Improper Classifications?

[Guest IBlameTedBenna]

I have seen where you can have classifications based on age and years of service. Is it permissible to have the following?

Group 1 - Owner (HCEs)

Group 2 - Employees making more than $50,000 and with more than 5 years of service (this could be a problem because some in this class may be considered HCEs as well)

Group 3 - Employees making less than $50,000 with less than 5 years of service

Thanks

[Tom Poje]

probably better would be

1. owners (not by attrubution)

2.employees not described above making more than 50,000....etc

3. etc.

remember, if you say owners, the kids are in.

but my understanding is that you can put people into whatever groups - it creates a definitely determinable allocation group.

[AndyH]

BTW, if you want to Blame Ted for something else, Blame Him for your probably disqualified plan. You've created an eligibility condition of 5 years of service - a 410(a) violation- for somebody making $50,001.

Everybody needs to be in at least one group.

Also, I wonder if the use of comp as a class could be challenged as being improperly within the discretion of the employer, i.e. not definitely determinable.

[himt4]

I ain't no expert, but sometimes I ask questions on this board and get an answer, so I feel I must try now and again to try to help with someone else's questions. Hopefully someone with a better understanding will fine tune what I am about to write:

I recently asked my technical support a similar question and my understanding of the answer is:

If ever during your testing you need to pass the "average benefits test" you will need to pass something called the Reasonable classification test {begin cut & paste}

"Section 1.410(b)-4(b) of the Regulations provides that a classification will be

reasonable if, based on all of the facts and circumstances, the classification is

reasonable and established under objective business criteria that identify the category

of employees who benefit under the plan. Reasonable classifications include specified

job categories, nature of compensation (i.e., salaried or hourly), geographic location,

and other similar bona fide business criteria."

{end cut & paste}

Since the classes are not safe harborish ones like the one's mentioned above (i.e. job type, salaried/hourly,geographical) then there is a question about whether you would pass this reasonable classification test. So, basically my tech support told me that if you are not using the safe harborish classes (i.e. job type, salaried/hourly,geographical) then you should file the Document with the IRS to get their approval and then you will be ok.

[Guest gdburns]

If you are not using a "safe harborish" classification, you might want to ask around to see if anyone has ever received approval for whatever classification you are thinking of using so as to save yourself the time and trouble etc of filing.

I always thought that you needed to have reasonable (acceptable) classification regardless of and before any other test.

[Tom Poje]

the reasonable classification test applies to coverage, so if plan failed ratio percent test, you could not proceed to avg ben test.

the IRS has confirmed at different conferences, etc, that you can indeed put people into classes for purposes of allocation. thus, unless you are excluding a class from getting something, the reasonable classification test should not come into play.

Technically an ee receiving a penny would be considered benefiting, I suspect there would be a point the IRS would say, it has to be meaningful, just like they have with the issue of usin short term employees to pass testing.

[himt4]

Let's say classes are based on comp - class1: earned "$1-$20,000", class2 "earned $20,000-$40,000", etc

if its a non top heavy plan, you can decide to give class2 $0. so, if by class2 not benfitting the ratio percentage is over 70%, then you would just need to pass the rate groups, and all is ok.

but if by giving class2 $0, what if the ratio percentage is less than 70% but more than the safe harbor percentage. Don't you then try to pass the average benefits test, and then wouldn't you have a potential problem with the reasonable classification test?

[AndyH]

That is right. If the creation of a class and giving them $0 has the same effect as if the class were excluded from the plan by definition then IRS reps have opined that you have an unreasonable classification for purposes of coverage; in such case you would need to pass the ratio/percentage test for coverage testing.

So, to summarize, the eligibility definition does not need to be reasonable provided that ratio/percentage is satisfied for coverage testing.

Allocation groups do not need to meet the reasonability standard so long as either (a) no class gets nothing or (b) the plan passes coverage by using the ratio/percentage test.

