Lori Chavez-DeRemer v. Medical Staffing of America, LLC

Opinion

PUBLISHED

UNITED STATES COURT OF APPEALS FOR THE FOURTH CIRCUIT

No. 23-2176

LORI CHAVEZ-DEREMER, Secretary of Labor, United States Department of Labor,

Plaintiff – Appellee,

v.

MEDICAL STAFFING OF AMERICA, LLC, d/b/a Steadfast Medical Staffing, a limited liability company; LISA ANN PITTS, individually and as owner and officer of the aforementioned company,

Defendants – Appellants.

No. 23-2284

LORI CHAVEZ-DEREMER, Secretary of Labor, United States Department of Labor,

Plaintiff – Appellee,

v.

MEDICAL STAFFING OF AMERICA, LLC, d/b/a Steadfast Medical Staffing, a limited liability company; LISA ANN PITTS, individually and as owner and officer of the aforementioned company,

Defendants – Appellants.

2

Appeals from the United States District Court for the Eastern District of Virginia, at Norfolk. Raymond A. Jackson, Senior District Judge. (2:18-cv-00226-RAJ-LRL)

Argued: September 27, 2024 Decided: July 17, 2025 Amended: July 17, 2025

Before KING and RICHARDSON, Circuit Judges, and FLOYD, Senior Circuit Judge.

Affirmed by published opinion. Judge King wrote the opinion, in which Judge Floyd joined. Judge Richardson wrote a dissenting opinion.

ARGUED: Abram John Pafford, MCGUIREWOODS LLP, Washington, D.C., for Appellants. Anne Warren King, UNITED STATES DEPARTMENT OF LABOR, Washington, D.C., for Appellee. ON BRIEF: Francis J. Aul, MCGUIREWOODS LLP, Washington, D.C., for Appellants. Seema Nanda, Solicitor of Labor, Jennifer S. Brand, Associate Solicitor, Fair Labor Standards Division, Rachel Goldberg, Counsel for Appellate Litigation, Office of the Solicitor, UNITED STATES DEPARTMENT OF LABOR, Washington, D.C., for Appellee

3

KING, Circuit Judge:

Defendants Medical Staffing of America, LLC, d/b/a Steadfast Medical Staffing, and Lisa Ann Pitts, individually and as owner and executive officer thereof (collectively, “Steadfast,” or the “Defendants”), pursue these consolidated appeals from a judgment entered against them in the Eastern District of Virginia. As determined by the district court, Steadfast contravened multiple provisions of the Fair Labor Standards Act (the “FLSA,” or simply the “Act”), by failing to properly classify approximately 1100 nurses as Steadfast “employees” (as the Plaintiff Secretary of Labor maintains), rather than as “independent contractors” (as Steadfast maintains). Steadfast has thus been adjudged liable for unpaid overtime compensation to its nurses in a sum of almost five million dollars, plus a nearly equal sum of liquidated damages.

As explained herein, the district court’s disposition of the Secretary of Labor’s enforcement action against Steadfast was resolved after a bench trial conducted by the court in Norfolk in 2021. The findings of fact and conclusions of law made by the court — the “Bench Verdict” — underlie the contentions of Steadfast on appeal. Put succinctly, our rulings reject each of the contentions of error interposed by Steadfast, and we therefore affirm the judgment.

I.

A.

We begin by briefly identifying certain legal principles that underlie these appeals.

At the forefront is the issue of whether the Certified Nursing Assistants, Licensed Nurse

4

Practitioners, and Registered Nurses who worked for Steadfast at some point from approximately August 2015 until January 2023 were “employees” of Steadfast under the FLSA.1 1.

As our Court recently emphasized, “Congress enacted the FLSA in 1938 — in the midst of the Great Depression — to combat the pervasive ‘evils and dangers resulting from wages too low to buy the bare necessities of life and from long hours of work injurious to health.’” See Salinas v. Commercial Interiors, Inc., 848 F.3d 125, 132 (4th Cir. 2017) (quoting S. Rep. 75-884 at 4 (1937)). By enacting the FLSA, Congress established several elementary protections for those working as “employees” in this Country, such as the requirement that an employer pay overtime for hours worked in excess of 40 hours per workweek, as well as the requirement that employers maintain proper records of the “persons employed” and “the wages, hours, and other conditions and practices of employment.” See 29 U.S.C. §§ 207(a)(2), 211(c).

To advance the FLSA’s “remedial and humanitarian” purpose of protecting “the rights of those . . . who sacrifice a full measure of their freedom and talents to the use and profit of others,” Congress created sweeping definitions of the terms “employer,” “employee,” and “employ.” See McFeeley v. Jackson Street Ent., LLC, 825 F.3d 235, 240

1

The Certified Nursing Assistants, Licensed Nurse Practitioners, and Registered Nurses who worked for Steadfast as “nurses” during the relevant time periods are identified in Schedule A of the Complaint.

5

(4th Cir. 2016) (internal citations omitted). As codified in Title 29 of the United States Code, the FLSA broadly defines an “employer” as “any person acting directly or indirectly in the interest of any employer in relation to an employee.” See 29 U.S.C. § 203(d). In a similarly broad fashion, an “employee” is defined by the Act as “any individual employed by an employer.” Id. § 203(e)(1). And the Act “defines the verb ‘employ’ expansively to mean ‘suffer or permit to work.’” See Nationwide Mut. Ins. Co. v. Darden, 503 U.S. 318, 326 (1992) (quoting 29 U.S.C. § 203(g)).

The Supreme Court has thus emphasized that the “striking breadth” of those FLSA’s definitions serve to place under the “employee” umbrella even those American workers “who might not qualify as [employees] under a strict application of traditional agency law principles.” See Nationwide Mut. Ins. Co., 503 U.S. at 326. The breadth of the FSLA’s protections of American workers, therefore, is “not confined exclusively to employees within the traditional legal distinctions separating them from independent contractors.” See NLRB v. Hearst Publ’ns., 322 U.S. 111, 126 (1944).

2.

For those workers who have been improperly classified as “independent contractors,” and consequently deprived of their entitlement to compensation for overtime hours worked, the FLSA provides an individual right of action for the recoupment of the full compensation illegally withheld from them by their employers. See 29 U.S.C § 207(a)(1) (defining overtime compensation). More specifically, an employer that fails to compensate a worker for overtime hours — that is, those hours worked in excess of 40

6

hours per workweek — “shall be liable to the employee or employees affected in the amount of . . . their unpaid overtime compensation.”2 Id. § 216(b) (emphasize added).

Section 216(b) of Title 29 “adds to that compensation” an award of “liquidated damages of an amount equal to the unpaid . . . overtime compensation,” which may be rebutted by an employer’s showing of “good faith.” See United States v. Edwards, 995 F.3d 342, 346 (4th Cir. 2021) (discussing FLSA in context of court-ordered restitution for “full amount of victim’s losses”); see also 29 U.S.C. § 260 (providing for good faith defense). Put simply, if an employer can demonstrate “that the act or omission giving rise to such action was in good faith,” and that the employer “had reasonable grounds for believing that his act or omission was not a violation of the [FSLA],” a court may, in its discretion, “award no liquidated damages or award any amount thereof,” up to an amount equal to the unpaid overtime compensation. See 29 U.S.C. § 260.

In addition to ordering the reimbursement of unpaid overtime compensation and making an award of liquidated damages, a district court possesses equitable jurisdiction “to restrain violations” of the FLSA — that is, “to enjoin acts and practices made illegal by the Act and to force compliance with the Act.” See 29 U.S.C. § 217; Mitchell v. Robert DeMario Jewelry, Inc., 361 U.S. 288, 291 (1960) (discussing “inherent equitable powers”

2

In describing what the FLSA calls “unpaid overtime compensation,” several other terms have been used by the lawyers and the district court during this litigation. See 29 U.S.C. § 216(b)-(c). As a result, terms like “back wages,” “back wage compensation,” “back pay,” and “back pay compensation,” have been used to describe what we generally refer to herein as “unpaid overtime compensation.”

7

of district courts to award relief under FLSA). And the broad jurisdictional power conferred by the FLSA is not limited to a private, individual right of action. Rather, the Act empowers the Secretary of Labor to enforce compliance with the FLSA and to recover unpaid overtime compensation withheld unlawfully “on behalf of any employee” by “bring[ing] an action in any court of competent jurisdiction to recover the amount of unpaid . . . overtime compensation and an equal amount as liquidated damages.” See 29 U.S.C. 216(c); see also Mitchell, 361 U.S. at 294.

With this legal framework in mind, we turn to the factual predicate of these appeals, which is drawn from the very substantial appellate record presented in these proceedings. That underpinning includes, in particular, the trial conducted in the district court and the Bench Verdict thereafter rendered.

B.

Steadfast is a Virginia medical staffing services business that was founded in Norfolk in 2015. Its business model has been to supply licensed nurses to multiple healthcare facilities located in Virginia and surrounding states. In order to do so, Steadfast maintains what it calls a “registry” of its nurses, and it connects those nurses with work opportunities at its client healthcare facilities.3

3

As the Secretary emphasizes, “Steadfast [labelled] its business as a ‘registry’ but this label [was] not determinative of the economic realities of Steadfast’s relationship with its nurses.” See Br. of Appellee 3 n.1.

8

1.

To qualify for inclusion in Steadfast’s so-called registry, a nurse was required to complete an “application for employment” — a document that referred to Steadfast as the “employer,” and to the nurse as the “employee.” This employment application included, inter alia, questions about the nurse’s credentials, skills, and employment history. After the employment application was completed, Steadfast paid for and executed several pre- hiring requirements, including a background investigation, a drug screening, and a tuberculosis test. Steadfast also verified that the applicant possessed a valid nursing license.

A nurse found qualified after submitting an application for employment and completing Steadfast’s pre-hiring requirements would then enter into what Steadfast called its “independent contractor” agreement, and the nurse would be included in Steadfast’s registry. Notably, Steadfast’s agreements with its nurses contained a “non-compete clause,” which recited that a Steadfast nurse had agreed that:

Commencing from the first date of the [nurse’s] engagement by [Steadfast]

and continuing from [sic] a period of twelve (12) months from the effective date of expiration or termination of this Agreement, [the nurse] agrees that [the nurse] shall not directly provide that [sic] type of services provided to [Steadfast] to any competitors of [Steadfast] without first obtaining the express written permission of [Steadfast]. [The nurse also] agrees that during the term of this Agreement and for twelve (12) months thereafter, [the nurse]

will not enter into, be engaged in, or be interested in any capacity whatsoever,

9

directly or indirectly in any business or undertaking which competes with that of [Steadfast].

See J.A. 1796.

Being included in Steadfast’s registry allowed the nurses to be notified of available placements — referred to as “shift opportunities” — at Steadfast’s client healthcare facilities. After receiving a staffing request from a client facility, Steadfast unilaterally decided which of its nurses were eligible to fill the available shift opportunity. Although Steadfast’s client facilities identified the services being sought and the qualifications of the nurses required for an available placement, those facilities did not have access to the Steadfast nurses’ individual contact information. Steadfast provided its nurses with training on topics such as patient confidentiality, substance abuse, and sexual harassment. The client facilities thus relied on Steadfast to ensure that its nurses were capable, prepared, licensed, and otherwise qualified.

Upon receipt of a staffing request from a client healthcare facility, Steadfast would notify eligible nurses in its registry of an available shift opportunity. That notification was generally made by phone, text message, email, or by use of Steadfast’s mobile application for communicating with its nurses, called the “Zira app.”4 Steadfast nurses could then accept or decline the shift opportunity. If a Steadfast nurse accepted, he or she would report

4

Steadfast hired a business named Zira Technologies to develop the Steadfast scheduling and pay system, which was called the “Zira app.” And Steadfast implemented the “Zira app” as its primary scheduling tool in 2021. Steadfast had declined, however, to include an important feature offered for the “Zira app” that would have tracked all overtime hours the Steadfast nurses worked, in excess of 40 hours per workweek.

10

to the client facility as directed by Steadfast. When a nurse was late for a shift, wanted time off, was ill or otherwise unable to complete the shift, the nurse was obliged to notify and obtain approval from Steadfast — not from the relevant client facility.

During their shifts at client healthcare facilities, Steadfast nurses would generally perform the tasks of a typical nurse, as directed by a supervising physician — i.e., administering medications, treating wounds, and otherwise caring for patients. Steadfast, however, controlled other aspects of the nurses’ workplace conduct. Steadfast, for example, required its nurses to wear identification, consisting of a badge issued by Steadfast and bearing the Steadfast name. And although the nurses’ technical tasks would typically be supervised by the client facility’s physicians, Steadfast promulgated written standards for its nurses’ workplace conduct, including the nurses’ attire, punctuality, and timekeeping. Failure to adhere to those internal standards and policies could result in the nurses being disciplined by Steadfast.

Steadfast not only addressed and resolved violations of its internal policies with its nurses, it also administered sanctions for poor work or disciplinary infractions reported by the client healthcare facilities. Although Steadfast requested that its client facilities provide feedback concerning each nurse’s performance, those facilities were contractually barred from disciplining the Steadfast nurses. Client facilities were instead required to contact Steadfast regarding problems with its nurses. Steadfast, rather than the client facility, would then address a disciplinary matter — including, inter alia, punishment for unprofessional conduct, declining or cancelling shifts, working for competitors, discussing

11

compensation with coworkers or the client facilities, or even contacting a client facility concerning work schedules or pay rates.

2.

Steadfast unilaterally dictated its nurses’ hourly pay rates, and it negotiated fixed hourly rates for the work its nurses performed at the client healthcare facilities. In turn, Steadfast would retain a percentage of the hourly rate paid to Steadfast by the client facility. As such, the Steadfast nurses could not directly negotiate their hourly pay rates with the client facilities where they worked. Rather, the Steadfast nurses could earn more only one way — by working additional hours.

Although Steadfast did not impose hourly requirements on its nurses, Steadfast paid them “straight time pay,” i.e., their regular rates for all hours worked, even overtime hours. In other words, Steadfast nurses were not paid the FLSA mandated one and one-half times their regular rate for overtime hours. On the other hand, Steadfast’s competitors often paid overtime compensation for the hours their nurses worked in excess of 40 hours per workweek. Steadfast was thus able to charge its client facilities lower hourly rates for its nurses’ services than its competitors would charge.

Steadfast was solely responsible for compensating its nurses and for the handling of their compensation-related issues, such as wage garnishments. Steadfast was paid directly by its client healthcare facilities for all work performed by the Steadfast nurses. In order to be paid for their completed work shifts, the Steadfast nurses were required to prepare and submit timesheets directly to Steadfast. Those timesheets — created and maintained by Steadfast — required the nurses to track their working hours and obtain an approving

12

signature of a client facility’s staff member to verify their hours worked. Steadfast would then pay the nurses from its own accounts. If a client facility failed to pay Steadfast for the services rendered by its nurses, Steadfast paid them in any event. In addition to maintaining its nurses’ timesheets, Steadfast also maintained what it called “employee change” records, which documented each nurse’s “hire date” or “term date,” and “tax jurisdiction.”

Steadfast provided insurance coverage for its nurses, including workers compensation policies for injuries the nurses sustained while working at the client healthcare facilities. And Steadfast covered its nurses under what are called general liability insurance policies. The nurses were not required by Steadfast to carry any of their own liability insurance. However, the client facilities required Steadfast to maintain various types of liability insurance coverage for the Steadfast nurses. Steadfast thus assumed responsibility for its nurses’ work at the client facilities, as the nurses’ “employer,” in multiple contracts with those client facilities. And Steadfast referred to its nurses as “employees” or “employed personnel” in those contracts.

The contracts between Steadfast and the client healthcare facilities also contained provisions that prohibited the client facilities from recruiting Steadfast nurses to fill employment positions at those facilities. Steadfast sometimes enforced this contract prohibition through a “buyout” clause, which required a client facility to purchase the nurse’s contract from Steadfast, if the facility had an interest in doing so. Otherwise, a Steadfast nurse recruited by a client facility could be disciplined or terminated by Steadfast.

13

II.

A.

In 2017, the Department of Labor (the “DOL”) launched an investigation into Steadfast’s business practices concerning the Steadfast nurses, in order to assess whether FLSA’s statutory requirements were being complied with. As a result of that investigation, the DOL concluded that Steadfast was violating the Act by misclassifying its nurses as “independent contractors,” failing to pay them overtime compensation, and failing to maintain proper records. Steadfast was so informed by the DOL, and Steadfast was advised in early 2018 by the DOL to properly classify its nurses as “employees.” Put simply, Steadfast was requested to comply with the FLSA.

Steadfast failed to act as requested by the DOL, however, and instead continued to classify its nurses as “independent contractors.” As a result, Steadfast continued to pay its array of nurses “straight time pay” for all their hours worked. That is, Steadfast failed to pay its nurses in accordance with FLSA’s overtime compensation mandate, i.e., one and one-half times the regular rate for hours worked over 40 hours per workweek. Steadfast also disregarded DOL’s request that it comply with other relevant provisions of the FLSA, including classification of its nurses as “employees,” and the maintenance of proper records.

On May 2, 2018, then-Secretary of Labor Acosta initiated this FLSA enforcement action in the Eastern District of Virginia against Steadfast, alleging in his Complaint that it was willfully misclassifying its nurses as “independent contractors.” Schedule A of the Secretary’s Complaint identified the Steadfast nurses who were to be beneficiaries of the

14

enforcement action. The Complaint specified that Steadfast was failing to properly classify and compensate its nurses as “employees” when they worked more than 40 hours per workweek. And it alleged that Steadfast had also failed to comply with FLSA’s recordkeeping requirements. As a result, the Secretary’s Complaint sought unpaid overtime compensation, plus liquidated damages and injunctive relief, on behalf of the Steadfast nurses identified in Schedule A.5 In March 2020, following discovery and other pretrial proceedings, Steadfast and the Secretary filed cross-motions for summary judgment. On July 23, 2020, the district court resolved those motions by denying each of them. In making those denial rulings, the court concluded that genuine disputes of material fact had been presented, and thus that neither party was entitled to a judgment as a matter of law. As a result — and because neither party sought a jury trial — a bench trial was scheduled before the presiding district judge.

B.

The bench trial of this litigation was conducted in Norfolk — over a period of about seven trial days — in August and September of 2021. Evidence was fully presented to the district court by both the Secretary of Labor and by Steadfast, and the court heard from approximately 30 witnesses. The court also received multiple evidentiary exhibits

Schedule A of the Complaint was revised during the course of these proceedings

5

and identified the 1105 Steadfast nurses who are the beneficiaries of the challenged judgment.

15

introduced by the parties. The Secretary, as the Plaintiff herein, initiated the evidence presentations and primarily sought to prove the Complaint’s allegation that the Steadfast nurses were “employees” of Steadfast under the FLSA. In pursuing that effort, the Secretary called about 25 witnesses. In response to the Secretary’s evidence, Steadfast called about six witnesses.

