Policy 04: Disintermediate healthcare and employment.

Researched: 2026-07-22. Status: draft.

The plank

README.md (verbatim):

Policy 4. Disintermediate healthcare and employment.

Incentivize employers to hire Americans by removing the burden of paying for employees' healthcare.

Drift bug: app/templates/index.html titles this plank "Separate healthcare and employment." while README.md says "Disintermediate healthcare and employment." The body text matches; the titles do not. Per the working agreements, one of the two files should be corrected. ("Separate" is plainer English; "disintermediate" is more precise about the mechanism — pick one and sync both.)

Honest note on scope: the plank names a goal (employers stop being the healthcare intermediary) and an incentive rationale (hiring), but does not specify the replacement mechanism — what the ~154 million people now covered through work would have instead. That is the plank's biggest open question and is treated as such below, not papered over.

The problem

Employer-sponsored insurance (ESI) is a historical accident, not a design. During WWII, the Stabilization Act of 1942 froze wages, and the National War Labor Board ruled that employer contributions to insurance and pensions did not count as wages — so employers competed for scarce labor with health benefits instead of pay. In 1943 the IRS ruled those employer premium payments were not taxable income to the worker, and Congress codified the exclusion in the Internal Revenue Code of 1954 (now IRC §106). Group coverage grew from ~12 million people in 1940 to over 70 million by about 1950 (Institute of Medicine, Employment and Health Benefits, 1993). Eighty years later the wartime workaround is the backbone of American health coverage.

What it costs today:

• Coverage scale: ESI covers roughly 154 million people under age 65 (KFF analysis of American Community Survey data, 2023).
• Price per job: in 2025, average annual premiums were $9,325 for single coverage and $26,993 for family coverage — family premiums rose 6% in one year. Workers contributed $6,850 of the family premium on average; the employer paid roughly the remaining $20,000 (KFF, 2025 Employer Health Benefits Survey). A $27k family premium is a per-employee cost on the order of a second part-time salary, attached to the act of hiring an American.
• It comes out of wages anyway: the economics literature finds workers, not employers, ultimately bear premium costs. Baicker & Chandra (Journal of Labor Economics, 2006) estimate a 10% premium increase is offset by a 2.3% wage decrease, reduces the probability of employment by 1.6%, cuts hours ~1%, and makes low-wage workers 3.8% less likely to be offered coverage at all. So "employer-paid" healthcare is really worker-paid, invisibly, with employment losses at the margin.
• The tax subsidy is enormous and hidden: the exclusion of employer premiums from income and payroll tax is the largest single federal tax expenditure (JCT, Estimates of Federal Tax Expenditures for Fiscal Years 2025–2029, 2025). Treasury estimates it will reduce income tax revenue by $3.9 trillion over 2025–2034, and $5.9 trillion including forgone payroll taxes (Treasury tax expenditure estimates, as summarized by the Bipartisan Policy Center, 2025). CBO put the combined annual figure at $641 billion in 2032 (CBO, Reduce Tax Subsidies for Employment-Based Health Insurance, budget options, 2022).
• It locks people into jobs: see Evidence below. The canonical estimate is a 25% reduction in voluntary job turnover among workers dependent on their employer's plan (Madrian, QJE, 1994).
• It's regressive: the exclusion's value scales with your marginal tax rate — a lawyer in the 37% bracket gets a bigger subsidy on the same plan than a warehouse worker in the 12% bracket. (Arithmetic of any exclusion; see Tax Policy Center briefing book, "How does the tax exclusion for employer-sponsored health insurance work?".)

How it shows up in an ordinary life: your raise disappears into a premium increase you never see; you stay in a job you'd otherwise leave because your kid's treatment is mid-course; you don't start the business because you'd lose coverage; and the person who lost their job during a recession lost their family's health insurance in the same week.

