California Proposition 38 Explained: Immunology Research Bonds

Recommendation: NO — moderate confidence. Proposition 38 would authorize $8.4 billion in California general-obligation bonds for immunology and immunotherapy research. The strongest case for YES is real: public research helps build the scientific knowledge behind new treatments, and long-term, stable funding can support work that private investors may not fund early. But this particular measure puts half of its bond proceeds into one qualifying institute, restricts the other half to a council-selected group of represented institutions, and makes California taxpayers responsible for bond repayment. The promised royalties and drug discounts may help, but their size and timing are unknown.

The institute is not named in the statute. Its eligibility criteria, however, are unusually specific and appear tailored to the California Institute for Immunology and Immunotherapy (CIII), an institute affiliated with UCLA. CalMatters reported that UCLA confirmed CIII was the only UCLA-affiliated nonprofit that met the requirements. Its co-founder and board chair, Gary Michelson, is also the measure’s principal financial backer. That relationship is relevant to the measure’s design and transparency; it does not by itself establish wrongdoing or negate the institute’s scientific potential.

This choice also intersects with Proposition 37. Prop. 38’s text says that if both measures appear on the same statewide ballot and Prop. 38 receives more affirmative votes, Prop. 38 prevails over Prop. 37 in its entirety. Prop. 37 would authorize a separate revenue-bond program for homebuyer loans. Read the two measures as separate choices with a specified legal conflict rule, not as a combined pool of bond money.

This analysis is part of the Los Angeles County Voter Guide 2026 and follows the sherafy.com Civic Outcomes Standard.

What a YES or NO vote actually does

The California Secretary of State lists Proposition 38 as an initiative statute on the November 3, 2026, statewide ballot. A YES vote authorizes the state to issue up to $8.4 billion in general-obligation bonds. A NO vote rejects this bond authorization; it does not end existing state, federal, university, philanthropic, or private medical research funding.

The bond money is divided equally after authorized bond and administrative costs: 50% to a single nonprofit medical research institute selected by the California Department of Public Health (CDPH), and 50% to individualized research grants for public and nonprofit universities and research institutions represented on a new Advisory and Accountability Research Council. At least $4.2 billion in total must support research related to cancer, heart disease, and Alzheimer’s disease. Those are minimum allocations, not a promise that the research will produce treatments for those diseases.

The initiative also establishes a California Immunology and Immunotherapy Medical Research Fund, a finance committee to issue bonds, reporting and audit requirements, a 10% revenue-sharing requirement, a California-patient price rule for qualifying products, and a California-supplier purchasing goal. Its legal text, rather than campaign shorthand, controls those obligations.

A direct conflict with Proposition 37

Proposition 38’s final text contains a specific “Conflicting Ballot Measures” section. It says that if Prop. 38 and another measure authorizing revenue bonds for middle-class housing loans appear on the same statewide ballot, the other measure is in conflict; if Prop. 38 receives more affirmative votes, Prop. 38 prevails in its entirety and the other measure is void. The described housing-loan measure is Proposition 37. California Constitution article II, section 10(b) also provides a general rule for conflicting voter-approved measures: the provisions of the measure receiving the higher affirmative vote prevail to the extent of the conflict.

Prop. 37 would authorize up to $25 billion in revenue bonds for repayable homebuyer loans. It is not the same kind of debt as Prop. 38’s general-obligation bonds. The Secretary of State’s guide presents Prop. 38’s direct conflict clause; the operative effect if both pass should be understood from the full texts and constitutional rule. This article evaluates Prop. 38 on its own merits and does not assume which measure will receive more YES votes.

The baseline: research funding already exists

Medical research is funded through several systems. The Legislative Analyst’s Office (LAO) describes U.S. medical research and development as a multistage process: basic research often occurs at universities, companies conduct much of the applied and clinical work, and regulators assess whether treatments are safe and effective. Many projects do not progress through every stage. The LAO estimates total U.S. medical R&D funding at about $300 billion in 2023, with companies and government agencies among the major sources.

California already has a state bond-financed medical research institution. Voters approved Proposition 71 in 2004 and Proposition 14 in 2020, authorizing a combined $8.5 billion in general-obligation bonds for stem-cell research through the California Institute for Regenerative Medicine (CIRM). That program supports a different research field and does not automatically substitute for an immunology program. It does show that rejecting Prop. 38 would not mean California has never used bonds for medical research, nor would it end the existing CIRM program.