1.

a.

The Secretary’s trial evidence focused on the controlling issue of worker classification, i.e., whether the nurses were “employees” of Steadfast, as the Secretary maintained, or whether they were “independent contractors,” as Steadfast argued. For its part, Steadfast emphasized and pursued a “good faith defense” against the Secretary’s misclassification contentions. That is, Steadfast maintained that the Secretary’s proof concerning the nurses’ employment status was lacking, and that its “independent contractor” classification of its nurses — even if erroneous — had been made in good faith. Pursuant to the FLSA, that is, § 260 of Title 29, proof of a good faith classification of workers can be an affirmative defense to a liquidated damages award.

Because a substantial amount of unpaid overtime compensation for the Steadfast nurses was being sought by the Secretary, the trial evidence also involved extensive presentations concerning the unpaid overtime compensation issues. The Secretary’s primary supporting documentary evidence regarding damages computations were presented as “Exhibit PX-21” and “Exhibit PX-22” (collectively, the “Initial damages computations”). Exhibit PX-21 was itself predicated on thousands of data entries obtained

16

from Steadfast’s payroll records, invoices, and timesheets. And the related Exhibit PX-22 is properly described as a summary of PX-21. The calculations and methodologies explaining the Initial damages computations were discussed thoroughly by counsel and the court and explained extensively from the witness stand.

The Initial damages computations were presented by DOL officials during the bench trial. One of the DOL officials who testified in that regard was a Mr. Melendez, who was the DOL supervisor of the primary investigator on the Steadfast investigation. Melendez explained, inter alia, how a “regular rate” had to be manually calculated for each Steadfast nurse, because the payroll data that Steadfast provided did not include it. He also explained how the relevant overtime rates had been determined, based on each of the Steadfast nurses’ regular rates.

A Mr. Xiong, another DOL official, testified at length regarding preparation of the Initial damages computations, which were primarily predicated on Steadfast’s payroll and other records. Xiong explained that, although an initial manual transcription of the relevant data was necessary, automated formulas had been utilized for the computations, and those computations detected and corrected “excessive” hours worked in some workweeks. In sum, the Initial damages computations — consisting of more than 600 pages — were utilized to support the Secretary’s damages requests at trial and later to support the district court’s damages awards.

b.

Steadfast’s witnesses included Ms. Pitts, the owner and executive officer of Steadfast, and a Virginia lawyer named Bredehoft. Steadfast had retained Bredehoft for

17

legal advice — subsequent to the DOL investigation — regarding Steadfast’s alleged misclassification of the nurses as independent contractors. And Pitts had consulted with Bredehoft on two occasions after the DOL investigation had concluded. Pitts testified in some detail about the business operations of Steadfast and her meetings with Bredehoft, and she asserted that Steadfast’s classification of its nurses as independent contractors was made in good faith.

Lawyer Bredehoft testified as a fact witness — rather than as an expert — and was questioned extensively about his post-DOL investigation consultations with Ms. Pitts and Steadfast. In rendering his advice to Steadfast concerning the classification issue, Bredehoft relied solely on the information Pitts supplied to him. And he advised Pitts during their second consultation — based on the information Pitts had provided — that it was likely that the Steadfast nurses were being properly classified as independent contractors. Bredehoft made his classification determination in part on his belief that Steadfast did not have any authority to discipline its nurses. Bredehoft also understood, based on the information Pitts provided him, that the “non-compete clause” had never been enforced, had never been threatened to be enforced, and that all of the Steadfast nurses were aware of it. Bredehoft confirmed, however, that if the “non-compete clause” was ever enforced or threatened to be enforced, his view of the classification issue would differ.

Bredehoft had advised Ms. Pitts that Steadfast should immediately cease any use of its “non-compete clause,” and that it should also stop using “employment” terminology in its application documents. In Bredehoft’s view, Steadfast’s use of the “non-compete clause” and the “employment” terminology in its application materials was inconsistent

18

with an independent contractor classification of the Steadfast nurses. Pitts acknowledged that she did not make the changes advised by lawyer Bredehoft.

2.

About two months after the bench trial concluded, on November 3, 2021, the lawyers presented the district court with their proposed findings of fact, arguing their separate positions on the pertinent issues. Two months later, on January 14, 2022, the court filed its Bench Verdict, detailing its findings of fact and conclusions of law. See Su v. Medical Staffing of Am., LLC, No. 2:18-cv-00226 (E.D. Va. Jan. 14, 2022), ECF No. 324. Put simply, the Bench Verdict ruled that the nurses were “employees” of Steadfast, and that Steadfast was thus liable to the Secretary for misclassifying its nurse employees and for violating the FLSA. The Bench Verdict also assessed and rejected Steadfast’s contentions with respect to the “good faith” defense that it asserted at trial. As a result, Steadfast was found by the Bench Verdict to be liable to the Secretary, on behalf of the nurses, for unpaid overtime compensation during the pertinent time periods.

In the Bench Verdict, the district court generally characterized what we call unpaid overtime compensation as “Back Wages.” And the court made findings of fact that approved the Secretary’s Initial damages computations that had been introduced concerning the amounts of unpaid overtime compensation due to the Steadfast nurses, including the methods utilized by the Secretary to correct errors discovered in Steadfast’s payroll data. Based on the Initial damages computations sponsored by the Secretary, the

19

Bench Verdict awarded the Secretary, on behalf of the Steadfast nurses — for the period from about August 2015 through June 2021 — approximately $3.6 million.

In its Bench Verdict, the district court explained that, both prior to and during the bench trial, the court had urged and directed Steadfast to review and assess the accuracy of the Initial damages computations. The Bench Verdict found, however, that Steadfast had failed to conduct those court-ordered reviews. Indeed, Steadfast had not sought to either refute or contest the Initial damages computations. And the Bench Verdict emphasized that Steadfast had not presented any counter-computations of unpaid overtime compensation to oppose those presented by the Secretary.

The Bench Verdict thus determined that the Secretary had established that the Steadfast nurses were “employees” of Steadfast, and that Steadfast had failed to meet its burden of a good faith defense for its misclassification. That is, the Bench Verdict determined that Steadfast’s misclassification of its nurses was neither in good faith nor objectively reasonable, and its good faith defense was therefore rejected. That ruling exposed Steadfast to an additional award of liquidated damages, pursuant to 29 U.S.C. § 216(b) (“Any employer who violates the provisions” of the FLSA that mandate overtime “shall be liable to the . . . employees affected in the amount of . . . their unpaid overtime compensation . . . and in an additional equal amount as liquidated damages.”).

The Bench Verdict also addressed the recordkeeping issues that had been presented at trial and ruled that Steadfast had failed to satisfy the FLSA’s recordkeeping requirements. See 29 U.S.C. § 211(c) (requiring employers to “make, keep, and preserve such records of the persons employed by him and of the wages, hours, and other conditions

20

and practices of employment maintained by him, and [to] preserve such records . . . .”). The Bench Verdict then awarded injunctive relief — which had been sought by the Secretary pursuant to 29 U.S.C. § 217 — enjoining Steadfast from further violations of the FLSA.

Turning to the issues concerning the amount of unpaid overtime compensation that would be awarded against Steadfast, the Bench Verdict ordered the Secretary to provide the district court with post-trial updates of the Initial damages computations, reflecting the unpaid overtime compensation owed to the nurses, as well as the potential liquidated damages awards, within 60 days of the Bench Verdict. Although the Secretary sponsored a preliminary award by the court of the unpaid overtime compensation and liquidated damages specified in the Initial damages computations, those trial submissions only accounted for Steadfast’s FLSA violations from August 2015 through June 2021. More specifically, the Initial damages computations did not account for Steadfast’s FLSA violations after June 2021.6

6

Because Steadfast had not at trial disputed the Secretary’s methodology for ascertaining the unpaid overtime compensation reflected in the Initial damages computations, the Bench Verdict directed Steadfast to cooperate with the Secretary in securing the post-trial updated computations. And Steadfast was instructed in the Bench Verdict to supply the records necessary to complete the court-ordered updated computations.

21

C.

1.

About two months after the Bench Verdict, on March 11, 2022, the Secretary, in compliance with the district court’s directive, submitted what was called an “Updated Back Wage Computations Notice” for the court’s consideration. Two days later, on March 13, 2022 — for the first time in this litigation — Steadfast sought to dispute the Secretary’s Initial damages computations, specifically seeking to challenge the methodologies and accuracy thereof.7 And the very next day, on March 14, 2022, Steadfast noticed an interlocutory appeal to this Court.

Steadfast’s interlocutory appeal related solely to the injunction aspect of the Bench Verdict. Appellate jurisdiction was predicated on 28 U.S.C. § 1292(a)(1) (providing jurisdiction of appeals from “[i]nterlocutory orders of the district courts . . . granting . . . injunctions . . . .”). On April 8, 2022, the district court entered an order delaying the proceedings in Norfolk and deferring to the appeal proceedings in this Court. The court explained that, “in the interest of judicial economy,” it would “defer action” on Steadfast’s pending March 13, 2022 motion for relief from Secretary’s Initial

7

On March 13, 2022, Steadfast replaced its counsel in this litigation with a new set of lawyers. The new lawyers, who yet represent Steadfast in these proceedings, were the first Steadfast lawyers to dispute the Initial damages computations, and thus the first to dispute the resulting court-ordered unpaid overtime compensation and liquidated damages computations. Put most simply, Steadfast did not dispute the Initial damages computations until at least six months after the bench trial concluded, and not until two months after the Bench Verdict had been rendered.

22

damages computations and from its updated computations, pending resolution of Steadfast’s interlocutory appeal.

Three months thereafter, however, on July 14, 2022, the Secretary moved in the district court for an updated computations order, seeking to finalize the amounts of unpaid overtime compensation and liquidated damages to be awarded “as a result of [Steadfast’s] continued violations of the overtime provision” of FLSA. That request for updated and final computations remained pending in the district court during the interlocutory appeal.

On April 20, 2023, however, before the appeal was resolved, the Secretary submitted to the court a “Second Updated Back Wage Computations Notice,” again seeking to account for Steadfast’s continuing FLSA violations. The Secretary’s April 20, 2023 “Second Updated Back Wage Computations Notice” was supported by about 250 pages of what became the “Final damages computations.” Using the same methodologies utilized to compute the damages amounts in the Initial damages computations, the Secretary’s April 20 submission of the Final damages computations also accounted for two additional time periods, from June 2021 to January 2022, and then from January 2022 to January 2023.

The injunction appeal pursued by Steadfast was resolved by this Court on May 31, 2023, and our decision did not reach or resolve the merits of the Bench Verdict. Su v. Med. Staffing of Am., LLC, No. 22-1290, 2023 WL 3735221, at *1 (4th Cir. May 31, 2023). In an unpublished opinion, we addressed the injunction issues only. And we therein concluded that the district court had “erroneously fail[ed] to comply with the requirements of Rule 65(d) of the Federal Rules of Civil Procedure.” Id. That is, the court had not

23

adequately explained why the injunction was issued; it had failed to specify the terms thereof; and it had not described in reasonable detail the act or acts restrained or required. Id. Our decision thus indicated that the court had also failed to properly address the injunctive relief factors established by the Supreme Court. In sum, the injunction aspect of the Bench Verdict was vacated and remanded to the district court.

2.

a.

Following our May 31, 2023 remand, the district court promptly conducted further proceedings. The court carefully reassessed the injunction aspect of the Bench Verdict, addressed various challenges and issues concerning the Initial damages computations, and finalized its unpaid overtime compensation and liquidated damages rulings and awards. The injunction issues were again addressed and disposed of in an Opinion filed on September 7, 2023. See Su v. Medical Staffing of Am., LLC, No. 2:18-cv-00226 (E.D. Va. Sept. 7, 2023), ECF No. 395. The court therein entered another injunction against Steadfast, and that ruling complied with the terms of Rule 65(d). Shortly thereafter, on October 1, 2023, the Secretary moved for an updated order and final judgment, requesting final damages rulings regarding unpaid overtime compensation and liquidated damages awards. Id. at ECF No. 396. Before the court resolved that motion, however, on November 3, 2023, Steadfast noticed yet another appeal, seeking interlocutory review of the court’s second injunction ruling. Id. at ECF No. 404.

24

b.

On December 7, 2023, the district court filed an Opinion that resolved Steadfast’s then pending motion for relief from the Secretary’s Initial damages computations. See Su v. Medical Staffing of Am., LLC, No. 2:18-cv-00226 (E.D. Va. Dec. 7, 2023), ECF No. 411.8 And the court therein rejected Steadfast’s challenges to the Initial damages computations. Those efforts of Steadfast had sought relief under a “shotgun array” of the Federal Rules of Civil Procedure, including Rules 52(b), Rule 54(b), and Rule 60(a)-(b).9 Carefully explaining the various defects that the court identified in Steadfast’s efforts to secure relief from the Initial damages computations, the court rejected each of Steadfast’s efforts.

8

In addition to challenging the Initial damages computations, Steadfast’s motion for relief, filed on March 13, 2022, had also opposed the Secretary’s March 11, 2022 court- ordered “Updated Back Wage Computations Notice.” The district court’s December 7, 2023 Opinion advised the parties that all issues concerning the updated damages computations would be addressed separately by the court. The Opinion of December 7 thus only addressed Steadfast’s challenges to the Initial damages computations, which had been relied upon in the Bench Verdict.

9

A brief explanation of “shotgun array” of the Civil Rules that Steadfast relied on in its March 13, 2022, motion is warranted. First, Rule 52(b) provides that “[o]n a party’s motion filed no later than 28 days after the entry of judgment, the court may amend its findings — or make additional findings — and may amend the judgment accordingly.” Fed. R. Civ. P. 52(b). Second, Rule 54(b) provides a court with authority to modify an interlocutory judgment “at any time before the entry of a judgment adjudicating all the claims and all the parties’ rights and liabilities.” Fed. R. Civ. P. 54(b). Third, Rule 60(a) allows a court to “correct a clerical mistake or a mistake arising from oversight or omission when one is found in a judgment, order, or other part of the record.” Fed. R. Civ. P. 60(a). And Rule 60(b) provides that a party may request relief from a final judgment, order, or proceeding on the basis of multiple specific reasons, including “mistake, inadvertence, surprise, or excusable neglect,” or “any other reason that justifies relief.” Fed. R. Civ. P. 60(b)(1), (6).

25

c.

After denying Steadfast’s multiple challenges to the Initial damages computations on their merits, the district court’s Opinion of December 7, 2023, went further. The court therein specified that, even if Steadfast could have identified a valid basis for relief under one of the multiple Civil Rules it relied on, Steadfast had simply “waived [its] right to attack [the Initial damages computations].” See Su v. Medical Staffing of Am., LLC, No. 2:18-cv-00226 (E.D. Va. Dec. 7, 2023), ECF No. 411 at 10. That waiver had occurred, as the court explained, when Steadfast “failed to make any arguments regarding the [unpaid overtime compensation] computations before or during the trial,” even after the court had advised Steadfast that those “computations would be based on the record and not anything presented after trial.” Id. at 9, 11. Despite the court’s multiple directives in that regard, Steadfast had failed to present any evidence that contested the Initial damages computations in any respect. Nor did Steadfast identify any flaws in — or even make objections concerning — the Initial damages computations during the bench trial, or during the four-month post-trial period between the trial and the Bench Verdict. Id.

The only semblance of an evidentiary objection to the Initial damages computations mentioned by Steadfast prior to the Bench Verdict is found in a pretrial order of March 10, 2021, where the parties were to identify pretrial objections. Steadfast’s position on Exhibit PX-21, which is the foundation of the Initial damages computations, read: “Defendants object to the relevance and computations contained in this exhibit but agree that if liability is established to work with plaintiff’s counsel to reach a stipulation as to the calculations.” See Su v. Medical Staffing of Am., LLC, No. 2:18-cv-00226 (E.D. Va. March 10, 2021),

26

ECF No. 261. No follow-up statement explaining that proposition, however, was ever made by Steadfast. But its position in that regard was referenced at trial when the issue of damages computations was discussed.10 At that point, Steadfast’s counsel referenced its pretrial assertion about the accuracy of the Initial damages computations, asserting that, “if liability is established, then [Steadfast] would work with the plaintiff to verify the accuracy.” In response, the trial court specified for the parties how it would determine the damages issues, stating:

[l]et’s put it this way: If liability is determined, the [c]ourt will [have] to make the determination about what the calculations will be . . . the damages need to be established in this case, and the [c]ourt, based on the information it gets from the record and what is presented, will determine what the damages will be.

See J.A. 522. The trial judge also confirmed that, if the parties had desired to stipulate to damages if liability was established, they would have done so before trial — indicating that there would be no post-trial opportunity to challenge how the damages were calculated.

In these circumstances, the district court thoroughly assessed and rejected the merits of each of Steadfast’s post-verdict motions. See Su v. Medical Staffing of Am., LLC, No.

10

We have consistently recognized that “[t]o preserve an issue for appeal, an objection must be timely and state the grounds on which it is based. See Kollsman, a Div. of Sequa Corp. v. Cohen, 996 F.2d 702, 707 (4th Cir. 1993) (citing Fed. R. Civ. P. 46). We have explained that “a party does not go far enough by raising a non-specific objection or claim.” See In re Under Seal, 749 F.3d at 287 (4th Cir. 2014). Rather, “if a party wishes to preserve an argument for appeal, the party must press and not merely intimate the argument during the proceedings before the district court.” Id. (internal quotations omitted). We identify these precedents to show that Steadfast failed to properly object to the Initial damages computations.

27

2:18-cv-00226 (E.D. Va. Dec. 7, 2023), ECF No. 411 at 5-10. In addition to rejecting the merits of Steadfast’s multiple challenges under the Civil Rules, the court determined that Steadfast had waived its contentions regarding the Initial damages computations, in that Steadfast failed to make proper objections before or during trial. Id. at 11. In sum, the court denied relief on all grounds interposed by Steadfast, and it separately concluded that Steadfast had waived its right to attack the Initial damages computations.

d.

Soon thereafter, on December 8, 2023, the district court also resolved the Secretary’s October 1, 2023 motion for a final damages ruling that included unpaid overtime compensation and liquidated damages awards, that is, for approval of the “Final damages computations” supported in the Secretary’s April 20, 2023 Notice. That motion was resolved in yet another Opinion, in which the court “granted in part and dismissed in part” the Secretary’s pending motion. See Su v. Medical Staffing of Am., LLC, No. 2:18- cv-00226 (E.D. Va. Dec. 8, 2023), ECF No. 412.

