Module 16: The Engine That Looks Like Seven Engines
How One Management Firm Runs $1.5 Billion A Year Through Seven Nominally Independent Nonprofits — And Why The Architecture Is Not A Loophole
The Engine That Looks Like Seven Engines
Shadow Patriots · Module 16 · Project Milk Carton · 501(c)(3) · EIN 33-1323547
A note before we begin.
Module 16 opens Part Four — The Big Players. The previous fifteen modules walked the moves. Modules Three through Ten walked the financial moves: donor-advised funds, entity laddering, fiscal sponsorship, foreign pass-through, Delaware opacity, educational classification, personnel overlap, the Schedule B vault. Modules Eleven through Fifteen walked the information moves. Now the series turns to the actual machines — the institutional megafunders that make every prior move scale.
Module 16 is the blue half of a paired drop. Module 17 — the red half — opens with the same analytical lens applied to a structurally identical machine on the opposite political coat. The two modules are drafted side-by-side and ship sequentially. The pairing is the bipartisan firewall doctrine of this series: the only way to prove the analysis is the analysis and not the politics is to land it on both sides with the same standard, the same depth, and the same closing register.
A reader who comes into Module 16 expecting a partisan exposé will not find one. The product of this article is not a verdict on a network. The product is a citizen who can walk into a Schedule I, identify a fiscal sponsor’s footprint, and describe the architecture in their own words. The same skill works on a nonprofit in your county. That portability is the point.
A reader sees seven names on seven 990s.
New Venture Fund. A 501(c)(3) public charity — the largest of the group.
Sixteen Thirty Fund. A 501(c)(4) — the political-muscle entity.
Hopewell Fund. A 501(c)(3) — an “incubator” for new projects.
Windward Fund. A 501(c)(3) — another incubator.
North Fund. A 501(c)(4) — issue advocacy.
Telescope Fund. A 501(c)(3) — the most recent addition.
Impetus Fund. A 501(c)(3) — the smallest of the seven.
Seven different names. Seven different missions on paper. Seven different IRS filings. The reader was meant to see seven engines.
There are not seven engines. There is one engine that looks like seven engines.
The engine has a name. Until November 2025 the name was Arabella Advisors. As of November 2025 the name is Sunflower Services. The rebrand changed nothing structural. The engine is the same.
I. What Arabella Built
Arabella Advisors was founded in 2005 by Eric Kessler, a former Clinton administration official. The founding business idea was legal, sensible, and unremarkable: be the back-office for nonprofits. Run the finance, the human resources, the legal compliance, the technology, the communications. Free the nonprofits to focus on their programs. Charge a fee for the service. The model has dozens of legitimate competitors. The model itself is not the move.
The move is what gets stacked on top.
The move is this: Arabella Advisors did not stop at being a service vendor for nonprofits run by other people. Arabella also operates seven nonprofits whose senior staff, board overlap, and back-office functions all flow through Arabella itself. The seven nonprofits pay Arabella consulting fees. The fees, accumulated across roughly seventeen years, total approximately two hundred and thirty million dollars. The fees are paid by 501(c)(3) entities to a private for-profit consulting firm. The 501(c)(3)s and the consulting firm share office space. The 501(c)(3)s and the consulting firm share senior personnel. The 501(c)(3)s and the consulting firm operate, in plain English, as a single coordinated enterprise.
Each piece of that arrangement, taken alone, is legal. The fee structure clears the IRS test for arm’s-length transactions on every individual filing. The personnel overlap clears the disclosure threshold. The office sharing clears the rules. The 501(c)(3) status of the seven entities clears the rules. The 501(c)(4) status of the political-muscle entities clears the rules. The DAFs that fund the seven entities clear the rules.
The DC Attorney General opened an investigation in September 2023 into whether the totality of the arrangement clears the rules — specifically, whether the consulting fees constitute self-dealing under DC nonprofit law. That investigation is ongoing.
The reader will notice something the rest of this article will return to: every individual move is legal, the totality is under legal investigation, and the totality has been legal long enough to grow to nine point two billion dollars in combined revenue.