The realistic NO alternative is the existing mix of NIH and other federal grants, university and nonprofit research, philanthropy, private-sector R&D, and state funding through the ordinary budget or existing programs. A NO vote does not guarantee that the Legislature will replace Prop. 38 with another immunology appropriation, or that federal support will remain stable. A YES vote likewise does not guarantee a particular number of grants, clinical trials, cures, or jobs.

What the measure funds—and how it selects recipients

Half for one institute

CDPH is to select one nonprofit institute that existed by January 1, 2025, specializes in immunology and immunotherapy, and had an affiliation agreement by that date with a UC campus meeting specified research, clinical-trial, inpatient, and outpatient-volume tests. The institute must have a research site of at least 200,000 square feet, senior UC and health-system officials on its board, at least $250 million in philanthropic commitments, and at least $50 million committed to each of a rapid-vaccine program and a microbiome program. If CDPH finds no qualifying willing institute, it must modify the criteria by the minimum amount necessary to identify one.

The measure does not name CIII, and the statute leaves CDPH a formal selection role. But the criteria are not a broad statewide competition among institutes. CalMatters compared the criteria with UC campus volumes, CIII’s UCLA relationship, facility plans, board composition, and fundraising. It reported that UCLA confirmed CIII was the only UCLA-affiliated nonprofit that could satisfy the requirements. This supports describing CIII as the apparent intended beneficiary, while preserving the distinction between investigative reporting and the formal statutory selection that would occur only if voters approve the measure.

CDPH must review the institute agreement every five years. A material failure can support termination only after a finding that it substantially frustrates the law’s purpose, notice, and a reasonable opportunity to cure. This is a meaningful compliance safeguard, but it is not a recurring voter or Legislature decision about whether to continue funding the institute. The institute may also receive an initial General Fund loan of up to $100 million, repayable from bond proceeds within 12 months or as soon as practicable thereafter.

Half for grants selected by an interested council

The other half is awarded as project-specific grants. The council includes representatives from seven named UC campuses and 10 to 15 additional representatives of universities or nonprofit research institutions. It selects grant recipients only from organizations represented on the council. The institute receiving the other half must be consulted on final grant decisions, and grant recipients must offer it a substantial opportunity to participate in funded research. UCLA must be offered an opportunity to participate in any funded clinical trials.

The council must establish conflict standards based on National Academies best practices and update them at least every four years. Peer-review panelists with a collaborative or commercial relationship with an applicant are disqualified from reviewing that application. These provisions provide real process protections. At the same time, the council’s represented organizations are also the eligible applicants, and the statute gives the funded institute a consultative role over the grant half. The text’s conflict standards leave important implementation detail to the council rather than spelling out comprehensive recusal rules in the initiative itself.

The priorities emphasize projects with the greatest potential to yield immediate, accessible, and affordable therapies. This favors near-term translational work, but it may put pressure on basic or exploratory research whose value is less immediate. The statute sets no measurable cure target, distribution formula among diseases or institutions, or external evaluation comparing funded projects with alternative investments.

Fiscal effects and the public-return claims

The bond is a state general-obligation bond, not a project revenue bond. The statute pledges the state’s full faith and credit, and the official voter guide says repayments come from the General Fund. The final guide estimates $500 million to $600 million in annual General Fund debt service for about 20 years, about 0.25% of the state’s General Fund budget. It says that, after adjusting for inflation, total bond costs would be about 10% more than paying up front with money already available.

An earlier LAO initiative review and the signed argument against the measure used an estimate of approximately $500 million annually for about 25 years. The final Secretary of State voter-guide analysis gives the later $500–$600 million for about 20 years estimate. Those published figures differ; this article uses the final official voter-guide estimate for the current ballot summary and does not multiply annual payments into a total-cost figure. The bond principal authorization is $8.4 billion, while actual issuance, interest, and repayment timing determine the realized debt-service schedule.

The measure directs recipients to pay California 10% of revenues they derive from commercializing, licensing, selling, or otherwise monetizing intellectual property that arises from funded research. The payment applies even if the research was only partly supported by the initiative. Revenue first repays the state’s bond cost, including interest; any additional receipts go back to the research fund. This is a potential recovery mechanism, not evidence that the state will recover all principal and interest. The LAO says revenue could be significant but is uncertain in amount and timing and could take decades.

