State Auto-IRA Landscape

States and even municipalities across the country are taking an increasingly active role in addressing the nation’s retirement preparedness crisis. From California’s CalSavers Retirement Savings Program originating as early as 2012 to the Maine Retirement Savings Program enacted only a few weeks ago, many states – and even municipalities like New York City and Seattle, Washington – have implemented or at least proposed “auto-IRA” programs aimed to provide a retirement savings vehicle for employees without access to a retirement savings vehicle through an employer.

Generally, these programs are designed to overcome a small employer’s objections to offering a retirement plan that would otherwise be subject to the Employee Retirement Income Securities Act (ERISA).  These programs provide an automatic enrollment payroll deduction into an individual savings account, typically a traditional pre-tax and/or Roth IRA. There are no employer contributions, and there are no ERISA reporting and regulatory compliance requirements.  These programs operate using professional private management of investments that are not readily available to individuals with a private IRA.  While default employee contribution amounts vary, employers subject to such programs are usually required to either demonstrate that they are exempt by providing proof of an employer sponsored plan, or to enroll eligible employees, facilitate applicable deferrals through payroll and transmit amounts to the program sponsor. Noncompliance may result in penalties for the employer.

CalSavers: An ERISA Preemption Case Study

Traditionally, employer sponsored benefit plans are subject to ERISA, a federal law historically held to preempt conflicting state law. Thus, the question naturally arose whether state sponsored auto-IRA programs were preempted under ERISA. As of May 6, 2021, it appears we have an initial answer from the US Court of Appeals for the Ninth Circuit, which upheld a lower court decision finding that California’s CalSavers Retirement Savings Program was not preempted by ERISA.

In so finding, the court concluded that the CalSavers Program was not an ERISA plan for several reasons. First, CalSavers is not an employer sponsored plan. It is a program established and maintained by the state of California, not in any capacity as an employer of CalSavers participants. Second, the court found that CalSavers does not require employers to maintain an ERISA plan. Rather, it requires nonexempt employers to maintain administrative functions in order to facilitate deferrals made into the CalSavers Program through employer payroll. Third, CalSavers is not impermissibly connected to ERISA, nor does it interfere with ERISA’s purpose. In fact, CalSavers exempts employers maintaining ERISA retirement plans from participation in CalSavers, and nonexempt employers that are subject to CalSavers are responsible only for ministerial requirements that do not rise to the purview of ERISA.

Employer Fallout

The CalSavers decision clearly does not provide for a blanket exemption to other state programs and thus, ERISA preemption will likely continue to be considered on a program by program basis. However, in light of the CalSavers decision, it is likely that more states and municipalities will implement similar programs if Delaware’s legislative proposal in May 2021, Maine’s adoption of an auto-IRA program in June 2021, and New York state legislature’s expansion of its voluntary auto-IRA program to a mandatory program in June 2021 are any indication. This patchwork of programs that vary from jurisdiction to jurisdiction will present compliance challenges that are not insignificant for multistate employers, particularly those with nontraditional workforces who historically have not been eligible to participate in an employer sponsored retirement savings plan.

Of note, multistate employers are likely to be required – or at least encouraged – to register with each state’s program regardless of whether they are required to participate in the auto-IRA program. For example, CalSavers regulations require eligible employers to register, but specifically provide that exempt employers may, but are not required to, inform the program’s administrator of their exemption. However, unless an exempt employer provides notice to the program’s administrator, one can fairly assume that CalSavers will continue to send notices requesting registration of the employer and threatening potential penalties for noncompliance. This leaves the employer’s compliance and risks associated with potential penalties in limbo until notice of exemption is provided and accepted by the program’s administrator.

A related question that persists and which will likely have to be addressed program by program is the requirements for employers to demonstrate exempt status. While most known state auto-IRA programs or proposals exempt employers simply if they maintain or have recently maintained a qualified employer sponsored retirement savings plan, at least one state legislative proposal would only provide exemption if the employer provides “each” eligible employee the opportunity to participate in a plan and another state requires the eligible employees who are eligible to participant in the plan to be located within that state. Obviously, if states begin to require all employer plan sponsors to enroll employees who are ineligible to participate in the employer’s plan, the administrative burden to multistate employers would be increased dramatically. For example, employers with a large part-time or variable employee population may not provide eligibility to those employees working less than 1,000 hours a year in part due to the transient nature of that portion of their workforce. Under a state auto-IRA program requiring coverage for all employees in order to be exempt, employers may find themselves with a significant administrative burden in addition to their administrative scheme required for their own retirement savings plan. Given the Ninth Circuit’s reliance in part on the fact that the CalSavers program exempts employers maintaining ERISA retirement plans from participation in CalSavers, however, it is unclear whether a court would find such a program to be preempted by ERISA or not.

Further, it is unclear how controlled group rules applicable to ERISA would apply in regard to state auto-IRA compliance. Depending on each jurisdictions interpretation of regulations, employers who are active in mergers and acquisitions or who may utilize a controlled group structure in which employees may periodically move between entities, may find themselves having to account for a controlled group member who they are not required to provide eligibility for a retirement savings plan. For example, ERISA provides for a transition rule in the merger and acquisition context under which an acquiring company may temporarily test a recently acquired entity separately from its pre-acquisition controlled group. This rule effectively relieves the acquiror temporarily from being required to immediately add a recently acquired entity to its plan without failing applicable nondiscrimination requirements, giving the acquiror time to transition the new entity into the controlled group. However, under a state auto-IRA program, such acquiror would need to account for the entity immediately upon acquisition since no similar transition period applies under state law. A similar example may occur for businesses relying on a franchise model. In that context, it is not unusual for one company to step in and take over an entity for a short period of time in order to preserve the entity’s business operations while it is transitioned to a new company. Again, due to the temporary nature of such an employee relationship, the acquiring company would typically not extend retirement savings plan eligibility to employees of the entity during the period of time in which the entity is transitioned to a new company due to the short period of time involved. Under most state-auto IRA program regulations, this would result in the disruption of access to a state auto-IRA for employees of the entity during the period of time in which the company with a retirement plan employs those employees.

