The size of the universe
How many there are, where they are and what they do
For decades, thousands of Indigenous businesses moved this country without appearing in a single official figure. The Indigenous Registry puts them on the map: economic initiatives gathered from public and institutional sources, counted by region, by people and by sector. That is what makes it possible to say this economy exists with a figure behind it, rather than an impression. Counting it is the first step for it to grow: an economy nobody measures is an economy nobody budgets for.
17.221Indigenous economic initiatives identified in ChileA floor, not a total: it is built from freedom-of-information requests, and not every institution answers, nor in the same detail.
- Surveys
- 2022 · 2025
- Next
- October 2026
- Data
- Download the CSV
- Licence
- CC BY 4.0
- Source
- catastroindigena.com
What has been counted
- Registered ventures
- 16.922
- Legally constituted
- 579 · 3%
- Led by women
- 14.274 · 86%
- Regions with records
- 16
- Peoples represented
- 9
Calculated over the ventures that report this field in the source, not over the total.
The unit counted is the initiative, not the company: it may be an unregistered venture, a formal company, a community that produces, or a funded project. Counting all four as «companies» is what produces figures that cannot later be defended.
The figures are recalculated whenever a new survey is loaded. The Registry is permanent work, not a one-off load.
And there is a limit worth stating with the figure rather than in a footnote: this number is what has been identified, not the total Indigenous economy. It was built from 389 freedom-of-information requests to regional governments, ministries and municipalities, and not all of them answer, nor record the same things, nor record them in the same detail. Against a conservative estimate of the universe (built on SII's small-business register, DIPRES's entrepreneur figures and the 12.8% of the population who identify as Indigenous in the 2017 Census) the Registry sees between 45% and 56% of what exists. That estimate is indicative, not a census, and its assumption is optimistic: it takes for granted that the share of Indigenous businesses matches the share of Indigenous population.
This figure comes from the 2025 Annual Report, signed on 25 May 2026, and it is the one the Foundation uses in every external document. The tables below count 16,922, and the difference is not an error: the report publishes records as they arrived, while the database counts unique initiatives after merging 338 that came in twice from two different institutions and setting aside 23 that failed validation. The downloadable file adds up to 17,268, the cut prior to final homologation. These are three moments of the same count, and the full chain is published alongside the dataset.
The October 2026 survey corrects what this one leaves open: it adds productive-development agencies to sources that today are mostly social-development ones, and asks for five fields this registry never had. If the next number is higher, it will not mean the economy grew: it will mean the asking got better.
Why formalization comes first
The portrait the registry leaves undoes several assumptions at once: who starts businesses, how formal those businesses are, and how long they had remained invisible.
Most of what has been counted is real economic activity that is not constituted as a company. That is not a flaw in the count: it is the finding about the sources behind it, which are mainly social development programmes. A large company cannot buy from someone who cannot invoice, which is why the first stretch of the path the Foundation offers is not certification, it is legal constitution.
By region
- Arica985
- Tarapacá1.004
- Antofagasta469
- Atacama579
- Coquimbo367
- Valparaíso455
- Metropolitana2.204
- O'Higgins209
- Maule276
- Ñuble411
- Biobío1.334
- La Araucanía5.509
- Los Ríos590
- Los Lagos2.190
- Aysén391
- Magallanes295
The sixteen regions in geographic order, from Arica y Parinacota to Magallanes, with the 2025 survey. The depth of green accompanies the figure, it does not replace it: every cell carries its number. This is not a scale map, and deliberately so: on a faithful one La Araucanía would look smaller than Magallanes and say the opposite of the data.
- La Araucanía7925.509
- Metropolitana2792.204
- Los Lagos1622.190
- Biobío3041.334
- Tarapacá1581.004
- Arica y Parinacota36985
- Los Ríos103590
- Atacama170579
- Antofagasta234469
- Valparaíso192455
- Ñuble—411
- Aysén35391
- Coquimbo40367
- Magallanes—295
- Maule65276
- O'Higgins11209
- Otras (bajo el umbral de 5)12—
Each category shows its two surveys: 2022 in the lighter tone, 2025 in the solid one. Both bars share one scale, so their length can be compared across categories and across years. A dash means the category did not exist in that survey.
