LTC Bullet:  Why Don't Consumers Insure Against LTC Expenses?
LTC Embed Report # 9

Friday, October 7, 2011

Washington, DC--

LTC Comment:  Can you think of a more timely or important question than the one in today's title?  It's critical to our DC project and to everyone who ever will need, provide or pay for long-term care.


LTC Comment:  Why Don't Consumers Insure Against LTC Expenses?  That's the question I was asked to address on a webinar yesterday sponsored by the Society of Actuaries' LTC insurance "Think Tank."

My job was to respond to research findings presented by Dr. Gopi Shah Goda, Ph.D., an economist with the Stanford Institute for Economic Policy Research.  Samples of Dr. Goda's published work can be found here and here.

In a nutshell, Dr. Goda's research has led her to conclude that people fail to purchase long-term care insurance for many reasons, but the expectation that government will pay for such care is not among the most important of those reasons.

She also concludes from her research that tax deductibility for LTC insurance does not save Medicaid more than the cost of the subsidy because people incentivized to purchase LTC insurance by tax deductions tend to be more affluent and less likely to "spend down" to Medicaid anyhow.

I don't have access to Dr. Goda's presentation, and some of the findings on which she reported and I replied yesterday, have not yet been published.  So, to be fair, please withhold any conclusions about her work until you can review it in full and in context.

For now, all I can offer is my explanation of why I believe mistaken assumptions about Medicaid LTC eligibility rules account for the common belief that Medicaid does not significantly discourage the purchase of LTC insurance. 

I also observe that if Medicaid really did require catastrophic asset spend down, then people would buy LTC insurance in much greater numbers and tax deductibility would save Medicaid considerable sums.


Following are my 20-minute opening remarks for yesterday's webinar:

Let me begin by thanking the Society of Actuaries' "Think Tank" for inviting me to participate in this important conversation.

Over the years, I've read the research on who buys long-term care insurance, who doesn't, and why with great interest.

I've followed the debates and industry appeals concerning tax deductibility for the private LTC insurance product closely as well.

I confess I've been puzzled by the research findings.  People tend not to mention Medicaid as a reason for failing to purchase LTC insurance?  Tax deductibility appears to have a relatively small impact on the likelihood people will insure for long-term care?  Peculiar.

How can it be that a government program which pays for most expensive long-term care in the United States does not discourage people from buying private insurance?  Why wouldn't offering a tax credit or deduction create a major incentive to insure?

I've puzzled over these questions for a long time, so I appreciate this opportunity to learn from experts in the field as well as to think through, firm up and articulate my own views.

I do not have the social science credentials of Dr. Goda.  I am duly impressed with her expertise and both the quality and clarity of her work.  Far be it from me to challenge the technical aspects of her research or the accuracy of her findings.

So let me stipulate at the outset that given the assumptions on which her work is based, I accept her findings and analysis without question.  It is with her assumptions about Medicaid LTC eligibility and spend down, however, that I will take some issue.  That's where my expertise lies. 

It seems to me that if different assumptions were used with the same analytical model, the findings would be very different, would clarify more definitively how Medicaid crowds out the LTC insurance market, and would indicate how tax deductibility could save Medicaid considerably more money than otherwise expected.

I'll begin by citing a couple examples of assumptions I'd question in the articles published by Dr. Goda which I reviewed to prepare for this discussion.  Later I'll explain why I think those assumptions are incorrect.

Example #1 regarding Medicaid savings from tax deductibility of LTCI:

On page 744 of her article in the Journal of Public Economics, Dr. Goda writes:  "The results indicate that the average tax subsidy raises coverage rates by 2.7 percentage points, or 28%. However, the response is concentrated among high income and asset-rich individuals, populations with low probabilities of relying on Medicaid.  Simulations suggest each dollar of state tax expenditure produces approximately $0.84 in Medicaid savings, over half of which funnels to the federal government."  (Emphasis added.)

She continues on page 755:  "To have a larger effect on the allocation of long-term care financing, tax incentives would need to increase private insurance coverage for those who are at higher risk of spending down to Medicaid eligibility."

Observation #1:  What if "high income and asset-rich individuals" are not "populations with low probabilities of relying on Medicaid"?  What if income rarely interferes with Medicaid LTC eligibility?  What if asset-rich individuals can qualify easily for Medicaid LTC even without consulting Medicaid planning attorneys?  Wouldn't that change everything? 

Example #2 regards whether Medicaid crowds out demand for private LTC insurance: 

In an article titled "Why Don't Retirees Insure Against Long-Term Care Expenses? Evidence from Survey Responses," Dr. Goda writes (with co-authors Jeffrey R. Brown and Kathleen McGarry):

"[B]eliefs about Medicare and Medicaid do not appear to be systematically related to rates of long-term care insurance ownership. . . .  These results suggest that beliefs about Medicare and Medicaid are not a large factor in ownership decisions . . .." p. 18

Observation #2:  I think the mistaken assumption here is that asking people about Medicare and Medicaid will elicit useful answers to the question of why they don't buy LTC insurance.  Most people do not know who pays for long-term care.  When asked, they guess Medicare.  The more important question for us to ask is:  If long-term care is such a big risk and high cost, why don't people know more about who pays for it?  That's the issue I'll try to elucidate.

