The 2028 policy primary has kicked off. In the realm of healthcare, the Searchlight Institute and the Center for American Progress recently put out white papers proposing tweaks to the system. We will likely see many more of these proposals in the coming months, including from the left. To best understand these competing agendas, it is useful to take a step back and take account of what our current healthcare system actually looks like and what sort of problems it has. This is what I aim to do in this post.
1. The System at a Glance
It is difficult to capture the US healthcare system in a single diagram, but the following illustration is a good place to start.
Money initially flows from households in the form of taxes, foregone wages, and premiums into a middle layer of governments, employers, and insurers. Money then bounces around within that middle layer until some portion of it eventually pops out the other end and makes its way to healthcare providers like doctors and hospitals. Money also flows directly from households to providers in the form of self-pay, copays, coinsurance, and deductibles, i.e. out-of-pocket expenditures.
This system is tremendously expensive. The United States spends over 17 percent of its GDP on healthcare while peer nations spend one-third less than that.
This higher price is not because the United States uses much more healthcare. According to the OECD data, which is a little bit old, healthcare utilization in the US does not, in aggregate, differ that much from peer nations.
What does differ is how much money the United States spends per unit of medical care. This excess spending leaks out in the form of needless administrative expenses, high drug prices, and provider rents.
2. Households
One way to understand all of the problems with this system is to move from node to node in the diagram above, starting with households. In one way or another, households ultimately suffer from all of the consequences of the system, but in this section I will focus on a few of the direct harms.
2a. Uninsurance
Despite spending much more than peer nations on healthcare, the United States fails to even formally insure its entire population. Most estimates of the uninsurance rate either provide an annual uninsurance rate, which defines someone as uninsured only if they were uninsured for the entire calendar year, or a point-in-time uninsurance rate, which typically asks people if they are currently uninsured. The following graph uses the Survey of Income and Program Participation to show these two figures.
These annual and point-in-time figures can lead you to believe that uninsurance is not that common. But this is not the case. Using the Survey of Income and Program Participation (SIPP), we can follow individuals for four years to see what percent of them faced at least one month of uninsurance over that period. The below graph shows this figure for all ages (solid lines), for all ages except the elderly (dashed lines), and for all ages except the elderly and children (dotted lines).
In a typical four year period, around 1 in 5 people experience uninsurance at some point. Medicare and Medicaid keep uninsurance down for the elderly and children. Thus, for non-elderly adults, around 1 in 4 experience uninsurance.
2b. Insurance Churn
One of the basic design principles of the US healthcare system is year-long health plans. The idea is that individuals will generally enroll in a plan during an annual open-enrollment period and stick with it for the next year. In reality, insurance churn — the name given to moving in and out of different insurance types and plans — is very prevalent, even within a calendar year.
Outside of the elderly, healthcare in America is just a constant nightmare of insurance churn. Lose or change your job? Insurance churn. Get your hours cut to part-time? Insurance churn. Employer decides to change or eliminate their health benefits? Insurance churn. Grow your income beyond Medicaid or ACA subsidy cutoffs? Insurance churn. Turn 26 years old? Insurance churn. Graduate from college? Insurance churn. Primary policyholder dies? Insurance churn. Divorce? Insurance churn. Move to a different area? Insurance churn.
In addition to this churning being a huge hassle, it also collides with other basic design principles of the US healthcare system. Most significantly, it collides with the idea of annual deductibles and out-of-pocket maximums. Someone who is forced to switch insurance in the middle of the year sees their deductible and out-of-pocket maximum start over from zero even if they have already spent thousands of dollars towards those limits in that year.
2c. Healthcare Affordability
Even if you have stable insurance, that does not necessarily mean that healthcare is always affordable. Insurance plans typically come with cost-sharing rules that make it difficult for many to actually use the healthcare system.
Healthcare affordability problems can present as foregone treatment, medical debt, or large bills. One particularly alarming way to see the effect of large bills is to look at out-of-pocket medical spending across the SPM income distribution. SPM income, which is used to calculate the supplemental poverty measure, deducts out-of-pocket health spending from each family's income. In general, this sort of spending rises with income. But there is an exception to this: at the very bottom of the income distribution, out-of-pocket health spending is massive.
That spike on the left is being driven by people whose healthcare bills were so large that they completely wiped out their annual income. One upshot of this graph is that, within the supplemental poverty metric, the poorest people each year are overwhelmingly people who got hit with a catastrophic medical bill.
3. Governments
Federal and state governments, especially the former, are responsible for structuring the whole system and so, in some sense, they are to blame for everything. However, in this section, I will focus on tax expenditures and the outsourcing of Medicaid and Medicare to private insurance companies.
3a. Tax Exclusion for Employer-Sponsored Health Insurance
Most know that the federal government finances Medicare and Medicaid. But less understood is the fact that the federal government also finances a large portion of employer-sponsored insurance (ESI) premiums as well. This financing is done through the tax code: ESI premiums are exempt from income tax, Medicare tax, and Social Security tax.
