The Concentration Economy: Who Wins Canada's Federal Contracts

Two markets wearing one budget line. Federal procurement is a single administrative system financing two economies that barely touch. The first is a megaproject economy: defence primes, shipbuilders, and systems integrators holding a small number of enormous, multi-year commitments. The second is a repetition economy: hundreds of thousands of modest awards — the under-$25K ticket, the annual service renewal, the local purchase — distributed across a vendor population of 183,024 registered names. The top of the market sets the headlines; the rest of it sets the experience of nearly every firm that has ever registered as a federal supplier.

The measurement. Ranking every vendor by cumulative value across the 2004–2026 corpus, the largest 100 firms account for 62.6% of all dollars and 12.8% of all awards. Inverted: 87.2% of federal contracts — the overwhelming majority — flow to firms outside the top 100, yet those contracts carry only 37.4% of the value. The implied average award differs by a factor of 11: $4.5M at the top versus $396K below it, against a corpus-wide average of $923K. Concentration of this shape is not cronyism or failure; it is what procurement for aircraft, ships and enterprise IT systems produces anywhere. But it does mean that "who wins federal contracts" has two correct answers, and most public discussion uses the wrong one.

The megaproject tier. The top of the table is a defence and infrastructure ledger. Parsons ($38.7B across 188 awards), the Irving Shipbuilding entries, Vancouver Shipyards, SkyAlyne ($22.4B across just 2 awards — aircrew training), Lockheed Martin and I.M.P. Group: seven of the ten largest vendors are defence or aerospace primes, reflecting the federal shipbuilding and fleet programs that have structured capital procurement since the early 2010s. Note the data's own honesty problem in this tier: the same firm appears under multiple registered names — two Irving Shipbuilding spellings ($44.4B combined over 8 awards), two Vancouver Shipyards registrations ($37.0B over 155), even two Groupe Signature entries — so true concentration at the very top is somewhat higher than any single-row ranking shows.

The recurring-vendor tier. Below the primes sits a different structure: firms that win constantly at mid-size. Bell (the telephone company) holds 21 awards totalling $11.5B; Microsoft 1,621 for $17.5B (an average of $10.8M per award); IBM 1,217 for $36.5B; Telus 1,290 for $10.7B. These are not megaproject wins but institutional positions — telecom, software and services relationships renewed across governments for decades. Their competitive significance is durable incumbency at scale, not any single award.

What concentration means for entry. For a firm considering the federal market, the strategic reading of these numbers is blunt. The value summit is effectively closed: it is won through program capture over years and defence-industry positioning, not through responding to tenders. The accessible market is the 87.2% of awards outside the top 100 — averaging $396K, overwhelmingly below $1M — where the cost of participation is modest, the buyer count is large, and the successful pattern in the vendor data is repetition: hundreds of firms hold federal positions built from exactly this kind of small-award accumulation. Concentration at the top and viability at the bottom are the same fact seen from two sides.

Policy reading. Concentration metrics of this kind are the baseline against which supplier- diversity and mid-market policies should be judged, because they define what the distribution actually looks like before intervention. Any policy aimed at "increasing competition" in federal procurement operates either on the thin head of this distribution — where entry is a decade-long proposition — or on its long tail, where competition is already the norm and the constraint is contract size, not access. The two require entirely different instruments, and the corpus makes it possible to say which is which.