I Have a Dream… of a Post-Network World
The carrier gangsters have spent decades making sure you can't leave. Here's how that ends.
I have a dream of a post-network world.
Not in a think tank’s this-will-never-happen-but-we’ll-still-write-white-papers-about-it sense. I mean a real world in which employers - the ones running the health plans that cover most Americans - can buy health care the way they buy everything else: by comparing prices, negotiating directly, and refusing to overpay when a seller is out of line.
Because right now, that’s obviously not how this market functions.
Carrier networks operate less like marketplaces and more like Hotel California.
Employers are allowed to choose between networks, but they are not meaningfully allowed to operate outside of them. When they try, they discover very quickly that the system has been engineered to make that choice cartoonishly painful.
Once you’ve chosen a network, it operates like a plantation.
Employers - the ones funding the entire enterprise - are held hostage to that network. This isn’t an accident - it’s the result of specific contractual hostage-taking, honed over years of dominating the health care landscape.
But I’m increasingly running into people building the underground railroad, allowing a few, obscure, invite-only escape routes. Some have been doing this for a couple years. Some are just starting. The tipping point isn’t even showing yet, but it’s on its way. I’m starting to partner with these insurgents, and so are others.
At this point, we’re all just hobbits taking on Mordor (but you know how that story ends).
To understand how we get out, let’s walk through how the system works to trap us in the first place.
[Industry people: Yes, we’re doing the 101. Feel free to skip ahead.]
Every employer-sponsored health plan has three core functions (even if they are bundled together in practice for most):
The first is plan administration. Someone has to receive the bills (“claims”) from doctors, hospitals, or pharmacies (we’re going to call them “providers” in this article even though that word makes them want to burn me at the stake, I know), figure out if the claims are for covered services, confirm the patient is actually enrolled on the plan, calculate what they’ve already paid against a deductible, and write the check to the provider. Someone also has to keep records of every transaction, pay the other vendors on the plan, and send the client regular reports. The companies that do this are third party administrators (TPAs) - and they can be either carriers themselves or they can be separate companies (“indie TPAs”).
The second is risk assumption. In other words, whose bank account is the TPA reaching into to write those checks to providers? For plans in the individual and small group market, the TPA itself is bearing risk - this is called a “fully-insured plan,” meaning the TPA is the true owner of the plan, it’s assuming your risk. If you cost them more than the premiums you pay, they lose and you win. I mean, until they raise rates for everyone next year, but you get the point. That’s why these markets are the most expensive - those insurance companies aren’t going to lose money on you - they’re going to build a huge buffer against that in the form of overpriced premiums. These plans are regulated partly by the ACA as well as state insurance commissioners.
Bigger companies (though they can be as small as 50 or 100 employees) usually choose a different model, one where the employer itself bears all or most of the risk. This is why this type of plan is called a “self-funded” plan - not because you the patient are funding things, but because the employer is funding its own risk. When you pay your premiums with payroll deductions, you’re actually paying them to your employer’s account, and your employer pays premiums each month to itself in that same account. Then the TPA writes checks to providers out of that account. Most employers that are smaller than 5,000 or so will actually buy some stoploss reinsurance for themselves, just in case of a really bad year or a particularly high-cost patient. These plans are regulated federally, by the Department of Labor (and a little bit by the ACA).
The third is pricing - or more accurately, re-pricing. Of the bills from doctors, hospitals, and pharmacies, to be specific. For some sick reason, health care is billed at a sticker price that is inflated far above what would be a reasonable cost-plus-some-profit rate. You need a way to access discounts off that list price, as a result. A network is nothing more than a set of contracts with providers for a certain contracted rate (usually, though not always, below the sticker price). There are other ways to get discounts, which we’ll discuss later.
So the claim goes FIRST to the network, which applies its contract with the provider, and sends the contracted rate, along with the claim, to the TPA for processing and payment. The big insurance companies all have their own network. So for most health plans, all this happens under one roof, either because the plan is fully-insured and the carrier is doing all three core functions, or because the plan is self-funded, and hires the carrier to both administer the plan as the TPA and to provide its network (even though that employer is paying all its own claims).