Who profits from the status quo

• Large group insurers. UnitedHealth Group booked $400.3 billion in revenue in 2024, of which its insurance arm UnitedHealthcare accounted for $298.4 billion (UnitedHealth Group 2024 results, SEC 8-K, January 2025). Employer group business is the commercial core for UnitedHealthcare, Elevance, Cigna, CVS/Aetna, and the Blues. The tax exclusion channels ~160 million customers to them through payroll departments, pre-sold and auto-enrolled, with no need to win individual customers on price.
• The trade association. AHIP (America's Health Insurance Plans) spent $11.77 million on federal lobbying in 2024; 31 of its 51 lobbyists previously held government jobs (OpenSecrets, 2024). The broader health sector spent $743.9 million on lobbying in 2024, the most of any sector (OpenSecrets, 2024).
• Benefits brokers and consultants. Brokers are typically paid commissions of roughly 2–10% of total premium (averaging ~4–6%), paid by the carrier and baked into the employer's premium (industry compensation summaries, 2024–2025). On a mid-size firm's multimillion-dollar premium, that is a recurring six-figure toll for intermediation that exists only because employers are the buyers. The Council of Insurance Agents & Brokers maintains a standing position paper defending the ESI tax exclusion (CIAB, 2025 Key Issues) — the trade group's revenue depends on the plumbing this plank would remove.
• Large incumbent employers. Big firms use rich benefits as a retention moat: job-lock is a subsidy to incumbent employers' retention budgets, paid for by the Treasury. When the ACA's "Cadillac tax" threatened the open-ended exclusion, the Alliance to Fight the 40 — Cigna, Pfizer, Exxon Mobil, American Airlines, New York Life, the Council of Insurance Agents & Brokers, plus major unions (United Brotherhood of Carpenters, LIUNA, IAFF) — lobbied until the tax was repealed on December 20, 2019, without ever taking effect (Business Insurance, 2015; ERISA Practice Center, 2019). That coalition is the honest preview of who fights this plank.
• Unions with negotiated plans also benefit from the status quo — covered honestly under losers and opposition below, because they are workers' organizations, not villains.

The proposal, concretely

The plank's mechanism, stated concretely, is: end the tax code's preference for employer-purchased insurance so that compensation is paid as wages and individuals buy their own coverage in a functioning individual market.

Statutory targets:

1. IRC §106 (exclusion of employer-provided coverage from income tax) and the parallel payroll-tax exclusion — cap it, phase it down, or repeal it, redirecting the ~$5.9 trillion/decade subsidy (Treasury estimate, 2025–2034) to individuals rather than to employer plans.
2. A portable, person-based subsidy replaces it. The main design options, all previously proposed or in current use:
3. Tax-credit swap (McCain 2008): repeal the exclusion, give every household a refundable credit — $2,500 individual / $5,000 family in the 2008 version — to buy coverage anywhere (Tax Foundation analysis, 2008).
4. Defined-contribution default (ICHRA): since a 2019 federal rule (effective 2020), employers may fund an Individual Coverage Health Reimbursement Arrangement — tax-free dollars the employee spends on an individual-market plan of their choosing. ICHRA adoption grew 34% from 2024 to 2025 among employers with 50+ full-time employees; roughly 450,000 employees and dependents were offered ICHRA/QSEHRA for the 2025 plan year, with total enrollment estimated at 500,000–1 million (HRA Council, 2025; Healthcare Dive, 2025). This is the existing off-ramp: the plank would make the ICHRA-style defined contribution the norm rather than the exception, e.g., by equalizing tax treatment so a dollar of wages or ICHRA contribution beats a dollar of group premium.
5. Cap first, repeal later (CBO menu): CBO's standing budget options tax employer contributions above a premium percentile (e.g., the 50th percentile premium), which erodes the exclusion gradually (CBO, Options for Reducing the Deficit 2025–2034, December 2024).
6. Where it already works: the Netherlands (since 2006) and Switzerland (since 1996) achieve universal coverage with no employer-sponsored insurance in the American sense — an individual mandate, standardized basic benefits, regulated competing private insurers, community rating with risk adjustment, and income-based premium subsidies. Uninsured rates: under 1% in Switzerland, ~1.5% in the Netherlands (Commonwealth Fund, The Swiss and Dutch Health Insurance Systems, 2009). These are the proof that private insurance without the employer intermediary is a solved problem, not a leap of faith.

