Chart of the Day: Increases in Healthcare Costs

The chart of the day comes from the Bureau of Labor Statistics. Pictured above is a chart of the year-on-year percentage changes in healthcare costs in the United States from 1936 to date. As I see it there are two takeaways from this chart. First, look at 1966. What could possibly have caused the massive increase in healthcare costs starting in 1966? Of course, I’m being facetious. We all know what caused that increase and it’s not greater use of technology.

Second, look at the year-on-year increases from roughly 1980 to date. Healthcare is increasing in cost at roughly the rate of inflation. The problem is that practically everything else has seen its costs battered down by overseas competition. All it takes is a steady rate of increase and by the magic of compounding costs become unsustainably high.

17 comments… add one
  • First, look at 1966. What could possibly have caused the massive increase in healthcare costs starting in 1966? Of course, I’m being facetious. We all know what caused that increase and it’s not greater use of technology.

    So…if the advent of Medicare led to a substantial increase in the price growth of health care…and we repeat that policy with the non-elderly……

    Now granted the non-elderly are not going to be using health care resources as intensively, but still, the logic is the same. We’d expect to see another jump in the growth rates of health care prices–i.e. health care costs. The current legislation wont make the problem better it will make it worse.

    But what the Hell, lets just call health care a right and let the chips fall where they may. If you are in your late 50’s or early 60’s I don’t envy you. You are likely to be the ones who get the shaft first….and likely with no KY since that will likely cost too much.


  • First, look at 1966. What could possibly have caused the massive increase in healthcare costs starting in 1966? Of course, I’m being facetious. We all know what caused that increase and it’s not greater use of technology.

    Why focus on 1966? I’m more interested in 1953, because that appears to be when the increase in health costs started being higher than inflation, a trend that remains constant with only a couple of exceptions.

  • steve Link

    How much chart and data reading do you do on your day job? What I see is medical costs leading from 52 to about 64. Actually, starting in about 52, they almost permanently lead. The spike in 66 is followed by a delayed lower spike in all items. The biggest separation n costs comes from about 80 to the early/mid 90s. Since 96, there has been a steady small separation. The more appropriate/important question then is, what happened in 1952? The second question should be what happened in 1980? The 66 spike is important, but if you were looking at this not knowing what is was about, you would look at those two data points first.

    I have to go to a meeting and figure out more ways to screw the medical system, errr enhance revenue, or I would try to print this out for you and figure out the areas under the curves. It would also be helpful to have a real statistician look at the raw data. I have certainly screwed up chart interpretation before.

    BTW, my wife has been researching time lines on medical developments and technology really started to change in the 50s and 60s. Remember that Patrick Kennedy died after being born 5 weeks early in 1963. Joint replacements started in bigger numbers in the late 60s. The first successful mitral valve was done in 1960.


  • Andy Link

    Hmm, I’m not so sure Dave. While I’m sure medicare did result in increased costs through more utilization, it looks to me like health-care costs don’t begin to significantly outpace general inflation until the 1980’s.

  • Andy Link

    Also, isn’t the early 1950’s when employer-sponsored health care really gets rolling?

  • Alex,

    1966 is when Congress created Medicare.


    You are kidding right? In most years since 1953 medical care costs outpace overall price increases.

    Everyone else (save maybe Dave),

    As for the spikes in all items in the 1970s, c’mon people did you fall asleep in history class, weren’t alive back then…what? Oil shocks. Stagflation. High unemployment and inflation.

    As for the 1950s, I’be be willing to bet it was the policy developed during WWII where employers started to offer insurance benefits as part of workers compensation packages. That and with the end of WWII, the demand for other goods likely dropped leading to lower growth in prices over time.

  • Medicare was enacted into law in July of 1965.

    Note also that “all costs” is inclusive of healthcare costs and that at this point healthcare costs comprise roughly one sixth of all costs. Also this chart doesn’t show the change in healthcare costs it shows the year-on-year percentage change.

