As they do each year, the Kaiser Family Foundation released its analysis of employer-sponsored health insurance premiums to significant fanfare. At four percent growth over the past year, premiums are continuing to grow at historically low rates.
Some – like Jonathan Chait at New York Magazine – have taken this as an opportunity to continue beating the drum on the ACA’s alleged cost-savings. In his article, Chait defends the President’s promise of $2,500 in savings per family under the ACA. The crux of his argument is that the ACA is not only expanding insurance coverage, but that it’s also keeping down costs. (For a solid discussion of important implications of the KFF report, see Chris Conover’s excellent discussion.)
“When it comes to some of the most important measures of health care costs, the ACA appears to have done little to nothing.”
The question of whether the ACA has saved $2,500 in premiums (that is what the promise related to) is difficult to answer. In his article, Chait links to another piece by Ezra Klein. In that article, health economist David Cutler is quoted saying that the savings promised have actually materialized. But whether you agree with this depends a lot on your baseline and what you use to measure savings. Cutler refers to estimates from 2009, and looks at health spending as a share of GDP. If instead, we look at estimates made in 2010, and look at health expenditures per capita, the savings aren’t really there. CMS estimated $10,048 for 2014, while the actual number turned out to be $9,695 – a far cry from $2,500 in savings per family!
Because of the difficulty in answering this question (partly because candidate Obama’s promise was intentionally vague) it makes sense to look at different measures of health care costs that are less vague. In this article, I’ll examine two of them – premiums for employer-sponsored coverage, and medical inflation.
1) Employer-sponsored (ESI) premiums
Source: Kaiser Family Foundation EHBS
The slowdown in growth of employer-sponsored premiums is a decade-long trend. Looking at average growth rates over a period of time rather than observing the entire trend is misleading because it glosses over changes in the rate of change.
Rather than being a product of the ACA, it’s likely that the slowing growth rate of ESI premiums is related to the growing share of workers enrolled in health plans with annual deductibles. Cost sharing is inversely related with premiums and can reduce the utilization of health care services. This in turn leads to even lower premiums, and possibly slower growth in premiums. Therefore, more workers enrolled in plans with deductibles can reasonably be expected to lead to lower premiums.
Indeed, the number of workers facing deductibles has grown from 55 percent (in 2006) to over 80 percent in 2015. While data before 2006 is hard to come by (neither Kaiser’s data nor the more academically rigorous Medical Expenditure Panel Survey has this data), it isn’t unreasonable to assume that the trend looked similar before then.
Conclusion #1: Given that the slowdown in growth of ESI premiums trend pre-dates the ACA, the health care law shouldn’t be credited with causing a trend that began a decade ago.
2) Medical Inflation
Has the ACA slowed medical inflation? While premiums paid by employers are related to underlying health care costs, they are also affected by the generosity of plans and the risk pool covered by the plan. Medical inflation, on the other hand, looks at changes in actual prices.
While medical inflation is growing at record-low rates over the past few years, this slowdown has to be taken in context.
The reason that we’re at all concerned about inflation in health care – and not so much concerned about inflation in the automobile industry, electronics, or agriculture – is that medical inflation has typically outpaced overall inflation. This still wouldn’t necessarily be an issue if we had some semblance of real prices (we don’t) that communicated actual cost and value to patients. Lastly, roughly half of our health care system is now paid for through taxpayer money, which means that unnecessarily high cost growth crowds out government funding that would otherwise be available for other uses.
This is all to reiterate that context matters. If the economy was growing at 5 percent annually (in real terms), wage growth was strong, inflation was running at 2 percent, and medical inflation hit 2.5 or 3 percent – that wouldn’t be such a big problem. But with the economy and wages growing sluggishly, and inflation hitting sub 2 percent growth, 2.5 to 3 percent growth in medical costs is significant!
So focusing on medical inflation in absolute terms is misguided. There are two ways to think of the relationship between overall inflation and medical inflation.
The first, as a ratio of indices. It’s instructive to keep in mind that inflation measures are indexed to a basket of goods, and to a particular year. This allows us to compare apples to apples. What the ratio tells us is whether the difference between medical and overall inflation is narrowing. Unfortunately, it isn’t.
Source: Bureau of Labor Statistics
While medical inflation has slowed dramatically over the years, it still hasn’t slowed to the point where it’s growing at least on par with overall inflation. Moreover, inflation overall has slowed. (Some have attributed this to the secular stagnation hypothesis advanced by economist Larry Summers.) What this means, simply, is that even as medical inflation is at record lows, so is overall inflation. The gap between the two is still growing.
But what about inflation growth rates? It’s possible that while the gap between inflation measures is growing, it’s growing at a slower rate. We can answer that question by comparing not the indices of inflation, but by comparing growth rates of the indices.
With the exception of two years – 2007 and 2010 – the growth rate of medical inflation has continued to outpace overall inflation. The gap in growth rates is indeed lower than in the first few years of the Bush presidency. But the decline in the gap between growth rates, again, pre-dates the ACA.
Conclusion #2: The ACA likely has had no effect on the gap between inflation and medical inflation.
The point here is simply that health care trends are immensely complicated. There are many moving parts in the health care system, and the ACA is a relatively small part of that. It’s within the realm of possibility that the ACA’s cost-moderating attempts (primarily the Cadillac Tax, ACO pilot programs, and the imposition of readmission penalties) will lead to lower health care spending. But when it comes to some of the most important measures of health care costs, the ACA appears to have done little to nothing.
This piece originally appeared in Forbes.com