In its December 8, 2023 Opinion, the district court emphasized that its Bench Verdict had already “accepted” the Secretary’s Initial damages computations and its “calculations and methodology.” It also explained that the Bench Verdict had instructed the Secretary to provide the court with updated computations for “any continuing violations.” The Secretary’s October 1, 2023 motion for a final damages ruling had requested that the court adopt the unpaid overtime compensation and liquidated damages amounts supported in the Final damages computations filed with the April 20, 2023 “Second Updated Back Wage Computations Notice.” The December 8, 2023 Opinion then

28

relied on and approved nearly all of the Secretary’s Final damages computations submitted with the April 20, 2023 Notice. The Final damages computations had relied on and utilized the methodology accepted by the Bench Verdict, and it was accompanied by a supporting Declaration of the DOL official that prepared those computations.

In making its December 8, 2023 decisions, the court sensibly divided into three separate time periods its unpaid overtime compensation awards and liquidated damages rulings. Those time periods were as follows:

• The “First Period” (August 2015-June 2021) was covered by the Initial damages computations.

• The “Second Period” (June 2021-January 2022) was from the Initial damages computations until approximately the Bench Verdict.

• The “Third Period” (January 2022-January 2023) ran from shortly after the Bench Verdict to the completion of the Final damages computations.

By its December 8, 2023 Opinion, the district court assessed and approved the Secretary’s Final damages computations of damages and awards for the First and Second Periods. For the First Period, the unpaid overtime compensation award against Steadfast was the sum of $3,619,716.49. The court then awarded the Secretary an equal amount of liquidated damages, for an aggregate First Period award against Steadfast of $7,239,432.98.

For the Second Period, the unpaid overtime compensation award against Steadfast was the sum of $917,926.34. The court again awarded an equal amount of liquidated damages to the Secretary, for an aggregate Second Period award against Steadfast of $1,835,852.68.

29

For the Third Period, the district court awarded the Secretary the sum of $271,377.57, confined to unpaid overtime compensation. The Secretary’s effort to secure liquidated damages from Steadfast for the Third Period was denied. As the court explained:

[T]he court will not award liquidated damages for the Third Period. A court can deny or reduce liquidated damages if the defendants can show they acted in good faith. 29 U.S.C. § 260. Here, Defendants began complying with the [c]ourt’s Order on February 8, 2022.

See Su v. Medical Staffing of Am., LLC, No. 2:18-cv-00226 (E.D. Va. Dec. 8, 2023), ECF No. 412 at 6.

3.

Three days thereafter, on December 11, 2023, the district court entered its final judgment, completing the court proceedings and record underlying these appeals. See Su v. Medical Staffing of Am., LLC, No. 2:18-cv-00226 (E.D. Va. Dec. 11, 2023), ECF No. 413. The final judgment was that Steadfast owed the Secretary, on behalf of the Steadfast nurses, the aggregate sum of $9,346,663.23 in unpaid overtime compensation and liquidated damages, with interest. That award represented $4,809,020.40 in unpaid overtime compensation, plus $4,537,642.83 in liquidated damages. In addition to those damages, the judgment again enjoined Steadfast from further violations of the FLSA, specifically incorporating the provisions of the injunction the court had imposed on September 7, 2023.

On December 11, 2023, Steadfast also noticed its appeal from the final judgment.

The two Steadfast appeals addressed herein, that is, Nos. 23-2176 and 23-2284, were then consolidated by our Court on December 19, 2023. We possess jurisdiction of those appeals pursuant to 28 U.S.C. § 1291.

30

III.

On appeal, Steadfast pursues five contentions of error. Those contentions are, first, that the district court misallocated the burden of proof on the worker classification issue, and in turn committed clear error on certain findings of fact; second, that the Bench Verdict’s analysis of the “control factor” conflicts with our worker classification precedent; third, that each of the other “economic realities” factors favor Steadfast; fourth, that Steadfast proved its good faith affirmative defense against an award of liquidated damages; and finally, that the court erred in accepting the Initial and Final damages computations.

Although Steadfast frames its appeal as a series of five contentions, it also recognizes that “the threshold liability question” presented here is whether the Steadfast nurses are “employees.” We are satisfied to first assess the district court’s ruling on the Steadfast nurses’ status as employees, addressing therein Steadfast’s contentions that the court misallocated the burden of proof, clearly erred in certain factual findings, and also erred in its assessment of our McFeeley factors. We will then assess and resolve Steadfast’s contention relating to its claim to a good faith affirmative defense to its misclassification of its nurses. Finally, we evaluate Steadfast’s contentions concerning the Initial and Final damages computations and other issues relating to unpaid overtime compensation and liquidated damages.

As explained herein, we reject each of Steadfast’s contentions of error and affirm the challenged judgment.

31

A.

1.

As a “remedial and humanitarian statute,” the FLSA seeks to “protect all covered workers from substandard wages and oppressive working hours.” See Salinas v. Commercial Interiors, Inc., 848 F.3d 125, 133 (4th Cir. 2017); Barrentine v. Arkansas- Best Freight Sys., Inc., 450 U.S. 728, 739 (1981). To address the concerns posed by such substandard wages, the Act mandates that employers pay overtime compensation — that is, at least one-and one-half times the regular hourly rate for hours worked in excess of 40 hours in a workweek — to those who are “covered workers.” See 29 U.S.C. § 207(a)(1); see also Helix Energy Sols. Grp., Inc. v. Hewitt, 598 U.S. 39, 44 (2023). “The FLSA, however, exempts certain categories of workers,” including those who are independent contractors, “from its protections, including the overtime-pay guarantee.” See Salinas, 848 F.3d at 149; Helix, 598 U.S. at 44; see also 29 U.S.C. § 207(a)(1). The FLSA conditions a putative employer’s liability for overtime compensation “on the existence of an employer-employee relationship, and the employee bears the burden of alleging and proving the existence of that relationship.” See Kerr v. Marshall Univ. Bd. of Govs., 824 F.3d 62, 83 (4th Cir. 2016).

To determine a worker’s status as an employee, the “courts look to the ‘economic realities’ of the relationship between the worker and the putative employer.” See McFeeley v. Jackson Street Ent., LLC, 825 F.3d at 241. This assessment is a fact-intensive inquiry, of which “[t]he focal point is whether the worker is economically dependent on the business to which he renders service or is, as a matter of economic reality, in business for

32

himself.” See Shultz v. Capital Int’l Sec., Inc., 466 F.3d 298, 304 (4th Cir. 2006) (internal quotation marks and alterations omitted); see also Bartels v. Birmingham, 332 U.S. 126, 130 (1947) (explaining that “in the application of social legislation employees are those who as a matter of economic reality are dependent upon the business to which they render service.”).

“This FLSA diagnosis is not always easy to perform.” See Brock v. Mr. W Fireworks, Inc., 814 F.2d 1042, 1043 (5th Cir. 1987); see also McFeeley, 825 F.3d at 241 (explaining that “the employee/independent contractor distinction is not a bright line but a spectrum”). Although there are no neat guidelines that exist, our Court has consistently endeavored to “capture the economic realities of the relationship between the worker and the putative employer” by applying a test designed to balance six factors, which are “derive[d] from the Supreme Court’s [1947] opinion in United States v. Silk.” See Salinas, 848 F.3d 125, 150 (4th Cir. 2017) (citing 331 U.S. 704 (1947)). As we specified in our McFeeley decision, those factors are:

(1) the degree of control that the putative employer has over the manner in which the work is performed;

(2) the worker’s opportunities for profit or loss dependent on his managerial skill;

(3) the worker’s investment in equipment or material, or his employment of other workers;

(4) the degree of skill required for the work;

(5) the permanence of the working relationship; and

(6) the degree to which the services rendered are an integral part of the putative employer’s business.

33

See 825 F.3d at 241.

Importantly, “[n]o single factor is dispositive — all six are part of the totality of the circumstances presented.” Id.; see also Schultz, 466 F.3d at 305. That is, each McFeeley factor must be considered with an eye towards the ultimate question: whether, against the backdrop of “the particular working relationship, the particular workplace, and the particular industry,” the “underlying economic realities lead to the conclusion” that the workers are actually employees. See McFeeley, 825 F.3d at 241; Rutherford Food Corp. v. McComb, 331 U.S. 722, 727 (1947).

2.

After carefully assessing the economic realities of the nurses’ relationship with Steadfast, the Bench Verdict concluded that they were employees of Steadfast. As a result, it then ruled that the Steadfast nurses were protected by the FLSA. Because the district court rendered its decision following a bench trial, we review its judgment “under a mixed standard of review — factual findings may be reversed only if clearly erroneous, while conclusions of law . . . are examined de novo.” See Universal Furniture Int’l, Inc. v. Collezione Europa USA, Inc., 618 F.3d 417, 427 (4th Cir. 2010) (internal quotation marks omitted); see also Fed. R. Civ. P. 52(a).

In applying this mixed standard of review, an appellate court is obliged to consider all of the “component parts” of the district court’s decision, “each under the appropriate standard.” See U.S. Bank Nat’l Ass’n ex rel. CWCapital Asset Mgmt. LLC v. Village at Lakeridge, LLC, 583 U.S. 387, 393 (2018). Bearing that principle in mind, we will initially

34

describe the three issues relating to worker classification — two factual and one legal — that were assessed by the district court. See, e.g., Mr. W Fireworks, 814 F.2d at 1044 (explaining that there are “three types of findings involved in determining whether one is an employee within the meaning of the Act”).

a.

To begin its analysis of the nurses’ employee status, the Bench Verdict appropriately made historical findings of fact. Those findings “address[ed] questions of who did what, when or where, how or why.” See U.S. Bank Nat’l Ass’n, 583 U.S. at 394. And we assess the court’s historical findings of fact only for clear error, affording the court “the highest degree of appellate deference.” See Evergreen Int’l, S.A. v. Norfolk Dredging Co., 531 F.3d 302, 308 (4th Cir. 2008); see also Fed. R. Civ. P. 52(a). Put simply, we are not entitled to reverse a factual finding unless we are “left with the definite and firm conviction that a mistake has been committed.” See Anderson v. Bessemer City, 470 U.S. 564, 573 (1985) (citing United States v. U.S. Gypsum Co., 333 U.S. 364, 394-95 (1948).

An application of “clear-error” review by a reviewing court “does not entitle [it] to reverse [a] finding of the trier of fact simply because it is convinced that it would have decided the case differently.” See Anderson, 470 U.S. 564, 573 (1985). Nor does our use of clear-error review entitle us to “duplicate the role of the lower court . . . to decide factual issues de novo.” Id. Rather, we assess only whether “the district court’s account of the evidence is plausible in light of the record viewed in its entirety.” See Walsh v. Vinoskey, 19 F.4th 672, 677 (4th Cir. 2021). And “[i]n cases in which a district court’s factual findings turn on assessments of witness credibility or the weighing of conflicting

35

evidence during a bench trial, such findings are entitled to even greater deference.” See Evergreen Int’l, 531 F.3d at 308.

b.

Next, the Bench Verdict made findings as to each of the McFeeley factors. See McFeeley, 825 F.3d at 241. Resolving those issues required the court to “marshal and weigh evidence, make credibility judgments, and otherwise address” factual disputes. See U.S. Bank Nat’l Ass’n, 583 U.S. at 396. Put otherwise, the court engaged in extensive “factual work” — it took “a raft of case-specific historical facts, consider[ed] them as a whole, and balance[d] them one against another.” Id. at 397. The court’s findings as to “[t]he existence and degree” of the McFeeley factors are therefore primarily factual, and we thus review them only for clear error. See, e.g., Brock v. Superior Care, Inc., 840 F.2d 1054, 1059 (2d Cir. 1988). And we afford those findings significant deference, applying the principles of clear-error review spelled out above. See, e.g., Hobbs v. Petroplex Pipe & Constr., 946 F.3d 824, 829 (5th Cir. 2020).

c.

Finally, the district court in the Bench Verdict resolved the legal question of whether the Steadfast nurses were, in the final analysis, “employees” or “independent contractors” of Steadfast under the FLSA. And that final inquiry — as we have held — is “ultimately a legal question” that is “subject to de novo review.” See McFeeley, 825 F.3d at 240.

B.

The district court determined, based on the pertinent historical facts, that the totality of the McFeeley factors weighed in favor of employee status for the Steadfast nurses. See

36

Bench Verdict 27. Although the Bench Verdict recognized that “the degree and skill (via education and licensures) required for the nurses’ work suggests that the nurses could work in an independent capacity,” it concluded that the other McFeeley factors “outweigh[ed] this factor significantly.” Id. On balance, the Bench Verdict found that the employment relationship between the nurses and Steadfast was such that, “[u]ltimately, under the FLSA, the nurses on [Steadfast’s] registry” were “employees, not independent contractors.” Id.

Steadfast, as the defendant here, and the Secretary, as the plaintiff, have offered competing narratives of the relationship between Steadfast and its nurses. Overarching Steadfast’s various contentions, however, is its bottom-line assertion that the trial court misallocated and misapplied the burden of proof applicable to the “threshold liability question” being presented: That is, whether the Steadfast nurses are “employees” or “independent contractors.” Steadfast is correct when it says that the Secretary bore the burden of establishing an employment relationship with respect to the Steadfast nurses that gives rise to liability under the FLSA. See Kerr, 824 F.3d at 83. And Steadfast maintains on appeal that the court erred in that respect and instead imposed upon Steadfast the burden of proving that its nurses were independent contractors. See Br. of Appellants 24. Based on this contention, Steadfast urges us to disregard the Bench Verdict’s factual findings and its liability ruling against Steadfast.

Steadfast argues, inter alia, that the district court erred concerning the burden of proof, which was “invited by” an erroneous assertion in the Secretary’s proposed post-trial conclusions of law. See Br. of Appellants 21-22. The Secretary therein had suggested that “[a]s an affirmative defense, [Steadfast] bears the burden of proving” that the nurses were

37

independent contractors. See J.A. 1976. But the Bench Verdict neither referenced nor adopted this erroneous proposition. As a result, the Secretary’s claim of error on this point is irrelevant.

Among Steadfast’s other contentions on appeal are that the court erred (1) by relying on “outlier testimony” or “isolated incidents” from witness testimony; (2) by failing to rely on controlling authorities supporting the Bench Verdict’s factual analysis; and by (3) failing to “even mention” the burden of proof. Our review of the record, however, readily dispels these contentions. They are simply an effort by Steadfast to magnify purported “errors” well beyond their scope. Steadfast’s burden of proof contention, therefore, amounts to nothing more than a disagreement with the Bench Verdict’s ultimate conclusion. Indeed, Steadfast fails to identify any specific instance — in the Bench Verdict or otherwise — where the district court failed to apply or allocate the proper burden of proof.

Put simply, our review establishes that the district court made its findings of fact in the Bench Verdict on the trial evidence, carefully assessed those facts under the McFeeley economic realities test, and then concluded that the Steadfast nurses were employees under the FLSA. That analysis was exactly what the court was obliged to conduct in order to properly resolve the issues and render its Bench Verdict. See Fed. R. Civ. P. 52(a)(1) (“[T]he court must find the facts specially and state its conclusions of law separately.”). In the circumstances, we reject Steadfast’s contention concerning the burden of proof, and thus turn to our separate assessment of the court’s findings and conclusions in the Bench Verdict.

38

1.

In separately assessing the McFeeley economic realities factors, we will initially consider the degree of control Steadfast exercised over the manner in which its nurses’ work was performed. See Schultz, 466 F.3d at 307. Central to this inquiry is whether the employer “control[s] the meaningful economic aspects of the business,” such that “the worker has a viable economic status that can be traded . . . to other companies.” See Mr. W Fireworks, 814 F.2d at 1049; Parrish v. Premier Directional Drilling, L.P., 917 F.3d 369, 381 (5th Cir. 2019) (internal quotation marks omitted).

“To guide this evaluation,” a court must “ask whether the company ‘retains the right to dictate the manner’ of the worker’s performance.” See Acosta v. Off Duty Police Servs., Inc., 915 F.3d 1050, 1060 (6th Cir. 2019) (quoting Donovan v. Brandel, 736 F.2d 1114, 1119 (6th Cir. 1984)). Importantly, a “lack of supervision over minor regular tasks cannot be bootstrapped into an appearance of real independence.” See Parrish, 917 F.3d at 381. Rather, a court conducting such an analysis should assess the control factor through the “focal point” of economic dependence: That is, “whether the worker has control over such a meaningful portion of his labor that he operates as a separate economic entity.” See Schultz, 466 F.3d at 304; Brant v. Schneider Nat’l, Inc., 43 F.4th 656, 666 (7th Cir. 2022).

a.

After recognizing that “[n]o single factor in the six-factor [McFeeley] test is dispositive,” the Bench Verdict stated that “[t]he degree of control a putative employer has over the way an alleged employee’s work is performed is the most important factor” in the economic realities test. See Bench Verdict 18 (citing Smith v. CSRA, 12 F.4th 396, 413

39

(4th Cir. 2021)). As a clarification, however, our Court has only recognized that control may be the most important factor in an economic realities test. See, e.g., McFeeley, 825 F.3d at 241 (“Most critical on the facts of this case is . . . the degree of control that the putative employer has over the manner in which the work is performed.”). We readily emphasize, however, that the McFeeley economic realities test “allows for flexible application” to the facts of each employment relationship. In the context of these proceedings, there is ample evidence that the control factor weighed heavily in favor of the Steadfast nurses’ employee status. We therefore cannot rule that the district court erred in placing too much emphasis on the degree of control that Steadfast exercised over its nurses, or in its balancing of the control factor under our economic realities test. See Bench Verdict 21-26.

We thus agree with the district court that the control factor points towards an employment relationship between Steadfast and its nurses. Although Steadfast lacked control over certain specific aspects of the nurses’ work — such as day-to-day patient care tasks supervised by a client facility’s medical professionals — the evidence amply supports the proposition that the control factor weighs heavily in favor of their employee status. As explained herein, the court’s control finding was not clearly erroneous.

b.

To further explain, the Bench Verdict found numerous “plain manifestations of [Steadfast’s] control over the nurses.” See McFeeley, 825 F.3d at 241-42. Perhaps most significantly, the court found that Steadfast unilaterally fixed the nurses’ hourly wages. As the Bench Verdict explained and emphasized, Steadfast’s wage-setting protocol meant that

40

the nurses’ ability to negotiate their pay rates with Steadfast was “limited by the extent to which changing the pay rate would benefit [Steadfast.]” See Bench Verdict 23.

Indeed, the evidence established that Steadfast’s wage-setting practices precluded any negotiations or alternative arrangements by the nurses. That is, the nurses lacked any opportunity to set or negotiate their hourly rates with Steadfast or its client facilities. Instead, Steadfast unilaterally determined the hourly rates that the client facilities paid to Steadfast for the nurses’ work. And, in turn, Steadfast retained a portion of each nurses’ hourly rate.

To be sure, the evidence was that the Steadfast nurses “could not bid for jobs or negotiate the prices for jobs.” See Scantland v. Jeffry Knight, Inc., 721 F.3d 1308, 1315 (11th Cir. 2013). And that fact weighs heavily in favor of an employment relationship. So, too, does the fact that Steadfast “set the rate at which the [nurses] were paid.” See Acosta, 915 F.3d at 1060. Indeed, our court has heretofore recognized that setting a worker’s wages, without affording those workers an opportunity to negotiate, is evidence that an employer exercises “significant control” over its workers. See McFeeley, 825 F.3d at 241-42.

