Keyboard shortcuts

Press or to navigate between chapters

Press S or / to search in the book

Press ? to show this help

Press Esc to hide this help

The Three-Circle Model

Concept

Vocabulary that names a phenomenon.

The Davis-Tagiuri map of family enterprise as three overlapping systems: family, ownership, and business, with seven role sectors where conflict and authority concentrate.

Also known as: Davis-Tagiuri model, three-circle model of the family business system, family-ownership-business model.

The Three-Circle Model is simple enough to sketch in thirty seconds and durable enough to explain a large share of family-enterprise conflict. Draw three overlapping circles. Label them family, ownership, and business. Then ask which circle a person is speaking from. The answer often explains why the same sentence can sound reasonable to one branch, self-serving to another, and operationally impossible to the office staff asked to carry it out.

What It Is

The Three-Circle Model was developed by Renato Tagiuri and John A. Davis at Harvard Business School in 1978 and later elaborated in Davis’s doctoral work and the family-business literature that followed. It depicts the family enterprise as three systems that overlap but do not collapse into one another.

The family circle includes relatives whose identity, history, and belonging give them a legitimate claim on family matters. The ownership circle includes people or entities that hold economic rights: shares, partnership interests, trust beneficial interests, voting control, or other ownership claims. The business circle includes the operating system: executives, employees, directors, family-office staff, foundation staff, and other people accountable for running something.

Where those circles overlap, the model creates seven role sectors:

SectorRole positionTypical voice in the room
1Family member only“This affects who we are as a family.”
2Owner only“This affects my economic rights.”
3Business participant only“This affects how the organization runs.”
4Family owner, not employee“I inherit and vote, but I don’t work here.”
5Owner-employee, not family“I own a stake and work here, but I’m not kin.”
6Family employee, not owner“I work in the system, but I don’t yet hold economic control.”
7Family owner-employee“I am kin, capital, and management at once.”

The point is not the diagram. The point is role clarity. A cousin can be a beloved relative, a minority owner, and an underqualified job applicant in the same week. A founder can speak as parent, controlling shareholder, board chair, and CEO in the same meeting. If no one names which voice is active, the family mistakes structural conflict for personality conflict.

For a family office, the “business” circle may be the operating company that created the wealth, the office itself, a foundation, a private trust company, or a portfolio of closely held entities. That variation matters less than the structural fact. The family office is rarely a single-circle institution. It serves family members, reports to owners or trustees, and runs as an organization with staff, budgets, policies, and risk controls.

Why It Matters

The model matters because many wealthy families try to solve three-circle problems with one-circle language. They say “the family decided” when only the founder decided. They say “the owners want” when the loudest family branch wants. They say “the office thinks” when staff are repeating what they believe the principal expects. That vagueness feels polite until the first serious conflict arrives.

The model gives the family a neutral way to name the conflict. A family member who is not an owner can have a legitimate moral concern and no formal vote. A non-family executive can have operational authority and no claim on family purpose. A family owner who doesn’t work in the business can have economic rights but no right to direct employees. Naming those differences doesn’t remove tension. It keeps the family from pretending every tension is disloyalty.

The model also explains why the governance stack needs more than one room. The Family Council is not a softer version of the board; it is the body for the family circle. The investment committee is not the family council with spreadsheets; it is a delegated owner or office body. The Family Constitution is not a sentimental document; it is the family-level agreement that explains which circle has authority over which kind of question. The Decision Rights Charter is where the model becomes routing logic.

Without the model, a family office often falls back to founder gravity. The founder can sit at the center of all three circles: parent, controlling owner, and business leader. That may work for a while. It doesn’t scale across branches, spouses, trustees, rising-generation members, or non-family executives. The model makes the bottleneck visible before the founder’s health, patience, or judgment becomes the only routing system the family has.

How to Recognize It

You are using the model well when role language becomes ordinary in meetings:

  • “Are we discussing this as family, as owners, or as foundation trustees?”
  • “This is a business decision below the CEO threshold, not a council decision.”
  • “The cousin has a family voice here, but the employment decision sits under the family employment policy.”
  • “The trustee has legal authority, and the family council has consultative authority. Those aren’t the same thing.”

You are missing the model when meetings blur those roles. The clearest signal is the phrase “the family” doing too much work. The family wants a liquidity event. Which family members? Owners or non-owners? Voting trusts or individual beneficiaries? Adults or the founder’s branch? Staff often know the difference before principals do, because staff have to execute the ambiguity.

A second signal is decision drift between bodies. The family council debates manager selection. The investment committee adjudicates cousin employment. Foundation trustees become the forum for unresolved sibling conflict. The founder’s assistant becomes the real decision register because every formal room is confused about its own jurisdiction. The model doesn’t solve the drift by itself, but it shows which room is being misused.

A third signal is resentment at an overlap sector. Family employees who don’t yet own feel watched by relatives who do. Family owners who don’t work in the office feel shut out by those who do. Non-family executives with real competence feel second-guessed by family members with no operating role. These resentments are not incidental. They are the seven-sector map showing up in lived form.

How It Plays Out

Consider a $900M family enterprise built around a regional food-distribution company, a $160M foundation, and a twelve-person family office. G1 still chairs the holding-company board. G2 includes three siblings: one CEO of the operating company, one foundation chair, and one passive owner who lives across the country. G3 includes nine adults, two of whom work in the family office and none of whom own voting shares yet.

