HUGHES, Circuit Judge.

Mark Shapiro appeals the Merit Systems Protection Board's finding of good cause to remove him from his position as an administrative law judge. Because the Board did not err in concluding that Mr. Shapiro's production was dramatically lower than similarly situated ALJs, and that this failure to manage his caseload constitutes good cause for removal, we affirm.

I

Mr. Shapiro began working for the Social Security Administration as an administrative law judge in the New York Hearing Office in 1997. As early as 1998, the Agency informed Mr. Shapiro that his performance was lacking. In March and April 1998, the Agency informed Mr. Shapiro that he was scheduling too few hearings and his total number of case dispositions was below expectations. Mr. Shapiro's poor performance continued and, in early 2000, "a tremendous backlog" of cases in his docket became apparent. Accordingly, the Agency began to take an active role in assisting Mr. Shapiro, including reviewing cases, drafting decisions, and returning them for signature. Mr. Shapiro's performance, however, did not improve. In 2003, Agency management provided continuing assistance, but Mr. Shapiro issued only a limited number of dispositions, causing his backlog to grow deeper.

From March 2006 to March 2007, Mr. Shapiro received more counseling from then-Acting Region II Chief Administrative Law Judge Wright. J.A. 338-39. ALJ Wright eventually discontinued these meetings in March 2007, as he did not see an adequate improvement in Mr. Shapiro's productivity. J.A. 341.

In February 2007, the New York Hearing Office Chief ALJ sent Mr. Shapiro a memorandum outlining his concern over Mr. Shapiro's failure to process cases in a timely fashion and to produce an adequate number of decisions. J.A. 293-94. The letter chronicled Mr. Shapiro's history of poor performance, and indicated that he had over 70 percent of the 1000-day-old cases in the New York Office. J.A. 294 (72.8%); J.A. 332 (over 75%).

In October 2007, the then-Chief ALJ for the Agency, Frank Cristaudo, sent a memorandum "asking each of our [ALJs] to manage their dockets in such a way that they will be able to issue 500-700 legally sufficient decisions each year." J.A. 281. In an effort to facilitate meeting this goal, the Acting Regional Chief Judge sent Mr. Shapiro a memorandum emphasizing Chief ALJ Cristaudo's message that each ALJ must "manag[e] his/her docket effectively." J.A. 297. As such, the Acting Regional Chief Judge directed Mr. Shapiro to attend a series of formal meetings with ALJ Wright. Id.

During these meetings, ALJ Wright and Mr. Shapiro engaged in "a frank discussion of [Mr. Shapiro's] low disposition productivity, recommended efficiencies, and a clear explanation of management's productivity expectations...." J.A. 174; J.A. 174-80 (summarizing contents of each meeting). As found by the presiding ALJ below, the "intent and substance of each `improvement meeting' was essentially the same: [Administrative Law] Judge Wright and [Mr. Shapiro] discussed [Mr. Shapiro's] cases and why many were lingering in certain pre-or-post hearing statuses without resolution or action." J.A. 180.

Mr. Shapiro's productivity, however, did not materially change following the improvement meetings. And for Fiscal Years 2008, 2009, and 2010, Mr. Shapiro disposed of drastically fewer cases as compared to his peers in the New York Hearing Office and across the entire Region II:

                                      FY 2008 Dispositions     FY 2009 Dispositions     FY 2010 Dispositions

Mr. Shapiro                                149                      122                      111

New York Hearing Office (Average)          567                      611                      630

Region II (Average)                        613                      608                      622

J.A. 254; J.A. 256-58. Thus, in those three years, there was a disparity of over 400 cases from the average.

Mr. Shapiro's supervisor concluded that "despite the extraordinary efforts to assist him, to mentor him, [and] to train him," Mr. Shapiro could not manage his docket to meet expectations. J.A. 355-56. Accordingly, pursuant to 5 U.S.C. § 7521, the Agency filed a complaint with the Board seeking a finding of good cause to remove Mr. Shapiro based on two charges: (1) "unacceptable performance" and (2) "neglect of duty." Charge I contains three separate specifications relating to the 2008-2010 time period: (1) failure to provide timely hearings; (2) failure to provide timely dispositions; and (3) failure to "acceptably manage [] cases." J.A. 39. Charge II contained substantially similar specifications in the context of a "Neglect of Duty" charge.

