Background

For as much as everyone complains about employer-sponsored health care as the “original sin”¹ of the American system, it can sometimes seem like employers offering their employees healthcare is an emergent quality of modern, developed markets. Despite France’s lauded Sécurité sociale single payer program, Paris-based startup Alan covers more than 500k people with close to $200m raised and a valuation close to $3b. Employer-sponsored coverage plays a role in Germany and Singapore, two other countries known for their efficient health care models, and employers the world over offer health-adjacent perks. It’s a sort of carcinisation that gives credence to a claim that Russell makes below: employers are a natural group. 

But as health systems around the world seemingly converge on employer-sponsored health to one degree or another, there’s discontent among the employers footing the bill in the United States. This discontent has everything to do with rising prices with employers “facing the biggest increase in health costs in more than a decade — almost 9 percent on average.” 

So what is an employer to do? Market forces, regulations, and tradition all but require them to offer health insurance, but increasingly unsustainable cost growth makes it another problem they need to manage, and one that is outside their core business. A clever and increasingly popular choice is to self-fund insurance. Cutting out insurers and their administrative fees, along with putting employee skin in the game, seems like an attractive choice, although it’s tantamount to running your own insurance company internally, which comes with its own trade-offs. Yuzu and the Plan Designers it supports are trying to change the calculus for employers considering self-insurance by making it radically easier and less expensive for employers to build their own insurance. In this wide-ranging conversation with co-founder Russell Pekala, we talk about market structure, reinsurance, and why, in his view, employer-sponsored is the optimal choice for the American system.

Questions

Q1: Self-insurance and third-party administrators (TPA) are niche concepts in the health care world. When I talk to folks outside of places like Health Tech Nerds, they generally assume the name on their insurance card is the person who is collecting their premiums and paying their claims, despite it being a large and growing part of the group insurance market. What role does Yuzu play? Are you infrastructure for modern TPAs or a TPA yourself? Do you focus just on the medical benefit, or do you do pharmacy benefits as well?

Q2: In this TechCrunch article back in 2023, it mentioned that Yuzu views its primary competitors as UHC, Humana, and Aetna. I’d be curious to hear how your view of the competitive landscape has changed over time. When companies are considering Yuzu as their TPA, who are you being evaluated against?

Q3: I’ve always thought of self-insured and fully-insured as binary options, but there seems to be a growing number of plans with “mixed-funded”. Can you talk about what this means, practically, and what options are available for a self-insured curious employer beyond this self/fully insured binary?

Q4: It appears that in response to the rapidly increasing cost of healthcare, employers are more willing to consider alternative plan designs to help manage costs – i.e., co-pay only plans, as discussed in this HFMA article. How have you seen employer interest changing in plan types, and how does Yuzu support these alternative plan designs?

Q5: Especially in the PBM space, although I’d imagine it’s true on the medical benefits side, too, we’ve seen a lot of innovative new companies stall out because it’s difficult for new entrants to negotiate better prices against legacy competitors with significant economies of scale. How do you overcome this size disadvantage?

Q6: We’ve read some stories about ways TPAs can play to improve their margins at the expense of their customers. Can you talk to us about the economics of your business? How do TPAs usually get paid, and is that different from your model? Is it purely fee-based, or do you participate in shared savings/risk for high-cost claims?

Q7: There’s a Sunlife report going around about high-cost claims for self-insured employers that are pretty eye-watering, with average costs for the top claims in the hundreds of thousands and highest cost claims like a congenital anomaly that cost $12.7m and a $10.3m transplant. I’d imagine an important part of your pitch is educating potential customers about re-insurance and managing these outlier risks. How do you approach that? What other worries or misconceptions do you hear from the self-insurance curious?

Q8: Many insurance-adjacent companies have tried to put primary care at the center of cost management (DPC, HMOs, capitated models), but member churn makes return on investment tough. How does Yuzu approach primary care?