The office is preparing a $40M liquidity event from a minority recapitalization. The CEO sibling wants most proceeds retained by the operating company for warehouse automation. The foundation-chair sibling wants a $15M transfer into the foundation to fund a rural-health initiative. The passive-owner sibling wants distributions to branches because several cousins are carrying personal debt after a real-estate loss. The founder wants to “keep everyone aligned” and asks the family office COO to draft a recommendation.

The COO can’t draft a coherent recommendation until the circles are separated. The warehouse decision belongs to the business board. The distribution policy belongs to ownership. The foundation transfer belongs to the philanthropy governance stack and may require trustee action. The branch-debt problem belongs partly to the family council and partly to the family bank policy, if one exists. Treating all four as one “family alignment” question gives the founder a chance to broker peace, but it gives no body a rule it can apply again.

The family maps the issue using the three-circle model. The board receives the warehouse-automation proposal and approves a $22M reinvestment. The owners approve a one-time $8M distribution under a written liquidity policy. The foundation trustees receive a separate $7M gift proposal after counsel reviews the tax and control issues. The family council refers the branch-debt question to the Family Bank committee and directs staff to draft a policy before any cousin receives a loan.

No one gets everything. That is the point. Each question goes to the room with authority to answer it, and the minutes record which circle was active. The founder still has voice, but the founder is no longer the routing system.

A second example is smaller and more common. A 29-year-old rising-generation member asks for a paid investment role in the family office. She is highly capable, with a CFA and six years at an institutional allocator. Two cousins object that she is being favored. The CIO says she would be a strong hire, but only if she reports through the ordinary analyst ladder. The family council wants to encourage rising-generation participation and worries that a rejection will read as exclusion.

The model separates the claims. As family, she deserves a clear participation path and an answer that preserves dignity. As a prospective employee, she needs to meet the same role requirements as any analyst. As a future owner, she should understand the office’s work but should not receive employment as a proxy for ownership education. The Family Employment Policy carries the business-circle decision. The rising-generation education program carries the family-circle path. The ownership education module carries the owner voice. One request becomes three decisions, each in the right room.

Caveats and Open Questions

The model is a map, not a governance design. It will not tell you whether a council should have seven seats or nine, whether a trust protector should hold a veto, or whether G3 should receive observer rights before age 25. It tells you which system a question belongs to so the family can design the body that answers it.

The model also comes from family-business scholarship. In a family-office setting, “business” can be ambiguous. Sometimes it is the operating company. Sometimes it is the office itself. Sometimes it is a foundation or private trust company with staff and fiduciary duties. The article’s operating rule is to name the business system explicitly in each context rather than treating the third circle as always meaning an operating company.

Newer practitioner extensions add more rooms. Banyan Global’s Four-Room Model, for example, maps decisions into owner, board, management, and family rooms. Davis’s later work adds the individual as a fourth circle in some treatments. Those extensions are useful. They do not replace the original model’s simpler diagnostic: role conflict in family enterprise begins when family, ownership, and business are confused.

Consequences

The benefit of the Three-Circle Model is practical de-escalation. It gives families a way to say, “We are not having one argument; we are having three different arguments in the same room.” That sentence lowers the emotional temperature. People can stop accusing one another of bad faith and start asking which body has authority.

The model also improves document quality. A constitution drafted with the circles in mind names the authority of family members, owners, trustees, directors, and staff separately. A decision-rights charter built from the model sends decisions to the right body instead of sending every hard question to the founder. A family employment policy can distinguish kinship rights from workplace standards without pretending either side is imaginary.

The liabilities come from overuse. Families can turn the model into a consultant’s sorting exercise and avoid the harder work: choosing thresholds, writing charters, training successors, and enforcing consequences when someone ignores the routing. A diagram that never changes a meeting agenda is decoration.

The deeper liability is that the model can make role conflict visible before the family is ready to govern it. A passive owner may discover they have fewer operational rights than they assumed. A family employee may discover kinship does not entitle them to a role. A founder may discover that the family has been calling personal preference “governance” for twenty years. That discomfort is the work. The model is useful because it makes the work harder to dodge.

Sources

  • John A. Davis, Three-Circle Model of the Family Business System — Davis’s canonical practitioner statement of the model, including its 1978 Harvard origin with Renato Tagiuri, the three systems, and the seven sectors created by their overlaps.
  • Renato Tagiuri and John A. Davis, “Bivalent Attributes of the Family Firm,” Family Business Review, 1996 — the scholarly family-business lineage behind the model, developed from Davis’s doctoral work and the earlier Harvard Business School working-paper stream.
  • Kelin E. Gersick, John A. Davis, Marion McCollom Hampton, and Ivan Lansberg, Generation to Generation: Life Cycles of the Family Business, Harvard Business School Press, 1997 — the developmental family-enterprise treatment that applies family, ownership, and business system distinctions across founder, sibling-partnership, and cousin-consortium stages.
  • International Finance Corporation, IFC Family Business Governance Handbook, 4th ed., 2018 — an open-access governance handbook that translates family, owner, board, and management distinctions into family institutions and decision processes.
  • Banyan Global, The Four-Room Model for Family Businesses — a practitioner extension that maps related decisions into owner, board, management, and family rooms, useful as a governance-design companion to the three-circle diagnostic.

This entry describes a structural pattern and is not legal, tax, or investment advice. Consult qualified counsel and tax advisors licensed in your jurisdiction before adopting any structure described here.