Alimony Fixed as Child Support

The incentives are strong to use alimony in divorce for what is really support of children. Congress doesn't like that. The result is a tangled system of principles that can sting an unwary taxpayer.

The test for alimony fixed as child support is actually three separate tests. Let's talk about them in increasing order of complexity:

• You can't call it child support
• You can't change it based on something that happens to a child
• You can't change it based on something that's associated with something that happens to a child

The first two are fairly straightforward. The third gets funky fast, because of something I call the "Weird and Wonderful" test.

You can't call it child support

When the divorce decree or separation agreement identifies a specific amount of continuing support as child support, the amount so designated will not be treated as alimony. Payments can be characterized as child support even though they are for the support of an adult child.

You can't change it based on something that happens to a child

A payment will be treated as child support to the extent it is subject to reduction on the occurrence of a specified contingency relating to the child. For example, you can't get alimony treatment for a payment that stops "when Julie gets a job," or gets reduced "when Salikka graduates from college." Other things we know you can't base it on include a child's marrying, dying, leaving home, leaving school, or reaching a certain income.

You can't change it based on something that's associated with something that happens to a child

The statutory language in the Internal Revenue Code (Section 71(c)(2)(B)) could arguably be interpreted to mean that any reduction in alimony that might fall anywhere near any date of significance related to a child could be enough to change the tax treatment. Fortunately, however, the regulations break the issue down to two tests, both of which can be calculated with certainty. Of the two tests, the first is simple to explain and apply. The second is comically complex.

Test #1 - the Birthday Test

Alimony cannot be reduced within six months of the 18th or 21st birthday of a child, or the local age of majority.

Test #2 - the "Weird and Wonderful" Test

The second test comes into play only if there are at least two children and at least two reduction dates. It says that alimony can't be reduced on two or more occasions that occur within a year before or after two or more children attain a particular age between 18 and 24 years. The measuring age must be the same for each child, but it doesn't have to be one of whole years. See what I mean? Cute, huh?

I think there are two strategies you can take to deal with the Weird and Wonderful test:

1. The first is to digest it, understand it, and make sure you think through how your proposed reductions of alimony will interact with it. Being an idiot and a glutton for punishment, and having this inexplicable yearning to bury myself in IRS regulations, I've done this. I've committed the Weird and Wonderful test to a spreadsheet and can test for it quickly and easily.
2. The second is simply to resolve not to reduce alimony on more than one date, and if you ever do, call some idiot who's chosen number 1.

If a reduction satisfies one or both of the tests, the payment will be rebuttably presumed to be child support to the extent the reduction coincides with the contingency related to the child. Rebutting the presumption requires a showing (either by the taxpayer or by the IRS) that the date of the reduction is set independently of a contingency related to a child. In short, don't count on it.