The 20% price rule applies to technology or drugs developed through research funded by Prop. 38 and sold or licensed for California patients, hospitals, or insurers. For the first 12 months of sales, the California price cannot exceed the price charged elsewhere in the United States. After that, the price must be at least 20% below the national average calculated without California, using a defined sales period and annual updates. It is not a 20% discount on existing medicines or a guarantee of lower drug spending overall. Whether a product is developed, licensed, sold in California, and priced under the formula depends on future scientific, commercial, and legal events.

The initiative caps state administrative costs at 2% of total bond funds. That cap does not mean that all remaining dollars are direct research spending: authorized bond costs are paid first, grant recipients may charge indirect costs up to 20% of a grant (with a matching-funds exception), and the institute has its own agreement and reporting obligations. The initiative also provides for a temporary $100 million General Fund startup loan. The final voter guide’s annual debt-service figure is the central public fiscal exposure; royalties and indirect Medi-Cal effects are uncertain offsets or costs, not booked savings.

What comparative evidence can and cannot establish

The cross-sectional study in JAMA Health Forum identified NIH-funded research associated with 354 of 356 new drugs approved by the FDA from 2010 through 2019 and estimated $187 billion in NIH research funding associated with those products and their targets. The study indicates that public investment is deeply connected to the knowledge base for approved medicines. The figure is not a causal estimate that NIH funding produced 99.4% of approvals, a return-on-investment calculation for each drug, or a forecast of Prop. 38 royalties. The article’s full methods could not be independently inspected in this research session; the study is therefore contextual rather than decisive.

CIRM offers a limited California comparison. Its audited fiscal-year 2023 statements report $198.4 million in research-grant expense and $19.7 million in state operations. Licensing and royalty revenue was $125,000 for that fiscal year, compared with $15.6 million in the prior year. Those figures come from a different stem-cell program and one year’s financial statements; they do not establish CIRM’s lifetime return, the performance of any particular award, or what immunology research would yield. They do illustrate why annual commercialization receipts can be volatile and why a promised revenue share should not be treated as a reliable debt-service forecast.

There is no direct evaluation of Proposition 38’s exact design because it has not been implemented. Nor does this record establish the incremental number of discoveries or treatments that $8.4 billion would produce compared with other public or private funding. The evidence supports investment in biomedical research as a public good more strongly than it supports this particular allocation, governance, and borrowing structure.

Applying the ten Civic Outcomes lenses

1. Human welfare

Immunology research may improve prevention, diagnosis, and treatment of serious illness. The potential benefit is substantial, but the measure’s health outcomes are not guaranteed, and the LAO notes that most research projects do not pass every development stage. No independent estimate shows how many additional patients would benefit because of this bond.

2. Distribution and inequality

The state would bear General Fund repayment costs broadly, while research funding is concentrated in one institute and eligible member organizations. Any eventual price benefit is limited to products arising from funded research and California purchasers; the measure does not require that treatments be covered by insurance or available to every patient who needs them.

3. Civil liberties and equal treatment

The initiative does not materially change individual civil rights. It does require funded researchers to provide proposed publications for review at least 30 days before release. Recipients may propose changes for patent or other specified reasons and request reasonable delays to protect patentable information. That is not a general ban on publication, but the phrase “other reasons” leaves discretion that merits careful implementation and public oversight.

4. Economic and material effects

Successful research could support jobs, private investment, and California biotechnology. Those are plausible downstream benefits, not guaranteed employment or tax-revenue forecasts. The statutory goal of more than half of grant purchases from California suppliers applies only when reasonably possible and is not a hard local-content quota.

5. Fiscal reality and opportunity cost

The $500–$600 million annual repayment estimate is a direct, long-term General Fund commitment. Royalties may lower the net burden, but timing and amount are unknown. Borrowing and earmarking reduce future budget flexibility; the proposal does not compare its expected incremental health benefit with other state health or research uses.

6. Institutional integrity and democratic accountability

Public reports, independent financial audits, Controller review, and five-year CDPH agreement reviews are meaningful safeguards. Concerns remain about the highly specific single-institute eligibility path, selection of eligible grantee organizations from council members, and the institute’s role in grant decisions and research commercialization. The relationship among the measure’s largest backer, CIII, and the likely statutory eligibility path warrants disclosure, not an assumption of misconduct.

7. Evidence of effectiveness

The broader evidence shows that public biomedical research contributes to foundational knowledge and that some discoveries become therapies. It does not demonstrate that the exact Prop. 38 program would cause a particular number of cures, jobs, savings, or royalties. The comparative studies and CIRM records cannot substitute for an evaluation of this initiative.

