The Dependent Care FSA 55% Average Benefits Test

Question: How should employers address the many challenges associated with the 55% average benefits test component of the dependent care FSA nondiscrimination testing?

Short Answer: The best practice approach is to test early, consider using the top-paid group election (top 20%) to avoid or mitigate a testing failure, and correct any failures by reducing highly compensated employees’ dependent care FSA elections to a passing level prior to the end of the plan year.

General Rule: Section 125 Cafeteria Plan Nondiscrimination Testing

Employees’ elections to pay the employee-share of the premium for health and welfare plan coverage on a pre-tax basis or make pre-tax FSA contributions are governed by Section 125 of the Internal Revenue Code. The Section 125 cafeteria plan rules include nondiscrimination testing (“NDT”) requirements designed to ensure that the plan does not discriminate in favor of highly compensated employees. 

There are multiple components of the cafeteria plan NDT and multiple definitions of which employees qualify as highly compensated.  This post focuses on the most (and only) commonly failed test, which is the 55% average benefits test (“ABT”) component of the dependent care FSA NDT under Section 129.

For more details:


The Dependent Care FSA 55% Average Benefits Test: Why It’s So Difficult to Pass

Although the ABT is just one component of a wide variety of cafeteria plan-related NDT, it is by far the most notorious.  This is because of the practical reality that virtually every other cafeteria plan-related NDT under Section 125 (POP), Section 105(h) (health FSA), or Section 129 (dependent care FSA) does not present a practical concern that the typical plan approach could result in a failed NDT.

In other words, the ABT is the only component of cafeteria plan NDT that employers have any significant chance of failing based on a quantitative analysis of utilization rates.  All other components are either based on a) structural plan design (as opposed to utilization rates) that can easily be addressed by making the plan available on a nondiscriminatory basis, or b) utilization rates that are essentially never in jeopardy of failing.  So, while employers still need to perform all of the other NDTs to confirm a passing result, there generally is no drama or suspense tied to the result.

The design of the ABT, on the other hand, conveys the sense that Congress intended for a large swath of employers to fail the test.  Employers await their ABT results back from their FSA TPA with bated breath, knowing that even a dependent care FSA made available on completely equal terms to both highly and non-highly compensated employees still has a high chance of failure based on utilization rates.

The reason the ABT is so difficult to pass is that the test requires at least 55% of the dependent care FSA benefits be for non-highly compensated employees (“NHCEs”).  The test looks to the average benefit received by highly compensated employees (“HCEs”) as a ratio compared to the average benefit received by NHCEs, requiring that the NHCE average be at least 55% of the HCE average.  HCEs tend to disproportionately participate at higher levels and make higher dependent care FSA elections, which is exactly what the ABT is designed to prevent.  That makes the ABT very hard to pass without adjustments to reduce HCE elections.


Who Qualifies as a HCE for Purposes of the Dependent Care FSA?

Individuals are considered highly compensated as an HCE for purposes of the dependent care FSA NDT if they are:

  1. A more-than-5% owner of the employer in the current or preceding plan year; or

  2. An employee who earned more than $135,000 (2023 testing) or $150,000 (2024 testing) in the prior plan year.

Employers will generally look to Box 1 of the Form W-2 to determine compensation levels for the prior year.  Employees hired mid-year in the prior year may have an annualized salary in excess of the applicable limit but still not qualify as an HCE because they did not earn over that limit in their initial partial year of employment.

The Section 125 NDT rules direct employers to use current-year compensation for new hires not employed in the prior year, but it is not clear whether this special current-year-hire rule applies to dependent care FSA testing.  Many FSA TPAs reasonably take that employers can simply exclude all current-year hires from HCE status unless they are more-than-5% owners.


Alternative HCE Determination Method: The Top-Paid Group Election (Top 20%)

Many employers face particularly difficult ABT prospects because a large amount of their population are employees who earn more than the applicable annual threshold.  These employers will have an unusually high number of HCEs, making it even more difficult to pass the ABT.  In other words, the more HCEs in the population, the more unlikely it is that 55% of the dependent care FSA benefits will accrue to the (smaller) NHCE population.