The Aftermath

The fallout of the CalSavers decision has some groups lobbying for Congress to provide tax credits for businesses who participate in these state and city auto-IRA programs, much like the tax credits that are available to businesses starting an ERISA plan.  While there is no fee on its face for an employer to participate in these state and city auto-IRA programs, employers do incur out-of-pocket costs in coordinating the data and payroll deductions with the state or city program as well as distributing information about the program to its employees.  One way for an employer to recoup these out-of-pocket costs is for Congress to provide federal tax credits associated with these programs.

Regardless of whether federal assistance becomes a reality, the unanswered questions and administrative burdens resulting from the emerging patchwork of state and city auto-IRA programs, specifically to multistate employers, seemingly cloud the impact of an otherwise clear Ninth Circuit decision. As a result, it is fair to assume that the increase of state auto-IRA programs will result in a corresponding increase in related employee claims. While an employer may only be responsible for the administrative payroll scheme in the eyes of the Ninth Circuit, it is conceivable that participants will include employers in related litigation. Assuming such state auto-IRA program is not subject to ERISA, participants are not subject to ERISA claims and appeals requirements or limited to federal court to litigate such claims, meaning employers could potentially face a myriad of claims in state or even municipal courts across the country from mild Maine to sunny California.

The 2021 Advisory Council on Employee Welfare and Pension Benefit Plans has announced that it will examine brokerage windows in participant-directed individual account retirement plans that are covered by ERISA.  The work of the Council is designed to assist the Department of Labor’s effort to determine whether more guidance would be appropriate and necessary to ensure that plan participants who have access to brokerage windows are adequately informed and protected.   ERISA does not define what is a “brokerage window” or “self-directed brokerage window,” but a common definition is an investment option in a participant-directed 401(k) plan that gives participants and beneficiaries the capabilities to buy and sell investment securities through a brokerage platform, above and beyond the core lineup of investment options offered by the plan.

Past DOL Guidance

In 2012, the DOL issued a revised Field Assistance Bulletin clarifying the disclosure requirements relating to brokerage windows offered in ERISA plans.  Apart from disclosure requirements, the FAB did not address how ERISA’s fiduciary standards might apply to brokerage windows.

In 2014, the DOL issued a Request for Information to increase its understanding of the prevalence and role of brokerage windows in self-directed individual account plans, including why and how often brokerage windows are offered and used in ERISA-covered plans.

Industry Support for Brokerage Windows

In connection with the Advisory Council’s June 2021 meetings, various industry participants provided input on brokerage windows in participant-directed individual account retirement plans.

The U.S. Chamber of Commerce provided written comments under a heading that says it all, “Don’t break brokerage windows.”  The Chamber believes that “[b]rokerage windows allow plan sponsors to meet the unique investing needs of certain participants” and that the DOL “should make it easier, not more difficult, for plan sponsors to offer this option if a plan sponsor feels it appropriate.”  The Chamber asked the DOL to issue formal guidance along the following lines:

  • a plan fiduciary is not liable for monitoring each underlying investment
  • tips on if and how to offer a brokerage window
  • sample language that describes what is involved with investing through a brokerage window (proposed model language was attached to the Chamber’s submission).

The ERISA Industry Committee (ERIC), a national trade association that advocates exclusively for large employers on health, retirement and compensation public policies, also spoke in support of brokerage windows.  ERIC’s view is that the current guidance on brokerage windows is sufficient for plan sponsors, and the DOL should not impose additional fiduciary requirements on plans with brokerage windows.

Specific Issues Addressed by Industry

  Prevalence of Offerings and Rate of Uptake

The Chamber cites a 2015 AON study showing that about 40% of plans offered a brokerage window and a 2019 Vanguard report showing that about 19% of its 401(k) plans offered a brokerage window.  The AON study says that about 3-4% of participants participate in the brokerage window option, while the Vanguard report says only 1% of participants use them.

ERIC conducted a survey of its own members (who are plan sponsors).  About half responded, and of that cohort, about 60% reported that they provide a brokerage window as part of their investment lineup.  As for the percentage of participants who actually use the brokerage window, ERIC states that a little more than half of the plans report a usage rate of 0-5%.  Another 24% of the plans report a usage rate between 6-15%, with 10% of the plans reporting a usage rate of 16% or higher.

In sum, only a fraction of plans offer brokerage windows, and an even smaller fraction of participants actually invest in them.


The Chamber says that although participant uptake is “low,” brokerage windows are still an important tool for those participants who want more varied investment options beyond the plan’s core lineup.  For example, some participants may wish to engage in Shariah investing or overall ESG investing, and while it may not make sense for a particular plan to offer these specialized kinds of investments as part of the core menu of investment options, the brokerage window can provide an avenue for these particular participants to meet their investing goals.

A common theme running through the industry comments is that the plans that offer brokerage windows do so because participants request them.

  Types of Participants Who Use Brokerage Windows

One of the express goals of the Advisory Committee’s study is to determine “who” is currently using brokerage windows.  The industry submissions provide some glimpses at the answer.

According to the Chamber, those who use brokerage windows tend to have higher account balances and are generally more sophisticated investors and more highly educated individuals who often work in finance, investing, law or engineering.

ERIC states that highly compensated employees were not the majority of participants in the brokerage windows their members offered.

According to ERIC, of those participants who invest in a brokerage window, it “seems to be rare,” that a participant invests 100% of their retirement account balance through the brokerage window.

  Concerns Over Potential Litigation

The Chamber asked the DOL to issue guidance that would help prevent the current wave of ERISA excessive fee litigation from spilling over to brokerage windows.  Specifically, the Chamber asked that the DOL “clarify that if a fiduciary otherwise meets the requirements under ERISA Section 404(c) and the applicable regulation, including the required disclosures under 29 CFR § 2550.404a-5, the fiduciary is not liable for any losses that a participant or beneficiary may incur from investing in a brokerage account.”

ERIC expressed an oft-repeated concern of plan sponsors about uncertainty over the issue of whether the plan fiduciaries have a duty to vet and monitor each and every underlying investment option being made available to participants through a brokerage window.  As stated by ERIC, “Any guidance from the DOL that would seek to impose fiduciary responsibilities over specific brokerage window investments would be unwieldy, if not impossible, to satisfy; potentially putting plan fiduciaries in the position of having to evaluate the thousands of investments and their appropriateness with respect to the investing plan participant and the plan.”

