ADP REFUNDS based on "bad" census data for 2015

[Pammie57]

A client provided census data on their plan back in February. It turns out that they included some "historical" data (overstated some compensation and deferrals). This of course, affects the ADP test results, and it changes the refunds that should have been made to the HCE's.

One HCE's refund is only overstated by $1.94, but the other one was paid $989 too much in corrective refunds. At this point in 2016, what are the options to get this corrected?

[hr for me]

Step back and look at the overall totals versus individual....What is the net change of refunds? I was able to argue successfully (albeit years ago) that the difference was de minimus overall and didn't have to re-do refunds since in effect the refunds to the group of HCEs were enough to make the plan nondiscriminatory. Not sure that would work now. We did have to file under a correction program.

In essence we were just going to be moving around who got the refunds rather than not enough refunds overall.

And it was our (the TPA/recordkeeper's mistake) so we ate the full cost of the re-testing and the IRS filing. We had a very inexperienced young employee doing the test and the person who checked him was much further up the corporate ladder and hadn't done a nondiscrim test in decades. It was a mess. I got to be the one to clean it up. While challenging, I wouldn't recommend it!

[BG5150]

I would look to EPCRS and the overpayment sections