Q9: My sense is that self-insurance makes the most sense for larger employers, and the numbers prove that out to a certain extent, with “28.2% of plans with 100-199 participants were mixed-funded or self-insured… compared with 89.7% of health plans with 5,000 or more participants.” First, is that true? And if it is, is this because it’s administratively taxing for a smaller employer, so it only makes sense when you get up to a certain size? Or is it a risk-pooling problem where it only becomes attractive when you’ve got enough participants to make it work actuarially?

Q10: There’s a writer I like, Dan Davies, who shared an exchange he had with a car insurance actuary who mentioned that they didn’t cover people who worked in theater professions because of “tendency of backstage staff to give lifts to visiting musicians, and that ‘you only need to have a claim once for a Fiat 500 with a Stradivarius on the back seat and it puts you off the whole industry’” Are there types of employers where, due to the nature of work, it’s harder to get reinsurance and therefore harder to go self-funded? Or conversely, is it easier to go self-funded because a traditional group insurer might misprice policies for an incorrectly perceived risky set of employees?

Q11: The employer-sponsored benefit space has a reputation for being a bit of a slog for startups with a lot of intermediation from brokers and consultants. What have you seen as part of your go-to-market motion? I’ve heard of some startups attempting to sell directly to employer HR teams, with the risk there being that you upset brokers/consultants who then box you out of the process.

Q12: Self-insurance is a popular focus for new and innovative businesses in healthcare because, in theory, the incentive alignment is perhaps closer than anywhere else in the third-party payer system. Employers want healthy, productive employees; employees want affordable, accessible care. How does this play out in practice? Have you seen any employers who are doing this really well?

Q13: Unlike ACA plans, Medicare, and Medicaid, self-funded coverage is regulated by the Department of Labor and ERISA, along with any applicable state laws. To what degree are regulatory compliance and reporting requirements important parts of your offering? How do you manage the regulatory patchwork between the state and federal government?

Q14: The WSJ highlighted how employer healthcare costs are rising as quickly as they have in fifteen years, according to data from the benefits consultants. With all the digital health innovation that has happened in the employer market in the last decade, with companies like Livongo lauded as massive successes, what does it say about the last wave of innovation in employer-sponsored insurance that they’re still in this moment of crisis?

Q15: In the past few months, we’ve heard folks like Oscar’s CEO Mark Bertolini argue that employer-sponsored insurance shouldn’t exist, with it often discussed as the “original sin” of American healthcare. How do you process this theoretical conversation while building a new business that supports employers offering insurance? Would we not just all be better off with employees buying insurance coverage through the exchanges via a mechanism like ICHRA?

Interview

Q1: Self-insurance and third-party administrators (TPA) are niche concepts in the health care world. When I talk to folks outside of places like Health Tech Nerds, they generally assume the name on their insurance card is the person who is collecting their premiums and paying their claims, despite it being a large and growing part of the group insurance market. What role does Yuzu play? Are you infrastructure for modern TPAs or a TPA yourself? Do you focus just on the medical benefit, or do you do pharmacy benefits as well?

This is a great question that we answer frequently. Answering this question requires a contextual framing that will pay off. A health plan is really composed of three parts, whether a company is fully insured or self-funded:

1. Plan Design. This includes what benefits are covered, which providers are in network, which services require prior authorization, etc. The plan design also influences what vendors are used for things such as pharmacy benefits.

2. Insurance. This is the part of health insurance that takes financial risk and ensures that the insured employer knows its maximum cost. 

3. Operations. This is the party that adjudicates and pays claims according to the plan design. This requires integrating data with all vendors in the plan to continuously process claims and communicate benefits, eligibility, claims, and payment information to members, employers, advisors, providers, and vendors.

Most people in healthcare have a mental model in which the “insurance” function above is the central thing to pay attention to, and that plan design and operations are more like departments within an insurance company.  More than that, traditional people in healthcare think that operations and plan design are exercises in finding arbitrage opportunities: network discounts, guaranteed ROI, tax-optimal savings, exploitation of proprietary data and relationships, and mining for loopholes.