By people
- Mapuche1.82410.339
- Sin información4214.583
- Aymara671.212
- Diaguita38730
- Lickan Antai92154
- Colla41129
- Quechua—86
- Kawésqar—15
- Rapa Nui8911
- Chango—9
- Otras (bajo el umbral de 5)21—
20222025
Rows without this field in the source fall outside the breakdown, so this table sums to less than the overall total.
By sector
- Arte, artesanías, platería y cultura4324.727
- Comercio (menor, mayor y electrónico)2463.847
- Agricultura, agroalimentos, ganadería, silvicultura y pesca4272.746
- Sin información812.151
- Turismo y gastronomía3921.937
- Servicios técnicos y profesionales371.003
- Construcción e inmobiliaria36396
- Servicios de salud35111
Show the remaining 13 categories
- Otras actividades26998
- Asesorías y consultorías en gestión—72
- Transporte—54
- Servicios de educación1352
- Generación de energía—35
- Textil, vestuario y accesorios11219
- Tecnología, informática y comunicaciones—13
- Decoración, diseño y muebles747
- Bares, cafetería, gastronomía y restoranes369—
- Belleza y estética30—
- Bodegas, cervecerías y viñas24—
- Empresa Social7—
- Otras (bajo el umbral de 5)9—
20222025
The categories are the Foundation's own twenty, not the CIIU sections: what the official classification calls «manufacturing» is split here between crafts, textiles and agrifood, because in the Indigenous ecosystem that black box hides precisely what needs to be seen. Crafts and retail lead. Two categories from the 2022 survey do not appear in 2025, and it is not that they vanished: the taxonomy merged them, and the clearest case is food services, now counted together with tourism.
It did not grow sixfold in three years. It was always there
Between 2022 and 2025 the count went from 2,593 to 17,221 ventures, and the tempting reading is that the Indigenous economy grew almost sevenfold. It did not. What grew was the survey: more institutions responded, several improved their records, and sources that had not been approached in 2022 came in. The ventures were already there; what changed is that they can now be counted.
The proof that this is a matter of record keeping and not real change lies in where each figure came from. The 2025 survey brought in social development sources that on their own contributed close to twelve thousand ventures, none of them incorporated, because that is the population those programmes serve. That pushed formalisation down from 14.8% to 3.4% without a single business closing: leaving that source aside, the 2025 share stands at 10.9%, close to the 2022 figure. Female leadership rose for the same reason, since those programmes serve mostly women. The indicators did not move because the economy changed, but because who was asked did.
Saying this matters because the large figure invites the opposite conclusion. A count presented as growth reads as an economy that took off and no longer needs anything. What it shows is an economy that had been operating for decades without anyone counting it.
Why there are no names here
None of the people in this registry declared it to the Foundation. The data was recorded by whoever carried out each survey, which makes it identity attributed by a third party, not declared identity. Belonging to an Indigenous people is sensitive data under Law 21,719, and processing it requires prior, free, informed and specific consent. Publishing these records would use thousands of people’s data for a purpose none of them authorized.
For the same reason, no figure in these tables corresponds to fewer than five people. The smallest categories are grouped together, because in a small territory «three ventures from this people in this sector» is a first and last name to anyone who knows the place.
There is one more gap worth declaring: 26.5% of ventures have no identified people. It is not that they lack one, it is that the source they were drawn from did not record it. Future versions of the registry aim to close that gap, and until then any breakdown by people should be read against the remaining 73.5%.
Where this data comes from
From institutional surveys of the Indigenous entrepreneurship ecosystem, loaded by a process that records the source file, its hash and the date of each load. Every row keeps its source and its reference: the Registry rests on being able to say where each figure came from.
And there is a limit to that provenance worth declaring, because it changes how the informality figure reads. Sources do not contribute initiatives of every kind: each contributes the ones its programme serves. Social development programmes, which are most of what was gathered, register initiatives that are not constituted, because that is the population they serve; commercial registers and buyer supplier lists, by contrast, bring companies already formally constituted. That is why the 97% describes the initiatives these programmes identify, and not the Indigenous economy as a whole. The 2026 survey adds business development agencies to close that gap.
