A lot for a little: The best 8% of our healthcare dollar

By Peter Kolchinsky

POLICY | BIOTECH

Photo by Toa Heftiba on Unsplash

We are trying something new here at RApport: an AI-generated podcast overview of ​“The best 8% of our healthcare dollar”. It’s a different way to experience this article, and we’d love to hear what you think!

November 27, 2024

Drugmakers get a lot of attention from politicians determined to demonize the industry for ripping off the American public, so you’d think we’re spending a significant chunk of our healthcare dollar on branded prescription drugs. Ask an American what fraction of healthcare spending goes to paying for branded drugs and you might hear ​“Half?” Or even ​“80%?” If they are well informed, they might guess 25%. But it turns out that number is roughly 8%.

Can it be possible that less than a dime of every premium dollar and tax dollar that goes toward private health insurance and federal programs like Medicare and Medicaid is paying for all the medicines that the drug industry has launched over the past 15 years or so (since most of the ones launched earlier have already gone generic)? 

People seem genuinely surprised when they learn this. Among those people is noted basketball enthusiast and former Microsoft CEO Steve Ballmer, who aside from starring in his own genre of excited software executive gifs also stars in his own video series called ​“Just the Facts.” A recent (heavily statistic-laden) episode from just last month is about healthcare, and it inspired us to dig into the numbers. Our back-of-the-envelope math is below, so feel free to check it. 

Consider what that 8% does. Firstly, it pays for the expensive novel branded medicines that are out there today treating cancer and autoimmune disease and depression and countless other conditions. The high prices of today’s branded drugs generate revenues and profits for the biopharmaceutical industry, which signal to investors and innovators that there could be significant reward for them if they bring new treatments and cures in the future. And as today’s branded drugs steadily go generic, typically about 14 years after coming to market, that 8% we spend shifts towards paying for newer drugs, generating profits for their inventors and their investors. The end result is the expansion of our generic armamentarium, which continues to work for all of us. It’s not ​“out with the old and in with the new.” It’s the new plus the old.

It’s like fueling a team of builders to erect buildings, one after the other, paying off the mortgage of each building over about 14 years. The builders keep building to keep earning our mortgage payments and what they leave in their wake is an ever-growing city. That’s our medicine city. There are now thousands of generic medicines that we’ll always have for as long as we need them. So many ailments are now tended to quickly and easily with the prescription of a generic drug. 90% of prescriptions are actually for generic medicines; typically low cost, typically low fuss. 

So in a sense, what we spend on branded drugs for a short period of time is really an investment towards owning them indefinitely for the rest of time. That’s the intent of the patent system; it’s how all innovation is meant to work. 

With many of the best drugs dropping in price by 95% or more when they go generic, it really is like paying off the mortgage on a house and living in it for merely the cost of maintaining the house. The reason the mortgage analogy is important is because we think of mortgages as investments, not expenses. On the other hand, hospitals are expenses; they don’t go generic. Every year, you pay more and more for the same thing with no end in sight.

Hospitals, doctors, nurses, clinics, nursing homes and other healthcare services are priced mostly as a function of land and the cost of human labor, which only go up. Healthcare services make up the vast majority of what we spend on healthcare: about 80%. And unless we invent new products for them to use (drugs, diagnostics, medical devices), healthcare services don’t really change and get better. Surgery stays the same.

So unlike our healthcare spending that goes to services and should be considered an expense, the mere 8% we spend on novel medicines doesn’t just pay for the same things year over year – it’s an investment in progress. It powers the conveyor belt that moves those drugs onto the massive and growing pile of inexpensive generic medicines that already make up about nine out of every 10 prescriptions Americans pick up at the pharmacy counter, forever upgrading our standards of care. And that 8% is not only constrained by the fact that almost all medicines will eventually go generic but it generally lowers spending elsewhere in the system, because those medicines can keep us all healthier, which means we remain productive and out of hospitals. Eight cents on the dollar does all that!

That 8% drives a lot of investment. 

