Two Reform Efforts, One Structural Problem
Two reform conversations are moving through Washington right now. One concerns pharmacy benefit managers. The other concerns retirement plan service providers. They are usually treated as separate policy areas with separate constituencies and separate committee jurisdictions, but structurally, they are the same problem.
In both markets, the nominal purchaser is a fiduciary buying services on behalf of beneficiaries. In pharmacy benefits, a health plan or employer hires a PBM to manage prescription drug coverage. In retirement plans, a plan sponsor hires a recordkeeper and an investment advisor to manage participant accounts. Both fiduciaries have a legal duty to act in beneficiaries' interest and to pay only reasonable amounts for services received. Neither can function as a buyer, because the mechanics of being a buyer have been engineered out of both markets.
PBM compensation flows through three principal channels. Rebates from drug manufacturers are paid to the PBM in exchange for preferred formulary placement, with the largest rebates going to the highest list price drugs. Spread pricing captures the gap between what the plan pays the PBM for a prescription and what the PBM pays the pharmacy that filled it, with the PBM retaining the difference. Administrative fees and various indirect payments make up the remainder, often tied to drug pricing rather than to defined services. Antonio Ciaccia, whose data analytics work helped uncover $244 million in spread pricing in Ohio Medicaid, summarized the structural feature this way. With mystery there is margin.
The economic effects have been documented with increasing precision. Researchers at the USC Schaeffer Center, led by Neeraj Sood and Karen Van Nuys, found that on average a $1 increase in manufacturer rebates paid to PBMs is associated with a $1.17 increase in the drug's list price. Geoffrey Joyce, Director of Health Policy at the Schaeffer Center, has estimated that delinking PBM and other intermediary compensation from list prices would reduce annual net drug spending by $95.4 billion, roughly 15 percent of the national drug bill. The Federal Trade Commission's 2024 interim staff report documented that the three largest PBMs now process roughly 80 percent of all US prescription claims, up from 52 percent in 2004, and that the Big 3 retained approximately $1.6 billion in excess revenue on just two cancer drugs across less than three years. Joyce has been explicit about the implication. Reforms aimed at curbing consolidation or simply requiring more disclosure are too modest to change PBM behavior, because the industry has shifted tactics faster than disclosure regimes can keep up. Structural change to the compensation mechanism is what is required.
In retirement plans, the dominant compensation mechanisms are either asset charges or revenue sharing, depending on the arrangement. The asset charge is a stated percentage of plan assets deducted directly from participant accounts on a recurring basis, paid to advisors and recordkeepers. Revenue sharing arrangements pay recordkeepers and advisors from fund expense ratios, with payments triggered by fund selection and asset levels rather than by services performed. Some plans rely on one, some on the other, and many on both. The percentages are disclosed, but the dollar amounts those percentages produce year over year are not invoiced. Plan sponsors almost never write a check for total compensation. They do sometimes write checks for admin fees specifically, but the admin invoice is typically the smallest fee in the plan, and a sponsor who sees the admin invoice and assumes it represents what the plan pays is missing most of the cost.
Revenue sharing has declined in large plans where competitive RFP pressure has driven scrutiny, but the structural opacity has not gone away. It has migrated. Recordkeepers that once collected through revenue sharing now collect increasingly through stable value fund spread income, where the recordkeeper earns the difference between what a stable value fund generates and what it credits to participants. The spread appears on no direct fee invoice. Twenty seven lawsuits were filed in 2025 challenging stable value arrangements in large plans, compared to fewer than five in 2024, according to the Encore Fiduciary ERISA Fiduciary Litigation report. Investment committees were evaluating direct fees without seeing what their recordkeepers collected through the spread. Opacity moved, but it did not disappear.
The architecture is the same across both industries. Compensation moves automatically through product pricing or account balances, and the provider's income arrives whether participants were served or not, whether prescriptions were rational or not, whether advice was given or not. In one documented case in retirement plans, fees continued to flow to a financial advisor who had been dead since 2014. In another, a plan sponsor inherited a relationship from a departed CFO and could not identify who the advisor was. The asset charge ran for years without anyone in the company knowing the relationship existed.
In both markets, the fiduciary purchaser frequently relies on intermediaries presented as independent advisors while the underlying compensation structure rewards steering toward providers whose economics remain difficult for the purchaser to evaluate directly.