An employer can also fire the carrier as the TPA, and hire an indie TPA to do the plan administration, while still keeping a network, by renting a carrier’s network. Indie TPAs partner with a few carriers to offer their networks to these employer clients. Most clients still want a network on their plan, even if they don’t want a carrier administering their plan. So indie TPAs contract with the carrier networks as a distributor of sorts for their network. The employer then signs a network access agreement when it hires the TPA, essentially renting the network.
Again, networks are simply a set of contracts between the carrier and providers. So doctors, hospitals and pharmacies all have their own contracts with networks, if they “take insurance.” That’s what makes them “in-network.” Some providers couldn’t come to agreeable terms in a contract or they didn’t even want to try, and those are considered out-of-network. That means that there’s no agreement as to how much they get paid for members of that plan.
So, regardless of how the pieces are arranged, almost every employer in the country ends up in the same position: dependent on a carrier network for access to care.
And once you are inside that network, you are not just getting a price list. You are entering into a set of contractual obligations that govern how care can be purchased, and, critically, what alternatives are off-limits.
Most employers do not realize how restrictive those obligations are until they try to step outside them.
The Obvious Question: Why Not Just Eliminate Networks?
Some employers have asked that question. A small number have gone further and actually done it. This is what is commonly referred to as reference-based pricing (RBP) plans.
Instead of relying on a carrier network, an RBP plan sets its own allowable reimbursement rate for providers, typically indexed to some percent of Medicare. As we’ve discussed in my post about how not to get financially screwed by the medical system, anything up to about 150 percent of the Medicare rate is considered pretty good. 150-200 percent is fair. 200-250 is iffy, depending on the type of service. 250 percent and above is rapey (it’s also the nationwide average carrier rate, which is precisely why RBP plans exist - to pay less than they would with a carrier network). On an RBP plan, patients can, in theory, go to any provider, and the plan pays a defined amount, usually somewhere between 125-180 percent of Medicare rates.
At first glance, this looks like the cleanest possible solution. You remove the middleman. You define a rational price. You let the market work.
In practice, it introduces a set of tradeoffs that are often minimized in sales pitches and only fully understood once a plan is live. I have built and run this type of plan, so I’m not throwing stones, just being honest.
The first problem an RBP plan faces and must find a solution for is balance billing. By definition, there’s no contract/network. So, providers have not agreed to your Medicare-based rate (“plan allowable”), and, absent a contract, they retain the right to bill the patient for the difference between what the plan pays and their billed charges. There are vendors who specialize in negotiating those balances down after the fact, often landing somewhere above the plan’s allowable but below commercial rates. That process can work when your vendor is good, but it is reactive, and it puts the patient in the middle of a negotiation they did not initiate, sometimes even being sent to collections while the parties in the background are negotiating (or playing hardball) - sometimes for months and months.
That is the visible problem.
The more serious problem is one step earlier in the process, and it is less discussed because it is harder to solve: patients often can’t get the appointment in the first place.
I learned this the hard way with a small group of Catholic nuns. I’ve mentioned them in previous posts. They were so desperately tired of being bankrupted by Blue Cross, they begged me for a solution that saved significantly. So we went with my first (and last) RBP plan.
Recently, a desperately sick sister needed to see a specialist at NYU-Langone Medical Center. Historically, that system had seen patients from this plan without issue. But something shifted a year ago or so - subtle changes in how they describe on their website about accepted insurance coverage - a tightening of language that signaled a change in posture.
When Sister called to schedule, she was told they did not take her plan. She tried to explain, correctly, that she did not have a network and could go anywhere. The response did not change. They would not schedule her.
The only option they offered was for her to pay cash upfront.
That might sound like a workable fallback until you remember who we are talking about. She had no income. She did not have a bank account. “Pay upfront” was not a logistical inconvenience; it was a hard stop.
We tried to engineer around it. Could the TPA send funds in advance? Yes, but only to the sister, they can’t pay a hospital for care not yet rendered. But she didn’t have a bank account, couldn’t receive a check. Could the TPA just pay with a credit card and be reimbursed by the plan later? This TPA - like most - wasn’t set up for that. Typically, auditors frown on payments for services not yet rendered, as well as on TPA transfers to themselves from client bank accounts.