All right, we have two questions before us.  Why doesn't tax deductibility for LTC insurance generate larger savings for Medicaid? and Why don't more people buy LTC insurance?  Dr. Goda's answers to both questions are premised on the assumption that people with wealth have to spend down their assets before qualifying for Medicaid.  She concludes that tax deductibility for LTC insurance saves Medicaid relatively little because people most likely to buy LTC insurance are the least likely, because of their higher income and assets to qualify for Medicaid.  People don't buy LTC insurance for a lot of reasons Dr. Goda eloquently analyzes but evidently, according to her findings, not because they are planning to get Medicaid to pay. 

Medicaid LTC Eligibility

But let's consider some alternative explanations for the observed phenomena.  I've spent nearly 30 years studying how Medicaid long-term care eligibility actually works, as opposed to the conventional wisdom displayed in both the popular and academic media, that qualifying for Medicaid LTC benefits requires low income and catastrophic asset spend down.  I want to suggest that Medicaid eligibility is actually quite easy to obtain without significant asset spend down for most Americans and that, therefore, most people don't worry about long-term care until they need it.  After that, they qualify easily for Medicaid. 

Consequently, under the current system tax deductibility has relatively little influence on the public's likelihood to purchase LTC insurance.  Neither does planning to rely on Medicaid significantly influence their decision to purchase or not.  If Medicaid did operate as most commentators and academic writers assume it does . . . if Medicaid did require catastrophic asset spend down and if long-term care frequently resulted in devastating financial consequences for families, then people would buy LTC insurance out of concern for those consequences and tax deductibility would translate into major savings for Medicaid.  All I want to suggest today, is that Dr. Goda and her colleagues try applying their analytical model with a set of assumptions that more closely reflect actual Medicaid eligibility rules and practices.

What are those rules and practices?  Assuming an individual is over the age of 65 and has a long-term care level of medical need, he or she must qualify based on income and asset means tests.  So let's look at those tests and to what extent they restrict access to Medicaid LTC benefits.

Medicaid Eligibility Means Tests


Despite the frequently stated claim that Medicaid LTC eligibility requires "low income," income is almost never an obstacle to qualification.  In a 2010 study of Medicaid and LTC financing that I conducted in Rhode Island, that state's LTC policy specialist told me that in his more than three-decade tenure with the program, Medicaid had only disqualified two people for LTC benefits based on excess income.  How can that be?

Income eligibility for Medicaid is determined in two different ways.  Most states (around 35) have "Medically Needy" income eligibility systems.  They deduct medical expenses, including the cost of nursing home care, from applicants' incomes before determining whether or not their remaining income is low enough, based on some percentage of the poverty level, to qualify.  Thus, people with large incomes who have comparably large medical and LTC expenses, qualify easily in such states.  One does not need "low income" to qualify for Medicaid LTC benefits, only a "cash flow" problem after paying for medical and LTC costs not covered by Medicare.

Other states have "income cap" eligibility systems.  In those states, people with over 300 percent of the Supplemental Security Income level (currently $2,022), do not qualify for LTC benefits.  But ever since the Omnibus Budget Reconciliation Act of 1993, such individuals can divert their excess income into a Miller Income Diversion Trust in order to qualify.  For all intents and purposes, anyone with income below the cost of a nursing home, which averages around $75,000 per year, qualifies for Medicaid LTC benefits in either a "Medically Needy" or "Income Cap" eligibility system.

Dr. Goda writes in a May 2010 "Policy Brief":  "I find evidence that although tax subsidies raised insurance coverage by 30 percent, they did so mostly among wealthy and high-income populations who were unlikely to rely on Medicaid in the first place. Therefore, tax subsidies are unlikely to reduce Medicaid expenditures for long term care by more than the cost of providing the subsidy." (SIEPR Policy Brief, pps. 1-2)

Query:  If high-income populations are not actually excluded from Medicaid LTC eligibility based on income, how would this fact affect the conclusion that tax subsidies are unlikely to reduce Medicaid expenditures more than the cost of the subsidy? 

The Asset Test

Does Medicaid really require catastrophic asset spend down as a condition of eligibility for long-term care benefits?  Hardly. 

First of all, there is no requirement under Medicaid that people spend down assets for long-term care.  As long as one does not give away assets for less than fair market value for the purpose of qualifying for Medicaid, it is perfectly acceptable to purchase anything for value.  Applicants are often advised by attorneys, but also by state Medicaid eligibility workers, to reduce their countable resources by purchasing exempt assets such as home improvements, furniture, a more expensive automobile, or prepaid burial plans.  Sometimes such recommendations extend to purchasing world cruises or throwing big parties.  All that matters is that the buyer get value and the goods purchased are either exempt or have been consumed already by the time of the Medicaid application.

