Has U.S. Health Care Spending Finally Stabilized? An Outlook for 2019


The official 2017 statistics from the U.S. Department of Health and Human Services (DHHS) are out, and there are some good news: The annual growth rate of health care spending is slowing down, and is the lowest since 2013 at 3.9%—it was 4.3% for 2016 and 5.8% for 2015. The bad news is that our health care cost increases are still well above inflation, and that we spent $3.5 trillion in this area, or 17.9% of GDP. Americans spent $10,739 on health care in 2017, more than twice as much as of our direct economic competitors: This per capita health care spending was $4,700 in Japan; $5,700 in Germany; $4,900 in France; $4,200 in the U.K.; $4,800 in Canada; and an average of $5,300 for a dozen such wealthy countries, according to the Peterson -Kaiser health system tracker from the Kaiser Family Foundation, and OECD data. Spending almost a fifth of our GDP on health care, compared to 9-11% for other large developed economies (and much less in China), is like having a chain tied to our ankles when it comes to our economic competitiveness.

Could 2019 be the year when our health care spending actually decreases, or at least grows at a slower pace than inflation? Or will we see instead an uptick in costs for health care consumers?

To answer these questions, we need to look in more detail at the largest areas of health care spending in America, and at the recent but also longer term spending trends in these areas. Using the annual statistics from the DHHS, we can compare the growth in spending in half a dozen critical health care categories with the growth in total spending, and this for the last three years as well as the last decade. Over the last decade, since 2007, these costs grew 52% in aggregate (from $2.3T to $3.5T) and 41% per capita (from $7,630 to $10,740).

1)  Private Health Insurance costs ($1.2T in 2017) grew 4.2%, 5%, and 6.9% in 2015, 2016 and 2017 respectively. This is faster than total health care costs, in part due to the strong average cost increases that followed the implementation of the ACA (“Obamacare”) marketplaces and subsidies. On the other hand, over the last decade, private health insurance costs only grew 40%, less than total health care spending. So we should be able to anticipate some stability here, but no significant change in the business models used and reductions of patients’ out of pocket cost, co-pays, deductibles, etc.

2)  Medicare costs ($706B in 2017) grew 5.1%, 3.5% and 3.5% in 2015, 2016 and 2017 respectively. Over the last decade, they grew 37%. Good news, Medicare cost increases lag total health cost increases. Bad news, the 5.1% cost increase in 2017 is high. What will we have in the next 5-10 years? Unfortunately, Medicare costs will likely increase faster than total health costs, due to the aging of the U.S. population, and this trend will continue at least until the last “Baby Boomer” reaches the Medicare eligibility age of 65.

3)  Medicaid costs ($582B in 2017) only grew 2.9% in 2017 and 4.2% in 2016. On the other hand, with the ACA’s Medicaid expansion, they grew 9% in 2015 and 11.8% in 2014, as several new states expanded Medicaid under the Obamacare guidelines. This will be repeated in 2019, following the November 2018 Midterm elections: Three newly elected Democratic governors promised to bring Medicaid expansion to their constituents. This has already been done in Maine, with Kansas and Wisconsin next. Plus, the citizens of Idaho, Nebraska and Utah will also get Medicaid expansion, following the success of ballot initiatives during the recent Midterms. In total, around 800,000 people are poised to gain access to Medicaid for the first time in these six states. This is very good news for national health coverage statistics, and of course for the new enrollees, but it also means that Medicaid costs are likely to increase at a strong rate in 2019 and 2020. The impact on overall health care spending will be mitigated, though, by fewer visits to expensive emergency departments from people without health insurance. From a patient’s standpoint, this will be the strongest positive impact of the ACA in 2019.

4)  Hospital spending ($1.1T in 2017) and Physician and Clinical Services costs ($694B in 2017) have grown in line with the overall health care costs, in recent years as well as over the last decade. With increased merger activity among health care delivery providers, their pricing power increases with industry concentration, and it is unlikely that we will see spending in these areas grow much more slowly than total health costs.