What this plank does not decide (flagged, not hidden): whether the replacement is a credit swap, ICHRA-by-default, a Swiss/Dutch-style regulated individual market with mandate and subsidies, or a public option. Each has different fiscal and coverage consequences. The plank as written is a direction, not yet a bill. See Open questions.

Note the platform interlock: Policy 2 (eliminate employer payroll tax) and Policy 8 (decouple HSA/FSA/401k from employment) share this plank's logic — stop taxing and routing life infrastructure through the employment relationship.

Evidence

Job-lock — strong, old, and replicated:

• Madrian, "Employment-Based Health Insurance and Job Mobility: Is There Evidence of Job-Lock?" (QJE 109(1), 1994): using the 1987 National Medical Expenditure Survey, job-lock reduced voluntary turnover of workers with employer coverage by 25% (from 16% to 12% per year).
• Gruber & Madrian, "Health Insurance, Labor Supply, and Job Mobility: A Critical Review" (NBER WP 8817, 2002): survey of the literature concludes the evidence for reduced mobility is substantial, though welfare-cost estimates vary.
• Garthwaite, Gross & Notowidigdo (QJE, 2014): when Tennessee abruptly disenrolled ~170,000 adults from TennCare in 2005, labor supply rose sharply among the disenrolled — quasi-experimental evidence that people take and keep jobs for the insurance ("employment lock").

Entrepreneurship-lock — moderate:

• Fairlie, Kapur & Gates, "Is Employer-Based Health Insurance a Barrier to Entrepreneurship?" (Journal of Health Economics 30(1), 2011): business-ownership rates jump discontinuously just past age 65 (Medicare eligibility), and workers without spousal coverage are less likely to become entrepreneurs — consistent with ESI suppressing business formation.
• ACA-era: one study finds ACA Medicaid expansion raised self-employment by 1.1–1.7 percentage points among low-income childless adults (Small Business Economics, 2021); but Leung & Mas find the expansion increased Medicaid coverage 3.0 points with no significant employment effect (Princeton working paper, 2016). Post-ACA evidence is genuinely mixed — the ACA's guaranteed issue already removed the worst of pre-existing-condition lock, which likely shrank the remaining measurable effect.

Incidence — strong: Baicker & Chandra (JoLE, 2006), figures in "The problem" above: premiums come out of wages and marginal jobs.

International — strong on feasibility, weaker on transferability: Switzerland and the Netherlands (Commonwealth Fund, 2009) show universal private coverage without employer intermediation. Caveat honestly: both are small, both impose an individual mandate and heavy insurer regulation (community rating, risk adjustment, standardized benefits), and both subsidize a large share of households. "Remove the employer" only worked there because those other institutions exist.

Overall strength of evidence: moderate-to-strong. Strong that ESI suppresses mobility and depresses wages and that non-employer systems can achieve universal coverage; weaker on exactly how large the job-lock cost is post-ACA, and no modern precedent for unwinding an ESI system at US scale.

Who wins, who loses

Transition duration: wage incidence research implies employer premium savings pass into wages over several years of raises, not overnight — a multi-year phase (e.g., cap the exclusion, then lower the cap) is the realistic path, meaning transition pain and political exposure last through at least two election cycles.

Does the compensation compensate? Only if the replacement subsidy is (a) refundable (so low-bracket workers gain rather than lose relative to an exclusion) and (b) risk-adjusted/community-rated (so sick workers can actually buy coverage). The McCain 2008 flat credit failed test (b) politically; ICHRA + post-ACA guaranteed issue passes (b) but currently fails scale. A plank that ends the exclusion without nailing both is a real tax increase on the middle class — concede this.