8. Implementation and administrative capacity

The measure specifies CDPH, bond-finance officials, a research council, peer review, audits, and grant agreements. It also includes a 90-day selection timeline and a fallback requiring CDPH to minimally modify criteria if no institute qualifies. The speed and complexity of selecting a recipient, establishing a council, setting conflict rules, and administering a new program remain untested.

9. Unintended consequences and behavioral response

The institute and council could draw researchers, facilities, and partners into California, but a narrow preferred pathway could also shift opportunities toward organizations connected to the named campus and selected institute. The collaboration requirement may create useful clinical-trial capacity while also making participation in other institutions’ funded work more dependent on the institute and UCLA.

10. Reversibility, resilience, and future lock-in

Bond issuance and debt service create obligations lasting years, and the initiative reserves substantial funding by statute. Five-year reviews can address institute noncompliance, but they do not give voters an ordinary sunset or let the Legislature freely redirect unissued funds. California could change the law later, but it would still owe validly issued bond debt.

The strongest case for YES

Public medical research has produced knowledge that private firms later develop into medicines, and foundational discoveries may not have an immediate commercial payoff. A dedicated, multiyear funding source can protect research from annual budget cycles and may be especially valuable when federal grants are unstable. Prop. 38 would also require open scientific exchange, peer review, public reporting, independent financial audits, and revenue sharing. The supporters’ case is not simply “cures are good”; it is that a long research pipeline needs stable public investment and California can capture some returns while reducing prices on qualifying products.

Those safeguards and benefits deserve weight. A state research institute might coordinate multidisciplinary teams, and the other half’s competitive grants could spread some funds among universities and nonprofits. If federal research cuts persist, a state source could protect ongoing work. A NO vote does not ensure that California will appropriate replacement funds.

The strongest case for NO

California can support biomedical research without creating this particular bond structure. The initiative commits General Fund debt service for about two decades while concentrating half the bond proceeds in one institute whose criteria appear to identify CIII, and narrowing the remaining grant pool to organizations represented on a council that consults with that institute. The statute’s audit and conflict provisions reduce some risks but do not change the concentrated allocation or make the program’s medical or fiscal returns predictable.

The measure’s public-return assurances also exceed what is established. Ten percent of commercialization revenue and a future California price rule could be valuable if qualifying products emerge; neither promises that the bond will repay itself or reduce the state’s total health spending. The voters would be locking in financing and allocation rules before independent evidence demonstrates the marginal benefit of this program over a more open, budget-based research strategy.

Campaign claims audit

Funding and interested parties

As of October 6, 2026, the Secretary of State reported $64,215,472 in contributions to support Proposition 38 and $3,782 to oppose it. The support total comprised $13 million reported by Michelson Center for Public Policy, Inc. (a 501(c)(4) committee) and $51,215,472 by the sponsored Yes on 38 committee. These are reported contributions, not necessarily the committees’ final receipts or total spending.

CalMatters reported that Gary Michelson, CIII co-founder and board chair, and the Michelson Center for Public Policy were among the measure’s principal financial backers; it also reported that co-founder Meyer Luskin contributed at least $5 million. The institutional relationship is relevant because the initiative’s single-institute criteria appear likely to direct half of the bond to CIII. It is not proof that the institute lacks merit or that any donor acted improperly. The difference in campaign resources also means the absence of a substantial opposition campaign should not be mistaken for broad consensus.

What remains unknown

• How much of the $8.4 billion would be issued, the issuance schedule, interest rates, and the eventual debt-service schedule.
• Whether CIII or another institute would formally qualify, and how CDPH would apply or modify each eligibility requirement.
• The final council membership, conflict rules, peer-review design, and practical degree of institute participation in grant decisions.
• How many additional research projects would occur because of this bond rather than existing grants, philanthropy, or private investment.
• Whether research would produce marketable treatments, what revenue they would generate, and whether 10% receipts would repay any meaningful share of debt service.
• How the 20% price formula would work for products with rebates, insurance intermediaries, limited launch data, or changing prices, and how compliance would be enforced.
• Whether the publication-review provision’s “other reasons” would be narrowly applied to protect patentable information or used more broadly.
• The exact effect on Medi-Cal and other state health spending; the LAO says it could increase or decrease.
• The current voter-guide debt schedule differs from the earlier LAO initiative analysis and argument language, which cited about 25 years rather than about 20 years.