We simply cannot say — as Steadfast would have it — that the Bench Verdict’s alleged “preoccupation with pay” contravenes our precedent. See Br. of Appellants 38. We instead emphasize that a proper application of our McFeeley economic realities test required the trial court to examine the “totality of the circumstances,” including all aspects of the economic relationship between the employer and its workers. See McFeeley, 825

41

F.3d at 241; Schultz, 466 F.3d at 305. In these circumstances, the district court did not err in the emphasis it placed on the evidence of Steadfast’s control over the nurses’ pay.

c.

The district court also found that Steadfast exercised extensive control over the scheduling of its nurses’ shifts, completely controlling “what information, if any, nurses receive[d] regarding available shifts at client facilities.” See Bench Verdict 22. The trial record demonstrates that Steadfast unilaterally determined which individual nurses it would notify about available shift opportunities. Steadfast then communicated directly with those nurses through phone calls, emails, text messages, or its “Zira app.” Id. at 6-7. Apart from establishing what shift opportunities were available to the nurses, the Bench Verdict found that Steadfast retained control over portions of the nurses’ day-to-day schedules. Nurses were prohibited from communicating with client facilities, and were instead required to notify Steadfast if they could not complete a shift or if they would be late to a shift. Id. at 22.

Nevertheless, Steadfast argues on appeal that its scheduling system functioned as a type of open marketplace, where its nurses were able to review lists of available shifts at various client facilities, and then determine whether to accept a shift opportunity. See Br. of Appellants 27. It also claims that this free-market system allowed the Steadfast nurses to have significant freedom in setting their own work schedules, in that the nurses could independently choose to accept or decline shifts. But Steadfast’s characterization of its scheduling system glosses over the very compelling fact that, ultimately, Steadfast controlled the nurses’ access to available shifts, thereby controlling when and where the

42

nurses worked. And Steadfast ignores the proposition that day-to-day supervision is not a prerequisite for control. See, e.g., Brandel, 736 F.2d at 1119 (explaining that the control inquiry asks whether an employer “retains the right to dictate the manner in which” work is executed).

d.

Beyond the control that Steadfast exercised over the nurses’ schedules, the Bench Verdict found that the evidence demonstrated that Steadfast retained control over various aspects of the nurses’ work — even though the work was performed at the client facilities and Steadfast personnel were not present at those client facilities when the nurses were on duty. But supervision doesn’t require a shadow. Although Steadfast was not “look[ing] over [its] workers’ shoulders every day,” the evidence clearly demonstrates that Steadfast engaged in other actions consistent with exercising significant supervisory control over its nurses. See Brock, 840 F.2d at 1060; see also Acosta, 915 F.3d at 1061. It required its nurses to track their work hours using Steadfast’s timesheets, obtain client-facility signatures to verify those work hours, and submit those timesheets directly to Steadfast. Steadfast also issued explicit instructions to its nurses regarding professional conduct and workplace expectations and disseminated written guidance to its nurses covering topics such as proper attire, punctuality, and timekeeping procedures.

To be sure, we have recognized that “a company that engages an independent contractor seeks to exert some control . . . over the performance of the contractor’s duties,” such as defining the terms of the work and agreeing to the price of the work or the method of payment. See McFeeley, 825 F.3d at 242-43. But we have distinguished that type of

43

control — specifically, that which is generally accepted as “part and parcel” of a bargain “between parties whose independent contractual status is not in dispute” from the very “significant control” that Steadfast exercised here. See McFeeley, 825 F.3d at 242-43.

The record amply supports the Bench Verdict’s finding that Steadfast exercised extensive control over the nurses’ manner of work. Steadfast’s client facilities were contractually obligated to provide feedback concerning the Steadfast nurses’ work performance, including any reports of deficiencies or noncompliance with Steadfast’s policies. Indeed, it was Steadfast — and not the client facilities — that ultimately monitored the nurses’ performance at client facilities and enforced performance standards. See Bench Verdict 25 (“Testimony from several nurses and client-facility representatives establish that [Steadfast is] responsible for supervising and disciplining nurses.”).

If a nurse was the subject of a complaint from a client facility — regardless of whether that client facility requested their removal — Steadfast alone decided whether that nurse could continue working. In those circumstances, Steadfast routinely disciplined nurses for infractions of its internal policies. Indeed, Steadfast personnel revealed that Ms. Pitts regularly disciplined the nurses — and even in some instances verbally berated them — for conduct such as discussing compensation with other nurses, contacting client facilities directly, or declining shifts.

This finding is further supported by trial testimony from Steadfast nurses, who explained that the Steadfast nurses were subject to discipline not only for failing to comply with Steadfast’s standards of performance, but also for declining to accept a shift opportunity or cancelling a shift. The nurses detailed the various disciplinary measures

44

that Steadfast used to reprimand those nurses, including placing them on a “do not return” list at client facilities, cancelling shifts or blocking access to assignments, or removing a nurse from the registry entirely. Id. at 26 (“[S]everal nurses testified that [Steadfast] disciplined nurses by cancelling nurses’ shifts or otherwise diminishing their ability to obtain work.”). At bottom, the economic realities of the entire arrangement — as evidenced by the trial testimony and the Bench Verdict’s findings — establish that Steadfast conditioned continued access to shift opportunities on its nurses’ strict adherence to Steadfast’s rules and directives.

Such a level of disciplinary oversight is entirely inconsistent with Steadfast’s contention that it merely facilitated connections between its client facilities and its nurses and, in doing so, did not exercise the type of operational control characteristic of an employer-employee relationship. Our review of the record convinces us that, as the Bench Verdict determined, Steadfast controlled the “meaningful economic aspects of the business,” such that the Steadfast nurses were not considered as “separate economic entities.” See Hopkins v. Cornerstone Am., 545 F.3d 338, 343 (5th Cir. 2008). In this respect, the trial evidence overwhelmingly demonstrated that Steadfast maintained direct and continuous oversight over its nurses, including their assignments, their pay, their work performance, and their discipline. See McFeeley, 825 F.3d at 242. As the Bench Verdict correctly recognized, the authority Steadfast wielded over its nurses was fundamentally incompatible with an independent contractor relationship, where a worker typically has control over his own business and client relationships. Id. at 241.

45

Thus, in light of that record evidence, the Bench Verdict did not clearly err in finding that the degree of control factor favored the employee status of the Steadfast nurses.

2.

We will next assess the Bench Verdict’s findings concerning the other McFeeley factors — that is, the nurses’ opportunities for profit or loss; the nurses’ investment; the nurses’ degree of skill; the permanence of the nurses’ working relationship with Steadfast; and the degree to which the nurses’ work was integral to Steadfast’s business. Of those factors, the Bench Verdict found that only one — the nurses’ degree of skill — plausibly weighed in favor of an independent contractor status. See Bench Verdict 27. The Bench Verdict found, however, that the remaining four factors weighed in favor of the nurses being properly classified as “employees” of Steadfast. And in assessing the totality of the nurses’ economic dependence on Steadfast, the Bench Verdict determined that those other factors significantly outweighed the single factor relating to the degree of skill.

At bottom, we discern no error with the employee ruling made by the Bench Verdict, and we therefore agree with the district court that the trial evidence established that virtually all of the McFeeley factors weigh in favor of the nurses’ employee status.

a.

The second factor under the economic realities test, as set forth in McFeeley, is whether “the worker’s opportunities for profit or loss is dependent on his managerial skill.” See Schultz, 466 F.3d at 305. Put simply, “[t]he more the worker’s earnings depend on his own managerial capacity rather than the company’s,” the “less the worker is ‘economically dependent on the business’ and the more he is ‘in business for himself’” — or, in other

46

words, the more likely that he is an independent contractor. See McFeeley, 825 F.3d at 243 (quoting Henderson v. Inter-Chem Coal Co., Inc., 41 F.3d 567, 570 (10th Cir. 1994)).

Importantly, “[a]n individual’s ability to earn more by being more technically proficient is unrelated to an individual’s ability to earn or lose profit via his managerial skill.” See Scantland, 721 F.3d at 1316. The same is true when a “putative employer’s control over the worker’s schedule and pay ha[s] the effect of limiting the worker’s opportunity . . . for profit or loss.” See Hobbs, 946 F.3d at 832. As such, when a worker’s “opportunity for profit [is] limited to their ability to complete more jobs than assigned,” it can hardly be said that the worker “operates his own business” or utilizes “the initiative, judgment or foresight of the typical independent contractor.” See Scantland, 721 F.3d at 1317; see also Rutherford, 331 U.S. at 730.

Consistent with these principles, the Bench Verdict determined that the nurses had no opportunities for profit or loss in their relationship with Steadfast. As the Bench Verdict explained, there was “no evidence that the nurses [had] an interest in Steadfast beyond their role as workers for a set hourly pay rate.” The Bench Verdict also recognized that the non- compete clause in the nurses’ contracts “significantly hinder[ed] the nurses’ ability to accumulate profit independent of [Steadfast].” See Bench Verdict 26. It thus found that the Steadfast nurses’ lack of opportunity for profit or loss weighed in favor of their employee status. We are constrained to agree.

Steadfast, on the other hand, contends on appeal that its nurses had real opportunities for profit and loss by negotiating their compensation according to market conditions — such as the availability of shift opportunities or specific assignments. But the trial evidence

47

was that the nurses could not “increase their take-home pay through their own ingenuity” or managerial skill. See Hall v. DirectTV, LLC, 846 F.3d 757, 774 (4th Cir. 2017). Indeed, as the Bench Verdict found, the Steadfast nurses only means of increasing income was by working more hours. See Bench Verdict 26. As we have recognized, the argument “that [nurses] can ‘hustle’ to increase their profits” — that is, take on more shifts — “has been almost universally rejected.” McFeeley, 825 F.3d at 243. Such conduct is simply “not managerial.” Id.

Moreover, the evidence established that the nurses “did not share in [Steadfast’s]

profits.” See Dole v. Snell, 875 F.2d 802, 809 (10th Cir. 1989). Rather, the Bench Verdict specifically found that the Steadfast nurses were paid fixed hourly rates that were unilaterally determined by Steadfast. See Bench Verdict 23. As we have explained, the Steadfast nurses could not negotiate their wages, nor could they bargain for compensation based on their efficiency, skill, or business acumen. The Steadfast nurses, as a result, “had no input whatsoever into the amount which they were to receive” for each shift opportunity, “except on the rare occasion where an unusual [assignment] was involved.” See Snell, 875 F.2d at 809. Put otherwise, Steadfast “ultimately controlled a key determinant — pricing — affecting [the nurses’] ability to make a profit.” See McFeeley, 825 F.3d at 243.

Steadfast also disputes whether its nurses were permitted to work for other healthcare agencies and healthcare providers. Although Steadfast acknowledges that its contracts contained the non-compete clause, it argues that “the testimony at trial established [that] this clause was not enforced by Steadfast.” See Br. of Appellants 9 n.2. Steadfast thus contends that the district court failed to adequately analyze the “economic realities of

48

the actual working relationship” in determining that the non-compete clause weighed in favor of finding that the nurses did not have an opportunity to meaningfully profit from their relationship with Steadfast.

Steadfast’s argument concerning its non-compete clause fails, for “it is not what the [parties] could have done that counts, but as a matter of economic reality what they actually do that is dispositive.” See Mr. W Fireworks, 814 F.2d at 1047. The trial evidence of the Steadfast nurses established that they believed this clause restricted them from working for other healthcare agencies or providers. See, e.g., J.A. 301-03. And while Steadfast claims that some nurses also testified that they worked for other provider registries — either close in time or parallel to their work with Steadfast — the Bench Verdict’s factual findings reflect its careful consideration of the witnesses’ credibility. As a result, the Bench Verdict’s determination that the non-compete clause prevented the Steadfast nurses from seeking other employment must be credited. Embracing Steadfast’s argument on the non-compete clause would, therefore, be inconsistent with our approach to the economic realities of the nurses’ relationship with Steadfast. See, e.g., McFeeley, 825 F.3d at 241. It would also contravene the significant deference we must afford to the district court’s factual findings. See, e.g., Evergreen Int’l, 531 F.3d at 308.

In sum, both the trial evidence and the Bench Verdict’s thorough evaluation thereof demonstrate that the Steadfast nurses “are far more closely akin to wage earners toiling for a living, than to independent entrepreneurs seeking a return on their risky capital investments.” See Mr. W Fireworks, 814 F.2d at 1051. That evidence not only establishes that Steadfast largely controlled the nurses’ opportunity for profit and loss, it shows

49

convincingly that the “lion’s share of managerial skill . . . came from [Steadfast].” See McFeeley, 825 F.3d at 243. We therefore agree with and accept the ruling of the Bench Verdict that this factor illustrates the nurses’ economic dependence on Steadfast and supports their classification as “employees.”

b.

Another hallmark of independent contractor status is whether a worker makes significant investments in equipment or hires others to assist in performing the work. See McFeeley, 825 F.3d at 241. To be sure, the more a worker “is personally invested in the capital and labor of the enterprise, the less the worker is ‘economically dependent on the business.’” Id. (quoting Henderson, 41 F.3d at 570). But an individual worker’s investment is only one-half of the inquiry: “[I]n order to determine the degree of economic dependence,” the worker’s “relative investment” must be compared with the investment of the putative employer. See Mr. W Fireworks, 814 F.2d at 1052.

Steadfast contends that the investment factor weighs in favor of its nurses’

independent contractor status, because the nurses paid for their “initial and continuing education, licensure, and equipment” plus “the financial investment” necessary for “establishing and maintaining their ability to provide nursing services.” See See Br. of Appellants 47. However, the fact that the Steadfast nurses had been licensed by their respective state entities is “immaterial” to this inquiry, because such licenses are required for all nurses “regardless of whether they are employees or independent contractors.” See Schultz, 466 F.3d at 308.

50

Nevertheless, we recognize that the evidence of certain Steadfast nurses shows that some of them preferred to use their own personally-procured tools, such as stethoscopes and blood pressure cuffs. We know, however, “that the fact that a worker supplies his or her own tools or equipment does not preclude a finding of employee status.” See Snell, 875 F.2d at 810 (collecting cases); see also Rutherford, 331 U.S. at 725 (finding that meat boners were employees despite “own[ing] their own tools”). Indeed, it is not at all uncommon for workers to have invested in negligible “equipment or materials required for the task,” like a stethoscope or blood pressure cuff. See Schultz, 466 F.3d at 308 (“Although some agents chose to use their own firearms . . . . The agents were thus not required to invest in any equipment or materials.”). But those purchases were not the type of “large capital expenditures, such as risk capital and capital investments” that are dispositive of our classification inquiry. See Snell, 875 F.2d at 802; see also McFeeley, 825 F.3d at 243 (explaining that employer’s investment included rent, utility bills, business liability insurance, and advertising costs).

Rather, in the context of the relative-investment inquiry, “substantial investments”

means expenses associated with, inter alia, “maintaining corporate offices, printing brochures and contracts, providing accounting services, and developing and underwriting insurance products.” See Hopkins, 545 F.3d at 344. The trial evidence established that the Steadfast nurses made no such investments. The Bench Verdict, therefore, correctly found that this factor weighed in favor of the nurses’ employee status. See Bench Verdict 26-27.

And our review of the trial record leads us to agree with the Bench Verdict’s relative-investment finding. Steadfast — and not its nurses — handled all administrative

51

costs, plus costs of advertising, providing the nurses’ insurance coverage, and contracting with client facilities. And Steadfast alone bore the costs associated with renting corporate office space. See, e.g., Silk, 331 U.S. at 716 (clarifying that “investment” factor refers to “investment in facilities”). Moreover, Steadfast “prohibit[ed] nurses from hiring other nurses, employees, or contractors to work for them at client-facilities.” See Bench Verdict 27; see also Schultz, 466 F.3d at 308 (observing that employees “could not hire other workers to help them do their work”).

As a matter of economic reality, the minimal relative investment required of the Steadfast nurses is in stark contrast to the significant capital investments normally made by independent contractors. See, e.g., Brock v. Superior Care, Inc., 840 F.2d 1054, 1059 (2d Cir. 1988) (finding home healthcare workers to be employees where they made no substantial capital investment). Steadfast’s “greater overall investment in the business scheme” convinces us that the Bench Verdict properly ruled that this factor weighs in favor of the Steadfast nurses’ employee status. See Hopkins, 545 F.3d at 344; see also Schultz, 466 F.3d at 308.

c.

Our economic realities test next assesses “the degree of skill required for the work.”

See McFeeley, 825 F.3d at 244. As an initial matter, there is no dispute that the Steadfast nurses are highly-skilled workers. Indeed, the Bench Verdict found that “[t]he degree of skill (via education and licensures) required for the nurses’ work suggests that the nurses could work in an independent capacity.” See Bench Verdict 27. And it would defy common sense to disagree with such a determination, as the Steadfast nurses were required

52

to complete extensive education and obtain licensing. We therefore agree that this factor — standing alone — tends to weigh towards independent contractor status.

We underscore, however, that the worker’s economic dependence on the putative employer is central to the degree of skill inquiry. That is, it is not only whether the workers “have some unique skill set” but also whether they possess “some ability to exercise significant initiative within the business.” See Parrish, 917 F.3d at 387 (collecting cases). In other words, the relevant inquiry includes an assessment of whether the workers use their skills “in any independent way” to secure business opportunities or “find job assignments.” See Superior Care, 840 F.2d at 1060.

Based on the trial record, we cannot say that the Steadfast nurses exercised their nursing skills in a way that reflected business initiative. As we have recognized, the trial evidence demonstrated that Steadfast controlled all information regarding available shift opportunities and restricted the Steadfast nurses from communicating directly with the client facilities. The nurses were thus economically dependent on Steadfast for “job assignments” or shift opportunities. See Superior Care, 840 F.2d at 1060. And, in any event, the Bench Verdict explained that “the economic realities of the nurses’ relationship with [Steadfast] under the other factors outweigh this factor significantly.” See Bench Verdict 27. Assessing the record evidence concerning the degree of skill factor, we discern no error with the district court’s determination.

d.

The final factors assessed under our McFeeley economic realities test are the permanence of the working relationship, and the degree to which the services rendered by

53

the worker are an integral part of the putative employer’s business. See 825 F.3d at 241. The Bench Verdict determined that these two factors weighed in favor of employee status on the part of the Steadfast nurses. See Bench Verdict 27. But because Steadfast does not contest that the nurses’ work is integral to Steadfast’s business, we need only address the Bench Verdict’s finding concerning the permanence of the nurses’ relationship with Steadfast. See Br. of Appellants 44, 49.