ERIC stated that if the DOL seeks to impose fiduciary liability over the underlying individual investments in a brokerage window, it would have a chilling effect on plan sponsors, potentially causing them to drop brokerage windows.  Or it could cause participants who rely on brokerage windows to “abandon” the employer retirement system in favor of IRAs or non-retirement accounts in which an open investment structure would remain available.


It remains to be seen what the Advisory Council will recommend to the DOL on the subject of brokerage windows and whether the DOL will take any action.  Will the DOL take steps to give plan sponsors more comfort, or less comfort, around offering brokerage windows?  Or will it continue to take a relatively hands-off approach on the subject, preserving the status quo in which a portion of retirement plans decide to offer brokerage windows among their investment options and an even bigger portion decide not to? will continue to track these questions and will report back after the Advisory Council announces its findings.



Retirement plans may have thousands of participants and billions of dollars in plan assets. Unfortunately, these large sums of money are attractive to bad actors who look to prey on unknowing victims by fraudulently accessing funds. Plan administrators, as fiduciaries of retirement plans, are wise to understand their legal obligations and best practices related to the security measures they must implement and maintain to protect these funds from cybercrimes.

Recent Cyber Attacks Against Retirement Plans

Earlier this year, in Bartnett v. Abbott Laboratories, et al. a retirement plan participant (Heide Bartnett) filed a lawsuit against her employer, Abbott Laboratories, the plan administrator, and the plan’s recordkeeper, Alight Solutions, LLC. According to the complaint, an individual impersonating the plaintiff attempted to access her retirement account by selecting the “forgot my password” prompt on the plan’s online recordkeeping platform. After requesting that a one-time security code be sent to the participant’s email account, which the impersonator had already improperly accessed, the impersonator gained access to the participant’s online retirement account and changed its password. Soon after, a new bank account was added to the participant’s retirement plan profile to which funds could be directly deposited from the participant’s retirement plan account. Two days later, the impersonator called Abbott’s service center to inquire about the transaction that he or she was (illegally) facilitating and was told that a distribution could not be made to the new bank account for seven days. Meanwhile, instead of attempting to contact the participant via phone or email (which was the plaintiff’s preferred method of communication), Abbott sent her a “snail mail” notice of the newly added bank account. By the time the participant received the notice, the impersonator had already looted her retirement account. Only a small fraction of the funds taken were recovered and the plaintiff filed a lawsuit seeking to recover $245,000, plus interest and other fees for the alleged breaches of fiduciary duty.

What Can Be Done To Stop Cybercrimes?

Although Abbott Laboratories is still a pending case, the plaintiff’s allegations are a stark reminder of the danger and risk that cybercriminals pose to retirement plans. Accordingly, plan administrators should ensure that the technical, physical, and administrative safeguards they have implemented to protect the confidentiality and integrity of plan assets satisfy basic legal requirements and meet industry security standards. Here are five areas that can serve as a starting point for a cybersecurity review in the retirement plan context:

First, plan fiduciaries should question the cybersecurity policies and procedures of their retirement plan recordkeepers and be aware of the liabilities they face for the shortcomings of their recordkeepers. Inquire about the recordkeeper’s cybersecurity capabilities and the safeguards in place to deter losses due to bad actors. In particular, inquire as to the access controls the recordkeeper has implemented to limit and verify access to an individual’s account. How are the controls created? How often are they tested? Have they ever been compromised, and if so how? What is the recordkeeper’s password policy for account access? Does the recordkeeper require multifactor authentication?

Second, identify whether the plan fiduciaries and the recordkeeper have an adequate level of cybersecurity insurance. It is also worth determining whether any existing insurance or fidelity bond coverage will provide financial relief in the case of a cybersecurity breach. If basic insurance coverage does not apply to forgery, consider a rider for additional coverage.

Third, request a copy of the recordkeeper’s data breach response plan and identify how often the recordkeeper undertakes table-top exercises or similar activities to test its response capabilities. It is important to identify where the plan sponsor aligns within the recordkeeper’s plan and even consider joint data breach-type exercises. If permitted, seek to identify any outside service providers and counsel that the recordkeeper has retained for such emergencies and ensure that they are qualified and capable to respond to data breaches upon a moment’s notice.

Fourth, require the recordkeeper to undergo third-party security and vulnerability testing so they can identify and remediate any aspect of their security program that presents a risk. It is especially important to ensure that high or critical risk vulnerabilities are resolved within hours or days (and not weeks or months). Accordingly, ensure that the recordkeeper has identified (in writing) an official who is fully responsible for the security of the plan’s assets. Accountability is a key aspect of any security program.

Fifth, educate plan participants. Let them know they can take an active role in protecting their own plan assets. As basic as it may seem, remind participants not to share their login or personal information with anyone. The allegations against Abbott Laboratories explain that an email account was compromised which allowed the bad actor to request authentication to the compromised email. Once the false authentication was made, the recordkeeper processed the request to have an additional bank account added. Savvy participants can help play an active role in the protection of their own account assets.

A Tough Road Ahead

Careful considerations by plan administrators have become especially important in light of the COVID-19 pandemic because there has been a steady increase in certain cyber-related crimes during this time. The recently enacted CARES Act provides many retirement plan participants with the opportunity to take large in-service distributions and loans, and such distributions and loans are ripe for the nefarious acts which were the basis for the Abbott Laboratories case. As a result, plan administrators need to stay vigilant and ahead of the curve when it comes to cybersecurity protections.

The Seventh Circuit has issued its decision in the much-anticipated case of Divane v. Northwestern.  The district court below had refused to allow plaintiffs to proceed with breach of fiduciary duty and prohibited transaction claims based on the recordkeeper’s use of participant data for purposes of “cross-marketing” non-plan services to plan participants.  The issue arose in a unique procedural posture, a motion for leave to amend the complaint near the close of discovery. The district court found that the proposed new counts, including the cross-marketing claim, were untimely (for being raised six days before the close of discovery) and futile (for failing to state a claim).

In affirming the district court decision, the Seventh Circuit agreed that the allegations based on cross-marketing were untimely and “failed to state claims for relief.” This is the first time the issue of cross-marketing participant data has been decided at the circuit court level. Going forward, this precedent will pose a significant obstacle for plaintiffs who wish to pursue cross-marketing claims.