Fiscal impact

• The status quo subsidy: $3.9 trillion in forgone income tax over 2025–2034; $5.9 trillion including payroll tax (Treasury tax expenditure estimates, via Bipartisan Policy Center, 2025). CBO: $641 billion in 2032 alone (CBO budget options, 2022). JCT independently ranks it the largest single tax expenditure for 2025–2029 (JCT, 2025).
• CBO's December 2024 Options for Reducing the Deficit: 2025–2034 includes capping the exclusion at the 50th-percentile premium (taxing contributions above it); CBO scores these options as raising substantial revenue, partially offset by higher federal outlays as some workers shift to marketplace/Medicaid coverage. (Exact alternative-by-alternative 10-year figures not verified in this pass — pull them from CBO option 60953 before the ebook draft; the site blocked retrieval during research.)
• Net effect depends on the unspecified replacement. Three honest scenarios:
• Cap-and-phase-down with no new subsidy: large net revenue gain (hundreds of billions per decade at a modest cap; trillions at full repeal) — but functions as a broad tax increase and is politically dead on its own.
• Revenue-neutral swap (McCain-style credit or universal ICHRA credit): roughly $0 net by construction; the fiscal win is indirect — a visible, capped, budgetable subsidy replaces an invisible, open-ended one that grows with premium inflation.
• Swiss/Dutch-style subsidies: new on-budget outlays for income-based premium support, funded by exclusion repeal; net fiscal effect near zero to positive, but converts hidden tax expenditure into visible spending (expect the score to look "worse" while the true cost falls).
• Back-of-envelope (labeled as such): even freezing the exclusion's growth — taxing only future premium growth above inflation — captures a growing share of a subsidy that CBO projects reaching ~$641B/year by 2032, without touching anyone's current plan value.

The platform's core fiscal claim holds: this is the single largest pile of money in the tax code, currently spent propping up an accident of 1943.

The opposition's best case

1. "You are disrupting coverage 154 million people have and like." (AHIP; U.S. Chamber of Commerce; every incumbent politician since "if you like your plan.") Industry-commissioned polling finds ~93% of covered employees satisfied with ESI (AHIP-sponsored survey, reported 2023 — treat as advocacy polling, but independent surveys also find majorities satisfied). Response: satisfaction is real but measured against a hidden price — workers don't see the ~$20k (2025, KFF) of forgone wages or the $5.9T decade subsidy. Still, concede: any reform that involuntarily moves tens of millions off plans they chose will be punished politically; the 2008 McCain plan and the Cadillac tax both died on exactly this. The viable path is voluntary migration (ICHRA-style) plus a slowly tightening cap — not a cliff.
2. "The individual market will suffer adverse-selection collapse." (American Academy of Actuaries risk-pooling analyses, 2009 and later.) Employer groups pool risk incidentally to employment; individual markets attract the sick first, risking premium spirals. Response: this was decisively true pre-ACA. Post-2014 guaranteed issue, community rating, risk adjustment, and (in the Swiss/Dutch template) a mandate and subsidies are precisely the technology that fixed it — Switzerland and the Netherlands run individual markets with under 1% and ~1.5% uninsured respectively (Commonwealth Fund, 2009). Concede: sequencing matters absolutely; dissolving employer pools before the individual market's guardrails are funded and enforced would recreate the pre-2014 disaster.
3. "This is a middle-class tax increase." (Deployed by the 2008 Obama campaign against McCain's exclusion repeal; by unions against the Cadillac tax.) Taxing previously excluded premiums raises taxes on anyone whose plan value exceeds the replacement credit. Response: the exclusion is regressive — biggest subsidy to the highest brackets — so a refundable flat credit redistributes downward on net. Concede: for upper-middle workers with generous plans, this genuinely raises taxes; say so rather than pretending everyone wins.
4. "Unions traded wages for these benefits for decades; repeal confiscates that bargain." (IAFF, LIUNA, Carpenters — the labor wing of the Alliance to Fight the 40, which won full Cadillac-tax repeal, December 20, 2019.) Response: the wage-incidence literature implies future compensation rebalances toward cash, and union members gain bargaining mobility like everyone else. Concede: past concessions can't be retroactively repriced; grandfathering collectively bargained plans through current contract terms is a fair and probably necessary accommodation.
5. "Employers are competent purchasers; individuals aren't." Employers negotiate rates, curate plan menus, and handle enrollment friction; individuals face a confusing market and worse per-unit prices. Response: this is an argument for good defaults, standardized plan tiers, and auto-enrollment (all present in the Dutch/Swiss systems and in ICHRA platforms), not for keeping your boss in charge of your doctor. But concede: the individual-market shopping experience today is worse than a good HR department, and pretending otherwise loses the argument with anyone who has used both.