II. A Tour Of The Seven Rooms
The reader who has read Modules Three through Ten of this series will recognize every move below. Module 16 is not introducing new moves. Module 16 is showing the reader what happens when a single management firm runs all of the moves at once at billion-dollar scale.
New Venture Fund is the largest of the seven. A 501(c)(3) public charity. Annual revenue typically in the high nine figures. The defining feature of New Venture Fund is fiscal sponsorship — the move Module 5 walked. New Venture Fund hosts dozens of “projects” inside its single 990. The projects do not file their own 990s. The projects do not appear in IRS public records as independent entities. A donor who funds a New Venture Fund project funds, in the public-record sense, New Venture Fund itself. The project’s name, mission, and staff are administrative details inside one larger nonprofit. Module 5’s title for this move was the backpack trick. New Venture Fund is the largest single fiscal sponsor in the United States.
Sixteen Thirty Fund is the 501(c)(4) — the political-muscle entity. A (c)(4) is permitted unlimited political activity provided the activity is not its primary purpose. The (c)(4) form was the IRS category Lyndon Johnson’s 1954 amendment carved around — Module 1 walked that history. Sixteen Thirty Fund spent approximately three hundred and eleven million dollars in 2024, twice the prior year’s spend. Among its 2024 disbursements: substantial funding for the RFK Jr. presidential campaign earlier in the year. Among its later 2024 disbursements: substantial funding for the campaign opposing his confirmation as Secretary of Health and Human Services. Both disbursements were legal. Both used the same (c)(4) structure. The donor who funded both disbursements through a single Sixteen Thirty Fund contribution did not, in the public-record sense, take a position on either candidate or campaign — the donor funded Sixteen Thirty Fund. Sixteen Thirty Fund took the positions.
Hopewell Fund and Windward Fund are 501(c)(3) “incubators.” The named function is to host new projects in their early stages — the same fiscal-sponsorship mechanism New Venture Fund operates at scale. The incubator model means a new project can begin operating, raising money, and conducting its program without ever filing its own 990. When the project matures, it can spin out into an independent entity, remain inside Hopewell or Windward indefinitely, or migrate to New Venture Fund for permanent sponsorship. In November 2025, Hopewell and Windward jointly financed the acquisition of Arabella Advisors itself by Sunflower Services. Two 501(c)(3) entities financed the acquisition of the consulting firm to which they pay management fees. The transaction was disclosed and legal. It is the kind of transaction that, in a publicly traded company, would require shareholder approval. (c)(3) entities have donors, not shareholders. The donors were not asked.
North Fund is the smaller (c)(4) — a sister entity to Sixteen Thirty Fund operating at lower scale. North Fund’s defining recent moment was a Nebraska state lawsuit filed in November 2025 alleging approximately ten million dollars in foreign-source contributions to North Fund’s Nebraska electoral activity. The contributions, if confirmed, would not constitute foreign-influence violations under federal law because they were made to a (c)(4) and not directly to a campaign. They would, if confirmed, raise structural questions about whether the (c)(4) form has become a vector through which foreign money reaches American electoral activity by laundering through one layer of domestic nonprofit. The Nebraska litigation is ongoing.
Telescope Fund and Impetus Fund are the lesser-known two. Telescope Fund is the most recently created of the seven, structured as a (c)(3) for project incubation similar to Hopewell and Windward. Impetus Fund is the smallest, with revenue typically in the low eight figures. Each plays a role in the system — but the reader who has tracked the prior five entities now understands the pattern. The seven are not seven engines. The seven are one engine subdivided seven ways for legal-vehicle optimization.
The optimization is the architecture. It is also the reason the architecture exists.
III. The Math Underneath
The numbers below are drawn from publicly filed 990s and from established reporting in The New York Times, ProPublica, The Washington Post, Politico, and The Atlantic. The reader can verify each figure on ProPublica Nonprofit Explorer or in the cited reporting.