At Yuzu, we think of the Plan Design as the central thing to pay attention to. Insurance and operations, if done well, become invisible suppliers to the management of the plan design. We don’t believe that arbitrage is how a health insurance plan becomes high-quality. We believe that high-quality health insurance starts with a high-quality framework for making trade-offs that balance access, quality, and cost. The plan design is the social contract that finances the taking-care-of-oneself-and-one-other. Operations and insurance are just means to implement this social contract.

Now it’s time to describe who Yuzu serves and what we do for them. Yuzu’s customers are Plan Designers, and Yuzu’s role is to manage operations. Legally speaking, we are a TPA; however, we do not sell bundled strategies or proprietary networks like most TPAs. Instead, we differentiate by having our own fully vertical software stack that lets us be maximally flexible and transparent so that Plan Designers can work with more tools and fewer constraints in building their own “re-bundled” Health Plans.

You might think that if Plan Design is the “essence” of health insurance, then its supplier (operations) would be a commodity. This is a logical fallacy that is far from the truth. In fact, there are lots of plan design strategies that everyone agrees handle trade-offs between access, quality, and cost well, but that incumbent TPAs cannot administer. Some examples include cash payment for outpatient services (faster access, lower cost), subscription primary care (better access, higher quality), and high-performance networks (higher quality, lower cost).

In thinking about how Yuzu works with Plan Designers, we have roughly the following task split that we outline during our program implementation period.

• Plan Designer (Strategy + Explaining)

  • [Strategy] Provider contracting and relationship management.

  • [Strategy] Stoploss insurance procurement (choice of supplier).

  • [Explaining] Member services; care operations model.

  • [Explaining] End-employer sales, marketing, and account management.

  • [Strategy] Benefit design decisions.

• Yuzu (Commodity Operations + Truth)

  • [Truth] Manage connections with networks, PBMs, clearinghouses, etc.

  • [Truth] Portals for employer, member, broker, care navigator, provider, carrier.

  • [Commodity Operations] Approve/deny/process health insurance claims.

  • [Commodity Operations] Manage invoicing and payments.

  • [Commodity Operations] Maintain eligibility with employers.

  • [Commodity Operations] Provider services/phone line.

  • [Commodity Operations] Stoploss claim filing.

  • [Truth] Records maintenance.

Yuzu administers medical benefits and integrates with PBMs that manage most pharmacy benefits. Yuzu does administer some pharmacy-related claims across a number of our plans (DPC-dispensed drugs and drugs purchased on Yuzu’s cash pay cards).

Back to Questions.

Q2: In this TechCrunch article back in 2023, it mentioned that Yuzu views its primary competitors as UHC, Humana, and Aetna. I’d be curious to hear how your view of the competitive landscape has changed over time. When companies are considering Yuzu as their TPA, who are you being evaluated against?

When Yuzu was featured in 2023, we saw ourselves as a vertically integrated finished-goods producer, selling and marketing a complete health plan directly to employers. At that time, we had no customers whatsoever and were frankly humbled by how hard it is to design and administer a whole health plan. We’ve since pivoted to being a technology-enabled TPA that is essentially a supplier to health plans. The health plans we supply compete against bundled plans offered by Blue Cross, United, Cigna, and Aetna (collectively BUCA).

What’s important to note, though, is that many of the parts that go into the plans that Yuzu powers could be owned by the big BUCA brands.  More than half of the plans we currently administer include access to the Cigna or Aetna networks.  A member on one of these health plans might think their health insurance is Cigna, and this isn’t exactly incorrect, since from the member's point of view, the Cigna network is probably the most important part of the plan for them to understand. Check out the Yuzu blog post “Parts of a Health Plan” to learn more about how employers can mix and match unbundled parts like a network, PBM, and utilization management.  

Health Plans typically evaluate Yuzu against other independent but legacy TPAs that have BPO divisions, including Boon Chapman, Aither, Lucent, Sage TPA, Centivo, Loomis, S&S, and others. We’ve had several customers evaluate us against building an in-house TPA.

Back to Questions.

Q3: I’ve always thought of self-insured and fully-insured as binary options, but there seems to be a growing number of plans with “mixed-funded”. Can you talk about what this means, practically, and what options are available for a self-insured curious employer beyond this self/fully insured binary?