Much of this is not clearly written anywhere (actually unless Tom put it in his book-I haven't checked there); it is the interpretation that IRS reps have expressed and generally accepted practice.

p.s. which is why requesting a determination letter is a good idea in a situation where you push the envelope a bit

[Guest gdburns]

But can a class be based on $ rather than salary/hourly ?

[Tom Poje]

The IRS has gone as far as to say you could actually have each person in his/her own class, thus net effect would be, yes you could define classes to chieve the same effect as being by $ or whatever.

(Note: all this was actually possible years ago, but you would have had to set up multiple plans and exclude people from the plan and then aggregate the plans for testing - so the IRS has simply made things easier on us)

now, all that being said, if the result by naming classes by $ would be to provide 'short term' employees who have small comp a large % of pay vs others could very well be a problem. The IRS has made that clear. (in some ways this would be no different than using a bottom up QNEC to satisfy the ADP test)

[Guest gdburns]

Tom

Where can I find this? If it is a case of one of those anonymous or unofficial statements, I would still like to know When and Where, if possible. I will track it from there. I need it for a specific important purpose.

[Tom Poje]

well the one cite I used in the Coverage and Nondiscrimination Answer Book is Q3 from the 2001 ABA Joint Committee on Employee Benefits. This might be listed as ABA (American Bar Association). I know you can find the Q and As on their website.

I am sure there have been other sources that have said the same thing, this is simply the one can easily reference.

[Guest Partly Cloudy]

Andy,

So, to summarize, the eligibility definition does not need to be reasonable provided that ratio/percentage is satisfied for coverage testing.

Are you saying then that the 5 year service requirement mentioned in the earlier post would be okay if the plan passed the ratio percentage test?

I have a situation that would be ideal if the plan could define eligible employees as (1) owners, and (2) non-owners with 5+ years of service at the beg. of the plan year.

That is like having a 5 year eligibility requirement for non-owners. (There are some HCE's in that non-owner group.) The plan would be aggregated with a DC plan and together they pass the ratio % test.

What do you think? I think it is a 401(a) violation and I need to find another way to define that group of employees with 5+ YOS. I would really like to hear otherwise because the class defined above works perfectly for the owners goals to reward long service employees (and manage total costs).

Thanks.

[Tom Poje]

Take a look at some of the examples in 1.401(a)(4)-8

the general discussion there is avoiding the gateway minimum, but the examples are clear you can put people into groups based on age / service.

I would expect if you go to the extreme you indicate, the plan would be top heavy, thus you would have to provide a 3% minimum, which would then become 5% gateway. even if plan is not top heavy, you would have a bunch of zeroes on the test, making it harder to pass testing.

remember, we are not talking eligibility - we are talking allocation groups. the people still come into the plan after 1 year. but it sounds like you are saying you want to give the group with < 5 years a 0% contribution. If plan fails ratio percent test, you can't use the avg ben test to pass coverage.

odd how you indicate you want to reward long term employees, but if you give zippo to people for 5 years I doubt many would stay anyway so you never retain possible good help. You might as well go with the 5 year cliff vesting, and that would reward people who stayed for 5 years.

I would also add that one of the pension bills would require minimum schedules of 2/20 or 3 year cliff.

[Guest Partly Cloudy]

Tom,

Thanks for your response.

So if we let everyone in, but, the "< 5 people" get a $0 benefit accrual (this is a DB plan), isn't that the same as having no group at all in the eyes of the IRS? So are we back to having a 5 year eligibility requirement?

Also, good point with repsoect to the reward issue, but the whole picture is that the DC plan is the "primary" plan for the non-owners. They get a DC contribution of 5% if they have less than 5 years and a larger contribution if they have greater than 5 years. The very small DB benefit is just a kicker.

[AndyH]

PC, I said 410(a) NOT 401(a). Section 410(a) sets eligibility standards. It lays out diferent requirements. Read the examples. If 410(a) is violated, how you propose to pass 410(b) or 401(a)(4) are irrelevant.

Disguising a violation of 410(a) with a 0% allocation group is transparent.

[Guest Partly Cloudy]

Typo, sorry.