Combined revenue across all seven entities, 2006–2023: approximately nine point two billion dollars. Combined expenditures across the same period: approximately seven point eight billion dollars. Estimated annual revenue, current year: one point five billion dollars or above. Total consulting fees paid by the seven nonprofits to Arabella Advisors over the ~seventeen-year operating period: approximately two hundred and thirty million dollars.
A single donor — Swiss billionaire Hansjörg Wyss — has donated approximately two hundred and forty-five million dollars to Sixteen Thirty Fund alone. The donations were legal. Wyss is a permanent U.S. resident. The disclosures were filed. The (c)(4) form does not require donors to be U.S. citizens — there is no statutory provision excluding lawful permanent residents from political contributions to (c)(4) entities. A reader who finds the size of the figure notable, or who finds the donor’s national origin notable, is observing a fact the law has determined is permitted. The reader’s observation and the law’s determination are separate questions. The article holds them separate.
In August 2025, the Gates Foundation — one of the largest private philanthropic funders in the world — withdrew approximately four hundred and fifty million dollars in funding commitments from the Arabella network. The Gates Foundation is a (c)(3) and is bound by IRS rules that require its grants to advance charitable purposes. A donor of Gates’s scale withdrawing a half-billion dollars is a disclosure of evaluation. The evaluation is the donor’s. The reader can draw their own conclusions about why a major philanthropic funder, after years of grant relationships, voted with its feet on a four-hundred-and-fifty-million-dollar timescale.
In March 2025, the Capital Research Center — a research organization whose primary function is to study the nonprofit sector — briefed the White House on the structure and operations of the Arabella network. The briefing is a matter of public record. Capital Research Center is itself a (c)(3); the reader who applies the audit skill this article exists to teach can pull Capital Research Center’s own 990 and trace its funding architecture, which is also public.
The reader who has now seen the consulting fees, the Wyss donations, the Gates pullback, the DC AG investigation, the Nebraska lawsuit, and the Capital Research Center briefing has the structural picture. None of these facts is criminal. All of them are allowed. Several of them are under formal legal review for whether the totality is allowed. The investigation may resolve in either direction. The architecture predates the investigation by twenty years.
IV. The Mirror
Module 17 — The One Point Six Billion Dollar Donation You Never Voted On — applies the same analytical lens to a structurally identical machine on the opposite political coat. The reader will see, in M17, a single donation of approximately one point six billion dollars routed into a 501(c)(4) trust. The reader will see a cascade through a Schwab Charitable Fund DAF passthrough, into an entity called the 85 Fund, into the Concord Fund (formerly Judicial Crisis Network), into DonorsTrust, and outward into hundreds of grantee organizations. The reader will see a trustee paid three hundred and fifty thousand dollars per year for twenty-five hours a week of work. The reader will see no DC AG investigation parallel to Arabella’s, partly because (c)(4) trusts disclose less by design and partly because the trust is not headquartered in DC.
The reader will also see, when the two modules are placed side-by-side, the structure underneath both: a wealthy donor or donors, a passthrough vehicle, a stack of (c)(3) and (c)(4) entities, a management entity earning fees on the float, and an output that looks like civic activity and functions as political infrastructure. Strip the political coats off and the boxes-and-arrows look the same.
This is the part of the article that requires the reader’s attention. The mirror is not a coincidence the Shadow Patriots discovered. The mirror is structurally inevitable under the rules in force. The next section of this article explains why.
V. Why The Pattern Is Inevitable
The reader who has watched the architecture mirror across the political spectrum has the right next question: if both sides do this, is one side simply copying the other?
The answer is almost certainly no, and the reason matters.
The tax code is shape-agnostic. 501(c)(3), 501(c)(4), 527, donor-advised funds — these categories define what an entity may do. They do not care about the entity’s politics. If a wealthy donor wants to move a billion dollars and produce political-adjacent output at scale, the IRS code forces the donor into a specific stack of entities. The (c)(3) entity is required for the educational program. The (c)(4) entity is required for the issue-advocacy program. The DAF is required for donor anonymity. The fiscal sponsor is required to host the small project that does not yet have its own 990. There is no other legal architecture available. Both political coats arrive at the same architecture because the architecture is the only one the law permits.