Plaintiffs’ Counsel Has Been Focusing on Cross-Marketing Claims

For a few years now plaintiffs have been challenging fiduciaries who allow service providers, usually recordkeepers, to utilize participant data to offer non-plan financial services to participants. For example, in their second amended complaint in Cassell v. Vanderbilt University, plaintiffs alleged that plan fiduciaries breached their fiduciary duties by allowing the plan’s recordkeeper “to use its position as the plan’s recordkeeper to obtain access to participants, gaining valuable, private and sensitive information including participants’ contact information, their choices of investments, the asset size of their accounts, their employment status, age, and proximity to retirement, among other things.” Further, plaintiffs alleged, the plan fiduciaries allowed the recordkeeper to use this valuable and confidential information to sell the recordkeepers’ products and wealth management services to the plan’s participants and “failed to even attempt to determine the value of this marketing benefit.”

Similar cross-marketing claims were brought against Johns Hopkins, MIT and Northwestern.

The District Court Rejects a Claim Based on Cross-Marketing Participant Data

In many ways Divane v. Northwestern was a typical ERISA fee case, with plaintiffs challenging the (allegedly) excessive fees and underperformance of various investment options in the university’s defined contribution plan. Near the close of discovery, however, the plaintiffs tried to add a claim that the plan fiduciaries should be liable for allowing the plan’s recordkeeper to market products to plan participants using participants’ contact information, their choices of investments, the asset size of their accounts, their employment status, age, and proximity to retirement.

The district court refused to allow plaintiffs to pursue this cross-marketing claim, finding that it should have and could have been raised earlier in the case. In addition to the issue of timeliness, the district court addressed the futility of allowing an amended pleading based on allegations of cross-marketing:

  • It is not imprudent to allow the recordkeeper to have access to this kind of participant information.
  • Disclosure of such information to the recordkeeper does not implicate ERISA fiduciary functions.
  • Not a single court has held that releasing confidential information or allowing someone to use confidential information is a breach of fiduciary duty, and “[t]his Court will not be the first, particularly in light of Congress’s hope that litigation would not discourage employers from offering plans.”

The district court also found that the recordkeeper’s use of participant data for cross-marketing was not a prohibited transaction because the court was “not convinced” such information is a plan asset. Thus, plaintiffs’ proposed claim based on using participant data for cross-marketing failed to state a claim.

Plaintiffs’ counsel appealed.

Some Defendants Settle and Agree to Prohibit Cross-Marketing

While the Divane v. Northwestern appeal was pending, and with a lack of controlling precedents specifically addressing cross-marketing issues in this context, defendants in some other ERISA cases hedged their bets by settling the cross-marketing claims asserted against them. For example, Vanderbilt University settled its case by paying $14.5 million and agreeing, among other things, to prevent any future cross-marketing by the plan’s recordkeeper. Specifically, Vanderbilt agreed that going forward the plan fiduciaries shall contractually prohibit the recordkeeper from using information about plan participants acquired in the course of providing recordkeeping services to market or sell products or services unrelated to the plan to plan participants unless initiated by a plan participant.

Subsequently, Johns Hopkins and MIT settled their ERISA fee cases by agreeing, among other things, to forbid cross-selling by their plans’ service providers.

The Seventh Circuit Affirms, Adopting the District Court’s Reasoning

After these settlements, the Seventh Circuit issued its opinion in the Divane v. Northwestern case. The Court of Appeals affirmed the district court decision to not allow plaintiffs’ leave to amend their complaint to add the “participant data” claims on the eve of trial.  The appellate court laid out the district court’s rationale (too late, and no precedent for treating participant data as a plan asset) and concluded, “We agree.”  Importantly, the Seventh Circuit wrote that the proposed data claim “fails to state a claim for relief.”  Such a conclusion is independent of the timeliness of the proposed claim and provides a significant hurdle for future plaintiffs to overcome if they seek to bring cross-marketing claims against plan fiduciaries.


Plaintiffs’ counsel are unlikely to give up their quest to pursue claims against fiduciaries who do not prevent service providers from cross-marketing participant data just because one circuit court has rejected such claims. This is especially true if defendants continue to show a willingness to settle such claims for a mixture of monetary and non-monetary concessions. Nevertheless, the Seventh Circuit’s opinion in Divane v. Northwestern will be a significant impediment to cross-marketing claims unless plaintiffs get it reversed on rehearing or in the Supreme Court. (On April 22, 2020, plaintiffs filed a petition for rehearing or rehearing en banc, which did not specifically address their proposed cross-marketing claim, but instead focused on the pleading standard in ERISA fee/investment cases.)



In a prior post, we commented on the growing trend of fiduciaries making non-monetary concessions to settle ERISA fee litigation cases. We observed that certain “onerus” non-monetary settlement features – such as obligating fiduciaries to provide plaintiffs’ counsel with customized reports on plan operations and performance during a years-long “monitoring” period — are significant and intrusive and that they are sowing the seeds of potential future disputes.

Well, the trend continues. In recent months ERISA class action settlements were announced involving MIT and Duke University.  Most of the headlines focused on the fact that, MIT and Duke agreed to pay $18.1 million and $10.65 million, respectively, to settle those cases.  That’s definitely noteworthy.

But that’s just the tip of the iceberg. As with Johns Hopkins and others, MIT and Duke also agreed to “substantial non-monetary terms,” which the parties claimed materially added to the total value of their settlements.