Researchers led by Harvard’s Amitabh Chandra published an analysis in August that pegged global biopharmaceutical R&D investment in 2021 at $276 billion across 4,191 companies (not just the two dozen big pharmas that usually get all the attention). Chandra and others (including my colleague Tess Cameron) discussed that analysis, where all that funding comes from, and how to sustain it, here. Additional coverage of that discussion is here.

As you’ll see below, each year the US spends about $365B on branded drugs and yet each year the drug industry invests collectively about $276B into R&D. Sure, some of that funding comes from profits earned from selling medicines in other countries, but all that R&D is of interest and benefit to Americans, so we should really be focused on what we pay and what we get. What others pay is secondary (though it sure would be decent of other wealthy countries not freeride on the US quite as brazenly as they do). 

So consider that novel medicines cost Americans eight cents of every healthcare dollar and six of those cents go into R&D to improve healthcare and our lives. Pretty remarkable.

Crunching the Numbers

In case you’re doubting this math, we debated Ballmer’s numbers and can confirm that they make sense. Here’s how we squared our math with his.

Ballmer’s analysis appears based on CMS’s most recent national health expenditure data (which is from 2022) and doesn’t break out branded drugs from generics. So he reports a total of 11% of our health spending is on prescription drugs, but uses a smaller denominator than we do (he peels off $760 billion from the $4.5 trillion total spend that year because it didn’t go directly to ​“personal healthcare”) though is actually consistent with our estimate that branded drugs alone are 8% of overall healthcare costs. 

That’s because we’re not inclined to take out these real costs. Insurance administration is a real cost. Pandemic preparedness is a real cost. Medical capital expenditures are real costs. NIH is a real cost. Just as hospital stays are a real cost. Some of these are more ​“investments” – pandemic preparedness, capital expenditures, NIH – akin to our investments in medicines. But for the purposes of this analysis, let’s just call it all ​“spending,” whether investment or expense.

You can take a look at all the numbers for yourself. CMS’s national health expenditure (NHE) data (PDF exec summary here, visualization here) itself says 9% is spent on prescription drugs, using that larger $4.5 trillion denominator. But, like Ballmer, CMS doesn’t break out branded vs generic drugs. HHS says here that between 2016 and 2021, 80% of prescription drug spend is on branded drugs and 20% on generics (and that the ⁸⁰⁄₂₀ split has ​“little change over time”). The generics lobby pegs it at an 87%/13% split in 2022. So conservatively, taking the higher split provided by AAM, that’s still just a shade under 8%.

But we need to make another adjustment. CMS’s figures include only ​“retail” drug spend, which are all the prescription drugs people can get at a pharmacy but not those administered by doctors (e.g., most infused medicines). We want to make sure we’re including all branded prescription drug spend, and that will include drugs covered under the medical benefit (like drugs administered via IV in the hospital). 

For that we’d need to turn to a higher authority than even the NHE data: IQVIA. 

The IQVIA number for total drug spend net of manufacturer rebates and discounts in 2022 is $429 billion. That’s only a bit more than the NHE’s $406 billion estimate, which seems odd, but it’s possible that the NHE data underestimates rebates (IQVIA estimates 2022 retail net sales of $267 billion and non-retail of $161 billion). 

So do IQVIA’s figures still yield 8%? Yes. 

Continuing to use AAM’s 87% brand/​13% generics split, $365 billion of the $429 billion is on branded drugs. Dividing that by the $4.5 trillion total spend gets you around 8.2%. (IQVIA pegs brand spend at 85% of the total, which would lower it a bit more.)

Alright, so give or take a tenth of a percent, 8% of what we all spend on healthcare goes toward branded drugs.

Where does $4.5 trillion come from? Mostly from premiums and taxes. The NHE breakdown of spend by financial sponsor is $1.2 trillion household, $0.8 trillion private business, $0.3 trillion other private, $1.5 trillion federal, $0.7 trillion state/​local. Of the household spending, $0.75 trillion is premiums and $0.5 trillion is out of pocket.

