There is a settled assumption in procurement analysis that data quality follows income: that the older and richer the administration, the more complete its public record. It is worth testing, because the field where it matters most produces the opposite answer. The number of bids a tender received is the single most diagnostic fact about a public market, and the places that publish it reliably are not the ones the assumption predicts.
Four records that answer the question
Ghana records a bidder count on every sampled notice and a value on every one as well. Liberia does the same and names the winner too, with its record denominated in US dollars, which removes even the currency problem that complicates comparison elsewhere. Guatemala reaches 99% on bid counts, the Dominican Republic likewise. These are complete records in the specific sense that matters: they allow an outsider to ask how contested a public purchase actually was, and to get an answer, which is a question most of the European Union cannot answer about itself.
Ranked by how often the bid count is actually recorded, the familiar reference markets sit at the bottom of the table.
| Market | Bid count recorded | Value recorded | Winner named |
|---|---|---|---|
| Ghana | 100% | 100% | — |
| Liberia | 100% | 100% | 100% |
| Guatemala | 99% | 68% | 100% |
| Dominican Republic | 99% | 100% | 21% |
| Lithuania | 22% | 45% | 86% |
| Germany | ~100% but 84% are placeholder zeros | 15% | 58% |
| Spain, Poland, France, UK, South Africa | 0% | varies | varies |
What the European record does instead
Set against that, Spain, Poland, France, the United Kingdom and South Africa record no bid counts at all. Lithuania manages 22%, which is thin but real and is currently the best figure in the union. Germany appears to be the exception at nearly full coverage until the distribution is examined, at which point 84% of the values turn out to be zeros standing in for ‘not stated’. The pattern is not that European records are worse overall — they carry deep histories, structured classifications and reliable identifiers that the four markets above do not. It is that they omit the one field that describes competition, which is the thing procurement rules exist to produce.
Why the assumption gets it backwards
The likely explanation is not diligence but sequence. Records built recently, often with development-bank involvement and a transparency mandate written into the design, capture the fields that modern practice considers essential — including bid counts — because the system was specified after the question became standard. Records built decades ago capture what mattered then, and adding a field to a mature national system is an expensive, slow, low-priority reform. Age produces depth and it produces inherited omissions, and in this particular field the omission is total. The consequence for anyone studying procurement competition across borders is uncomfortable: the strongest evidence comes from Accra, Monrovia, Guatemala City and Santo Domingo, and it cannot be compared with Madrid or Paris, because Madrid and Paris do not collect it.
The strongest evidence on procurement competition comes from Accra, Monrovia and Santo Domingo — and cannot be compared with Madrid, because Madrid does not collect it.
WHY IT MATTERS
Every cross-country statement about procurement competition is constrained by which markets record the bid count, and that set does not resemble the set of markets usually treated as the reference standard. Analysis that quietly drops the countries with the best data because they are unfamiliar is not being cautious; it is discarding the evidence. The practical use is narrower and more immediate: in four markets outside Europe it is possible to establish, before bidding, how many suppliers a buyer’s tenders typically attract — and in most of the European Union it is not possible at all.
Data: Otnox — live procurement intelligence across 56 markets. Basis: field completeness by market, measured on recent windows with distributions inspected, 2026-06-26..2026-07-26.