5)  Administrative costs, like a snake in the grass, represent the biggest outlier in U.S. health care costs. What good can come out of excess administrative costs? Employment! Yes, but this is not the type of employment that adds to our economic competitiveness…The numbers in this area are tragic: $259B in 2017, with well above average cost increases of 6.5%, 5.6% and 5% in 2017, 2016 and 2015 respectively—and a whopping 65% over the last decade. We spent more on administrative costs than the United Kingdom spends on total health care. Let’s do some quick math here: The dozen large developed economies mentioned above spend an average of 2% of total health care costs on administration, versus 7.4% in the U.S. All other things being equal, if we too spent 2% on admin. costs, we would save $190 billion per year, enough to insure pretty much all our uninsured at today’s costs. Unfortunately, this sad state of affairs is the reflection of the enormous complexity of our health care system, and as long as we cannot surmise the political will to do a complete overhaul of how we provide health care to our citizens, this will continue. A potential sign of improvement in this area will be when technology innovations lead for the first time to a demonstrated ability to cut health care costs. Thus far, and unlike in any other industry, the opposite has proven true, and administrative costs have grown hand in hand with ever more complex electronic health records, patient accounting systems, the use of big data and other IT innovations. Complexity reigns supreme in U.S. health care, and even Silicon Valley is unable to do anything about it.

6)  Drug Costs: I have saved this most often mentioned culprit in health care costs for last. And yes, drug spending over the last decade has outpaced total health care spending in America, principally in the 90s and early 2000s, where it grew in double digits annually. Americans spend about twice as much in this area than residents of other rich developed countries. Surprisingly, however, spending on prescription drugs in pharmacies and other retail outlets only grew at 0.4% in 2017, at $333B, and 1.2% in 2016. After increases of 9% in 2015 and 15% (!) in 2014, could galloping pharmaceutical prices be behind us? To have more certainty in this area, one would have to be able to track the aggregate costs of drugs used in hospitals (not reported in the DHHS statistics), often very expensive ones, in part because these costs are well reimbursed by both private and public payors. Because it has become such a political rallying cry, we could very well see some bipartisan legislation, not just to contain but reduce prescription drug costs in our country. We all know what should be done: Eliminate the absurd clause within Medicare Part D that does not allow Medicare, the largest purchaser of drugs in the country, to negotiate prices directly with pharmaceutical companies; and allow the import of lower-priced drugs from reputable countries such as Canada, Germany and Japan. Any partial movement in this direction, such as the recent proposal by the Trump administration to allow the import of lower cost pharmaceuticals in certain areas for Medicare, would likely lead to a reduction in prescription drug costs for U.S. patients.

In summary, U.S. health care costs will continue to grow at a rate above inflation in 2019. The aging of the population, Boomers reaching the Universal Care promised land (called Medicare here), will ensure continuing growth of Medicare costs; Medicaid costs will increase too, with six new states adopting its expansion; hospital costs growth may slow because more insured patients means fewer uninsured write-offs; administration costs will remain intractable; the one potential good surprise for U.S. patients (and those of us who like political bipartisanship) could come from the area of prescription drugs, with corresponding reductions in out of pocket costs—fingers crossed. But since even a 20% reduction in drug costs—an amazing achievement—would only lead to a 2% reduction in total health care costs, it is hard to project an outlook that does not see at least a 2-3% increase in U.S. health care costs in 2019.

Author of “Untangling the USA: the Cost of Complexity, and What Can Be Done About It,” Etienne Deffarges has counseled, created, and invested in countless organizations during his professional life as a management consultant, business executive, and entrepreneur.

10 replies »

  1. “Competition has always been the only thing that can actually have a meaningful impact on prices”

    How can drug companies compete when their stock holders will punish them for seeking lower prices? Drugs are patented, where is the competition for that? This whole idea that health care can be just like a normal marketplace is a fallacy. It will only be like a marketplace when we can choose to not buy the things that keep us healthy and alive.

