Surely No One Dies on Schedule: Split-Interest Gifts and Mortality Tables

Much of our work in planned giving – in particular, with split-interest gift arrangements – involves the use of mortality tables and the related estimates of life expectancy. When calculating the charitable income tax deduction for a split-interest gift, for example, we are required to use specific mortality tables dictated by the IRS. If we wish to estimate the remaining liability for a life-income gift, we must use some type of mortality data. And should we attempt to estimate the ultimate charitable remainder value for a life-income gift, the use of mortality data is critical.

But what are the different mortality tables, and when do we choose to use a specific one?

The biggest distinction to make at the outset is the difference between mortality tables based on United States government data and mortality tables based on insurance industry data. The national census is conducted every 10 years and attempts to count every person in the U.S. It is not perfect, but it is representative of the overall population. Insurance industry mortality tables, on the other hand, represent only a subsection of the total population, and they tend to be more conservative. By that, we mean longer life expectancies, which means charities should expect to wait longer for the remainders of split-interest gifts.

A couple of things to keep in mind on the government tables: U.S. Census data represents the entire population. There is no filter for those who are more affluent or for those who are more likely to be donors of charitable gifts. Moreover, the census-based mortality tables are gender-neutral: men and women have the same life expectancies. The most recent census-based mortality table is the 2010CM table, based on the 2010 census; the previous table was the 2000CM table, based on the 2000 census.

What does this mean for split-interest gifts? When we run an income tax charitable deduction calculation for a charitable gift annuity or a charitable remainder trust, the IRS mandates that we use a specific census-based mortality table. The 2000CM table was released in early 2009 and was required for determining charitable deductions for gifts made from July 1, 2009, through April 30, 2019. Beginning with gifts made on or after May 1, 2019, the IRS allowed use of either the 2000CM table or the 2010CM table. The IRS requires the 2010CM table to be used exclusively for gifts made on or after June 2, 2023.

The latest of those dates was established only recently, and there was, in fact, an unusually long transition period allowing for use of either the 2000CM table or the 2010CM table. Let us be clear: for all split-interest gifts made on May 1, 2019 through June 1, 2023, charitable deduction calculations may be computed with either table. But this begs the question, does it really matter which table is used? And if so, how are the results different?

For gifts made during this transitional period, the use of the 2000CM table typically results in a larger charitable deduction for all life income gifts than does use of the 2010CM table (all other things being equal). On the other hand, specifically for charitable gift annuities, the use of the 2000CM table typically results in a smaller amount of tax-free income than does use of 2010CM. A gift annuity donor who does not itemize deductions on their income tax return, or who is unable to use any additional income tax deductions, might prefer calculations that use 2010CM to maximize the tax-free portion of their payments.

The mortality tables created by the insurance industry are decidedly different from the census-based tables. Foremost, the tables are based on a select subset of the U.S. population – persons who purchase life insurance products. Second, the tables are gender-biased, resulting in different life expectancies for males and females of the same age.

These two differences can have a profound impact on calculations for life income gifts. Forget about the charitable deduction, because the IRS requires use of a census-based table for that calculation, but for estimates of remaining liability and estimates of ultimate remainder values, the differences can be dramatic. The two most popular insurance industry tables are the Annuity 2000 and 2012 IAR tables, but we can summarize by saying that liability estimates based on insurance industry tables tend to be much larger than ones computed using census-based tables. This is because insurance industry tables predict much longer life expectancies than the government tables discussed above – especially for older females, who typically represent the largest group of CGA and CRT annuitants and beneficiaries.

The 2012 IAR table tends to predict longer life expectancies than the Annuity 2000 table and, hence, results in greater estimated liabilities. In contrast, there is no expected pattern regarding which table will result in greater estimated remainder amounts. The outcome will depend on whether a gift’s principal is expected to grow or shrink over time.

Here is a table we put together showing the life expectancies for males and females at ages 70, 80, and 90, using the four mortality tables mentioned above:

A last point: state reserve calculations for gift annuities are also based on mortality tables, but here the planned giving professional has no choice. The states that mandate reserve calculations dictate which mortality tables must be used for each grouping of gifts. For gift annuities funded today, all the regulating states require use of the 2012 IAR table for computing reserve requirements.

We cannot avoid the reality that most split-interest gifts last for the lives of the beneficiaries. Typically, the charitable benefit is derived upon the passing of one or more persons. We use mortality tables in many ways, but the values we calculate using those tables, while precise for purposes of taking a charitable deduction or reporting a reserve amount, are merely estimates. Nobody dies according to a mortality table, and that is a good thing, all around!