Industries / Government, education and community / Tribal enterprises
Governance
How Indigenous nations and their peoples set up, own and oversee the businesses they run, and how far elected leaders reach into day-to-day management.
6
Milestones achieved, from 1934 to 2025.
0
To watch for: announced, not yet confirmed outside.
2
Open deliverables. Nobody has shipped them.
The road
AchievedWatch forOpen
Achieved
1934
Federal charters for tribes
Section 17 of the US Indian Reorganization Act let the Secretary of the Interior issue a charter of incorporation to a tribe, giving the chartered corporation power to buy, own, manage and sell property once the tribe ratified it. A 1990 amendment let a tribe itself petition for a charter and its governing body ratify it, where before one-third of adult members had to petition and a reservation-wide vote had to approve. The Bureau of Indian Affairs (BIA) now lists three forms for a tribally owned business: a Section 17 corporation, a corporation chartered under the tribe's own law, or one chartered under state law, each with different consequences for tax, for protecting tribal assets and for what creditors and partners can see.
Credit
US Congress, Indian Reorganization Act of 18 June 1934
1971
Corporations in place of reservations
The Alaska Native Claims Settlement Act (ANCSA), signed on 18 December 1971, extinguished aboriginal land title in Alaska. In exchange it transferred 44 million acres and $962.5 million to Alaska Native corporations it created: twelve regional corporations and more than 200 village corporations, all for-profit and owned by enrolled Alaska Native shareholders, each of whom received 100 shares. A thirteenth regional corporation, for Alaska Natives living outside the state, followed in 1975, and sections 7(i) and 7(j) make the corporations share revenue from natural resources developed on their lands.
LessonThe law let shares be sold after 20 years, on the assumption that the corporations would be well established by 1991; when profits fell short of what Congress expected, amendments passed in 1988 kept the shares unsellable unless a corporation's shareholders vote otherwise, and the ANCSA Regional Association says none has.
Credit
US Congress; the Alaska Federation of Natives, whose delegates negotiated the settlement
1990
Boards apart from councils
In June 1990 Stephen Cornell and Joseph Kalt of the Harvard Project on American Indian Economic Development surveyed 18 tribal chairs at Northern Arizona University about their nations' enterprises. By the chairs' own reports, 34 of 39 enterprises run by independent management were profitable, odds of 6.8 to 1, against 20 of 34 run by the tribal council, odds of 1.4 to 1. A 2000 study by Miriam Jorgensen and Jonathan Taylor, using data on Indian-owned enterprises, found that enterprises under undue political influence, especially from elected officials sitting on their boards, frequently failed to thrive.
Credit
Stephen Cornell and Joseph P. Kalt; Miriam Jorgensen and Jonathan B. Taylor
1998
A settlement under iwi control
In 1996 work by the Ngāi Tahu Māori Trust Board produced Te Rūnanga o Ngāi Tahu, a tribal council with its own legal personality, accountable to Ngāi Tahu rather than to New Zealand's government. In September 1998 Parliament passed the Ngāi Tahu Claims Settlement Act, which gave the iwi (tribe) NZ$170 million, returned the mountain Aoraki, vested the pounamu (greenstone) in its territory in the iwi, and gave it first refusal on many Crown properties. Ngāi Tahu put its economic losses from the Crown's land purchases at more than NZ$20 billion. The iwi invests through separate commercial arms such as Ngāi Tahu Capital and Ngāi Tahu Farming, and in the 2017 financial year Te Rūnanga Group reported a net worth of NZ$1.36 billion and distributed almost NZ$50 million to tribal programmes and grants, by The Spinoff's account of the group's figures.
Credit
Ngāi Tahu; Tā Tipene O'Regan, chair of the Ngāi Tahu Māori Trust Board for more than a decade
2002
A First Nation certified
Membertou, a Mi'kmaw First Nation in Sydney, Nova Scotia, achieved certification to ISO 9001:2000, the international standard for quality management systems, in 2002. The Nova Scotia government calls it the first Aboriginal community in the world to achieve ISO certification. Chief Terrance Paul said the certification helped secure outside investors, because companies know it requires management processes to be in place.
Credit
Membertou First Nation; Chief Terrance Paul and chief executive Bernd Christmas
2025
Tribal businesses counted
In April 2025 the Center for Indian Country Development (CICD) at the Federal Reserve Bank of Minneapolis published first results from its Native Entity Enterprises Dataset: 5,559 business establishments owned by 344 federally recognized tribes. The list was built from tribes' own websites and other public lists between May 2021 and August 2023, with sector and employment taken from a subscription business database. Fewer than 5% of the 359 federally recognized tribal entities in the contiguous states did not appear; the median tribe owned nine businesses and one owned 184. CICD publishes only aggregated, anonymized results under its data principles.
