Running eleven separate 501(c)(3) entities under one operational umbrella was not a plan I drew up on a whiteboard. It grew out of necessity, out of hard legal lessons and a few governance failures that cost me real time and real credibility. As Executive Director of the TheraPetic® Healthcare Provider Group, I have spent the better part of fifteen years building what I now call a federated governance model. This post is my attempt to document what that actually looks like. The structure, the failures, the fixes, and the principle I call the rule of three that governs every expansion decision I make.
If you are a nonprofit executive, a board chair or a healthcare attorney trying to understand how multiple 501(c)(3) organizations can share infrastructure without collapsing into a single legal and financial risk pool, this is written for you. I am not going to give you a textbook overview. I am going to tell you what I built, what broke and what I rebuilt.
Why I Built a Federated Model Instead of One Big Nonprofit
The question I get most often from other nonprofit executives is a simple one: why not just create one large 501(c)(3) with multiple programs? It is a fair question and for a long time it was my operating assumption too. One board, one Form 990, one audit. Clean, simple, administratively efficient.
The problem is that simplicity at the governance level creates complexity, and exposure, at the operational level. When TheraPetic® began expanding into distinct service verticals, I realized that a single entity structure creates a single point of legal vulnerability. A complaint, a regulatory action or a funding dispute in one program area can freeze assets and distract leadership across every other program simultaneously.
Federated governance solves that problem by treating each distinct mission function as a legally separate organization with its own board, its own 501(c)(3) determination letter from the IRS and its own EIN. The entities can share staff, facilities and operational infrastructure through formal intercompany agreements. But their legal identities, their fiduciary obligations and their risk profiles remain distinct.
For TheraPetic®, this meant that our service animal training operations, our clinical documentation functions, our veteran services programs and our patient advocacy work each live in their own legal containers. That separation is not cosmetic. It is structural protection for every stakeholder involved.
Separate Boards Are Not Optional. They Are the Architecture
I want to be direct about something that took me longer to fully internalize than it should have. Separate boards are not a formality you complete to satisfy the IRS. They are the load-bearing structure of the entire federated model. Without genuine board independence at each entity, what you actually have is one de facto organization wearing multiple legal costumes. That creates interlocking directorship problems, it creates the appearance of self-dealing and it gives regulators grounds to treat all your entities as a single integrated enterprise.
Each of my eleven entities has its own board of directors. Those boards have different compositions. They hold their own meetings with their own minutes. They approve their own budgets, their own executive compensation structures and their own conflict-of-interest policies. I serve in an executive capacity across entities through formal employment or contractor agreements. Not by sitting on every board simultaneously.
The IRS Form 990 asks directly about relationships between related organizations. If your answers to Schedule R look like a family tree where one person controls every branch, you are inviting scrutiny. I recommend working with a nonprofit governance attorney to map your board composition before you file, not after.
The National Council of Nonprofits publishes useful guidance on board responsibilities that I reference with new board members: councilofnonprofits.org. It is not a substitute for legal counsel but it establishes a shared vocabulary that makes early board orientation much more efficient.
Shared Operations Without Shared Liability
Here is where the federated model gets genuinely interesting from an operational standpoint. Eleven separate legal entities could mean eleven separate HR departments, eleven separate accounting systems and eleven separate vendor contracts. That is obviously not sustainable for organizations that are, in aggregate, a mid-sized healthcare nonprofit rather than eleven small ones.
The solution is the management services organization, or MSO. One of my entities functions as the operational backbone. Handling HR, payroll, technology infrastructure, facilities management and vendor relationships. And provides those services to the other ten entities through formal written management services agreements. Those agreements specify the services rendered, the pricing methodology and the terms under which either party can exit.
The pricing methodology matters enormously. The IRS requires that transactions between related nonprofit entities reflect fair market value. If the MSO undercharges affiliate entities, it may be subsidizing organizations with different missions in ways that violate the MSO's own exempt purpose. If it overcharges, it creates inurement risk. I work with our CPA firm annually to benchmark our intercompany pricing against comparable commercial rates for each service category.
This structure also means that when I hire a clinical coordinator or an IT specialist, that person is employed by the MSO entity and deployed across the group through the service agreements. It simplifies benefits administration, it centralizes HR compliance and it gives me actual visibility into total labor costs across the portfolio in one place.
Compliance Per Entity: The Part Nobody Warns You About
Running eleven 501(c)(3) organizations means eleven Form 990s, eleven state charitable registration renewals in every state where we solicit, eleven sets of state corporate annual reports and, in the case of our healthcare-adjacent entities, eleven distinct relationships with state healthcare regulatory bodies.
I will be honest: in year three of running multiple entities, I missed a state charitable registration renewal in two states because I was tracking deadlines in a shared spreadsheet that nobody owned. We received cure notices. We paid late fees. We disclosed the lapse on subsequent 990s. It was embarrassing and entirely preventable.
