GRIFFITH, Circuit Judge:
After agreeing to a representation election in which the union prevailed, employer SSC Mystic challenged the results. For the reasons set forth below, we reject each of Mystic’s arguments and affirm the decision of the National Labor Relations Board upholding the outcome.
I
SSC Mystic (Mystic) operates Pendleton Health & Rehabilitation, a nursing home in Mystic, Connecticut. On February 25, 2013, the Service Employees International Union, Local 1199 (Union), filed a petition with the National Labor Relations Board (NLRB) seeking to represent nurses at the facility. In response, the NLRB Regional Director issued a Notice of Election. The Union and the company entered a Stipulated Election Agreement that, among other things, provided that either party could ask the Board to review any decision the Regional Directors made. See 29 C.F.R. § 102.69(c). .
Mystic vigorously opposed the Union. Its campaign included posting anti-union material in the workplace and sending the material by mail to employees’ homes. Mystic also held meetings at work to make the case against the Union to its employees, who were required to attend. It also distributed anti-union bracelets for employees to wear.
Separately, a supervisor named Diane Mackin engaged in a campaign of urging employees to sign Union authorization cards and to vote for the Union in the election. She frequently discussed the virtues of organizing. To those who opposed the Union, Mackin would speak coldly or refuse to speak at all. Mackin also claimed that the Union would help her get her job back if Mystic fired her for her advocacy.
After an employee reported Mackin’s pro-union conduct to management, the company reprimanded her on March 12, 2013. Mystic explained to Mackin that her conduct violated her professional responsibilities as a supervisor and, more seriously, might be illegal pressure on employees in violation of the National Labor Relations Act (NLRA). Mystic warned Mackin that she would be fired if she did not end her support for the Union. Mystic then posted a notice in the workplace acknowledging, without identifying Mackin by name, that a supervisor had been involved in electioneering advocacy on behalf of the Union. In an effort to limit any effect Mackin’s conduct may have had on employees’ plans to vote, the notice explained that neither the company nor its supervisors intended to place pressure on employees. Despite all this, Mackin continued to openly advocate for the Union in the election and Mystic fired her on March 19, 2013.
The election continued for the next sixteen days. On April 4, 2013, the Union won the election. Of the 112 employees in the bargaining unit, 104 voted in the elec[306]*306tion: 64 supported the Union while 40 opposed.
Mystic filed objections to the election with the NLRB arguing principally that Mackin’s conduct had tainted the election so thoroughly that its result should be set aside. Mystic also alleged that Mackin was acting as an agent of the Union when she “polled” employees, or interrogated them regarding their support for the Union in a way that could coerce them and infringe on their free choice. Because Mackin .was allegedly acting as a Union agent, the company argued that the Union should be held responsible for that misconduct.1 Finally, Mystic insisted, relying on our decision in Noel Canning v. NLRB, 705 F.3d 490 (D.C.Cir.2013),2 that the NLRB lacked a quorum because three of its members had been placed in their posts through unconstitutional recess appointments and so had no authority to conduct the election at all.
On May 8 and 9, 2013, an NLRB Hearing Officer held a hearing to consider Mystic’s objections. A party to a representation proceeding may apply for and receive a subpoena for the production of any evidence. 29 C.F.R. § 102.31. Exercising that power, Mystic subpoenaed any records of telephone calls between Mackin and the Union organizer assigned to the election. The Union opposed this subpoena. Mystic argued that it needed these records to prove that Mackin was a Union agent when she coercively interrogated employees regarding their support for the Union. The Hearing Officer refused to enforce the subpoena, concluding that records could not prove that Mackin was acting as the Union’s agent. Instead, the Hearing Officer directed the Union to produce the organizer himself to testify about his relationship with Mackin. The Union did not do so. Neither the parties nor the Hearing Officer mentioned the subpoéna or the organizer again on the record.
At the close of the hearing, the Hearing Officer upheld the election result, concluding that even though Mackin had exerted impermissible pressure on employees, her misconduct had not materially affected the outcome of the election. The Hearing Officer also rejected Mystic’s argument that Mackin was acting as a Union agent, reasoning that the company had failed to present any evidence supporting its claim. Finally, the Hearing Officer concluded that the Board should continue conducting elections and adjudicating disputes until the Supreme Court decided the legality of the Board’s composition in Noel Canning.
