How Nonprofit Insurers Compare in Claim Denials

You might think not-for-profit health insurers have a lower claim denial rate, but 2024 research shows that’s not true. I pulled the latest CMS Transparency in Coverage data available for the 2024 plan year, published in the PY2026 public-use file. CMS reports roughly 451 million in-network Marketplace claims in 2024; about 85 million were ultimately denied, for a 19% denial rate. Across individual reporting insurers, denial rates ranged from 3% to 36%. (KFF)

Large insurers: nonprofit/mutual vs. for-profit

Below are the 19 parent organizations processing more than 5 million HealthCare.gov claims in the KFF/CMS analysis. I’ve classified organizations by parent ownership structure; “nonprofit/mutual” includes policyholder-owned mutual organizations rather than publicly traded companies.

The claim counts come from the CMS-derived KFF parent-company dataset; KFF reports the same range of 8% for Elevance to 25% for Oscar among the large parent organizations. (Becker’s Payer Issues | Payer News)

Ownership classifications have a few nuances. For example, GuideWell explicitly describes itself as a not-for-profit, policyholder-owned organization, while HCSC is a mutual legal reserve company rather than a publicly traded corporation. BlueCross BlueShield of Tennessee explicitly identifies itself as not-for-profit. (Florida Blue)

Here’s the surprising result

Using those claim counts rather than simply averaging insurer percentages, I calculate:

So among these 19 very large Marketplace parent organizations, nonprofit/mutual insurers had a slightly higher weighted claim-denial rate—approximately 19.4% versus 18.5% for for-profits.

That’s only about 0.9 percentage point, so I would not interpret it as evidence that nonprofits are systematically worse. But it certainly undermines an assumption that nonprofit status by itself leads to fewer claim denials.

Differences among individual companies are far greater than the nonprofit/for-profit difference. Elevance was at roughly 8%, while nonprofit GuideWell/Florida Blue was around 22% and nonprofit BCBS Tennessee around 21%. Oscar, a for-profit, was highest among the large organizations at 25%. (Becker’s Payer Issues | Payer News)

What about medical-necessity denials?

This is where the data become more interesting—and more problematic.

Across HealthCare.gov plans in 2024, CMS received approximately 79 million reported denial reasons. Of those reasons:

• 5% involved medical necessity.
• 9% involved failure to obtain prior authorization or a referral.
• 13% involved an excluded service.
• 25% were administrative.
• 36% were simply classified as “other.” (KFF)

The percentages are shares of reported denial reasons, not percentages of every claim received. A claim can also have more than one denial reason, so you can’t simply add them to get a patient-level denial rate. (Insurance News | InsuranceNewsNet)

And individual insurers vary enormously. One of the most striking findings is Molina Healthcare of Mississippi, a for-profit insurer. About 38% of its reported denial reasons were medical-necessity denials, versus roughly 5% nationally. (KFF)

That means Molina Mississippi was roughly 7–8 times the Marketplace average in the proportion of reported denial reasons attributed to medical necessity.

But CMS does not give us enough information to say that “Molina rejected 38% of treatments as medically unnecessary.” That would be inaccurate. The denominator is reported denial reasons, not treatment requests.

Prior authorization produces an even more extraordinary outlier

The standout is Blue Cross Blue Shield of Arizona. Nationally, about 9% of reported in-network claim-denial reasons involved lack of prior authorization or referral. For Blue Cross Blue Shield of Arizona, approximately 97% of its reported denial reasons fell into that category.(KFF)

That deserves emphasis:

National average: 9% of denial reasons attributed to missing prior authorization/referral
BCBS Arizona: approximately 97%

But this does not mean BCBS Arizona denied 97% of prior-authorization requests. CMS’s 2024 dataset primarily covers post-service claims—claims submitted after care was delivered. It does not provide the true pre-service prior-authorization approval/denial rate we’re interested in when asking, “Did the insurer allow the patient to get treatment?” (kff.org)

That’s actually the biggest transparency problem

Suppose a physician orders a biologic drug, cancer imaging, surgery or specialty procedure. There are two fundamentally different questions:

Question 1: Did the insurer authorize the treatment before it happened?

Question 2: After the treatment happened, did the insurer pay the claim?

The CMS dataset we’ve been discussing does a reasonably good job with #2. It does a poor job with #1. And #1 may be far more important from a patient-access standpoint. CMS itself says its Transparency in Coverage PUF contains issuer- and plan-level claims and appeals information; the 2026 PUF reports data from the 2024 plan year. (CMS)

Nonprofit status is therefore a weak predictor of denials

The data reveal something I think is important if you’re examining the role of nonprofit health insurers. Consider the range:

Nonprofit/mutual

• GuideWell/Florida Blue — 22%
• BCBS Tennessee — 21%
• BCBS North Carolina — 19%
• BCBS Alabama — 19%
• HCSC — 18%
• CareSource — 18%
• BCBS Michigan — 17%
• Arkansas BCBS — 16%
• BCBS South Carolina — 15%

For-profit

• Oscar — 25%
• Molina — 22%
• Cigna — 21%
• UnitedHealth — 19%
• Centene — 18%
• Elevance — 8%

There’s massive overlap. The corporate tax/ownership designation appears much less predictive of denial behavior than the individual insurer’s claims-adjudication processes, utilization-management rules, network design, prior-authorization requirements, and administrative systems. And perhaps the most counterintuitive number is this:

Roughly 61% of denial reasons had nothing explicitly to do with whether the treatment was medically appropriate.

That’s 36% “other” + 25% administrative. Only about 5% of reported denial reasons were explicitly medical necessity. (KFF)

That doesn’t mean only 5% of disputes concern access to appropriate care—the CMS categories and reporting methodology aren’t good enough to draw that conclusion. It means the publicly available federal data can’t adequately tell us.

One more statistic I wouldn’t overlook

Patients appealed less than 1% of approximately 85 million denied in-network claims. When patients did appeal, insurers upheld about 66%, meaning roughly 34% were not upheld at internal appeal. (KFF)

That is potentially more consequential than the nonprofit/for-profit distinction: an enormous number of claims are denied, almost nobody formally appeals them, yet a meaningful share of the relatively tiny number that are challenged are reversed.

Bottom line: based on the large-insurer CMS Marketplace data, there is no evidence that simply being a nonprofit insurer produces substantially fewer claims denials. In this sample, the weighted denial rate is actually about 19.4% for nonprofit/mutual organizations versus 18.5% for for-profits. The much bigger story is the extraordinary variation within both groups—and how little publicly available data reveal about actual treatment and prior-authorization refusals.

2024 Marketplace claim denial rates

Large parent organizations processing more than 5 million HealthCare.gov claims. Ownership shown in labels; percentages are in-network claims ultimately denied.