Talking points

• One line: Your boss shouldn't own your health insurance — pay people wages, let them buy coverage that follows them, and hiring an American stops costing an extra $27,000 a year. (Family premium: $26,993, KFF 2025.)
• Thirty seconds: Employer health insurance is a 1943 tax accident, not a plan. It's now the largest loophole in the tax code — about $5.9 trillion over the next decade — and it makes every hire carry a five-figure premium, which studies show comes straight out of your wages. Cut the cord: convert the subsidy into a portable credit every worker keeps whether they change jobs, start a company, or retire early. Switzerland and the Netherlands cover essentially everyone this way, privately, with no employer middleman. (Treasury 2025–2034 estimate; Baicker & Chandra 2006; Commonwealth Fund 2009.)
• Two minutes: adds the villain and the honest cost. The status quo's defenders are specific: insurers whose group divisions book hundreds of billions a year (UnitedHealthcare alone: $298B in 2024), brokers skimming 2–10% commissions off every employer premium, and the coalition — Cigna, Pfizer, Exxon Mobil, plus, honestly, several major unions — that lobbied the Cadillac tax to death in 2019 without it ever taking effect. The evidence that the system hurts workers is decades deep: job-lock cuts voluntary job-changing by about a quarter (Madrian 1994), business formation jumps the moment people hit Medicare at 65 (Fairlie–Kapur–Gates 2011), and premium increases show up as wage cuts (Baicker–Chandra 2006). The honest cost: people with gold-plated plans — including union members who traded pay for them — would pay more or get less unless we grandfather contracts and make the replacement credit refundable and risk-adjusted; and the individual market must keep its ACA guardrails or this fails the way the pre-2014 market failed. We say all of that out loud.

Open questions

1. The replacement mechanism — the plank's central unspecified choice. Credit swap (McCain 2008), ICHRA-by-default, Swiss/Dutch regulated individual market with mandate + income-based subsidies, or public option? These have wildly different fiscal, coverage, and political profiles. The platform must pick before this plank is scoreable. Recommended framing for the next pass: "make the ICHRA-style defined contribution the tax-preferred default, cap the exclusion, and revisit in five years."
2. Exact CBO scores. CBO option 60953 (December 2024) alternative-by-alternative revenue over 2025–2034 could not be retrieved this pass (site blocked); pull and insert before ebook draft. The Treasury $3.9T/$5.9T figures were sourced via a secondary summary and should be checked against the Treasury tax expenditure tables directly.
3. How big is job-lock after the ACA? The canonical 25% figure is from 1987 data, pre-HIPAA and pre-guaranteed-issue. Post-ACA evidence (Leung & Mas) is mixed. An economist reviewer should be asked first: what is the best current estimate of ESI's mobility and entrepreneurship cost, and does it still justify the plank's weight?
4. Wage pass-through speed. How many years until employer premium savings actually appear in paychecks, and does pass-through reach low-wage workers or get captured? This determines whether the transition story ("your wages go up") is honest on a political timescale.
5. Interaction with Policy 2 and Policy 8. Eliminating employer payroll tax (P2) and decoupling HSA/FSA/401k (P8) alongside this plank amounts to a full decoupling of compensation from benefits — should the three be scored and presented as one package?
6. ICHRA trajectory as natural experiment. ICHRA enrollment (est. 0.5–1M in 2025, HRA Council) is small but compounding; by 2027–28 there should be credible data on premiums, selection, and satisfaction in employer-funded individual coverage. That evidence could confirm or kill the gradualist path.