MIT’s Non-Monetary Concessions

In addition to paying $18.1 million (up to one-third of which may go to plaintiffs’ counsel), MIT agreed to the following terms:

  • Training. MIT shall provide training to Plan fiduciaries on how to discharge their duties under ERISA. (Thus converting a best practice into a contractual obligation.)
  • RFP’s. MIT agreed to submit RFP’s to at least three qualified recordkeepers, requesting that the proposed fees be expressed on a per-participant basis, not on a percentage of plan assets basis. After reviewing the RFP’s, MIT may keep its current recordkeeper or select a new one. The final bid amounts shall be provided to class counsel on a confidential, no-names basis as to those not selected. (Nothing in ERISA specifically requires RFP’s on a regular basis, but now MIT has obligated itself to conduct them.)
  • Revenue Sharing. Here is how the revenue sharing provision is described in the court filings: “Any revenue sharing related to Plan investments will be deposited in the Plan trust and, to the extent not seasonably used to defray lawful Plan expenses, be returned to Plan participants according to a method of allocation approved by Plan Fiduciaries and permitted by ERISA not less frequently than on an annual basis.”
  • Allocation of Expenses. The plan fiduciaries will determine a method of allocating recordkeeping and administrative expenses that it determines is fair, equitable, and appropriate for plan participants “separate from the flat fee negotiated with the recordkeeper and based on the number of plan participants.”
  • No Cross-Selling. MIT and the plan fiduciaries agreed to allow the plan’s recordkeeper to communicate with participants only at the direction or with the authorization of plan officials, and agreed to prohibit “any communications to Plan participants (in their capacities as such) concerning non-Plan products and services.” In other words, no cross-marketing of non-Plan services by the recordkeeper. (No court has found cross-selling to be an ERISA violation, but MIT has agreed to prohibit it.)
  • Independent Investment Consultant. MIT agreed to continue its practice of using an independent investment consultant. (So what the fiduciary once voluntarily chose to do, it is now obligated to continue doing.)
  • Costs. MIT and plaintiffs agreed that the costs relating to the use of an independent consultant and the costs of conducting an RFP are expenses properly paid by the plan under applicable law.

Duke’s Non-Monetary Concessions

In addition to paying $10.65 million, Duke agreed to the following “non-monetary terms”:

  • Disclosures to Class Counsel. For two years Duke will provide to Class Counsel a list of the Plan’s investment options and fees, and a copy of the Plan’s Investment Policy Statement.
  • Disclosures to Participants. Duke will communicate in writing to current participants and inform them of the investment options available in a new lineup, and provide a link to a web page containing fees and performance information for the new investment options, as well as contact information to facilitate funds transfers out of a frozen annuity account.
  • Possible RFP. During the third year of the settlement, the fiduciaries shall retain an independent consultant to provide a recommendation on whether they should conduct a request for proposal for recordkeeping and administrative services.
  • Factors to Consider in Investments. In considering Plan investment options during the Settlement Period, the fiduciaries agreed to “consider” various enumerated factors, such as costs and availability of rebates.
  • No Use of Plan Assets. During the Settlement Period, Duke shall not cause Plan assets to be used to pay salaries, benefits and expenses incurred by Duke for services performed by Duke employees.

Differing Terms; Same Recipe for Future Disputes?

Johns Hopkins, MIT and Duke each agreed to substantial non-monetary settlement terms, but those terms differed for each university (as did the amount of their cash payments).  This indicates that the parties are separately negotiating the non-monetary terms of their respective settlements, and are not treating them as boilerplate.  Nevertheless, there is some overlap in the non-monetary terms agreed to by all three universities.  For example, all three agreed to conduct or consider RFP’s on an agreed timetable, to report certain information to plaintiff’s counsel on an agreed timetable, and to utilize independent investment consultants.  Two of the three, Johns Hopkins and MIT, agreed to forbid cross-selling by their service providers, but Duke’s settlement does not include such a prohibition.

These obligations carrying into the future are mostly, if not completely, one-sided — running from the fiduciaries to the participants and/or class counsel.  For example, Johns Hopkins agreed to provide customized “reporting” to plaintiffs’ counsel and to subject itself to “monitoring” by plaintiffs’ counsel for a three-year period.  MIT did not agree to those, but it did agree to provide fiduciary training (which is a best practice in any event) and to have its revenue sharing deposited in the plan and shared with plan participants to the extent not used to defray lawful plan expenses. MIT also agreed to determine a method of allocating recordkeeping and administrative expenses among participants separate from a per participant flat fee.  Duke agreed to “consider” certain enumerated factors when making investment decisions.

In agreeing to a particular course of conduct going forward, these fiduciaries are, essentially, taking general standards under ERISA, like prudence or reasonableness, and agreeing to convert them into more specific rules. For example, nothing in ERISA requires a fiduciary to conduct RFP’s on a regular basis, but by agreeing to do so going forward, a new market standard is perhaps being shaped through litigation.

This seems like fertile ground for future disputes.  What happens if the fiduciaries don’t comply with their ongoing non-monetary obligations?  Or fail to document their compliance?  Plaintiffs’ counsel have already sued these fiduciaries once.  Are these future entanglements with plaintiffs’ counsel worth the risk?

Strategic Change: Good or Bad?

The trend of plaintiffs’ counsel extracting concessions from fiduciaries regarding operational changes in the administration of defined contribution plans continues. Whether this is good for participants, fiduciaries, and plans as a whole remains to be seen.   But it is doubtful that regulation through litigation is the most efficient way to set standards of conduct in a multi-billion dollar industry.

As the bellwether cases in the ERISA actuarial assumptions litigation approach the end of the motion to dismiss stage, this is a good time to step back and assess how they are proceeding.

Test cases of this sort tend to unfold in phases.

  • In the first phase, plaintiffs file a series of test cases, floating new legal theories and challenging relatively wide-spread behavior on a selective basis. Defendants respond by filing motions to dismiss, testing the sufficiency of plaintiffs’ legal theories and factual allegations.
  • In the second phase, at least in cases where the plaintiff survived a motion to dismiss, the parties engage in fact and expert discovery, culminating in summary judgment practice. Here the facts are developed case-by-case, and one or both parties seek to end the litigation before trial by convincing the court that on the undisputed facts that have been developed in discovery, one of them is entitled to judgment as a matter of law.
  • If the court finds that there are genuine disputed issues of material fact, motions for summary judgment will be denied, and the cases will proceed to the third phase: trial.

Of course, settlement can occur during any of these phases.

Sometimes defendants manage to defeat the early test cases in a decisive fashion, snuffing out the incentive for future litigation by obtaining early dismissals that dissuade plaintiffs’ attorneys from continuing to invest time and resources, usually on a contingent basis, in a losing cause. Other times the bellwether litigation progresses in a less decisive manner, with both sides scoring some early victories and absorbing some early setbacks, as they wrestle their way through the multiple phases of litigation.

