Top 20% of Employees

[imchipbrown]

What are you doing with "fractional people"? I've got 8 employees. 20% of them is 1.6 people. Do I have 1 or two people making over $80k and Top 20%?

[Dan]

You have two people. This is a situation where you always round up.

[Guest ndt123]

The regs do not require you to round up. Whatever you do must be "reasonable, nondiscriminatory and uniformly and consistently applied."

[Dan]

If you did not round up the top 20%, in this situation would you have the top 12.5% instead of the top 20%?

[Disco Stu]

Yes, and if you round up, you're using the top 25%. At seminar I attended, the speaker told us to always truncate a fractional person. The reason being that if you ever rounded up, you'd be using greater than 20%. Frankly, that makes sense to me, but absent any specific guidance to the contrary, I share NDT123's opinion.

[Wessex]

Although the prior message correctly quotes the regulation, I would be reluctant to round down to one person, particularly if the applicable tests would fail if you used two HCEs. I'm not sure that in the one-or-two person context that it is reasonable to round down. It's diffult to say what size number would be reasonable, although rounding down to ninety-nine, for example, probably would be.

If there is a history of using only one person and the plan would have passed discrimination tests for those prior years even if two HCEs had been used, I would be much less concerned.

[Tom Poje]

"any reasonable method of rounding or tie-breaking is permitted as long as it is uniformly and consistently applied"

1.414(q)-1t-Q3 2b

in other words, if you have rounded down for 3 straight years, don't all of a sudden round up (to pass) just cuz you fail if you round down. That is not consistent.

also, on your body count, remember the guidelines are minimums. if you want to count everybody (e.g. rather than exclude those with less than 6 months) you can

[imchipbrown]

Thank you all for the input. I still pass the tests with 2 HCEs, BTW. Liked Tom's comment about consistancy.

Also, to add some fuel to this dying fire (and as I've mentioned elsewhere), IRS has no problem rounding down their COLAs!

Then again, those regs are specific.

[MWeddell]

The truncating or "rounding down" of the COLA figures was (1) determined by Congress (not the IRS) and set forth in the Code itself, and (2) designed to raise revenue not to be mathematically pure. Hence, it's not a very applicable precedent if we're ever in a position of trying to convince the IRS that truncating is a reasonable interpretation in a completely different context.