With respect to the Bench Verdict’s finding that the nurses’ relationship with Steadfast was “permanent in nature,” Steadfast argues that the nurses’ relationship with Steadfast was instead “open-ended” and “transient in nature.” See Br. of Appellants 48- 49. Steadfast asserts that the nurses “could accept or reject shifts as they pleased” and move between “Steadfast and competing agencies” or registries, suggesting a lack of permanence. Id. at 32, 48.

It is true that “[t]he more permanent the relationship, the more likely the worker is to be an employee.” See Schultz, 466 F.3d at 309. And “[g]enerally, independent contractors have variable or impermanent working relationships,” with employers.” See Keller v. Miri Microsystems LLC, 781 F.3d 799, 807 (6th Cir. 2015). As a result, those independent contractors “‘often have fixed employment periods and transfer from place to place as particular work is offered to them, whereas employees usually work for only one employer and such relationship is continuous and indefinite in duration.’” Id. (citing Baker v. Flint Eng’g & Constr. Co., 137 F.3d 1436, 1442 (10th Cir. 1998)).

But while “the length and regularity of the working relationship between parties” is important, it “is not dispositive of independent contractor status.” See Baker, 137 F.3d at

54

1442 (explaining that “the short duration” of employment is “clearly due to the intrinsic nature of” the work); Superior Care, 840 F.2d at 1060 (finding that nurses’ “lack of permanence is due to operational characteristics intrinsic to the industry”). Nor does a permanent relationship require exclusivity: Courts have routinely concluded that “employees may work for more than one employer without losing benefits under the FLSA.” See Keller, 781 F.3d at 808; see also Acosta, 915 F.3d at 1058.

To be sure, the evidence was that many of Steadfast’s contracts with the nurses included a term limit on their employment. But the evidence also shows that the Steadfast nurses depended on Steadfast “for their continued employment” during the term of their contracts. See Halferty v. Pulse Drug Co., 821 F.2d 261, 267-68 (5th Cir. 1987) (citing Mr. W Fireworks, 814 F.2d at 1054; see also Sec’y of Labor v. Lauritzen, 835 F.2d 1529, 1537 (7th Cir. 1987) (“[H]owever temporary the relationship may be it is permanent and exclusive for the duration of that [term].”). Indeed, the Bench Verdict explained that the nurses’ pay records and testimony demonstrated a lasting relationship, “even if a term limit is stated in a nurse’s contractor agreement.” See Bench Verdict 27. Our review thus concludes that the “consistent theme” of the trial evidence plainly established that the Steadfast nurses’ work “is not the kind of itinerant work that independent contractors ordinarily perform.” See Acosta, 915 F.3d at 1058-59.

Moreover, as we have discussed — and as the Bench Verdict recognized — the non-

compete clause in the Steadfast nurses’ contracts limited their ability to work for other agencies. Apart from contradicting Steadfast’s assertion that its nurses purportedly moved between agencies with some regularity, this fact reinforces the existence of a permanent

55

relationship — because “even short, exclusive relationships between the worker and the company may be indicative of an employee-employer relationship.” See Keller, 781 F.3d at 807. The economic realities of the Steadfast nurses’ employment relationship, therefore, support the Bench Verdict’s finding that the nurses’ relationship with Steadfast was permanent in nature. We do not find any error with the Bench Verdict’s permanence finding, and we agree that this factor supports employee status for the Steadfast nurses.

3.

Recall that we are tasked not only with assessing whether the trial court erred in its factual findings as to our McFeeley factors, but also with determining whether the workplace and the particular industry presented in these proceedings lead to the conclusion that the Steadfast nurses were employees of Steadfast. See McFeeley, 825 F.3d at 241. This analysis requires us to balance each factor of our Court’s McFeeley economic realities test against the others, keeping in mind that “no single factor is dispositive” of an employee-employer relationship. Id. at 241. Ultimately, the “determination of that relationship does not depend on [those] isolated factors but rather upon the circumstances of the whole activity.” See Rutherford, 331 U.S. at 730.

In undertaking this analysis, we also heed the Supreme Court’s warning that the FLSA “must not be interpreted in a narrow, grudging manner.” See Tenn. Coal, Iron & R.R. Co v. Muscoda Local No. 123, 321 U.S. 590, 597 (1944), superseded in part by statute, 29 U.S.C. § 254(a) (1947). Instead, we must be guided by the FLSA’s “remedial and humanitarian” purpose, which demands that it be “broadly interpreted and applied” to fulfill its goal of “protect[ing] the rights of those who toil, of those who sacrifice a full

56

measure of their freedom and talents to the use and profit of others.” See Salinas, 848 F.3d at 140; Benshoff v. City of Va. Beach, 180 F.3d 136, 140 (4th Cir. 1999) (internal citations omitted).

As explained herein, at least five of the six economic reality factors in our McFeeley test support the existence of an employment relationship between Steadfast and the Steadfast nurses. The trial evidence demonstrates that Steadfast exercised an enormous degree of control over the Steadfast nurses, thereby retaining “the right to dictate the manner of the nurses’ performance.” See Brandel, 736 F.2d at 1119. It further establishes that the Steadfast nurses could not “exercise or hone” their managerial skills to increase their pay, that the nurses invested minimally in capital expenditures or equipment for their work, that the relationship between Steadfast and its nurses was permanent in nature, and that the nurses’ work was integral to Steadfast’s business. See Schultz, 466 F.3d at 307; see also McFeeley, 841 F.3d at 240-41. And although we recognize that the degree of the nurses’ skill may weigh in favor of an independent contractor status, we readily agree that this single factor does not — and cannot — outweigh the other factors. Nor does it compel a contrary conclusion.

Our balancing of the McFeeley economic realities factors leads us to the very conclusion made by the Bench Verdict: The totality of the circumstances clearly establishes an employer-employee relationship between the Steadfast nurses and Steadfast. We conclude, therefore, that the nurses are entitled to overtime wages under the FLSA.

57

C.

Having assessed and resolved Steadfast’s liability as an employer under the FLSA, we turn to Steadfast’s contentions regarding the district court’s damages assessments and the final judgment. Steadfast’s challenges to the damages rulings are twofold. Steadfast first contends that it was entitled to the good faith defense against liquidated damages. Second, Steadfast challenges the court’s adoption of the Initial damages computations and the Final damages computations. We review each of those contentions for an abuse of discretion.

1.

Steadfast maintains that the district court erred when it ruled that Steadfast was not entitled to a good faith defense for the First and Second Periods, i.e., August 2015 to June 2021 and June 2021 to January 2022. That ruling subjected Steadfast to an assessment of statutory liquidated damages for each of those Periods. Because the court’s determination that Steadfast acted in good faith in the Third Period is not challenged by the Secretary, we evaluate the good faith issue only as to the First and Second Periods.11 a.

The FLSA is unambiguous: An employer that violates the Act’s overtime provisions is liable not only for unpaid overtime wages but also for “an additional equal amount as liquidated damages.” 29 U.S.C. § 216(c). This provision is intended to ensure

11

Due to Steadfast’s efforts to comply with the Bench Verdict in early 2022, as confirmed by a third-party auditor, the district court awarded Steadfast the good faith defense for the Third Period.

58

that employees receive full redress for unpaid wage violations — a policy that reaches to the heart of the FLSA’s remedial design. See United States v. Edwards, 995 F.3d 342, 346 (4th Cir. 2021) (explaining that liquidated damages serve to ensure that employees receive full compensation for FLSA violations); see also Brooklyn Sav. Bank v. O’Neil, 324 U.S. 697, 707 (1945). As we have explained, the FLSA’s “provision for liquidated damages is an additional penalty on non-compliant employers.” See McFeeley, 825 F.3d at 245. The good faith provision exists not just to make employees whole, but also to deter employers from gambling on noncompliance — that is, as we have said, from “rolling the dice by underpaying employees, reasoning all the while that it would be no worse off even if the employees eventually prevailed in court.” McFeeley, 825 F.3d at 245. The good faith provision is designed to make such a strategy a losing bet.

Though the stakes are high, liquidated damages under the FLSA are not an extraordinary remedy; they are the “norm” for FLSA overtime compensation violations. Edwards, 995 F.3d at 346 (internal quotation marks omitted). The FLSA nevertheless leaves room for an exercise of a court’s discretion. Such a court may reduce or deny an award of liquidated damages if the employer can sustain its burden on two requirements — that the employer acted in “good faith,” and that it “had reasonable grounds for believing that [its] act or omission was not a violation of the [FLSA].” See 29 U.S.C. § 260; see also See Burnley v. Short, 730 F.2d 136, 140 (4th Cir. 1984) (explaining that employer bears “plain and substantial burden” of demonstrating good faith defense). Because a decision to reduce or deny a liquidated damages award rests in the district court’s discretion, our review is limited to assessing whether such discretion has been abused. See Perez v.

59

Mountaire Farms, Inc., 650 F.3d 350, 375 (4th Cir. 2011). (“We review the district court’s decision denying liquidated damages for abuse of discretion.”).

Critically, an employer’s “good faith” and “reasonable grounds for believing” it had not violated the FLSA are each measured objectively. See 29 C.F.R. § 790.22(c); see also Clifton D. Mayhew, Inc. v. Wirtz, 413 F.2d 658, 661-62 (4th Cir. 1969). That means “an employer may not simply remain blissfully ignorant of FLSA requirements” by taking “an ostrichlike approach to the Act.” See Burnley, 730 F.2d at 140; Roy v. Cty. of Lexington, S.C., 141 F.3d 533, 548-49 (4th Cir. 1998). Indeed, “[i]f mere assumption amounted to good faith and reasonable belief of compliance, no employer would have any incentive to educate itself and proactively conform to” the FLSA. See McFeeley, 825 F.3d at 245.

Our sister circuits have consistently agreed that good faith under the FLSA demands more than a shrug of the shoulders — that is, “more than ignorance of the prevailing law or uncertainty about its development.” See, e.g., Reich v. S. New England Telecomms. Corp., 121 F.3d 58, 71 (2d Cir. 1997) (citing Martin v. Cooper Elec. Supply Co., 940 F.2d at 908 (3d Cir. 1984)); Williams v. Tri-County Growers, Inc., 747 F.2d 121, 129 (3d Cir. 1984) (ruling that “ignorance alone” is not enough to establish good faith defense); Marshall v. Brunner, 668 F.2d 748, 753 (3d Cir. 1982); Barcellona v. Tiffany Eng. Publ’n, Inc., 597 F.2d 464, 468-69 (5th Cir. 1979). An employer invoking good faith must show not only that it tried, but that it took “active steps to ascertain the dictates of the FLSA” and made a genuine effort “to comply with them.” See Reich, 121 F.3d at 71; see also Renfro v. City of Emporia, 948 F.2d 1529, 1540 (10th Cir. 1991) (“The good faith

60

requirement mandates the employer have an honest intention to ascertain and follow the dictates of the [FLSA].”).

Pursuant to the foregoing principles, an employer seeking to show good faith cannot just plead confusion and call it a day. One practical — and judicially recognized — way to satisfy the good faith showing is for the employer to prove that it sought out and adhered to legal advice regarding compliance with the FLSA. See Burnley, 730 F.2d at 140; see also McFeeley, 825 F.3d at 245 (affirming good faith defense because defendants consulted lawyer and adhered to legal advice); Perez, 650 F.3d at 375-76 (affirming denial of liquidated damages because employer sought legal assistance on FLSA compliance issues); Roy, 141 F.3d at 548-49 (concluding that modification of compensation structure could establish good faith defense, even though legal interpretation that employer received was later rejected). Another route that can satisfy the good faith defense is for the employer to rely on the DOL’s interpretation of the FLSA. See 29 U.S.C. § 259.

b.

After carefully assessing the trial evidence, the Bench Verdict ruled that Steadfast had failed to prove its assertion of the good faith defense. In that regard, the Bench Verdict first explained that Steadfast “could not have classified their nurses as independent contractors in good faith prior to seeking legal counsel in June 2018.” See Bench Verdict 27. Prior to that date, Steadfast had neither sought legal advice on the classification of its nurses nor taken any proactive steps to educate itself on the FLSA.

The Bench Verdict then resolved that Steadfast’s reliance on lawyer Bredehoft’s legal advice after June 2018 was neither in good faith nor objectively reasonable. As the

61

district court related, Steadfast had by then been informed by the DOL of its FLSA violations, and it was advised by the DOL in early 2018 to comply with the FLSA. The court also emphasized that Steadfast, in first consulting Bredehoft in June 2018, had failed to provide him with the pertinent information needed for a “fully informed, reasonable legal opinion” concerning Steadfast’s employee classification practices. See Bench Verdict 28.

On appeal, Steadfast contends that — even assuming its nurses were employees under the FLSA — it proved its good faith defense. See Br. of Appellants 49. And Steadfast argues that it reasonably believed that it had complied with the FLSA in consulting with lawyer Bredehoft. Steadfast also challenges the district court’s ruling that its reliance on Bredehoft was unreasonable, arguing that the court made several “clear errors” in weighing the evidence related to the good faith issue. Id. at 51.

Steadfast urges us, on appeal, to accept a theory of the good faith defense under which an employer can withhold important facts from its lawyer, decline to follow the attorney’s advice, and nevertheless successfully claim the good faith defense provided for in § 260 of Title 29. But as the Bench Verdict recognized, a brief or incomplete consultation with a lawyer regarding FLSA conformity does not satisfy the standards of the good faith defense. See Bench Verdict 29. In fact, the FLSA’s reasonableness standard is in place to prevent such a façade — one which could allow a phone call to a lawyer to justify a good faith defense. See 29 U.S.C. § 260.

To be sure, Steadfast did more than make a phone call seeking lawyer Bredehoft’s advice. But the limited engagement that Steadfast had with Bredehoft failed to satisfy the

62

FLSA’s requirements. Notably, Steadfast sidesteps the fact that it only sought Bredehoft’s advice after the DOL conducted its investigation. Even then, Steadfast did not provide Bredehoft the relevant facts. As the evidence demonstrated, Bredehoft did not interview any of the Steadfast nurses, did not visit any of the client facilities, and did not review any of Steadfast’s memoranda regarding best practices. Because Steadfast failed to provide Bredehoft with the relevant facts regarding its business operations, Bredehoft could do little more than offer a general opinion — predicated on incomplete facts — that Steadfast might have been correct in its classification decision.

Ignoring the aspects of Bredehoft’s advice it did not agree with — such as removing the “non-compete clause” and avoiding use of “employment” terms — Steadfast continued doing business as usual. See, e.g., Bench Verdict 29 (“[Steadfast] did not adhere to Mr. Bredehoft’s advice to forgo the non-compete clauses in its contractor agreements.”). And Steadfast did so in spite of the DOL investigation and its awareness of a 2018 DOL guidance bulletin which, as the Bench Verdict explained, “expressly identifie[d]” several of Steadfast’s pay practices as violative of the FLSA. Id. at 28.

We might agree that Steadfast did not take an “ostrichlike” approach to the FLSA by remaining “blissfully ignorant” of the FLSA’s requirements. See Roy, 141 F.3d at 549- 50. But good faith is not about merely “ascertain[ing] the dictates of the FLSA.” See Reich, 121 F.3d at 71. The good faith defense requires an employer to take serious and informed steps to adhere to the applicable law. And Steadfast did not do so.

63

In these circumstances, the district court did not abuse its discretion in ruling that Steadfast failed to satisfy its burden of proof with respect to the good faith defense for the First Period and the Second Period.

2.

We next turn to Steadfast’s contentions about the district court’s damages computations, which primarily relate to the Initial damages computations.12 Steadfast argues that the district court erred in adopting the Initial damages computations because they were “rife with false entries and prejudicial data entry errors.” See Br. of Appellants 54.

Steadfast challenged the Initial damages computations for the first time in its post-

trial motion of March 13, 2022, seeking relief under Federal Rules of Civil Procedure 52(b), 54(b), and 60(a)-(b). In its Opinion of December 7, 2023, the court sustained its reliance on the Initial damages computations, rejecting the merits of the March 13, 2022 motion invoking each of those Civil Rules. The court also then ruled that Steadfast had waived any right to contest the Initial damages computations because it had declined to contest them in a timely manner, that is, before or during the bench trial.13 And Steadfast

12

As previously explained, the Initial damages computations were presented by the Secretary in the bench trial and were predicated on Steadfast’s records. The Initial damages computations and their underlying methodologies were approved by the district court on January 14, 2022, when liability under the FLSA was established against Steadfast by the Bench Verdict.

13

In its appellate submissions, the Secretary maintains that the district court’s ruling that Steadfast waived any right to contest the Initial damages computations and their (Continued)

64

had opted not to present evidence to counter the Initial damages computations. Furthermore, Steadfast had failed to timely object in any respect to those damages computations or their underlying methodologies.

In assessing Steadfast’s challenges to the Initial damages computations and the updates thereto, we review each challenge — interposed under Rules 52(b), 54(b), and 60(a)-(b) — for abuse of discretion. See Carlson v. Boston Sci. Corp., 856 F.3d 320, 325 (4th Cir. 2017) (reviewing motion to reconsider pursuant to 54(b) for abuse of discretion); Sartin v. McNair L. Firm PA, 756 F.3d 259, 265 (4th Cir. 2014) (“Rule [60(a)] gives the district court discretionary authority, and a court of appeal’s review is for an abuse of discretion.”); Leber v. United States, No. 92–2461, 1994 WL 564753, at *1-2 (4th Cir. Oct. 17, 1994) (reviewing motion to amend judgment pursuant to 52(b) for abuse of discretion); Schwartz v. United States, 976 F.2d 213, 219 (4th Cir. 1992) (refusing to overturn denial of relief under Rule 60(b) absent abuse of discretion). As explained below, there was no abuse of discretion by the district court in denying Steadfast’s motion for relief under Rule 52(b), Rule 54(b), or Rule 60.

underlying methodologies precludes us from assessing the merits of Steadfast’s damages contentions. And the Secretary makes a very compelling argument in that regard. Steadfast consistently ignored the court’s directives, its suggestions, and its direct orders with respect to the Initial damages computations. And Steadfast’s trial lawyers may have undermined Steadfast’s challenges to the computations by their inaction and failure in that regard. In any event, however, the district court assessed and resolved each of the post- trial contentions made by Steadfast’s new set of lawyers, lodged under Rules 52(b), 54(b), and 60(a)-(b) of the Civil Rules. In these circumstances, we will decline to address or adopt the waiver contentions of the Secretary, and we will resolve those appellate challenges of Steadfast.

65

a.

As a threshold matter, we begin with the timeliness of Steadfast’s Rule 52(b)

motion. Civil Rule 52(b) provides that, “[o]n a party’s motion filed no later than 28 days after the entry of judgment, the court may amend its findings — or make additional findings — and may amend the judgment accordingly.” See Fed. R. Civ. P. 52(b). The 28-day deadline is firm — indeed, Rule 6(b)(2) prohibits a court from “extend[ing] the time to act” under Rule 52(b), no matter the reason. Despite that strict deadline, Steadfast moved to amend the district court’s findings of fact on the Initial damages computations 58 days after the Bench Verdict and a corresponding judgment were entered.