The FEC rules are shape-agnostic. Hard-money limits, soft-money rules, super-PAC structure, coordination thresholds — same mold for everybody. A donor on either coat who wants to be effective at scale faces the same constraints, the same vehicles, and the same optimization problem.
The DAF and Schedule B layer is shape-agnostic. A donor who values anonymity has, under current law, exactly one route to that anonymity: the donor-advised fund. Both political coats use DAFs at scale because DAFs are the only legal anonymity vehicle in the system. The (c)(4) Schedule B redaction applies symmetrically. Both political coats benefit from the redaction. Both political coats would lose the redaction simultaneously if Schedule B disclosure became mandatory.
The “back office” innovation propagates. Once one operator discovers that managing seven nonprofits as a portfolio is more efficient than running seven independent nonprofits, every other operator working at the same scale either copies the model or independently arrives at it. This is convergence under optimization. It is what happens when competitors face identical legal constraints and want to maximize output per dollar.
There are honest nuances to the symmetry, and the article will not pretend otherwise. Different activities push different vehicles — judicial-confirmation campaigns naturally cluster in (c)(4) trusts, get-out-the-vote ground games naturally cluster in super-PACs. Different donor topologies — one mega-donor versus many high-net-worth donors aggregated — produce slightly different cascade shapes. Different jurisdictions produce different legal-pressure profiles. Strip the labels off, however, and the macro pattern — donor → passthrough vehicle → fund stack → mission output → management fees on top — is structurally inevitable for any large-scale political-adjacent operation playing by the rules.
This is the keystone observation of Module 16, and of Module 17 when it lands. The shape is the rules. The rules are the shape. If you played this game correctly under current law, at scale, on either side, you would arrive at exactly this architecture. The architecture is not a partisan invention. The architecture is the legal envelope optimization within current statute generates.
That observation has consequences the rest of the series will keep returning to. The most important one comes next.
VI. The Tool
Every name in this article — New Venture Fund, Sixteen Thirty Fund, Hopewell Fund, Windward Fund, North Fund, Telescope Fund, Impetus Fund, Arabella Advisors, Sunflower Services — is a legal entity operating inside a legal architecture under publicly filed disclosures. Every move in this article — fiscal sponsorship, the (c)(4) form, the DAF passthrough, the consulting-fee structure, the rebrand — is a tool the United States tax code and FEC framework explicitly permit. The system is a toolset.
Tools do not pick sides. The side that shows up owns the tool. If you do not show up, someone else will — and they may not have your town in mind. Will they have the safety of your children in mind?
The reader will hear that line again in Module 17. The reader will hear it in every Big Players module that follows. The line is the closing register of this section of the series, and it is not decorative. It is the empowerment thesis underneath the architectural diagnosis. Diagnosing the architecture without offering the citizen a way back into it produces despair. Diagnosing the architecture and showing the citizen the architecture is participation-shaped produces literacy.
The Arabella network was built by people who showed up. The judicial-confirmation infrastructure of Module 17 was built by people who showed up. Every (c)(3) in the reader’s county was built by people who showed up. The citizen who learns to read a Schedule I has joined the audience these architectures were built to operate around. The citizen who learns to read a Schedule R has joined the audience these architectures were built to operate around. The citizen who learns to do both for a nonprofit in their own town — a hospital foundation, a youth-services organization, a civic-development corporation, an education advocacy group — has joined the audience these architectures were built to operate around in the place those architectures touch the citizen most directly.
The skill is portable. The skill is the product. Module 16’s citizen-action card is structured for the skill, not for the network. The card teaches the reader to read a fiscal sponsor’s 990 in ten minutes and then to apply the same reading to one nonprofit in the reader’s own zip code. The reading takes ten minutes once. The skill becomes a reflex within five repetitions. The reflex becomes a literacy that does not depend on this article, this series, or this organization to be useful.