Scorecard on the First Phase

Though the first phase is not complete for all the actuarial assumptions cases that have been filed, it is not too early to check out the scoreboard. Of the 11 cases that have been filed, three are still in the first phase (motions to dismiss pending or expected to be filed soon), one has settled, and the rest have entered the second phase (discovery and summary judgment underway, but no rulings on summary judgment yet). The 11 test cases were not all filed at the same time. In fact, the first case was filed in December 2018, and the last was filed 13 months later. In light of this lag time – and the fact that federal cases typically proceed at different paces depending on the judge, the lawyers and the parties –the cases are not moving through the litigation phases in unison.

However, as the first phase approaches conclusion, we can take an early look at the scoreboard and draw certain conclusions.

Cases Outcome of Phase 1
6 Defendants’ motions to dismiss denied.
1 Defendants’ motions to dismiss granted. (Plaintiff immediately filed a motion for reconsideration, which was granted, and the case settled within days on undisclosed terms.)
1 Defendants’ motions to dismiss granted in part, and denied in part.
3 Defendants’ motions to dismiss still pending or expected to be filed soon.

So far the defendants have not delivered a knockout punch against these cases, but they have at least three more attempts.

In addition to these numerical outcomes, we can also make a few additional observations about the earliest phase of this litigation.

  1. Geographic Diversity. Plaintiffs are taking pains to file the early cases in a variety of federal appellate circuits, including some that are typically more plaintiff-friendly and some that are not. So far they have commenced cases in eight different circuits.

Undoubtedly this is designed to diversify the test case portfolio and soften the blow in the event of a poor outcome in one district court (or on appeal, in one circuit court).

  1. Procedural Basis for Many of the Denials. A procedural theme is emerging from the early opinions in which courts have denied the defendant’s motion to dismiss: courts are reluctant to embrace defendants’ theories as a matter of law and are inclined to allow plaintiffs to proceed to the discovery phase. In other words, defendants’ theories are not being rejected out of hand in most cases; rather, courts are finding that it would be premature to embrace those theories at the motion to dismiss stage, where plaintiffs are entitled to a presumption that their factual allegations are true and need only satisfy a plausibility standard on their claims.
  2. Risky to Read Too Much into the Early Opinions. The emphasis on procedural issues in the early opinions makes it difficult to draw too many hard and fast conclusions about what the courts think about the relative merits of the parties’ substantive positions. There are a few statements in the opinions that sound more substantive than procedural. But any such statements should be taken with a grain of salt because they were written in the context of a motion to dismiss, where plaintiffs enjoy favorable presumptions. All of the merits-based elements in the motion to dismiss opinions will be revisited, rehashed, scrutinized, debated, and refined in the summary judgment phase.

In conclusion, these cases are a long way from providing plan sponsors, fiduciaries, actuaries and participants any kind of definitive guidance on the challenged behavior and various issues teed up in these bellwether cases. It will be interesting to see if courts are comfortable granting summary judgment on the claims and defenses raised by the parties, or whether they will send a test case or two to trial – and, eventually, to appeal.

Stay tuned to for updates when phase two nears completion.

Many Americans use prescription drugs on a daily basis to control or prevent a wide variety of illnesses. However, the increasing costs of prescription drugs make it hard for many Americans to obtain the medications they require. One way to combat this is through the use of manufacturer coupons. The process is simple. A manufacturer sells a high-priced prescription drug and pairs it with a coupon to allow an individual to obtain the drug at an affordable price. Pretty simple, right? Well, for plan administrators of self-insured group health plans, it may not be so simple. Plan administrators must determine whether the total cost of the drug counts toward the individual’s annual out-of-pocket maximum or just the amount actually paid out-of-pocket by the individual.

For example, let’s say the annual cost of a prescription drug is $10,000 and a plan participant purchases a yearly supply with a $9,500 coupon and $500 of his or her own money. Should the plan administrator credit $10,000 or $500 towards the annual out-of-pocket maximum?

Recent Guidance Is Here to Help!

In order to help answer this question, the Centers for Medicare & Medicaid Services (“CMS”) released the Health and Human Services Notice of Benefit and Payment Parameters for 2020 on April 25, 2019 (“the Notice”). The Notice provides that for plan years beginning on or after January 1, 2020, any amounts paid using direct support offered by drug manufacturers to eliminate or reduce out-of-pocket costs (coupons) of a brand-name prescription drug are not required to be counted toward the annual out-of-pocket maximum when there is a therapeutically equivalent generic drug available. In these situations, plan administrators may, but are not required to, ignore the value of the manufacturer coupons when calculating the plan participant’s annual out-of-pocket maximum.

But what happens when there is no therapeutically equivalent generic drug available? Although not explicitly stated, the language of the Notice (which is supported by the preamble to the Notice) suggests that the value of a brand-name drug coupon must be counted towards the out-of-pocket maximum where there is no therapeutically equivalent generic drug available.

Wait, We Forgot to Consult the Tax People

However, such a reading conflicts with the rules relating to high deductible health plans (“HDHPs”) (and could interfere with an individual’s ability to contribute to a health savings account). These rules provide that a HDHP cannot pay claims (other than claims for preventive care services) until the annual deductible has been satisfied.

Way back in 2004, the IRS released Notice 2004-50. Addressing the issue of manufacturer coupons, Q&A-9 requires a HDHP to disregard manufacturer discounts when calculating whether the minimum annual deductible has been satisfied. The rationale for this Q&A is that a HDHP should not begin to pay claims (other than claims for preventive care services) until a plan participant has paid the high deductible in full out-of-pocket. Otherwise, allowing for coupon amounts to be credited towards the deductible would allow a plan participant to circumvent the HDHP rules. Using the example from above, if all $10,000 is credited towards the plan participant’s deductible, the HDHP will begin paying claims when the plan participant has only paid $500 out-of-pocket (assuming the deductible is $10,000 or less).

These conflicting rules could be read to require a plan administrator to credit coupons for brand-name drugs without a therapeutically equivalent generic drug available as required by HHS while simultaneously requiring a plan administrator not to credit coupons for brand-name drugs without a therapeutically equivalent generic available as required by the IRS.

A Possible Solution?

Responding to this possible conflict, the Departments of Labor, Health and Human Services, and Treasury (collectively, the Departments) released FAQs About Affordable Care Act Implementation Part 40. These FAQs provide that the Departments intend to undertake rulemaking regarding the Notice and its possible implications for 2021. In the interim, the Departments will not initiate any enforcement action if a group health plan excludes the value of drug manufacturers’ coupons from the annual out-of-pocket maximum, even in situations where there is no therapeutically equivalent generic drug available.