The same architecture produces the same outcomes regardless of whether the underlying product is a prescription drug or a target date fund.
Both markets have tried disclosure based reform. PBMs face transparency requirements at the state level, with at least 166 state prescription drug pricing laws enacted in recent years, and have been subject to FTC inquiry. The Department of Labor has proposed a rule that would require PBMs serving self insured employer plans to disclose detailed information about rebates, fees, and other compensation. In retirement plans, ERISA Section 408(b)(2) has required service providers to disclose compensation arrangements to plan sponsors since 2012. Both disclosure regimes have produced compliance burdens. Neither has produced market discipline.
The 2015 to 2024 retirement plan data is the empirical test. I have built and maintained a longitudinal dataset since 2009 covering 1,299 unique Chicago area retirement plans across 6,423 plan year observations, sourced entirely from sworn Form 5500 public filings. From 2015 to 2024, the median per participant fee in small and mid sized plans doubled from $681 to $1,381. The mean more than doubled from $875 to $2,163. Within that period, the mean spiked to $1,994 in 2021 when equity markets rallied and pulled back to $1,827 in 2022 when markets fell. Advisory services did not expand in 2021 and did not contract in 2022. Fees moved because assets moved. Compensation is being determined by asset balances, not by services performed.
The PBM data tells the same story. The Schaeffer Center has documented that state level transparency measures have largely failed to produce reductions in patient out of pocket costs, because the disclosures are structured around what PBMs report rather than around the transaction prices patients actually face. The FTC's 2024 interim staff report describes contracts between PBMs and pharmacies as extraordinarily opaque and notes that the disclosure regimes that exist have not constrained the steady growth of PBM market share or the parallel growth in PBM affiliated specialty pharmacy revenues.
This is not a disclosure problem in the conventional sense. The fee information exists, scattered across Form 5500 filings, 408(b)(2) disclosures, fund prospectuses, state level PBM filings, and FTC investigative materials. Some 408(b)(2) disclosures even show dollar amounts, but a disclosure showing dollars is not the same as an invoice. The dollars are projected rather than billed, paid automatically rather than approved, and presented without the recurring renewal moment that produces market discipline in every other professional service relationship. The information is not the missing piece. The invoice is.
The reform that addresses the structural problem is the same in both markets. Delink compensation from the product. Replace automatic collection with invoicing. Require the service provider to bill the fiduciary directly for total compensation from all sources, in actual dollars, on a recurring basis, tied to a defined scope. The invoice is the price. Once it exists as a separate transaction, the fiduciary can compare, negotiate, switch providers, refuse renewal, and discipline the market.
In pharmacy benefits, this reform has begun to take legislative form. The Consolidated Appropriations Act of 2026 includes provisions requiring 100 percent pass through of manufacturer rebates to group health plans, the most significant federal PBM reform since the modern industry took shape. The Federal Trade Commission's 2026 settlement with Express Scripts requires changes to formulary design, reduced reliance on high list price drugs when lower cost alternatives exist, and increased transparency around compensation and drug spending. These steps move in the right direction. The structural fix is to delink PBM compensation from list prices entirely and replace it with a stated administrative fee for services rendered.
In retirement plans, the same reform takes the form of a quarterly invoice in actual dollars covering all service provider compensation from all sources. Asset charges and revenue sharing arrangements are replaced with explicit dollar billing tied to defined scope. Embedded payments through 12b-1 fees, sub transfer agent arrangements, commission share classes, and stable value spread income are either eliminated or disclosed in dollar terms on the same recurring invoice. Compensation no longer depends on which funds participants happen to hold or what fraction of assets the advisor extracts, and the fiduciary can see what the service costs.
The legislative vehicle is a one paragraph amendment to ERISA 408(b)(2) requiring quarterly invoicing in actual dollars covering all service provider compensation from all sources. Several existing 408(b)(2) provisions become redundant once invoicing exists, because the invoice replaces the projection based disclosure those provisions were designed to require. An administrative path exists in parallel. EBSA can issue a Field Assistance Bulletin clarifying that fee reasonableness under ERISA requires evaluation of total compensation in dollar terms against defined services. The structural target is the same in both markets. Make compensation visible as a separate transaction that the fiduciary must approve.
Disclosure cannot price a service. An invoice can.