  2. The unrelenting decline of Social Capital has an inherently adverse effect on almost all forms of Unstable HEALTH. And, now we have seemingly random micro-mass shootings, “Bowling Alone” personified.

  3. Agreed, especially on pricing and competition and markets. Unfortunately, as Porter & Teisberg discuss in their book “Redefining Healthcare, Creating value-based competition on results” in the present day US healthcare system (‘HEALTHcare’ and ‘system’ being called an oxymorons by some as it relates to present day US) the competition based on price is not taking place over the right things at the right level within the healthcare delivery continuum.

  4. I don’t think we can assume stabilization given a very small data set. The accelerated costs are merely resting. Still concerning that costs are rising relative to GDP, especially with so many trying to reverse that trend.

  5. But, of course, Parkinson’s Law continues to apply: as in “Work expands to use the available resources.” It is likely that Parkinson’s Law will continue as long as we lack a community by community means to assure that equitably available and ecologically accessible, enhanced Primary Healthcare is offered to each citizen. In addition, this goal should be integrated with the collective thrust projects chosen to assure social mobility for each citizen, a generational effort. Finally, a Master Disaster Planning Strategy should be annually updated to integrate the community’s needs for the mitigation strategies applicable for their own risk management and it connections with regional, state and federal systems.

    We continue to ignore our nation’s severely declining level of Social Capital as expressed by its deficient social cohesion. We continue to burden our nation’s healthcare industry with Unstable HEALTH borne out of the daily burden of survival.

    Papanicolas et al JAMA Special Communication 2018 is still the best analysis of HEALTH SPENDING. https://doi:10.1001/jama.2018.1150 See page 1026 Figure 1 Spending, and its “Inpatient Care” line item. As you will note, cutting Inpatient Health Spending has no future as a means to decrease overall HEALTH SPENDING. It only represents 19% of the annual outlay. Only Canada is lower at 17%. The other 9 OECD nations are between 32% for the Netherlands and 21% for Sweden.

    As noted in the citation, the data of the Special Communication uses the OECD as its source. They have internal strategies to be sure that the line items reflect standardized usage among the 35 associated nation’s. As a reminder of our nation’s declining Social Capital, think again about worsening maternal mortality, childhood obesity, adolescent suicide/homicide, narcotic overdose mortality, homelessness, mass shootings, mid-life depression and declining longevity (now 4 years in a row).

  6. I am not sure I see things slowing down significantly unless we hit another recession. But, following that recession, the growth rate will continue. Some things contributing to these hikes:
    (1) Drugs: Precision medicine is still in its infancy, but we are bound to see numerous new drugs enter the market in the coming years. These products are priced, justifiably, high given the small markets they target. I would image pharma costs to continue to rise rapidly.
    (2) Provider prices: Consolidation has been all the rage in healthcare over the last decade. This has given some firms outsized market power, which they are using to further increase prices, particularly in hospitals. Barring any new regulation or the scaling of a disruptive entrant, we currently have no major weapons to fight this trend. https://www.techprescribed.com/all-posts/hospitals-kill-01142019
    (3) Aging: You reference it, but I think it bears stating once again. The population is going to shift significantly towards elderly in the coming years. Total healthcare costs will increase, but we also should look closely at per capita costs as well here.

    With that said, there are some bright lights of hope. The value-based programs launched by CMMI several years ago are starting to take hold (e.g., ACOs), and I expect they will help address these costs to a degree. In addition, Medicare Advantage continues to grow in terms of proportion of population. While it isn’t necessarily cheaper today, I foresee this program putting downward pressure on costs into the future. https://www.techprescribed.com/all-posts/total-cost-of-medicare-12262018

    In the end, to quote the late Uwe Reinhardt, it’s the prices stupid! Competition has always been the only thing that can actually have a meaningful impact on prices, and we need to do more to encourage it in our industry.