Where To Go From Here?

While we await further guidance, plan sponsors should determine how their plans will credit coupons in 2020 and whether any changes to plan language are required. Should your group health plan include or exclude the value of manufacturer coupons in the annual out-of-pocket maximum for 2020? For HDHPs, it seems clear that the value of manufacturer coupons should be excluded from the annual out-of-pocket maximum. For all other group health plans, plan sponsors can determine to include or exclude the value of all or some coupons from the annual out-of-pocket maximum. For 2021, who knows?

There always seem to be enough open important questions to keep ERISA practitioners operating in some uncertainty. When new legislation or regulatory guidance is not forthcoming, ERISA practitioners only have the Supreme Court and the lower federal courts to look to for assistance. Although the Supreme Court usually takes either zero or one ERISA cases per year, beginning earlier in 2019, the Supreme Court has shown a renewed interest in ERISA.

At least four ERISA-related cases will be argued before the Supreme Court this year. It is possible that more will be added, though if so, they likely will not be fully briefed and argued until next Fall.

First, here are the four that are already accepted for review by the Supreme Court:

1.     Retirement Plans Committee of IBM v. Jander

This case will likely shed more light on the Supreme Court’s prior holding in Fifth Third Bancorp v. Dudenhoeffer, focusing on what must be pled to state a claim that a fiduciary has failed to take an action that would not do more harm than good. After Dudenhoeffer set forth that standard, with really only one exception, ERISA claims involving alleged breaches of fiduciary duty in stock drop cases have been dismissed.

This case may be more about pleading standards under Federal Rule of Civil Procedure 12(b)(6) than ERISA by the time it is written, but it arises in the context of a type of case – – stock drop cases – – that are of great interest to ERISA practitioners.

This case is to be argued on Wednesday, November 6, 2019.

2.     Intel Corp. Inv. v. Sulyma

ERISA’s hybrid statute of limitations breach of fiduciary duty claims includes one prong that applies when an individual has actual knowledge of an alleged breach. In that case, an individual must bring an action within three years. This case poses a question what is “actual knowledge.” If the alleged breach is apparent from ERISA-required disclosures, but the Plaintiff claims they cannot recall if they read them, is that actual knowledge?

This case is set to be argued on Wednesday, December 4, 2019.

3.     Maine Community Health Options v. United States (and two other cases)

This case is not directly about ERISA, but it does interest ERISA practitioners in that it involves congressional appropriations of amounts to pay insurers under the Affordable Care Act. Congress included an appropriations rider that barred Department of Health & Human Services from using funds to pay a statutory obligation to insurers as part of the ACA. This issue will affect the financial health of these insurers.

This case is set to be argued on December 10, 2019.

4.     Thole v. U.S. Bank, N.A.

This case involves the issue of whether a claim for breach of fiduciary duty regarding a defined benefit plan can be maintained when the plan is fully funded. In a defined-benefit plan, no individual has a right to a portion of the trust funds. Instead they have a right to the timely (eventual) payment of their amount due under the Plan. Therefore, some courts have said that as long as the plan is fully funded by ERISA’s standards, there is no standing to bring a claim for breach of fiduciary duty as no participant has been harmed.

The argument date for this case has not yet been set, but it will be no earlier than January 13, 2020.

There are two other cases awaiting the Supreme Court’s decision as to whether to hear the case where the Court has asked for the view of the government’s top lawyer – the Solicitor General of the United States. While the Court’s request for the Solicitor General’s view does not guarantee that they will take the case – indeed sometimes the Solicitor General recommends that the Court not grant review – it is an indication of heightened interest by the Court.

Putnam Investments, LLC v. Brotherston

This involves the issue of which side (the participants who are suing or the fiduciaries who are defending) has the burden of proof to show that a loss did or did not result from an alleged breach of fiduciary duty. The Solicitor General has not yet filed its brief indicating its opinion as to whether the Court should grant review.

Rutledge v. Pharmaceutical Care Management Association

This is a throwback to the old days of preemption litigation. This is about Arkansas’ drug reimbursement rate law and whether it is preempted by ERISA. In this case too, the Court is still awaiting the view of the Solicitor General.

There could be additional petitions for review regarding ERISA that catch the Court’s eye later in the term. None of those are likely to be heard this term as the last set of arguments is in April 2020. However, at least for now the Supreme is more focused on ERISA than it has been for years.

Part 2: Partial Plan Terminations

Workforce reductions seem to be an inescapable consequence of economic downturns. Whether this occurs through the sale of a business, layoffs or plant closures, employers too often overlook the potential impact on their employer-sponsored retirement plans. Unfortunately, failure to recognize and timely address the retirement plan implications of a reduction in force can jeopardize the plan’s tax qualified status and lead to costly and time-consuming corrections.

Earlier this month, we kicked off our blog series focused on the benefits implications of an economic downturn in Winter Is Coming: Employee Benefits Planning for the Eventual Economic Downturn. With Part 2, we now shift our focus to partial plan terminations. Fear not – as night gathers, our watch begins.


Section 411(d)(3) of the Internal Revenue Code requires tax qualified retirement plans—both defined contribution plans (including 401(k) plans) and defined benefit plans—to fully vest affected participants upon the occurrence of a partial plan termination.

Failure to timely recognize and properly administer this obligation may result in the improper forfeiture of accrued benefits of terminated participants, correction of which may grow more-costly and administratively burdensome with the passage of time. For example, affected participants may include those who have terminated employment and received distribution of their account balances. Correction would likely require establishing a new account for the participant, crediting the amount improperly forfeited (with earnings), and communicating the error to the participant. That is, of course, if you can locate the former employee. In addition to potentially onerous correction, these types of failures also create a potential litigation risk from former participants related to claims for impermissibly reduced accrued vested benefits.

Has a Partial Termination Occurred?

A partial termination generally occurs when a sufficiently large number of plan participants are terminated over a specified period of time. The IRS has also indicated that a partial termination can occur for reasons other than turnover, such as plan amendments that exclude a group of employees that has historically been covered by the plan.

