Open Letter to President Obama About His JAMA Paper

Dear President Obama,

If I were to tell you that alligators and southern accents are correlated and that alligators cause southern accents, what would you say? You’d say, “Yes, Kip, there is a correlation, but it’s weak. But more importantly, even if the correlation were strong, there is no plausible mechanism by which alligators could influence accents. Therefore I reject your conclusion.”

I offer the same rejoinder to your argument in your August 2 JAMA article  that the Affordable Care Act has reduced health care inflation. In that paper, you claimed the ACA has “contributed to a sustained period of slow growth in per-enrollee health care spending,” and you cited the low average inflation rate of the five-year period 2010 to 2014. But that period correlates only loosely with the period of very low inflation that began in 2008 and ended in 2014. The fit is even worse if we define the “sustained period of slow growth” as 2004 or 2005 to 2014 as some do. To give you the benefit of the doubt, I’ll assume you were referring to the 2008-2013 inflation lull.

Of far more importance, even if the correlation were strong, there is no plausible mechanism in the ACA that could have caused more than a tiny fraction of the 2008-2013 slowdown.

Your post-presidency sabbatical

The purpose of this letter is to urge you to find the time after you leave office to discover for yourself there is no cost containment in the ACA for the non-Medicare population and very little for the Medicare program, and the net effect of the ACA on national health spending is inflationary, not deflationary. If the ACA’s so-called cost-containment provisions don’t cut costs (or even raise costs when all costs are measured), but the ACA raises inflation by expanding coverage, its net effect has to be inflationary.

You of all people need to acknowledge that fact and do something about it. Your name will forever be tied with the ACA. The ACA has never been on a solid footing, and now that inflation is resuming, it should be clear even to you and other diehard ACA supporters the ACA is in trouble. If nothing changes, the ACA could become what Henry Aaron predicted in 2010 it might become – “zombie legislation, a program that lives on but works badly.”

Just because you’re leaving office doesn’t mean you must sit on the sidelines and watch the ACA take a slow nose dive. If you set aside some time after you leave office to immerse yourself in health policy, and if you give high priority to finding the truth and low priority to making the ACA look good, you will conclude as I have that the ACA has little cost containment in it. You’ll conclude, as I have, that Peter Orszag, Zeke Emanuel, Jeanne Lambrew and the other advisors who told you “accountable care organizations” and other pay-for-performance fads could cut costs were merely regurgitating groupthink developed over the last half-century by the managed care movement.

And once you have determined that ACOs and the other cost-containment nostrums in the ACA are not lowering inflation and some (notably ACOs and “medical homes”) may actually be inflationary, you can then use your influence to educate Congress and the health policy elite about why the ACA is aggravating inflation and what should be done about it.

In this letter and two comments I’ll post shortly, I will offer a quick tutorial on where you went wrong in claiming the ACA is anti-inflationary. Obviously I don’t have the space here to teach a graduate course in health policy. But I do believe that in three essays I can make you suspicious of the groupthink you were exposed to in the White House, and I believe I can whet your appetite for the antidote to groupthink – evidence-based health policy.

It’s the economy, stupid!

There is near-universal consensus that the 2008-2013 inflation lull began in 2008 because that was the first full year of the Great Recession, and the lull ended in 2014 because the recession’s worst effects were over by then and because the most inflationary provisions of the ACA – the coverage expansion provisions – kicked in that year. Other factors contributed to the lull, but the recession was far and away the most important.

The latest report on the National Health Expenditure Accounts, a document you cited four times in your paper, makes it crystal clear the inflation lull came to a halt in 2014. Here is the very first sentence in the summary of that report:

“In 2014, U.S. health care spending increased 5.3 percent following growth of 2.9 percent in 2013 to reach $3.0 trillion…. The faster growth experienced in 2014 was primarily due to the major coverage expansions under the Affordable Care Act, particularly for Medicaid and private health insurance.” [1]

How did you miss that statement? Inflation almost doubled in 2014 primarily because of the ACA. Inflation is projected to rise gradually to 6 percent by 2020

Why did you fail to tell us the “sustained period of slow growth” ended two years ago?

There is widespread agreement that the enormous loss of income and wealth caused by the 2007-2009 Great Recession is the most important reason why the inflation lull began and ended when it did. The reason for that is the existence of a well-documented correlation you neglected to mention – a consistent correlation over time, over space, and over income classes between income and health care spending. Richer nations spend more on health care than poorer nations; upper-income Americans spend more on health care than lower-income Americans; and Americans as a group spend more on health care during good times and less in bad times.

