With the cost of health care just continuing to spiral higher and higher as the years go by, it becomes increasingly difficult to advise clients about how much to save to handle those future costs in retirement. On the one hand, it's crucial not to undersave, such that ongoing health care costs devastate and deplete the retirement portfolio; on the other hand, excess conservatism can be bad too, forcing clients to unnecessarily constrain their lifestyle with more saving than is necessary, or working longer and retiring later than was actually needed. So just how much do you assume your mass affluent clients will pay in projected future health care costs during retirement?

The inspiration for today's blog post comes from an email I received from a reader, who was struggling to decide what "reasonable" assumptions would be for projecting health care costs in retirement. Her approach - which I have seen many other planning firms use as well - was to project health care costs separate from all other retirement costs, specifically so that a higher inflation rate could be attached to the health care expenses. Her assumptions? 3.5% for general inflation, and 8% for health care costs.

As the saying goes, trends that can't continue forever, don't. At some point they end. And I have to admit, I think our inflating health care costs fall into this bucket. After all, a spread of 8% health care inflation and 3.5% general inflation leads to some pretty wacky results when projected over a multi-decade period of time. Eventually health care begins to crowd out other parts of the economy. In the long run, the only jobs we'd all have is providing health care services to each other, as that portion of expenditures - nationally - would eventually consume the overwhelming majority of the household budget. And that's just not realistic in my opinion; from the lack of financial viability of Medicare, to the direct household impact, health care costs just cannot inflate indefinitely at these kinds of levels. At some point, it just becomes unaffordable for everyone, and the system must change.

On the other hand, it's also worth noting that for most retirees, especially those "mass affluent" or wealthier with some retirement asset base, the driving cost of retirement health care costs isn't really the cost of underlying care, per se; it's the cost of the insurance that covers it. After all, most retirees are not paying raw health care costs out of pocket; they're covered by Medicare, and what they're paying for is Medicare Part B, a Part D prescription drug plan, and some co-pays and costs that slip through the cracks (or a MediGap supplemental policy to cover those, too). Thus, in essence, the "cost of retirement health care" for most retirees probably looks more like the cost of Part B + Part D + MediGap supplemental insurance, and whatever the cost of that coverage inflates at over time.

For instance, a married couple who earns less than $170,000/year of income will pay up to $115.40/month/person for Medicare, or a total of $115.40 x 2 x 12 = $2,769.60/year. We'll assume a general Part D policy for the couple costs $75/month/person, and a reasonable MediGap supplemental insurance policy costs $200/month/person, for a total annual cost of $1,800 and $4,800, respectively. At this point, we're at a total cost of $9,369.60/year, but the MediGap policy should cover the overwhelming majority of any remaining co-pay and co-insurance amounts, leaving very little if any out-of-pocket costs from this point forward.

Now I'll grant that almost $10,000/year is not cheap. And for families living on a limited, fixed income, this can be a crushing cost (and most likely, such families will spend only about half this amount, forgoing the MediGap policy or buying a much less expensive "bare bones" version, and retaining some of the remaining risk). But at the same time, for those with some level of additional savings and a reasonably comfortable retirement, these are not catastrophic costs to the retirement plan. And because virtually all medical costs have been insured at this point (although long-term care costs are another story), there's little uncertainty left (aside, perhaps, from the inflation rate on insurance costs, as mentioned earlier).

But insurance costs - at least, for Medicare - aren't inflating anywhere near 8%. In fact, due to the flat no-change cost-of-living adjustments for Social Security benefits in recent years, some people are still paying the $96.40/month/person Medicare Part B premiums from two years ago! Insurance premium inflation has been 0% for these people!

So the bottom line is that while the cost of health care is a serious issue to contend with, I'm skeptical of the idea that health care costs will just spiral upwards forever at levels significantly higher than general inflation, such that it overwhelmingly consumes the budget of the average mass affluent retiree. Not only is that not sustainable, but when viewed from the perspective of the cost of Medicare and other insurance, that doesn't seem to be entirely happening even today. Granted, Medicare itself may go through some significant changes in the future, as its sustainability is questionable as well. But indirectly, that just makes the point again - health care costs that inflate excessively can't go on forever. It can't, so it won't.

So what do you think? How do you handle projecting health care costs for your retired clients? Do you separate out health care costs in the retirement analysis, or just include it as part of the overall package? Do you look at the costs of health care itself, or the costs of the insurance it takes to protect against those costs?

  • http://www.bonadio.com Cindi Turoski

    Great topic! We have grappled with this over the years too. We typically used $5k per yr for a couple once on Medicare and a 6% inflation rate. We then left vacation expenses in the analysis for lifetime knowing as they get older they likely won’t vacation like that, but the costs could instead be applied to increased medical need as they age.

    However, we recently increased the estimate to $10k per yr for a couple (to include all you mention, dental, eye care, procedures, possible hearing aids & other things, etc.), still indexing it separately at 6%. We are concerned, though, about the possibility that Medicare will shift more costs to the consumer due to budgetary issues. We recently read about a legislative proposal that would have the consumer increase their current 25% cost-sharing to 68% cost-sharing on Medicare. Is that where things are going, what is the right number, what is reasonable and justifiable to a client?