People use terms like “mixed-funded”, “level-funded”, and “partially self-funded” to describe self-funded employer Health Plans with a lot of stoploss insurance (and thus minimal financial risk). Financial risk is one of two dimensions that self-insured curious employers should think about, the other being benefits customizability. When a product is designed to cap risk as low as possible, the employer is typically invoiced monthly at the maximum they will ever have to pay. After a run-out period, they will sometimes be able to get a small refund if they had good plan performance.

What’s important to understand, though, is that there is a continuous spectrum of self-funded plans in this two-dimensional space of “financial risk” and “custom benefits”.  On one end of the extreme (low financial risk, low benefits configurability), a company could explore a product like United’s AllSavers or one administered on Yuzu, such as Arlo Clear, which has access to the Cigna network and very low financial variability.

Yuzu also administers plans that are built with custom benefits (e.g., DPC built-in) but still very low financial risk, such as DPC-based level-funded plans offered by HIPnation or First Primary Care. Larger employers typically prefer to take some amount of financial risk, but smaller employers typically prefer plans that cap their financial risk to be as low as possible. What’s great about self-funded is that it lets employers explore both dimensions.

Back to Questions.

Q4: It appears that in response to the rapidly increasing cost of healthcare, employers are more willing to consider alternative plan designs to help manage costs – i.e., co-pay only plans, as discussed in this HFMA article. How have you seen employer interest changing in plan types, and how does Yuzu support these alternative plan designs?

Cost is top of mind for employers, as recent premium increases are unsustainable. Many employers and their advisors correctly understand that Plan Design is the right lever to pull to lower costs without reducing quality by enabling members and providers to make better trade-offs. As I discussed in my answer to the first question, many of the plan design strategies that employers and brokers want access to are definitively good investments and plan design changes that incumbent TPAs cannot administer.

Incumbent TPAs cannot administer these new strategies because they do not control their technology stack. Without controlling their technology stack, they cannot innovate as fast as employers are demanding. Yuzu’s custom software positions us to be able to find ways to administer new strategies in a way that feels integrated to end-employers, members, and care navigators.

Examples:

• Yuzu integrates subscription-based direct primary care (DPC) into health plans by supporting several different fee strategies DPCs typically use (PMPM, PEPM, pricing by age) into our invoicing logic. We also allow care navigators and employers to manage member assignment to DPCs. We have custom portals where DPCs can view remittance information and eligibility. Since we build all the software portals for members, employers, and providers in-house, we can manage standards around this assignment without being limited by what other software can handle.

• Care navigators working within Yuzu plans can get access to cash-pay rates by using Yuzu cards to pay for medical expenses and have them auto-filed as claims.   

• Yuzu administers a copay-only plan, Arlo Clear. These plans give members simplicity and capped risk by also giving them incentives to shop around for providers with lower costs and get rewarded with lower copays.  

Most health plans that Yuzu supports are designed to control costs more effectively than traditional plans. They each use a different mix of plan design strategies to do this, and Yuzu prides itself on our ability to support these new plan design strategies natively.

Back to Questions.

Q5: Especially in the PBM space, although I’d imagine it’s true on the medical benefits side, too, we’ve seen a lot of innovative new companies stall out because it’s difficult for new entrants to negotiate better prices against legacy competitors with significant economies of scale. How do you overcome this size disadvantage?

Scale is not nearly as much of an advantage as people think. Many strategies work better when a health plan is smaller. Some, like internationally sourcing high-cost drugs, would not even work at scale, where higher regulatory scrutiny would complicate efforts to save members and employers money. Others, like having high-touch local care navigation, are difficult to scale but work well when plans are small.

Scale has other disadvantages, too, that I believe will be exacerbated by AI. When plans operate at scale, they standardize and outsource processes to entities that lack context on the local dynamics at play or customer and member preferences, and work based on checkboxes and defined procedures. This outsourcing already bureaucratizes the health plan experience (lowering customer satisfaction) but also makes it possible for sophisticated AI-powered RCM players on the provider side to reverse-engineer processes like prior authorization and medical code editing.

