The slowdown in increases often happens during and shortly after recessions, but this time the trend may continue as the American economy perks up
Frank Diamond
Managing Editor

About a decade ago, economists noticed something strange about the rate at which health care spending increased: the relentless climb started to level off. Then came the economic slowdown of 2007—some call it the Great Recession—and the slower rates of increase continued because that’s what typically happens in recessions. But then another strange thing: The flattening of the trend line persisted. Will this last? And might there come a point—unimaginable only a short while ago—when American health care spending would actually decrease?

In December, CMS actuaries reported that total spending on health care increased by just 3.6% in 2013, the smallest annual increase in national health care spending on record, which goes back to 1960. The growth in aggregate private health insurance premiums was 2.8% in 2013, down from 4.0% in 2012. Medicare spending was also part of the notable downshifting. It increased by 3.4% in 2013, down from 4.0% the year before.

How slow can it grow?

Total national health expenditures (NHE), annual percent change from the previous year, 2000–2013

Sources: CMS, National Bureau of Economic Research

For years, economists, some lawmakers, and a small slice of the public have fretted over health care gobbling up a larger and larger chunk of the country’s economy. But CMS actuaries reported that from 2009 through 2013, health care spending’s share of GDP has remained steady at 17.4%.

In an article published last month in Health Affairs that explains the spending numbers, the actuaries said the slowing down of health care spending can be viewed either as a temporary aftermath of the Great Recession or “the beginning of a new era.”

Ah, a new era. The CMS actuaries ( note that the crucial question is whether health spending growth will accelerate once economic conditions improve significantly —and that historical evidence suggests that it will.

Forget the historical evidence, says David Cutler, PhD, a health economist at Harvard: “We’re not seeing anything that looks like a ticking up of cost increases even as the economy has done better.” That’s not definitive, he adds; talking about the future of health care spending is like talking about the weather—there’s an element of chaos. Still, Cutler has more than a hunch that things aren’t going to reverse themselves as soon as the economy peps up. The Great Recession officially ended in June 2009, and last year real GDP increased by 2.4%.

Medicaid’s spurt the only deviation from a leveling off

Annual percent change from the previous year, 2007–2013

Private health insurance

Spending on private health insurance premiums has decelerated faster than NHE because of lower overall enrollment, high deductibles, and provisions of the ACA such as the medical-loss ratio requirement.

Out of pocket

Out-of-pocket spending has increased in recent years because of the overall improvement in the economy and high-deductible health plans (CDHPs). Employees have more financial responsibility in all plans, not just CDHPs.


The growth in Medicare spending tracked with the NHE in 2012 and 2013. Overall Medicare spending was influenced by slower growth in enrollment, provisions of the ACA, and the 2013 federal budget sequester.


Annual expenditures rose by 6.1% in 2013, partly because of Medicaid growth and increases in spending per enrollee. Spending growth will continue if Medicaid is expanded in more states.

Source: CMS

What caused it?

Economists are debating the reasons behind the deceleration of health care spending. Many say that at least 70% of the slowdown was caused by the recession, according to Uwe Reinhardt, a Princeton health economist. “I belong to that group. The reason we say that is because spending also started to fall in Europe. We said there has to be a common thread here, and it is the recession.”

Don’t underestimate the political climate, says economist Uwe Reinhardt. Republicans “have never hesitated to beat up on the health care industry” to reduce spending.

Other economists say that it was because of structural changes, high-deductible policies especially. Reinhardt’s side says that those structural changes did not occur in Europe, and yet it too saw spending decline.

“So we have a big fight about it and the honest answer is that economists really can’t quite agree,” says Reinhardt. “They say it was probably all of these factors and that’s the safest thing. We always end up with that. We have these heated debates among ourselves, and then we wind up always saying that it was probably all of the factors.”

“We’re not seeing anything that looks like a ticking up of cost increases” even as the economy recovers, says economist David Cutler. More Sovaldis could change that.

Cutler views Reinhardt’s 70/30 demarcation as adjustable: Where you draw the line depends on when you think the slowdown in health care spending really started to occur. Do you start from 2007 and the country’s financial unwinding, or do you go back to 2002, which is when many economists think the spending curve started to level. Cutler believes it started before the recession but became more noticeable during and after. The leveling off has gotten more robust and the economy has recovered somewhat, and yet we still don’t see a return to spending increases, he notes.

