Out-of-Pocket Health Care Expenditures at the End of Life

Featured in print Bulletin on Aging & Health

While virtually all Americans age 65 and above are covered by Medicare, they may nonetheless face significant out-of-pocket health care expenditures. Medicare includes numerous deductibles and copayments, and many people do not have a "medigap" policy to cover these costs. Home health care and nursing home expenditures must be paid for out-of-pocket, unless individuals are poor enough to qualify for Medicaid or have purchased an expensive long-term care insurance policy. Even those with generous insurance coverage may face the costs of specialized food or medical equipment to manage their health conditions.

As health generally declines with age, these out-of-pocket expenditures may be particularly high towards the end of life. High out-of-pocket medical expenditures may strain the finances of older households, for example by draining a couple's assets when one spouse gets sick and leaving little for the other spouse later in life. The risk of high expenditures may also be an important factor in household decisions regarding how much to save and how quickly to spend assets in retirement.

In The Risk of Out-of-Pocket Health Care Expenditures at End of Life (NBER Working Paper 16170), researchers Samuel Marshall, Kathleen McGarry, and Jonathan Skinner estimate the magnitude of these expenditures. They focus not only on the average amount but also on the distribution of expenditures, to better understand the risks that older households face.

The authors use the Health and Retirement Study (HRS) for their analysis, a data set which includes detailed information on the health care expenditures of older individuals. Notably, the HRS conducts an "exit interview" after the death of a survey respondent in order to collect data on the period up to the person's death from a surviving spouse or other family member.

Despite the rich data available in the HRS, the authors nonetheless face a number of challenges in estimating these costs. First, individuals may not answer questions about some expenditure categories or may only be able to provide a range rather than an exact amount. Second, as the survey is administered at fixed two-year intervals but respondents may die at any time, information on spending since the last survey will cover different periods of time for different individuals. Third, some respondents report expenditure amounts that appear to be implausibly large. The authors may overstate the risk of medical expenditures if they use this data but may understate the risk if they do not. Much of their analysis focuses on their strategies for dealing with these challenges.

The authors find that out-of-pocket expenditures toward the end of life are large - the average amount is $11,618 in the last year of life. There is also considerable variation in end-of-life expenditures, with ten per-cent of respondents spending more than $29,000 per year and the top one percent of respondents spending more than $94,000 per year. These values are large relative to the non-housing wealth of a typical older household, which is about $25,000.

The authors find that the largest single category of spending is nursing home and hospital expenditures, which average $4,731 in the last year of life; roughly two-thirds of that amount is for nursing home care. Here too the distribution is skewed, with the top one percent spending more than $75,000 per year in this category. Other big expenditure categories include insurance ($1,746 on average), prescription drugs ($1,496), home health care and helpers ($1,966) and spending to make houses accessible ($721).

The authors also explore whether out-of-pocket health care expenditures vary with the respondent's income or wealth. They find that they do, particularly with respect to wealth - the richest 20 percent of households spend an average of $18,232, versus $7,173 for the poorest 20 percent of households. These differences appear to be driven mostly by greater spending for nursing homes, home health care, and items related to making houses accessible. The greater spending by wealthier households in these categories may reflect their ability to "buy" both the ability to stay at home longer and more comfortable living arrangements once they move to nursing home care, but could also indicate that low-wealth individuals do not receive needed health care.

In short, the authors find that out-of-pocket health care expenditures "represent a numerically large and potentially important drain on financial resources, particularly for households as time of death nears." They note that their results are "thus consistent with the view that many elderly may continue to hold (or even accumulate) wealth as a hedge against uninsured costs surrounding expensive end of life care-giving for themselves or their spouse."

The authors acknowledge financial support from the NIA through grants number R01 AG16593 and P01 AG19783 and from the U.S. Social Security Administration through grant #10-P 98363-1-05 to the NBER as part of the SSA Retirement Research Consortium.