That literacy is the only thing the architecture cannot route around.
VII. What Is Happening Now
A reader who has reached this point of the article and asks what should I expect to happen with the Arabella network specifically deserves an honest answer.
The likely range of outcomes from the DC AG investigation is narrow. The investigation may produce a settlement, a structural-remediation order, or a finding of no actionable violation. None of those outcomes will dismantle the seven-entity architecture. The architecture is legal under current rules. Any remediation will adjust the consulting-fee disclosures, the personnel-overlap disclosures, or the fiscal-sponsorship transparency rules, and the network will continue.
The likely outcome of the Nebraska lawsuit is a state-court ruling on whether North Fund violated Nebraska state campaign-finance law. That ruling, in either direction, will not affect the federal-level (c)(4) status of North Fund or any sister entity.
The Sunflower Services rebrand has already happened. The Hopewell-and-Windward-financed acquisition of Arabella by Sunflower has already closed. The branding will continue to evolve. The architecture will not.
The Gates Foundation pullback represents a half-billion-dollar evaluation by a sophisticated philanthropic funder. The pullback does not legally affect the Arabella network. It does affect the network’s revenue runway by approximately the size of the pullback. The network will continue at lower scale or will replace the lost revenue from another source. (c)(3) revenue substitution at the half-billion-dollar level is non-trivial; the network has eighteen months to two years of operational runway to address it.
The most consequential ongoing factor for the network is none of the above. The most consequential factor is whether American citizens become literate enough in the architecture to evaluate, on their own evidence, whether the rules that produced the architecture are the rules they want in force. That question is not an outcome a court can produce. That question is the question Modules Twenty-One through Twenty-Five — the Citizen’s Playbook section of this series — exist to give the reader the tools to ask, answer, and act on. The Big Players section, of which Module 16 is the opening, exists to give the reader the architectural literacy that makes the Citizen’s Playbook usable.
That sequence — diagnose the architecture, then equip the citizen — is the spine of the entire series. Module 16 is doing diagnosis. Module 21 forward will do equipping. The two halves operate together. Either alone is incomplete.
Module 16: CITIZEN ACTION CARD, Read A Fiscal Sponsor’s 990 In Ten Minutes
VIII. The Next Move
Module 17 — The One Point Six Billion Dollar Donation You Never Voted On — opens next. The same analytical lens. The opposite political coat. The same closing register. The same Citizen Action Card skill — applied to a different stack with different cascade dynamics — landing in the reader’s hand at the end. By the time Module 17 closes, the reader will have walked both the blue-coded and the red-coded versions of the same architecture, with the same depth, the same standard, and the same call to participation.
The bipartisan firewall is not a stylistic preference. It is the structural requirement of an analysis that cannot be reduced to its politics without ceasing to be the analysis. The Shadow Patriots ship M16 and M17 paired because the alternative — shipping one without the other — would convert this series into the thing this series is mapping. We are not the cluster. We are the citizens reading the cluster’s filings.
For Module 16: the engine that looks like seven engines is, by every available structural metric, one engine. The engine is legal. The engine is under formal legal review for whether the totality of its operations is legal. The engine is one of two architecturally identical engines — the other engine is Module 17. The shape is the rules.
The reader, having walked the seven rooms, the math, the mirror, the inevitability, the tool, and the news-of-the-network, now has the structural picture. The reader’s next move is the citizen-action card paired with this module — a ten-minute reading skill the reader can apply to any 990 in any zip code in the United States. The reading produces the citizen the architecture cannot route around.
The shape is the rules. The rules are the shape. The side that shows up owns the tool.
Module 17 walks the same architecture in a different coat.
Sources And Citations
Arabella Advisors / Sunflower Services 990 filings — publicly filed and available on ProPublica Nonprofit Explorer, the IRS Tax Exempt Organization Search, and Capital Research Center’s InfluenceWatch profile. Filings cited cover fiscal years 2006 through 2023; the 2024 filings are accessible as they are released.
