EBRI Criticizes HSAs (Again)

In a widely publicized April 2010 Note, the Employee Benefit Research Institute (EBRI) deploys confusing calculations based on questionable estimates to demonstrate that Health Savings Accounts (HSAs) are inadequate vehicles for saving for health care costs in old age. Its trail of endnotes eventually leads to an estimate of the present value of lifetime Medicare benefits for a husband and wife that is attributed to a “personal communication.”

The Obama Administration is now writing regulations that could eliminate health plans that use HSAs to lower premiums, control costs, and improve health care. Given the explosive growth in consumer directed health insurance plans prior to the passage of the ObamaCare legislation (an estimated 23 million people now have them), eliminating HSA plans will be easier to do if government officials and the public can be convinced that they are an unimportant innovation.

In this context, EBRI concludes that

“Health savings accounts (HSAs) are often touted as a vehicle for funding future retiree health care costs. However, statutory contribution limits mean that they are unlikely to plan more than a minor part in savings for health care costs in retirement.”

How minor is minor? EBRI says that an average man could accumulate “just” 16% of his future health care costs by contributing the maximum amount to his HSA in the ten years from 55 to 65. This is considered “minor” even though conventional health insurance, and the type of policy envisioned by those who wrote ObamaCare, would result in no savings at retirement at all.

To calculate the amounts available at retirement, EBRI assumed that someone aged 55 in 2009 put $4,000 a year into his health savings account (HSA), maximum payments plus catch-up payments, for just 10 years at interest rates of 2 or 5 percent. At 2 percent, EBRI states that he would have $48,300 available for retirement medical costs. At 5 percent he would have $55,100 dollars available for medical costs. Yet limiting saving to only 10 years biases the calculation from the outset.

Assuming no inflation, someone putting $250 into a HSA each month (which yields the maximum contribution of $3,000 a year) over a working life of 40 years at a 2% interest rate compounded monthly would have $183,600 to pay for medical expenses at age 65. This amount would be higher if “catch-up” contributions were included.

Whether the EBRI claims about health savings accounts are reasonable depends in part on whether its estimates of retirement health spending are reasonable. Three other methods of estimating the amount needed at retirement are discussed below. All three yield estimates that are much less than EBRI’s.

A 2004 paper by Alemayehu and Warner used a combination of Medicare and commercial claims data to estimate per capita lifetime health expenditures in the US at $268,700 for men and $361,200 for women in 2000 dollars. Nearly half of these expenditures were incurred after age 65. The estimated average expenditure after age 65 for those who survived that long was $188,658. People who survived to age 85 were estimated to spend 36 percent of their lifetime expenditures between age 85 and their death.

Far from being “minor,” the savings available after funding a health savings account for 40 years under conservative assumptions are close to the Alemayehu and Warner estimates. This suggests that HSA based insurance, given sensible reform, could come close to fully replacing Medicare entitlement expenditures in 40 years even if one does not include the effects of the 10 to 20 percent expenditure reductions that have been documented to occur when health savings account policies replace standard third party payment arrangements.

Converted to 2005 dollars, the Alemayehu and Warner estimate of spending after 65 is $213,966. This amount it consistent with after 65 spending estimates derived from the Medicare Payment Advisory Commission’s estimate of total annual medical spending as reported in a Heritage Foundation report. In 2005, the Commission estimated that Medicare enrollees had total annual medical expenses of $12,157. Spending $12,157 a year for 30 years creates expenditures of $364,710 per person. But using 30 years of the Commission based estimate is an upper-bound estimate of the amount needed at retirement, because US life expectancy at 65 was just 18.7 years in 2008.  Spending $12,157 for 18.7 years creates expenditures of  $227,336, within about $13,000 of the Alemayehu and Warner estimate converted to 2005 dollars.

The present value of the lump sum of cash needed to fund 30 years of the Commission estimated payments at a 2 percent interest rate with annual compounding is $272,274, roughly $50,000 less than the EBRI estimate even though the EBRI estimate, as will be explained later, does not appear to include hospitalization. The present value of 18.7 years of payments is approximately $198,000, well within the range of value that can be amassed through 40 years of HSA contributions.

The savings needed at retirement can also be estimated by looking at the payments that would be needed to buy current health care coverage. In 2003, data from the Medical Expenditure Panel Survey suggested that the median out-of-pocket health expenditure for those 65 and over with Medicare and private insurance was $881 a year or $74 a month. The average expenditure among those with expenses was $129 a month. It is important to control for any additional insurance purchased by Medicare beneficiaries as spending totals vary between those who had no supplemental insurance, private Medigap insurance, or Medicaid. It is also important to recognize that the passage of prescription drug coverage in 2006 significantly reduced out-of-pocket expenditures and that 2003 out-of-pocket costs are likely to have been higher than those faced by Medicare beneficiaries after prescription drug coverage was added..

The estimates from the Medicare Payment Advisory Commission and Alemayehu and Warner include hospitalization, which Alemayehu and Warner estimated were about a third of lifetime spending. The EBRI estimates do not include payments for hospitalization. EBRI calculated the amount needed to have a 50 percent chance of being able to cover “premiums and out-of-pocket expenses for Medigap and Medicare Part D.”

Reducing the Commission and Alemayehu and Warner lifetime estimates by a third yields spending estimates in the $181,000 range.

Combining premium payments for average Medigap coverage and Medicare Parts B and D with average 2003 MEPS estimates of out-of-pocket expenses yields a monthly payment of slightly more than $450. In present value terms, a lump sum of $106,169 would be sufficient to support a payments stream of $450 a month for 25 years assuming a 2 percent interest rate with monthly compounding. A lump sum of $121,747 would be required to make 30 years of payments.

All three of the alternate methods for roughly estimating the amount of savings currently needed to provide health care in retirement produce estimates of the amount needed at retirement that are considerably lower than the $144,000-$290,000 estimated by EBRI, and, more importantly, are within range of the amounts that could be accumulated by regular contributions to a health savings account over a working lifetime.

Comments (4)

Trackback URL | Comments RSS Feed

  1. Ken says:

    The problem with EBRI is that they really believe in private sector socialism. When they are not advocating public sector socialism, that is.

  2. Larry C. says:

    Right now, HSA are the ONLY vehicle people have to save for post retirement care. What is EBRI’s problem?

  3. Bruce says:

    I agree with Larry. What’s the point?

  4. Joe S. says:

    EBRI has never liked consumer driven health care. I don’t believe they have ever had one favorable thing to say about it.