The health-cost slowdown isn’t just about the economy

By David Leonhardt

New York Times, December 5, 2014

It’s one of the most important economic questions today: Is the snail-like growth of health costs over the last several years a real trend, or is it merely a temporary part of the Great Recession’s aftermath?

The data experts who compile the government’s official numbers on health spending lean toward the more pessimistic view. They think the slowdown – to the lowest level of growth on record – stems in large part from Americans skimping on medical care during tough times.

“The pattern observed in recent years is not unique and is consistent with historical patterns,” Anne Martin of the Centers for Medicare & Medicaid Services said after that agency released new numbers this week. The agency’s report put the argument this way: “The key question is whether health spending growth will accelerate once economic conditions improve significantly; historical evidence suggests that it will.”

I do think the government’s analysts are putting an overly pessimistic spin on the numbers.

It’s simply not true that G.D.P. and health costs have historically moved in tandem.

The annual rate of growth in total national spending on health care tumbled in 2009 and continues to be very low. This is happening primarily because our country is struggling to recover from the worst recession since the Great Depression, and secondarily because insurance coverage and income inequality are getting worse.

This is not what managed care advocates like David Leonhardt would have you believe. Leonhardt argues there is simply no correlation between recessions and health care inflation and that managed care fads, particularly those endorsed by the Affordable Care Act, deserve much of the credit for the current slowdown in health care inflation.

Leonhardt cites the pay-for-performance, aka value-based-purchasing, fad. He argues that insurers, including Medicare, have begun to “make payments based less on the quantity of health care provided and more on the quality.” He claims the ACA’s provision requiring Medicare to punish hospitals that have “excess” readmissions is a successful example of the fad at work.

The actuaries at the Centers for Medicare and Medicaid Services (CMS) do not agree. In their last two annual reports on national health spending (for 2012 and 2013) they attribute the unusually slow growth in national health expenditures (NHE) to the recent recession and give no credit at all to managed care nostrums, including “accountable care organizations” and other fads endorsed by the ACA. CMS’s position is correct. Claims by Leonhardt and other ACA proponents, such as White House adviser Jason Furman, that the ACA’s managed care provisions deserve some credit are wrong.

The graph below is taken from CMS’s report on NHE for 2013.

Health Aff   hlthaff.2014.1107, Exhibit 3 (1)


The graph, which covers the last quarter-century, gives you a sense of how strong the correlation between NHE and recessions is.

The graph compares changes in NHE with changes in Gross Domestic Product (GDP). (GDP is the yardstick used to determine when the U.S. has entered into and exited from a recession. When GDP falls for two consecutive quarters, we are said to be in a recession.) The graph breaks the last 25 years into two groups: Years that contained a recession plus the two years that followed those recessions; and all succeeding years up to the start of the next recession.

The graph reveals an obvious pattern: The three recessions that have occurred during the last 25 years have all been followed by sustained declines in health care inflation. Recessions cause NHE inflation to drop, but only after the passage of a few years. And, moreover, the dampening effect that recessions have on NHE persists long after the recession ends. Leonhardt’s statement that “It’s simply not true that GDP and health costs have historically moved in tandem” is wrong. Astonishingly wrong.

Because the Great Recession was the worst recession since the Great Depression, and because NHE and GDP are strongly related, it follows that the decline in health care inflation following the last recession would be unusually deep and unusually prolonged. That is in fact what we are seeing.

Other research on the inflation slowdown supports CMS’s position. The most convincing paper on this subject appeared last August in Health Affairs. Rather than compare national GDP to national health care spending, the authors, David Dranove et al., compared employment rates at the metropolitan level to health care spending at the metropolitan level. Using a very large sample size (47 million non-elderly people insured by Aetna, Humana and United HealthCare), they found a clear correlation between the employment rate and health care inflation for the entire sample across all cities. Where the employment rate fell only slightly, inflation in health spending declined only slightly, and where employment fell precipitously, health care inflation fell precipitously.