Fortunately, many such employers do have an alternative approach available through an option to use the “top-paid group election” that may alleviate this predicament.  The top-paid group election determines HCE status by reference to whether employees are in the top 20% highest-paid employees instead of the specific annual compensation level under §414(q) ($135,000 for 2023 testing). 

This can result in significantly fewer employees meeting HCE status because employees who earned $135,000 (2023 testing) in the prior year but are not in the top 20% highest paid employees (and are not more-than-5% owners) will be moved from HCE to NHCE status.  The more NHCEs in the employer’s population, the better the chance employers have to pass the ABT. 

Important Limitation to the Top-Paid Group (Top 20%) Election: Must Also Apply to 401(k) Plan

The top-paid group election cannot be used for the dependent care FSA NDT unless it is also applied to the employer’s other employee benefit plans that rely on the §414(q) HCE definition, which includes the employer’s retirement plan.  

Accordingly, employers seeking to utilize the top-paid group election must coordinate with their 401(k) plan nondiscrimination testing to confirm that the top-paid group election is also in place for the 401(k) plan for that plan year.

Action Item: Employers that fail the ABT pre-test should work with the FSA TPA to run the pre-test again utilizing the top-paid group election—provided that election is also in place for the 401(k) plan.  The top-paid group approach may move the plan from a failing to passing level.  However, even where the top-paid group election does not move the plan to a passing level (at or above 55%), it often will reduce the degree of the failure by moving the result closer to 55%.  The degree of the failure is still meaningful because the closer the test results are to 55%, the less severe the reductions needed to HCE elections to pass the ABT.


When to Test: The Advantages of Performing Nondiscrimination Testing Early

As a technical matter, the NDT results that matter are only those as of the last day of the plan year.  In other words, the cafeteria plan NDT rules look to the last day of the plan year to determine whether the plan passes.

However, it is best for employers to run a pre-test mid-year to determine whether the plan is on track to pass or fail as of the last day of the plan year.  For example, a mid-year pre-test might show the plan’s ABT is running at a failing result of 45%, and that the employer needs to reduce HCE elections by 20% to pass.  In that situation, the employer would need to cap HCEs who elected the full $5,000 at a reduced $4,000 limit.  Reducing HCE elections ensures that the plan will pass the ABT as of the last day of the plan year—which, again, is the key date for all the cafeteria plan NDT requirements.

The earlier the employer runs the cafeteria plan NDT pre-test, the more likely the employer will discover any ABT failure before the HCEs have reached the reduced limit needed to ensure a passing result.  It is far simpler to correct a failed ABT result before HCEs reach the reduced cap by stopping HCE contributions at the reduced amount indicated by the pre-test (as opposed to the alternative of having to correct excess contributions, as discussed below).  This is generally as easy as informing HCEs of the reduced limit and directing payroll to implement the reduced cap on HCE contributions.

Action Item: Engage with the FSA TPA early in the year to run a pre-test and determine the direction of the plan’s ABT results, preferably prior to the last quarter of the plan year.  For example, consider running the pre-test in summer for a calendar plan year.  This early testing approach will greatly simplify the process if corrections are needed.


Correcting an ABT Failure Where HCEs Have Already Exceeded the Reduced Limit

In some situations, employers will not discover an ABT failure in time to impose a reduced HCE contribution limit prior to HCEs contributing to the dependent care FSA in excess of that limit. 

For example, suppose the ABT pre-test results show that HCE elections must be reduced by 20%, resulting in HCEs who elected the $5,000 maximum having to drop to $4,000.  If those HCEs have already contributed $4,375, there is a $375 excess that must be made taxable income before the last day of the plan year.

There are two basic approaches to converting excess HCE dependent care FSA contributions to taxable income:

  1. Refund/Return: The employer can distribute the excess contributions back to the HCEs through payroll as taxable income subject to withholding and payroll taxes by the end of the year, thereby reducing the amount available in the HCEs’ dependent care FSA account balance.  Note that this approach will not work for HCEs that have already received reimbursement of the excess amount.