Pursuant to 5 U.S.C. § 3105 and 5 C.F.R. § 1201.140, an ALJ was designated to conduct a hearing and issue a recommendation regarding the complaint. The presiding ALJ heard six days of evidence and argument across two sessions, with a break in between to permit Mr. Shapiro an opportunity to conduct more discovery. During the hearing, the Agency presented evidence of the average number of case dispositions for ALJs in Mr. Shapiro's office and across the region. This evidence was supported by testimony from ALJs who reviewed Mr. Shapiro's cases and concluded that his cases were no different than theirs or any other case before the Agency.

After the hearing, the presiding ALJ found that the Agency failed to prove specifications 1 and 2 because the Agency had not established "an enforceable timeliness standard ... attributable solely to a judge." J.A. 156. In other words, the Agency could not prove that Mr. Shapiro "was solely responsible for the processing times referenced in the cases cited in the... complaint." J.A. 156; see also J.A. 166. As for the third specification, the presiding ALJ found that the Agency showed by a preponderance of the evidence that Mr. Shapiro failed to "acceptably manage his cases" and that such conduct constituted good cause for removal.

The Board found that the presiding ALJ correctly determined that the Agency defined the phrase "acceptably manage" in the third specification of Charge I by comparing the number of cases Mr. Shapiro either scheduled for hearing or disposed of with the cases scheduled or disposed of by his peers in the New York Hearing Office and Region II. J.A. 6-7. The Board found that even if a portion of Specification 3 related to the timeliness of Mr. Shapiro's decision, Mr. Shapiro had failed to show that the presiding ALJ erred by sustaining the other portions of the specification showing a failure to manage his cases acceptably. Thus, the Board sustained Charge I and found good cause for the removal of Mr. Shapiro.

Mr. Shapiro appeals. We have jurisdiction under 28 U.S.C. § 1295(a)(9).

II

"Our review of Board decisions is limited. We may only reverse a Board decision if we find the decision to be arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law; obtained without procedures required by law; or unsupported by substantial evidence." Kahn v. Dep't of Justice, 618 F.3d 1306, 1312 (Fed.Cir.2010) (citing 5 U.S.C. § 7703(c)). "Under the substantial evidence standard, this court reverses the Board's decision only `if it is not supported by such relevant evidence as a reasonable mind might accept as adequate to support a conclusion.'" Abrams v. Soc., Sec. Admin., 703 F.3d 538, 542 (Fed.Cir.2012) (quoting Haebe v. Dep't of Justice, 288 F.3d 1288, 1298 (Fed.Cir.2002)).

The Agency may remove Mr. Shapiro "only for good cause established and determined by the [Board]...." 5 U.S.C. § 7521(a). "Congress intentionally failed to define `good cause'" leaving it "to be given meaning through judicial interpretation." Brennan v. Dep't of Health & Human Servs., 787 F.2d 1559, 1561-62 (Fed. Cir.1986). "And we have made clear that, as a general matter, we defer to the Board's reasonable interpretation of `good cause' because `the Board has exclusive rulemaking and adjudicatory authority with respect to section 7521.'" Berlin v. Dep't of Labor, 772 F.3d 890, 894 (Fed.Cir. 2014) (quoting Long v. Soc., Sec. Admin., 635 F.3d 526, 534 (Fed.Cir.2011)). Because Mr. Shapiro does not contest that "unacceptable performance" or a lack of production can constitute good cause for removal, we need only determine whether substantial evidence supports the Board's conclusion that the Agency showed by a preponderance of the evidence that this charge was met. See Brennan, 787 F.2d at 1561.

A

Mr. Shapiro first argues that the Agency failed to establish "good cause" for removal because the Agency did not prove the precise charge asserted. However, this argument is predicated on the mistaken assumption that removal of an ALJ for "good cause" under § 7521 must comply with the precedent and requirements of § 7512 cases governing removal of employees under the "efficiency of the service" standard. But that is plainly not the case. See 5 U.S.C. § 7512(E) ("This subchapter... does not apply to ... an action initiated under section 1215 or 7521 of this title."). Under the governing regulations, a complaint seeking removal of an ALJ need only "describe with particularity the facts that support the proposed agency action," 5 C.F.R. § 1201.138, and the Board may only discipline an ALJ for "good cause," id. at § 1201.140(b). This is in contrast to the regulations governing § 7512 actions, which require a notice to "state the specific reason(s) for the proposed actions," 5 C.F.R. § 752.404(b)(1), and state that the agency may "consider only the reasons specified in the notice of the proposed action ...," id. at § 752.404(g)(1). Thus, in the context of § 7521, "[t]he purpose of an agency's notice of charges is to put an employee on notice of the allegations against him in sufficient detail to apprise him of the allegations he must refute or acts he must justify." Brennan, 787 F.2d at 1561.