Whether a partial termination has occurred is generally measured over the plan year, though a longer period may apply if the reductions are in connection with a “corporate event.” Corporate events may include reduction in workforce (RIF) programs, plant closures, sales of businesses, or similar events.

To determine whether a partial termination has occurred, the following ratio must be calculated:

Plan Participants Terminated
During the Period (both Vested and Unvested)

Plan Participants at the Beginning of the Period +
New Participants Hired During the Period

If the resulting ratio for this period meets or exceeds 20%, a partial termination is presumed to have occurred.

There has been confusion in the past regarding who must be counted for purposes of calculating the turnover ratio. IRS guidance has clarified that both vested and nonvested participants are included in the calculation. The guidance has also clarified that employees who terminate employment other than those who die, become disabled, or retire at normal retirement age during the period must be included unless the employer can demonstrate that the employee would have resigned even if the corporate event had not occurred. This, of course, may be a difficult showing the make and generally leads to inclusion of all terminations other than those resulting from death, disability or retirement.

Who Must Be Fully Vested?

The IRS takes the position that all participants who terminate during the period covered by the partial termination must be vested. While it is possible to take the position that an employee who was terminated for cause or other unrelated reasons, or who voluntarily resigned during the period, should not be vested, most employers follow the IRS’s position and vest all employees who terminate during the applicable period.

Navigating a Partial Termination

While a plan termination may be “partial”, the plan sponsor should be fully engaged in planning for the event to avoid the burdensome correction previously mentioned. Advance planning will also allow the employer to anticipate future qualification issues that may increase the cost of plan compliance following the event. For example, if a plan sponsor of a 401(k) plan terminates a significant number of non-highly compensated employees, the plan may experience non-discrimination testing issues in the future, necessitating additional employer contributions to the plan or unwelcome benefit reductions among highly compensated employees.

While this blog has primarily focused on implications to 401(k) plans, partial plan terminations also may occur in defined benefit plans. We will focus on partial plan terminations in defined benefit plans—and other implications of reductions in force unique to defined benefit plans—in an upcoming blog.

The Wrap

In the event a plan sponsor proceeds with widespread layoffs, closing of a facility, reducing the pool of eligible employees, or similar action, care must be taken to comply with the resulting requirements, namely the accelerated vesting of participant retirement plan accounts. However, with a strategic plan, the risks associated with a partial plan termination can be anticipated and mitigated, resulting in the successful cost-savings that are necessary in an economic downturn.

Part 1: Introduction

Some economists are now predicting a global economic downturn as soon as 2020, as indicators from bonds, interest rates, currencies, and commodities signal declining growth, including the recent inversion of the yield curve. While there is not a consensus on this point (as George Bernard Shaw once said, “if all economists were laid end to end, they would still not reach a conclusion”), it’s hard to ignore the growing chorus. There is no doubt that the ebbs and flows of the economy and business trends impact employee benefit plans and the employers that sponsor them, often leading to hard decisions to reduce or eliminate benefits. Poor execution can turn already difficult decisions into potential liabilities and result in a devastating increase to employer costs.

This blog series will explore common benefits cost reduction decisions and associated compliance pitfalls. Future blog posts will consider, among other topics—

401(k) Plans

  • Reductions in Employer Contributions to a 401(k) Plan: Employers are generally free to make changes to employer contributions to a 401(k) plan but must do so in accordance with a variety of legal requirements. Failure to properly implement changes can create fiduciary risk and potential loss of tax-qualified status of the plan.
  • Terminating a 401(k) Plan: Plan termination may result in significant cost savings, but also must be accomplished in accordance with legal requirements. Nor is plan termination a quick and painless process: plan documents must be updated, and missing participants must be addressed. Missteps in properly communicating and implementing a termination may also carry fiduciary risks.
  • Partial Termination of a 401(k) Plan: Even without an intent on the part of the employer, significant decreases in plan participation as a result of layoffs, reorganizations, or amendments limiting plan eligibility may lead to a partial plan termination requiring full vesting of affected participants. Failure to recognize the occurrence of a partial plan termination may result in fiduciary exposure and qualification problems for the plan.
  • Company Stock in a 401(k) Plan: Risks associated with plan investments in employer stock can be exacerbated in a down economic environment. Plan fiduciaries may have to disclose certain information to participants or decide whether to discontinue the plan’s investment in company stock.

Pension Plans

  • Freezing plan participating and benefit accruals: Freezing a pension plan requires employers to carefully navigate anti-cutback rules and participant notice requirements.
  • Terminating a Pension Plan: Pension plans generally must be fully funded before they may be terminated. And the involvement of the PBGC—the government insurer of private pension plans—adds additional complexity to the termination process.
  • PBGC Reportable Events: Pension plans are required to notify the PBGC of certain events that indicate a pension plan or employer may be in severe financial distress.
  • Withdrawal from Multiemployer (or Union) Pension Plan: Multiemployer pension plans have increasingly faced funding challenges, which may be exacerbated by a difficult economic environment. Certain employer actions may result in a withdrawal from the plan and trigger liability for the employer’s proportionate share of the plan’s unfunded vested liabilities.

Health and Welfare Plans

  • Health Plan Contributions and Unpaid Claims: Employers in extreme financial distress may be tempted to view obligations under health and welfare plans as those of a creditor but doing so may cause fiduciary exposure and trigger costly prohibited transactions.
  • COBRA Obligations Resulting from a Reduction in Force: COBRA requires the continuation of coverage for a limited time under group health plans for most employers upon certain qualifying events.
  • Terminating a Health and Welfare Plan: While an employer generally may terminate a health or welfare plan, such action may be a recipe for significant tax penalties arising from the employer’s failure to offer group medical coverage to its full-time employees and their children.

Severance Plans

Many employers inadvertently create ERISA plans when providing former employees severance payments payable over time under an administrative scheme. If the severance plan is an ERISA plan, there are additional legal requirements that must be met, including annual Form 5500 filings. Plan counsel should be consulted to determine whether severance benefits are not subject to ERISA.

Over the coming months, the series will dive deeper into these and other possible decisions precipitated by an economic downturn. Each blog will identify lessons learned and opportunities for strategic employer cost-savings and provide employers with a road map to navigate associated legal risks and obstacles. Winter is coming, but we can help plan sponsors be better prepared for these uncertain economic times.