  2. Recharacterize: The employer can recharacterize the excess contributions as taxable income subject to withholding and payroll taxes without directly refunding the excess to HCEs.  The downsides of this approach are that the employer will need to a) take the withholding and payroll taxes from other income, and b) inform the HCEs that they may take a distribution of the excess contributions (which no longer have pre-tax status) from the FSA without the need to submit qualifying dependent care expenses.

With either approach, the employer will need to coordinate with the FSA TPA to ensure proper administration of the correction.  As always, the employer will need to take action before the end of the year to ensure a passing result as of the last day of the plan year.


Relying on Pre-Test Results and Corrections: The General Industry Standard

By running a mid-year pre-test and making any corrections needed to address failures, the employer can be confident that the result as of the last day of the plan year will reflect a passing grade. 

Although it is not entirely clear in official IRS guidance, the typical FSA TPA approach has essentially set the industry standard that if the IRS were to audit an employer’s NDT results, pre-test results showing a clear passing score—including any adjustments made to reduce HCE contributions—should always be sufficient.  If the IRS was not willing to rely on pre-test results (including, if applicable, any necessary corrections), the employer should still be able to perform an after-the-fact NDT as of the last day of the plan year to confirm the passing result.

Many FSA TPAs do not perform a second year-end NDT as of the last day of the plan year for this reason.  The additional NDT is viewed mostly as an unnecessary formality.

Employers that are concerned about demographic and participation changes potentially affecting the result after the pre-test is performed can consider a) closing the dependent care FSA to new HCE participation for the remainder of the year for new hires and mid-year permitted election change events, and/or b) correcting the pre-test results up to a level higher than 55%, such as 57.5% or 60% to provide additional buffer to serve as a margin for error.


Corrections Required by the Last Day of the Plan Year (December 31 for Calendar Plan Year)

Employers must correct any NDT failure, including ABT results below 55%, by the last day of the plan year.  Once the plan year closes, there is no option to correct because the plan will have failed as of the last day of the plan year—the only relevant date for NDT purposes.

For example, employers with a calendar plan year cannot wait until January’s Form W-2 preparation process to address any testing failure through an after-the-fact recharacterization process.  HCEs must have their pre-tax contributions reduced to the passing level through payroll as of December 31 to pass the ABT.

Where the employer fails to make corrections before the end of the year to result in a passing ABT as of the last day of the plan year, all the dependent care FSA contributions for all HCEs must be recharacterized as taxable income.  In other words, HCE elections must be reduced to “0” because the entire pre-tax contribution amount becomes taxable.  The Form W-2 would accordingly reflect “0” in Box 10 for dependent care benefits (because the full amount would have to be moved to standard Box 1, 3, 5 income).

For example, suppose the ABT pre-test results show that HCE elections must be reduced by 20%, resulting in HCEs who elected the $5,000 maximum having to drop to $4,000.  If the employer fails to correct HCE elections by the last day of the plan year, the HCEs will not receive any dependent care FSA benefit.  Those HCEs who elected $5,000 will not receive a $4,000 reduced benefit, but rather the complete elimination of their benefit down to zero (because the full amount will have to be recharacterized as taxable income).

Action Item: The ABT cannot be satisfied through corrections made after year-end.  Make sure all corrections required to HCE elections are implemented and processed through payroll prior to the last day of the plan year.  Employers must reduce HCE elections prior to the end of the plan year to preserve the reduced tax-advantaged benefit for all HCEs.

_ _

Strategies to Avoid ABT Failures: The Good, the Bad, and the Ugly

Employers clearly become frustrated with failing the ABT year after year, and in many cases they overcompensate with approaches that unnecessarily harm HCEs.  The following are a few of those strategies:

The Good: Offer an Employer Matching Contribution to Non-HCEs

The only approach to reducing the risk of testing failure that is recommended is to incentivize NHCEs to participate in greater numbers and to a greater degree.  Although it is uncommon because of the increased employer cost, a few employers have taken the smart approach of offering NHCEs additional employer contributions as a carrot to encourage greater participation.

One particularly effective approach is to offer an employer matching contribution to the dependent care FSA for NHCEs that operates in a similar manner as a 401(k) plan matching contribution.  For example, the employer might offer a dollar-for-dollar matching contribution for NHCEs of up to $500.  This will entice greater participation and higher elections from NHCEs, which will in turn significantly improve the employer’s chances of passing the ABT.