Sources

1. JCT, Estimates of Federal Tax Expenditures for Fiscal Years 2025–2029 (x-45-25), 2025. https://www.jct.gov/getattachment/8c830c45-1680-4f7e-a649-2a0106f6b6e3/x-45-25.pdf — primary.
2. Bipartisan Policy Center, "Paying the 2025 Tax Bill: Employer-Sponsored Health Insurance," 2025 (summarizing Treasury tax expenditure estimates: $3.9T income tax / $5.9T with payroll, 2025–2034). https://bipartisanpolicy.org/explainer/paying-the-2025-tax-bill-employer-sponsored-health-insurance/ — secondary (verify against Treasury tables).
3. CBO, "Reduce Tax Subsidies for Employment-Based Health Insurance," Budget Options 58627, 2022 ($641B in 2032). https://www.cbo.gov/budget-options/58627 — primary (figure via search excerpt; re-verify).
4. CBO, Options for Reducing the Deficit: 2025 to 2034, December 2024 (option 60953). https://www.cbo.gov/budget-options/60953 — primary (alternative-level figures unverified this pass).
5. KFF, 2025 Employer Health Benefits Survey (premiums $9,325 single / $26,993 family; worker contribution $6,850; 154M with ESI per ACS 2023). https://www.kff.org/health-costs/2025-employer-health-benefits-survey/ and summary PDF https://files.kff.org/attachment/Employer-Health-Benefits-Survey-2025-Annual-Survey-Summary-of-Findings.pdf — primary.
6. Institute of Medicine, "Origins and Evolution of Employment-Based Health Benefits," in Employment and Health Benefits: A Connection at Risk, 1993. https://www.ncbi.nlm.nih.gov/books/NBK235989/ — primary (history: 1942 Stabilization Act, 1943 IRS ruling, 1954 codification, 12M→70M coverage growth).
7. Madrian, B., "Employment-Based Health Insurance and Job Mobility: Is There Evidence of Job-Lock?" Quarterly Journal of Economics 109(1): 27–54, 1994. https://academic.oup.com/qje/article-abstract/109/1/27/1850075 — primary.
8. Gruber, J. & Madrian, B., "Health Insurance, Labor Supply, and Job Mobility: A Critical Review of the Literature," NBER WP 8817, 2002. https://www.nber.org/papers/w8817 — primary (link pattern standard; number verify on second pass).
9. Garthwaite, C., Gross, T. & Notowidigdo, M., "Public Health Insurance, Labor Supply, and Employment Lock," QJE, 2014 (TennCare 2005 disenrollment). https://sites.bu.edu/talgross/files/2017/08/garthwaite-gross-notowidigdo.pdf — primary.
10. Fairlie, R., Kapur, K. & Gates, S., "Is Employer-Based Health Insurance a Barrier to Entrepreneurship?" Journal of Health Economics 30(1): 146–162, 2011. https://rfairlie.sites.ucsc.edu/files/2022/08/jhe-2011-entrepreneurship-lock.pdf — primary.
11. Baicker, K. & Chandra, A., "The Labor Market Effects of Rising Health Insurance Premiums," Journal of Labor Economics 24(3), 2006 (NBER WP 11160). https://www.nber.org/papers/w11160 — primary.
12. Leung, P. & Mas, A., "Employment Effects of the ACA Medicaid Expansions," working paper, 2016. https://www.princeton.edu/~amas/papers/aca_060116.pdf — primary.
13. Small Business Economics, "State Medicaid Expansion and the Self-Employed," 2021 (self-employment +1.1–1.7pp). https://link.springer.com/article/10.1007/s11187-021-00559-5 — primary.
14. Commonwealth Fund, The Swiss and Dutch Health Insurance Systems: Universal Coverage and Regulated Competitive Insurance Markets, 2009. https://www.commonwealthfund.org/publications/fund-reports/2009/jan/swiss-and-dutch-health-insurance-systems-universal-coverage-and — secondary (dated; uninsured figures are 2009-era).
