Investigation · Series
How much Uruguay spends on gender and diversity policy
The procurement feed has no "gender policy" label: it has to be rebuilt from the text each civil servant typed. Doing so surfaces three things the public debate rarely mentions: the measurable spend is small, it is concentrated in a single victim-support network, and eight in ten contracts in the field carry no amount at all.
The starting point
Why you cannot just look this up
State procurement publishes what was bought, from whom and for how much — but not which policy it served. There is no "policy area" field, no buying body called Inmujeres (the National Women's Institute buys through the MIDES executing unit), and no catalogue heading that groups the subject. All that is left is the free text a civil servant typed when loading the call.
So it is rebuilt from there, with a published, auditable term list. Rebuilding from text has concrete traps that must be killed before adding up a single peso: in Uruguayan procurement Spanish "género" also means CLOTH, "trans" hides inside "transporte" and "transferencia", the "Plan de Equidad" was an unrelated MIDES cash transfer, and "diversidad" on its own drags in the National Biodiversity Directorate.
Finding 1
The spend is small and concentrated in one network
Most of what was detected is not training or campaigns: it is support for women experiencing violence. Montevideo's ComunaMujer network and the gender-violence response system account for the vast majority of the measurable money. A single buyer — the Montevideo city government — explains most of the total.
Set against what those same bodies spent on everything else, the subject is a very small fraction of their procurement budget. Intendencia de Montevideo: 77%. Violence victim services + ComunaMujer network (IM): 96%.
Composition of the measurable spend
Finding 2
Five organisations take almost all of it
On the supplier side the concentration is sharper still: five civil-society organisations receive the vast majority of the measurable spend. They are not companies but specialised NGOs and foundations, winning calls addressed to civil-society organisations by the Montevideo city government and MIDES. 98%.
This does not prove irregularity: in calls addressed to specialised CSOs the pool of possible bidders is small by definition. But it is a concentration figure worth reviewing, and every supplier links to its full record on the site.
Top suppliers
When a contract has more than one awardee the amount is charged in full to each: a supplier row is a ceiling, not a settlement.
Finding 3
The hole: eight in ten contracts carry no amount
This is the uncomfortable finding, and it is not about gender policy: it is about transparency. Most contracts in the topic appear in the feed with no amount loaded at all. The total we publish is therefore a floor, not a final figure — and there is no honest way to estimate the rest from open data.
421 / 528
contracts carry no amount in the feed
The gap is not evenly spread: it clusters in the older records and in the historic MIDES executing unit. Any figure circulating about "how much the state spends on gender" — higher or lower — is built on this same hole.
The political cross-reference
What can and cannot be said about parties
Each contract can be tagged with who governed the body in the year it was recorded, from the public electoral record. That is context, not attribution: nobody signs a contract "for" a party.
And one trap invalidates the raw ranking: the Montevideo city government explains most of the measurable spend and was governed by the same party throughout. A peso ranking by party measures who governs Montevideo, not who spends on gender. So the comparison is published normalised: how much of those bodies' own spending went to the topic.
- Frente Amplio
10.4of every 10,000
- Partido Nacional
2.3of every 10,000
Right now
What is out to tender this week
Closes 26/10/2026
Closes 29/10/2026
The method
How it was built, and what was thrown out
Two stages. First rules: a term list with context guards — the thing that stops "esterilla de género" or "metres of género for curtains" entering as public policy. Then a model reads the whole contract and decides whether it really belongs and in which category. Only what survives both stages counts.
The discards are kept and published, because a topic rebuilt from text is only auditable if you can see what was left out. Among the false positives the classifier removed: a purchase of worms "(GENERO EISENIA FOETIDA)", an air conditioner for a section called "Género" inside an HR directorate, and cloth bags.
Tell it where you can be protected
We publish what is in the open data. What isn't — the file, the message, what was said in the meeting — goes to Uruguay Leaks, the anonymous inbox run by la diaria, PODER and DATA Uruguay. The message is written for you: copy it, add what you know and paste it into their form.
Do not send it from your work computer or work email. Uruguay Leaks protects the submission, not the machine you send it from.
Uruguay Leaks is a project by la diaria, PODER and DATA Uruguay. We are not affiliated with them and receive no copy of what you send.
How to read this investigation
Every figure comes from State-procurement open data (OCDS) and is recomputed every Monday. The total is a floor: most contracts in the topic carry no amount in the feed.
The page does not say whether this spending is good or bad, nor whether the policies work. Procurement data holds purchases, not policy outcomes.
The governing party is shown as electoral context for the year of each contract. It does not imply a party decided, caused or executed the purchase.
The topic classification uses a language model in its second stage. It can be wrong; that is why each contract keeps the reason for its classification and the discard list is public.