Here, Mr. Shapiro had ample notice of the charges brought against him. Charge I, labeled "Unacceptable Performance," plainly put Mr. Shapiro on notice that his "performance has been unacceptable, in that ... [i]n or about FY2008-2010, [he] did not acceptably manage his cases." J.A. 242. And Mr. Shapiro testified that he understood this to mean that the Agency was seeking to remove him from his position due to his perceived mismanagement of cases, including his failure to "produce very many decisions." J.A. 365. The Board, on the presiding ALJ's recommendation, found good cause for removal because Mr. Shapiro was unable to effectively manage his docket. There is no indication that this did not comply with the governing regulations. And given that the hearing was recessed to allow for Mr. Shapiro to conduct additional discovery, it is beyond debate that Mr. Shapiro was "afforded notice `both of the charges and of the employer's evidence' and an `opportunity to respond' before [he was] removed from employment," as required to satisfy due process. Ward v. U.S. Postal Serv., 634 F.3d 1274, 1279 (Fed.Cir.2011) (citing Stone v. Fed. Deposit Ins. Corp., 179 F.3d 1368, 1374-76 (Fed.Cir.1999)).

B

Mr. Shapiro next argues that the Board erred in its good cause determination by relying on a comparison of his production statistics to regional averages, in contravention of the rule announced by the Board in Social Security Administration v. Goodman, 19 M.S.P.R. 321, 331 (1984). Although Goodman was decided over thirty years ago, we have not yet had occasion to review its rule beyond noting that we are not bound by it. See, e.g., Abrams v. Soc., Sec. Admin., 703 F.3d 538, 544 n. 2 (Fed.Cir.2012). And we do not adopt it today. We agree with Goodman to the extent that it requires a proper foundation for the type of comparative statistics employed here. But to the extent Goodman requires some type of heightened evidentiary proof before an agency can rely upon comparative production statistics to prove good cause for removal, we decline to follow it. When an agency establishes that an individual ALJ's case disposition rate is so significantly lower than the rate of similarly situated ALJs in his own region, that evidence, absent some contradictory showing that the statistics do not present a valid comparison, can support a finding of good cause.

In Goodman, the Board was concerned with the use of comparative statistics to establish that an ALJ's "performance was sufficiently below a reasonable level of productivity to warrant his removal." Goodman, 19 M.S.P.R. at 331-32. The Board found that, absent evidence that "the same amount of time was required to render most final dispositions" and "the complexities presented by the mix of cases assigned to the respondent mirrored the complexities of those included" in the compared disposition rates, such a comparison would be based only on speculation and therefore could not support a finding of good cause. Id. Because "no evidence was offered regarding the time required to render dispositions or comparing respondent's assignments with those included in the national average," the Board found that the agency failed to meet its burden of establishing good cause for removal. Id. at 332.

Similarly, in Social Security Administration v. Brennan, 19 M.S.P.R. 335 (1984), decided the same day as Goodman, the Board rejected evidence that an ALJ's case disposition rate was "not within the range of the average productivity of all ALJs employed by the agency." Id. at 337 n. 2. This was because "the national average included different types of dispositions, such as dismissals, short form reversals, and affirmances both on the written record and after a hearing" and there was no evidence presented that the ALJ's caseload mirrored that case makeup. Id. at 337.

Goodman and its progeny have been described as a "pyrrhic victory for the [Agency]" because it "won the right to bring charges against low producing ALJs but was handed a virtually insurmountable burden of proof." Jeffrey S. Lubbers, The Federal Administrative Judiciary: Establishing an Appropriate System of Performance Evaluation for ALJs, 7 Admin. L.J. Am. U. 589, 599-600 (1994). The proceedings below reflect that sentiment. The record reveals a long history of poor performance and the Agency's "unprecedented and extraordinary efforts" to assist Mr. Shapiro. See, e.g., J.A. 143, 209. Indeed, the presiding ALJ found the Agency "demonstrated extraordinary patience and expended extraordinary effort to assist [Mr. Shapiro] ... in the performance of his duties." J.A. 211. Similarly, counsel for the Agency stated at oral argument that the Agency delayed initiating removal proceedings because of a perception that Goodman did not, in fact, permit the removal of an ALJ for productivity reasons.1 It is therefore not surprising that, when it came time to evaluate the statistics presented during the removal proceedings, the presiding ALJ found: "Doubtless because of Goodman, the Agency went to exceptional lengths to establish that during [the relevant time period], the cases assigned to [Mr. Shapiro] were essentially the same or similar, in terms of file size, complexity, legal and evidentiary and/or factual issues, and time requirements as those cases assigned to all other judges in the New York City and Region II hearing offices." J.A. 187.