  7. I agree.

    First, with respect to expanded Medicaid, we need to realize that many if not most Medicaid beneficiaries don’t have adequate transportation and / or can’t get time off from work or afford to take it if it means losing pay. So, they are not as willing as higher income people to first visit a primary care doctor then be referred to a specialist if necessary and then maybe have to go somewhere else for imaging or labs. They see the ER as a one-stop shop and they perceive the quality of care they get in the hospital as better so the ER is their destination of choice more often than for the rest of the population.

    Regarding the VA formulary, it is indeed highly restrictive. Some years ago, there were about 3,800 drugs on the VA formulary vs. 10,000-11,000 on a typical private sector formulary. When Lipitor was still patent protected, it wasn’t on the VA formulary nor was the blood thinner, Plavix. In the addition, the VA health system serves only 8-9 million veterans out of a total veteran population today of 20-22 million. Also, veterans are seen by the drug industry as a very sympathetic group because of their service to the country and the fact that many of them got hurt defending the rest of us. As a result, the pharmaceutical industry is more willing to give the VA system a significant price discount beyond what it would be willing to give to other payers. Finally, drug prices for the rest of us are negotiated by three large payer groups that together account for 90% of the market. Each of those groups negotiates on behalf of close to 100 million people with insurance. Why should we think Medicare could do a better job without resorting to dictated prices which could easily be low enough to have a significant adverse impact on future drug innovation? Moreover, most private sector patients, especially seniors who consume twice as many drugs per capita as the non-senior population would never accept a drug formulary anywhere near as restrictive as the VA’s.

    The main drug issue payers need to fix is subjecting people with high deductible insurance plans to full drug list prices within their deductible. Historically, payers used rebates to reduce insurance premiums from what they would have otherwise been which means, in effect, that the sick are subsidizing the healthy. That’s the opposite of the way health insurance is supposed to work and, as I understand it, the insurance industry is in the process of fixing it so insured members pay the net drug price after rebates within their deductible starting this year. For self-funded large employers, the employer will control how that issue is handled, not the insurer or PBM.

  8. A few points.

    First, your comment regarding expanding the roles of Medicaid illegible that “The impact on overall health care spending will be mitigated, though, by fewer visits to expensive emergency departments from people without health insurance.” doesn’t hold water according to at least two well done published studies. Nikpay et al in “Effect of the Affordable Care Act Medicaid Expansion on Emergency Department Visits: Evidence From State-Level Emergency Department Databases.” Ann Emerg Med. 2017 Aug;70(2):215-225.e6. doi: 10.1016/j.annemergmed.2017.03.023. Epub 2017 Jun 19 and Finkelstein et al in Effect of Medicaid Coverage on ED Use — Further Evidence from Oregon’s Experiment (NEJM) both found increased Medicaid coverage correlated with increased visits to the emergency department – not decreased visits as you state.

    Second regarding drug costs, drug re-importation is not the panacea some would have you believe. Re-importation brings up the issue of counterfeit supply and pedigree and the increased risk of counterfeit drugs entering the US market to say nothing of other countries not wanting to act as medicine cabinet to the US (Canadians have been fairly vocal on this point) and the likely response of the drug manufacturers and distributors to re-importation.

    Third, as far as the ability of Medicare to directly negotiate drug prices – prices are already negotiated in the current market based system. To those who point to disparity between prices the VA pays vs. what the government pays in Medicare as evidence the government could get a better deal – a better deal for what? Price is but part of the equation – the other is access. The VA formulary and any other that focuses on price (like Medicaid) have as part of their design limited access to branded products – so comparing current Medicare to current VA prices without looking at what you’re buying in terms of a PDL is a rather specious argument. Direct negotiations will take away one of the more highly rated aspects of MPD as it stands now – consumer choice based on market incentives in a market based system.