The challenge that Yuzu and the plans we support have is that we need to figure out ways to scale plans while maintaining the local touch and practical common-sense solutions that have made our plans run well so far. We believe that negotiation is not going to be the approach to this problem. Our approach is instead to find ways to distribute accountability and ownership so that the plans we run can continue to be empowered to make common-sense trade-offs that benefit members and employer sponsors.

Back to Questions.

Q6: We’ve read some stories about ways TPAs can play to improve their margins at the expense of their customers. Can you talk to us about the economics of your business? How do TPAs usually get paid, and is that different from your model? Is it purely fee-based, or do you participate in shared savings/risk for high-cost claims?

Yuzu only makes a flat PEPM fee that is specific to each plan designer we serve. Our fee starts at $50 PEPM with significant discounts available for customers willing to simplify their plan administration and agree to Yuzu-preferred processes and vendors.

Yuzu does not make any shared savings, kickbacks, or at-risk fees because we view simplicity as one of our value props. We are not morally opposed to fee models like shared savings that can align incentives and pay for performance. In fact, several of the plans we work with collect these types of fees from employer-customers. Shared savings programs work when they are transparent, simple to administer, easy to measure, and don’t have operational side-effects (e.g., don’t affect other processes like reporting or stoploss filing).

Back to Questions.

Q7: There’s a Sunlife report going around about high-cost claims for self-insured employers that are pretty eye-watering, with average costs for the top claims in the hundreds of thousands and highest cost claims like a congenital anomaly that cost $12.7m and a $10.3m transplant. I’d imagine an important part of your pitch is educating potential customers about re-insurance and managing these outlier risks. How do you approach that? What other worries or misconceptions do you hear from the self-insurance curious?

All of Yuzu’s customers have stoploss insurance designed to exactly reinsure the benefits of the plan. The Plan Designers that Yuzu partners with help employers procure and understand their stoploss policy, and in general, employers feel comfortable with the mechanics of stoploss since it is so abstracted away from them. Yuzu’s role with respect to stoploss is to make sure that we file stoploss claims in a timely manner and administer the plan in a way that does not leave gaps (paid claims not reimbursable under stoploss).

Eligibility is the trickiest part of this, because Yuzu relies on employers to keep us updated on when employees and dependents are added and removed from the plan. Yuzu has built many features to help employers stay up to date with eligibility, including HRIS integrations, automatic eligibility questionnaires sent out when high-cost claims come in, and detailed invoicing showing adds and terms. Employers still occasionally struggle to keep their eligibility current, which can impact stoploss reimbursement.

Carriers in particular press Yuzu to have strategies to contain these million-dollar-plus claims. Carriers are thankful that Yuzu gives them real-time data access to claims (even pre-adjudicated claims) and notices of potential high-cost claimants. Yuzu can operationalize creative solutions and contracts for these high-cost events.

Back to Questions.

Q8: Many insurance-adjacent companies have tried to put primary care at the center of cost management (DPC, HMOs, capitated models), but member churn makes return on investment tough. How does Yuzu approach primary care?

I am tired of hearing that it takes many years before a primary care investment pays off. An investment in enhanced primary care pays off in the first year if incorporated natively into a health plan. So, what does a native integration between something like DPC and a Health Plan look like?

• The Health Plan leans on DPC in its case management.

• The Health Plan works around and learns from DPC relationships with cash-pay providers in their region to manage downstream care.

• DPCs get paid automatically over ACH for an entire population of employees.

• DPCs encourage patients to trust plan steerage.  

There are two things that I believe hold back the ROI of enhanced primary care in other plans:

• “Conditional” trust with primary care doctors based on a techno-data-bureaucracy mindset. This is the trap that a lot of the ACO world has fallen into, where payment models are based on quality metrics set by insurance companies. This increases the documentation load on primary care, especially, and reduces the agency that on-the-ground clinicians have in solving problems. I believe it is possible for engaged employers to evaluate the quality of primary care doctors without relying on vast quantities of structured data.