Mystic filed objections to the Hearing Officer’s ruling with the Board, arguing that the Hearing Officer’s findings and conclusion were in error. Nonetheless the Board ratified the Hearing Officer’s legal and factual determinations and certified the election result. SSC Mystic Operating Co., No. 01-RC-098982, 2013 WL 6252453 (Dec. 3,‘ 2013) (unreported). The Board agreed with the Hearing Officer that Mac-kin’s impermissible conduct had not affected the outcome of the election, especially on the ground that Mackin’s activities were offset when Mystic “engaged in an extensive [anti-union] campaign that included a string of mandatory meetings [307]*307during the critical period, the dissemination of [anti-union] literature via mailings, handouts, and postings, and the distribution of [anti-union] bracelets.” Id. at *1 n. 2.
Once the Board had certified the election result, the Union asked Mystic to bargain, but the company refused. Accordingly, the Union filed an unfair labor practice charge against Mystic, alleging that its refusal to bargain violated the NLRA. See 29 U.S.C. § 158(a)(1), (5) (prohibiting an employer from refusing to bargain with representatives of its employees or interfering with employees’ rights to organize). The Board’s General Counsel issued a complaint and moved for summary judgment. In response, Mystic argued that the Hearing Officer erred in refusing to enforce Mystic’s subpoena and should have held that Mackin’s conduct impermissibly contaminated the election. For the first time, Mystic also raised the argument that the Regional Director, as opposed to the Board itself, had no power to conduct the representation election because he could not exercise the Board’s delegated authority when the Board had no quorum and could not act itself.
The Board granted summary judgment against Mystic on March 31, 2014. SSC Mystic Operating Co. LLC d/b/a Pendleton Health & Rehab. Ctr., 360 N.L.R.B. No. 68 (2014). The Board rejected Mystic’s arguments that the Hearing Officer had made substantive and procedural errors, finding that Mystic had not produced any arguments or evidence not already made and rejected when the Board certified the election result. The Board also rejected Mystic’s new argument that the Regional Director lacked authority to administer this representation election because the Board lacked a quorum. The Board interpreted the statute to mean that the Regional Directors “remain vested with the authority to conduct elections,” pursuant to the Board’s original delegation of that authority in 1961, “regardless of the Board’s composition at any given moment.” Id. at *1 n. 1.
Mystic filed a timely petition for review of the Board’s order, and the Board cross-applied for enforcement. We have jurisdiction under 29 U.S.C. § 160(e), (f).
On appeal, Mystic raises three challenges, each with its own standard of review. First, Mystic argues that the Board could not interpret the NLRA to permit Regional Directors to continue conducting elections when the Board lacked authority to act due to lack of a quorum. Absent plain meaning to the contrary, a court is obliged to defer to an agency’s reasonable interpretation of its statutory jurisdiction pursuant to the familiar Chevron doctrine. City of Arlington v. FCC, — U.S. -, 133 S.Ct. 1863, 1870-71, — L.Ed.2d-(2013).
Second, Mystic argues that substantial evidence did not support the Hearing Officer’s decision to certify the election results. We review the substance of NLRB decisions under a “highly deferential standard” and will set them aside only “if the Board ‘acted arbitrarily or otherwise erred in applying established law to the facts at issue, or if its findings are not supported by substantial evidence.’ ” Waterbury Hotel Mgmt., LLC v. NLRB, 314 F.3d 645, 650 (D.C.Cir.2003) (quoting Plumbers & Pipe Fitters Local Union No. 32 v. NLRB, 50 F.3d 29, 32 (D.C.Cir.1995)).
Finally, Mystic challenges the Hearing Officer’s refusal to enforce its subpoena. We review refusals to enforce subpoenas for abuse of discretion. Joseph T. Ryerson & Son, Inc. v. NLRB, 216 F.3d 1146, 1153 (D.C.Cir.2000).