  9. I’ll several thoughts on this.

    First, the standard Medicare Part B premium, which is set to cover 25% of Part B costs, increased from $115.40 per month in 2011 to $135.50 per month in 2019. That’s an increase of only 17.4% over eight years or 2.0% per year compounded. Part B accounts for 44% of total Medicare spending including about $23 billion of specialty drugs paid for under Part B instead of Part D. It’s also worth noting that the $706 billion figure cited for 2017 Medicare spending is a gross number whereas the Congressional Budget Office (CBO) reports Medicare spending on a net of offsetting receipts basis which reflects Part B premiums and state payments on behalf of Dual-Eligible beneficiaries. That brings the net figure to below $600 billion which is the cost to taxpayers.

    Second, Medicare’s reported low administrative costs are misleading. First, when people age into Medicare at age 65, they apply for it through the Social Security Administration, not CMS. So those costs don’t count. Medicare’s employees are housed in buildings owned or leased by the General Services Administration. Those costs don’t count either. The money needed by CMS to pay its employees is raised and collected by the Treasury Department. Those costs don’t count either. If properly accounted for, Medicare’s true administrative costs are probably closer to 6%-8% on a base of per capita healthcare claims that are roughly twice as much as the per capita claims of the under 65 population. For reference, large self-funded employers have administrative costs in the high single digits on claims costs of about $6,000 per covered life.

    Third, I’ve never seen any definitive evidence that the accounting conventions used to calculate total medical costs in the U.S. and the OECD countries are consistent from one country to the next. We already know there are significant differences in the definition of a live birth used to calculate infant mortality statistics. Maybe there are significant accounting differences here too. I don’t know.

    Finally, Medicare imposes significant documentation requirements on providers which many of them complain loudly about. They absorb the costs in their practice expenses and don’t get reflected anywhere in Medicare’s administrative costs. Also, there is generally believed to be a much higher incidence of fraud in both Medicare and Medicaid than in the private health insurance sector. Since fraud is impossible to quantify with any precision, it just gets ignored in healthcare discussions like this.

    In short, any administrative costs savings or other potential savings from a single payer Medicare for All system is illusory at best and don’t even get me started on its potential to have a significant adverse effect on medical innovation. Moreover, doctors, nurses and other healthcare system employees earn significantly more money than their counterparts in other countries so it’s likely that our total healthcare costs will always be higher than elsewhere as a percentage of GDP.

    If we could ever reform our medical tort system to get malpractice cases out of the hands of juries and substitute specialized health courts with judges empowered to hire neutral experts to evaluate conflicting scientific claims, we could reduce defensive medicine over time. If more people execute a living will or advance directive for healthcare and we evolve a more sensible cultural approach to end of life care, we could save lot of money there. If we invest more in data analytics we could make a sizeable dent in healthcare fraud. If we could outlaw confidentiality agreements between payers and providers which make it impossible for doctors and patients to identify the most cost-effective high quality providers in real time, we could steer more care away from the most expensive hospitals whose prices are currently high because of their market power and not the quality of their care and we need special rules that limit how much can be charged for care that must be delivered under emergency conditions. The changes needed to drive down the growth of healthcare costs in the U.S. are both cultural (torts and end of life care) and contractual (confidentiality agreements between payers and providers.)

  10. Good analysis. Thanks. Efforts to lower costs in health care are almost pointless in a system that 1. cannot have prices that mean anything—because there is no real shopping—and 2. cannot have oligopsonic or monopsonic purchasing which politicians are deathly afraid of.

    And these are the only two ways to lower prices. And we have to pick one as they are mutually exclusive.

    So we have to wait until the system grinds to a halt before we can fix it…which it will because of low total demand. The marginal benefits will be less than the marginal costs for many of its transactions. “I can’t afford sixty grand to spend a night in the hospital.”

    Talking about costs in health care is thus like talking about parachute function in a vacuum until we can have at least one of two aforementioned tools

    Incidentally, expanding Medicaid seems to be associated with more ER visits, rather than less.