Dranove et al. reported a lag time of only one year for the nation as a whole between the initial decline in unemployment and health care inflation: The employment rate dropped precipitously in 2008, and in 2009 the health care inflation rate fell sharply. They also reported that both the employment and health care inflation rates remained depressed two years after the recession officially ended in 2009 (2011 was the last year examined by the study).

Dranove et al.’s findings confirm, and shed more light on, the correlation between recessions and the rate at which health care spending increases. When the unemployment rises, people postpone medical care. And demand for medical care stays depressed even after recessions end because the employment rate takes a while to recover.

This finding is especially relevant today. It may be true that the Great Recession has been officially over since 2009, but it is also true that the recovery from that recession has been unusually slow and has disproportionately benefited the wealthy. The unemployment rate has taken longer than usual to come down, and incomes for average Americans have stagnated.

According to a recent Wall Street Journal analysis of income and spending by the middle class (which the Journal defined as the 60 percent of households earning between $18,000 and $95,000), middle-class income has declined in inflation-adjusted terms since 2007.

Other evidence indicates that the effect of the Great Recession has been aggravated by the rising cost of health insurance coupled with rising copays and deductibles, and that this trend has gotten worse during 2014 despite the drop in the nation’s uninsured rate caused by the ACA. A poll released by Gallup in November reported that the percentage of Americans who have put off medical care because of cost ticked up 3 percentage points between last year and 2014, and that the increase was far worse (9 percentage points) among those with private insurance.


Two New York Times/CBS polls, one conducted in December 2013 and the other this December, confirm Gallup’s finding. The 2013 poll reported that 31 percent of respondents said paying for “basic medical care” is a hardship. That number had risen to 46 percent by the end of 2014. The poll also reported that out-of-pocket costs have soared because coverage is getting worse and health care fees and prices (as opposed to utilization) have gone up.

As the latest CMS report notes, a slowdown in Medicare inflation contributed to the reduction in the rate of growth in NHE during 2013. Medicare expenditure growth dropped slightly in 2012 as well. Some ACA and managed care proponents have asserted that this is evidence that some or all of the managed care fads imposed upon Medicare by the ACA must be working. They argue that Medicare beneficiaries are more insulated from the effects of recessions than the non-elderly and for that reason the recession cannot have contributed to reduced Medicare inflation. This argument no doubt exaggerates the isolation of Medicare from the general economy, but for the sake of discussion let’s accept it at face value.

This “Medicare is insulated from recessions” argument ignores an obvious explanation for the recent reduction in Medicare inflation: Congress has ordered several cuts in Medicare spending in recent years. The cuts authorized by the ACA in 2010 and the Budget Control Act of 2011 (the “sequestration” bill) were large by any standard. In its last two reports, CMS said these cuts explain the bulk of the recent reduction in Medicare inflation. A study just released by the Kaiser Family Foundation confirms CMS’s explanation.

One does not have to have a PhD in economics to grasp the significance of the facts I have just presented. These facts – the longstanding correlation between recessions and NHE inflation, the severity of the last recession, the rising cost of health insurance along with rising out-of-pocket costs, and the devastating and prolonged effects of the last recession on the incomes of average Americans – explains why we are seeing historically low rates of health care inflation. We need not consult the Book of Managed Care Theology to find other reasons.

If income inequality continues to worsen, and if insurance premiums and out-of-pocket costs continue to grow faster than the incomes of the vast majority of Americans, we may see a weakening of the longstanding correlation between GDP and NHE. If that happens, it won’t be evidence that managed care fads are finally working. It’ll be evidence that GDP is no longer a reasonably accurate proxy for the income of the average U.S. household and that economists need to find a better measure of income than GDP.

But rest assured that if annual growth in national health spending continues to stay low despite growth in GDP, managed care and ACA proponents will claim managed care fads are working. They will do this even though research on the individual fads, be it ACOs or “medical homes” or electronic medical records, fails to find these nostrums save more money in reduced medical costs than the nostrums themselves cost to implement.

Kip Sullivan, J.D., is a member of the steering committee of the Minnesota chapter of Physicians for a National Health Program. His writing has appeared in The New York Times, The Nation, The New England Journal of Medicine, Health Affairs, the Journal of Health Politics, Policy and Law, and the Los Angeles Times.