Beth Umland, the director of research for Mercer Health & Benefits, says pace of employment-based health care spending started to ease up about 10 years ago. From 2004 to 2007, her company’s survey found increases stabilizing at 6.1%.

The “magic” began about 10 years ago when employers held health care spending increases to about 6% for three years, says Beth Umland of the consulting firm Mercer.

“We survey about 3,000 employers every year so to have it come out that consistent every year, it was starting to feel like something magical was going on,” she says.

The magic, as Reinhardt notes, was, in part, simple pushback. Employers decided that 6% annual increases were about all they were willing to tolerate, so they figured out how to hold costs down. Umland says employers have more tools now to control health care costs than they did in the past. One of those is the consumer-directed health plan. It’s no coincidence, says Umland, that spending is slacking off during a period of rapid growth of coverage that features higher deductibles. The fact that increases in health care spending continued to taper off in 2011 and 2012 after the country started to emerge from the economic doldrums is evidence, in Umland’s view, that “health care consumerism has finally taken root.”

Cutler, Reinhardt, and Umland, track from on high, but things look different in the trenches. Alan Muney, MD, Cigna’s chief medical officer, warns that if we are indeed entering a new era of health care spending it’s just lower trend on top of baseline spending that’s still too high from years of double-digit increases.

“We’re all happy to see that moderation, but we have a long way to go to actually transform the health care system to be something we all want it to be, which is affordable with better outcomes,” says Muney. He points to antibiotic prescriptions as an example. Despite guidelines and other efforts to curb them, “We haven’t gotten the results we want.”

“It’s just lower trend” on top of baseline spending that’s too high, says Alan Muney, Cigna’s CMO, of the recent slowdown. Health plans still have a lot of work to do.

Comparing antibiotic prescribing with medical imaging may seem to be a stretch, except that it illustrates just what tighter management can do, according to Muney. In the case of imaging, insurers now measure requests against guidelines. “That kind of management has brought down radiology costs,” says Muney. “It also has doctors focusing more on it.”

In addition, where MRIs and other imaging tests are done is important. An MRI that might cost $600 at a freestanding radiology unit could cost four times as much at a hospital, Muney says. “It’s not just managing according to guidelines. It’s also the site of care.”

High baseline

The regional differences also matter. For more than 20 years, Dartmouth researchers have been documenting wide regional variation in practice patterns. The headline is that they’ve found no correlation between how much money is spent on health care and the outcomes of that care. “If care is either too much too often or too little too late because clinical guidelines aren’t being followed, you’re going to get a high degree of variability clinically and therefore a wide variation in cost,” says Muney.

Plenty of fat

Muney believes that between 25% and 30% of all health care spending is inappropriate and wasted. “We spend $3 trillion on health care, and $1 trillion is either over- or underutilization. If there’s an inability to improve variability and outcomes, that’s a lot of baseline.”

In other words, an increase of a meager 3.6% on top of wasteful spending is still excess cost.

Umland understands Muney’s hesitancy to celebrate, but she sees real change. “We can quibble about whether we want to call it a new era. Are things different now then they were 10 years ago? It looks like it.” Nearly half (48%) of employers now offer consumer-directed plans and almost a quarter (23%) of covered employees are enrolled in one, she notes.

Even with PPOs, employers have more arrows in the quiver, such as penalties for smokers and surcharges for employees whose spouses have other means of insurance. “Yes there’s cost shifting involved,” says Umland, “but we’re also seeing employers doing more sophisticated things.” She mentions high-quality networks and getting employees to use transparency tools that make it possible to compare provider price and the quality of health care services.

Any conversation about health care spending these days invariably includes specialty drugs, with Sovaldi as Exhibit 1. Reinhardt, however, predicts that the “Sovaldi moment” will be the last hurrah of high health care spending. Payers, so far, don’t really know how to manage these breakthrough medications, but they’ll figure it out, he says. A $1,000 pill is a big attention getter because it seems so outrageous, says Reinhardt, although he notes there are actually many drugs costing a lot more both by dose and total spending and that even at $1,000 a pill, Sovaldi appears cost-effective relative to what we were willing to spend on hepatitis C already.

Besides, drug spending overall is not a big-ticket item relative to other health care spending, Reinhardt notes. Total spending on prescription drugs is still slightly below 10% of total national health expenditures. “There is not much mileage in beating up biotech and pharmacy if the goal is to control total health spending,” says Reinhardt. “My overall take is that it’s a headline-grabbing thing.”