- Social development programme11.8640%
- Municipalities3.921fewer than 5
- CONADI8950%
- Commercial directory304100%
- A buyer supplier list138100%
- Our own 2022 survey104100%
- ProChile, business development42100%
formally constitutednot constituted
Each bar shows that source composition, not its size: the number of initiatives sits next to the name and every bar spans the full width. There are 7 sources and 17.283 records, and none of the 7 contributes both kinds.
The first three are programmes serving people who have not yet constituted a company, and they account for 96% of what was surveyed. The last four are registers that by definition hold only formally constituted businesses. The 97% informality figure does not describe the Indigenous economy: it is the weighted average of which agencies were asked.
This is the only breakdown on this page that crosses two dimensions. It is justified because the first one does not describe the people surveyed but the agency that registered them, and the smallest of the seven holds 42 records.
Surveys loaded: 2025
How this data is built
The registry does not come from a survey or from voluntary sign-ups. It is built by combining freedom of information requests to public institutions and municipalities with direct fieldwork, and that is why it reaches ventures no formal register captures.
How records are matched, classified, and what share is verified
An artificial intelligence process runs over that raw material, standardising and cleaning every record: region, municipality, people, gender, formalisation and sector. Without that layer the same municipality appears spelled five different ways and the count becomes useless.
Sector is classified twice on purpose: under CIIU, the official classification of the Chilean tax authority, and under the Foundation's own twenty categories, which are legible and culturally relevant. The first allows cross-referencing with state data; the second allows an artisan to recognise her own trade on the list.
22.5% of the sector data is verified against the tax registry, and 3.4% of ventures have formally registered activity. Both figures are low, and stating them is part of the method: a registry that does not declare how much it could verify is not a registry, it is an estimate that looks like a census.
Who funded each registry
The first registry, in 2022, was carried out under the project «U.S. Models Promote Indigenous-Owned Business Development», with support from the Embassy of the United States in Chile. The second, in 2025, was funded by Fundación Empresas Indígenas with its own resources.
Why the distinction is not a technicality
The distinction is not an administrative technicality. The first count of Chile's Indigenous economy was paid for by a foreign embassy; the second was paid for by the Indigenous institution itself, and it counted more than six times as much. It is the same economic autonomy this ecosystem advances, applied to the body that runs it.
From data to autonomy
Counting is not the end. The registry is the first step of a path the Foundation walks with each initiative, and every stretch ends in something its owner can use.
- Know
- Count, register and characterise every Indigenous economic initiative in the territory.
- Assess
- Identify gaps and opportunities in management, formalisation and market access.
- Develop
- Co-design action plans, training and financing with cultural relevance.
- Deliver
- Certify, connect with buyers and report the impact generated.
What to do with this evidence
The registry exists to inform decisions, and each actor has a different one at hand.
- The State
- Design an Indigenous public procurement programme and a simplified, territorially grounded route to formalisation. The 97% informality recorded by social development programmes is not a lack of economic activity: it is a gap in access.
- Private sector
- Bring Indigenous suppliers into value chains through supplier development programmes and in-house fairs.
- Support ecosystem
- Create a mixed public, private and Indigenous fund to finance the growth of Indigenous businesses at every stage.
Download the data
The count is available as an open file, aggregated by region, people, sector, gender and formalisation, covering both surveys. It is published so that anyone can work with the figure without having to request it.
What you download are counts, never individual rows, and no cell with fewer than five ventures: the file goes through the same k-anonymity rule as this page. The source data holds no names and is still not publishable, because crossing municipality, people, gender and sector leaves one thousand six hundred and seventy four cells with a single venture, and in a small municipality that amounts to a name.
Download the CSVRead the methodological note
The data is published under a CC BY 4.0. And the code that produces it is versioned and public: the figure does not have to be taken on trust, it can be reproduced.
To appear by name, that is the Directory
The Registry counts; the Directory identifies. The difference between the two is consent: an initiative appears in the Directory because its owner decided to, chose what is published about it, and can withdraw it at any time.