We find all those breakdowns confusing. For example, private business spending is for premiums paid on behalf of workers, which is a form of compensation and therefore arguably is akin to household expenditure. Taxes come out of all of us, too. So maybe a simpler way to look at where all this money is coming from is to consider what we all pay on average over the course of our lives in premiums and taxes, directly and indirectly, (let’s call that ​“Premiums”) and separately ask what only the few of us who are patients seeking treatment pay out of pocket. 

Because it’s kind of that simple: there are the payments we all make regularly spread out over time (i.e., Premiums) and the payments that hit us suddenly when we are sick and trying to afford proper treatment (i.e., Out-of-Pocket costs).

That breaks the $4.5 trillion into a simpler $4 trillion in premiums and $0.5 trillion out of pocket, ​“merely” 11% of the total $4.5 trillion. We say ​“merely” because those out-of-pocket costs can be ruinous to some people because of how concentrated they can be on the weak and vulnerable among us.

Addressing affordability by rounding 8% up a little

While the $365 billion America pays for branded drugs could be readily affordable to all Americans if we paid for them out of our premiums (after all, that’s just 8% of all healthcare spending), the trouble is that we make just the patients among us pay about $28 billion out of pocket for branded drugs. 

In fact, to be precise, we should say that Americans pay $337 billion for branded drugs out of the total $4 trillion that they pay in Premiums. That’s about 8.4%. But that $28 billion out-of-pocket spending isn’t spread out across everyone like Premiums and is often what makes something unaffordable to a patient. (Note, our analysis is about drugs. We’re not even talking about the ruinous out-of-pocket costs for healthcare services, which make up over $400 billion per year.)

So what if we rolled those $28 billion into what we paid in Premiums? That would mean we would spend $365 billion out of $4.03 trillion in Premiums (adding $28 billion to the $4 trillion we spend in Premiums). That’s 9%. 

But the $28 billion is misleading because this is what Americans actually pay and therefore, by definition, CAN afford. What we can’t see is what Americans don’t pay because patients just give up on trying to get the medicine they need. What’s the total unpaid cost of under-utilization of medicines? What are the costs of treatments that Americans find unaffordable? That’s very hard to gauge. 

It’s been well documented that out-of-pocket costs deter healthcare utilization, including appropriate use of prescription drugs. Without out-of-pocket costs, it would stand to reason that a lot of that under-utilization would go away, and we’d need to pay a bit more for branded drug spend. 

Based on oft-cited stats that about ~20% of patients are under-insured, we might surmise that for every eight patients being properly treated, two are not because they can’t afford their medicines. If all patients could readily afford the treatments they need, then we would be paying to treat 10 patients, not eight. That suggests that total drug spending would go up by 25%. In truth, we know that companies provide assistance to patients to help them cover their out-of-pocket requirements, and so odds are there isn’t this much under-utilization, but let’s say there is and so utilization would go up by 25%.

Yet it’s still unlikely that the drug industry would capture the full amount of that market expansion. PBMs and insurers would use those higher volumes to negotiate deeper rebates in any competitive drug class (and most are competitive). Technically, any amount of revenue over the cost of producing those extra 25% of doses would be profit to the bottom line (and in many cases, those extra doses are already being produced and provided via patient assistance programs to under-insured patients, so it’s not that 20% of patients are going without treatment but no doubt some fraction of them are). So instead of $365 billion of drug spending surging to $456 billion, which would be 11% of a total of $4.12 trillion in premiums, we would estimate that $365 billion would go up by half that increase to about $410 billion, which is 10% of the $4.08 trillion we would be paying in premiums so that all patients could afford the branded medicines they needed. 

Overutilization and underutilization

But some might say ​“Wait! If patients don’t have to pay anything, there will be rampant over-utilization! We’ll pay far more in premiums!”

