0.07B | 1,749 | 18% | 1.6% |
National Defence is the largest sole-source spender in dollars, and it is also the department that competes most of its work. It signed 19,780 contracts worth $17.07 billion, more than any other department, and only 12% of that money went out without competition. Its $2.09 billion is large because the department is large.
Shared Services Canada is the most concentrated. It buys and runs the government's information technology. Fifty-five percent of its spending is awarded without competition, through only 269 contracts, and 97% of that value is claimed under exclusive rights. That is the signature of enterprise software and hardware licensing, where the vendor of a proprietary product is the only supplier who can legally provide it.
The reasons given point two ways
Departments must record why competition was skipped. Those reasons invert depending on whether you count dollars or contracts.
| Stated reason | Contracts | Value | Share of value |
|---|---|---|---|
| Exclusive rights | 2,318 | $2.215B | 50.3% |
| No reason recorded | 16,457 | $1.091B | 24.8% |
| Extreme urgency | 54 | $0.724B | 16.5% |
| Exceptionally advantageous conditions | 77 | $0.308B | 7.0% |
| All other reasons | 341 | $0.065B | 1.5% |
By value, half of all non-competitive spending is claimed under exclusive rights, meaning the supplier is the only one who can legally provide the thing. That is the strongest justification available, and it is the largest single category.
By count, the picture reverses. 16,457 of the 19,247 non-competitive contracts, or 86%, record no limited-tendering reason at all. They are small: $1.09 billion between them, an average of about $66,000 each.
Federal sole-sourcing is therefore two different practices wearing one label. A few thousand contracts with a strong stated legal basis account for most of the money, and a long tail of small ones with nothing stated accounts for most of the paperwork.
Sole-source vendors are more likely to be registered lobbyists
Contracts and the Registry of Lobbyists are published separately and are not normally read together. Joined on the organisation, a pattern appears.
| Vendors | In the lobbying registry | |
|---|---|---|
| Holding a sole-source contract | 9,270 | 186 (2.01%) |
| Competitively awarded only | 8,957 | 77 (0.86%) |
Sole-source vendors are 2.33 times more likely to appear in the registry.
The obvious objection is size: large firms both lobby more and win more work. So the same comparison, banded by each vendor's total contract value.
| Vendor size | Sole-source | In registry | Competitive only | In registry | Ratio |
|---|---|---|---|---|---|
| $10,000 to $100,000 | 6,909 | 0.91% | 6,157 | 0.44% | 2.08× |
| $100,000 to $1 million | 1,660 | 2.71% | 1,963 | 1.07% | 2.53× |
| $1 million to $10 million | 542 | 8.67% | 683 | 2.64% | 3.29× |
| Over $10 million | 153 | 20.26% | 87 | 10.34% | 1.96× |
The gap holds in every band. It is widest among vendors between $1 million and $10 million, where a sole-source holder is more than three times as likely to be a registered lobbying client. It narrows again at the top, where a fifth of sole-source vendors and a tenth of competitive ones are in the registry: at that size most firms have a government relationship worth registering either way.
What this does not show
It does not show that lobbying causes sole-sourcing. A plainer explanation fits the same data. Sole-sourcing usually happens because a supplier is uniquely positioned, through proprietary technology, incumbency or security clearance, and firms in that position are also the firms with a government relationship worth registering. The reasons data supports that reading directly: half the value is claimed under exclusive rights.
It is an association measured over one window, and not a sequence. This archive holds lobbying from 2025 onward. Federal procurement lead times are longer than that, so most contracts starting inside the window were tendered before the lobbying record begins. Nothing here establishes what came first, and the archive cannot currently answer that question.
Vendor names are matched by normalised string, not resolved identity. The 2.01% and 0.86% figures are floors. Proper entity resolution would raise both, and would most likely leave the ratio between them broadly intact.
A blank reason field is not the same as no reason. It may reflect a dollar-value threshold, a reporting practice, or a genuine omission. The archive records what departments published, and cannot say which of those applies.
Contract value is not spending. Values are contract totals, often multi-year, and the month a contract is disclosed is not the month the money moved.
Why the totals here are lower than a spreadsheet gives
This is the one figure worth checking against any other account of the same data, because the disclosure invites a specific mistake.
A contract gets a new row every time it is amended, and each new row repeats the whole value of the contract rather than the amount it changed by. Adding the rows therefore counts the same money once per amendment. The 67,405 published rows sum to $57.7 billion; the 59,637 contracts behind them are worth $34.5 billion. The difference, $23.2 billion, is the same money counted more than once.
One agreement shows the size of the effect. Shared Services Canada holds a software licensing contract with IBM Canada worth $814 million. It appears in the records ten separate times, and those ten rows add up to almost $8 billion. Counted by rows, that single contract would be larger than all non-competitive spending in this report.
Counting rows also moves the conclusions, not only the totals. On a row count Shared Services Canada looks like two-thirds of all federal sole-sourcing. Counted once per contract it is a quarter, and National Defence is the largest.
To count contracts properly, group the rows by procurement_id, which
identifies the contract. The records also carry reference_number, which
identifies the filing and changes with every amendment.
Method
Counted from 67,405 contract disclosure rows in the KnowDrive Canada store, reduced to 59,637 distinct contracts by grouping on procurement identifier, department and vendor, and taking each contract's latest reported value. Procedure and limited-tendering codes are decoded from the Government of Canada contracts data schema published at open.canada.ca.
The lobbying comparison uses 3,177 organisations from the Registry of Lobbyists' Monthly Communication Reports, 2025-01-01 onward, matched to vendors by normalised organisation name. One size band, vendors under $10,000, is omitted as too small to interpret.
Contains information licensed under the Open Government Licence – Canada, and under the Open Data Licence Agreement, Office of the Commissioner of Lobbying.