But setting the timeliness issue aside, Steadfast’s motion — as the district court ruled — fails on the merits. Importantly, Rule 52(b) is meant to “is to correct manifest errors of law or fact or, in some limited situations, to present newly discovered evidence.” See Fontenot v. Mesa Petroleum Co., 791 F.2d 1207, 1219 (5th Cir. 1986). It is not an invitation “to rehash old arguments already considered and rejected by the trial court.” See Nat’l Metal Finishing Co. v. BarclaysAmerican/Comm., Inc., 899 F.2d 119, 123 (1st Cir. 1990). Nor does it allow a party “to introduce evidence that was available at trial but was not proffered, to relitigate old issues, to advance new theories, or to secure a rehearing on the merits.” See Fontenot, 791 F.2d at 1219.

As the district court explained, Steadfast “point[ed] out no errors of law or fact”

warranting relief under Rule 52(b). See Su v. Medical Staffing of Am., LLC, No. 2:18-cv- 00226 (E.D. Va. Dec. 7, 2023), ECF No. 411. Instead, its motion simply attempted to “relitigate issues that were decided at trial” — despite having ample opportunity to raise

66

those objections earlier. Id. The court had repeatedly directed Steadfast to review and respond to the Initial damages computations, which were drawn from Steadfast’s own records and explained through witness testimony. Yet Steadfast failed to do so, either before or during the bench trial. Nor did Steadfast offer alternative computations or any newly discovered evidence.

In short, Steadfast’s motion did none of what Rule 52(b) requires. We therefore discern no abuse of the district court’s discretion in denying Steadfast’s motion under Rule 52(b).

b.

When a district court has entered an interlocutory order “that adjudicates fewer than all of the claims,” it retains the discretion to revise that order under Rule 54(b), “at any time before the entry of a judgment adjudicating all the claims.” Fed. R. Civ. P. 54(b). Our Court, however, has emphasized that a court’s discretion under Rule 54(b) “is not limitless.” See Carlson, 856 F.3d at 325. We have narrowed its application to include “(1) a subsequent trial produc[ing] substantially different evidence; (2) a change in applicable law; or (3) clear error causing manifest injustice.” Id. A Rule 54(b) motion to reconsider an interlocutory ruling is thus not a “vessel” for the losing party to submit evidence that it could have presented earlier. Id. at 325-26.

Even a showing that “legal precedent could support an alternative resolution” is insufficient for an award of relief under Rule 54(b). See Bethany Boardwalk Grp. LLC v. Everest Sec. Ins. Co., No. 20-2319, 2022 WL 12324609, at *6 (4th Cir. Oct. 21, 2022). Indeed, we have ruled that Rule 54(b) requires that a challenged decision be so “dead

67

wrong” that it “strike[s] us as wrong with the force of a five-week-old, unrefrigerated dead fish.” See U.S. Tobacco Coop. Inc. v. Big S. Wholesale of Va., LLC, 899 F.3d 236, 258 (4th Cir. 2018) (quoting TFWS, Inc. v. Franchot, 572 F.3d 186, 194 (4th Cir. 2009)).

In assessing the Rule 54(b) issue presented here, we must evaluate whether the trial court abused its discretion by ruling that Steadfast’s consistently tardy challenges to the Initial damages computations failed to demonstrate a “clear error causing manifest injustice.” See Carlson, 856 F.3d at 325. And as the district court pointed out, Steadfast’s new set of lawyers simply emphasized additional evidentiary material that had been readily available to their predecessors in the earlier litigation proceedings. See Su v. Medical Staffing of Am., LLC, No. 2:18-cv-00226 (E.D. Va. Dec. 7, 2023), ECF No. 411 at 7-8.

Yet Steadfast argues that the trial court’s reliance on the Initial damages computations — and the methodologies utilized therein — in making and approving its subsequent computations is sufficient to satisfy the Rule 54(b) “manifest injustice” requirement. In Steadfast’s Rule 54(b) challenge, however, there is no claim presented of “a change in the applicable law.” Nor is there a claim of “substantially different evidence” that was only discovered during the litigation. Steadfast simply says that it has now identified clear errors “infecting” the Initial damages computations, and that a “manifest injustice resulting from these errors is equally clear, especially given the district court’s order on liquidated damages.” See Br. of Appellants 57.

The assertedly clear errors that Steadfast sought to correct relate solely to the Initial damages computations that were readily available to Steadfast both before and during trial. Again, Steadfast was directed on multiple occasions by the trial court to examine and

68

scrutinize the Initial damages computations. Steadfast knew in a timely fashion of the Initial damages computations. And Steadfast always knew that a finding of liability based on its “employer” status meant that liquidated damages would likely be awarded. As the district court emphasized, Steadfast failed to satisfy Rule 54(b) and instead simply sought to challenge a previously litigated issue after receiving an adverse ruling. In these circumstances, the district court did not abuse its discretion when it declined to grant relief to Steadfast under Rule 54(b).

c.

Steadfast next contends, without further specification, that it was “an abuse of discretion, and a manifest injustice,” for the district court “to deny relief under Rule 60.” See Br. of Appellants 58. Steadfast is not entitled, however, to use Rule 60(a) or Rule 60(b)(1) as a tool for correcting asserted mathematical errors, nor can it plausibly argue that the damages computations made and awarded by the court have resulted in an “extraordinary circumstance” that warrants relief under Rule 60(b)(6).14 See Aikens v. Ingram, 652 F.3d 496, 500 (4th Cir. 2011).

Under Rule 60(a) of the Civil Rules, a district court is entitled to “correct a clerical mistake or a mistake arising from oversight or omission” in a “judgment, order, or other

14

More specifically, Rule 60(a) allows a court to “correct a clerical mistake or a mistake arising from oversight or omission when one is found in a judgment, order, or other part of the record.” Rule 60(b)(1) permits a party to seek relief from a final judgment, order, or proceeding due to “mistake, inadvertence, surprise, or excusable neglect.” And Rule 60(b)(6) allows a court to grant relief from a final judgment, order, or proceeding for “any other reason that justifies” it.

69

part of the record” — so long as the correction reflects the court’s original intent. See Fed. R. Civ. P. 60(a); see Sartin, 756 F.3d at 265 (explaining that Rule 60(a) “is properly utilized to perform a completely ministerial task” (internal quotation marks omitted)). Likewise, Rule 60(b)(1) authorizes a court to relieve a party of a final judgment based on a “mistake,” among other grounds. See Fed R. Civ. P. 60(b)(1); see also Kemp v. United States, 596 U.S. 528, 537 (“Rule 60(a) covers a subset of ‘mistake[s]’ — e.g., ‘clerical’ ones — whereas Rule 60(b)(1) covers ‘mistake[s]’ simpliciter.”). And to obtain such relief, the moving party is required to show that it was not at fault. See Home Port Rentals, Inc. v. Ruben, 957 F.2d 126, 132 (4th Cir. 1992).

What Steadfast sought here, however, was not a correction of the record, but further litigation concerning how the district court’s damages calculations had been made. Steadfast was not asking the court to fix a “slip[] of the judicial pen.” See TDK Elecs. Corp. v. Daiman, 321 F.3d 677, 679 (7th Cir. 2003). Instead, Steadfast invoked Rule 60(a) to characterize the court’s substantive rulings on damages as merely being “clerical errors” or “mistakes.” But that is not the purpose of Rule 60(a). Put simply, a court is not authorized to use Rule 60(a) to reconsider an issue that has already been decided, or to revisit the court’s decisions already made. See Sartin, 756 F.3d at 265; see also Hylind v. Xerox Corp., 632 F. App’x 114, 116 (4th Cir. 2015) (rejecting Rule 60(a) challenge to court’s method for calculating backpay). Nor can a court grant relief where a party’s Rule 60(b)(1) motion “is nothing more than a request that the district court change its mind.” See United States v. Williams, 674 F.2d 310, 313 (4th Cir. 1982).

70

In this situation, the district court had clearly explained to the parties and their lawyers during the bench trial that its damages calculations and awards would be based on the trial evidence. And Steadfast had multiple opportunities to object or offer competing computations. Steadfast is thus unable to demonstrate that it was not “at fault” for failing to object to the Initial damages computations, or for not presenting alternative computations. See Home Port Rentals, Inc., 957 F.2d at 132.

Steadfast’s Rule 60 contention would have us decide that the district court abused its discretion by declining to label its damages computations and awards — for 1105 nurses across multiple years — as “clerical errors” under Rule 60(a), or as a “mistake” under Rule 60(b)(1). And Steadfast had first requested the court do so at least six months after those computations had been fully litigated. In that context, we cannot say that the court abused its discretion in declining to award Steadfast relief under Rule 60(a) or 60(b)(1).

Nor can we say that the district court abused its discretion in denying relief to Steadfast under Rule 60(b)(6), which allows a party to seek relief from a final judgment, order, or proceeding for “any other reason that justifies relief.” See Fed. R. Civ. P. 60(b)(6). To prevail, a movant must act in a timely fashion, demonstrate a lack of prejudice to the nonmoving party, and proffer a “meritorious defense.” See Augusta Fiberglass Coatings, Inc. v. Fodor Contracting Corp., 843 F.2d 808, 811 (4th Cir. 1988). And even then, the bar is high — Rule 60(b)(6) requires “extraordinary circumstances.” See Aikens, 652 F.3d at 500; see also Kemp, 596 U.S. at 540 (Sotomayor, J., concurring) (explaining that Rule 60(b)(6) is available “to reopen a judgment in extraordinary circumstances.”).

71

In this situation, there were no extraordinary circumstances presented to the district court. Steadfast’s new set of lawyers simply disagreed with some strategic choices that were made by their predecessors. As we have explained, “strategic decisions made during the course of litigation” — even if later second-guessed — will not justify relief under Rule 60(b)(6). See Schwartz v. United States, 976 F.2d 213, 218 (4th Cir. 1992). Put simply, Rule 60(b)(6) is not a vehicle for a do-over, which Steadfast now seeks. See Williams, 674 F.2d at 313.

In these circumstances, we are satisfied that the district court did not abuse its discretion in denying relief to Steadfast under any of the specified provisions of Rule 60.

IV.

Pursuant to the foregoing, we reject each of Steadfast’s contentions of error and affirm the judgment.

AFFIRMED

72

RICHARDSON, Circuit Judge, dissenting:

Today’s decision well illustrates the peril of judicially contrived, many-factor tests.

After a seven-day bench trial, the district court and the parties had thousands of pages of evidence to filter through just such a test. By plucking a few conclusions from this vast record, the majority concludes that nurses on Steadfast’s registry are employees, not independent contractors. And because one factor can always outweigh another, the majority’s decision to set aside overwhelming contrary evidence comes with a veneer of respectability.

But the majority is wrong on the facts and wrong on the law. On the facts, the majority ignores crucial testimony and glosses over clear mistakes from the district court that form the heart of Steadfast’s argument on appeal. On the law, the majority dismisses careful distinctions in our precedent, relies instead on out-of-circuit cases, and stretches the factor test so far that it could cover most any economic arrangement. I respectfully dissent. I. Nurses On Steadfast’s Registry Are Independent Contractors, Not Employees Steadfast’s claim is simple. Nurses on its registry, it says, could do more or less as they pleased. Faced with a menu of options, they could choose any Steadfast shifts they wanted—or none. In choosing, they could optimize for pay rate, type of work, travel distance, hours, and myriad other factors. As they did, they could also look beyond the registry, crafting schedules that suited their interests from a combination of Steadfast and non-Steadfast shifts. And on the job, the nurses operated on their own. Without Steadfast training, supervision, or performance evaluations, their only task was to provide medical

73

care that satisfied the client facilities. Their fortunes then rose, or fell, based on how well they did.

The testimony leads inexorably to those conclusions. And on those facts, the right legal result is clear. Nurses use Steadfast’s registry as one of many ways to find workplaces and pick up shifts. That service is valuable. But the nurses do not depend on it. Economically, the nurses are in business for themselves. For that reason, Steadfast rightly treats them as contractors rather than employees.

A. Our Precedents Focus On Economic Dependence As the majority recounts, we determine whether a business relationship is employment within the meaning of the Fair Labor Standards Act by applying a six-factor test. I will march through the factors soon and explain why none supports the majority’s conclusion. But first, it’s important to understand what the factors are for—and what they are not.

Start by considering how we got here. FLSA requires that courts distinguish between employees and independent contractors, but the definitions it offers for “employer” and “employee” are unhelpful. The statute defines the former as “any person acting directly or indirectly in the interest of an employer in relation to an employee.” 29 U.S.C. § 203(d). And it defines the latter as “any individual employed by an employer.” Id. § 203(e)(1). Definitions like these are, of course, “completely circular.” Nationwide Mut. Ins. v. Darden, 503 U.S. 318, 323 (1992). And the circularity is not much abated by the definition of “employ,” which is “to suffer or permit to work.” 29 U.S.C. § 203(g). Yet then-Assistant Attorney General Robert Jackson testified in Congress that FLSA’s

74

open-ended definitions were part of the point—that the “draftsmen have been painstaking to make the standards as definite as the conditions with which they have had to deal permit.” 1 Joint Hearings Before the S. Comm. on Educ. and Lab. and the H. Comm. on Lab. on S. 2475 & H.R. 7200, 75th Cong. 10 (1937).

What are we to make of this? Normally, “when Congress has used the term ‘employee’ without defining it, we have concluded that Congress intended to describe the conventional master-servant relationship as understood by common-law agency doctrine.” Cmty. for Creative Non-Violence v. Reid, 490 U.S. 730, 739–40 (1989). But FLSA, unlike many federal employment statutes, was not written to capture common-law notions of agency. It was instead supposed to govern not just immediate principals but also those who sat higher up long chains of command: “any person acting directly or indirectly in the interest of an employer in relation to an employee.” Salinas v. Com. Interiors, Inc., 848 F.3d 125, 133 (4th Cir. 2017) (quoting 29 U.S.C. § 203(d)); see also Rutherford Food Corp. v. McComb, 331 U.S. 722, 728–29 (1947) (noting this breadth and collecting cases).

In other words, FLSA is broad, but it is broad in a particular way—and it doesn’t mean that just anything counts as employment. Intent on capturing this legislative purpose, our cases apply the six-factor test the Court articulated in United States v. Silk, 331 U.S. 704 (1947). See McFeeley v. Jackson St. Ent., LLC, 825 F.3d 235, 241 (4th Cir. 2016) (enumerating the Silk factors). Like any test, this one is supposed to measure something. “The aim of the Act was to remedy the inequality of bargaining power in controversies over wages, hours and working conditions.” Silk, 331 U.S. at 713; see also 1 Joint Hearings Before the S. Comm. on Educ. and Lab. and the H. Comm. on Lab. on S. 2475 &

75

H.R. 7200, 75th Cong. 10 (1937) (explaining that the Act aimed at “difficult and complex industrial situations”). And the aim of the test is the same, the Court has told us, which means we must grapple with those complex situations—and determine whether workers are “employees . . . as a matter of economic reality,” not under “the technical concepts” of agency law. Silk, 331 U.S. at 713 (quotations omitted); see also N.L.R.B. v. Hearst Publ’ns, 322 U.S. 111, 129 (1944) (looking to “underlying economic facts rather than technical[] . . . legal classifications”).

What does it mean to be an employee as a matter of “economic reality?” Because it glossed statutes that took aim at disparities in bargaining power, the Supreme Court wove a test focused on whether workers were, “as a matter of economic reality . . . dependent upon the business to which they render service.” Bartels v. Birmingham, 332 U.S. 126, 130 (1947) (emphasis added); accord Schultz v. Capital Int’l Sec., Inc., 466 F.3d 298, 305 (4th Cir. 2006) (asking “whether the [workers] were, as a matter of economic reality, dependent on the business they served, or, conversely, whether they were in business for themselves”), and McFeeley, 825 F.3d at 241 (same). That is, the test is designed to differentiate (1) workers who are “economically dependent on a putative employer” from (2) workers “whose profit or loss depend upon [their] own creativity, ingenuity, and skill.” Salinas, 848 F.3d at 150.1

1

As this discussion should make clear, to apply the test from the Supreme Court’s decision in Silk is to engage in a now-discredited task: effectuate Congress’s unenacted purposes rather than its enacted text. Panel precedent requires us apply the Silk test here. And as long as this test is our law, we should follow it faithfully. Payne v. Taslimi, 998 (Continued)

76

F.3d 648, 654 (4th Cir. 2021). This is why one of my complaints with the majority is that it stretches this test further than it should—ignoring what the Supreme Court first designed it to measure and converting it into a meaningless catchall.

But binding as it is on this Court, this line of cases is ripe for correction. For one, it makes little sense that we apply Silk’s test in FLSA cases. That decision interpreted the Social Security Act, not FLSA. And though Silk took some cues from a National Labor Relations Act case, Hearst Publications, both of those decisions have been overruled for 30 years. See Darden, 503 U.S. at 324–25. If any, the Supreme Court decision that should control in these cases is Rutherford Food, 331 U.S. 722. That case interpreted the right statute (FLSA), announced a controlling test (which is not the same as Silk’s), and has never been overruled. See id. at 730 (identifying relevant considerations and announcing a test that considers the “circumstances of the whole activity”). So why do our cases choose to invoke Silk instead, an overruled case that interprets a different statute?

Not that Rutherford Food is much better. Rather than consider FLSA’s text on its own terms, that case borrowed a generic, purposive approach from Silk and Hearst Publications because FLSA “is a part of the social legislation of the 1930’s of the same general character as the [NLRA] . . . and the [SSA].” 331 U.S. at 723. Rutherford Food shares those cases’ faulty foundations. And its own slender analysis is no better. It focuses on the legislative “purpose” “to eliminate low wages and long hours [and thus] free commerce from the interferences arising from production of goods under conditions that [are] detrimental to the health and well-being of workers.” Id. at 727. The one bit of textual analysis that decision offers is that FLSA’s text “derives from the child labor statutes.” Id. at 728. And though the test it gives is distinct from Silk’s, it is no more helpful: Rutherford Food just asks whether, in light of “the circumstances of the whole activity,” “the work done, in its essence, follows the usual path of an employee.” Id. at 730, 729.

There is a better way. When it comes to other statutes, the Supreme Court has already “abandon[ed] Silk’s emphasis on construing th[e] term [employee] ‘“in light of the mischief to be corrected and the end to be attained.”’” Darden, 503 U.S. at 325 (quoting Silk, 331 U.S. at 713 (in turn quoting Hearst, 322 U.S. at 124)). Today, “[t]he starting point for our interpretation of a statute is always its language.” Cmty. for Creative Non- Violence v. Reid, 490 U.S. 730, 739 (1989). Rather than follow Silk’s lead, we should respect the fact that FLSA defines employment in different language than the other New Deal-era social legislation. See Darden, 503 U.S. at 326 (noting differences).