To put that last point more precisely, we spend less during periods that include recessions and recovery from recessions, and we spend more during periods following recovery from recessions. Because the Great Recession was brutal (it was the worst recession since the Great Depression), its effect on medical spending was immediate and its effect lasted beyond 2009.

Perhaps the single best paper documenting the tight link between the recession and the inflation lull is one by Dranove et al.  entitled, “Health spending slowdown is mostly due to economic factors, not structural change in the health care sector.” Your former advisor Peter Orszag cited that paper in his editorial commenting on your paper.

Alligators don’t cause accents

The correlation between recessions and health care inflation is not only consistent over many years and reasonably tight, but it has a plausible explanation. Conversely, the correlation you asked us to accept (the 2008-2013 inflation lull and the short lifespan of the ACA) is loose and, with one possible exception that I will discuss and dismiss below, has no plausible explanation.

The explanation for the correlation between recessions and health care inflation is not just plausible, it is obvious. Medical care is expensive (it accounts for a sixth of our Gross Domestic Product) and much of it can and will be put off when more basic needs like food and shelter become harder to pay for. Conversely, when most households have recovered from recessions, national spending on health care returns to levels approximating pre-recession levels. By contrast, with one remotely possible exception (the cuts to Medicare), your claim that there is something in the ACA that would explain the last four years of the inflation lull (2010-2013) defies commonsense.

I urge you to read the annual reports on national health spending during the post-2010 portion of the inflation lull by CMS’s Office of the Actuary (OACT). You’ll see that OACT gave the ACA virtually no credit for the inflation lull and instead attributed the lull to the Great Recession. For example, in a paper  on national health expenditures for 2011, OACT stated: “Although some provisions of the Patient Protection and Affordable Care Act … were in effect in 2010 and 2011, the impact on aggregate health spending growth was minimal in these years. The most prominent provisions of the act will not be implemented until 2014.”

In its report on national expenditures for 2013 ,OACT again attributed the continued low inflation to the recession. OACT mentioned the ACA only in passing, and then merely to say that the ACA contained inflationary and deflationary provisions. [2] Of the four deflationary provisions OACT mentioned, only two – the cuts in Medicare’s FFS program and in the Medicare Advantage program – constitute remotely plausible mechanisms by which the ACA could have made a dent in inflation during the 2010-2013 period. But upon examination, those cuts do not constitute a plausible mechanism because they did not take effect until 2012, and they did not become substantial until 2014. [3] To repeat, by 2014 the inflation lull was over.

To your credit, you did not attempt to argue in your JAMA paper that ACOs and other “alternative payment model” (APM) programs authorized for Medicare by the ACA should take any credit for the inflation lull. That would have been foolish for two reasons. First, like the cuts to Medicare, the APM demonstrations weren’t implemented till late in the 2008-2013 period. Second, those demos are saving little or no money and may be raising costs when all costs incurred by all payers (doctors, hospitals, foundations, and public and private insurers) are taken into account.

For the record, the Medicare inflation lull ended in 2014 as abruptly as the system-wide inflation lull ended. [4] Medicare inflation resumed in 2014 despite the big cuts inflicted on Medicare by the ACA and the 2011 sequestration legislation.

Electronic medical records: An overlooked cause of inflation

Oddly, you made no mention of electronic medical records (EMRs) as a mechanism that deserves credit for the inflation lull. You have been a huge fan of EMRs. You believed the folklore promoted by the Institute of Medicine and the computer industry that EMRs would cut medical costs by more than the EMRs cost to install and maintain. Because you believed that folklore, you enthusiastically supported legislation (the HITECH Act, the ACA, and MACRA) that put financial pressure on doctors and hospitals to buy EMRs.

The evidence indicates that the campaign to induce doctors and hospitals to buy EMRs (which began under Bush II) has raised total health care spending. A paper published in 2005 concluded that the cost of installing EMRs in all hospitals and clinics will raise national health expenditures by 2 percent. [5] Not every clinic and hospital has purchased one of the clunky EMRs available for sale these days, so we can’t say the cost of buying and maintaining EMRs has reached 2 percent of our $3 trillion health care bill yet. But we’re getting there. By 2013, according to the CDC, eight in ten clinics and six in ten hospitals had purchased an EMR. Meanwhile, the evidence indicates EMRs are not cutting costs. We may conclude, therefore, that the ACA has contributed to health care inflation by putting pressure on doctors and hospitals to buy EMRs.

