Medicare Spending Projections Drop By $511 Billion


As I have posted before, our debt problem is directly tied to our healthcare problem; we spend way too much on substandard care, when compared with the rest of the advanced world. Some good news has emerged on that front, however. The Center for Budget Policies and Priorities (who thinks of these names?) has a new forecast of Medicare spending through 2020.

The takeaway from this chart is simple: healthcare spending is America is slowing, and because of that, Medicare is expected to spend $511 billion less than it would otherwise.

CBO has reduced its projections of Medicare spending in response to a pattern of very low spending growth in the past three years.  Medicare spending per beneficiary in fiscal year 2012 increased by only 0.4 percent — well below the 3.4-percent growth in gross domestic product (GDP) per capita.  Over the 2010-2012 period, Medicare spending per beneficiary grew at an annual rate of 1.9 percent, while GDP per capita increased by 3.2 percent a year.  CBO projects that Medicare cost growth will remain subdued for the rest of the decade.

Medicare spending growth has slowed even more than costs in private health insurance, according to Standard & Poor’s and Medicare’s actuary.  Although some of the slowdown stems from the recession, CBO Director Douglas Elmendorf and other experts have concluded that a substantial part reflects structural changes in the health care system.  Professional associations, hospitals, and doctors are taking steps to curb costly and ineffective procedures and treatment.

No one knows how long this good news will continue, but even if cost growth remains moderate, Medicare spending will keep rising as more baby boomers become eligible for benefits.  Making the U.S. health care system more efficient thus remains our biggest budget challenge.  Yet the recent spending data and CBO’s new projections belie claims that policymakers must radically restructure Medicare to put it on a sound financial footing.

This entry was posted in Politics and tagged , , , , . Bookmark the permalink.

11 Responses to Medicare Spending Projections Drop By $511 Billion

  1. First paragraph ” 202o” instead of “2020” second paragraph “healthcare spending IN America…”. Really like Tracy’s comment. Very insightful.

  2. I am a bit skeptical of this here. I actually work as a medicaid forecaster and I am familiar with this area. My primary concern is that a change in trends for 3yrs is not necessarily a good gauge of long term trends. They are literally saying because 2010-2012 costs grew slower than expected then 2013-2020 costs will do the same. Think about that, they take 3 yrs in which cost are different then project based on that out 7yrs. If they were my forecasts I would be especially wary of utilizing recent growth in long term projections. Rather I would prefer a model that maintains the low growth in the short term (2 or 3yrs) with a return toward a historical trend in the later years. Though I could accept a lower long term forecast if it is based on a model that has been well proven for the data series and has been compared to a variety of other models. That is the model produced that answer and that particular model has been shown historically to be more accurate than other options I would be more accepting of a long term reduction due to statistical support. But I don’t know if that is what is happening of if they wanted to make it lower by using a model which would weigh the recent low trends much greater than historical trends such as an exponential smoothing model which may not be appropriate. Also I am skeptical because it is coming out of CBPP and not CMS. CBPP may be just fine but the actuaries in CMS live and breath those costs day in and day out. I trust those who’s entire job is to focus on those costs more than a group which deals with a huge variety of costs, forecasts and policy impacts.

    • I love that your comment is longer than my entire post! I always appreciate your insight, and what you wrote makes a lot of sense. Why do they never publish the range of forecasts instead of just one scenario. After reading Nate Silver’s book, I try to look at forecasting with a skeptical eye, like you pointed out. I should have added that when I wrote this piece, but luckily I have amazing readers like you!

      Whats the CMS, by the way? I can’t recall if I know the acronym.

      • CMS is the Centers for Medicare and Medicaid Services. That is the federal agency that oversees all Medicare and Medicaid. Of anybody they have the best data on Medicare and Medicaid as well as the fact that their sole focus is on those two programs. That is why I would trust their numbers best because they are in the best positions to analyze the numbers.

        As for publishing a range of forecasts that could be possible but the issue still is how are those forecasts developed. Unfortunately without being directly involved in the forecasting it is hard to tell. They could be using an objective method based on historical performance and a competition between a variety of models which may produce good numbers. Or they could be building a model to give them the answer they want. If you hand me a data set like Medicare costs with 3yrs of lower trends at the end I could produce models with very low trends or with trends that are much higher. In fact you would give me the data set and tell me the trends you want to see and I probably could make that for you. Then you could claim that the forecast is based on a model when in fact it is based on what you want to see and simply produced by a model.

        So what I am saying is that even with a range of forecasts they can be manipulated. It all depends on the process utilized to develop the forecast not just a range of projections. When I produce forecasts I run between a dozen and two dozen models for every series. I do this quarterly and I have about a decade worth of past forecasts to evaluate. Then I measure error on each model produce in several different manners. Based on that I am able to identify models which have been proven historically to be accurate and I can use those for my forecasts.

        BUT then comes the big caveat, Hume’s problem of induction. ALL forecasting (and predictions for that matter) are based on the assumption that the future will reflect the past. That is historical behavior is representative of future behavior. That is simply isn’t true all the time such as when there is major policy change or the conditions in the country change or other differences that effect the series being forecast. In those cases you can attempt to account for the change if it has been seen elsewhere and try to apply that to the current situation but it can run into problems of generalizability. Or you can simply guess at the effect that the change will have which is probably more common than you think.

      • I’m always so impressed by your comments. Thanks for giving me an education!

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s