    I think this chart is a nearly perfect graphical representation of the conclusions from the study I’ve cited here before that showed that for the last 45 years the cost increases in healthcare have proceeded in three phases: 1966-1972, 1972-1984, and 1984 to the present. Remarkable. In the study they attribute the cost increases 1966-1972 to increased utilization, those 1972-1984 to price increases, and 1984 to the present to inflation. A great chart.

    I also think it’s interesting that the furor over healthcare costs in the early 1990’s were subsequent to a sharp spike in costs while the current concern is not. I think we’ve reached the point at which the unsustainability has become obvious.

  • PD Shaw Link

    I don’t see this healthcare cost spike in ’52.

    Starting in 1940, I see three spikes in the overall CPI, which healthcare mirrors though generally as a trailing indicator. AND the healthcare costs don’t subsequently fall as far as CPI after each of the three spikes. That suggests to me that healthcare costs in ’36 to ’56 are primarily wage based. Wages are sticky and don’t rise as fast, and more importantly don’t fall as much.

    The fourth spike in the CPI (around ’58) creates the same pattern, except this time healthcare costs rise higher than the overall CPI. That appears to be a relic of wage-stickiness from the previous three spikes. A breakdown on goods and services components of the CPI might bear this out. We have found ways to reduce widget costs (or introduce substitutes). We haven’t found ways to get people to continue to work for a lower wages.

    But the fifth CPI spike around ’70 is different. This is the first time healthcare costs spike before CPI. So, I think Dave has a point.

  • steve Link

    I am not sure that one leading the other is determinant of costs. What is clear is that starting in the early 50s, health care costs began to exceed over all costs. I think Insurance probably played a part.

    PD-It is a divergence that starts in about 52. The problem is the differential.
    Dave-Why do you ignore that period from 52 to 64?

    The other part I forgot to mention earlier is VA and DoD spending on health care. That hit at about 4.8% of health care costs in 1966. remember that in 65 we only had about 180k in country. That doubled next year and went up to about 530k in 1968. It dropped bellow 150k in 1972. We had a total of about 300,000 wounded and most of those would have occurred after 1964 as we still only had about 20k in country in 64.


  • PD Shaw Link

    If there is any differential in ’52, it’s because the CPI is acting differently. From ’40 to ’55, CPI is averaging peaks of +12 and valleys of -11. For medical costs, the peaks average about +4.33 and valleys of -3.67. After ’52, the CPI stabilizes from such shifts, and begins to look more like the gentle shifts of the medical costs.

    When I look at ’52, I wonder why overall CPI changed its behavior. Likely, it’s because of the ’53 recession, followed by the low-inflationary policies of the Eisenhower era, and increased productivity in goods.

  • I’m interested as to why there was such a downturn in costs during the 90’s. Thoughts?

  • PD Shaw Link

    E.D. Kain, I’ve read a report that said healthcare costs for the elderly have significantly improved due to increased availability of nursing home alternatives in the last 10 years.

  • PD Shaw Link

    Also, there is an OECD study on comparative renumeration for physicians that indicates that American primary care physicians have experienced negative real growth of annual renumerations (IIRC) around 1998-2004.

  • There was no downturn. The rate of increase stayed reasonably steady.

  • steve Link

    I believe there was a change in physician reimbursement in 98-99 so that there was an actual drop in Medicare spending.

    PD- I still think that the key is not what the CPI does, but what medical costs do in comparison. Starting in the early 50s medical costs have nearly always exceeded the CPI. As Verdon noted, compounding that means medical costs eventually consume everything.

  • I’m interested as to why there was such a downturn in costs during the 90’s. Thoughts?

    Uhhhm, what down turn in costs? The line is everywhere above zero since 1942 meaning that costs have gone up in every year since 1942. The lines are year on year changes. That is (using guesses based on the graph) costs go up for each year as:

    1943 2%
    1944 3%
    1945 4%

    As such costs were constantly rising and there was no downturn at any time at all since 1942.

Leave a Comment