Please note I do not dispute those portions of your JAMA paper which argued that the uninsured rate has fallen and that millions of Americans have access to medical care they didn’t have before. I have focused on your cost-containment claims in order to make it clear to you the ACA will never “bend the cost curve.” If it can’t reduce health care inflation, it is unlikely any Congress, even a Congress consisting of 100 percent Democrats, will ever raise taxes high enough to maintain the lower uninsured rate the ACA has achieved to date. I want you to understand the ACA is headed toward “zombie-law” status unless ACA proponents get it through your heads that the ACA’s faddish “value-based purchasing” nostrums will have little or no effect on health care inflation and might even make it worse.

Analyzing the bad advice you got

I believe you cling to the myth that the ACA deserves substantial credit for the inflation lull, and you cling to the hope that the ACA will lower health care inflation in the future, because you bought the wrong diagnosis of the US health care crisis. You bought the diagnosis peddled by the managed care movement since 1970, namely, that US health care costs are high because patients get too much medical care, and this alleged overuse problem is in turn caused by the fee-for-service method of payment. Having bought this diagnosis, you naturally bought the “solution” proposed by the managed care movement – the fee-for-service incentive must be turned upside down by exposing doctors and hospitals to insurance risk, and doctors must be micro-managed. That diagnosis is wrong, and so is the solution.

In my next two posts I will criticize both the managed care diagnosis (overuse) and the managed care solution (risk-shifting and micro-management). I will do so by focusing on three managed care proponents who influenced you deeply: Atul Gawande, Elliot Fisher and his colleagues at the Dartmouth Atlas, and Peter Orszag. If I can show you how they misled you, I believe you’ll be more open to my argument that the ACA is inflationary.

As you can tell, I really don’t like your health policy. But I want you to know I voted for you twice, and I have great respect for you. Thank you for being our president in such tumultuous times.

[1] Here are the inflation figures (increase in national health expenditures) for 2007 through 2014, beginning with 2007: 6.5 percent, 4.6, 3.9, 4.0, 3.9, 3.8, 2.9, and in 2014, 5.3. (Source: National Health Expenditure Accounts https://www.cms.gov/Research-Statistics-Data-and-Systems/Statistics-Trends-and-Reports/NationalHealthExpendData/NationalHealthAccountsHistorical.html )

[2] Here are the inflationary and deflationary factors OACT mentioned in its report http://content.healthaffairs.org/content/early/2014/11/25/hlthaff.2014.1107.full on 2013 spending: “A few key provisions [of the ACA] exerted downward pressure on health spending growth in 2013, including the productivity adjustments to Medicare fee-for-service payments, reduced Medicare Advantage base payment rates, increased Medicaid prescription drug rebates, and the medical loss ratio requirement for private insurers. At the same time, other provisions – such as early Medicaid expansion initiatives, a temporary increase in Medicaid primary care provider payments, reducing the size of the Medicare Part D doughnut hole, and the implementation of drug industry fees – exerted upward pressure on health spending growth.”


[3] The Congressional Budget Office projected ACA-authorized Medicare cuts would total $9 billion in 2012, $14 billion in 2013, and $40 billion in 2014 (see Table 2 of this 2011 CBO report https://www.cbo.gov/sites/default/files/112th-congress-2011-2012/costestimate/hr22.pdf ). The $40 billion cut equaled about 1 percent of national spending in 2014, a significant amount. But by 2014, the inflationary coverage expansion provisions swamped the Medicare cuts.


[4] Medicare inflation rose from 3.0 percent in 2013 to 5.5 percent in 2014 (see Exhibit 3 in this paper http://content.healthaffairs.org/content/early/2015/11/25/hlthaff.2015.1194 behind a paywall) , and is projected to remain in the 5 to 7 percent range over the next decade (see Exhibit 2 in this paper http://content.healthaffairs.org/content/early/2016/07/12/hlthaff.2016.0459, also behind a paywall).

[5] Here is a quote from that paper: “To achieve an NHIN (national health information network) would cost $156 billion in capital investment over 5 years and $48 billion in annual operating costs. … $156 billion is equivalent to 2 percent of annual health care spending for 5 years.” (Rainu Kaushal et al., “The costs of a National Health Information Network,” Annals of Internal Medicine 2005;143:165-173, 165.)

Categories: Uncategorized