  • Tom Davison

    A tabulation I find interesting is provided by Fidelity at

    There is a small update for 2011 but does not appear as a table of details. More information on their website about chronic disease may be useful as well.

    We do use a separate number for healthcare, also noting that vision, dental, hearing aids, and a variety of other costs come in. We do use a healthcare inflation rate of about 3.5% over the regular inflation rate, which as Michael points out, would not be sustainable for very long times. (that spread is not as wide as the 4.5% spread in the post).

    The more discretionary income the client has, and the more current health issues, the higher the costs are likely to be. On the other side, one of the more expensive lifetime costs for wealthier folks is 24×7 home health care for an extended time. But that is largely due to semi- or even fairly low skilled labor costs, which should track more closely to wage inflation. Like most things, the shorter the timeframe and the more that is known about the specific client, the better the prediction can be.

    At some point, personal bankruptcy caps the costs, which happens to a fair number of lower income people). A very difficult item to predict, with significant unknowns.

  • http://bonadio.com Jean Bedell

    What we are trying to figure out is if we are starting our medical expense figure out at $10K which includes the Medigap policy, there are many things that may not be covered (hearing aides, dentures, other dental procedures, eyecare, etc.). As need increases with age, will using a 6% inflation rate account for those extra costs that someone may incur? In the past we have used the vacation expense as their “cushion” for increased need, but maybe the 6% accounts for it. What are your thoughts?

    • Michael Kitces

      I appreciate the concern that other health expenses may come into the picture, but I’m not a fan of managing that by just assigning a higher inflation rate. High growth rates compounded over multi-decade periods of time becomes very distorting.

      I’d probably be more inclined to add an extra flat dollar amount – e.g., at age 75 add another $5,000/yr of health costs on top of existing projected costs.

      Some higher health care inflation rate seems appropriate over the coming few decades (5-6% ?) but I’m more inclined to add flat dollar amounts (also inflating) rather than push the exponential growth/inflation rate itself to a higher level.

  • Doug Keegan

    I actually did a lot of research on the subject when I was with Harris SBSB. We wanted to make sure we were properly calculating health care costs for our retired (and not retired) clients. We would model health care costs over and above their lifestyle costs.

    From my research, I have found five reliable sources: EBRI (Paul Fronstin has done the best work on the subject),Center for Retirement Research at Boston College, Kaiser/Hewitt 2006 Study, Fidelity (as mentioned above), and the Milliman Medical Index.

    The challenge for me was that these studies tended to provide present value lump sum needs to cover medical costs at retirement, as opposed to using a build-up method. Like, for example, at Fidelity where the study will total cost for a 65 year old couple.

    In addition, some studies did not have employer-sponsored retiree health care coverage or expenses associated with Medicare Part B, D or Medigap.

    So, what I decided to do was create a calculator (excel) whereby each planner could calculate annual medical costs using a build-up method, from employment-based health care, Part B, Medigap, Part D, dental, hearing, vision, and out-of-pocket expenses.

    This way, if a client had employer-based health care, then the planner didn’t have to assume a Medigap policy of Part D premiums. The calculator allowed you to be flexible. However, I also created a default for out-of-pocket, dental, hearing and vision (based on research) and a low, average and high cost assumption for Part B, D and Medigap.

    Now, the next step for me was to see if my annual assumptions were correct based on the lump-sum research available (EBRI and Fidelity). To do that, I ran a present value calculation of those annul costs. To my surprise (and relief), those numbers were very much in line with the research. I used a 6% health care cost inflation rate.

    Using the calculator, assuming no employment –based health care, with Part B (income at 100,000), Medigap F (average cost), Part D (average cost), dental hearing and vision, and out-of pocket expenses, the annual cost was around $12,000 for a couple age 65.

    Now, I did all the work and analysis for a number that most people use anyway just as a rule of thumb! But at least I knew I was in the ballpark. The calculator though allows planners to be more flexible and individual when modeling the expense with their clients.

    I agree that 6% inflation may seem high, but that has been the average. Studies show that it should decrease to about 5% in 20 years.

    Hope this helps.

    • Cindi Turoski

      Wow, that does help very much, Doug! We want to work from assumptions that are reasonable and justifiable – not just out of the air where we can’t support it. We hope our analysis is as realistic for a client as possible without forcing them to be too frugal or to underestimate things either (like Michael said too).

  • Pingback: kitces.com | Nerd's Eye View()

One Tweet / Trackback

  • Tuesday, February 28th, 2012
    Estimating retirement expenses over the entire duration of a client's retirement years is a fundamental part of retirement planning. Yet ... kitces.com | Nerd's Eye View

Michael E. Kitces

I write about financial planning strategies and practice management ideas, and have created several businesses to help people implement them.

For ConsumersFor Advisors

Blog Updates by Email

Nerd’s Eye View Praise

@MichaelKitces Twitter

Out and About

Wednesday, October 7th, 2015

*CPWA Training on Retirement Planning @ IMCA

Wednesday, October 7th, 2015

*Cutting Edge Tax Planning Developments & Opportunities *Social Security Benefits Planning for Couples *Life & Death Tax Planning for Annuities @ Dupage County Estate Planning Council

Tuesday, October 13th, 2015

*Understanding Longevity Annuities and their Potential Role in Retirement *Generating Tax Alpha with Effective Asset Location @ FPA Southern Wisconsin