Cutler, at Harvard, is not so sure: “If there are a lot more Sovaldis, that will materially impact health care spending,” he says. In Cutler’s opinion, the reason health care spending has leveled off is that the number of new technologies, including medications and devices, introduced into the health care system has been has been relatively low.

Muney does not take the Sovaldi moment lightly. If everyone who could benefit from Solvaldi got it, the cost would be about $300 billion a year, he says. Today, we spend about $270 billion on all prescription drugs combined.

However, Muney is also an optimist about many specialty drugs eventually helping to reduce health care spending, even if they push up spending initially. With good adherence, the medications stand to lower the spending on hospital care, physician services, and other services because chronic diseases will be better managed.

Sovaldi vs. WalMart

Others are optimistic about technology and innovation helping cut costs. After all, in other parts of the economy, inventiveness leads to greater efficiency and lower costs, not greater expense.

Technology fans believe, for example, that 3-D printing could reduce the cost of research and development for pharmaceutical companies by billions of dollars. New techniques could make for cheaper, more convenient blood tests. And, as Umland points out, expensive as specialty drugs are, they could reduce health care spending in the long run if people are cured or spared years of treatment with medications that are less expensive but not as effective. For instance, there’s a full-on assault against AIDS. The drug vorinostat is seen as a possible weapon in the “shock and kill” approach, which seeks to unmask dormant HIV-carrying cells and kill them.

“I think the future of spending increases as sort of Sovaldi versus WalMart,” says Cutler. “Sovaldi on the one hand is expensive and adds to short-run spending. Who knows about the long-run?”

Then there’s the WalMart effect: Efficiency that drives prices down through scale, standardization, and tough-minded purchasing. “It’s really a question about which will be bigger: the Sovaldi effect, or the WalMart effect?” says Cutler.

Historically, American health care has benefited little from the WalMart effect and experienced heaps of the Sovaldi one. Cutler says that lopsidedness is changing, particularly since the signing of the Affordable Care Act in 2010.

The ACA’s effects on spending are just beginning to be felt, says Cutler, directly, in the form of payment reductions to hospitals, home health agencies, and Medicare Advantage plans but also indirectly: “The ACA has fostered a climate where people understand that the medical system is going to change in a way that cost savings are going to be rewarded relative to volume increases,” says Cutler. “The extent to which that continues is going to depend to a great extent on policy.”

Coordination matters

Drug costs aren’t likely to derail the new era, agrees Umland. The wild card—for employers—is the ACA employer mandate, which requires businesses with 100 or more workers to offer coverage to all who clock in for 30 or more hours per week. Most employers already offer coverage to their full-time employees but under this new rule, about a third of large employers have had to extend coverage to workers who were formerly not eligible, says Umland. “If many of these employees take up coverage, employers could spend significantly more on health benefits.”

What will Congress do?

That wild card may be offset by another one: politics and Republican control of both houses of Congress. Reinhardt says that when it comes to controlling health care spending, Republicans have tended to walk the walk. “They have never hesitated to beat up on the health care industry when they get fed up with it.” He ticks off examples: President Nixon, price controls; President Reagan, DRGs; President George H. W. Bush, resource-based relative value scales.

Reinhardt also mentions the Balanced Budget Act of 1997, a Republican idea to reduce payments to providers and hospitals. “Democrats are like drunken lovers at the bar,” says Reinhardt. “They talk about a lot of lovemaking but it’s all talk and little action. Republicans pretend they hate regulation, but they don’t hesitate in the least to use it. Think about it: The pro-market Reagan basically introduced Soviet pricing for Medicare—the DRGs—where the central government sets prices for the whole country.”

Even the Harvard faculty feels the effects of cost shifting, with some professors griping over having to pay a modest deductible for coverage.

What will matter are policies that affect providers and patients, says Cutler. “If we stay with the fee-for-service world, you make money by doing more. In the bundled payment world you make money by being more efficient. That will absolutely affect how much care is given.”

Harvard has recently made a modest move toward coverage with higher deductibles for its faculty and staff, and some professors have objected. So even the Harvard faculty is feeling the effects of the changes in American health care and how the dollars flow through the system. Who pays for what, where, and when—it’s all up for grabs.

“In reality, we’re not just observers,” says Cutler. “We can actually affect things through policy and through the incentives we create. It is possible that we are entering an era of very low cost increase, if not cost reduction. On the other hand, that’s going to depend to a great extent on policy. And policy can do things that can make that be more likely or make that be less likely.”