By abrogating Silk and Hearst Publications, the Supreme Court allowed lower courts to apply two important statutes based on what they say. But as long as the Silk test lives on in FLSA cases—where it should never have applied in the first place—we must continue deciding important labor cases without a thought for the law Congress actually made. It is time to finish the job Darden started decades ago.

77

The majority begins by gesturing toward this idea. But then it gets bogged down in the factors themselves and forgets what they are supposed to measure. And as it does, the majority downplays—or ignores altogether—evidence that decisively answers this question.

B. Overwhelming Testimony Showed That The Nurses Are Not Economically Dependent On Steadfast

Start with the evidence the majority gives short shrift. Two key sets of facts rule out its legal conclusion that the nurses depended on Steadfast in the way our precedents require.

First, it was the nurses, not Steadfast, who decided when and where they would work. Just being on the registry did not oblige a nurse to work at any particular place or time. Instead, a nurse on the list would receive “a general blast” whenever a facility requested staff. J.A. 1000. Every nurse with the qualifications to “satisfy the facilities’ needs” got the same notifications, and each could “accept or reject” any available shift. Walsh v. Med. Staffing of Am., 580 F. Supp. 3d 216, 221 (E.D. Va. 2022).2 Nurses could also contact Steadfast before a blast went out, asking to work at specific facilities or during specific hours. And if they preferred, nurses could go straight to the source and use an app called Zira to talk directly with client facilities about work opportunities.

2

Client facilities dissatisfied with a nurse’s performance could ask Steadfast to avoid offering that nurse its shifts again. After a client placed such a nurse on its “Do No Return” list, that nurse could no longer take shifts at its facility.

78

Though nurses could make these decisions for any reason—or no reason—many testified that they did so specifically to optimize their work experiences. That is, they picked shifts that paid the most in relation to other considerations like type of work, relationship with the client facility, travel distance, hours, and so forth.

As they performed this calculus, the nurses considered not just options available through Steadfast’s list but from other sources too. Seventeen nurses testified on this point at trial, and 14 said that they did or could take nursing jobs outside Steadfast’s network while they also worked on the registry. And as one nurse explained, this is common practice among nurses: “[A] lot of nurses or healthcare workers are on more than one registry in case one has less work or one is going to a facility that you like better.” J.A. 693. Another explained that this choice carries financial ramifications: “Steadfast’s not the only company I work for, so if I work for two other companies, there may be a company, a facility that’s paying more. . . . Or they’re giving more hours that you’re looking for.” J.A. 870.

This testimony is hard to square with the idea that appearing on Steadfast’s registry put a nurse at Steadfast’s mercy. Instead, it shows that each nurse was free, based on a menu of options, to “determine his hours, his income, and who he worked for.” Cilecek v. Inova Health Sys. Servs., 115 F.3d 256, 261 (4th Cir. 1997). And these facts “are core incidents to a work relationship that are inconsistent with employee status.” Id.3

3

Although Cilecek considered a different statute, Title VII, it applied a factor test much like the one we use for FLSA—a test that also begins with control. See id. at 260.

79

Second, Steadfast did no meaningful management. Steadfast did not provide “any orientation,” “a handbook,” “[p]olicies and procedures,” or “any training at all.” J.A. 973.4 Nurses could only appear on the registry if they already had their own training, licensure, and certifications. And having those medical skills, nurses used them on the job free from any Steadfast control. Steadfast never deployed personnel at client facilities to supervise or manage nurses. Client facilities instead expected that nurses would arrive fully prepared to do their jobs and then do them, using professional judgment and “discretion in the manner in which they practice nursing.” J.A. 449–50. When they wanted nurses to do something different, the client facilities—not Steadfast—gave nurses instructions “oriented to the unit that they are working,” including “documents based on the assignment that they have.” J.A. 498. And while doing those assignments, nurses reported to doctors at client facilities, not Steadfast. For its part, Steadfast’s expectations were minimal: “act professional, . . . go to work like you’re supposed to, show up on time and do what you’re supposed to do as whatever you are, a CNA or a nurse.” J.A. 258–59.5

4

To be sure, Steadfast sometimes trained nurses on matters like substance abuse and sexual harassment. But no one testified that Steadfast told nurses how to provide medical care.

5

In this vein, Steadfast sometimes issued short memoranda. Such as: “Please be on time for all shifts,” J.A. 1030, “[n]o smoking on facility grounds,” J.A. 598, “[n]o sleeping in the residents’ rooms,” J.A. 1029. When nurses committed this sort of misconduct, or failed to show up for work altogether, Steadfast might remove them from its registry.

But that is just what we would expect for any independent contractor. A “worker”

is not “automatically” “an employee covered by the FLSA the moment a company exercises any control over him.” McFeeley, 825 F.3d at 242. If you hired a plumber to fix (Continued)

80

Nurses’ success with client facilities depended on this medical skill. When nurses did their jobs well, client facilities might ask them to return. The facilities could and did “negotiate the rate for a nurse just because the nurse is really good at what she does.” J.A. 998. And when nurses did subpar work, client facilities could instead put them on a “Do Not Return” list—which instructed Steadfast to avoid giving the underperforming nurse future shifts at the complaining facility. Either way, nurses’ opportunities and pay depended on how well they did their work. And how well they did was a question for the clients to answer, not Steadfast.

The majority all but ignores this critical testimony—testimony that consumed days of trial and runs for hundreds of transcribed pages. Yet with this testimony in mind, it’s hard to see how the nurses were economically shackled to Steadfast. Instead, this testimony makes clear that it was the nurses, not Steadfast, who controlled the work—the when, where, and how. The nurses’ fortunes rose or fell based on which opportunities they pursued and what the clients, not Steadfast, thought of their work. Joining the registry just expanded the set of shifts that the nurses could choose from, placing Steadfast options on

your kitchen sink, but he never showed—or turned up hours late, then smoked a cigarette in your living room before taking a nap on your couch—you understandably might not retain him for future jobs. After all, “[i]t is rather hard to imagine a party contracting for needed services with an insouciant ‘Do whatever you want, wherever you want, and however you please.’” Id. So under our cases, it is legal error to treat instructions like these as going to control. Though they may show control in some colloquial sense, it’s not the kind of control our cases instruct us to focus on.

Apparently recognizing this, the Secretary conceded at argument that contractors and employees alike often receive instructions like these. So rather than focus on the substance of these instructions, the Secretary now asks us to consider their “tone.” Oral Arg. at 35:00–35:40. I leave what to make of this argument to the reader.

81

the table alongside other opportunities. (For nothing compelled nurses to work Steadfast shifts,6 and they could still work non-Steadfast shifts.) And on either kind of shift, nurses’ task was satisfying the client, not Steadfast. The nurses’ economic situations, in short, depended on their own managerial and medical skills—skills that the nurses developed without any Steadfast input and used every day without any Steadfast supervision.

With this in mind, my trouble with the majority runs deeper than a quibble about how to analyze this factor or that. To be sure, like the majority, I will trudge through the factors, and I will show why its analysis is wrong at every turn. But the deeper problem is that by minimizing or ignoring the testimony I have recited, the majority deprives Steadfast of its best arguments—arguments that should be decisive when it comes to gauging the economic reality.

The majority cannot defend its decision to ignore this evidence by urging deference to the district court either. For the district court did not make any findings of fact contrary to any of what I have said so far. On this evidence, there is nothing to defer to. And of course, considering whether this evidence shows that Steadfast wielded the relevant kind of control—and then weighing that factor against others—are matters of law on which we cannot defer to a district court at all.

6

Unless, of course, a nurse had signed up for a Steadfast shift. Once a nurse committed to a shift, Steadfast reasonably expected the nurse to work it, just as anyone who signs a contract with an independent contractor would expect them to abide by their commitment. But enforcing that expectation is very different from requiring nurses to take certain shifts to begin with.

82

II. The District Court’s Contrary Conclusion Rests On Clear Errors For its part, as it weighed the factors, the district court mostly ignored this evidence and instead relied on two findings of fact that it thought cut the other way. But although the majority rightly remarks that we afford these findings deference on review, clear-error review is not no review. And in this case, anyone who reads the trial transcript would be “left with the definite and firm conviction that a mistake has been committed.” Ante at 34 (quoting Anderson v. Bessemer City, 470 U.S. 564, 573 (1985)).

First, the district court found that “[t]he nurses do not exercise independent judgment and perform activities typical of nurses in the medical industry including, but not limited to, administering medications, treating wounds, taking notes, and otherwise caring for patients.” Medical Staffing, 580 F. Supp. 3d at 224 (emphasis added). This is manifestly false. The nurses uniformly testified that Steadfast did not manage their medical work. Not one could recall a situation where Steadfast told them how to care for patients— much less supervised them while they did it or punished them for doing it badly.

And how could a nurse nurse without making independent judgments? A nurse at a hospital may need to triage patients to decide whether a head wound or anaphylaxis warrants quicker treatment. During treatment, she may need to check vital signs and gauge the patient’s status. Then, she must select a course of action: There are many types of procedures, diagnostic techniques, medications, and so on to choose from. And of course, nurses must also decide when to summon other medical professionals—from phlebotomists to brain surgeons. The nurses in this case testified that this is what they did, day in and day out:

83

You have to know how to read a doctor’s order, know what medications . . . do for the patient[s]. You have to know some of the critical values of labs . . . You have to know trachs [tracheotomy tubes] and G-tubes [gastro tubes,] and how to treat dialysis ports and things like that.

J.A. 749. Each of these tasks plainly requires judgment.

The district court’s own citations back this up. To support its finding, the district court cited testimony from a nurse who said, “basic nursing care . . . are the services I provide,” J.A. 151, plus testimony from two client facilities’ administrators who corroborated that account. No doubt, Steadfast sometimes gave nurses guidance on matters like deportment—do not show up late or drunk, do not sleep on the job. But the district court did not cite this testimony; it cited only the testimony that directly contradicted its finding. And not one scrap of evidence, cited or not, suggests that Steadfast told the nurses how to provide medical care.

If more were needed, the district court not only found that the nurses exercised no independent judgment but that they didn’t even provide medical care—that they did not “perform activities typical of nurses in the medical industry.” Medical Staffing, 580 F. Supp. 3d at 224. No “administering medications,” no “treating wounds,” no “taking notes.” Id. But the nurses testified that they did. No one testified that they did not. And the Secretary’s own litigating position was that the nurses work as, well, nurses—and that this arrangement warranted employee status. In light of all this, finding that the nurses are not nurses but apparently something else strikes me not so much as a reasonable disagreement on how to weigh the evidence, but as a clear mistake. Yet given this finding,

84

it’s no surprise that the district court thought Steadfast wielded tight control over the nurses—the foremost factor among six. See J.A. 1209–10.

Second, the district court found that “Steadfast’s agreements with the nurses include ‘non-competition’ clauses, in which the nurses are prohibited from working for Steadfast’s competitors without Steadfast’s express, written consent.” Medical Staffing, 580 F. Supp. 3d at 223. This, at least, is true. As the majority observes, the noncompete provisions forbid nurses to “directly provide” medical services “to any competitors of Steadfast.” Ante at 8–9 (quoting J.A. 1796).

But the district court, and now the majority, make a subtler factual mistake by misunderstanding how the noncompete works. In the district court’s view, these “clauses . . . significantly hinder the nurses’ ability to accumulate profit independent of [Steadfast]” and so tilt the noncontrol factors toward employment. Medical Staffing, 580 F. Supp. 3d at 234 & n.3. Not at all. To be sure, the clauses require nurses to avoid appearing on other registries while they serve on Steadfast’s, for those registries are Steadfast’s competitors. But the clauses say nothing about the nurses’ ability to work for client facilities on their own terms or outside Steadfast’s network. Those facilities are not Steadfast’s competitors but its customers; they are buyers, and Steadfast is a supplier. And as 14 out of 17 nurses testified, many nurses on the registry did just that and therefore could make as much or as little of their total income via Steadfast’s network as they pleased.

This makes the district court’s, and the majority’s, reliance on the noncompetes perplexing. The surest sign of the nurses’ independence is the fact that they could and did work elsewhere—that they didn’t earn their pay from Steadfast alone and chose, as they

85

made their schedules, between Steadfast and non-Steadfast shifts. And the text of the noncompete places no limit on this freedom. Tellingly, despite concluding that the noncompetes hindered the nurses’ independence, the district court never said how—it never found, for instance, that nurses couldn’t or didn’t work off-registry. Yet given what the noncompete says, this is no surprise. More still, no one testified that Steadfast ever enforced the noncompete—and Steadfast representatives testified that they did not. As the majority concedes, our economic realities test is concerned with what defendants “actually do” rather than what they “could have done.” Ante at 48 (quotation omitted). So this should be the end of the matter.

Without either the erroneous supervision-and-control finding or the noncompete, the district court’s conclusion is impossible to defend. For this reason, the right result in this appeal comes down to how we resolve Steadfast’s objections to the district court’s findings of fact.

III. The Majority’s Approach Ignores These Mistakes, Plus Steadfast’s Arguments, The Evidence, And Our Precedent

Rather than engage with this testimony, or Steadfast’s arguments about it, the majority takes a new approach to our FLSA test. That approach is wrong, and it portends trouble in future cases.

A. The Majority Does Not Engage Steadfast’s Clear-Error Argument Steadfast’s lead argument is not that the district court misapplied our law to the facts; it’s that the district court “committed clear error on key findings of fact.” Op. Br. at 19 (cleaned up). Its main complaint with the decision below is that “the district court

86

repeatedly . . . cite[d] anecdotal evidence about discrete instances of alleged ‘control’ by Steadfast, while failing to make evidence-based findings about Steadfast’s customary operational practices that prevailed in the aggregate across its relationships with a referral network of more than 1100 nurses.” Id.

As the majority observes, this argument faces an uphill climb. A district court’s findings of fact get “great weight with the appellate court,” and such a finding is only “‘clearly erroneous’ when although there is evidence to support it, the reviewing court on the entire evidence is left with a definite and firm conviction that a mistake has been committed.” United States v. U.S. Gypsum Co., 333 U.S. 364, 395 (1948). The majority would be right to hold Steadfast to this demanding burden. But although the majority perseverates about the standard of review, it does not then scrutinize the record and resolve Steadfast’s argument.

Steadfast clears this high hurdle. The district court’s finding that nurses made no medical judgments—and indeed did not nurse at all—beggars belief. And its finding that Steadfast prevents nurses from working elsewhere fares no better. Both cave in under the weight of the evidence; neither is “plausible in light of the full record.” Cooper v. Harris, 581 U.S. 285, 293 (2017) (quotation omitted). That makes all the difference. Without the plainly wrong findings that Steadfast controls how nurses do their work and prevents them from working elsewhere, I don’t see how we could conclude that the nurses labor under Steadfast’s thumb.

And even if you set aside my view of the merits and assume that the majority could persuasively explain why we should rely on these findings, its decision makes a more

87

fundamental misstep too. It abdicates our duty to consider and resolve the parties’ contentions. Steadfast argues first and foremost that the district court made clear errors. As I have explained, that is far from a frivolous claim. Yet the majority doesn’t engage it. The majority offers no analysis of the record to determine whether any of the challenged findings is correct; it gives no serious discussion of these points; and when it reaches the merits, it cites only the district court’s findings of fact—the very findings that Steadfast says are clearly erroneous—without so much as pausing to explain why it thinks those findings withstand scrutiny. Reciting one side’s facts and ignoring the other’s is not clear- error review. It is not review at all.

B. The Majority Instead Takes A Detour Around Our Precedent Without any decisive arguments, the majority is left with no choice but to refocus on minutiae. Yet for all the time it spends marching through the factors, it marshals little factual support for its views. And even when it can identify some definite fact, the majority strains to explain—or worse, does not try to explain—why that fact matters under our precedents. The result is a stretched-out economic realities test that covers dynamics our precedents have never thought consequential. To set the record straight, I will consider each factor in turn.7

7

To reiterate, the factors are: “(1) [T]he degree of control that the putative employer has over the manner in which the work is performed; (2) the worker’s opportunities for profit or loss dependent on his managerial skill; (3) the worker’s investment in equipment or material, or his employment of other workers; (4) the degree of skill required for the work; (5) the permanence of the working relationship; and (6) the degree to which the services rendered are an integral part of the putative employer’s business.” McFeeley, 825 F.3d at 241.

88

1. Control over the manner in which work is performed The majority’s approach to the control factor marks a serious departure from our precedent. Our cases do not ask us to scour the record for anything that might be called control. Instead, they home in on control over “the manner in which the work is performed.” McFeeley, 825 F.3d at 241; accord Schultz, 466 F.3d at 304–05. In McFeeley, we focused on a nightclub’s “coach[ing] dancers” about “attitude,” “dictat[ing] each dancer’s work schedule,” and “impos[ing] written guidelines” about how to dance and what to wear. 825 F.3d at 242. But we set aside ancillary matters like “banning drinking while working,” because terms like these “are part and parcel of bargaining” between any worker and those for whom he works—independent or not. Id. at 242–43. Likewise in Schultz, we focused on whether the putative employer enacted detailed policies that “strictly dictated the manner in which” security guards were to operate—there, how often to patrol, which doors to use, and even how to open doors were all micromanaged. 466 F.3d at 307; see also Hall v. DIRECTV, LLC, 846 F.3d 757, 774–75 (4th Cir. 2017) (“DIRECTV . . . extensively controlled Plaintiffs’ day-to-day—indeed, hour-to-hour— work.”).

The majority does not try to find analogous facts. Instead, it ignores all this and focuses on three other concerns that find little support in our precedent and make less practical sense.

First, the majority insists that Steadfast unilaterally set nurses’ pay rates. See ante at 40 (citing McFeeley, 825 F.3d at 241–42). But as a matter of law, the majority oversells how much (if at all) this point should matter. If it kept reading McFeeley, it would have

89

seen that that case says “conditions” like “the terms of . . . compensation[] are part and parcel of bargaining between parties whose independent contractual status is not in dispute.” 825 F.3d at 243.

And on the facts, the majority oversimplifies. No doubt, Steadfast told nurses how much it would pay for any given job. (Though even then, some testified that they angled for higher rates when the job was tough or the facility distant.) But relying on this fact alone overlooks important details. For starters, Steadfast did not “unilaterally” set those rates; it negotiated them with client facilities based on what the facilities were willing to pay. The district court expressly found as much. Medical Staffing, 580 F. Supp. 3d at 225. And during those negotiations, as explained, client facilities would and could offer to pay more for nurses who did better work. Plus, even if nurses had little latitude to adjust the rate for a given job, that alone makes no difference since the nurses picked shifts from a list of many jobs—all paying different rates. So in no real sense did Steadfast pluck rates from thin air or compel nurses to accept them. By ignoring that it was the nurses who chose which shifts to take, the majority conjures an illusion that Steadfast controlled pay. But understanding the menu-of-options model that Steadfast used dispels this mirage.