No fewer than four ALJs provided testimony that they reviewed Mr. Shapiro's docket during the relevant timeframe. The region chief "personally reviewed ... [Mr. Shapiro's] assigned cases" and concluded that the cases were the same or similar to those handled by other judges within the region. J.A. 187. Mr. Shapiro's supervisor gave similar testimony, testifying that the Agency's ALJs "see the same types of cases `over and over again.'" J.A. 188 (citation omitted). Indeed, Mr. Shapiro himself conceded that "I have no reason to think that any of us [ALJs] are given a different type" of case. J.A. 364. Especially in light of this concession, it is particularly inappropriate to require the Agency to undergo the herculean effort of providing testimony from four ALJs that an ALJ's caseload was the same or similar to that of any other ALJ in the region before the fact finder can rely on any comparative statistics, as the Agency perceived to be required by Goodman.

We agree that there must be a showing that the proffered production statistics are relevant to the determination — i.e., that they reflect the average disposition rate for a particular region across the same time period. But in extreme cases like this, where Mr. Shapiro's production is, at best, roughly a quarter of that performed by the rest of the ALJs in his region, that standing alone is highly relevant and potentially preponderant evidence that he failed to manage his cases acceptably. Absent some suggestion that the character of an individual ALJ's docket dramatically differs from that of his colleagues, or any other factors that might undermine the reliability of the comparative statistics, the Board is free to give such statistics appropriate weight when determining whether the Agency has met its burden to prove the charges alleged by a preponderance of the evidence. See Brennan, 787 F.2d at 1561.

In this case, the presiding ALJ assessed the credibility of the witnesses and concluded that the cases assigned to Mr. Shapiro were "essentially the same or similar... as those cases assigned to all other judges in the New York City and/or Region II hearing offices." J.A. 189. The Board adopted these findings, recognizing that they were based in part on a credibility determination. J.A. 8-9; see also J.A. 15. We see no reason to deviate from these findings. See, e.g., Long, 635 F.3d at 530-31.2

Mr. Shapiro also argues that even if the identified cases are the same or similar, the comparative statistics relied on by the Board are flawed because they ignore the role played by support staff in the rate of disposition; and they fail to account for the fact that Mr. Shapiro had far fewer cases than the average ALJ used in the comparison. Neither of these arguments is sufficient to undermine the Board's findings.

First, Mr. Shapiro did not identify any evidence of record suggesting his support staff played a major role in his decreased performance numbers. At best, there is some evidence in the record suggesting that support staff played a role in the timely disposition of cases. But nothing in the record provides a legally sufficient basis for overturning the presiding ALJ's finding that timeliness is not necessarily tied to production numbers. J.A. 171 n. 29 ("Although it might be anecdotally true that an ALJ who is untimely in scheduling hearings or producing dispositions might also schedule fewer hearings or produce fewer dispositions, one does not necessarily follow the other."). Moreover, even if Mr. Shapiro were correct that support staff played a role in his inability to manage his docket, there is no evidence in the record to support an assumption that his support staff was so far below the norm as to account for a disparity of over 400 cases from the average for three straight years. Indeed, as the Board recognized, Mr. Shapiro conceded that "because of his techniques and approach to case processing, he might be able to produce only 200 cases per year, far below the [Agency's] goal of 500-700 dispositions per administrative law judge per year." J.A. 6-7.

Second, even if there were a bottleneck caused by support staff, the Board did not disturb the presiding ALJ's finding that "a prudent ALJ, aware that staff deficiencies contributed to slow case development and a low disposition rate, should ask for additional cases to compensate for those delays ...." J.A. 180 n.35. And, as a result of the numerous counseling sessions and improvement efforts by the Agency, Mr. Shapiro was no doubt aware of the Agency's expectation that he should have decided more cases per year. But during these sessions, Mr. Shapiro only once asked for additional cases. The Agency complied but his production numbers continued to be unacceptable. Thus, substantial evidence supports the Board's adoption of the presiding ALJ's conclusion that "even if [Mr. Shapiro] had been given a full complement of cases, he could not (or would not) have been able to meet the desired quantity." J.A. 195; J.A. 6-7.

III

We find no error in the Board's removal of Mr. Shapiro based on a charge of "unacceptable performance." The Board's decision is supported by substantial evidence. Moreover, the Agency properly used comparative statistics in this case to establish that Mr. Shapiro's performance was substantially below that of his peers. And, the Board complied with the governing regulations, as Mr. Shapiro was given adequate notice of the charges alleged along with an opportunity to respond. Accordingly, the Board's decision to remove Mr. Shapiro for good cause is affirmed.

AFFIRMED