[308]*308II
A
Mystic insists that the Regional Director did not have authority to conduct this election because the Board had no quorum at the time the representation election took place. We disagree; as we recently explained in UC Health v. NLRB, No. 14-1049, slip op. at 8-19 (D.C.Cir.2015), 803 F.3d 669, 673-79, 2015 WL 5474171, we must defer to the Board’s reasonable interpretation that the lack of a quorum at the Board does not prevent Regional Directors from continuing to exercise delegated authority that is not final because it is subject to eventual review by the Board.
As an initial matter, the Board argues that Mystic waived this argument by failing to raise it during the representation proceeding. The Board made the same argument in UC Health, and we rejected it there. We do so here for the same reasons: Our precedents make clear that a challenge1 to agency action based on the agency’s lack of authority to take any action at all need not be raised below and may be made for the first time on appeal. See UC Health, No. 14-1049, slip op. at 6-7, 803 F.3d at 672-73.3 Nor do we agree with the Board that Mystic abandoned this argument when it executed the Stipulated Election Agreement. Id. at 7-8, 803 F.3d at 672-73. Nonetheless, just as in UC Health, we disagree with Mystic on the merits of its claim. The Regional Director had authority to conduct this election even though the Board had no quorum. See id. at 8-19, 803 F.3d at 673-79.
Mystic makes one additional argument on this score that we did not confront in UC Health. Mystic insists that Regional Director Jonathan Kreisberg did not have authority to conduct this election even if the Regional Directors as a class could do so. In 2010, Kreisberg was appointed as the Regional Director for Connecticut, which was at that time Region 34 of the NLRB’s regions. In 2012, while the Board lacked a quorum, the NLRB reorganized the regions and Kreisberg’s jurisdiction expanded to cover both Connecticut and Massachusetts, now identified as new Region 1. Mystic insists that because the Board had no quorum in 2012, it could not validly- appoint Kreisberg to his new post as the Regional Director of new Region 1 at that time.
The Board again argues that Mystic waived this argument because it was never made until the opening brief in this appeal. We disagree. Because this challenge and the argument that Regional Directors may not conduct elections while the Board lacks a quorum are both premised on the Board’s lack of authority to act, we believe both are properly before us no matter when they were first raised. Nonetheless we reject Mystic’s argument on the merits here as well. Mystic’s nursing home is located in Mystic, Connecticut, inside the boundaries of old Region 34, which covered Connecticut alone. Mystic does not and could not contest that Kreisberg was validly appointed to administer old Region 34. There may be some question whether the Board had authority in 2012 to expand [309]*309Kreisberg’s jurisdiction to include Massachusetts, but that seems irrelevant to the question of whether he continued to have authority to conduct elections in Connecticut as he had since 2010. Surely adding Massachusetts to his jurisdiction or renaming the region he administered did not impair his preexisting authority. Therefore we believe that his ability to conduct this election remains beyond dispute.4
B
Mystic argues that Diane Mackin’s supervisory misconduct tainted the outcome of the election. We find that substantial evidence supports the Board’s conclusion to the contrary.
Section 7 of the NLRA secures the rights of employees “to form, join, or assist labor organizations, to bargain collectively through representatives uf their own choosing, and to engage in other concerted activities for the purpose of collective bargaining,” as well as to refrain from all such activities. 29 U.S.C. § 157. To ensure that employees are fully able to exercise their section 7 rights, the Board requires that elections take place under “laboratory conditions” free from coercion by the union or the employer. Harborside Healthcare, Inc., 343 N.L.R.B. 906, 909 (2004). Neither employers nor unions may “interfere with, restrain, or coerce employees in the exercise” of their .section 7 rights. 29 U.S.C. § 158(a)(1), (b)(1)(A). Supervisors, defined as individuals with authority to direct, reward, or punish employees, id. § 152(11), do not hold section 7 rights. To the contrary, supervisors may not participate in or try to influence the outcome of an election any more than an employer itself is permitted to do so: “Election campaign statements by supervisors which reásonably cause [pro-union] employees to fear reprisal or to expect reward if they exercise their section 7 rights will ordinarily be attributed to the employer and found objectionable.” Harborside, 343 N.L.R.B. at 906.