• Thinking of DPC as an add-on that doesn’t affect other benefits. It is magical thinking to believe that members will use DPC as designed when it exists outside of what they think of as their insurance. Employers that succeed with Direct Primary Care know how to put it at the middle of the health plan (raising copays on other primary care, engaging members continuously, finding creative ways to introduce the DPC doctors to members at informal company events).

In short, I think that primary care engagement works when it can make use of real personal relationships between primary care providers and plans. Making this model work at scale is an unsolved problem. Yuzu’s approach is to think of DPCs as true partners in the health plans we administer with them, rather than just vendors to manage.

Back to Questions.

Q9: My sense is that self-insurance makes the most sense for larger employers, and the numbers prove that out to a certain extent, with “28.2% of plans with 100-199 participants were mixed-funded or self-insured… compared with 89.7% of health plans with 5,000 or more participants.” First, is that true? And if it is, is this because it’s administratively taxing for a smaller employer, so it only makes sense when you get up to a certain size? Or is it a risk-pooling problem where it only becomes attractive when you’ve got enough participants to make it work actuarially?

I get this question a lot. The general trend of larger companies being more likely to self-fund is true; however, the reason for this has less to do with risk-pooling and more to do with regulations and administration.

• Stoploss insurance is regulated at the state level, and many states do not allow small businesses to buy stoploss insurance if they have fewer than some number (often between 25 and 100) of employees. States do this because they want to protect the risk pool of the small group marketplace, and they worry that if healthy small groups were able to buy stoploss insurance, then only unhealthy groups would be left on the fully insured market (leading to a death spiral).

• Administratively, it is more complicated for businesses to self-fund benefits than to fully-insure those benefits. Companies like Yuzu are working to close that gap, but there are parts of self-funded that always take more work for employers, including getting an underwritten stoploss quote and opening a bank account for the use of claims processing.

Many carriers specialize in level-funded underwriting and are very comfortable with the risk of these smaller groups. They have a large enough block of business across all their small groups that they can give rates to individual small groups based on the group’s expected claims, not based on the group’s size.

Back to Questions.

Q10: There’s a writer I like, Dan Davies, who shared an exchange he had with a car insurance actuary who mentioned that they didn’t cover people who worked in theater professions because of “tendency of backstage staff to give lifts to visiting musicians, and that ‘you only need to have a claim once for a Fiat 500 with a Stradivarius on the back seat and it puts you off the whole industry’” Are there types of employers where, due to the nature of work, it’s harder to get reinsurance and therefore harder to go self-funded? Or conversely, is it easier to go self-funded because a traditional group insurer might misprice policies for an incorrectly perceived risky set of employees?

Yuzu does not do any underwriting and can support any carrier that has a good reputation for paying claims. Each of the plans we work with has its own underwriting relationships and occasionally keeps two or so carriers in the mix to ensure that quotes for individual employers remain competitive.

There are many different stoploss carriers and managing general underwriters (MGUs) out there. Anecdotally, I have learned they each do have different industries they might like or not like to underwrite. Some carriers won’t underwrite medical cannabis companies because there is no clear federal guidance establishing the legality of processing financial transactions (like claims) for these companies. I’ve also met an insurance company that has actively told us they love underwriting cannabis companies because their employees tend to be young males (favorable demographics for health insurance). There’s another carrier we work with that won’t underwrite law firms, believing that they would be likely to face expensive litigation in the event of an adverse claim outcome.  Another carrier doesn’t like underwriting public entities (schools, municipal governments, etc) because they believe that public-sector employees lack frugality and common financial sense because public entities don’t have a profit motive and are already (in the carrier’s opinion) not responsibly spending taxpayer money making it unlikely that they would embrace strategies to control costs like case management and steerage.

In general, there are many options out there for employers, and Yuzu can work with a wide variety of carriers.

Back to Questions.

Q11: The employer-sponsored benefit space has a reputation for being a bit of a slog for startups with a lot of intermediation from brokers and consultants. What have you seen as part of your go-to-market motion? I’ve heard of some startups attempting to sell directly to employer HR teams, with the risk there being that you upset brokers/consultants who then box you out of the process.