The Bad: Limit HCE Contributions to a Reduced Level from the Outset

A common approach that is not recommended is for employers to simply cap HCE elections from the beginning at a reduced level.  For example, HCEs may elect only up to $3,000 instead of the standard $5,000 limit available to NHCEs.

Although this approach does improve the chances of passing the ABT because it will decrease HCE dependent care FSA benefits by design, it does so in a very haphazard manner.  In that example, allowing HCEs full elections and pre-testing may have resulted in a required reduction of $5,000 HCE elections to $4,000 to reach a passing 55% level.  However, by capping HCE elections at $3,000, the employer would have effectively preemptively precluded HCEs from enjoying an additional $1,000 pre-tax contribution.

This approach is therefore not recommended because it will always either a) cause HCEs not to be able to take advantage of the maximum permitted pre-tax election, or b) require only a slightly smaller correction to HCE contributions than would have otherwise been required if they initially could have elected up to the full $5,000 maximum.

The Ugly: Permit Only NHCEs to Participate in the Dependent Care FSA

This approach has the advantage of guaranteeing the plan will pass the ABT because 100% of the benefits will be for NHCEs.  However, needless to say this approach does not go over well with the HCEs who are blocked entirely from dependent care FSA participation.

Best Practice Approach

Employers should not preemptively cap HCE elections at the outset, but instead run an early pre-test (e.g., summer for a calendar plan year) to determine any reduced contribution limit required to pass the ABT.  This approach ensures that HCEs receive the maximum pre-tax benefit that can be made available.

Although it is common for employers not to pass the ABT in pre-test results, running the test early will typically catch the issue before HCEs have contributed up to the reduced limit needed to achieve 55%.  The administrative burden is relatively minor where the adjustment simply requires a payroll contribution cap for HCEs.


Mid-Year Reduction to HCE Limit May Permit Spouse to Enroll/Increase Dependent Care FSA

The IRS has informally stated that the spouse’s employer may permit the spouse to enroll in (or increase an existing election in) the dependent care FSA sponsored by the spouse’s employer mid-year in response to an HCE being informed their contribution limit must be decreased to satisfy the ABT. 

If the spouse’s cafeteria plan permits this mid-year adjustment approach, it will in some cases allow a married couple affected by a failed ABT to still achieve the maximum $5,000 in combined dependent care FSA contributions.


Should ABT be Performed Based on All Employees or Only Those Participating?

There is debate among benefits professionals and the FSA TPA industry as to whether the ABT averages for HCEs and NHCEs are based on only those employees participating in the dependent care FSA, or based all employees regardless of participation.

The dependent care FSA rules refer to “the average benefits provided to employees” (emphasis added) who are NHCEs vs. HCEs, rather than the average benefits provided to eligible employees or participants.  Therefore, the best practice approach is to include all HCEs and NHCEs in the average regardless of participation or eligibility. 

Some FSA TPAs offer employers the ability to choose how to perform the ABT based on these different approaches, which can be reasonable for employers to consider based on their risk tolerance.


What Happens if Employers Fail to Perform the ABT?

The reason employers run the NDT for cafeteria plans, including the dependent care FSA ABT, is not for the sake of testing itself.  There are no penalties for failure to perform the testing.

The reason employers perform the NDT annually is to know whether HCEs are permitted to retain their full pre-tax cafeteria plan elections, and to be able to prove to the IRS that they treated the taxation of HCE elections correctly in case of IRS audit.

In other words, the NDT is performed simply to determine whether HCEs are permitted to retain the full exclusion from income for their elections.  Not running the NDT is a bad practice because if the employer fails to complete the NDT, the employer will not be able to verify that they passed (or the required corrections needed by the end of the plan year to reach a passing level).  

Employers that fail the ABT unknowingly have a Form W-2 reporting issue for the affected HCEs.  Those HCEs should have received no tax-advantaged dependent care FSA benefits for the year of the failure.  That generally creates a potential liability of $290 for the Form W-2 to the employee, and $290 for the copy of the Form W-2 to the IRS, for a general total of $580 per return in 2023.