Let’s examine that at the level of specific drugs. Do we think anyone without HIV would end up taking HIV medicines? Is it so hard for insurance to confirm that someone is HIV+? Or in the case of using these drugs to prevent HIV infection, will people really be taking these drugs if they don’t consider themselves at risk? It’s not like they taste good or make you feel great. How about chemo? Will anyone fake cancer to score some chemo? Insulin injections? How about VEGF-inhibitors that need to be injected into your eyeballs. Do those seem like something anyone would want to joyride? 

My daughter suffers from atopic dermatitis. Despite liberal use of topical generics, she used to scratch herself bloody, something millions of Americans with this condition can relate to. Now she’s virtually free of her condition because of Dupixent, which requires an injection every two weeks. I deeply empathize with any parent whose child has atopic dermatitis but can’t afford the out-of-pocket cost their plan demands before covering Dupixent. If that plan uses copays to prevent over-utilization, I would reassure them that jamming two milliliters of fluid into a child’s arm is unpleasant for the child and the parent doing it and that anyone would sooner manage the disease with topical generics. We had to bribe her just to get her to take the first shot, after which she tolerated the biweekly ordeal because of how well it worked for her. She’s no drug seeker. So what’s with the copays?

GLP-1s? Yes, there’s a lot of demand for those drugs and many people want to use them off-label. But even there, insurance already relies on prior authorization to confirm that a drug is right for someone. So if a plan decides to only cover GLP-1s for people who are overweight and/​or have a comorbidity, they can have the physician confirm that before authorizing coverage. Height, weight, photo, whatever. It’s possible to objectively confirm if someone is overweight.

So we have the tools to prevent overutilization. It’s the under-utilization due to high out-of-pocket costs that we must solve.

By our estimates, we’re talking about collectively paying $75 billion more in premiums for branded prescription drugs so that the patients among us (and someday, each of us will be a patient) don’t have to face the burden of out-of-pockets and potentially forgo appropriate treatment. 

That’s a 1.9% increase in premiums, a 100% decrease in out-of-pocket costs for medicines, and a dramatic reduction of insurance injustice.

We should also point out that Americans spend a total of $64 billion per year on generic drugs (per IQVIA), of which $10 billion is baked into our premiums and $54 billion comes out of the pockets of patients. So truly we can say that we are paying a total of $335 billion + $10 billion = $345 billion for all drugs out of our premiums, which is around 9%. It would make excellent sense to cover generic drugs entirely out of premiums. They are the most cost-effective option for managing countless diseases. While most people can afford their generic drugs, all should. 

So if we rolled the out-of-pocket costs for generics into premiums and even assumed 25% higher utilization (though OOP costs for generics are much lower and under-utilization of generics due to unaffordability is likely much lower than 25%) and assumed that competition would erode half the extra spend, then we are talking about now paying a total of $80 billion in premiums for generic meds, bringing total drug spend to $490 billion ($410 billion for branded and $80 billion for generics), which is still around 11.9%. That’s only slightly higher than the 9% we pay out of premiums for all drugs and yet it would allow everyone to afford the medicines they need out of the premiums they pay over their lives.

But that’s not all. Because when people are able to take the medicines they need, they stay healthier and out of hospitals. The Congressional Budget Office estimated once that if Medicare reduced drug out-of-pocket costs, it would spend $10 billion more on drugs and save $20 billion on healthcare services.

So properly covering all medicines without out-of-pocket costs would likely increase our premiums but not nearly as much as one might think thanks to savings elsewhere. It might even eventually result in a net reduction in premium costs. 

Ironically, that would mean that more of each premium dollar was being spent on medicines. 

Let’s take this to an extreme. If increased spending on medicines coupled with a smaller overall total health spend raised the share of our spending on drugs to be 30% of all healthcare costs, would that be too much? I don’t think so. We need to look at our overall budgets. We are a wealthy nation. Yes, we spend about 17.5% of our GDP on healthcare, but if inventing better and better drugs were to cut our reliance on healthcare services in half, that’s $2 trillion of savings! That would mean we would be spending less on healthcare and have much more in our budget for other things.