Of course, as a general matter, supervisors may be more likely to urge employees to oppose union organization than to support it because their interests are more aligned with those of the employer than those of the employees who seek to organize. However, pro-union supervisory conduct is just as impermissible because it poses the same risk of interfering with the free choice of employees. Harborside, 343 N.L.R.B. at 906. In other words, the law always forbids a supervisor from trying to influence the free choice of employees in exercising their section 7 rights, regardless of what outcome the supervisor is seeking to achieve. “This is true whether or not the statements or actions of the supervisor are consistent with the views of the employer.” Id. at 907. After all, the average “employee is more concerned about the attitude of his immediate supervisor[s] than he is with the feelings of the company president,” as his immediate supervisors “control his day to day life.” Id. at 907 n. 3 (quoting Turner’s Express, Inc. v. NLRB, 456 F.2d 289, 292-93 (4th Cir.1972)).
[310]*310In Harborside, the Board established a two-step inquiry to determine “whether supervisory [pro-union] conduct upsets the requisite laboratory conditions for a fair election” such that the election result is invalid. 343 N.L.R.B. at 909. At the first step, the Board asks “[w]hether the supervisor’s ... conduct reasonably tended to coerce or interfere with the employees’ exercise of free choice in the election.” Id. If so, the Board moves to the second step and asks “[w]hether the conduct interfered with freedom of choice to the extent that it materially affected the outcome of the election.” Id. The effect of an individual episode of supervisory misconduct depends on “factors such as (a) the margin of victory in the election; (b) whether the conduct at issue was widespread or isolated; (c) the timing of the conduct; (d) the extent to which the conduct became known; and (e) the lingering effect of the conduct.” Id. In other words, even conduct that actually interferes with employee choice will not invalidate the election result unless it actually influenced the outcome. But if a supervisor’s pressure on employees played a meaningful role in the union’s victory or the union’s defeat, the Board will throw out the result and order a new election.
The Board measures the effect of a supervisor’s impermissible conduct by also taking into account any “mitigating circumstances” that may have “sufficiently negated” the coercive activities such that the election result was not materially affected. Veritas Health Servs., Inc. v. NLRB, 671 F.3d 1267, 1272 (D.C.Cir.2012) (quoting SNE Enters., Inc., 348 N.L.R.B. 1041, 1042 (2006)). For example, the employer can mitigate a supervisor’s conduct if it “ ‘takes timely and effective steps to disavow’ the conduct.” SNE Enters., 348 N.L.R.B. at 1043 (quoting Harborside, 343 N.L.R.B. at 914). That is, if the employer publicly announces that a supervisor lobbying on the union’s behalf is acting against the employer’s wishes, the employer limits the risk that employees will feel coerced. Employees will understand that the supervisor is simply a rogue agent and does not have the employer’s support.
Separately, the Board also determines whether any anti-union effort by the employer itself had the effect of counteracting a supervisor’s pro-union conduct. Harborside, 343 N.L.R.B. at 914. Of course, the NLRA forbids employer anti-union campaigns just as surely as it forbids pro-union lobbying by supervisors. However, the Board’s inquiry focuses on the validity of the election as a whole, not simply on whether inappropriate conduct took place during the election period. An employer’s effort to defeat a union does not violate the law if the union wins. More to the point, if the employer works at cross-purposes to a supervisor’s pro-union activity during an election, the employer may end up neutralizing the supervisor’s wrongdoing and inadvertently preserve the conditions necessary to reach a valid election result.
The record is clear and both parties acknowledge that Mackin’s pro-union conduct satisfies the first step of the Har-borside analysis. Nonetheless, at the second step of Harborside, the Board reasonably determined that Mackin’s efforts did not materially affect the election’s outcome because Mystic adequately made up for them by disavowing Mackin’s conduct and by running its own anti-union campaign.
Substantial evidence supported this determination. Mystic required employees to attend anti-union meetings, sent materials to their homes, posted materials in the workplace, and even distributed anti-union bracelets for employees to wear at work as a way of showing their opposition to the Union. Mystic’s campaign was much like another employer’s efforts to defeat a un[311]*311ion that the Board found neutralized pro-union conduct by supervisors. In Terry Machine Co., 356 N.L.R.B. No. 120 (2011), supervisors who oversaw the bargaining unit were “actively involved” in a union organizing drive. Id. at *2. At the same time, the employer “engaged in an extensive [anti-union] campaign,” including mandatory company-wide meetings, individual meetings with employees, anti-union videos, anti-union postings, home mailings, and distribution of anti-union buttons to wear at work. Id. at *3. The Board upheld the election result despite the supervisors’ substantial pro-union conduct, concluding that the employer’s, own anti-union campaign had adequately offset the supervisors’ efforts. Id. at *5. The Board here reasonably concluded, just as it did in Terry Machine, that the combination of tactics Mystic deployed in its extensive effort to defeat the Union cancelled out Mackin’s own attempt to help the Union prevail.