Yuzu reaches employers through the Plan Designers whose plans we power. Some of these Plan Designers have broker partnership models, and some of them go direct to employers.

I used to be dismissive of brokers but have come to appreciate their role more over the last year or so. I view them less as a distribution channel and more as a partner in operating the health plan who is uniquely positioned to help with account management and customer service. Yuzu’s role is to make them valuable by equipping them with what they need to explain insurance to members and employers. If we can make them successful at this, then everyone wins because plans operate as designed.

Back to Questions.

Q12: Self-insurance is a popular focus for new and innovative businesses in healthcare because, in theory, the incentive alignment is perhaps closer than anywhere else in the third-party payer system. Employers want healthy, productive employees; employees want affordable, accessible care. How does this play out in practice? Have you seen any employers who are doing this really well?

The beautiful thing about self-funded is that it can get employers and employees to understand and work to improve the trade-offs involved in their benefits because that incentive alignment is very close. 

The best employers all do a few things to make sure their benefits work:

• HR leaders enroll in the plan and publicly engage with the plan’s cost-containment strategies.

• Employers explain to employees why a health plan change is being made, e.g., “we’ve had unsustainable premium increases, we are trying a new strategy rather than increasing costs to you”.

• Employers encourage employees to lean into initiatives like DPC and give feedback. One employer CFO we work with has a policy where he’ll give any employee a $100 bill if they share their feedback (positive or negative) on the Direct Primary Care doctor the employer has contracted with.

Back to Questions.

Q13: Unlike ACA plans, Medicare, and Medicaid, self-funded coverage is regulated by the Department of Labor and ERISA, along with any applicable state laws. To what degree are regulatory compliance and reporting requirements important parts of your offering? How do you manage the regulatory patchwork between the state and federal government?

Getting into this industry from scratch, I spent a lot of time building a good mental framework for the tragically complicated health insurance legal landscape. Core elements of my framework are:

• Understand the difference between statutes, regulations, and guidelines.

  • If something is required by statute, then it’s unlikely to change since change would require new laws being passed or statutes being struck down by courts.

  • Regulations change constantly as they are controlled by executive branch bureaucrats. Regulations are frequently reversed and edited when administrations change. As a current example, Secretary Kennedy is changing which vaccines are considered preventive and thus required by the ACA to be covered at $0 to members.

• Understand what enforcement looks like for various laws and regulations.

  • The Trump administration is halting enforcement of parts of the Mental Health Parity law.

• Understand what is regulated at a state level versus at a federal level.

  • ERISA is regulated at a federal level and pre-empts state laws regarding benefits and claims procedures.

  • Insurance is regulated at a state level, which is why different states have slightly different stoploss laws, and TPA licensure is handled state-by-state.

Yuzu’s approach to compliance is to really try to build as many compliance guardrails into the product as possible such as:

• Our plan-building software ensures that plans are built and contracts signed that are compliant with laws like the ACA, the No Surprises Act, and Mental Health Parity.

• Some reporting tasks are automated, like sending plan eligibility to CMS over SFTP so that they can enforce Coordination of Benefits provisions. 

• Yuzu’s real-time portals give employers and their brokers the ability to download all data needed to fill out their IRS 5500 form in real time.

We still have many compliance tasks that we do manually, but being the source of truth for plan data, we can build these tasks into auditable workflows over time.

Back to Questions.

Q14: The WSJ highlighted how employer healthcare costs are rising as quickly as they have in fifteen years, according to data from the benefits consultants. With all the digital health innovation that has happened in the employer market in the last decade, with companies like Livongo lauded as massive successes, what does it say about the last wave of innovation in employer-sponsored insurance that they’re still in this moment of crisis?

I would not say that the last wave of innovation was all that impressive or beneficial to employers. The thesis behind many of yesterday’s innovators, like Livongo and Hinge Health, is that you can build point solutions or high-performance networks for specific conditions and then sell access to employers. There was this belief that each of these point solutions had some sort of guaranteed ROI, and that it was the role of employers and brokers to purchase ROI rather than construct coherent and holistic plans.