In addition to the $2,000 in cash or resources easily convertible to cash that Medicaid recipients may retain ($13,800 in New York), virtually unlimited additional resources are exempt from the asset limit test.  I've provided hyperlinks to federal regulations governing these exemptions in several of my state level reports.  See especially "Medi-Cal Long-Term Care:  Safety Net or Hammock?" at or on the Pacific Research Institute's website.

A home and all contiguous property is exempt up to at least $500,000 in equity value ($750,000 in New York, California, and a few other states.)

The following resources are exempt without any limit:

One business including the capital and cash flow; household goods and personal belongings; one automobile; prepaid burial plans for the Medicaid recipient and immediate family members; term life insurance; and individual retirement accounts as long as the Medicaid recipient is receiving periodic interest and principal payments. 

Over and above these generous asset limits, married couples can preserve additional income and assets for the healthy spouse in the community.  The Medicare Catastrophic Coverage Act of 1988 set these "spousal impoverishment" protections at up to $1,500 per month of income and $60,000 in assets.  They've increased with inflation to $2,739 per month of income and $109,560 of assets.

The vast majority of elderly Americans who need long-term care qualify easily based on these basic eligibility criteria.  Some people who are even more wealthy may consult Medicaid planning attorneys to self-impoverish artificially in order to qualify.  When GAO studied only one Medicaid planning technique--asset transfers--it concluded the cost to Medicaid was about $1 billion per year.  Hardly insignificant.  But asset transfers are only one method of Medicaid planning and not the most important one at that.  What might the cost be of commoner techniques such as special trusts, Medicaid friendly annuities, life-care contracts, reverse half-a-loaf strategies, life estates and purchasing exempt assets?  I'm not aware of any studies that estimate the actual impact on Medicaid of the full range of Medicaid planning methods.

Nevertheless, I believe this practice of intentional self-impoverishment, although very costly to Medicaid and destructive of Medicaid's mandate to protect the indigent, is only the tip of the iceberg.  The larger problem is that most people qualify without having to employ any sophisticated legal techniques. 


To conclude, I'll point out that according to data on sources of LTC financing cited in Dr. Goda's "SIEPR Policy Brief" article, Medicaid covers 48.5 percent; Medicare, 22 percent; and 8 percent comes from private insurance.  Out of pocket costs are only 21.5 percent of LTC financing.  It is important to understand that roughly half of those "out-of-pocket" costs for long-term care are actually contributions to their cost of care from their Social Security benefits of people already on Medicaid.  In other words, they represent "spend down" of income from another highly vulnerable government program.  They are not spend down of personal savings.  When you add up all the direct and indirect government payments for long-term care, you've accounted for over 80 percent of the total cost.  Add in private insurance payments and you're nearly up to 90 percent.  Thus, at most, only 10 or 12 percent of long-term care costs could come from asset spend down.

Bottom line, I want to suggest that the vast majority of all long-term care expenses are paid by government directly or indirectly.  Most of the remainder of such expenses are paid by third parties or they come from income, not assets.  Despite the conventional wisdom that life savings are at risk for catastrophic long-term care expenses, the reality is very different.  If people really did have to spend down savings, including most of their currently exempt resources, before qualifying for Medicaid . . .

  • They would be much more likely to purchase LTC insurance
  • Incentives to purchase LTC insurance such as tax deductions or credits would be much more effective
  • The availability of Medicaid after total asset spend down, whether consumers knew about it or not, would have much less impact on their decision to buy or not buy LTC insurance.  Who wants Medicaid if it doesn't save your estate?
  • Medicaid expenditures would plummet and the program would have more resources to expend for the care of truly indigent people.

Would I then advocate forcing people into total impoverishment before Medicaid helps?  Of course not.  All we need to do is implement some moderate and sensible reforms.

The most important such reform is to reduce Medicaid's home equity exemption from $500,000--13 times the level of asset protection allowed in England's socialized health care system--to something closer to what England allows or $38,000. 

If most home equity were at risk to fund long-term care, people would tap the equity in their homes to fund home and community-based care.  If they could no longer live at home, they could sell to family members or others to fund their institutional care.  The point is their need to rely on Medicaid would be delayed or completely eliminated.

Once home equity, the largest repository of seniors' wealth in the United States, is at risk for long-term care, families will pull together to find ways to protect their parents and their inheritances from LTC risk instead of fighting over the savings from putting a parent on Medicaid.

Unless and until Medicaid returns to its original purpose of providing a long-term care safety net for people in need, no amount of education, tax incentives, or LTC insurance salesmanship will convince many more people to buy protection they can get at a radical discount just by ignoring the problem.

Thank you.