Mystic argues otherwise by attempting to minimize the significance of each element of its own anti-union program. It insists that few employees saw the anti-union materials, attended the anti-union meetings, or understood the intent behind the anti-union bracelets. None of these challenges to the Board’s determination succeed. A number of employees testified that they received Mystic’s anti-union materials through the mail or saw them posted in the workplace, and one even testified that she knew of other employees who had discussed the materials during the election. Although some employees testified that the anti-union meetings were sparsely attended, there was also testimony that “a lot” of the staff attended a meeting at one point or another. And while one employee testified that she did not recognize Mystic’s anti-union bracelet, a number of other employees testified that they knew what the bracelets were for, wore bracelets themselves, and saw others wearing them. We cannot say that “no reasonable factfinder” could decide, as the Board did here, that Mystic’s campaign was effective at neutralizing Mackin’s pro-union advocacy. Kiewit Power Constructors Co. v. NLRB, 652 F.3d 22, 25 (D.C.Cir.2011) (quoting United Steelworkers of Am., AFL-CIO-CLC, Local Union 14534 v. NLRB, 983 F.2d 240, 244 (D.C.Cir.1993)).
Substantial evidence also supported the conclusion that Mystic limited the effect of Mackin’s conduct when it posted a public notice that disavowed her pro-union behavior, discussed the notice at mandatory employee meetings, and ultimately fired her. See, e.g., Terry Machine, 356 N.L.R.B. No. 120, at *3 (finding that an employer’s “explicit disavowals” and “widely disseminated termination threat ... relieved any potential continuing pressure employees might have felt” from pro-union supervisory conduct). Mystic argues otherwise by suggesting that few employees ever saw the notice, that most employees did not attend the mandatory meetings and so would never have heard it discussed, and that any employees who were aware that the notice existed would not have realized it referred to Mackin because it did not identify her by name. We think the Board could reasonably reach the opposite conclusion on each count. Five employees testified that they saw the notice, and we have already noted that there was testimony indicating that “a lot” of employees attended the mandatory anti-union meetings at which the notice was discussed. One employee who saw the notice indicated that she knew the notice applied to Mackin in particular. Another testified that she understood the notice to refer to all supervisors who may have been inappropriately discussing the election — obviously including Mackin. Most significantly, Mackin told a number of employees around the time the notice was posted that [312]*312Mystic had reprimanded her for advocating on behalf of the Union. In one case, she told an employee that the notice addressed her own behavior in particular. The Board could reasonably rely on all this evidence to conclude that employees knew of the notice and understood that Mystic was disavowing Mackin’s conduct.
Even if employees somehow missed the existence or significance of the notice, they could not have misunderstood that Mystic was disavowing Mackin’s pro-union behavior when it took the much more dramatic step of firing her. Though some employ^ ees testified that they did not know why Mackin was terminated, the Hearing Officer specifically found, and the Board subsequently agreed, that this testimony was not credible. See SSC Mystic, 2013 WL 6252453, at *1 n. 2. Mystic has not challenged that credibility determination on appeal. Thus the only credible .testimony before us comes from employees who said that they knew Mackin had been fired because of her pro-Union efforts. Joint Appendix 152. The Board was entitled to rely on this undisputed testimony to reach the commonsense conclusion that employees knew Mystic was conclusively disavowing Mackin’s conduct by firing her. See SNE Enters., 348 N.L.R.B. at 1043 (noting that an election result can be valid despite inappropriate pro-union supervisory conduct where an employer “ ‘takes timely and effective steps to disavow’ the conduct” (quoting Harborside, 343 N.L.R.B. at 913)).