As I do when I’m in a pickle with a hospital, I called my dear friend who is an executive with the hospital trade group. She put me in touch with an NYU executive. He didn’t understand what an RBP plan was, and kept suggesting that we were a health care sharing ministry or some other type of non-plan. In other words, a hospital executive was flummoxed in exactly the same way that most front-desk or scheduling staff are when patients call and try to explain their RBP plans.
She needed care. The system would not let her access it. The sisters scrambled to come up with the cash - which they raised from benefactors - and live without it somehow until the plan could reimburse them.
This is the tradeoff: reference-based pricing plans can be extremely inexpensive, but they shift friction onto the patient in ways that are not always tolerable.
“Cheap” is no help if you can’t get in the door.
The Hospitals Aren’t Confused.
Hospitals understand what a threat RBP plans are to them. They are making calculated decisions.
They understand that if they create enough friction - if patients encounter enough resistance, enough uncertainty, enough delay - those pressures will flow back to the employers, which, after all, are running plans that cover their own families. All it takes is one high-friction experience with the CEO’s wife - a delayed appointment, a collections notice, an aggravated doctor - and the calculus changes.
The employer reinstates a carrier network.
The broker who recommended the alternative is replaced.
And the conclusion, reinforced through experience, is that the system cannot be changed.
A particularly notorious case of this pattern was when University of Pittsburgh (UPMC) - the health system that monopolizes half of Pennsylvania - went gangster on its own community. A bunch of public sector health plans - the local school system and various municipalities - adopted the RBP model to try to save on costs. UPMC refused to schedule any patient covered by those plans. None of them lasted more than two years before giving up and buying into the local Blue network or UPMC’s own (overpriced) plan offering. UPMC became a hero and role model for other hospital systems across the country. NYU-Langone has apparently followed their lead, icing out the Catholic nuns’ plan that they had been taking for years.
This playbook makes it clear: try to pay less by dropping networks that pay confiscatory rates to hospitals and you’ll pay less alright - by not being able to get care at all. The hospitals are in league with the networks against their own communities.
But there are early signs that the ground is shifting.
Regulatory pressure has begun to expose pricing in ways that were previously not possible. New requirements were just passed by Congress on PBMs, but they also threw in a little gem that effectively applied them to TPAs, which would force transparent disclosure from all plan vendors about their rapey rates and conflicts of interest and secret revenue streams they’ve been hiding from their client employers. The Department of Labor (DOL), which oversees larger employer plans, doubled down on this concept, essentially saying that these schemes are illegal transactions that no employer is allowed to tolerate. It’s a one-two punch - first you have to disclose your business model. Then, your clients aren’t allowed to have a plan vendor with that kind of business model. I wrote about these fantastic policy innovations here. I called them a nuclear bomb on the status quo.
Equally encouraging, the Department of Justice (DOJ) has begun to pursue antitrust enforcement actions, targeting provider-side contracting practices. Cases like the one against OhioHealth (which I wrote about here), and a similar case that just dropped on New York Presbyterian, one of the truest Bond villains in the country, focus on provisions that require plans to include entire systems, regardless of cost or quality variation between facilities in that system, as a condition of access to any of their network rates.
Those provisions matter. Removing them would create more flexibility on the provider side.
I appreciate DOJ taking action. I tried to get them to pay attention to this stuff when I worked in the White House, and they said it was FTC’s job, and FTC said it was DOJ’s job, and neither of them did their own job. But even if these contract provisions were ruled illegal, we would still be a long way from the Holy Grail of a post-network world.
It’s not that better options don’t exist that could hasten us toward that future. It’s that they are contractually suppressed.
Let’s walk through exactly how carrier agreements govern payment behavior, why even independent TPAs are constrained in ways that disadvantage innovation-minded employers, and what I believe are illegal network contract provisions preventing the emergence of a true price-based market, strangling it before it can breathe.
Then I’ll describe why the three most popular work-arounds - reference-based pricing, direct contracting, and cash pay - all run into structural limits, even when executed well.
And finally, because I’m not in the business of handing you a burning building and walking away, I’ll lay out what would actually have to change, at the policy and enforcement level, for my dream of a post-network world to become reality.
That part is for paid subscribers.
If you’re trying to build something different, advising someone who is, or trying to regulate this market in any meaningful way - you need what’s behind this paywall more than you need whatever you’re doing next.