The seven Arabella-managed entity 990 filings — New Venture Fund, Sixteen Thirty Fund, Hopewell Fund, Windward Fund, North Fund, Telescope Fund, Impetus Fund. All are filed on Form 990 (or 990-PF) and are accessible through ProPublica Nonprofit Explorer.
DC Attorney General investigation announcement (September 2023) — the press release and the subsequent reporting on the investigation’s scope are available through the DC AG public records portal and contemporaneous reporting in The Washington Post and Politico.
Nebraska Attorney General lawsuit (November 2025) — the complaint is filed in Nebraska state court and is accessible through the Nebraska AG’s office and through reporting in The Lincoln Journal Star and Politico.
Gates Foundation grant withdrawal (August 2025) — disclosed in the Gates Foundation’s own grant-database transparency reporting and in subsequent coverage by The Chronicle of Philanthropy and The New York Times.
Capital Research Center White House briefing (March 2025) — documented through Capital Research Center’s own published materials and through contemporaneous press accounts.
Hansjörg Wyss donation history — disclosed in Sixteen Thirty Fund’s 990 filings (the (c)(4) form requires Schedule B donor disclosure to the IRS but permits public-record redaction; the figures are disclosed in Wyss-affiliated foundation filings and in established reporting). Cited reporting includes The New York Times, The Washington Post, Politico, and The Atlantic.
Sunflower Services acquisition of Arabella Advisors (November 2025) — announced in the firms’ joint public statement and subsequently reported across the philanthropic-sector trade press, The Chronicle of Philanthropy, and major outlets.
InfluenceWatch.org Arabella Advisors profile — Capital Research Center’s running profile, useful as a synthesis of the publicly available 990 data, public reporting, and lawsuit filings.
ProPublica Nonprofit Explorer — the primary research tool the citizen-action card paired with this module teaches the reader to use. Free, public, and operates against IRS-filed 990 data.
Module 1 — The Tax Code That Became A Battlefield — for the 1954 Lyndon Johnson amendment and the historical origin of the (c)(3) and (c)(4) categorizations that produced today’s architecture.
Module 5 — The Backpack Trick — for the fiscal-sponsorship mechanism New Venture Fund operates at the largest scale in the United States.
Module 9 — Three Hats, One Head — for the personnel-overlap analysis that applies inside the Arabella architecture across the seven entities and Arabella Advisors itself.
Module 10 — The Schedule B Black Hole — for the (c)(4) donor-anonymity architecture that protects the Wyss donations and equivalent-scale donations across the political spectrum.
Module 17 — The One Point Six Billion Dollar Donation You Never Voted On — the paired red-coat module landing next. The reader who has walked Module 16 will recognize every structural element in Module 17, in different proportions.
Shadow Patriots · Module 16 · Project Milk Carton · 501(c)(3) · EIN 33-1323547
This article is the first in Part Four — The Big Players — of the Shadow Patriots civic-investigations series. It is the blue-coat half of a paired drop with Module 17. Both modules apply the same analytical lens with the same depth and the same closing register. The pairing is the bipartisan firewall doctrine of this series.
Editorial discipline (Two-Tier Naming Doctrine, locked 2026-04-24): Module 16 operates under Tier 1 financial doctrine. Named legal entities are permitted because the symmetry between the M16 and M17 entities is the teaching. Where individuals are named — Eric Kessler, Hansjörg Wyss — they are named in the structural roles they hold under public filings, not as targets of evaluation.
Evidence standard: every factual claim in this article is verifiable through publicly filed 990s, court records, government press releases, and contemporaneous reporting in named outlets of record. PMC does not allege misconduct by any individual or organization. PMC describes the architecture in which 2026 citizens live and equips citizens with the literacy to identify it themselves.
Subliminal through-line (Series Bible §0): every Citizen Action Card in this series teaches a transferable skill the reader can apply to a nonprofit in the reader’s own town. The skill is the product. The architecture is the diagnosis. The citizen is the destination.