One consequence of this policy is that the federal subsidy for employer-sponsored insurance varies based on each tax unit's income. For the lowest earners, this federal subsidy covers only 14.2 percent of ESI premiums. For the highest earners, it covers 41.9 percent.
The same thing is true of the tax subsidies for Health Savings Accounts (HSAs) and Health Flexible Spending Accounts (FSAs) as well. For every $100 a low-earner places into one of those accounts, the federal government kicks back $14.20. For a very high-earner, the federal government kicks back $41.90.
3b. FSAs and HSAs
FSAs and HSAs, which allow workers to funnel "pretax money" into accounts that they can then use on certain medical expenditures, have additional problems beyond unequal federal subsidies.
FSA users are subject to strict use-it-or-lose-it rules that require them to forfeit any money they fail to use by the end of the plan year. This induces some FSA users to engage in wasteful medical spending near the end of their plan year purely to try to avoid forfeiture. On the other end of things, either due to a lack of need or forgetfulness, around half of all FSA users fail to use all of their money and thus face a forfeiture. These forfeitures, which are paid out to employers, cost workers around $4 billion every year. Thus, through the FSA program, workers gift about $4 billion of their own wages to their employers every year.
HSAs, which can only be used if an individual is enrolled in a high-deductible health plan (HDHP), are not subject to use-it-or-lose-it rules like FSAs are. HSAs are perpetual, like a 401k or an IRA. The perpetual existence of HSAs has led certain rich, savvy people to use them as backdoor tax-advantaged retirement accounts. Unlike regular retirement accounts, which get a tax advantage at contribution or distribution, HSAs enjoy a one-of-a-kind triple tax advantage: contributions, investment growth, and distributions are all untaxed. On the top end, then, HSAs are diluting the federal tax base in a way policymakers never anticipated or intended.
On the other end of things, the perpetual existence of HSAs has created a problem of orphaned accounts. Individuals often wind up with an HSA because it came along with an employer insurance option they selected. Their employer contributes small sums of money to the HSA each month, e.g. $50, but the employee never realizes they have an HSA or forgets about it, especially after they leave the company. Once an account is orphaned and contributions to it cease, whatever financial institution is responsible for the account will eventually drain the balance to $0 with maintenance fees. Thus, in the same way that FSAs facilitate the gifting of worker wages to employers, HSAs facilitate the gifting of worker wages to financial institutions.
In recent years, companies like Truemed have emerged to facilitate HSA and FSA fraud. Traditionally, HSA and FSA money could only be used to purchase certain medical items. Truemed has upended this by teaming up with companies that sell products that have some kind of marginal relationship to health and issuing Letters of Medical Necessity (LMN) to their customers so that they can pay for them with their HSAs or FSAs. For instance, Garmin sells this $3,200 watch and advertises that Truemed will issue an LMN for it. This would allow a very high-earner to essentially get a $1,300 discount on the watch, paid for by the federal government.
3c. Medicare Advantage and Medicaid Managed Care
Colloquially, "Medicare" and "Medicaid" are often used to refer to the traditional fee-for-service (FFS) plans run by federal and state governments. But Medicare and Medicaid have undergone substantial privatizations in recent years. No longer are these programs in which taxes flow through governments to medical providers. Nowadays, there is typically a private insurer sitting in between the government and the medical provider.
Adding a private insurer layer inside these programs is pointless and costly. Indeed, essentially every major health insurer (Kaiser, UnitedHealth, Cigna, Humana, Aetna, Anthem) that participates in Medicare Advantage has been targeted at one point or another for fraud, typically a kind of reimbursement fraud where they overstate the health risks of their customers to collect huge sums of money from the federal government.
4. Employers
One could imagine a hypothetical corporatist welfare system in which workers and their dependents received insurance through collective funds run by employers while everyone else received insurance through the state. This would not be a great system, and no country — not even the most corporatist-friendly like Switzerland — does this. But one could at least imagine it.
Sometimes, it seems like this is what the US is going for. But rather than have employer insurance run through some kind of collective fund where all the employers contributed some kind of predictable amount scaled to their payroll, we instead make it the individual responsibility of each particular employer to set up health insurance. Big companies often do this through self-funded plans that they hire a third-party to administer while others do it through contracts with private insurers.
Immediately, some obvious problems emerge from this model. The first is that the average cost of insurance varies from $9,300 for a single plan to $27,000 for a family plan. For someone who works 2,000 hours per year, this is equal to about $4.65/hr for the single plan or $13.50/hr for a family plan. For someone working 30 hours a week, which is when the employer mandate to provide insurance kicks in, the single plan is $6/hr while the family plan is $17.30/hr. For low-wage employers, this sort of head tax would be immense, especially given that the federal subsidy for employer-provided insurance is relatively small for low-earners, as noted in section 3a above.
