Main Abstraction2026-09-03

Public data · 5,343,805 awards

Where public contracts actually change hands

In Lithuania, 82.7% of public contracts go to a supplier the buyer has already used. In Switzerland, 17.5%. Same rules, same continent.

SourceEU public award notices
Period2016 to 2023
Awards5,343,805
Countries27, each with 20,000+ awards

A firm deciding whether to bid is really asking one question: does this buyer already have someone? Across 5,343,805 award notices the answer varies more by country than by anything else we measured, and it varies by a factor of four and a half.

The two numbers that decide whether bidding is worth it

20%40%60%80%10%20%30%40%50%60%Lithuania: 82.7% repeat, 23.6% single-bid, 130,505 awardsLatvia: 81.9% repeat, 24.7% single-bid, 104,108 awardsRomania: 79.9% repeat, 26.2% single-bid, 562,733 awardsSlovenia: 79.8% repeat, 57.2% single-bid, 221,281 awardsCzechia: 77.4% repeat, 48.7% single-bid, 213,399 awardsCroatia: 76.8% repeat, 27.0% single-bid, 73,362 awardsBulgaria: 69.5% repeat, 35.4% single-bid, 145,385 awardsEstonia: 68.5% repeat, 22.8% single-bid, 37,970 awardsSlovakia: 63.7% repeat, 28.3% single-bid, 32,129 awardsHungary: 60.1% repeat, 39.7% single-bid, 73,140 awardsPoland: 59.7% repeat, 51.1% single-bid, 1,044,803 awardsIreland: 52.9% repeat, 11.1% single-bid, 32,138 awardsUnited Kingdom: 51.5% repeat, 20.6% single-bid, 218,454 awardsDenmark: 50.9% repeat, 18.8% single-bid, 55,665 awardsSpain: 45.2% repeat, 27.6% single-bid, 296,877 awardsPortugal: 44.9% repeat, 28.5% single-bid, 50,823 awardsGreece: 44.3% repeat, 40.1% single-bid, 60,312 awardsFinland: 40.1% repeat, 12.6% single-bid, 64,661 awardsGermany: 38.9% repeat, 23.3% single-bid, 531,458 awardsSweden: 37.8% repeat, 8.8% single-bid, 138,738 awardsAustria: 37.2% repeat, 25.7% single-bid, 38,690 awardsItaly: 35.8% repeat, 33.0% single-bid, 190,702 awardsBelgium: 35.8% repeat, 22.0% single-bid, 59,063 awardsFrance: 31.7% repeat, 18.5% single-bid, 824,583 awardsNetherlands: 30.6% repeat, 17.4% single-bid, 74,062 awardsNorway: 27.5% repeat, 11.8% single-bid, 40,136 awardsSwitzerland: 17.5% repeat, 10.4% single-bid, 28,628 awardsLTLVROSICZHRBGEESKHUPLIEUKDKESPTGRFIDESEATITBEFRNLNOCH Share of awards going to a supplier the buyer had used before Share of awards with a single bidder
Each point is a country. Right means the buyer usually returns to a supplier it already knows. Up means nobody else turned up. The bottom left is where an outsider has a real chance.

The two measures move together, and the countries separate into groups that a bidder should treat completely differently. Where both numbers are high, a procedure is frequently a formality with one participant and a familiar winner. Where both are low, the buyer runs a genuine contest and an unknown supplier is reading the same documents as everyone else.

Portugal sits at 44.9%, in the middle, which is the more interesting position: enough repeat business to show that relationships matter, and enough turnover to show that the door is not shut.

Every country measured

CountryAwardsRepeat supplierSingle bidder
Lithuania130,50582.7%23.6%
Latvia104,10881.9%24.7%
Romania562,73379.9%26.2%
Slovenia221,28179.8%57.2%
Czechia213,39977.4%48.7%
Croatia73,36276.8%27.0%
Bulgaria145,38569.5%35.4%
Estonia37,97068.5%22.8%
Slovakia32,12963.7%28.3%
Hungary73,14060.1%39.7%
Poland1,044,80359.7%51.1%
Ireland32,13852.9%11.1%
United Kingdom218,45451.5%20.6%
Denmark55,66550.9%18.8%
Spain296,87745.2%27.6%
Portugal50,82344.9%28.5%
Greece60,31244.3%40.1%
Finland64,66140.1%12.6%
Germany531,45838.9%23.3%
Sweden138,73837.8%8.8%
Austria38,69037.2%25.7%
Italy190,70235.8%33.0%
Belgium59,06335.8%22.0%
France824,58331.7%18.5%
Netherlands74,06230.6%17.4%
Norway40,13627.5%11.8%
Switzerland28,62817.5%10.4%

How this was measured

An award counts as a repeat when the winning supplier had already won at least one earlier award from the same buyer inside the panel. Buyers and winners are identified by national registration number where the notice carries one and by name where it does not, because identifier coverage varies enormously between countries. Countries with fewer than 20,000 awards are excluded, since the rate is unstable below that.

What this does not show. A repeat award is not evidence of anything improper. A buyer returning to a supplier that performed well is ordinary and often correct. The number measures how often the situation occurs, not why. Name matching will merge two firms that share a name and split one that changed its own, and the error is larger in countries where registration numbers are rarely published. Contract values are not used anywhere on this page, because the source mixes currencies without conversion. The average number of bidders is also left out: a small number of procedures report implausibly large counts and the median would be the honest statistic, which this panel does not support.