Most FSA TPAs incorporate full cafeteria plan NDT into their standard service offering for the plan.  In situations where the FSA TPA does not offer NDT services, the employer should engage with a vendor that will offer the NDT services on a stand-alone basis for a fee.


The 55% Average Benefits Test is a regular pain point for employers who often express frustration with the frequent need to address ABT failures by reducing dependent care FSA elections for highly compensated employees.  However, employers can take three simple steps to alleviate at least some of these frustrations:

  1. Pre-test early;

  2. Consider using the top-paid group election (top 20%); and

  3. Catch and correct testing failures before HCEs reach the reduced contribution cap.

Admittedly, the process will always cause some level of small headaches in working with payroll to implement the reduced contribution limit and communicating the reduction to the affected HCEs.  However, consistently taking this approach has the benefit of avoiding the more burdensome administrative work, more restrictive preemptive limitations, and the last-minute recharacterizations before the end of the year that can put a damper in anyone’s holiday spirit.

For more details on cafeteria plan compliance considerations, see our Newfront Office Hours Webinar: Section 125 Cafeteria Plans.


Relevant Cites: 

IRC §129(d):

(d) Dependent care assistance program.

(1) In general.

For purposes of this section a dependent care assistance program is a separate written plan of an employer for the exclusive benefit of his employees to provide such employees with dependent care assistance which meets the requirements of paragraphs (2) through (8) of this subsection . If any plan would qualify as a dependent care assistance program but for a failure to meet the requirements of this subsection, then, notwithstanding such failure, such plan shall be treated as a dependent care assistance program in the case of employees who are not highly compensated employees.

(2) Discrimination.

The contributions or benefits provided under the plan shall not discriminate in favor of employees who are highly compensated employees (within the meaning of section 414(q)) or their dependents.

(8) Benefits.

(A)  In general. A plan meets the requirements of this paragraph if the average benefits provided to employees who are not highly compensated employees under all plans of the employer is at least 55 percent of the average benefits provided to highly compensated employees under all plans of the employer.


Treas. Reg. §1.414(q)-1, Q/A-9(b)(2)(iii):

(iii) Method of election. The elections in this paragraph (b)(2) must be provided for in all plans of the employer and must be uniform and consistent with respect to all situations in which the section 414(q) definition is applicable to the employer. Thus, with respect to all plan years beginning in the same calendar year, the employer must apply the test uniformly for purposes of determining its top-paid group with respect to all its qualified plans and employee benefit plans. If either election is changed during the determination year, no recalculation of the look-back year based on the new election is required, provided the change in election does not result in discrimination in operation.


Prop. Treas. Reg. §1.125-7(j):

(j) Time to perform nondiscrimination testing.

(1) In general. Nondiscrimination testing must be performed as of the last day of the plan year, taking into account all non-excludable employees (or former employees) who were employees on any day during the plan year.

ABA JCEB Q/A (May 2007):

  1. § 125 – Cafeteria Plans: Dependent Care

Husband has elected a $5,000 dependent care flexible spending account (FSA) through his employer’s Code § 125 cafeteria plan. Husband’s employer runs mid-year nondiscrimination tests and ceases husband’s contributions to his dependent care FSA. Is wife permitted to elect a dependent care FSA under her employer’s Code § 125 cafeteria plan for the remainder of the year?

Proposed Response: Yes, provided wife’s employer permits mid-year changes on account of and corresponding with a change made under another employer’s plan. See Treasury Regulation § 1.125-4(f)(4).

IRS Response: The Service representative agrees with the proposed response.

Brian Gilmore
The Author
Brian Gilmore

Lead Benefits Counsel, VP, Newfront

Brian Gilmore is the Lead Benefits Counsel at Newfront. He assists clients on a wide variety of employee benefits compliance issues. The primary areas of his practice include ERISA, ACA, COBRA, HIPAA, Section 125 Cafeteria Plans, and 401(k) plans. Brian also presents regularly at trade events and in webinars on current hot topics in employee benefits law.

Connect on LinkedIn
The information provided here is of a general nature only and is not intended to provide advice. For more detail about how this information may be treated, see our General Terms of Use.