To avoid this hit, low-wage employers create schedules that ensure that their workers get less than 30 hours a week, thereby avoiding the employer mandate threshold. This is rational for employers, but it wreaks havoc on the lives of workers who cannot get full-time hours and leads to an underutilization of labor factors in the economy. Even employers that might want to give their workers full-time hours and absorb the cost of health insurance end up having to think twice about putting themselves at such a disadvantage relative to their competitors. Entire sectors of the US economy end up with labor models that are completely distorted around the employer insurance system.
Another problem with this model is that employers end up spending some of their time organizing welfare benefits, whether they are good at it or not. This is true of all employer-based welfare systems, including pensions, leave, child care flexible savings accounts, life insurance, and so on down the line. Creating independent welfare bureaucracies within each corporation is redundant and inefficient and pulls business managers away from actual production.
Workers also suffer from this employer-based insurance model. Employer health insurance directly and indirectly chews into worker wages. Potential raises get reduced or eliminated altogether to finance escalating health insurance costs. Employer-sponsored insurance increases job-switching frictions, which makes it harder for workers to raise their wage by taking advantage of outside options. ESI increases the economic penalties associated with getting fired or striking, which makes it harder for workers to raise their wage through organizing and collective bargaining.
5. Private Insurance
Private insurers keep around 15 to 20 percent of every dollar that is paid to them. Traditional Medicare does the same thing private insurers do but only keeps around 2 percent. The gap between the two is pure waste. It serves no purpose. It adds no value. It is rent. It is a scam.
Private health insurers have always been an unnecessary menace in the system, but, after Obamacare, they became almost comically unnecessary.
In other sectors, insurers assess risk and charge premiums accordingly. Indeed, this is the general description of what the business of insurance actually is. However, Obamacare largely banned health insurers from doing this because it led to sky-high premiums for people with pre-existing conditions and certain other kinds of price discrimination, e.g. on the basis of gender.
In other sectors, insurers have a lot of leeway to determine what their products will actually cover. In theory, this leads to competition and consumer choice. Obamacare largely eliminated this flexibility as well by requiring all insurers to cover certain essential health benefits without lifetime or annual caps.
In other sectors, the amount that insurers charge for overhead and profit is not set by law but is theoretically determined by market competition. The Obamacare Medical Loss Ratio (MLR) rule effectively caps overhead and profit at 15 to 20 percent of revenue. This rule also reduces the incentive of the insurance sector as a whole to keep medical costs down: the higher the medical costs, the more profit insurers are allowed to book.
So American health insurers are largely not involved in assessing risk, establishing benefit packages, or even establishing their own prices. They are essentially third-party administrators of the Obamacare-regulated system. That is better than the alternative of having them actually act like insurers, but it calls into question why they should receive so much money for administrative work that the government can do at a fraction of the cost.
6. Providers
Finally, we get to the providers. For the providers, we have two main problems: rents and administration.
The United States pays more for drugs than any other country in the world. This is partially a function of the US patent system but also separately a function of a policy decision to pay that much. Other countries, which also respect US drug patents, force drug companies to charge much lower prices whether through formal price controls or monopsonistic bargaining. The US does neither and the drug companies are happy to charge massive prices far in excess of what would be necessary to induce them to sell the drug (i.e. rents), thereby driving US healthcare costs up.
Other healthcare providers, like doctors and hospitals, do much the same, primarily by taking advantage of the privately insured. But these other healthcare providers also have uniquely high costs that are associated with our fractured multi-payer health insurance system. Specifically, provider-side administrative costs are 19 percent, which is more than 50 percent higher than they would be in a single-payer system (12 percent).
Cutting these provider rents and administrative costs would save Americans a tremendous amount of money every year.
Conclusion
The left's answer to this is, by now, pretty well known. By collapsing the middle layer into a single public insurer offering a single public health plan with a high actuarial value, virtually all of these problems could be solved.
A universal public plan would eliminate uninsurance and insurance churn. A high actuarial value would eliminate the affordability problems of high out-of-pocket cost-sharing. Employer-sponsored insurance, FSAs, and HSAs would be eliminated along with the unequal tax subsidies that help fund them. Private insurers, including those operating through the Medicare Advantage and Medicaid Managed Care systems, would be eliminated, which would directly recoup the excess administrative costs of private insurers and indirectly recoup the excess provider-side administrative costs that are caused by our multi-payer system. Employers would be liberated from the burden of providing health insurance as well as the employer mandate rules that push them to immiserate low-wage workers by underscheduling them. Provider-side rents would be neutralized by monopsonistic bargaining or price-setting.
To the extent that this sort of ideal is not currently achievable or deemed too disruptive, things short of this that make steps in this direction might also be worth considering. In the past, I have endorsed Medicare for Kids as a logical gradualist stepping point. In subsequent pieces, I will elaborate on that as well as the current proposals offered by more centrist policy outfits.