We also agree with the Board that Mac-kin’s firing limited the effect of her conduct despite the fact that she assured employees that the Union would help her get her job back. Mystic insists to the contrary that these assurances “blunted the impact” of Mackin’s discharge by leading employees to believe that she would return to the workplace and regain the power to retaliate against the Union’s opponents. But the opposite seems to be true. The record shows that several different employees who were subject to Mackin’s pro-union pressure ended up opposing the Union by the time of the election, two weeks after Mackin was fired. Whatever the immediate effect of Mackin’s campaign, employees who were among its targets were unafraid to oppose the Union after her discharge. And Mystic produced no evidence indicating that employees she pressured to support the Union actually did so. The Board was entitled to conclude from this that Mackin’s firing had broken whatever hold she might have exercised over employees.
In short, substantial evidence supports the Board’s conclusion that Mystic’s efforts to limit Mackin’s effectiveness and its own anti-union campaign cancelled out Mac-kin’s efforts on the Union’s behalf and preserved the environment necessary for a valid representation election.
The Hearing Officer also noted that a number of other factors diminished the likelihood that Mackin influenced the election result. For example, Mackin was the sole pro-union organizer, naturally limiting the total amount of pressure that could be brought to bear on the Union’s side of the ledger. And the election was not a close one. The Union won by sixty-four votes to forty, or almost one quarter of the entire voting population, indicating that any influence Mackin might have wielded over a few employees could not possibly have altered the result. Nor did Mackin’s conduct “linger[ ],” Harborside, 343 N.L.R.B. at 909, as any influence she might have wielded at one time apparently dissipated before the end of the election period. The Board was entitled to rely on all these factors as part of its conclusion that Mac-kin’s campaign did not materially alter the election outcome.
[313]*313The Board was also entitled to conclude that the length of the time between Mac-kin’s discharge and the election further limited the impact Mackin’s efforts could have had on the outcome. Mystic insists that this decision was forbidden in light of Board decisions in which, it argues, the Board invalidated an election despite even longer intervals between the end of inappropriate supervisory conduct and an election. But in each of the cases Mackin cites, supervisors either continued to lobby for the union throughout the election period, or the interval was immaterial because other factors helped the supervisors’ influence linger. For example, in several of the cases, supervisors continued to campaign for the union “right up until the ... election” took place. Madison Square Garden CT, LLC, 350 N.L.R.B. 117, 122 (2007); see also Millard Refrigerated Servs., Inc., 345 N.L.R.B. 1143, 1144 (2005); Harborside, 343 N.L.R.B. at 913-14. And in the others, though the supervisors stopped campaigning before the election, the employer never publicly disavowed the supervisors’ conduct and the supervisors remained in the workplace, allowing their pro-union pressure to linger. See SNE Enters., 348 N.L.R.B. at 1044; Chinese Daily News, 344 N.L.R.B. 1071, 1072 (2005). This case is quite different. Mystic forcefully disavowed Mackin’s conduct and fired her, dispelling the influence she might otherwise have exercised. No case forbids the Board’s conclusion on this score. Absent such precedent, we cannot say the Board was wrong to decide that the effects of Mackin’s conduct had at least in part evaporated by the time the election took place, especially when considered in conjunction with the other factors we have already discussed that limited Mackin’s possible influence on the election.
Mystic points to Veritas Health Ser- . vices, arguing that much more is required to neutralize the impact of a supervisor’s pro-union activity than was present here. In Veritas, supervisors who pressured employees on behalf of the union ultimately switched sides and became fervent anti-union advocates, speaking directly to employees in the workplace and sending letters to most of the staff explaining that they no longer supported unionization. Veritas, 671 F.3d at 1273. This about-face, the Board found, neutralized the supervisors’ previous pro-union conduct. Id. Mystic relies on Veritas to argue that Mackin’s pro-union conduct was not mitigated here because Mackin herself never disavowed her past support for the Union. But Veritas does not suggest that the only permissible form of mitigation is personal disavowal by the supervisor. And the Board has elsewhere found that an employer’s own anti-union campaign can cancel out supervisory conduct like Mackin’s. See, e.g., Terry Machine, 356 N.L.R.B. No. 120, at *3, *5. The Board was entitled to do the same here.