LessonCICD calls the count almost certainly an undercount, because earlier data on these businesses were "scattered, incomplete, or focused solely on gaming".
Credit
Ava LaPlante, Vanessa Palmer and colleagues at CICD
Open
01
Code the published charters
A public reference set of the governance documents that Indigenous nations and their corporations have published themselves (corporate charters, business corporation and limited liability codes, articles of incorporation), each coded on fixed fields: legal form, who appoints directors, whether sitting elected officials may serve on the board, how profits pass to the government or shareholders, who may waive sovereign immunity, and what the enterprise must report and to whom. The set describes; it does not score or rank any nation.
The briefCredit · unclaimed02
Test AI on enterprise facts
A public test of at least 400 factual questions about how Indigenous-owned enterprises are set up and governed in the US, Canada, Australia and New Zealand (what a Section 17 charter does, who may own ANCSA shares, what limits apply to tribal tax-exempt bonds), each with an answer keyed to a statute or to a document the nation or corporation has published, plus a rubric that flags answers that drift from the legal question into generic cultural description. Run current AI models through it and publish the scores.
The briefCredit · unclaimed
Open 01 · the brief
Code the published charters
The first to publish is credited on this line, and the deliverable moves from Open to Achieved.
What to build
A public reference set of the governance documents that Indigenous nations and their corporations have published themselves (corporate charters, business corporation and limited liability codes, articles of incorporation), each coded on fixed fields: legal form, who appoints directors, whether sitting elected officials may serve on the board, how profits pass to the government or shareholders, who may waive sovereign immunity, and what the enterprise must report and to whom. The set describes; it does not score or rank any nation.
Why it is still open
The best-known figures on keeping politics out of enterprise management come from a 1990 survey of 18 tribal chairs, and CICD's 2025 dataset, the most complete list of tribally owned businesses yet, records sector, location and employment but not how each business is governed. Thomas Stratmann's Reservation Economic Freedom Index (Public Choice, 2024) classifies reservation institutions and scores reservations, but codes constitutions and laws, not enterprise charters. A search for a coded dataset of enterprise charters or business codes found none.
Built with what exists
Codes and charters that nations post themselves, such as those gathered in the Native Nations Institute's Constitutions Resource Center; the BIA's guide to tribal business structures as the starting list of legal forms; an AI model such as Claude to draft a first coding of each document, with every code checked by a person (the Note "Seven more flagged" covers an AI reviewer of long documents); Label Studio for two independent human coders; Python to measure agreement; the CARE Principles for Indigenous Data Governance for how the set is shared.
First step
Write the 12-field codebook from the three legal forms in the BIA guide and the management questions in Cornell and Kalt's survey, then code 10 business codes that nations have posted online and measure agreement between two coders.
Done looks like
At least 150 published documents from at least 60 nations or Indigenous-owned corporations in at least three countries, each coded on 12 fixed fields with page references; two coders agreeing on at least 90% of codes, with Cohen's kappa of 0.8 or more on every field; no scores or rankings; a published route for any nation to correct or withdraw its entries.
Open 02 · the brief
Test AI on enterprise facts
The first to publish is credited on this line, and the deliverable moves from Open to Achieved.
What to build
A public test of at least 400 factual questions about how Indigenous-owned enterprises are set up and governed in the US, Canada, Australia and New Zealand (what a Section 17 charter does, who may own ANCSA shares, what limits apply to tribal tax-exempt bonds), each with an answer keyed to a statute or to a document the nation or corporation has published, plus a rubric that flags answers that drift from the legal question into generic cultural description. Run current AI models through it and publish the scores.
Why it is still open
In April 2026 the Technology Policy Institute reported that AI models asked to describe tribal lands without detailed guidance wrote about heritage, spirituality and tradition instead of geography and infrastructure. The nearest published test, IndigiEval (2026), measures language and cultural proficiency in five Indigenous languages, not knowledge of how Indigenous-owned businesses work. No test of these business questions was found.
Built with what exists
Statute texts from Cornell's Legal Information Institute (25 USC 5124, ANCSA from 43 USC 1601, 26 USC 7871) and the official legislation sites of Canada, Australia and New Zealand; current models such as Claude, Gemini and Mistral Large 4 called through their APIs (the Note "A thousand questions, one pass" covers Gemini Decision 1, which answers up to 1,000 fixed-answer questions in one request, and "A trillion open parameters" covers Mistral Large 4); a Python grading script; two human reviewers.
First step
Write 50 questions with keyed answers from ANCSA, 25 USC 5124 and 26 USC 7871, run them through three models, and grade the answers by hand.
Done looks like
At least 400 questions with keyed answers and citations across four countries; at least five models scored, with accuracy per model and 95% confidence intervals; a separate drift rate per model; reviewer agreement of at least 90% on a 20% sample; no question about any nation's culture, membership or internal politics, or about its finances beyond what it has published.