Executive Summary: Level Funded in the Post-Employment Economy: Structural Adaptation, Regulatory Lag, and the Question of Relevance
LFP-12.06 — The AI Disruption#
Level funded was designed for the employer with 6 to 50 full-time employees and a stable enough workforce to anchor an annual plan year. The workforce AI is creating does not match that specification. The question is direct: can level funded adapt to serve the workforce resulting from AI-driven employment restructuring, or does its addressable market contract as that restructuring accelerates?
The architecture is adaptable. ERISA covers any private-sector employee welfare benefit plan regardless of employer size, with no statutory minimum group size. Stop loss carriers can underwrite any group size if the actuarial model supports it. TPAs can administer plans at any group size. In each case, the constraint is economic rather than structural, and economic constraints can move if the underlying pooling and cost architecture changes.
Four barriers define the gap between the current architecture and the fragmented workforce. The actuarial barrier is the most fundamental: below 10 lives, specific stop loss attachment points approach individual insurance pricing, capturing progressively less of the risk-pooling efficiency that makes group insurance superior to individual coverage. The administrative cost barrier operates on fixed-cost plan administration components, compliance documentation, plan document drafting, stop loss submission, Form 5500 filings, that do not scale linearly with group size. The adverse selection barrier operates asymmetrically: a one-to-three person business seeking self-funded coverage is statistically more likely to have a known high-cost health situation than a randomly selected employer. The member sophistication barrier is specific to self-funded context, where the micro-employer owner is simultaneously plan sponsor, plan administrator, and primary plan member.
Three paths toward adaptation exist. Pooling is the most structurally important: aggregating micro-employers into a stop loss pool large enough for actuarial stability changes the per-group underwriting logic from individual group pricing to pool-level pricing. Association health plans, if the Rand Paul Association Health Plans Act introduced as part of the Senate HELP Committee’s July 2025 portable benefits package is enacted, would address the pooling barrier directly for fractional professionals and micro-employers. Simplified underwriting using AI-assisted risk scoring and group-level signals rather than individual health questionnaires would reduce the per-unit underwriting cost and adverse selection information asymmetry.
The pessimistic scenario: level funded’s addressable market contracts as average group sizes decline below the stop loss threshold faster than product innovation extends viability downward. The optimistic scenario: product innovation and regulatory change expand the addressable market because the total number of small business entities is increasing even as average group sizes decrease. The realistic scenario sits between. The TPA that waits to see how this resolves is already falling behind. The workforce AI is creating is forming now.