Which is to say that when you are burdened with a high rent expense (healthcare services costs), then not paying a finite mortgage for a home to get out from under that rent expense is a terrible way to save money. What we spend on novel branded medicines is a finite mortgage precisely because they go generic, whereas hospitals and doctors and nurses do not; the cost of services rises forever unless we prevent the need for them.

A healthy hundred years

I shared the argument about drugs saving us money on services with Harvard’s Amitabh Chandra, and he made a really important point that saving money on healthcare services should not be the measure of whether we are getting good value from biomedical innovation. What matters is the actual value of those medicines to us. It’s worth paying more for a healthier life. We value living longer and being free of pain and disability.

Amitabh is entirely right. We shouldn’t be making drugs justify their value purely based on how much money they will save us. Consider if the average American could count on a healthy productive 100 years of life. What’s that worth to us? Might we elect to spend $5 trillion per year on healthcare (instead of $4.5 trillion) if that’s what we’ll someday get? Might we continue to delay full retirement, especially if we enjoy our jobs, and therefore live richer lives? Maybe knowing that one will be healthy into one’s 90s might motivate more people to switch careers when they are 50 or 60, finding more happiness in their work without compromising their income. 

Since I can’t entirely quit justifying the value of medicines with numbers, let’s look at productivity more quantitatively. If we’re living a healthy hundred years, that’s a lot more productive years. We already see the transformative power of medicines in the case of HIV, where roughly $13B per year of spending on HIV medicines keeps over a million Americans living healthy and productive lives despite being HIV+, which using US average data adds up to over $60 billion of wages per year. COVID vaccines and treatments cost what seemed like a lot, adding up to over $100 billion of spending during the pandemic, but it restored trillions of dollars of economic activity once everyone stopped distancing. Vertex’s drug Trikafta has restored most patients with cystic fibrosis to nearly full health, able to live productive lives and start families. These medicines not only liberate patients to live productive lives but also liberate their caregivers to contribute in some other way to society. 

GLP‑1 drugs are starting to reverse the ravages of obesity. Were everyone to have access, consider all the people who will be liberated from their joint pain and other comorbidities, able to breathe easier, enjoy life more, and work more productively. Consider as well the reduced rates of diabetes, heart failure, cancer, strokes, and many other diseases that stem from obesity. Researchers at USC Schaeffer Center for Health Policy and Economics estimated that coverage of these drugs in Medicare alone would ​“generate $1 trillion in social benefits over the next ten years.” Expanding that coverage to the commercial market would eventually result in $500 billion per year of societal benefits over 20 years (more than $10 trillion over 20 years and more than $17 trillion over 30 years) on top of all healthcare savings. 

As for what such a transformation would cost, consider that there are now scores of weight-loss drugs in development, including easy-to-make and easy-to-take oral pills that, as they come to market, will compete aggressively on price, as we’ve seen in other drug classes. At even $500/​year for every overweight American (~$40/​month), the drug industry would stand to make $100B in revenues, a prize so large that competition, especially from the inevitable generics that come down the line, would continue to drive prices even lower. And yet that $100B/​year of spend would be small compared to the value. So we shouldn’t focus on the cost of the drugs in isolation of the resulting savings and, as Amitabh Chandra emphasizes, their benefit to society. 

Medicines are equitable; services and hospitals are not

Our ability as a society to afford the quality of life that we enjoy today is a function of our total productivity. The more productive we are, the wealthier we are. How that wealth is distributed is of great concern to many Americans, so let’s consider this question of equity.

Relying on hospitals and doctors for care is the least equitable form of healthcare. People with low incomes often live far from good care, far from a hospital, and have low-paying jobs that they can’t readily miss for appointments to see a doctor. Consider that millions of Americans couldn’t afford to seek treatment from a hospital even if it were free. 