15. UnitedHealth Group, 2024 full-year results (revenue $400.3B; UnitedHealthcare $298.4B), SEC 8-K exhibit, January 2025. https://www.sec.gov/Archives/edgar/data/731766/000073176625000022/a2024q4exhibit991.htm — primary.
16. OpenSecrets, AHIP lobbying profile, 2024 ($11.77M; 31/51 revolving-door lobbyists) and Health sector lobbying totals, 2024 ($743.9M). https://www.opensecrets.org/orgs/america-s-health-insurance-plans/lobbying?id=D000021819 and https://www.opensecrets.org/industries/lobbying?cycle=2024&ind=H — secondary (compiled from primary disclosures).
17. Tax Foundation, "McCain's Health Credit: The Intersection of Health Policy and Tax Policy," 2008. https://taxfoundation.org/research/all/federal/mccains-health-credit-intersection-health-policy-and-tax-policy/ — secondary.
18. ERISA Practice Center (Proskauer), "'Cadillac Tax' on High-Cost Group Health Plans Repealed," December 2019. https://www.erisapracticecenter.com/2019/12/cadillac-tax-on-high-cost-group-health-plans-repealed/ — secondary.
19. Business Insurance, "Alliance takes on 'Cadillac' tax," 2015 (coalition membership: Cigna, Pfizer, Exxon Mobil, American Airlines, New York Life, Carpenters). https://www.businessinsurance.com/article/20150802/NEWS03/150739947/alliance-takes-on-cadillac-8217-tax — secondary.
20. Healthcare Dive, "ICHRA adoption grows as Congress mulls codifying the coverage into law," 2025 (HRA Council: +34% 2024→2025; ~450k offered; 0.5–1M enrolled). https://www.healthcaredive.com/news/ichra-adoption-growing-employers-congress-aca/750988/ — secondary.
21. Peterson-KFF Health System Tracker, "Explaining Individual Coverage Health Reimbursement Arrangements (ICHRAs)." https://www.healthsystemtracker.org/brief/explaining-individual-coverage-health-reimbursement-arrangements-ichras/ — secondary.
22. American Academy of Actuaries, "Critical Issues in Health Reform: Risk Pooling," 2009. http://actuary.org/sites/default/files/pdf/health/pool_july09.pdf — primary (actuarial analysis).
23. AHIP-sponsored polling on ESI satisfaction (93% satisfied), reported by TechTarget, 2023. https://www.techtarget.com/healthcarepayers/news/366603783/93-of-Employees-Satisfied-with-Employer-Sponsored-Health-Insurance — secondary/advocacy (industry-commissioned; use with that label).
24. Council of Insurance Agents & Brokers, "Employer-Sponsored Healthcare Tax Exclusion," position paper, 2025. https://www.ciab.com/government-regulations/2025-key-issues/employer-sponsored-healthcare-tax-exclusion-position-paper/ — primary (as evidence of broker-lobby position).
25. Tax Policy Center Briefing Book, "How does the tax exclusion for employer-sponsored health insurance work?" https://taxpolicycenter.org/briefing-book/how-does-tax-exclusion-employer-sponsored-health-insurance-work — secondary.
26. Broker commission ranges (2–10% of premium, avg 4–6%), industry compensation summaries: Nava Benefits, "Employee benefits broker commissions" and Mira Health, 2024–2025. https://www.navabenefits.com/resources/employee-benefits-broker-commissions — secondary/unverified (industry self-description; find a regulatory filing basis before ebook).