This makes the majority’s selective quotation from our decision in McFeeley perplexing. True enough, in that case we thought it relevant that a nightclub set rates for private dances; that was an aspect of control. See 825 F.3d at 241–42. But that case well illustrates why this one is different: There, by setting one rate across the board for private dances, the club prevented dancers from making more money by focusing on clients willing to pay more. Here, in contrast, clients offer different rates—and sometimes offer to pay

90

more for particular nurses—which gives nurses the very freedom of contract that the dancers in McFeeley lacked.

Second, the majority stresses that Steadfast told nurses what shifts were available through its network. In making this argument, majority seems to accept Steadfast’s characterization of its system as a “free[] market” where “nurses could independently choose to accept or decline shifts.” Ante at 41. Even so, the majority suggests, Steadfast wielded a kind of indirect control by limiting the options available to any given nurse. But this seems to be rank speculation. No one testified that Steadfast limits which shifts each nurse can see; the evidence was that Steadfast sent out “general blast[s]” to the nurses in its database. J.A. 1000, 1187. And the district court never made any finding that Steadfast artificially limited which shifts a given nurse could see. Perhaps Steadfast could do such a thing. But with neither evidence nor a finding that it did, the majority’s insistence on this point is surprising.

More to the point, why does this matter? The majority seems to accept that (1)

nurses had a menu of options, even if that menu was curated by Steadfast; (2) Steadfast didn’t compel nurses to accept shifts offered; and (3) nurses could discover and take non- Steadfast assignments too. Against all this, does the majority mean to say that the mere fact of advertising some jobs and not others suggests an employment relationship? I hope not. If I wanted to hire a plumber, I might advertise only one job (fixing my backed-up sink) and not every other bit of work I may want done someday (replacing all my pipes). This hardly shows that I employ the plumber. And tellingly, the majority’s contrary

91

analysis finds no support in our precedent. Just check—the majority cites nothing for its claim that a limited-offer system affects worker classification.

Third, the majority observes that nurses turned in their time sheets to Steadfast, not client facilities. Ante at 42. But of course they did. An independent contractor sends invoices to whoever pays him for the job. Insofar as independent contractors can exist at all under FLSA, this point can carry no weight.

At long last, the majority ends its control analysis with a point that does carry some weight under our precedents: worker discipline. But on this point, the majority can’t be bothered to find any record support for its claims. It just quotes a sentence from the bench verdict, uncritically equating a finding that Steadfast is “responsible” for discipline, ante at 43 (emphasis added) (quoting Medical Staffing, 580 F. Supp. 3d at 233), with a conclusion that Steadfast did discipline nurses, and did so in the way our precedents require.

Yet as explained, the district court’s cited facts don’t support that legal conclusion.8 And the majority makes no effort to add the missing support, despite Steadfast’s extended

8

It cited only two relevant findings of fact. First, that nurses needed to tell Steadfast, not client facilities, “when they are running late for a shift, want time off, are sick, or otherwise cannot complete a shift.” Medical Staffing, 580 F. Supp. 3d at 223; see also id. at 233 (citing findings of fact). Second, that Steadfast disciplined nurses for violating the noncompete clauses, discussing compensation with each other, failing to appear for shifts, and being drunk on the job. See id. at 224; see also id. at 233 (citing findings of fact). The first point makes sense: Until nurses arrived, client facilities did not know who had agreed to work. As the middleman, Steadfast knew that information. So we should not be surprised that nurses had to contact Steadfast. As for the noncompetes, as I have said, the testimony did not support the claim. And last, discipline for ancillary matters like no- (Continued)

92

argument that there isn’t any. It does not identify a single situation where Steadfast disciplined a nurse for unsatisfactory nursing work.

Instead, the majority switches from a claim about discipline for bad work to a claim that Steadfast stopped connecting nurses with shifts if they no-showed, drank on the job, or wound up on clients’ DNR lists. Contra McFeeley, 825 F.3d at 242–43 (focusing on control over work and setting aside ancillary matters like a ban on drinking). The reason is clear: Only the latter finds support in either the district court’s findings or the record. But that doesn’t get the majority where it needs to go. It was the clients who reviewed nurse performance and controlled DNR list placements; Steadfast exercised no control over the work itself. The only sort of “discipline” it did was stop working with a nurse who agreed to do a job but never showed up to do it. And that is compatible with independent- contractor status. For who, stiffed by a contractor on one job, would ask that contractor to do future jobs?

2. Opportunities for profit and loss The majority only encounters more trouble when it reaches the nurses’ economic opportunities. The district court purported to find, as a matter of fact, that “[t]he nurses do not have opportunities for profit or loss depending on their managerial skill.” Medical Staffing, 580 F. Supp. 3d at 222. But of course, this is not a finding of fact. It’s the label our cases attach to one of the factors, and concluding that this factor cuts in favor of

showing and drinking on the job is just the kind of control we rejected in McFeeley because it is not distinctive to employment relationships. See 825 F.3d at 242–43.

93

employment requires identifying facts and explaining why they matter under our precedents. The district court did not do that work. It is instead the majority that tries to fill in the missing links. But its effort is unavailing.

Just take its lead argument that the nurses did not own equity in Steadfast. See ante at 46. That’s true, but it’s an odd fact to rely on. The point of a registry business model is to help connect workers with customers who want to buy their work. The workers then decide which offers to take—when, where, how, and for whom they will work. This should make clear that the registry and the nurses are in two different businesses. The registry matches buyers and sellers; the nurses provide medical care. So it should be no surprise that the nurses do not tether their income to the success of the registry by owning shares of it and instead tether their income to the success of the nursing work.

For the same reason, I struggle to see why the majority adds that Steadfast doesn’t use a profit-sharing compensation system. Again, the nurses are in the business of nursing, not matching nurses with customers. Plus, even if they were in the same business, why would it matter how the nurses are paid? A solo lawyer who consults on a case surely does not become an employee just because he is paid by the hour rather than on contingency. Need he buy some shares of his client’s business to secure his independence? Our cases do not require us to ignore common sense. Instead, they tell us to “adapt” our “analysis to the particular working relationship . . . and the particular industry.” McFeeley, 825 F.3d at 241. And so far as I can tell, when it comes to nurses and registries, equity compensation is an unhelpful indicator.

94

The majority’s other affirmative arguments are worse. It next leans on the noncompete provisions. But as explained, it misreads them and ignores the evidence that they were never enforced—which should take pride of place seeing as our test focuses on economic realities.9 Then the majority recites another bald statement from the district court, this time that nurses had no way to increase their pay without working more hours. See ante at 47. But here, it repeats another mistake I have explained: It says Steadfast set the rate for each shift and then proceeds as though Steadfast set the same rate for all shifts. Not even the district court reached that conclusion, and it is belied by the record. See, e.g., J.A. 704, 733–34, 760, 872.

In the end, the majority resorts to ipse dixit, asserting that “the ‘lion’s share of managerial skill . . . came from [Steadfast].’” Ante at 49 (quoting McFeeley, 825 F.3d at 243) (alteration original). If only the majority followed McFeeley’s lead and explained

9

Perhaps sensing this problem, the majority reinforces its claim about the noncompetes in two ways. First, it reiterates that “the Bench Verdict’s factual findings reflect [the district court’s] careful consideration of the witnesses’ credibility.” Ante at 48. But if the district court performed a careful credibility analysis, it didn’t write it down; its opinion contains nothing of the sort. More to the point, although the majority seems to be implying that we shouldn’t trust the nurses’ testimony, it won’t commit to telling us which testimony it thinks we should set aside, or why. This is a curious way to handle Steadfast’s clear-error argument. Second, the majority speculates that even if Steadfast never enforced the noncompete, it may have hung like Damocles’s sword over the nurses and thus had a kind of chilling effect. But this cannot be reconciled with the overwhelming testimony that nurses did work outside Steadfast’s network while they were on the registry.

Plus, as Steadfast pointed out at argument, the only testimony that even suggests a chilling effect came from nurses who worked with Steadfast before 2018, when Steadfast adjusted the terms of its contracts. Oral Arg. at 44:40–45:25. So if the majority thinks the noncompetes are decisive, why does it think that all of Steadfast’s nurses worked as employees, rather than just the nurses who worked with Steadfast through 2017?

95

what evidence supports that conclusion. See McFeeley, 825 F.3d at 243 (summarizing the evidence). But it simply does not say what managerial tasks it thinks Steadfast performed. And neither did the district court. Presumably for the same reason: The record contains no evidence of any.

Meanwhile, the majority discounts the evidence I focus on—that the nurses’

opportunities for profit depended on their skill at nursing—by citing an out-of-circuit decision that, it says, looks only to “managerial skill.” Ante at 46 (quoting Scantland v. Jeffry Knight, Inc., 721 F.3d 1308, 1316 (11th Cir. 2013)). But that is not how our precedents approach the matter. See, e.g., Hall, 846 F.3d at 774 (asking whether workers could “increase their take-home pay through their own ingenuity or skill”). And here, it’s clear that the nurses could do just that—in part by entrepreneurial skill (arranging their schedules from a menu of Steadfast and non-Steadfast options) and in part by medical skill (impressing client facilities and thereby increasing pay).

3. Investment in the business The majority’s conceptual confusion deepens when it comes to the nurses’

investment in the business. For at the outset, even the district court’s findings of fact suggest that this factor should cut decisively for Steadfast. In order to work, the nurses need training, licensure, and equipment. Yet Steadfast handles none of that. Training and licensure are prerequisites to placement on its registry. Medical Staffing, 580 F. Supp. 3d at 222 (“Nurses must obtain and maintain their own licensure. Steadfast does not pay for or reimburse nurses for licensing or educational expenses.” (citation omitted)). And on the job, nurses use a combination of their own equipment (stethoscopes, blood pressure cuffs,

96

and so on) and client facilities’ equipment (machines that can’t easily be transported from job to job). “Steadfast does not provide the nurses with any equipment.” Id. Instead, the nurses—being “personally invested in the capital and labor of the[ir] enterprise”—are not “economically dependent on the business” and could easily strike out on their own. McFeeley, 825 F.3d at 241 (quotation omitted).

Finding no support there, the majority tries a different tack. It emphasizes that the nurses didn’t pay for Steadfast’s office or overhead costs. See ante at 50. But of course they didn’t; none of that helps provide nursing services. It may be that Steadfast’s business requires those investments. Yet this just underscores that Steadfast is not in the same business as the nurses. Rather than provide nursing services itself, Steadfast connects people who do sell those services with those who want to buy them. We should not be surprised that Steadfast invested more than the nurses in its own business—just as we should be unsurprised to hear that a restaurant invests more in its building, seating, and décor than do musicians who play gigs there on Friday nights. The majority cannot satisfy this factor by pointing to the absence of evidence about a point that makes no difference under our precedents.

What the majority really needs to put this factor to work is to show that Steadfast invested more than the nurses did in the nursing business—not the matchmaking business. See, e.g., McFeeley, 825 F.3d at 244 (“The focus . . . should remain on the worker’s . . . investment relative to the company’s.”). Yet it cannot show that. So instead, the majority retreats to generalities, saying that “large capital expenditures, such as risk capital” should be “dispositive,” ante at 50 (quoting Dole v. Snell, 875 F.2d 802, 810 (10th

97

Cir. 1989)), and that contractors “normally” make “significant capital investments,” id. at 51.

But this overlooks the nature of the industry. Nursing is indeed a capital-intensive business, but Steadfast provides none of that capital. And if the majority widened its lens to examine the cost of medical care in general, it would only reveal more reason to think the nurses do not depend on Steadfast. Nursing itself may often be a low-cost activity, but nurses often use expensive medical equipment—and they do so in expensive facilities. Yet these investments are borne neither by the nurses nor by Steadfast. Client facilities like hospitals and nursing homes instead buy the expensive machines and furnish the elaborate facilities that some kinds of medical care require. So to compare like with like, we would assess not the nurses’ investment relative to Steadfast but their investment relative to the clients. See McFeeley, 825 F.3d at 244.10

10

And if the majority thinks independent-contractor status requires capital-intensive facilities and equipment, then many industries couldn’t have independent contractors at all. Just consider online businesses that require little more up front than a laptop and internet service. Happily, FLSA does not require this counterintuitive result. After all, “[t]he opportunity to obtain profit from efficient management is not the same as exposing a stock of capital to a risk of loss.” Sec’y of Lab., U.S. Dep’t of Lab. v. Lauritzen, 835 F.2d 1529, 1540 (7th Cir. 1987) (Easterbrook, J., concurring). And “[t]his is true of many workers we would call independent contractors. Think of lawyers, many of whom do not even own books.” Id.

98

4. Degree of skill, permanence of the relationship, and importance of nurses to Steadfast’s business

The majority marches through the three remaining factors, but its heart isn’t in it.

For good reason: Each adds more length than luminance to the majority’s analysis, yet none cuts clearly in its favor.

When it comes to the nurses’ degree of skill, the majority all but concedes—as did the district court—that this factor supports an independent-contractor classification. Nurses are skilled workers who cannot ply their trade without formal education and State imprimatur. Those entry barriers, coupled with a liquid market, give them economic freedom that unskilled laborers may lack. See Silk, 331 U.S. at 714 n.8. As the district court put it, these “suggest[] that the nurses could work in an independent capacity.” Medical Staffing, 580 F. Supp. 3d at 234. And indeed, they do. The nurses “use their skills ‘in any independent way’ to secure business opportunities or ‘find job assignments.’” Ante at 52 (quoting Brock v. Superior Care, Inc., 840 F.2d 1054, 1060 (2d Cir. 1988)). Only by ignoring that the nurses often worked outside Steadfast’s network can the majority plausibly claim this factor for itself.

As for the permanence of the working relationship, the majority again accepts a key premise: The evidence showed that the nurses entered only short-term contracts with Steadfast. See ante at 54. While not dispositive, see McFeeley, 825 F.3d at 244, the short- term nature of these gigs points toward independent-contractor status. And indeed, this is the whole point of Steadfast’s business model. Facilities providing medical care often cannot fulfill all their needs through permanent employees alone. To fill the gap, they seek

99

out temporary help. Then Steadfast helps them find it, connecting them with nurses who would rather work short-term gigs than join any one facility on a permanent basis.

Neither of the majority’s two counterarguments persuades. First, it returns to the noncompetes and claims that within a nurse’s contract, she had little economic independence because of these provisions. But I have said why this reliance is misplaced. Second, the majority gives up on permanence as such and argues instead that “even short, exclusive relationships . . . may be indicative of an employee-employer relationship.” Ante at 55 (quoting Keller v. Miri Microsystems LLC, 781 F.3d 799, 807 (6th Cir. 2015)). Maybe so. But if that’s right, what is the point of our six-factor test? As the majority sees things, if the relationship is permanent, that carries some weight. If it isn’t, fear not—the factors are malleable enough to account for anything.

The final factor is whether the nurses are integral to Steadfast’s business model.

This one isn’t hard: No one doubts that the nurses are integral to Steadfast’s business. See Brief for Appellants 44; Brief for Appellee 49. For that is how registries work. But this is just one, nondecisive factor. Unless the majority thinks all registries work only with employees—more on that soon—this factor should carry little weight.

*

Two last points. First, the majority is seemingly more concerned with FLSA’s “‘remedial and humanitarian’ purpose” than messy facts. Ante at 55–56 (quoting Salinas, 848 F.3d at 140). That is unfortunate. Although factor tests are open-ended by nature, they are not license to do just anything. Instead, our precedents instruct that some facts matter and others don’t. Reasonable people could disagree about which way the facts cut

100

or how one factor weighs against another. But in the majority’s ostensibly humanitarian zeal, it relies on whatever facts are available—whether or not our precedent says that they matter. When our precedent offers support, the majority cites it with gusto. When it does not, the majority can always find another circuit that sees things differently. And when it can find no support beyond its intuitions, the majority apparently has no trouble relying on those. Even when our cases tell us to apply a balancing test, this is not how such a test should work.

Second, although the law the majority creates may seem purpose-built to deal with Steadfast’s case (and it is), make no mistake: The tools the majority forges will work just as well on other registry-style businesses. Today, this sort of decentralized business is everywhere, in industries from transportation to handyman work to graphic design. Many workers seemingly would rather embark on their own ventures than tether their fortunes to those of a single employer. Yet a lone worker seldom has the resources and expertise to make the most of the market. For these workers, services like Uber, TaskRabbit, and Fiverr can make all the difference—letting workers set their own hours and rely on their own equipment and skills yet benefit from scale and the services’ expertise when it comes to marketing, sales, and other administrative functions. And consumers benefit from these platforms too, getting easy access to many options through one clearinghouse.

That is not just my judgment; it is the judgment of the Department of Labor. Just two months ago, the Department explained that under the economic realities test, workers available through virtual marketplace services are often independent contractors. See U.S. Dep’t of Lab., Wage & Hour Div., Opinion Letter FLSA2025-02 at 7–8 (May 2, 2025); see

101

also U.S. Dep’t of Lab., Wage & Hour Div., Field Assistance Bulletin No. 2025-1 at 2 (May 1, 2025) (rescinding more-restrictive 2024 guidance). In reaching that conclusion, the Department stressed that such registries often do not “impose any duties” on workers to take one or another job yet do give “service providers the right to work simultaneously for competitors.” May 2 Opinion Letter, supra, at 8. The Department also emphasized that these services often do not tell workers how to work, supervise the work, or “control the particulars of that work.” Id. at 9. And it pointed out that workers “regularly use this flexibility to their own profit and personal advantage.” Id. at 8.

All that, of course, describes this case. So under the majority’s analysis, it is hard to see how the platforms the Department addresses work with contractors rather than employees. That too is unfortunate. As the Department’s recent guidance suggests, questions about how to regulate registries and virtual marketplaces are the subject of active debate in the political branches. As they should be: These economic innovations may bring promise for some and peril for others, and weighing those factors is a job for lawmakers, not law-interpreters. Yet after today, in the five states that form the Fourth Circuit, workers and consumers alike have fewer options—not because of a regulator’s policy judgment but because of a court’s.

In the name of helping nurses, of course. But does the majority even do that? We should not forget that nobody obligates nurses to work with registries. If they wish to exchange the flexibility and higher earning ceiling associated with independent work for the consistency, predictability, and nonwage benefits of working for only one buyer of labor, nurses could always move in-house—as some formerly independent nurses testified

102

that they did in this very case. The majority’s decision does not help those nurses; it just takes from them that choice.

* * *

To reach its conclusion, the majority sets aside Steadfast’s arguments, ignores inconvenient testimony, and distorts our precedent. And that holding sweeps well beyond this case, depriving workers, consumers, and regulators of the chance to work out these problems for themselves. Neither the facts nor the law requires that result. I respectfully dissent.