This model had two primary flaws:

• Distribution became the bottleneck for most of these solutions. Hinge Health spent more than 50% of revenue on sales and marketing in many quarters [src].

• Plans became harder for members and employers to understand. When a plan is complicated and not cohesive, it leads to higher administrative costs and lower member engagement in high-ROI strategies. 

Yuzu believes that the problem with health insurance is, to some extent, its complexity. We aim to support plans that are built on local relationships between plans and providers, which let people who have more context than Yuzu develop the right trade-offs in benefit design.  Our approach is much more “bottom-up” than the top-down approach of purchasing point solutions evaluated on heavily doctored ROI numbers.

Back to Questions.

Q15: In the past few months, we’ve heard folks like Oscar’s CEO Mark Bertolini argue that employer-sponsored insurance shouldn’t exist, with it often discussed as the “original sin” of American healthcare. How do you process this theoretical conversation while building a new business that supports employers offering insurance? Would we not just all be better off with employees buying insurance coverage through the exchanges via a mechanism like ICHRA?

I get this question all the time, and incorporated some of my thoughts in my blog post on ICHRAs. I’ll try to summarize those thoughts below.

I am sympathetic to Mark Bertolini’s argument. Indeed, it seems democratic and even market-optimal to let individuals shop for their own insurance plan. From a microeconomics perspective, there is certainly more consumer surplus obtained by letting healthy people choose cheaper plans and sicker people choose richer plans. However, this produces an issue at the macro-level if you insist on a single risk pool and do not allow insurers to underwrite each person’s risk (which the ACA forbade).

At the macro-level, having an individual marketplace with a single risk pool creates numerous bad incentives:

• Insurers will intentionally make their plans worse to control selection into their plans.

• Healthy people will leave the marketplace because the expected value of their insured benefits becomes too low for them to stomach the cost.

The ACA marketplace has tried to add band-aids like risk adjustment and subsidized premiums to help correct these two second-order effects. These band-aids have been ineffective and have ultimately contributed to healthcare becoming more administratively complex and expensive. The ACA market is still in a death spiral, and the government is spending an unsustainable amount of money trying to patch holes in this model.

When you analyze this problem from first principles, what becomes clear is that health insurance is not an individual product: it is a policy, or social contract, between everyone whose risk is pooled together (including healthy people whose primary policy interest is in desiring cheaper insurance). When you view health insurance as a policy chosen by the people who pay for that policy, it makes sense that it is unfair to allow individuals to choose their own insurance (just like it is unfair to let citizens choose which taxes they want to pay).

There are essentially three solutions to this problem that work at the macro level:

• Go back to a pre-ACA world where individuals can be underwritten. Then policies could be adapted to and sold to individuals with many different choices available. A downside of this model is that some individuals would be too risky to insure.

• Put everyone in the same risk pool (e.g., single payer). Allowing some degree of individual choice in this model is hard and potentially not possible. The evolution of the Medicare Advantage model has, to some extent proven this out: incentives are too strong for insurers to game risk-adjustment or design policies to optimize risk-selection rather than risk-management.

• Find a market-friendly way to create local risk pools that give essentially guaranteed-issue insurance and still enable plan design choice with low bureaucracy.

This third option is what employer-sponsored insurance is. Employers are not the only grouping that could be used to create risk pools, but overall, they are a pretty good grouping and are considered “natural groups”. Other “natural groups” of people that could be used to create risk pools include possibly stratifying people by birthdate, address, church affiliation, or social security number range.

Employers are convenient because:

• Individuals still choose their employer based on their benefits. Thus, employees still have the ability in this model to choose employers who offer desirable benefits.

• Employers typically operate in a defined region with people who have similar values, at least more so than other natural groups of people.

• Administrative tasks like premium collection and benefits education are simple for employers.

Even in other countries with single risk-pools, employer-sponsored insurance is taking off. Governments simply are not very efficient at administering insurance, for the same reason that communism might be a good idea, but doesn’t work to lift people’s standard of living.

Back to Questions.

1  a surprisingly popular metaphor