What fixed it was moving to a dedicated compliance calendar managed in a project management system, we use Asana, with entity-specific workspaces and automated deadline reminders assigned to a named compliance coordinator. Every filing has a due date, an owner and a review step before submission. No filing moves to "complete" without a second set of eyes.
For healthcare-specific compliance, each entity that touches patient data or clinical documentation has its own HIPAA compliance infrastructure. That includes entity-specific Business Associate Agreements with vendors, entity-specific workforce training records and entity-specific breach response procedures. HIPAA compliance cannot be federated the way HR can. The regulatory obligations attach to the covered entity, and each covered entity has to be able to demonstrate compliance independently.
The HHS Office for Civil Rights guidance on covered entity determinations is the authoritative source I use when onboarding new entities into the group: hhs.gov/hipaa.
The Rule of Three for Expansion
Every time someone proposes that TheraPetic® should spin up a new 501(c)(3) entity. Whether for a new program area, a new geographic market or a new service vertical. I apply what I call the rule of three. It is not a formal governance policy. It is a decision filter I developed after creating two entities that I later had to dissolve because they were not viable independent organisms.
The rule of three asks three questions before any new entity formation moves forward.
First: does this new entity have a mission that is genuinely distinct from every existing entity in the group? Not a subset of an existing mission. Not a rebranding of an existing program. A distinct exempt purpose that would justify its own 501(c)(3) determination letter on its own merits.
Second: does this new entity have an identified board of at least three independent directors who are committed before formation, not recruited afterward? I will not file Articles of Incorporation for an entity whose board is a placeholder. The board has to be real, it has to be independent and it has to understand what it is agreeing to govern.
Third: does this new entity have a three-year financial model that reaches operational sustainability without depending on a subsidy from another group entity? Entities that are structurally dependent on sibling-entity transfers from day one are not independent organizations. They are program departments wearing a legal costume. That distinction matters to the IRS and it matters to the fidelity of the federated model.
If any one of the three questions produces a no, the entity does not form. The program waits or gets absorbed into an existing entity that can support it. This rule has saved me from at least four formations that would have become liabilities.
Honest Failures and What Fixed Them
I mentioned the state registration lapse above. That was a compliance failure. The more consequential failures were governance failures, and they are harder to admit because they reflect judgment errors, not process gaps.
The first significant governance failure was allowing one entity's board to operate with a de facto single decision-maker for nearly eighteen months. A founding board member held deep subject-matter expertise in a specialized clinical area and the other board members consistently deferred to their judgment without genuine deliberation. On paper the votes were unanimous. In practice there was no real governance happening. When that individual resigned unexpectedly, the entity had no institutional knowledge in its remaining board, no documented rationale for its past decisions and no capacity to operate. We spent six months rebuilding that board before the entity could function independently again.
What fixed it was implementing a formal board evaluation process across all entities, conducted annually by an outside facilitator. That process surfaces governance dynamics that self-reporting never would. It is not inexpensive but it is far less expensive than a governance collapse.
The second failure was financial. I allowed an intercompany loan between two entities without documenting it as a formal loan with a written agreement, a repayment schedule and board approval at both entities. An auditor flagged it during our annual audit as an undisclosed related-party transaction. We disclosed it retroactively, documented it properly and repaid it on a formal schedule. But the audit finding was a real reputational cost with a foundation funder who had been considering a significant grant to the group.
Every intercompany transaction now moves through a formal approval workflow. Written agreement. Dual board approval. Disclosure on both entities' 990s. No exceptions. The administrative overhead is real. The alternative is worse.
What I Would Tell Myself Starting Over
If I could sit down with the version of myself who was forming the second entity in this group, I would say four things.
Hire a nonprofit governance attorney before you need one. Not a general business attorney who has handled a few nonprofit formations. An attorney whose practice is nonprofit governance, who understands exempt organization law under IRC Section 501(c)(3), who has handled related-organization structures and who will push back when your plan creates legal exposure you have not seen yet. The TheraPetic® group's relationship with specialized counsel has prevented more problems than any other single investment I have made.
Build your compliance infrastructure before you build your program. The temptation in mission-driven work is to get the services running and handle the paperwork later. Paperwork first. Always. A program that launches without its compliance architecture in place will eventually create a crisis that stops the program entirely while you fix what you should have built at the start.
Accept that the federated model creates overhead. Eleven entities cost more to govern than one entity running eleven programs. That overhead is the price of the structural protection the model provides. If you are not willing to pay it, do not use the model. A single-entity structure with strong program accounting is a legitimate alternative for many organizations. The federated model is the right choice when the risk differentiation between program areas is significant enough to justify the cost.
Document everything between related entities as if a regulator will read it tomorrow. Because sometimes they will.
The work of running a federated nonprofit portfolio is not glamorous. It is entity-by-entity compliance reviews, intercompany agreement renewals, board composition audits and a lot of conversations with attorneys and auditors. It is also the operational foundation that lets TheraPetic® and its affiliate operations including officialservicedog.com Training Plus function at a level of clinical and legal integrity that the people we serve deserve. That is the work. I would not trade it.