A medicine, especially a simple pill they can take themselves and could even be delivered to them by mail, is the easiest way to manage their disease. A self-administered injection isn’t far behind. If covered by insurance, they can afford this treatment and take it and stay healthy and not miss work. People with low incomes can’t afford to pay for caregivers when a loved one gets sick, and yet they also can’t afford to be a caregiver if they are trying to hold down a job. Two income households are therefore extremely vulnerable to catastrophic decline (e.g., medical bankruptcy) when one earner falls ill and requires the other to provide care. They lose one income and the second person might need to choose between their partner and their job. They may fall behind in mortgage payments and lose their homes. Having access to a medicine that prevents, manages, or cures the sick partner therefore saves the whole family. 

The rich can afford to live without better medicines, can afford to go to hospitals and see doctors as needed, and will merely lose out on joy and health. America’s poor literally cannot afford for us to stop inventing new medicines. Better medicines and proper insurance are the best way to help millions of working-class Americans remain solvent when they get sick. 

Protecting the big picture

So when you next read something that suggests that America is overpaying for branded drugs and that we absolutely must have price controls to make medicines more affordable, I hope that you’ll consider the 8% that we currently spend out of premiums for branded drugs. Is that really too much? Might not a better solution be to pay just a bit more in premiums so as to reduce out-of-pocket costs?

Let’s look beyond the conventional wisdom that people need financial skin-in-the-game to prevent them from joyriding all kinds of medicines they don’t need. To those who warn that we can’t afford to keep people alive longer because Social Security and Medicare will run out of money, let’s remind them that, as long as we’re not just alive but healthy and productive, retiring in our 60s is a choice, not a rule.

Think about the big picture. Insurance is what makes innovation both profitable enough for investors to fund and yet affordable to patients. There would be almost no biomedical innovation without insurance (at best, we might have things like wrinkle fillers and botox, since those are cash-pay consumer markets). Insurance works best when we all pay into it, but that doesn’t mean everyone has to pay the same, just as taxes are not the same. The very poor pay almost nothing for insurance (Medicaid), which is funded by all taxpayers. Those with slightly higher but still lower income (up to about $100,000 of household income for a family of four) get taxpayer-funded subsidies that shrink as their incomes climb. Above $100,000, people don’t get direct subsidies, but most get insurance through their employer as untaxed compensation. And those over 65 get taxpayer-funded Medicare supplemented with insurance premiums, which people pay but which can also be further subsidized by taxpayers. 

Making sure everyone has insurance is a function of getting those premiums and tax-funded subsidies right. To claim that a country as wealthy as America can’t afford healthcare is false. And to claim that it can’t afford the tiny fraction that branded drugs (8%) or all drugs (9%) represent is even more false. We can and we must afford them, especially in the case of novel medicines. Because what we spend on novel medicines is what drives the conveyor belt of innovation that allows us all to justifiably expect that our lives will only get healthier, more joyful, more productive, potentially longer, and less burdened by hospitalizations and caregiving.

So we’re paying 8% now. And paying a couple percent more, if even that much, so that patients (eventually all of us) aren’t burdened by out-of-pocket costs is an affordable and rational way to improve America’s experience of medicines.

Short epilogue: But what about!?

I know that some people will say ​“but what about…” and follow that up with something about patent monopolies (misunderstood since most drug classes are competitive), other countries paying less (damned freeriders should pay more), old drugs not going generic (solved with price controls that kick in after a patent period of market-based pricing, such as IRA’s ​“negotiation” provision 13 years after biologics launch), executive compensation (miniscule in the grand scheme and a function of market forces), direct to consumer advertising (misunderstood, FDA-regulated, and essential to innovation, even if a bit tedious), sales and marketing expenses being higher than R&D investment (actually, R&D in aggregate is higher), and other concerns. For more on 1) why all those issues are either not the problems they are made out to be or 2) solvable with careful policy that preserves incentives for innovation and, in any case, 3) are not a reason to ignore what I’m saying here, please visit www​.nopatientleft​be​hind​.org. 

I’ve also summarized answers to some common questions and concerns in this LONG ​“But What About!?” EPILOGUE.