These gaps limit the ability to estimate results or net costs. They do not erase the certain debt obligation or the statute’s allocation rules, which are enough to make a comparative judgment while keeping confidence moderate.

What would change this analysis?

The YES case would strengthen if independent evidence showed that the bond would finance materially additional research that would otherwise not occur, if the state published credible multi-scenario debt and royalty forecasts, and if CDPH disclosed a competitive selection record demonstrating that the criteria were applied openly. Clear, independently reviewed council conflict and grant procedures and a narrow, enforceable publication-review policy would also reduce governance concerns.

The NO case would strengthen if current official fiscal schedules showed materially higher debt costs, if the institute criteria were found to exclude capable alternatives without a research justification, or if implementation rules gave the selected institute excessive control over competing grant awards. Evidence of persistent federal cuts alone would make the need for state research funding more compelling, but would not by itself establish that this exact bond design is the best response.

sherafy.com recommendation: NO — moderate confidence

The strongest reason to support Proposition 38 is substantial: medical research is a public good, early-stage science can be underfunded by markets, and the state may reasonably invest in promising immunology research. NIH-linked studies and California’s existing CIRM program support that general premise. The NO recommendation does not treat research as unimportant or assume that other funders will fully replace these dollars.

The decision is whether to authorize this long-term, voter-created financing structure without a sunset. Half of the bond proceeds is reserved for a single institute under criteria that appear tailored to CIII; the other half is limited to represented organizations, with the institute consulted on final decisions and offered a role in funded projects. Audits, peer review, reporting, and periodic compliance checks matter, but they do not resolve the concentration and accountability concerns built into the allocation itself.

The state would be obligated to repay the bonds from its General Fund, while royalty recovery, price reductions, job growth, and health-care savings are uncertain. That imbalance is decisive: the public payment is a forecasted schedule grounded in bond financing, while claimed offsets depend on research progressing through a long, failure-prone development pipeline and reaching commercial use. The law’s 10% return and 20% price provisions are potential benefits, not a fiscal guarantee.

Because the choice involves a consequential bond authorization with narrow allocation rules and no demonstrated incremental outcome estimate, the burden is on the measure to show that its specific design produces enough additional public value to justify the lock-in. The reviewed evidence makes the case for public immunology research, but not strongly enough for this $8.4 billion structure. A transparent institute competition, stronger independent grant governance, and credible fiscal and outcome projections could change the recommendation. Moderate confidence reflects both the substantial public value of research and the uncertainty about what this alternative funding structure would produce.

Evidence Ledger

References and Further Reading

• California Secretary of State, Proposition 38 official analysis — current ballot summary, LAO fiscal estimate, program mechanics, background, and indirect-fiscal uncertainty.
• Full text of Proposition 38 — controlling provisions, including §§130352–130358 on allocations, institute criteria, council, grants, revenue-sharing, pricing, and §§130372–130376 on bond authority.
• California Secretary of State, official arguments and rebuttals — signed advocacy claims and opposition arguments; campaign statements are not impartial findings.
• California Secretary of State, Proposition 37 analysis — identifies the separate homebuyer revenue-bond proposal that Prop. 38’s conflict clause describes.
• California Legislative Analyst’s Office, initial analysis of the immunology initiative — December 2025 review of the proposed measure and research-funding context; its fiscal timeline differs from the final voter-guide estimate.
• CalMatters, “This CA proposition will send $4 billion to a single nonprofit. Who’s behind it?” — investigative comparison of eligibility criteria, CIII, UCLA, and backer relationships. Reporting supports an inference about likely beneficiary, not a formal selection decision or finding of misconduct.
• NIH research associated with new drug approvals, 2010–2019 — cross-sectional study abstract; useful evidence of public research’s broad contribution, not a causal estimate or forecast of this bond’s returns.
• CIRM 2023 audited financial statements — reports one year of grant expenses, operations, and licensing/royalty revenue for a different state research program; not a direct Prop. 38 comparator.

Editorial currency and research limits

Research currency: October 10, 2026. Ballot status and fiscal effects were checked against the final Secretary of State guide; campaign contributions are reported through October 6, 2026. The NIH study’s abstract and indexing information were reviewed, but its full methods were not independently inspected; it is used only as context. CIRM’s fiscal-year 2023 audit is a limited comparison, not a forecast. Before publication, refresh official ballot status, debt estimates, campaign filings, and any new legal analysis of the Prop. 37 conflict clause and the price rule.

Return to the Los Angeles County Voter Guide 2026.