Finally, Mystic insists that the Board unfairly showed more lenience toward Mackin’s pro-union conduct than it would have shown had Mackin successfully urged employees to vote against the Union instead. See Harborside, 343 N.L.R.B. at 906-07 (holding that both pro- and anti-union coercion áre equally impermissible). We need not engage with the hypothetical circumstance Mystic would have us imagine. Mystic has offered no support for its assertion that the Board displayed bias. The Board ruled that Mystic’s anti-union campaign made up for Mackin’s impermissible pro-union conduct, just as it has found in the past. There is no basis to criticize the Board’s conclusion regarding what actually transpired here.
Mackin’s campaign to help the Union succeed was inappropriate. However, a number of factors showed its limited effectiveness: Mackin acted alone; she was [314]*314dismissed from the workplace well before the election took place; her conduct apparently had little lingering effect; and the Union prevailed by a substantial margin. By disavowing Mackin’s pro-union advocacy and ultimately firing her, Mystic further minimized her impact on the result of the election. And Mystic’s own efforts to defeat the union provided a powerful counterbalance to Mackin’s lobbying. Based on this record, the Board was entitled to conclude that the election result chal- . lenged here was valid.
C
Finally, Mystic argues that the Hearing Officer erred by refusing to enforce Mystic’s subpoena of Mackin’s telephone records, which the company claims would show that Mackin was acting as the Union’s agent. Refusing to enforce this subpoena did not prejudice Mystic. See Ryerson, 216 F.3d at 1154 (noting that we will only reverse the Board’s decision not to enforce a subpoena “if prejudicial”). Even proving that Mackin was a Union agent would not have altered the Board’s determination that the election was valid. It is true that coercively interrogating an employee is yet another way in which employers and unions can violate the section 7 rights of employees. See Millard Refrigerated Servs., 345 N.L.R.B. at 1146. But we have already found that substantial evidence supports the Board’s determination that all of Mackin’s inappropriate conduct was adequately offset by Mystic’s own conduct with respect to Mackin in particular and the overall election in general. The Board’s ultimate conclusion as to the propriety of the election remains valid regardless of whether Mackin was acting as an agent of the Union.
Mystic insists otherwise and points to our recent decision in Ozark Automotive Distributors, Inc. v. NLRB, 779 F.3d 576 (D.C.Cir.2015). There, an NLRB hearing officer had refused an employer’s effort to subpoena documents it believed might show that employees who had advocated on behalf of the union during a representation election were union agents who had played a large role in influencing the election. We found that refusing to enforce the subpoena in that case was prejudicial because obtaining the records would have given the employer critical advantages that it otherwise lacked in putting on its case. But those considerations are not present here. The records would not have revealed any information other than the existence of conversations between Mackin and the Union organizer, two individuals already well known to Mystic. The records could not have served as new evidence or helped to identify new leads or witnesses. And because Mystic failed to call either Mackin or the Union organizer to testify, the records could not have helped impeach or examine them. Id. at 585. Admittedly, the Hearing Officer directed the Union to produce the organizer and the Union failed to do so. But Mystic also failed to remind the Hearing Officer of her instruction or to mention the organizer or the subpoena again in any way. Mystic cannot complain that it was prejudiced when it failed to call the only witness whose testimony might have made the records relevant.
Mystic also argues that the Board’s decision in this case is undermined by its decision in Voith Industrial Services, Inc., No. 09-CA-075496, 2012 WL 4169024 (Sept. 19, 2012). In Voith, the Board found an abuse of discretion where a Hearing Officer refused to enforce a subpoena for records regarding the relationship between an employer and the union that represented its employees because the records were at least “potentially relevant.” Id. at *1. But here, regardless of whether [315]*315the records were relevant to the question of Mackin’s status as a Union agent, they could not have altered the Board’s decision that Mackin’s lobbying did not contaminate the election result because it was offset by Mystic’s public discipline of Mackin and Mystic’s own anti-union conduct. Mystic was not prejudiced because these records simply could not have changed the outcome. And absent any prejudice we have no basis to reverse the Board with respect to the subpoena. Ryerson, 216 F.3d at 1154.
Ill
For the foregoing reasons, we deny Mystic’s petition for review and grant the NLRB’s cross-application for enforcement of its order.